<%BANNER%>

Experimental and Computational Study of Catalytic Combustion of Methane-Air and Syngas-Air Mixtures

Permanent Link: http://ufdc.ufl.edu/UFE0021035/00001

Material Information

Title: Experimental and Computational Study of Catalytic Combustion of Methane-Air and Syngas-Air Mixtures
Physical Description: 1 online resource (149 p.)
Language: english
Creator: Pathak, Saurav
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: catalytic, combustion, computational, deutschmann, experimental, fluent, gambit, gas, mechanism, methane, phase, reaction, surface, syngas
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Mechanical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Catalytic combustion and conversion of methane (CH4) and Syngas (in our case, a gas mixture of H2, CO, CO2 and CH4) is characterized by the complex transport of reacting gases and the chemical reactions on the catalyst surface and in the gas phase. The initiation and the kinetics of reactions on the catalyst surface depend largely on the surface coverage of species. Since the catalyst allows combustion at a low temperature resulting in a lower thermal NO formation, a detailed investigation on the effects of species surface coverage on catalytic combustion becomes an area of concerned research. In this work, we investigate the dependence of catalytic combustion of two fuel-air mixtures, namely: methane-air and methane + hydrogen-air mixture, based on their surface coverage on Platinum, using a detailed surface reaction mechanism model. We look into the changes brought about in the kinetics and thus on combustion, when different species interact with the platinum surface, individually and in the presence of one another. We further extend our numerical investigation, to understand the kinetics involved in the catalytic combustion of Syngas on platinum. Syngas today can now be made by the gasification of coal and from other fuel sources using detailed technologies, such that the efficient and low energy harnessing of the thermal properties of Syngas is an important area of research. Catalytic combustion of syngas would meet exactly these goals. Hence the study of a low energy reaction pathway of combustion of syngas on the surface of any catalyst in general has now become an important area of research. Catalytic reactor has been constructed to provide experimental data for comparison with model. Experimentally the ignition process is monitored by thermocouples and mass spectroscopy. Parameters such as equivalence ratio, temperature and the surface site density of the catalyst surface were varied in the computational works to bring out their influence on the kinetics of combustion.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Saurav Pathak.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Segal, Corin.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021035:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021035/00001

Material Information

Title: Experimental and Computational Study of Catalytic Combustion of Methane-Air and Syngas-Air Mixtures
Physical Description: 1 online resource (149 p.)
Language: english
Creator: Pathak, Saurav
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: catalytic, combustion, computational, deutschmann, experimental, fluent, gambit, gas, mechanism, methane, phase, reaction, surface, syngas
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Mechanical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Catalytic combustion and conversion of methane (CH4) and Syngas (in our case, a gas mixture of H2, CO, CO2 and CH4) is characterized by the complex transport of reacting gases and the chemical reactions on the catalyst surface and in the gas phase. The initiation and the kinetics of reactions on the catalyst surface depend largely on the surface coverage of species. Since the catalyst allows combustion at a low temperature resulting in a lower thermal NO formation, a detailed investigation on the effects of species surface coverage on catalytic combustion becomes an area of concerned research. In this work, we investigate the dependence of catalytic combustion of two fuel-air mixtures, namely: methane-air and methane + hydrogen-air mixture, based on their surface coverage on Platinum, using a detailed surface reaction mechanism model. We look into the changes brought about in the kinetics and thus on combustion, when different species interact with the platinum surface, individually and in the presence of one another. We further extend our numerical investigation, to understand the kinetics involved in the catalytic combustion of Syngas on platinum. Syngas today can now be made by the gasification of coal and from other fuel sources using detailed technologies, such that the efficient and low energy harnessing of the thermal properties of Syngas is an important area of research. Catalytic combustion of syngas would meet exactly these goals. Hence the study of a low energy reaction pathway of combustion of syngas on the surface of any catalyst in general has now become an important area of research. Catalytic reactor has been constructed to provide experimental data for comparison with model. Experimentally the ignition process is monitored by thermocouples and mass spectroscopy. Parameters such as equivalence ratio, temperature and the surface site density of the catalyst surface were varied in the computational works to bring out their influence on the kinetics of combustion.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Saurav Pathak.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Segal, Corin.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021035:00001


This item has the following downloads:


Full Text

PAGE 1

1 EXPERIMENTAL AND COMPUTATIONAL ST UDY OF CATALYTIC COMBUSTION OF METHANE-AIR AND SYN GAS-AIR MIXTURES By SAURAV PATHAK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 Saurav Pathak 2007

PAGE 3

3 Dedicated to my country (India)

PAGE 4

4 ACKNOWLEDGMENTS I am thankful to God for having given me an opportunity to fulfill my dreams of higher education. I am thankful to my parents and sister for their continued support and faith in me. I extend my thankfulness to my country India and Birla Institute of Tec hnology, Mesra, India from where I earned my Bachel or of Engineering. At the University of Florida I am thankful to my adviser Dr.C.Segal. Had it not been for the fait h and support of Dr C.Segal, Dr Mark Sheplak and Dr. W. Shyy, it would have been impossibl e for me to pursue my PhD program. I am thankful to all my committee members for th eir suggestions and suppor t. I also extend my acknowledgments to all my lab mates namely Jonas, Amit, Stepan, Aravind, Kyle, John and Jignesh for helping me out on numerous occasions. I am also thankful to my room mate Rakesh for helping me out whenever I had any problems.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES................................................................................................................ .........8 ABSTRACT....................................................................................................................... ............13 CHAPTER 1 INTRODUCTION..................................................................................................................15 Catalytic Combustion: Purpose and Application....................................................................15 NOx Control in Gas Turbine Combustion Chambers.............................................................17 Catalytic Combustion: a Subset of Catalysis..........................................................................19 Catalysis and Catalysts....................................................................................................20 Catalysts and Reaction Energetics...................................................................................20 Types of Catalysts...........................................................................................................21 Kinetics of Catalyzed Reactions......................................................................................21 Mechanism of Catalyzed Reacti ons: Making and Breaking Bonds................................23 Mechanism of Heterogeneous Catalysis.........................................................................25 2 EXPERIMENTAL STUDIES OF CATALYTIC COMBUSTION.......................................31 Combustion Diagnostics.........................................................................................................31 Mass Sampling for Species Concentration Measurement..............................................31 Optical Diagnostics-Laser Induced Fluorescence in Combustion..................................32 Computational Diagnostics............................................................................................32 Motivation..................................................................................................................... ..........33 Surface Chemistry Models.....................................................................................................34 One-Step Surface Reaction..............................................................................................34 MultiStep Surface Reaction Mechanisms.....................................................................35 Catalytic Combustion of Methane on Platinum......................................................................37 Catalytic Partial Oxidation of Me thane on Platinum-Hydrogen Assisted..............................38 Catalytic Combustion of Syngas on Platinum........................................................................40 Catalytically Stabilized Thermal Combustion (CST).............................................................41 Catalytic Activation Index (CAI)...........................................................................................41 3 EXPERIMENTAL AND CO MPUTATIONAL SETUPS.....................................................47 Experimental Setup............................................................................................................. ....47 Mass Sampling................................................................................................................48 Hardware..................................................................................................................48 Data processing........................................................................................................50

PAGE 6

6 Computational Setup............................................................................................................ ..52 Computational Setup in Gambit......................................................................................52 Computational Setup in FLUENT...................................................................................53 Program Capabilities.......................................................................................................54 Numerical Model.............................................................................................................54 Mathematical Model........................................................................................................55 Governing Equations.......................................................................................................55 Mass Diffusion in Laminar Flows...................................................................................56 Treatment of Species Trans port in the Energy Equation.................................................56 The Generalized Finite-Rate Form ulation for Reaction Modeling.........................................57 The Laminar Finite-Rate Model......................................................................................58 Boundary Conditions......................................................................................................62 4 RESULTS AND DISCUSSION.............................................................................................68 Experimental Results........................................................................................................... ...70 Methane-Air Combustion: Mass Samp ling in the Axial Direction.................................70 Syngas-Air Combustion: Mass Samp ling in the Axial Direction...................................70 Ccomputational Results......................................................................................................... .71 Non Reacting Flow..........................................................................................................71 Reacting Flow..................................................................................................................72 Effect of Catalyst Su rface Site Density on th e Formation of Products...................72 Catalytic Combustion of Methane on Platinum Catalyst........................................73 Catalytic Partial Oxidation of Me thane on Platinum: Hydrogen Assisted..............76 Catalytic Combustion of Syngas on Platinum.........................................................77 Effect of Equivalence Ratio on Su rface Coverages in Syngas Combustion...........80 Comparison between the Experimental and Computational Results of Syngas Combustion...........................................................................................................82 5 CONCLUSION.....................................................................................................................122 APPENDIX A DETAILED COMPUTATIONAL SETUP..........................................................................125 B GAMBIT PROGRAM FOR DESIGN AND MESH GENERATION.................................140 LIST OF REFERENCES.............................................................................................................144 BIOGRAPHICAL SKETCH.....................................................................................................1499

PAGE 7

7 LIST OF TABLES Table page 1-1. Deutschmann's Surface Reaction Mech anism, A is Pre Exponential Factor (cm.mol.sec), is Temperature Exponent and Ea (KJ/mol)..............................................30 3-1. Operating Range for Mass Flow Controllers.........................................................................67 4-1. Experimental and Co mputational Case Matrix...................................................................119 4-2. Deutschmann's Surface Reaction Mech anism, A is Pre Exponential Factor (cm.mol.sec), is Temperature Exponent and Ea is activation energy(KJ/mol)............120 4-3. Five Step Gas Phase Reaction Mechanism, A is Pre Exponential Factor (cm.mol.sec), is Temperature Exponent and Ea is activation ebergy(KJ/mol)......................................121 4-4. Methane Inlet Flow Rate Specification...............................................................................121 4-5. Syngas Inlet Flow Rate Specifications................................................................................121

PAGE 8

8 LIST OF FIGURES Figure page 1-1. Types of combustors...................................................................................................... ........28 1-2. Catalysis taking place at reduced activation energy..............................................................28 1-3. Steps involved in heterogeneous catalysis.............................................................................29 1-4. Steps involved in Langmuir-Hinsh elwood-Hougen-Watson (LHHW) Catalysis.................29 1-5. Steps involved in Eley-Rideal (E-R) catalysis......................................................................29 2-1. Surface Ignition Temperature decreases wi th increase in methane equivalence ratio..........43 2-2. Methane Combustion with no H2 and with H2 (a) maximum temperature (b) methane ignition temperature (c) methane conve rsion (d) oxygen conversion and (e) CO2 yield.......................................................................................................................... ..........44 2-3. Surface coverage of species in methane combustion with no H2 and with H2......................45 3-1. Model of the Catalytic Combustor: Zone 2 is the fuel-air mixing region, Zone 3 is the fuel-air mixture expansion in a CROSS a nd Zone 4 is the com bustor section holding the catalyst and the annulus. All dimensions are in mm....................................................65 3-2. 2D-axisymmetric schematic of the Co mbustor section including catalyst and mass sampling ports 1 and 2. All dimensions are in mm............................................................65 3-3. Mass Sampling Ports...................................................................................................... .......66 3-4. Mass Spectrometers for Species Concentration Measurement..............................................66 4-1. Mole Fraction Distribution of Syngas Combustion at Port One and Port Two on Platinum Smooth................................................................................................................86 4-2. Mole Fraction Distribution of Syngas Combustion at Port One and Port Two on Platinum Rough.................................................................................................................86 4-3. Comparison between Catalyst Platinum Smooth and Rough at Port 1 in Converting Reactants...................................................................................................................... ......87 4-4. Comparison between Catalyst Platinum Smooth and Rough at Port 2 in converting reactants...................................................................................................................... .......87 4-5. Comparison between Catalyst Platinum Smooth, Rough, A & B in Product Yield at Port One....................................................................................................................... ......88

PAGE 9

9 4-6. Comparison between Catalyst Platinum Smooth, Rough, A & B in Product Yield at Port Two....................................................................................................................... ......88 4-7. Schematic Diagram of the Computational Domain...............................................................89 4-8. Non Reacting Mole Fraction Distribution of Methane at Section ee, ff, Ports One and Two............................................................................................................................ ........89 4-9. Non Reacting Mole Fraction Distribution of Hydrogen in Syngas at Section ee, ff, Ports One and Two.............................................................................................................90 4-10. Non Reacting Mole Fraction Distribution of CO in Syngas at Section ee, ff, Ports One and Two........................................................................................................................ .....90 4-11. Mole Fraction Distributi on with respect to Site Dens ity for Methane Combustion at 1700K at Port 2................................................................................................................ ..91 4-12. Mole Fraction Distributi on with respect to Site Dens ity for Methane Combustion at 1700K at Outlet................................................................................................................ ..91 4-13. Mole Fraction Distributi on with respect to Site Density for Hydrogen Assisted Methane Combustion at 900K at Port 2.............................................................................92 4-14. Mole Fraction Distributi on with respect to Site Density for Hydrogen Assisted Methane Combustion at 900K at Outlet............................................................................92 4-15. Mole Fraction Distributi on with respect to Site Dens ity for Syngas Combustion at 1100K at Port.................................................................................................................. ...93 4-16. Mole Fraction Distributi on with respect to Site Dens ity for Syngas Combustion at 1100K at Outlet................................................................................................................ ..93 4-17. SEM images of Rough (a) and (b) and Smooth (c) and (d) magnified 10000 and 15000 respectively. Rough surfaces show more number of grains than the smooth surface per unit area. The exposed surface ar e is thus greater for rough surface and thus provides more active sites on platinum surface to allow adsorption. M: Magnification: S: Scale.....................................................................................................94 4-18. Surface Coverage Distributions along Catalyst for Methane Combustion at 400K............95 4-19. Surface Coverage Distributions along Catalyst for Methane Combustion at 600K............95 4-20. Surface Coverage Distributions along Catalyst for Methane Combustion at 700K............96 4-21. Surface Coverage Distributions along Catalyst for Methane Combustion at 900K............96 4-22. Surface Coverage Distributions along Catalyst for Methane Combustion at 1100K..........97 4-23. Surface Coverage Distributions along Catalyst for Methane Combustion at 1200K..........97

PAGE 10

10 4-24. Surface Coverage Distributions along Catalyst for Methane Combustion at 1500K..........98 4-25. Surface Coverage Distributions along Catalyst for Methane Combustion at 1700K..........98 4-26. Surface Temperature Distribution along Cata lyst for Methane Combustion at different Inlet Temperatures.............................................................................................................99 4-27. Mole Fraction Distributi on of Methane Combustion at Inlet, Port One, Two and Outlet at 1500K................................................................................................................ ..99 4-28. Mole Fraction Distributi on of Methane Combustion at Inlet, Port One, Two and Outlet at 1700K................................................................................................................100 4-29. Surface Coverage Distribution along Ca talyst for Hydrogen Assisted Methane Combustion at 400K........................................................................................................100 4-30. Surface Coverage Distribution along Ca talyst for Hydrogen Assisted Methane Combustion at 600K........................................................................................................101 4-31. Surface Coverage Distribution along Ca talyst for Hydrogen Assisted Methane Combustion at 700K........................................................................................................101 4-32. Surface Coverage Distribution along Ca talyst for Hydrogen Assisted Methane Combustion at 900K........................................................................................................102 4-33. Surface Temperature Distribution along Catalyst for Hydrogen Assisted Methane Combustion at different inlet Temperatures....................................................................102 4-34. Mole Fraction Distributi on of Hydrogen Assisted Methan e Combustion at Inlet, Port One, Two and Outlet at 400K..........................................................................................103 4-35. Mole Fraction Distributi on of Hydrogen Assisted Methan e Combustion at Inlet, Port One, Two and Outlet at 600K..........................................................................................103 4-36. Mole Fraction Distributi on of Hydrogen Assisted Methan e Combustion at Inlet, Port One, Two and Outlet at 700K..........................................................................................104 4-37. Mole Fraction Distributi on of Hydrogen Assisted Methan e Combustion at Inlet, Port One, Two and Outlet at 900K..........................................................................................104 4-38. Surface Coverage Distributions along Catalyst for Syngas Combustion at 400K............105 4-39. Surface Coverage Distributions along Catalyst for Syngas Combustion at 600K............105 4-40. Surface Coverage Distributions along Catalyst for Syngas Combustion at 700K............106 4-41. Surface Coverage Distributions along Catalyst for Syngas Combustion at 900K............106 4-42. Surface Coverage Distributions along Ca talyst for Syngas Combustion at 1100K..........107

PAGE 11

11 4-43. Surface Temperature Distribution along Ca talyst for Syngas Combustion at different inlet Temperatures...........................................................................................................108 4-44. Mole Fraction Distributi on of Syngas Combustion at Inle t, Port One, Two and Outlet at 400K........................................................................................................................ .....108 4-45. Mole Fraction Distributi on of Syngas Combustion at Inle t, Port One, Two and Outlet at 600K........................................................................................................................ .....109 4-46. Mole Fraction Distributi on of Syngas Combustion at Inle t, Port One, Two and Outlet at 700K........................................................................................................................ .....109 4-47. Mole Fraction Distributi on of Syngas Combustion at Inle t, Port One, Two and Outlet at 900K........................................................................................................................ .....110 4-48. Mole Fraction Distributi on of Syngas Combustion at Inle t, Port One, Two and Outlet at 1100K....................................................................................................................... ....110 4-49. Surface Coverage Distributions along Ca talyst for Syngas Combustion at 900K at = 0.47........................................................................................................................... ........111 4-50. Mole Fraction Distributi on of Syngas Combustion at Inle t; Port One and Two at 900K for =0.47........................................................................................................................111 4-51. Surface Coverage Distributions along Ca talyst for Syngas Combustion at 900K at = 0.72........................................................................................................................... ........112 4-52. Mole Fraction Distributi on of Syngas Combustion at Inle t; Port One and Two at 900K for =0.72........................................................................................................................112 4-53. Surface Coverage Distributions along Ca talyst for Syngas Combustion at 900K at = 1.0............................................................................................................................ .........113 4-54. Mole Fraction Distributi on of Syngas Combustion at Inle t; Port One and Two at 900K for =1.0..........................................................................................................................113 4-55. Surface Coverage Distributions along Ca talyst for Syngas Combustion at 900K at = 1.21........................................................................................................................... ........114 4-56. Mole Fraction Distributi on of Syngas Combustion at Inle t; Port One and Two at 900K for =1.21........................................................................................................................114 4-57. Temperature Comparison between Experi mental, Heterogeneous and Heterogeneous combined with Homogeneous Mechanisms....................................................................115 4-58. Temperature Distribution of Cooling Ai r for Methane, Hydroge n assisted Methane and Syngas Combustion at Port 1....................................................................................115

PAGE 12

12 4-59. Temperature Distribution of Cooling Ai r for Methane, Hydroge n assisted Methane and Syngas Combustion at Port 2....................................................................................116 4-60. Temperature Distribution of Cooling Ai r for Methane, Hydroge n assisted Methane and Syngas Combustion at Outlet....................................................................................116 4-61. Comparison between Experimental and Computational % Mass Fractions for Syngas Combustion at Port One and Two and at 700K...............................................................117 4-62. Comparison between Experimental and Computational % Mass Fractions for Syngas Combustion at Port One and Two and at 900K...............................................................117 4-63. Comparison between Experimental and Computational % Mass Mole Fractions for Syngas Combustion at Port One and Two and at 1100K.................................................118

PAGE 13

13 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EXPERIMENTAL AND COMPUTATIONAL ST UDY OF CATALYTIC COMBUSTION OF METHANE-AIR AND SYN GAS-AIR MIXTURES By Saurav Pathak August 2007 Chair: Corin Segal Major: Mechanical Engineering Catalytic combustion and c onversion of methane (CH4) and Syngas (in our case, a gas mixture of H2, CO, CO2 and CH4) is characterized by the complex transport of reacting gases and the chemical reactions on the cata lyst surface and in the gas phase. The initiation and the kinetics of reactions on the catalyst surface depend largely on the surface coverage of species. Since the catalyst allows combustion at a low temperatur e resulting in a lower thermal NO formation, a detailed investigation on the effects of species surface coverage on catalytic combustion becomes an area of concerned research. In this work, we investigate the dependen ce of catalytic combustion of two fuel-air mixtures, namely: methane-air and methane + hydrogen-air mixture, based on their surface coverage on Platinum, using a detailed surface reaction mechanism model. We look into the changes brought about in the kinetics and thus on combustion, when different species interact with the platinum surface, individually and in th e presence of one another. We further extend our numerical investigation, to unde rstand the kinetics involved in the catalytic combustion of Syngas on platinum. Syngas today can now be made by the gasification of coal and from other fuel sources using detailed tec hnologies, such that the efficient and low energy harnessing of the thermal properties of Syngas is an important ar ea of research. Catalytic combustion of syngas

PAGE 14

14 would meet exactly these goa ls. Hence the study of a low energy reaction pathway of combustion of syngas on the surface of any catalys t in general has now become an important area of research. Catalytic reactor has been c onstructed to provide experimental data for comparison with model. Experimentally the igni tion process is monitored by thermocouples and mass spectroscopy. Parameters such as equivalence ratio, temp erature and the surface site density of the catalyst surface were varied in the computationa l works to bring out their influence on the kinetics of combustion.

PAGE 15

15 CHAPTER 1 INTRODUCTION Catalytic Combustion: Purpose and Application Catalytic combustion is a pro cess that uses a solid catalys t to promote the combustion reaction by lowering the activation energy of certain reactions. Thes e reactions would take place at lower temperatures relative to homogeneous co mbustion and at lower e quivalence ratios, thus greatly reducing formation of NOx emissions and eliminating lower flammability limits respectively. The reactors become more compact because the catalytic reactions occur on the surface of the catalyst reducing the requi red volume but greater surface areas. Catalytic combustion research currently focu ses on reactor design, increase in efficiency, and the search for more efficient catalysts. Thus, there is an urgent need for a better understanding of the physical and chemical proce sses occurring on the cata lytic surface and their coupling with the surrounding flow field. In particular, it is importa nt to understand the ignition and extinction behavior of the oxidation of hydrocarbons. During the la st decade, detailed catalytic combustion models have been suggest ed [1, 2] including multi-step heterogeneous surface-reaction mechanisms which may provide gui delines for a better understanding and for the optimization of catalytic combustion. Since the ignition of methane on Pt occurs at relatively high temperatures at lean conditions [1, 3], there is a major interest to find a way to reduce the light-off temperature. The light off temperature is the temperature at which the first onset of ignition is noticed marked by a sudden rise in te mperature and an increase in the conversion of the reactants into products. This is true, in pa rticular for gas turbine applications where the catalytic combustion critically depends on the feasibility of a convenient light-off mechanism of the catalytic combustor.

PAGE 16

16 Catalyzed combustion can be defined as th e complete oxidation of a combustible compound on the surface of a catal yst. While conventiona l combustion occurs in the presence of a flame, catalyzed combustion is a flameless process occurring at lo wer temperatures and, therefore, emitting less nitrogen oxides. Furthermore, catalyzed combustion offers fewer constraints concerning flammability limits and re actor design. These advantages of catalyzed combustion determine its potential applicati ons. For example, a gaseous mixture of H2 and O2 on the platinum is a classical example of hetero geneous catalysis. The sticking probability of H atoms are higher than those of O atoms, hen ce the exothermic process of adsorption of H2 on the platinum surface is preferred over O2. The O atoms are loosely bound on the platinum surface; they combine with H atoms on the platinum su rface and form OH species which still remains adsorbed to the surface, this OH specie s further abstracts O atoms and forms H2O. This is an exothermic reaction too. The heat released shif ts the adsorption/desorption equilibrium towards the desorption side, such that H atoms would leave the platinum surface and become more available to abstract O atoms, thus favoring more OH formation and eventually forming more H2O. Using appropriate surface and gas phase re action mechanisms, the study of catalytic combustion may include Computational Fluid Dy namics (CFD) along with experiments. CFD may thus be a tool to provide early details of different technical de sign variants. Two main reasons for the application of CFD in combusto r studies are(1) Only CFD is capable of delivering detailed insight into the transport processes encountered. In flow, whereas measurements within the reacting flow of the co mbustor are physically limited; they can be used successfully to provide data to validate the m odels and numerical tool s applied and, (2) CFD offers a possibility to compare and evaluate different design vari ants already in the design phase

PAGE 17

17 of a new combustion system. The ma in objectives of CFD applicati on in the field of gas turbine combustion includes close monitoring of a numb er of processes including homogeneity at the burner inlet, the velocity profile at the burner inle t, the velocity and concentration profiles at the burner outlet using CFD. Considering the combus tor, the thermal wall heat load has to be calculated for the design of the liner cooling. Mo reover, for the development of high temperature turbines, a detailed knowledge of the velocity an d temperature profile has to be provided by CFD applications. Advantages of employing catalyt ic combustion extend to reduced combustion instabilities [4, 5] in practical burners, development of micropower supplier with high power density, reduced emissions from ground base d power plants and increased combustion efficiency. The present research includes both computational and experime ntal studies of catalytic combustion on model catalysts and is focused on the process of combusti on inside a combustor designed for a gas turbine engine motivated by the need to improve efficiency and reduce pollutants. NOx Control in Gas Turbine Combustion Chambers A combustor is a small component or area of a gas turbine, ramjet or pulsejet engine, where combustion takes place [6]. It is also known as a burner or flame can, depending on the design. In a gas turbine engine, the main combustor or combustion chambe r is located between the high pressure compressor and the high pressu re turbine of the gas generator. Combustor design requires that part of the airflow is directed to a region where the fuel and air are mixed and ignited. This area of the combustor is sometimes called a flame holder and allows a stable flame front to be established and maintained. The heat addition stage of a gas turbine cycle incurs a slight pressure drop a ssociated with Rayleigh losses with an increasing volume to accommodate the temperature rise. This in turn resu lts in an increase in the velocity of the gas

PAGE 18

18 flow. This is directed over th e turbine's blades provi ding power to the compressor and finally is passed through a nozzle generating thrust. There are two categories of combustors, annul ar and can as described by Figure 1-1. Can combustors look like cans and are mounted around the engine core. They can be easily removed for maintenance and provide convenient plumbing for fuel injection. Annular combustors are more compact and lighter. Modern jet engines usua lly have annular combustors but efficient can combustor arrangements can be used for ground based operations. Double annular combustors are being introduced to reduce emissions. The combustion efficiency [6] is a measure of the completeness of fuel combustion that takes place in the combustor chamber. If com bustion is incomplete not all of the energy potentially available is released and the non-co mbusted components are released as undesirable pollutants including NOx, SOx, Polycyclic Aromatic Hydrocar bons and char. The requirements for complete combustion include sufficient resi dence to complete mixing and chemical reaction completion to take place. Almost 100% efficiency is achieved in current engines; however as the temperature increases the efficiency tends to diminish. Under these c onditions nitrogen oxides are formed. Therefore measures to eliminate NOx emissions are needed. A range of new and existing technologies can be employed to help gas turbine operators and reduce NOx. The most common methods involve inte rnal changes to the combustion system of the machine and/or the additi on of a selective catalytic reduc tion (SCR) [7-9] systems to the exhaust [7]. For combustion systems, there are three major modifications available: Wet combustor [10]: a process that injects either steam or water into the combustion system and replaces nitrogen with steam dilu ent to lower the emissions levels, typically from 30ppm NOx Catalytic combustor: a new technology that employs a catalyst dire ctly in the combustion chamber to lower the emissions reportedly from 1.5ppm NOx

PAGE 19

19 Dry emissions reduction (DER) combustor [ 6, 11]: a method that utilizes staged combustion and fuel/air mixing to lower the em ission levels as low as 9ppm NOx without the introduction of steam or wa ter during natura l gas operation. An SCR system typically utilizes a heavy me tal catalyst injected with anhydrous ammonia to lower the emissions levels. There are two rang es of catalyst: a low-temperature catalyst that can be used in combined-cycle applications and high-temperature catal ysts, which are designed for simple-cycle applications. A DER combustor premixes the fuel and air in the combustion chamber to achieve a uniform and lean fuel/air mi xture. In these combustors the fuel and air are premixed to avoid stiochiometric fuel air mixtures that yield the highest temperatures hence large quantities of pollutants. The lean mixture result s in a cooler flame temperature when compared to a standard diffusion (non-premixed) combustor. Because NOx formation is an exponential function of the flame temperature, NOx levels can be reduced from approximately 150ppm in a standard combustor to less than 9ppm in a DER combustor. Catalytic Combustion: a Subset of Catalysis Catalytic combustion includes several essentia l processes: (1) diffu sion of the reactants from the gas phase to the catalytic surface, (2) adsorption of the reactants onto the catalytic surface, (3) movement of adsorb ed species, (4) reaction on the surface of the catalyst, (5) desorption of the products from the surface, and (6) diffusion of the products from the catalytic surface to the gas phase. Depending on the conditi ons, each of these processes can be rate limiting. Since measuring chemical activities ne ar or on a catalyst surface is difficult, experimental data of surface kinetics, temperature, or concentrations of gas phase species near the catalyst surface are scarce. As a result, catalytic combustors have conventionally been modeled as a black box that produces a de sired amount of fuel conversion. A few basic concepts of catalysis are described below:

PAGE 20

20 Catalysis and Catalysts The phrase catalysis was coined by Jns Jakob Berzeliu s [12, 13] in 1835 who was the first to note that certain chemicals speed up a reac tion. In chemistry and biology catalysis is the acceleration, i.e. increase in rate of a chemical reaction by means of a subs tance, called a catalyst and by reducing the activation energy. The catalyst in itself is left unchanged by the reaction. Figure 1-2 shows catalysis taking pl ace at reduced activation energy. In chemistry a catalyst is a substance that decreases the activation energy of a chemical reaction without itself being change d at the end of the chemical r eaction. Catalysts participate in reactions but are neither reactants nor products of the reaction they catalyze. An exception is the process of autocatalysis which is a form of catalysis in which one of the products of a reaction serve as a catalyst for the reaction. Examples of autocatalysis are ozone depletion, binding of oxygen by hemoglobin, reaction of permanganate w ith oxalic acid. More generally, anything that accelerates a chemical reaction can be defined as a "catalyst. A promoter is an accelerator of catalysis, but not a catalyst by itself. A reaction inhibitor inhibits the working of a catalyst. Catalysts and Reaction Energetics Catalysts work by providing an alternative m echanism involving a different transition state and lower activation energy. As a result more mol ecular collisions have the energy needed to reach the transition state. Hence, catalysts can perform reactions that, albeit thermodynamically feasible, would not run without th e presence of a catalyst, or perform them much faster, more specific, or at lower temperatur es. This means that catalysts reduce the amount of energy needed to start a chemical reaction. Catalysts cannot make energetically unfavorable reactions possible they have no effect on the chemical equilibrium of a reaction because the rate of both the forward and the reverse

PAGE 21

21 reaction are equally affected. Th e net free energy change of a reaction is the same whether a catalyst is used or not; the catalyst just makes it easier to activate. The SI derived unit for measuring the catalytic activity of a catalyst is the katal [14], which is a mole per second. The degree of activity of a catalyst can also be described by the turn over number or TON and the cat alytic efficiency by the turn over frequency, TOF. TON is defined as the number of moles of substrate that a mole of catalyst can c onvert before becoming inactivated. An ideal catalyst would have an infinite turnove r number in this sense, because it wouldn't ever be consumed, but in actual practice one ofte n sees turnover numbers which go from 100 to a million or more. The term turn over frequency is used to refer to the turnover per unit time. Types of catalysts It is possible to divide catalytic systems into two distinct categories. Homogeneous catalysis [15] is when the catalyst is of the same phase as the reactants and no phase boundary exists. Heterogeneous catalysis [15] is when a phase boundary separates the catalyst from the reactants. Kinetics of Catalyzed Reactions The primary effect of a catalyst on a chemical re action is thus to increase its rate, and this must mean to increase its rate co-efficient. The consequential effects may be analyzed in terms of either the collision theory or the absolute rate theory. According to the collision theory, the rate coefficient k is given by ) / exp( RT Ea Z P k (1-1) where P is the so-called steric factor which is introdu ced as a factor into the simple versions of collision theory of reactions to take care of the fact that the reaction probability depends on the

PAGE 22

22 certain mutual orientations of the reactant molecules, Z the collision frequency, Ea the activation energy, R the gas constant and T the absolute temperature. For example a unimolecular transformation (A B) in which the slow step is the adsorption of the reactant exhibits a collision frequency which is then the number of collisions per unit time between the reactant molecules and th e catalytic sites or species. The concentration of the latter will be much smaller, by a factor of some 1012 than the number of collisions between reactant molecules alone, which is relevant to the uncatalyzed reaction but irrelevant to the catalyzed reaction. Therefore if the catalyzed reaction is to compete effectively with the uncatalyzed reaction, then its e xponential term must be some 1012 larger [16], which means that its activation energy must be about 65 kJmol -1 less. There may be some relief in the form of a higher steric factor but this is unlikely to contribute more than a factor of 102 or 103 at most, and the main conclusion is not really altered. Negl ecting this effect, activ ation energy of 65 kJmol -1 only makes the rates of the catalyzed and uncatal yzed reaction equal: this scarcely represents efficient catalysis [16], for which activa tion energy difference must exceed 100 kJmol -1. In terms of the absolute rate theo ry, the rate coefficient is given by exp(/) KT kGRT h (1-2) where K is the Boltzmann constant, G is the Gibbs free energy of activation and h is the Plancks constant, and so the effect of a catalyst must be to decrease the free energy of activation of the reaction. The entropy of activ ation in a catalyzed reaction w ill usually be less than in the corresponding uncatalyzed because the transition state is immobilized on the catalyst surface with the consequent loss of translational fr eedom. There must therefore be a corresponding decrease in the enthalpy of activ ation to compensate for this, or more than to compensate if

PAGE 23

23 efficient catalysis is desired. Thus according to either theory, the activation energy for a catalyzed reaction ought to be less than for the same catalyzed reaction. Mechanism of Catalyzed Reactions: Making and Breaking Bonds Reactions catalyzed by a heterogeneous catal yst can be represented by a sequence of processes. Reactant molecules are adsorbed at ac tive sites onto the surface of the catalyst. This involves the formation of weak bonds between r eactant molecules and the catalyst which causes other bonds in the reactant molecu le to be stretched and weaken ed. The weakened structure is converted to another complex that is essentially the produc t attached to the catalyst. Finally, this complex breaks down to release the product molecu le which moves away to leave the catalyst surface ready to interact with another reactant molecule. Several processes occur during heterogeneous catalysis. They are as follows: Adsorption [15-17] is a process that occurs when a liquid or gas (called adsorbate) accumulates on the surface of a solid or liquid (adsorbent), forming a molecular or atomic film (adsorbate). It is diffe rent from absorption, where a substanc e diffuses into a liquid or solid to form a "solution". The term sorption encompasses both processes, while desorption is the reverse process. Adsorption is operative in most natural physi cal, biological, and chemical systems, and is widely used in industrial applications such as activated charcoal, synt hetic resins and water purification. Adsorption, ion exchange and ch romatography are sorption processes in which certain adsorptives are selectively transferred fro m the fluid phase to the surface of insoluble, rigid particles suspended in a vessel or packed in a column. Similar to surface tension, adsorption is a consequence of surface energy. In a bulk material, all the bonding requirements (be they ioni c, covalent or metallic) of the constituent atoms of the material are filled. But atoms on the (clean) surface experience a bond deficiency,

PAGE 24

24 because they are not wholly surrounded by other atoms. Thus it is energetically favorable for them to bond with whatever happens to be availa ble. The exact nature of the bonding depends on the details of the species involved, but the adsorbed material is generally classified as exhibiting Pysisorption or Chemisorption. Pysisorption [15-17] is a type of adsorption in which the adsorbate adheres to the surface only through Van der Waals force (weak intermolecular) interactions. It is characterized by: Low temperature, always under the crit ical temperature of the adsorbate Type of interaction: Intermolecular forces (van der Waals forces) Low enthalpy: H < 20 kJ mol -1 Adsorption takes place in multilayers Low activation energy Chemisorption [15-17] is a type of adsorption whereby a molecule adheres to a surface through the formation of a chemical bond, as opposed to Pysisorption. It is characterized by: High temperatures. Type of interaction: strong; covale nt bond between adsorbate and surface. High enthalpy: H 400 kJ mol -1 Adsorption takes place only in a monolayer. High activation energy Desorption [15-17] is a phenomenon and process oppos ite of sorption (that is, adsorption or absorption), whereby some of a sorbed substance is released. This occurs in a system being in the state of sorption equilibrium between bulk phase (fluid, i.e. gas or liquid solution) and an adsorbing surface (solid or boundary separating two fluids). When the concentration (or

PAGE 25

25 pressure) of substance in the bul k phase is lowered, some of the sorbed substance changes to the bulk state. Surface Coverage [14-17] is the number of adsorbed molecules on a surface divided by the number of molecules in a fill ed monolayer on that surface Fractional surface coverage [14-17] is the ratio of the amou nt adsorbed at a given pressure to the maximum amount that the surface can take up. Sticking Coefficient [14-17] is the ratio of the rate of adsorption to the rate at which the adsorptive strikes the total surface, i.e. covere d and uncovered. It is usually a function of the surface coverage, of temperature and of the deta ils of the surface structure of the adsorbent. Mechanism of Heterogeneous Catalysis The sequence of steps shown in Figure 1-3 c onstitutes the mechanism of heterogeneous catalysis. There are a number of theories proposed that capture these steps in sequence. The three most widely known and used in modeling surf ace chemistry and surface reactions are (1) Langmuir-Hinshelwood-Hougen-Watson (LHHW) [14,18,19] shown in Figure 1-4, (2) EleyRideal (E-R) [18,19] shown in Figure 1-5 and (3) the Quasi-Homogeneous [20] mechanism. Langmuir-Hinshelwood-Hougen-Watson (LHHW) proposes three steps to capture a surface reaction. They are (A) two molecules adsorb onto the surface, (B) they diffuse across the surface and interact when they are close and (C ) a molecule is formed which desorbs. The assumed sequence for the LHHW mechanism in an equation would take the following form Step 1: A + S = (AS) Step 2: B + S = (BS) Step 3: (AS) + (BS) = (CS) + (DS) Step 4: (CS) = C + S Step 5: (DS) = D + S where the letter S denotes a surface species. The assumptions made fro the LHHW derivations are:

PAGE 26

26 The rate determining step is controlled by the surface reaction-step 3. Uniformly energetic adsorption sites, Monolayer coverage, A dual site mechanism is used for the reaction. Eley-Rideal (E-R) also proposes a three steps mechan ism to capture surface reaction, including (A) molecules adsorb on the surface, (B ) another atom passes by which interacts with the one on the surface and (C) a molecule is formed which desorbs. The assumed sequence for the E-R mechanism in an equation would take the following form Step 1: A + S = (AS) Step 2: (AS) + B = (CS) + (DS) Step 3: (CS) = C + S Step 4: (DS) = D + S where the letter S denotes a surface species. The assumptions made fro the E-R derivations are: The rate determining step is controlled by the surface reactionstep 2, Uniformly energetic adsorption sites, Monolayer coverage, A single site mechanism is used for the reaction. Quasi-Homogeneous catalysis does not allow for any ad sorption of the molecules on the catalyst surface, in contrast to the LHHW and ER mechanisms. The solid catalyst is assumed in the same phase as the reactants. Catalytic combustion is a chemical process of the oxidation of fuel, initiated at the surface of the catalyst. This heterogeneous catalysis fo llows one or more of the above mechanisms of reactions. Based on LHHW, methane oxidation on Platinum (Pt) surface was proposed by Deutschmann [1, 21]. It involves a 24 step react ion mechanism and involves 7 species in the gas

PAGE 27

27 phase and 11 species in the adsorbed phase. Table 1-1 shows the equations involved in the mechanism proposed for methane oxidation on Plati num. This surface mechanism is used in this study. Equations 1, 3, 4, 5, 7, 8, 10, 15, 19, 20, 21 and 22 are classical examples of adsorption. Equations 2, 6, 9, 11, 16 and 17 are examples of desorption and Equations 12, 13, 14, 18, 23 and 24 represent complex product formation as a result of surface reactions.

PAGE 28

28 Figure 1-1 Types of combustors Figure 1-2. Catalysis taking place at reduced activation energy

PAGE 29

29 Step 1 Reactants + Catalyst Step 2 Reactants/Catalyst complex Step 3 Products/Catalyst complex step 4 Products + Catalyst Figure 1-3. Steps involved in heterogeneous catalysis Figure 1-4. Steps involved in Langmuir-Hinshelwood-Houge n-Watson (LHHW) Catalysis Figure 1-5. Steps involved in Eley-Rideal (E-R) catalysis

PAGE 30

30 Table 1-1. Deutschmann's Surface Reaction M echanism, A is Pre Exponential Factor (cm.mol.sec), is Temperature Exponent and Ea (KJ/mol) No. Reaction A (cm.mol.sec) Ea (KJ/mol) 1 H2+2Pt(s)=>2H(s) 4.60 x 10-2 0 0 2 2H(s)=> H2+2Pt(s) 3.70 x 1021 0 67.4 3 H+Pt(s)=>H(s) 1.00 0 0 4 O2+2Pt(s)=>2O(s) 1.8 x 1021 -0.5 0 5 O2+2Pt(s)=>2O(s) 2.30 x 10-2 0 0 6 2O(s)=> O2+2Pt(s) 3.70 x 1021 0 213.2 7 O+Pt(s)=>O(s) 1.00 0 0 8 H2O+Pt(s)=>H2O(s) 0.75 0 0 9 H2O(s)=>H2O+Pt(s) 1.00 x 1013 0 40.3 10 OH+Pt(s)=>OH(s) 1.00 0 0 11 OH(s)=> OH+Pt(s) 1.00 x 1013 0 192.8 12 O(s)+H(s)<=> OH(s)+Pt(s) 3.70 x 1021 0 11.5 13 H(s)+OH(s)<=> H2O(s)+Pt(s) 3.70 x 1021 0 17.4 14 OH(s)+ OH(s)<=> H2O(s)+O(s) 3.70 x 1021 0 48.2 15 CO+Pt(s)=> CO(s) 8.40 x 10-1 0 0 16 CO(s)=> CO+Pt(s) 1.00 x 1013 0 125.5 17 CO2(s)=> CO2+Pt(s) 1.00 x 1013 0 20.5 18 CO(s)+O(s)=> CO2(s)+Pt(s) 3.70 x 1021 0 105.0 19 CH4+2Pt(s)=>CH3(s)+H(s) 1.00 x 10-2 0 0 20 CH3(s)+Pt(s)=>CH2(s)+H(s) 3.70 x 1021 0 20.0 21 CH2(s)+Pt(s)=>CH(s)+H(s) 3.70 x 1021 0 20.0 22 CH(s)+Pt(s)=>C(s)+H(s) 3.70 x 1021 0 20.0 23 C(s)+O(s)=>CO(s)+Pt(s) 3.70 x 1021 0 62.8 24 CO(s)+Pt(s)=>C(s)+O(s) 1.00 x 1018 0 184.0

PAGE 31

31 CHAPTER 2 EXPERIMENTAL STUDIES OF CATALYTIC COMBUSTION Combustion Diagnostics The mechanisms of heterogeneous catalysis refe rred so far models or to say predicts the possible manners in which a combustion process would be initiated. However, the whole purpose of any combustion analysis is rather incomplete without adopting a suitable diagnostics approach. Three most widely diagnos tic approaches are (A) Mass Sampling, (B) Optical Diagnostics and (C) Com putational Diagnostics. (A) Mass sampling for species concentration measurement Two most widespread techniques of mass sampling are mass Spectrometry and gas Chromatography. Mass sampling techniques using mass spectro metry is during combustion. The technique gives a non intrusive measurement of radicals and reaction products dur ing combustion. Nakra [22] has used mass spectrometr y to predict the thermal deco mposition of JP-10, Seery [23] probed flame species of JP-10 and Korobeinic hev [24] has probed th e combustion and the thermal decomposition of GAP. The technique is now widely being used in measuring the concentrations of in situ radicals and gas phase reaction products formed during catalytic combustion. Schwiedernoch [50] has investig ated the cata lytic combustion and hydrogen assisted catalytic par tial oxidation of methane on platinum Horn [34, 57] carried out mass spectra measurements recorded during catal ytic partial oxidation of methane at 1500K on a Pt catalyst. The detection of CH3 radicals was successfully demonstrated the formation of which is the most important step in methane combustion on platinum.Grunwaldt [35] studied the structure of a heterogeneous catalyst in side a catalytic micro-reactor during the partial oxidation of methane and used mass spectrometry to analy ze reaction gas products. Yeh [36] used gas

PAGE 32

32 chromatography for the characterization and determination of gas products of methane combustion on hexaluminate catalyst. (B) Optical Diagnostics-Laser In duced Fluorescence in Combustion In Laser-Induced Fluorescence (LIF), an atom or molecule gets exc ited by the absorption of a laser photon. When the atom falls back into the ground state after a certain time (typically 1100nsec), it undergoes radiative emission, known as fluorescence. Since the fluorescence intensity is dependent on parameters such as ground state population, chemical environment, pressure, and temperature, laser-induced fluoresce nce measurements have become a vital tool in physical chemistry. For example, some applicati ons include the investigation of elementary chemical reactions and trace analytics down to sub-ppm concentrations. In combustion diagnostics, LIF measurements (also called PLIF in planar illumination) are widely utilized. They allow for the qualitative and quantitative detection of temperatur es, flame radicals and combustion intermediates such as OH, C2, CH, CH2O, as well as pollutants like NO and CO. Since these radicals are reaction in situ products in catalytic com bustions LIF is another diagnostics. Dogwiler [34] used PLIF to map the OH concentration field in the streamwise direction while investigating the homogeneous ignition of meth ane air mixture over platinum catalyst. (C) Computational Diagnostics Computational fluid dynamics (CFD) comp lements experiments with theoretical predictions as a research tool to produce multi-scale information which can not be obtained using any other technique. There are several possible solution techni ques in CFD investigation of reacting flows and a choice needs to be made for the proper technique that is to be applied in the present research. Combustion diagnostics is no w an investigative technique in catalytic combustion too. Modeling the surface reaction mechan ism, chemical kinetics of the reactions on

PAGE 33

33 the surface, predicting the su rface coverages of species on the catalyst surface along with modeling the transition from the heterogeneous ca talysis to homogeneous are some of the most fundamental areas of catalytic combustion that are studied using CFD techniques. Deutschmanns [1-3, 21] surface chemistry m echanism for the combustion of methane on platinum is among the first using CF D to model catalytic combustion. Motivation The motivation behind being able to employ cat alytic combustion is to reduce emissions and being able to create a pathway of low activ ation energy for the reactions to take place. The implementation of a noble-metal catalytic comb ustor in a natural-gas fired turbine for NOx (nitrogen oxides) reduction has drawn great attenti on in recent years (Dalla Betta [37]). Currently NOx emissions from stationary gas turbine syst ems are controlled either by lowering the combustion temperature with wate r injection or by removing NOx through exhaust gas treatment such as selective catalyt ic reduction. In a cata lytic combustor, a major portion of fuel conversion takes place on the catalyst surface ; consequently, the gas phase NOx production route via the prompt (or Fenimore [38, 39]) pathway is avoide d (Schlegel [38, 39]). In addition, the peak gas phase combustion temperature is substantially re duced leading to low th ermal (or Zeldovich [38, 39]) NOx formation rate. However, the use of a so lid catalyst then introduces two separate phases, namely, a gas phase wherein the particip ant reactants flow and a gas-solid (catalyst surface) interface phase, wherein th e reactants flow over the surface of the catalyst. Both these phases play a major role in the completion of combustion. Catalytic comb ustion has a number of potentially important and practical applications like improved fl ame stability, generation of lowtemperature process heat, and reduced pollutant emissions. It is expect ed that quantitative simulations will play an important role in accelerating the devel opment in this field. Therefore, extensive research has been started to understand the elementary chemical processes occurring at

PAGE 34

34 gas-surface interfaces [40], as well as their c oupling to the surrounding flow and the gas phase reaction mechanism. The aim is to get a qua ntitative understanding of catalytic combustion. Methane is the major constituent of natural gas, and reactions of the methyl radical are among the rate limiting steps in the gas phase com bustion of larger hydrocarbons. Thus, understanding of heterogeneous ignition and ex tinction of methane and the role of intermediate radicals at surfaces is crucial to use heterogeneous com bustion processes efficiently and safely. Hence attempts have been made to understand heteroge neous catalysis by way of the kinetic approach. The fuel used is no longer limited to the use of natural gases only. Atte mpts are made to use mixtures of combustible gases on a catalytic surface to combine the combustion powers of two or more fuels. One such fuel of intere st is syngas which is a mixture gaseous H2 and CO. Both these gases oxidize in the presence of oxidizer usually air to form H2O and CO2 respectively and both these oxidation are exothermic process. He nce the reaction kinetics of species in various fuels on a given catalyst is of utmost importance. An attempt to understand the reaction pathway followed by different species on a given catal yst is studied in our research work. Basically the method of chem ical kinetics involves the qua ntitative determination of reaction products and their depe ndence on time and temperature, th e expression of rates in terms of concentrations and temperat ure, and the construction of a mechanism to fit the observations41. Hence models for such a mechanism for various heterogeneous combustion processes have been tried for methane oxidation on Platinum surface. This thus leads us to understanding various surface chemistry models in details. Surface Chemistry Models One-Step Surface Reaction Due to limited knowledge of the elementary su rface reaction kinetics, numerical studies of methane catalytic combustion were often perfor med with a single step global surface reaction.

PAGE 35

35 Song [42] used a single step su rface reaction for predic tions of methane catalytic combustion in a stagnation flow. With this simple chemical kine tic model, their calculat ions showed success in predicting surface ignition/extinction temperatures of lean methane-air mixtures. Since surface ignition temperatures of methaneair catalytic combustion are low, ca. 600C, as predicted by Williams [43], the heterogeneous ignition process is dominated by the surface reaction due to its much lower activation energy compared to those of gas phase reactions. As calculations with a single-step reaction do not involve radicals, the well predicted ignition temperatures of lean methane-air mixture by Song [42] suggest th at, under fuellean condi tions, the interaction between catalytic and gas phase re actions is not important because the heterogeneous ignition is driven by the heat release of surface reaction. However, for high temperature conditions (> 1200 K), the interaction between catalytic and homogene ous reactions via radicals such as hydroxyl (OH) and O atom may potentially affect the igniti on process (Pfefferle, Griffin [44]). In order to include the radical interacti on between surface and gas phase at high temperatures, Markatou [45] modified the single step surface reaction model by introducing a coefficient to regulate the amount of OH desorbing from surface. The valu e of this coefficient was determined from experimental data. Their results show that th e OH desorbed from the surface enhances the gas phase reactions and, hence, the generation of radicals in the bounda ry layer for surface temperatures above 1300 K. Markatou suggested the need of detailed su rface kinetics and gassurface energy balance to properly couple phase and surface processes in catalytic combustion calculations. MultiStep Surface Reaction Mechanisms Previous numerical studies of catalytic methane oxidation by using multiple-step surface reactions have been reported by Hickman and Schmidt [46], by Deutschmann and Behredt [40]. These multi-step surface reaction mechanisms were developed with available kinetic and thermal

PAGE 36

36 data along with several assumptions, such as LangmuirHinshelwood type surface reaction mechanism, dissociative adsorption of O2 and CH4, perfect catalyst su rface, no substrate diffusion, and monolayer surface coverage. These re sulting mechanisms consist of several basic reactions: adsorption of reactants (O2 and CH4) and intermediate species (CO, H2, and OH), surface reactions of adsorbed specie s, and desorption of products (CO2 and H2O) and intermediate species. Details of these surface reactions are tabu lated in Table 1-1. The surface reaction rates are described by an Arrhenius expr ession or by an initial sticking coefficient for adsorption processes. With the assumption of Langmuir-Hinshel wood surface reaction mechanism, the onset of surface ignition is determined by the competition between O2 and CH4 for surface sites. The dominant limiting process changes from oxygen desorption to methane adsorption as the surface temperat ure increases. When the surface temperature is low, due to the O2 adsorption process, the catalyst surface is en tirely covered by adsorbed O atom. The heat generated by this reaction raises the surface temperature. As the surface temperature increases, the su rface coverage of O atom drops and the methane adsorption reaction starts to increase. When the surface temperat ure reaches a certain point, the methane adsorption rate exceeds the oxygen adsorption rate such that methane adsorption reaction becomes dominant. Consequentl y, more heat is generated and ignition soon takes place on the catalyst surface. Similar surf ace ignition processes have been postulated by Behrendt [40]. Surface ignition temperature is another important feature which should be properly predicted by the surface reaction mechan ism. Griffin and Pfefferle [47] measured methane surface ignition temperatures on a platin um wire. Their experimental results show the surface ignition temperature decreases with mixture equivalence ratio.

PAGE 37

37 Moreover, the surface temperature prediction provides useful engineering information, such as the maximum temperature and temper ature gradients, for better catalyst designs. Although some groups used Pt foils and N2 as diluent [48, 49], other groups Pt wires [47], or Pt coated monoliths and Ar dilution (this work), no effect of dilution or shape of catalysts was observed. Hence two cases of methane oxidation, namely catalytic combustion of methane in absence of hydrogen (power heat ed) and catalytic part ial oxidation of meth ane in presence of hydrogen (hydrogen assisted heating) is being studied on various flow configurations. The discussion beneath is based on platinum hone ycomb monolith for both case studies, by Schwiedernoch, although as prove n, the shape of the catalyst does not affect the studies. Catalytic Combustion of Methane on Platinum Research works based on the catalytic com bustion of methane on platinum was conducted by Schwiedernoch [50]. The chemical reactions on the platinum surface were modeled using 27 irreversible elementary chemical reactions amon g 9 gas phase species and 11 surface species [1, 50]. In order to estimate the potential influen ce of gas-phase reactions on the methane oxidation, single channel calculations using additional 14 gas-phase species and 168 gas-phase reactions were performed. The homogeneous reaction scheme was based on a reducti on of a large set of hydrocarbon oxidation reactions to C1 species However, for temperatures below 1200 K no significant influence on the conversion was obser ved. Hence, the presented simulations are performed using only surface reactions. The simula tions of this study were carried out by S. Tischer with DETCHEM versi on 1.4.2. [50, 51]. Our numerical simulation was conducted using FLUENT 6. Before ignition, the surface is primarily covered with oxygen, because the sticking probability of oxygen is higher than that of methane. With incr easing surface temperature a point will be reached where the adsorption/desorption equilibrium of oxygen shifts to desorption. This

PAGE 38

38 results in bare surface sites where CH4 can be adsorbed, followed by H abstraction, which leads to adsorbed C(s) and H(s) atoms reacting with the surrounding O atoms to form CO(s) and OH(s) immediately. Both molecules quickly form CO2 and H2O, which desorb leaving more free surface sites for CH4-adsorption. For CH4/O2-ratios below 0.1 (2 vol.-% CH4) the reaction is extinguished in a couple of minutes if the external heating is turned off. With the assumption of absence of gas-phase reactions, no CO was formed in the simulation. For CH4 feeds exceeding 4 vol.-% CH4/O2-ratio greater than 0.21), gas-phase reac tions become significant due to the high reaction temperature above 1200 K. Thus, simulations for CH4/O2-ratios exceeding 0.21, needs to include gas-phase reactions. Ho wever, in the study conducted in th is research a very fuel rich mixture was used. It was expected that methan e combustion on platinum for a rich mixture would also show similar results as when a lean mixture in Schwiedernochs [50] works were used. Catalytic Partial Oxidation of Meth ane on Platinum-Hydrogen Assisted There are two ways in which the catalyst surface is basically heated. Power heating [50] of the catalyst and hydrogen assi sted heating [1, 21, and 40]. The addition of hydrogen to the initial mixture for example may help to reduce the i gnition temperature, because the ignition of hydrogen on platinum occurs at room temperature. Deutschmann [21] investigated the dependence of the hydrogen-assisted light-o ff of methane on platinum and on methane concentrations. They suggested that the light -off was primarily determined by the catalyst temperature that is a result of the heat release due to catalyt ic oxidation of hydrogen. The addition of H2 to the feed gas for the catalytic combustion of CH4 on Pt has a huge impact on the reaction behaviour. H2 is very likely to contribute more th an only heat for the ignition to the oxidation. Since H2 removes O(s) from the surface by formation of H2O, it is likely that H2 has a great impact on the re action kinetics of CH4 combustion. Different ignition and maximum

PAGE 39

39 temperatures, conversions, and yields of CO2 indeed indicate an important role of H2 on the kinetics causing considerable differences betw een hydrogen-assisted a nd thermally initiated reaction .In both cases, the light-off temperat ure decreases Figure 22 with increasing CH4/O2 ratio. At 400 K the surface is mainly expected to be covered with O(s) and hardly any Pt vacancies are left. Increasing the wall te mperature (500 K) shows that now more CH4 reaches the surface. This is due to a very small amount of C and CH formed by dissociation of CH4, which is necessary for the combustion reaction. At 600 K the first product formation is i ndicated by increasing amounts of CO and H2O. However, it can be clearly seen that the addition of H2 has a positive effect on the ignition temperature, which is more than 100 K lowe r than in case of common combustion. Here, not only CH4 and O2 compete for free vacancies but also H2, which occupies the surface first. On a Pt catalyst, room temperature is sufficient to start exothermic H2 combustion, which heats up the monolith. The higher maximum temperature blue symbols in Figure 2-2 (a) of the classical catalytic combustion compared to the H2-assisted reaction (red) c ould be due to higher exothermic methane conversion as illustrated in (c). This phenomenon is best explained by surface kinetics. In case of H2 addition water will be formed immediately by oxidation of H2. Therefore, a certain amount of the surface area is always occupied by H2O(s). This can be seen clearly by comparing these results with the surface coverage calcu lations in Figure 2-3 at 900 K, where with H2 more surface sites are occupied by OH and H2O than in absence of H2. The adsorption of CH4, dissociation as well as the reaction to CO and CO2 respectively has to happen on the remaining vacancies downstream of the m onolith. The high flow rate and the short length of the catalyst result even for these lean mixt ures in a very short contact time of some milliseconds CH4 breakthrough occurs. The added H2 occupying the surface first under

PAGE 40

40 dissociation, reacts with the adsorbed O(s). The produced water leaves the surface continuously generating vacancies, which methane can occupy, too. Figure 2-3 (right side) shows that the coverage of the surface is completely different, if H2 is added. At 400 K already a huge amount of CO is formed, the surface is mostly covered by carbon, and CH and CH2 can be found on the surface (not shown), too. Compared to the catalytic combustion without H2, the number of free adsorption sites on Pt is increa sed by a factor of about 1000. Th e equilibrium calculations as depicted in Figure 2-2 (d) dashed lines show that in case of H2 addition to the combustion gas more O2 (in the order of 20%) should be convert ed; the experimental results support the calculation. Additionally, the trend of increasing conversion of O2 with increasing CH4/O2 ratio is observed, too. The reduced yield of CO2 in case of addition of H2 to the combustion gas compared to the conventional combustion can be explained with the equilibrium calculations, as well. Due to the higher concentr ations of reducing agents CH4 and H2 in presence of H2, more reduced species will be observed as reaction products. In this case more CO is formed. However, not only H2 is responsible for enhanced formation of CO, but also the higher temperatures. With increasing temperatures, the CO + O2 --> CO2 equilibrium favor the formation of CO. In both cases gas-phase reactions occurred at a critical CH4/O2 ratio. For pure combustion of methane, they started at a CH4/O2 ratio exceeding 0.2 and in case of the H2-assisted combustion at a ratio of 0.35. The gas-phase reactions were indicated by a sudden rise of the concentration of CO and temperature. Catalytic Combustion of Syngas on Platinum Syngas is a gaseous mixture of H2 and CO. From the hydrog en assisted catalytic combustion of methane on platinum performed in the works of Schwiedernoch [50], it is

PAGE 41

41 observed that H2 is very likely to contribute more than only heat for the ig nition to the oxidation. Since H2 removes O(s) from the surface by formation of H2O, it is likely that H2 has a great impact on the reaction kinetics of CH4 combustion. Due to reducing agent H2, more reduced species will be observed as reacti on products. In this case more CO is formed. In this present study, the syngas reaction mi xture comprises of H2, CO, CH4 and CO2. Due to the presence of H2, it is expected that for the given mixture, it would still act as a reduc ing agent and thus the formation of reduced species such as CO is more likely to be formed than the fully oxidized CO2. This prompted us to believe that the same su rface reaction mechanism as used by Deutschmanns [21] during the hydrogen assisted catalytic combus tion of methane would be able to model the catalytic combustion of syngas as well. Catalytically Stabilized Thermal Combustion (CST) In Catalytically Stabilized Combustion (CST) [52, 53] partial fuel conversion is accomplished heterogeneously in burners with a large surface-to-volume ratio, such as catalytically-coated honeycomb monoliths. Comple te fuel conversion is attained in a postcatalyst homogeneous combustion zone. This pro cess leads to substant ial reduction of NOx emissions (typically < 3 ppm) as NOx is produc ed exclusively from the gaseous (homogeneous) reaction path. Thus the overall predictions of CST combusti on take into account both the heterogeneous and the homogeneous phase reacti on schemes. The thermal interactions between the two schemes are instrumental in predicting the CST combustion accurately. Catalytic Activation Index (CAI) In our present work, combustion experiments we re performed on four catalysts of different chemical composition. A basis for comparing the pe rformance of the catalys ts is therefore very essential. An overall performance of any cat alyst can be judged by the amount of products formed. Since a catalyst actually reduces the ac tivation energy required for any reaction, it means

PAGE 42

42 that a catalyst increases the number density of reactants available for reaction by reducing the activation energy thereby increasing the probability of more reactants part icipating in a given reaction. However, a new index called the Catalytic Activation Index (CAI) [54] conceptualizes the performance of one catalyst over the other. Th e value of CAI compares different catalysts. This is a non-dimensional number calculated from the delay period ( delay), combustion duration ( cd) and the lean misfire limit ( LLT). The delay period is defined as the time taken for a 5% heat release. The combustion duration is assumed to be the time period between 5% and 90% heat release55. These times would be achieved by running th e reacting case in FLUENT and reporting the temperature changes with respect to tim e. The lean misfire limit was determined experimentally by varying the fuel-air ratio. cat LLT base LLT cat cd base cd cat delay base delayCAI ) /( ) ( ] ) /( ) ( ) /( ) [(5 0 CAI higher than unity indicates catalytic activation. Based on this index, different catalysts can be compared and the apparent surface activation temperatures are obtained from CAI Ru Ea Ru Eaglobal surface/ ) / ( ) / ( The characteristic surface re action rate is obtained from ] ][ )[ ( / 1Oxidiser fuel W chY Y T k Kmol/s Where )] /( exp( ) ( ) (W surface W WT Ru Ea T A T k A characteristic temperature fo r a catalyst can be defined as the temperature beyond which the catalytic ac tivity increases expon entially. This temperature can be determined by calculating

PAGE 43

43 the Damkohler number (Da), based on the ratio of diffusion rate and the surface reaction rate. The physical meaning of unity Da is that the rate of reaction will be equa l to rate of diffusion. For each catalyst, the temperatur e at which Da becomes unity differs. The surface reaction is expected to increase after the temperature corresponding to unity Da. Hence, we propose that the critical temperature of the catalyst may be the temperature at which Da is unity. Figure 2-1. Surface Ignition Temper ature decreases with increase in methane equivalence ratio

PAGE 44

44 Figure 2-2. Methane Combustion with no H2 and with H2 (a) maximum temperature (b) methane ignition temperature (c) methane conve rsion (d) oxygen conversion and (e) CO2 yield

PAGE 45

45 Figure 2-3. Surface coverage of spec ies in methane combustion with no H2 and with H2

PAGE 46

46 Table 2-1. Deutschmann's Surface Reaction Mechanism, A is Pre Exponential Factor (cm.mol.sec), is Temperature Exponent and Ea (KJ/mol) No. Reaction A (cm.mol.sec) Ea (KJ/mol) 1 H2+2Pt(s)=>2H(s) 4.60 x 10-2 0 0 2 2H(s)=> H2+2Pt(s) 3.70 x 1021 0 67.4 3 H+Pt(s)=>H(s) 1.00 0 0 4 O2+2Pt(s)=>2O(s) 1.8 x 1021 -0.5 0 5 O2+2Pt(s)=>2O(s) 2.30 x 10-2 0 0 6 2O(s)=> O2+2Pt(s) 3.70 x 1021 0 213.2 7 O+Pt(s)=>O(s) 1.00 0 0 8 H2O+Pt(s)=>H2O(s) 0.75 0 0 9 H2O(s)=>H2O+Pt(s) 1.00 x 1013 0 40.3 10 OH+Pt(s)=>OH(s) 1.00 0 0 11 OH(s)=> OH+Pt(s) 1.00 x 1013 0 192.8 12 O(s)+H(s)<=> OH(s)+Pt(s) 3.70 x 1021 0 11.5 13 H(s)+OH(s)<=> H2O(s)+Pt(s) 3.70 x 1021 0 17.4 14 OH(s)+ OH(s)<=> H2O(s)+O(s) 3.70 x 1021 0 48.2 15 CO+Pt(s)=> CO(s) 8.40 x 10-1 0 0 16 CO(s)=> CO+Pt(s) 1.00 x 1013 0 125.5 17 CO2(s)=> CO2+Pt(s) 1.00 x 1013 0 20.5 18 CO(s)+O(s)=> CO2(s)+Pt(s) 3.70 x 1021 0 105.0 19 CH4+2Pt(s)=>CH3(s)+H(s) 1.00 x 10-2 0 0 20 CH3(s)+Pt(s)=>CH2(s)+H(s) 3.70 x 1021 0 20.0 21 CH2(s)+Pt(s)=>CH(s)+H(s) 3.70 x 1021 0 20.0 22 CH(s)+Pt(s)=>C(s)+H(s) 3.70 x 1021 0 20.0 23 C(s)+O(s)=>CO(s)+Pt(s) 3.70 x 1021 0 62.8 24 CO(s)+Pt(s)=>C(s)+O(s) 1.00 x 1018 0 184.0

PAGE 47

47 CHAPTER THREE EXPERIMENTAL AND COMPUTATIONAL SETUPS Experimental Setup The test setup is modeling a small part of a cat alytic combustor for a stationary gas turbine fuelled by syngas mixtures using a singl e surface covered with catalyst. The test setup shown in Figure 3-1 consists of a steel tube coated with catalyst on the outside, which is enclosed in a larger coaxial tube fitted with thermocouples, gas sample ports and optical windows for observation A rich fuel mixture flows in the annular sp ace between the two tubes while the center of the catalyst tube ca rries cooling air. Downstream of the catalyst the cooling air is allowed to mix with the rich catalyst exhaust. The air flowing through the annul ar portion is called the mixi ng air while the air through the center of the catalyst is called the cooling ai r. The fuel is never heated while the mixing as well as the cooling air are heated as per th e experiment specification. The fuel and the hot mixing air mixes in a predefined mixing zone before flowing through the annular portion. Both these streams of air are heated using two Omega Electric Heaters -AHP-3741 and AHP7561and these two air temperatures are read using K-type Omega thermocouples KMTXL062E-6. The Omega heaters are rated as 200W and 750W respectively and their safe working operation is ensured by the use of tw o Solid State Relays from Omega SSRL240DC10. The air stream temperatures are controlled using two Pro portional-integral-derivat ive (PID) controllers. There are four other similar K-type Omega thermocouples KMTXL-062E-6 that read the catalyst surface temperature, an intermediate dow nstream temperature, the temperature at the location where the rich catalyst exhaust mixes with the cooling air and the temperature of the stainless steel mass sampling tube, respectively. The two thermocoupl es that register the catalyst surface temperature are located at x/L values of 0.3 and 0.73, where x is the axial distance from

PAGE 48

48 the start of the catalyst tube and L is the leng th of the catalyst tube. These two thermocouples touch the surface of the catalyst. The thermocouple that registers the gas phase temperature when the combustion exhausts mix with the cooling air stream is located at x/L value of 1.2. A 2Daxisymmetry schematic of the combusto r section is shown in figure 3-2. In order to ensure that water vapor form ed during combustion does not condense in the stainless steel mass sampling tubes described below a provision of heating this tube up to a set point temperature of 130C was made using the same principles of PID controllers that was applied to maintain the two air streams at their respective set point temperatures. Since only the catalyst part of th e combustor is studied in this setup, the exhaust is then further diluted with excess air to achieve a non-flammable mixtur e which is exhausted. The mass flows of the two incoming air stream s as well as those of the constituent gases of the syngas, i.e. CO, CO2, H2 and CH4 were controlled by Alicat Scientific 16 series mass flow controllers, allowing full flexibility in the gas composition te sted. The operating ranges for these mass flow controllers are shown in Table 3-1, where SLPM and SCCM are Standard Liters Per Minute and Standard Cubic Centimeters per Minute respectively. Mass Sampling Hardware The physical location of mass sampling ports alo ng the axis of the catalytic combustor is shown in Figure 3-2. The axial positions of the two sampling ports are also shown in the figure. The test section has provisions for two mass sampling ports along the axial direction of the catalytic combustor. Sampling is done from onl y one of these during a given experiment while the other port is being used to hold a thermocouple for measuri ng the surface temperature of the catalyst. Thus when the upstream port samples gas species the downstream port records surface temperature of the catalyst at th at location and vice versa. This set up allows investigating the

PAGE 49

49 change in temperatures and species concentratio n at two different locatio ns on the surface of the catalyst. The sampling ports are swapped when r unning a different experime nt. They are located at x /L = 0.3 and x /L = 0.73, where L is the length of the catalyst, L = 22.8 cm and the x distance is measured from the locati on where the catalyst starts. These ports are 0.8 mm inner diameter steel tubes and are at right angles to the direction of flow, i.e. they are located radially inside the cylindrical catalytic combustor such that they touch the surface of the catalyst. A schematic diagram of mass sampling from the catalytic com bustor and subsequent, real-time analysis by a mass spectrometer is shown in Figure 3-3. The species were analyzed by Stanford Research Systems RGA-300 mass spectrometer, which uses el ectron ionization to i onize the sampled gas, RF quadrupole filter to sort spec ies according to thei r mass-to-charg e ratio, and Faraday cup to detect ion currents. The instrument is controlled and operated by software and associated electronics. The ionizer, filter and detector ar e enclosed in a clean vacuum chamber and require an operating pressure range of 10-4 torr (1.3 x 10-7 atm) to ultra high vacuum. Such low pressure is attained in two stages. In first stage, a rota ry pump brings the inlet pressu re down to about 60 mtorr (7.9 x 10-5 atm). In a second stage, a diffusion pump and a rotary pump operate in series to bring the pressure further down to vacuum conditions. Th e diffusion pump is surrou nded by a water jacket carrying cold water to keep the diffusion pump cool The water is cooled using a chiller. For the optimum working conditions of the diffusion pum p, it is specified to receive water at a temperature between 15C to 25C at a flow rate of 0.6 l/s. The spectrometer can detect species up to a mass-to-charge ratio of 300 and has a re solution of 0.5 AMU @ 10% peak height. The sensitivity factor of the instrume nt, defined as the signal detected per unit partial pressure of a given species (Amp / torr), varies for different gases. Hence calibration of the instrument was

PAGE 50

50 performed for the following gases: helium, nitrogen, oxygen and argon. The sensitivity factor of nitrogen is used as the baseline and sensitivity factors of other gases are normalized with this baseline. One limitation with the use of compressed air as oxidizer in combus tion analysis using Quadruple mass spectrometry is that the peaks of N2 and CO overlap at an AMU value of 28. Hence concentrations of CO and N2 in the combustion mixture coul d not be reported separately. Assuming N2 inertness, the mass fraction of N2 after combustion remains the same as known from inlet conditions. The partial pressure of species with AMU 28 was obtained from MS which was then converted into mole fraction and mass fr action. This mass fraction is a result of the contribution made by both N2 and CO if present. The differen ce between the mass fractions after and before combustion gives the ma ss fraction contribution due to CO. However, it was seen in a number of experiments that the calculated valu e of CO by adopting this method resulted in unrealistic values of COs mole fraction. A possible solution would be replacement of N2 in air with a different species. Argon use, attempte d here required renewed calibration. Given the higher Ar molecular weight, i.e. 40, the cal ibration factor for li ghter species e.g. H2 led to large errors and the procedure was discarded. Hence, in what follows compressed air was used as the oxidizer. Over experiments described in chapte r 4 the uncertainty of species measurements ranged from 0.9-1.3% for N2, 0.3-0.7% for O2, 0.5-0.7% for CH4, 0.4-0.6% for CO2 and 0.71.1% for H2O for all methane combustion cases and 0.1-0.4% for N2, 0.4-3.3% for CO, 0.11.6% for O2, 0.1-0.6% for H2, 0.3-1.1% for CO2, 0.2-1.5% for H2O and 0.1-1.1% for CH4 for all syngas combustion cases. Data processing The composition of gas in the catalytic com bustor was analyzed by the mass spectrometer in partial pressure of species vs. time mode. Th e species scanned were hydrogen (m/z=1 and 2),

PAGE 51

51 methane (m/z=15 and16), oxygen (m/z=16 and 32), (water (m/z=17 and 18), nitrogen (m/z=14 and 28), carbon monoxide (m/z= 28) and carbon monoxide (m/z=28 and 44) for combustion experiments. The local mole fraction of a given species in the sample was determined from the partial pressures of all the component species recorded by the mass spectrometer. The timeaveraged fuel mole fraction at a port was obtai ned by averaging the mole fractions obtained over the sampling time period. The species mole fractions were corrected us ing calibration factors for individual gases. The sensitivity or the calibra tion factor for the mass spectrometer was established using the following the procedure. The specific inlet composition for the case of methane as well as syngas composition was passed through the test s ection at room temperature. This represented the non-reacting flow of the gas mixtures through the combustor section. Only one of the sampling ports, located at x/L value of 0.3 was chosen to sample this gas mixture into the mass spectrometer. The mole fractions of the constituent gases in the gas mixture calculated from the partial pressures registered by the mass spectrometer must match the exact inlet mole fractions. Since the mass spectrometer reading was slightly different a fact or was formulated for those species for which the mole fraction obtained from the inlet specifications and mass spectrometer did not match. This factor was either less than unity or more than unity de pending upon whether the species was over-predicted or under-predicted from the speci es inlet composition. A gas mixture of each species and nitrogen at room temperature and at 700K was passed through the test section. The calibration factors at both temperatures were the same, thus, indicating the fact that the calibration factor was te mperature independent.

PAGE 52

52 Computational Setup Tools: The computational tools used were GAMBIT 2.2.30 for ge nerating the design of the catalytic combustor as well as generating the me sh over this design, a nd FLUENT 6.2.16 as the computational solver. The computa tional setup is therefore divide d into two parts, namely (A) the design of the computational domain (cataly tic combustor) and grid generation, both using GAMBIT [34] and (B) setting up and running the CASE, using FLUENT [35] solver. Computational Setup in Gambit The GAMBIT software package is designed to he lp analysts and designers build and mesh models for computational fluid dynamics and ot her scientific applications. GAMBIT receives user input primarily by means of its graphica l user interface GUI. The GAMBIT GUI makes the basic steps of building and meshi ng a model simple and intuitive, yet it is versatile enough to accommodate a wide range of modeling applicat ions. GAMBIT allows creating drawing files, dividing the drawing domain into gr ids (meshing the model), defining the boundary condition type and defining the continuum of the domain, i.e. assigning the domain as either solid or fluid. Computational Domain-Creating Drawi ng Files and Generating Mesh Files In order to ease the computat ional expense, the catalytic co mbustor design is divided into four separate zones. They co llectively capture (i) the individu al fuel and oxidizer feed line zones, (ii) the premixing zone of the fuel and th e oxidizer, (iii) the expa nsion zone of the fueloxidizer mixture over a CROSS-JOINT and (iv) the zone where the fuel-oxidizer mixture flows through the annular portion of the catalytic combustor over the surfa ce of the catalyst. The last three zones are collectively shown in Figure 3-1. The first zone is the fuel line and the oxidizer line, each being a quarter inch outer diameter tube. It is long enough to result in a fully developed flow for both the fuel and oxidizer lines. The catalytic combustor has a circular cross section in all the four zones. The meshing mode l includes boundary layers at all the circular

PAGE 53

53 edges as well as at that portion in the premixi ng zone where the fuel comes in contact with the oxidizer for the first time, face mesh at all the inlet faces of a ll the four computational zones and finally volume mesh over the entire volume of all th e zones. A brief discussion on the types and procedures of generating the mesh mentioned above is Appendix A. Computational Setup in FLUENT Fluent serves as the solver software, general-purpose CFD so ftware for a wide range of industrial applications along with highly-automated, specifically-focused packages. FLUENT models fluid flow and heat transfer in complex ge ometries providing complete mesh flexibility, including the ab ility to solve flow problems using unstructured meshes that can be generated about complex geometries with re lative ease. Supported mesh types include 2D triangular/ quadrilateral, 3D te trahedral/hexahedral/ pyramid/wedge, and mixed (hybrid) meshes. FLUENT also allows us to refine or coarsen ou r grid based on the flow solution. FLUENT is written in the C computer language and makes full use of the flexibility and power offered by the language. The commercial CFD software FLUENT is a fu lly-unstructured finite-volume CFD solver for complex flows ranging from incompressible (s ubsonic) to mildly comp ressible (transonic) to highly compressible (supersonic and hypersonic) flows. The cellbased discretization approach used in FLUENT is capable of handling arbi trary convex polyhedral elements. For solution strategy, FLUENT allows a choice of two numerical methods, either segregated or coupled. With either method FLUENT solves the governing integral equati ons for conservation of mass, momentum, energy and other scalar s such as turbulence and chemical species. Both segregated and coupled numerical methods employ a similar finite-volume discretization process but their approach to linearization and so lution of the discretized equati ons is different. A point implicit (Gauss-Seidel) linear equation solver is used in conjunction with an Al gebraic Multigrid (AMG)

PAGE 54

54 scheme to solve the resultant linear system for th e dependent variables in each cell. In the present study, laminar flame dynamics has been inve stigated by applying the FLUENT solvers. Program Capabilites The FLUENT solver has the following modeling capabilities: 2D planar, 2D axisymmetric, 2D axisymmetric with swirl (rotationally symmetric), and 3D flows Quadrilateral, triangular, hexa hedral (brick), tetrahedral, prism (wedge), pyramid, and mixed element meshes Steady-state or transient flows Incompressible or compressible flows, incl uding all speed regimes (low subsonic, transonic, supersonic, and hypersonic flows) Inviscid, laminar, and turbulent flows Newtonian or non-Newtonian flows Heat transfer, including forced, natural, and mi xed convection, conjugate (solid/fluid) heat transfer, and radiation Chemical species mixing and reaction, including homogeneous and heterogeneous combustion models and surface deposition/reaction models. Numerical Model A brief description of the numerical model fo llows. It is a full elliptical two-dimensional axisymmetric model for a steady, laminar two dimensional gaseous reactive flow with the additional presence of surface reactions. For surf ace chemistry, the react ion mechanism proposed by Deutschmann (Table 1-1) fo r the catalytic combustion of methane on platinum was used which included 22 irreversible reactions, 8 gase ous and 11 surface species (including Pt). The catalyst site density was taken as 2.7 x 10-8 Kgmol/m2 simulati ng a polycrystalline platinum surface. The platinum coating on top of a non-porous Al2O3 suppor t closely resembles such a surface. For gaseous chemistry, the five step reaction mechanism is a reasonable simplification

PAGE 55

55 of this study. The mechanism is listed in table. The Chemkin and FLUENT database is used to evaluate gaseous transport and surf ace species thermodynamic properties. Mathematical Model (A) Governing Equations The governing equations solved are those of continuity, momentum, energy and the species transport. This conservation e quation takes the following general form: Continuity: 0 ) .( v t (3-1) Momentum: F g p v v v t ) .( ) .( ) ( (3-2) where p is the static pressure, is the stress tensor, g andF are the gravitational and external body forces respectively, the stress tensor is given by ] 3 2 ) [( I v v vT (3-3) where is the molecular viscosity, I is the unit tensor and the second term on the right hand side is the effect due to volume dilation. Energy: h j j jS v J h T k p E v E t )) ( .( )) ( .( ) ( (3-4) where k is the ther mal conductivity, jJ is the diffusion flux of species j The first three terms on the right hand side represent energy transfer due to conduction, species diffusion and viscous dissipation respectively. hS includes the heat of chemical r eaction and any other user defined heat source.

PAGE 56

56 When solving conservation equations for chemical species, FLUENT predicts the local mass fraction of each species, Yi, through the solution of a convection-diffusion equation for the ith species. ().().iiiiiYvYJRS t (3-5) where Ri is the net rate of production of species i by chemical reaction (described later in this section) and Si is the rate of creation by addition from the di spersed phase plus any user-defined sources. An equation of this form will be solved for N-1 species where N is the total number of fluid phase chemical species present in the syst em. Since the mass fraction of the species must sum to unity, the Nth mass fraction is determined as one minus the sum of the (N-1) solved mass fractions. To minimize numerical error, the Nth species should be selected as that species with the overall largest mass fraction, such as N2 when the oxidizer is air. Mass Diffusion in Laminar Flows In Equation 3-5, iJ is the diffusion flux of species i, which arises due to concentration gradients. By default, FLUENT uses the dilu te approximation, under which the diffusion flux can be written as iimiJDY (3-6) Here Di, m is the diffusion coefficient for species i in the mixture. For certain laminar flows, the dilute approximation may not be acceptable, and full multicomponent diffusion is required. In such cases, the Maxwell-Stefan equations can be solved. Treatment of Species Transp ort in the Energy Equation For many multicomponent mixing flows, the tran sport of enthalpy due to species diffusion

PAGE 57

57 1.n ii ihJ (3-7) can have a significant effect on the enthalpy fiel d and should not be negl ected. In particular, when the Lewis number i pimk Le cD (3-8) for any species is far from unity, neglecting this term can lead to significant errors. FLUENT will include this term by default and where k is the thermal conductivity, for any species is far from unity, neglecting this term can lead to significant errors. FLUENT will include this term by default. The Generalized Finite-Rate Form ulation for Reaction Modeling The reaction rates that are computed in FLUENT by one of three models: Laminar finite-rate model: The effect of turbulent fluctuations are ignored, and reaction rates are determined by Arrhenius expressions. Eddy-dissipation model: Reaction rates are assume d to be controlled by the turbulence, so expensive Arrhenius chemical kinetic calculations can be avoided. The model is computationally cheap, but, for realistic re sults, only one or two step heat-release mechanisms should be used. Eddy-dissipation-concept (EDC) model: Detail ed Arrhenius chemical kinetics can be incorporated in turbulent flames. Note that detailed chemical kinetic calculations are computationally expensive. The generalized finite-rate formulation is su itable for a wide range of applications including laminar or turbulen t reaction systems, and combustion systems with premixed, nonpremixed, or partially-premixed flames. We have adopted the finite rate formulation for setting up our combustion model.

PAGE 58

58 The Laminar Finite-Rate Model The laminar finite-rate model computes the chemical source terms using Arrhenius expressions, and ignores the eff ects of turbulent fluctuations. The model is exact for laminar flames, but is generally inaccurate for turbul ent flames due to highly non-linear Arrhenius chemical kinetics. The laminar model may, however, be acceptable for combustion with relatively slow chemistry and small turbulent fl uctuations, such as supersonic flames. The net source of chemical species i due to reaction iR is computed as the sum of the Arrhenius reaction sources over the NR reactions th at the species participate in: ,, 1RN iwiir r R MR (3-9) where i wM, is the molecular weight of species i and r iR, is the Arrhenius molar rate of creation/destruction of species i in reaction r. Reaction may occur in the continuous phase between continuous-phase species only, or at wa ll surfaces resulting in the surface deposition or evolution of a continuous-phase species. If we consider the rth reaction wr itten in general form as follows: ,,, 11fr brNN K iriiri K ii (3-10) N = number of chemical species in the system r i = stoichiometric coefficient for reactant i in reaction r r i = stoichiometric coefficient for product i in reaction r i = symbol denoting species i r fK, = forward rate constant for reaction r r bK, = backward rate constant for reaction r

PAGE 59

59 Equation 3-10 is valid for both reversible a nd non-reversible reac tions (Reactions in FLUENT are non-reversible by default.) For non-reversible reactions, the backward rate constant, brK is simply omitted. The summations in Equation 3-10 are for all chemical species in the system, but only species that appear as reactants or product s will have non-zero stoichiometric coefficients. Hence, species that are not involved will drop out of the equation. The molar rate of creation/destruct ion of species i in reaction r (,ir R in Equation 3-9) is given by ,,,,,,,,, 11 ()([][])rr jrjrNN iririrfrjrbrjr jjRKCKC (3-11) rN = number of chemical species in reaction r jrC = molar concentration of each reactant and pr oduct species j in reaction r (kgmol/m3) ir = forward rate exponent for each reacta nt and product specie s j in reaction r ir = backward rate exponent for each react ant and product species j in reaction r represents the net effect of third bodies on the reaction rate. This term is given by ,rN jrj jC (3-12) where jr is the third-body efficiency of the jth sp ecies in the rth reac tion. By default, FLUENT does not include thirdbody effects in the r eaction rate calculatio n. You can, however, opt to include the effect of third-body efficiencies if you ha ve data for them. The forward rate constant for reaction r, f rK is computed using the Arrhenius expression ,r rE RT frrKATe (3-13) where rA = pre-exponential factor (consistent units)

PAGE 60

60 r = temperature exponent (dimensionless) rE = activation energy for the reaction (J/kg mol) R = universal gas constant (J/kg mol-K) The user or the databa se provides values for, ir ,, ir ,, ir ,, ir ,r ,rA rE and r j during the problem definition in FLUENT. If the reaction is reversible, the backward rate constant for reaction r, r bK, is computed from the forward rate constant using the following relation: f r br rK K K (3-14) where rKis the equilibrium constant for the rth reaction, computed from ,, 100 ()exp()()N R jrjr ratm rr rp SH K RRTRT (3-15) where atmp denotes atmospheric pressure (101325 Pa). The term within the exponential function represents the change in Gibbs fr ee energy, and its components are computed as follows: 0 0 ,, 1()N i r irir iS S R R (3-16) 0 0 ,, 1()N i r irir ih H R TRT (3-17) where iS0 and ih0 are the standard-state entropy and standa rd-state enthalpy (heat of formation). These values are specified in FLUENT as properties of the mixture material. For gas-phase reactions, the reaction rate is defined on a volumetric ba sis and the rate of creation and destruction of chem ical species becomes a source term in the species conservation

PAGE 61

61 equations. The rate of deposition is governed by both chemical kinetics and the diffusion rate from the fluid to the surface. Wall surface reactions thus create sources (and sinks) of chemical species in the bulk phase and determine the rate of deposition of surface species. FLUENT treats chemical species deposited on surfaces as distinct from the same ch emical species in the gas. Similarly, reactions involving surface deposition are defined as di stinct surface reactions and hence treated differently than bulk phase reactions involving the same chemical species. Surface reactions can be limited so that they occur on only some of the wall boundaries (w hile the other wall boundaries remain free of surface reaction). The su rface reaction rate is defined and computed per unit surface area, in contrast to the fluid-phase reactions, which are based on unit volume. We consider the rth wall surface reaction written in general form as follows: i N i r i i N i r i N i i r i K i N i r i i N i r i N i i r iS s B b G g S s B b G gs b g r s b g 1 1 1 1 1 1 (3-18) where, Gi, Bi, and Si represent the gas phase species, th e bulk (or solid) species, and the surfaceadsorbed (or site) species, respectively. Ng, Nb, and Ns are the total numbers of these species., irg ,, irb and irs are the stoichiometric coefficients for each reactant species i, and, irg irb and irs are the stoichiometric coefficients fo r each product species i. Kr is the overall reaction rate constant. The summations in Equation (3-18) are for all chemical species in the system, but only species involved as reactants or produc ts will have non-zero stoichiometric coefficients. Hence, species that are no t involved will drop out of the equation. The rate of the rth reaction is ,,,,,, 111111g g bsbs rNN NNNN K iriiriiriiriiriiri iiiiiigGbBsSgGbBsS (3-18)

PAGE 62

62 ,,, 1[][]g irirN g s rfriwalliwall iRkGS (3-19) where, [ ]wall represents molar concentrations on the wa ll. It is assumed that reaction rate does not depend on concentrations of the bulk (solid ) species. From this, the net molar rate of consumption or production of each species i, is given by ,,, 1 ()rxnN igasirirr r R ggR i = 1, 2, 3 Ng (3-20) ,,, 1 ()rxnN ibulkirirr r R bbR i = 1, 2, 3 Nb (3-21) ,,, 1 ()rxnN isiteirirr r R ssR i = 1, 2, 3 Ns (3-22) The forward rate constant for reaction r (kf, r) is computed using the Arrhenius equation ,r rE RT frrkATe (3-23) where Ar = pre-exponential fact or (consistent units) r = temperature exponent (dimensionless) Er = activation energy for the reaction (J/kgmol) R = universal gas constant (J/kgmol-K) The user or the database provides the values of the coefficients. (B). Boundary Conditions A 2D axisymmetric model is solved for Zone 4. This is the only zone that we are solving the reacting case simulation for. Velocity: A no slip velocity boundary condition is applied at all wall surfaces. Temperature:

PAGE 63

63 The combustor wall is a zero heat flux zone. The catalyst thickness has a boundary condition of 1-D heat conduction. On one side of the catalyst wall, we have air in flow at 700K while on the other side we have the fuel air mixture in flow at 700k. Hence, boundary condition on either side of the catalyst is that of convective heat transfer. Wall surface reaction Boundary condition: As shown in Equations (3-153-18), the goa l of surface reaction modeling is to compute concentrations of gas species and site species at the wall; i.e., [Gi] wall and [Si] wall. Assuming that, on a reacting surface, the mass flux of each gas species is balanced with its rate of production/consumption, then ,,,iwall wallidepiwallwiigasY DmYMR n i = 1, 2, 3 Ng (3-24) ,[] iwall isiteS R t i = 1, 2, 3 Ns (3-25) The mass fraction Yi, wall is related to concentration by ,[]walliwall iwall wiY G M (3-26) ,ir R is the net rate of mass deposition or etch ing as a result of surface reaction; i.e. ,, 1bN depwiibulk imMR (3-27) [Si] wall is the site species concentration at the wall, and is defined as []iwallsiteiSz (3-28) where s ite is the site density and iz is the site coverage of species i. Using Equations (3-24) and (3-25), expressions can be derived for the mass fr action of species i at the wall and for the net rate of creation of species i per unit area. These expressions are used in FLUENT to compute gas

PAGE 64

64 phase species concentrations, and if applicable, si te coverages, at reacting surfaces using a pointby-point coupled stiff solver.

PAGE 65

65 Figure 3-1. Model of the Catalytic Combustor: Zone 2 is the fu el-air mixing region, Zone 3 is the fuel-air mixture expans ion in a CROSS and Zone 4 is the combustor section holding the catalyst and the annulus. All dimensions are in mm Figure 3-2. 2D-axisymmetric schematic of the Combustor section including catalyst and mass sampling ports 1 and 2. All dimensions are in mm

PAGE 66

66 Figure 3-3. Mass Sampling Ports Figure 3-4. Mass Spectrometers for Species Concentration Measurement x/L = 0.3 Port 1 x/L = 0.7 Port 2 Flow Direction

PAGE 67

67 Table 3-1. Operating Range for Mass Flow Controllers Mass Flow Controllers Operating Range Air 0 10 SLPM Air 0 60 SLPM CH4 0 50 SCCM CO 0 5 SLPM CO2 0 1 SLPM H2 0 5 SLPM

PAGE 68

68 CHAPTER 4 RESULTS AND DISCUSSION Experimental and Computational analysis of the catalytic combusti on was applied to the cases shown in Table 4-1. As seen in the table the computation using Deutschmanns surface chemistry model for CH4/Air and (H2+CH4)/Air and an additional five step gas phase reaction mechanism for Syngas/Air covered the plati num smooth experimental cases. Since the composition of catalysts A and B is not fully char acterized a surface reaction is not available at this time. The experimental work measured the mole fraction of the products of combustion of hydrogen assisted methane/air and syngas/air on platinum catalysts. Ma ss spectroscopy was employed to measure the mixture composition of combustion at two axial locations. The computational work provided detailed information including the mole fraction, velocity and temperature profiles at the two experimental axial locations. The reacting flow was modeled using Deutschmanns surface chem istry [1, 3, 21] shown in Table 4-2 and an additional five step gas phase reaction mechanism shown in Table 4-3. A five step gas phase reaction was added to the 27 step surface re action mechanism. The Deutschmanns model assumes the presence of atomic hydrogen and oxygen which required a pathway to their formation. Combustion temperat ures were expected to be above 1000K which could lead to the dissociation of O2 and H2. This justifies the inclusion of the first two equations in the gas phase model. To keep the number of reactions to a computationally manageable size a simple reaction was introduced fo r CO oxidation in the gas phase. The non-reacting flow was computed to infer mole fraction distribution, velocity and temperature profiles at various s ections of the comput ational domain. These sections include the two mass sampling ports named as Port One and Port Two throughout our study. The mass

PAGE 69

69 sampling experiments were carried out under steady premixed fuel-air mixture flow conditions. Two fuels and four catalysts were us ed for experimentation: methane (24 04CH ), and syngas, which was primarily a mixture of methane ( 002 04 CH ), carbon monoxide (29 0 CO ), carbon dioxide (06 02CO ), and hydrogen (16 02 H ). The inlet gas composition is shown in Table 4-4 and Table 4-5. The two catalysts used were stainless steel hypodermic tubes coated with smooth Platinum and rough platinum respectively. The mixing air was kept at room te mperature for all cold flow experiments with both fuels and on all cata lysts. The cooling air was not passed during cold flow experiments. The mixing and cooling air (allowed to pass during combustion/reacting flow conditions) were both h eated to 688 K during all experiments carried on all catalysts and with both the fuels. The fuels were always maintained at room temperature. Pressure conditions were atmospheric. The area weighted average velocity of the methane-air mixture at the entrance to the ca talyst section was 3.1 m/s and th at of the syngas-air mixture was 8.82 m/s respectively. These velocities lead to Mach Numbers of 0.01 and 0.026, thereby confirming the incompressible flow regime. The composition of the gas (re acting as well as non reacting flow) in the combustor over the catalyst section was analyzed by the mass sp ectrometer in Partial Pressure of Species vs. Time mode. The species sca nned were nitrogen (m/z=14 a nd 28), oxygen (m/z=16 and 32), methane (m/z=15 and 16) for non-reacting experiments with methane; nitrogen (m/z=14 and 28), oxygen (m/z=16 and 32), methane (m/z=15 and 16), hydrogen (m/z=1 and ), carbon monoxide (m/z= 28) and carbon dioxide (m/z=44) with non -reacting syngas and nitrogen (m/z=14 and 28), oxygen (m/z=16 and 32), hydrogen (m/z=1 and 2) methane (m/z=15 and 16), carbon monoxide (m/z= 28), carbon dioxide (m/z=44) and water (m/z=17 and 18) for combustion experiments

PAGE 70

70 with both fuels. There are two sampling ports at x/L = 0.3 and x/L = 0.73 respectively, where L = 21.8 cms denotes the length of the catalyst. The mole fraction of a given species in the sample was determined from the partial pressures of all the component species recorded by the mass sp ectrometer. Further, the timeaveraged fuel mole fraction at a port was obtai ned by averaging the mass fractions obtained over the sampling time period. Experimental Results For all reacting flow experime nts, the temperatures of the mixing and cooling were maintained at a fixed 700 K. The surface temperat ure of the catalyst is measured when the air flows have reached a steady state. Methane-Air Combustion: Mass Sa mpling in the Axial Direction Methane inlet conditions are specified in the Table 4-4.Methane did not ignite when the platinum catalysts were used. Syngas-Air Combustion: Mass Sampling in the Axial Direction Syngas inlet specifications are shown in Tabl e 4-5. The combustion of syngas was carried on platinum smooth and platinum rough Figures 4-1 th rough 4-6. Repeatability experiments were carried out performing two experiments at por t two for each of the catalysts. The standard deviation between 10%-14% for the pl atinum catalysts is observed. Unlike methane, syngas ignited on all the two cat alysts. The axial distribution of the time averaged mole fraction of the species obtaine d from the combustion of syngas on platinum smooth and platinum rough catalysts ar e shown in Figures 4-1 through 4-6. Methane was in trace amount in the syngas composition used here and th erefore for all practical purposes the methane presence can be neglected. The inlet conditions are shown in Table 4-5. Platinum rough shows better combustion traits than smooth. A consider able consumption of fuel and oxygen seems to

PAGE 71

71 take place from port one to two and increased production of water and carbon dioxide on port two suggest increased co mbustion downstream. A comparison between the performances of th ese two catalyst in converting reactants and yielding products during Syngas combustion is shown in Figures 4-3 through 4-6. An equilibrium calcula tion using STANJAN was performed for syngas combustion with the same experimental conditions. The e quilibrium mixture showed a mo le fraction of 0.075% of H2 suggesting a conversion of 95% a nd a mole fraction of 11.3% for CO suggesting 65% conversion for CO. From our experiments it is found that the percentage conversion of H2 on platinum smooth and 75% on platinum rough, whereas conv ersion of CO was found to 12% on platinum smooth and 30% on platinum rough are 39% and 42% respectively. Hence platinum smooth appears to convert H2 better than any other catalyst and pl atinum rough converts CO than others. Computational Results Non Reacting Flow The non reacting cases were computed to determine the mixture temperature and composition uniformity at the catalyst entrance. The computational domain is shown in Figure 47. The local mole fractions are shown in Figure 48 at selected cross sections as follows: Section ee is the exit from the computationa l zone 2 just after the fuel a nd air have mixed, section ff is the exit from the section of the cross which is computational zone 3, port 1 is section at experimental axial mass sampling location at x/L = 0.3 and port 2 is section at experimental axial mass sampling location at x/L = 0.73. The bulk of the methane distribution above th e catalyst surface show s a local equivalence ratio of 3.17 corresponding to a mole fraction of 0.249 as against a global input equivalence ratio of 3.04, thereby suggesting that the sampling port locations show an even fuel rich mixture. Other parts of the air passage show a local equivalence ratio of 2.95 corresponding to a mole

PAGE 72

72 fraction of 0.237. Thus the computational result s show a 5% non homogeneity. However, as the flow progresses downstream, the mole fraction distribution beco mes more homogeneous as the flow reaches port one suggesting improved mixing. The computational works accomplished so far w ith syngas, is shown as contours of mole fraction of H2 and CO and velocity profiles at vari ous sections. All these cases represent computational results for non reacting case. Secti ons are shown in figure 4-7. The mole fraction distribution of hydrogen and CO in Syngas at the sections ee, ff, port one and two are shown in Figure 4.9 and Figure 4.10. Reacting Flow Four separate computational reacting cases we re studied in this research work, namely, (a) Effect of catalyst surface site de nsity on the formation of products, (b) Catalytic Combustion of Meth ane on Platinum catalyst, (c) Catalytic Partial Oxidation of Meth ane on Platinum: H ydrogen Addition, (d) Catalytic Combustion of S yngas (a gas mixture of H2, CO, CO2 and CH4) on Platinum, (e) Effect of equivalence ratio on surface c overages of Syngas constituent species and product mole fractions, and (f) Comparison between the experimental and computational results of Syngas combustion. The basis of all the numerical investigations is a study of th e surface coverage behavior of different species on platinum, and the effects that they have on the chemical kinetics of the heterogeneous combustion process. It is believed that the different surface coverage behaviors of different species on platinum are the key fact or that initiates or truncates combustion. (A). Effect of Catalyst Surface Site Density on the Formation of Products Computations were also performe d with the different values of the surface site density i.e. 1 x 10-8 Kgmol/m2, 2 x 10-8 Kgmol/m2, 2.7 x 10-8 Kgmol/m2, 3 x 10-8 Kgmol/m2 and 6 x 10-8

PAGE 73

73 Kgmol/m2. Decreasing the site density from 2.7 x 10-8 Kgmol/m2 shows a small change in the product % mole fraction, while increasing the si te density shows almost no effect on product % mole fraction, since surface reactions are close to their mass transport limit as shown in Figure 411 through 4-16. The choice of temperature is rand om, since the trends of observations are the same at all temperatures. Only those temperatures are thus chosen in figures for which there were pronounced yield of the products Temperatures of 1700K for Methane combustion, 900K for hydrogen assisted methane combustion and 1100K for syngas combustion was chosen since at these temperatures ignition is ra pid and a trend of the reactions taking place with variations in site density can thus be established. The surface characteristics of the two types of platinum catalyst used in experiments are shown below in Figures 4-17. Platinum was depos ited on stainless steel tubes. Then one of the tubes was sand blasted. This tube is referred in what follows to as Rough platinum. The rough surface shows more grains per unit area and hen ce more exposed surface area than the smooth surface. The increased exposed surface area could result in an increased number of active free sites on the platinum which in tu rn would provide for increased probability of species adsorption and hence in turn increased surface reactions The smooth surface appears to have reduced surface area. However, the range of site density values selected for the simulations (see Chapter 4 below) seems to produce no pronounced effect in combustion for all cases studied numerically a range of 1x10-8 Kgmol/m2 to 6x10-8 Kgmol/m2 for comparisons with previous studies60. Here the baseline was selected to be 2.7x10-8 Kgmol/m2. (B). Catalytic Combustion of Methane on Platinum Catalyst Experimentally methane predominantly di d not ignite on the platinum for surface temperatures below 1100K. There are two reasons: the melting point of stainless steel, which was the platinum substrate, is 1200K-1400K. Ther efore the experiments were limited to 900K.

PAGE 74

74 At a reduced surface temperature the adsorbed oxygen atoms do not leave the surface indicating higher surface coverage, and thus allow lesser vacan cies for methane dissociation to take place. Hence, the surface temperature of the catalyst at higher equivalence ratio is not desirable for a surface ignition. Moreover, with more and more gas phase methane approaching the platinum surface and with surface temperature gradually decreasing the dissociation of methane into CH3, CH2, CH, H and C is greatly inhibited. Since the formation of these radicals is absolutely necessary for methane combustion on platinum, thei r inhibited formation at fuel rich conditions did not support combustion. This behavior requi res a detailed explanation of the effects of species surface coverage on initia ting and sustaining and/or extinguishing an ignition. This description is given below: The role of surface coverage can be explai ned by assuming that before combustion is initiated, the surface is primarily covered with oxygen because the sticking probability of oxygen is higher than that of methane. This can be noticed in Figures 4.18 through 4.25 where the platinum surface is totally covered with oxygen as long as the surface temperatures does not exceed 1100K. In these cases the surface coverage value is close to 1 for oxygen on platinum and hardly any platinum vacancies are left. He nce no O atom is available to oxidize CH4 and there is no CO2 and H2O formed along the entire catal yst length. With increasing surface temperatures conditions are reached where the adsorption/de sorption equilibrium of oxygen shifts to desorption as shown in Figures 4.23 through 4.25 where the reducing surfa ce coverage value of oxygen suggest desorption. As a result ther e are available bare surface sites where CH4 can be adsorbed, an absolutely necessary step in ini tiating combustion of meth ane on platinum followed by H abstraction, which leads to adsorbed C(s) and H(s) atoms reacti ng with the surrounding O atoms to form CO(s) and OH(s). Significant amount s of CO(s) and OH(s) seem to be formed

PAGE 75

75 only at and above 1100K, and hence their oxidation to form CO2 and H2O becomes predominant above 1100K followed by their desorption leaving more free surface sites for CH4 adsorption. The pronounced behavior can be seen in figures 4-21 through 4-25 along with the corresponding product formation is shown in figures. The su rface temperature distri bution over the catalyst surface shown in Figure 4.26 suggests that not until the surface temperat ure reaches 1200K that the adsorption/desorption equilibrium of oxygen begin to shift towards desorption and it is then when the onset of O atoms leave the platinum su rface, thus creating a possibility of vacant sites on the platinum surface upon which the adsorption of CH4 becomes possible. However, the literature56 indicates that the homogeneous i.e. ga s phase reaction scheme is based on the reduction of a large set of hydrocarbon oxidation reactions to C1 sp ecies. Also, with the assumption of the absence of gas phase reactions, no CO was formed in the simulation. In order to capture the production of CO the gas-phase reaction mechanis ms must be included above 1200K. Research works based on the catalytic combustion of methane on platinum was conducted by Schwiedernoch50 show that no significant in fluence on the conversion was observed. In the present computation the gas phase reactions set in not until the surface gets heated to 1200K. Therefore this numerical investigation confirms and explains why there was no ignition of methane on platinum in the experi ments described here. The surface temperature distribution shown in Figure 4-26 indicates that there is hardly any adsorption taking place below 900K. Adsorption is an exothermic process but a constant temperature below 900K suggests no adsorption. First significant si gns of adsorption can be obser ved at 1100K. Pronounced gas phase reactions seem to take over at temperatur es above 1500K and above. The mole fraction distribution of methane combustion at 1500K and 1700K are shown in Figure 4.27 and 4.28

PAGE 76

76 respectively indicating more c onsumption of CH4 and O2 as the mixture flows further downstream. There is a simultaneous increase in product formation. At 1500K over 66% conversion of CH4 is observed and the mole fr action of H2O in the combustion mixtures is noticed to be 4.3%. At 1700K over 67% conversi on of CH4 is observed and the mole fraction of H2O in the combustion mixtures is noticed to be 4.4%. The reactions taking place at 1500K and 1700K is largely contributed by the gas phase re actions than surface react ions. For all methane cases described here formation of molecular hydrogen was not observed. (C). Catalytic Partial Oxidation of Met hane on Platinum: Hydrogen Assisted The addition of hydrogen can have a major influence on the reaction kinetics further complicating the methane dissociation pathway. H2 is very likely to contribute more than only the heat for the ignition of CH4. Since H2 removes O(s) from the surface by formation of H2O it is likely that H2 has a great impact on the reaction kinetics of CH4 combustion by hydrogen. Different conversions and yields of CO2 indeed indicate an important role of H2 on the kinetics causing considerable differences between hydrogenassisted and thermally initiated reaction as seen in this work. Here, not only CH4 and O2 compete for free vacancies but also H2, which occupies the surface first as s hown in Figure 4.29 and 4.30. H2 combustion on platinum heats up the catalyst surface and provides the ade quate surface temperature to initiate CH4 combustion. In case of H2 addition water will be formed immediately by oxidation of H2. Therefore, a certain amount of the surface area is always occupied by H2O(s). This can be seen clearly by comparing these results with the surface coverage calculations in Figure 4.32 at 900 K, where with H2 more surface sites are occupied by OH and H2O than in absence of H2. The adsorption of CH4, dissociation as well as th e reaction to CO and CO2 respectively have to happen on the remaining vacancies downstream of the catalyst. This expl ains why more products are formed downstream

PAGE 77

77 i.e. more at port two than at port one. However, with the addition of hydrogen first signs of formation of CO, CO2 and H2O are already to be seen at 700K as shown in Figures 4.31. The added H2 occupying the surface first under dissociat ion, reacts with the adsorbed O(s). The produced water leaves the surface continuously generating vacancies, which methane can occupy as well. At 400 K already traces of CO ar e formed and the surface is mostly covered by carbon. Compared to the catal ytic combustion without H2 the number of free adsorption sites on Pt seems to have increased by a factor of about 105. In order to simulate H2-assisted catalytic combustion, the carbon coverage, C( s), plays an important role too. Here, also we see increase in the surface sites increase with th e increase in temperature. The temperature distribution on the surface of the catalyst is shown in Figure 4.33. The mole fractions at inlet, port one, port tw o and the combustor outlet is shown in figure 4-34 through 4-37. The surface temperature distribution shows no adsorption up to 600K and first signs of adsorption at 700K and above, sugg ested by an increase in temperature. The mole fraction distribution of hydrogen assisted me thane combustion at 400K, 600K, 700K and 900K are shown in Figure 4.34 through 4.37 respectively. It is observed that there is no consumption of fuel and hence no noticeable product formati on at 400K, 600K and 700K. At 900K it is observed that over 35% of CH4 and over 64% of H2 is consumed producing 10.6% and 1,3 % H2O and CO2 respectively. The trend established from the results is as expected, i.e. there is more consumption of reactants and more product fo rmation as the flow goes further downstream. (D). Catalytic Combustion of Syngas on Platinum Catalytic Partial Oxidation i.e. hydrogen assi sted as well as high temperature catalytic oxidation of methane shows a pathway of combusti on that forms syngas. This is an expected trend in the numerical analysis; hence, the s yngas catalytic combustion on platinum using the same Deutschmanns model. Since the pressure conditions are atmospheri c and the adsorption of

PAGE 78

78 methane on platinum is the rate determining st ep, this assumption appears reasonable. Syngas was a gaseous mixture of H2, CO, CO2 and CH4, thus offering a competition between species to occupy vacant sites on the platinum surface. It was noticed that for the case of catalytic oxidation of methane the sticking coefficient of oxygen to be more than methane, hence the whole platinum surface was initially covered with O atom s. In the case of partial oxidation of methane with hydrogen addition the H atoms initially cove red the whole of the Platinum surface. But in the case of syngas as the fuel, the H, O, CO, CO2 and CH4 molecules all compete to occupy the vacant sites on platinum surface. In terestingly, it can be noticed that at 400K see Figure 4.38 the entire platinum surface is covered with CO at oms thereby suggesting a greater sticking coefficient of CO than any other pa rticipating species in the mixtur e. A dominant coverage of the C(s) is noticed, but no noticeable formation of OH(s), H2O or CO2 seems to take place. In fact CO2 hardly adheres to the plat inum surface, suggesting that CO2 predominantly remains in the gas phase. At 600K see Figure 4.39, a similar trend is seen with no noticeable formation of OH(s), H2O or CO2; however, the formation of OH(s) and H2O(s) has increased 104 times. At 700K see Figure 4.40, however the first traces of OH(s) and H2O(s) formation are seen, the CO atoms still occupying the whole of pl atinum surface. This observation confirms the fact that the oxidation of CO to CO2 is relatively more difficult to achieve than the oxidation of OH(s) to H2O(s). Hence, with a syngas fuel one e xpects to see a greater yi eld of water due to the oxidation of H2 than the yield of CO2 due to the oxidation of CO. This beha vior is also observed from figure where traces of H2O have already form ed with no signs of in crease in the CO2 mole fractions. At 900K see Figure 4.41, the platinum surface c overage increases, suggesting that most of the other adsorbed species are loosely bound and have started leaving the surface. This is

PAGE 79

79 observed from the figure where surf ace coverage of CO has decrease d and that some of it has left the platinum surface. Thus H(s), OH(s) and CO(s) all competes with the available O atoms in the surrounding. However, with significant amount of H2O noticeable on the surface, it means that the formation of H2O from O and OH is pref erred over the oxidation of CO. More and more O(s) and OH(s) seem to cover the surface downstream from the mid point of the catalyst length, simultaneous decrease in the surface coverage of H(s) and H2O(s) takes place, suggesting more formation of H2O downstream. The bottle neck behavior can be explained as the transition between the heterogeneous and homogeneous reactions. At 1100K see Figure 4.42, the surface of free pl atinum has increased and that of CO has decreased even further than at 900K, a trend is also noticed at 1100K, however, more and more OH(s) combines with O(s) and hence more H2O is formed with more CO leaving the surface and an augmented amount of CO2 is formed at 1100K At this temperature, pronounced gas phase reaction is expected judging by the increased free sites of platinum and the bottle neck distribution of surface coverage. The surface te mperature distribution shown in Figure 4.43 indicates no adsorption up to 600K and the first signs of adsorption appear at 700K due to an increase in temperature. Pronoun ced heterogeneous and transition into gas phase reactions are observed in the profiles at 900K and 1100K. The mole fraction distribution of syngas combustion at 400K, 600K, 700K, 900K and 1100K are shown in Figure 4.44 th rough 4.48 respectively. It is observed that there is no consumption of fuel and hence no noticeable product formation at 400K, 600K and 700K. At 900K it is observed that over 65% of H2 is conve rted and over 11% of H2O is formed. There is no significant consumption of CH4, CO and CO2 at 900K. At 1100K 67% of H2 is converted and 12% of H2O is formed. The mole fraction of CO2 in the combustion mixture was found to

PAGE 80

80 be 7.4% suggesting a 2% increase in CO2 from inlet conditions. The tre nd established from the results is as expected, i.e. there is more cons umption of reactants and more product formation as the flow goes further downstream. (E). Effect of Equivalence Ratio on Sur face Coverages in Syngas Combustion The effect of temperature on the surface coverage is described below. The analysis was done for syngas mixture us ing Deutschmanns model21. Combustion occurred at 900K and above but no noticeable combustion was observed below 900K. An investigation of equivalence ratio effects on the combustion and surface coverage distribution of speci es at 900K followed including the surface behavior of species and th eir role in the reacti on kinetics. Equivalence ratios were varied from fuel lean cases to fuel rich values including = 0.47, 0.72, 1.0 and 1.2. One experimental case of equivale nce ratio of 2.28 was compared to the numerical results. It is observed that the surface behavior is vastly different from th e observations of our previous section where a fuel rich mixture was studied. Th e effects of equivalence ratios are shown in Figures 4-49, Figures 4-51, Figures 4-53 and Figure 4-55 and the corresponding mole fraction distribution is shown in Figur e 4-50, Figures 4-52, Figures 4-54 and Figures 4-56. At =0.47, shown in Figure 4-49, the fuel lean mixture results in altogether different surface coverages than the fuel rich conditions of =2.28, for example as noticed in Fi gure 4-41. It is observed that at =0.47, the entire platinum surface is covered mostly with O atoms and that CO is loosely bound to the surface. Traces of OH and H2O have already been formed. The loosely bound CO atoms suggest their availability to co mbine with O atoms and thus form CO2. Hence at this fuel lean condition there is a higher probability of CO meeting O a nd reacting. Thus both OH and CO now compete to abstract one O atom and oxidize to H2O and CO2 while at =2.28 it was observed that whole of platinum surface was covere d with CO, such that oxidation of CO to CO2 seemed difficult to achieve. This is also noticed from the mole fraction distribution of CO in

PAGE 81

81 Figure 4-47 where practically no conversion of CO to CO2 is observed at ports 1 and 2 as against Figure 4-50 where a conversion of over 27% CO at port 1 and over 55% at port 2 is observed. Thus at = 0.477 formation of both CO2 and H2O is observed whereas at = 2.28 oxidation of CO is suppressed and formation of H2O is dominant. From Figure 4-50 it is observed that there is a conversion of over 39% H2 at port 1, 70% at port 2, 24% O2 at port 1, 47% at port 2. 3.4 % and 6.2% H2O. An increase of 47% and 96% in the mole fraction of CO2 is observed at port 1 and port 2 respectively. The trend established from the results is expected, i.e. there is more consumption of reactants and more product fo rmation as the flow goes further downstream. The surface coverage and the mole fraction distribution at =0.77 are shown in Figure 451 and Figure 4-52 respectively. Th e behavior of these parameters is the same as that at =0.47 except that we have more productio n of CO2 at ports 1 and 2 is observed. So at a fuel lean condition, the oxidation of CO is favored because CO is loosely bound to the platinum and can thus abstract O atoms and oxidize. The percen tage conversion and product formation are the same at =0.47 and =0.73. However, a change in trend appears when mixture composition becomes stoichiometric and further richer. The surface coverage beha vior and the mole fraction distribution at stoichiometric and fuel richer condition are shown in Figures 4-53 and 4-54. At =1, the surface coverage of CO on platinum increases 1000 times and approaches unity s uggesting that now the whole of platinum is predominantly covered with CO. This therefore translates in a very high sticking coefficient of CO and suggests that CO is no longer available to abstract the O atoms. Hence an inhibition in the oxidation of CO is observed which mean s that the conversion of CO is largely inhibited. Less than 10% and 26% convers ion of CO is seen to occur at ports 1 and 2

PAGE 82

82 respectively. Most of the O atoms ar e now abstracted by OH to form H2O. This is suggested by over 6% and 10% production of H2O at ports 1 and 2 respectively. At =1.21, the surface coverage of CO is obs erved to increase further and hence the conversion of CO is further inhibited. This is shown in Figure 4-55 and 4-56 where practically no CO conversion occurs at port 1 and port 2. Al l the O atoms are now available for to oxidize OH to form H2O. (F). Comparison between the Experimental and Computational Results of Syngas Combustion The inlet specifications for both the experimental as well as the computational cases were identical. The mixture compositions are show n in Table 4-4 and 4-5. The study involved determining the mole fraction and the temperatur e distribution during both the experimental and computational studies and then comparing the two. The computational study was carried out in two stages. The first stage included the com bustion simulation with only the heterogeneous surface reaction mechanism and the second stag e included both the heterogeneous and the additional five step gas phase homogeneous reac tions. This was done to magnify the difference from just the heterogeneous reactions and to br ing about an improvement in the prediction of mole fraction and temperature distribution. The temperature distribution, shown in Figure 4.57, suggests good agreement between the experiment and the simulation that involved both surface and gas phase reactions. The temperatures at inle t, port one and port two ag ree within 5%, but at the outlet is affected by the com putational artifact. There is a 12% error in the experimental and computational temperature at the outlet. During simulation, FLUENT requires a user specified temperature boundary conditions at the outlet. Th e outlet face was chosen to be a pressureoutlet face, wherein FLUENT requires for a backfl ow temperature. This temperature has to be somewhat of a similar value as the one which the computational domain i.e. combustor section

PAGE 83

83 with catalyst in this case is expe cted to reach during combustion. If this is not done then there arise serious convergence issues due to a large gradient betwee n the temperature reached during simulation and a default backflow temperature chosen by FLUENT. In this study, a backflow temperature of 700K was chosen so that a slight decrease seems to appear in the temperature valu e from port two to the outlet. Cooling air is used in the experiments to keep the catalyst from over heating. The computation models this flow and the conjugate heat transfer. Figure 4.58 through 4.60 shows a radial temperature profile of the cooling air for several cases at port 1 (x/L=0.3), port 2 (x/L=0.7) and outlet (x/L=1) respectively. The initial coolin g air temperature is assumed the same as the initial reactant mixture temperature. These te mperatures are shown on the y = 0 axis of the Figure 4.58 through 4.60. In the experiment th e inlet temperature was limited to 700K. The computations were extended to higher values. The heat transfer to the cooling air is evid enced by the temperature rise noticed at the catalyst proximity. At higher temperatures the metal surface reaches high temperature values indicating that cooling c ould be insufficient. Clearly in prac tice cooling air would be maintained at lower values. The amount of heat transferred to the cooling air in th e case of 900K initial temperature was 75% for syngas and 90% for hydroge n assisted methane mixtures of the total heat released assuming equilibriu m is reached at the catalyst ex it. The difference in temperature along the catalyst from inlet to outlet was 331K for methane + hydrogen mixtures and 314K for syngas mixtures. This indicates a heat fl ux in an upstream direction of 23.4x103 W/m2 and 22.2x103 W/m2 respectively. Thermal conductivity of steel was taken to be 16.27 W/m-K. The experiment and the simulation using hete rogeneous reactions of syngas do not agree well, the later significantly lower than the values obtained in experiments. This validates the use

PAGE 84

84 of additional gas phase reactions, which proceed with much heat release, thereby increasing the temperature. The % mass fraction distribution co mparison between simulation and experiment at ports one and two is shown in Figure 4.61 throug h 4.63. There is a negligible yet noticeable decrease in the mass fractions of the incoming speci es during simulation. This is attributed to the fact that certain gas phase reac tion is taking place before the fuel -air mixture actua lly reaches the catalyst. The Figure shows that the combinati on of heterogeneous and homogeneous reactions during simulations predicts the oxidation of H2 quite well and thus provides an agreeable formation of water with the experiments. It does not however predict the oxidation of CO well. This can be attributed to the high sticking pr obability of CO amongst a ll the species in the syngas composition. Since CO does not leave the surface, its presence and oxidation in the gas phase to form CO2 is greatly inhibited. But in experime nts there is consider able production of CO2 and a corresponding consump tion of CO. Hence another m echanism that predicts the oxidation of CO to CO2 first on the catalyst surface followed by the desorption of CO2 into the gas phase and subsequently additional gas phase r eactions is needed to predict the formation of CO2. Another reason for a good agreem ent in the prediction of H2O can be attributed to the fact that the surface reaction mechan ism shows the H(s) atoms are relatively mobile on the surface than CO (suggested by lower surface coverage of H than CO), hence abstraction of H and O atoms on the surface is easier th an the abstraction of CO and O. Hence the mechanism supports the formation of H2O. The simulation however predicts good agreement at higher temperatures, both with and without addition of gas phase reactions. This can be attributed to the fact that at higher temperatures, the adsorbed species includ ing CO, tend to leave the surface and that the anticipation of the adsorbed sp ecies with an available O atom becomes more probable. Hence more CO2 seem to be formed at higher temperatures because of CO less st rongly adhered to the

PAGE 85

85 surface. H2O however is seen to be formed more than CO at any temperature. The mass fractions of H2O and CO2 at port two found from experiment is 7.6% and 22.2% respectively, whereas those found from simulation is 8.2% and 12.05%.W e thus find that the simulation predicts H2O within less than 8% error but appe ars 45% offset in predicting CO2. The robustness of the simulation scheme is best achieved at higher temperatures. Although the simulation does not validate experimental findings quantitatively, it does a good job to establ ish the correct trend qualitatively, namely, a decrease in the reacta nt composition with a si multaneous rise in the product composition while going downstream along the catalyst, thus providing valuable insight in the catalytic effect of the surface on the gas mixture combustion. A detailed study of the surface reaction behavior of CO on platinum as well as a reaction mechanism involving C1, C2, C3 etc species, which would captu re the gas phase reactions of CO oxidation, is thus desired.

PAGE 86

86 0 5 10 15 20 25 30 35 0.00.10.20.30.40.50.60.70.80.91.0x/LX (%) O2 H2 CH4 CO2 CO H2O Figure 4-1. Mole Fraction Distri bution of Syngas Combustion at Port One and Port Two on Platinum Smooth 0 5 10 15 20 25 30 35 0.00.10.20.30.40.50.60.70.80.91.0x/LX (%) O2 H2 CH4 CO2 CO H2O Figure 4-2. Mole Fraction Distri bution of Syngas Combustion at Port One and Port Two on Platinum Rough

PAGE 87

87 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 CO O2 H2 Species%-Conversion Platinum Rough Platinum Smooth Figure 4-3. Comparison between Catalyst Platinum Smooth and Rough at Port 1 in Converting Reactants 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 CO O2 H2 Species%-Conversion Platinum Rough Platinum Smooth Figure 4-4. Comparison between Catalyst Platinum Smooth and Rough at Port 2 in converting reactants

PAGE 88

88 0.0 5.0 10.0 15.0 20.0 25.0 CO2 H2O Species% Mass Fraction Platinum Rough Platinum Smooth Figure 4-5. Comparison between Ca talyst Platinum Smooth, Rough, A & B in Product Yield at Port One 0.0 5.0 10.0 15.0 20.0 25.0 CO2 H2O Species% Mass Fraction Platinum Rough Platinum Smooth Figure 4-6. Comparison between Ca talyst Platinum Smooth, Rough, A & B in Product Yield at Port Two

PAGE 89

89 Figure 4-7. Schematic Diagram of the Computational Domain. Figure 4-8. Non Reacting Mole Frac tion Distribution of Methane at Section ee, ff, Ports One and Two ee ff Port 1 Port 2

PAGE 90

90 Figure 4-9. Non Reacting Mole Fr action Distribution of Hydrogen in Syngas at Section ee, ff, Ports One and Two Figure 4-10. Non Reacting Mole Frac tion Distribution of CO in S yngas at Section ee, ff, Ports One and Two ee ff Port 1 Port2 ee ff Port 1 Port 2

PAGE 91

91 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0E+001.0E-082.0E-083.0E-084.0E-085.0E-086.0E-087.0E-08 Site Density (Kgmol/m2)% Mole Fraction H2O CO2 Figure 4-11. Mole Fraction Dist ribution with respect to Site Density for Methane Combustion at 1700K at Port 2 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.0E+001.0E-082.0E-083.0E-084.0E-085.0E-086.0E-087.0E-08 Site Density (Kgmol/m2)% Mole Fraction H2O CO2 Figure 4-12. Mole Fraction Distribu tion with respect to Site Density for Methane Combustion at 1700K at Outlet

PAGE 92

92 0.00 2.00 4.00 6.00 8.00 10.00 12.00 0.0E+001.0E-082.0E-083.0E-084.0E-085.0E-086.0E-087.0E-08 Site Density (Kgmol/m2)%Mole Fraction H2O CO2 Figure 4-13. Mole Fraction Distri bution with respect to Site Density for Hydrogen Assisted Methane Combustion at 900K at Port 2 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 0.0E+001.0E-082.0E-083.0E-084.0E-085.0E-086.0E-087.0E-08 Site Density (Kgmol/m2)%Mole Fraction H2O CO2 Figure 4-14. Mole Fraction Distri bution with respect to Site Density for Hydrogen Assisted Methane Combustion at 900K at Outlet

PAGE 93

93 0 5 10 15 20 25 30 35 0.0E+001.0E-082.0E-083.0E-084.0E-085.0E-086.0E-087.0E-08 Site Density (Kgmol/m2)% Mole Fraction H2O CO CO2 Figure 4-15. Mole Fraction Distribu tion with respect to Site Density for Syngas Combustion at 1100K at Port 2 0 5 10 15 20 25 30 35 0.0E+001.0E-082.0E-083.0E-084.0E-085.0E-086.0E-087.0E-08 Site Density (Kgmol/m2)% Mole Fraction H2O CO CO2 Figure 4-16. Mole Fraction Distribu tion with respect to Site Density for Syngas Combustion at 1100K at Outlet

PAGE 94

94 (A) M-10000 times: S5 m (B) M-15000 times: S2 m (C) M-10000 times: S5 m (D) M-15000 times: S2 m Figure 4-17. SEM images of Rough (A) and (B ) and Smooth (C) and (D) magnified 10000 and 15000 respectively. Rough surfaces show more number of grains than the smooth surface per unit area. The exposed surface ar e is thus greater for rough surface and thus provides more active sites on pla tinum surface to allow adsorption. M: Magnification: S: Scale

PAGE 95

95 Figure 4-18. Surface Coverage Di stributions along Catalyst fo r Methane Combustion at 400K Figure 4-19. Surface Coverage Di stributions along Catalyst fo r Methane Combustion at 600K Surface Coverage, x/L x/L Surface Coverage,

PAGE 96

96 Figure 4-20. Surface Coverage Di stributions along Catalyst fo r Methane Combustion at 700K Figure 4-21. Surface Coverage Di stributions along Catalyst fo r Methane Combustion at 900K x/L Surface Coverage, Surface Coverage, x/L

PAGE 97

97 Figure 4-22. Surface Coverage Di stributions along Catalyst fo r Methane Combustion at 1100K Figure 4-23. Surface Coverage Di stributions along Catalyst fo r Methane Combustion at 1200K Surface Coverage, Surface Coverage, x/L x/L

PAGE 98

98 Figure 4-24. Surface Coverage Di stributions along Catalyst fo r Methane Combustion at 1500K Figure 4-25. Surface Coverage Di stributions along Catalyst fo r Methane Combustion at 1700K Surface Coverage, Surface Coverage, x/L x/L

PAGE 99

99 Figure 4-26. Surface Temperature Distribution along Catalyst for Methane Combustion at different Inlet Temperatures 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction O2 CH4 H2O CO2 Figure 4-27. Mole Fraction Distri bution of Methane Combustion at Inlet, Port One, Two and Outlet at 1500K Surface Temperature (K) x/L

PAGE 100

100 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction O2 CH4 H2O CO2 Figure 4-28. Mole Fraction Distri bution of Methane Combustion at Inlet, Port One, Two and Outlet at 1700K Figure 4-29. Surface Coverage Di stribution along Catalyst for Hydrogen Assisted Methane Combustion at 400K Surface Coverage, x/L

PAGE 101

101 Figure 4-30. Surface Coverage Di stribution along Catalyst for Hydrogen Assisted Methane Combustion at 600K Figure 4-31. Surface Coverage Di stribution along Catalyst for Hydrogen Assisted Methane Combustion at 700K Surface Coverage, x/L Surface Coverage, x/L

PAGE 102

102 Figure 4-32. Surface Coverage Di stribution along Catalyst for Hydrogen Assisted Methane Combustion at 900K Figure 4-33. Surface Temperature Distribution al ong Catalyst for Hydrogen Assisted Methane Combustion at different inlet Temperatures Surface Coverage, x/L x/L Surface Temperature (K)

PAGE 103

103 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction H2 O2 CH4 H2O CO CO2 Figure 4-34. Mole Fraction Distri bution of Hydrogen Assisted Methane Combustion at Inlet, Port One, Two and Outlet at 400K 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction H2 O2 CH4 H2O CO CO2 Figure 4-35. Mole Fraction Distri bution of Hydrogen Assisted Methane Combustion at Inlet, Port One, Two and Outlet at 600K

PAGE 104

104 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction H2 O2 CH4 H2O CO CO2 Figure 4-36. Mole Fraction Distri bution of Hydrogen Assisted Methane Combustion at Inlet, Port One, Two and Outlet at 700K 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction H2 O2 CH4 H2O CO CO2 Figure 4-37. Mole Fraction Distri bution of Hydrogen Assisted Methane Combustion at Inlet, Port One, Two and Outlet at 900K

PAGE 105

105 Figure 4-38. Surface Coverage Di stributions along Catalyst for Syngas Combustion at 400K Figure 4-39. Surface Coverage Di stributions along Catalyst for Syngas Combustion at 600K Surface Coverage, Surface Coverage, x/L x/L

PAGE 106

106 Figure 4-40. Surface Coverage Di stributions along Catalyst for Syngas Combustion at 700K Figure 4-41. Surface Coverage Di stributions along Catalyst for Syngas Combustion at 900K Surface Coverage, Surface Coverage, x/L x/L

PAGE 107

107 Figure 4-42. Surface Coverage Di stributions along Catalyst for Syngas Combustion at 1100K Surface Coverage, x/L

PAGE 108

108 Figure 4-43. Surface Temperature Distributi on along Catalyst for Syngas Combustion at different inlet Temperatures 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction H2 O2 CH4 H2O CO CO2 Figure 4-44. Mole Fraction Distri bution of Syngas Combustion at Inlet, Port One, Two and Outlet at 400K Surface Temperature (K) x/L

PAGE 109

109 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction H2 O2 CH4 H2O CO CO2 Figure 4-45. Mole Fraction Distri bution of Syngas Combustion at Inlet, Port One, Two and Outlet at 600K 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction H2 O2 CH4 H2O CO CO2 Figure 4-46. Mole Fraction Distri bution of Syngas Combustion at Inlet, Port One, Two and Outlet at 700K

PAGE 110

110 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction H2 O2 CH4 H2O CO CO2 Figure 4-47. Mole Fraction Distri bution of Syngas Combustion at Inlet, Port One, Two and Outlet at 900K 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 0.000.200.400.600.801.001.20 x/L% Mole Fraction H2 O2 CH4 H2O CO CO2 Figure 4-48. Mole Fraction Distri bution of Syngas Combustion at Inlet, Port One, Two and Outlet at 1100K

PAGE 111

111 Figure 4-49. Surface Coverage Di stributions along Catalyst fo r Syngas Combustion at 900K at = 0.47 0 5 10 15 20 25 00.10.20.30.40.50.60.70.8 x/L% Mole Fraction H2 O2 H2O CO CO2 Figure 4-50. Mole Fraction Distri bution of Syngas Combustion at Inlet; Port One and Two at 900K for =0.47 Surface Coverage, x/L

PAGE 112

112 Figure 4-51. Surface Coverage Di stributions along Catalyst fo r Syngas Combustion at 900K at = 0.72 0 2 4 6 8 10 12 14 16 18 20 00.10.20.30.40.50.60.70.8 x/L% Mole Fraction H2 O2 H2O CO CO2 Figure 4-52. Mole Fraction Distri bution of Syngas Combustion at Inlet; Port One and Two at 900K for =0.72 Surface Coverage, x/L

PAGE 113

113 Figure 4-53. Surface Coverage Di stributions along Catalyst fo r Syngas Combustion at 900K at = 1.0 0 2 4 6 8 10 12 14 16 00.10.20.30.40.50.60.70.8 x/L% Mole Fraction H2 O2 H2O CO CO2 Figure 4-54. Mole Fraction Distri bution of Syngas Combustion at Inlet; Port One and Two at 900K for =1.0 Surface Coverage, x/L

PAGE 114

114 Figure 4-55. Surface Coverage Di stributions along Catalyst fo r Syngas Combustion at 900K at = 1.21 0 2 4 6 8 10 12 14 16 18 20 00.10.20.30.40.50.60.70.8 x/L% Mole Fraction H2 O2 H2O CO CO2 Figure 4-56. Mole Fraction Distri bution of Syngas Combustion at Inlet; Port One and Two at 900K for =1.21 Surface Coverage, x/L

PAGE 115

115 Figure 4-57. Temperature Comparison be tween Experimental, Heterogeneous and Heterogeneous combined with Homogeneous Mechanisms 0.0 0.5 1.0 1.5 2.0 2.5 60070080090010001100120013001400Temperature (K) Radial Position (mm) methane methane+hydrogen Syngas Figure 4-58. Temperature Comparison be tween Experimental, Heterogeneous and Heterogeneous combined with Homogeneous Mechanisms

PAGE 116

116 0.0 0.5 1.0 1.5 2.0 2.5 60070080090010001100120013001400Temperature (K) Radial Position (mm) methane methane+hydrogen Syngas Figure 4-59. Temperature Comparison be tween Experimental, Heterogeneous and Heterogeneous combined with Homogeneous Mechanisms 0.0 0.5 1.0 1.5 2.0 2.5 60070080090010001100120013001400Temperature (K)Radial Position (mm) methane methane+hydrogen Syngas Figure 4-60. Temperature Comparison be tween Experimental, Heterogeneous and Heterogeneous combined with Homogeneous Mechanisms

PAGE 117

117 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 H2O2H2OCH4COCO2 Species% Mass Fractions CFD-port1 CFD-port2 Exp-port1 Exp-port2 Figure 4-61. Comparison between Experimental and Computational % Mass Fractions for Syngas Combustion at Port One and Two 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 H2O2H2OCH4COCO2 Species% Mass Fractions CFD-port1 CFD-port2 Exp-port1 Exp-port2 Figure 4-62. Comparison between Experimental and Computational % Mass Fractions for Syngas Combustion at Port One and Two

PAGE 118

118 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 H2O2H2OCH4COCO2 Species% Mass Fractio n CFD-port1 CFD-port2 Exp-port1 Exp-port2 Figure 4-63. Comparison between Experimental and Computational % Mass Mole Fractions for Syngas Combustion at Port One and Two

PAGE 119

119 Table 4-1. Experimental and Computational Case Matrix Computation Experiment Pt-site density Temperature Mixture Gases PtSm Pt-R 1 2 2.7 3 6 400 600 700 900 Syngas, = 2.28 = 1.21 = 1.0 = 0.7 = 0.4 Methane+H2 =1.5 Methane =3.05

PAGE 120

120 Table 4-2. Deutschmann's Surface Reaction Mechanism, A is Pre Exponential Factor (cm.mol.sec), is Temperature Exponent and Ea is activation energy(KJ/mol) No. Reaction A (cm.mol.sec) Ea (KJ/mol) 1 H2+2Pt(s)=>2H(s) 4.60 x 10-2 0 0 2 2H(s)=> H2+2Pt(s) 3.70 x 1021 0 67.4 3 H+Pt(s)=>H(s) 1.00 0 0 4 O2+2Pt(s)=>2O(s) 1.8 x 1021 -0.5 0 5 O2+2Pt(s)=>2O(s) 2.30 x 10-2 0 0 6 2O(s)=> O2+2Pt(s) 3.70 x 1021 0 213.2 7 O+Pt(s)=>O(s) 1.00 0 0 8 H2O+Pt(s)=>H2O(s) 0.75 0 0 9 H2O(s)=>H2O+Pt(s) 1.00 x 1013 0 40.3 10 OH+Pt(s)=>OH(s) 1.00 0 0 11 OH(s)=> OH+Pt(s) 1.00 x 1013 0 192.8 12 O(s)+H(s)<=> OH(s)+Pt(s) 3.70 x 1021 0 11.5 13 H(s)+OH(s)<=> H2O(s)+Pt(s) 3.70 x 1021 0 17.4 14 OH(s)+ OH(s)<=> H2O(s)+O(s) 3.70 x 1021 0 48.2 15 CO+Pt(s)=> CO(s) 8.40 x 10-1 0 0 16 CO(s)=> CO+Pt(s) 1.00 x 1013 0 125.5 17 CO2(s)=> CO2+Pt(s) 1.00 x 1013 0 20.5 18 CO(s)+O(s)=> CO2(s)+Pt(s) 3.70 x 1021 0 105.0 19 CH4+2Pt(s)=>CH3(s)+H(s) 1.00 x 10-2 0 0 20 CH3(s)+Pt(s)=>CH2(s)+H(s) 3.70 x 1021 0 20.0 21 CH2(s)+Pt(s)=>CH(s)+H(s) 3.70 x 1021 0 20.0 22 CH(s)+Pt(s)=>C(s)+H(s) 3.70 x 1021 0 20.0 23 C(s)+O(s)=>CO(s)+Pt(s) 3.70 x 1021 0 62.8 24 CO(s)+Pt(s)=>C(s)+O(s) 1.00 x 1018 0 184.0

PAGE 121

121 Table 4-3. Five Step Gas Phase Reaction Mechan ism, A is Pre Exponential Factor (cm.mol.sec), is Temperature Exponent and Ea is activation ebergy(KJ/mol) No. Reaction A (cm.mol.sec) Ea (KJ/mol) 1 O2+H=>OH+O 2.00 x 1014 0.00 70.3 2 H2+O=>OH+H 5.06 x 104 2.67 26.3 3 H2+OH=>H2O+H 1.00 x 108 1.6 13.8 4 CO+OH=>CO2+H 6.00 x 106 1.5 -3.1 5 H+OH=>H2O 2.20 x 1022 -2.0 0 Table 4-4. Methane Inlet Flow Rate Specification Methane Air Composition Flow Rate in SLPM CH4 1.6 Mixing Air 5 Cooling Air 30 Table 4-5. Syngas Inlet Fl ow Rate Specifications Syngas Air Composition Flow Rte in SLPM H2 1.75 CO 3.04 CO2 0.664 CH4 0.224 Mixing Air 5 Cooling Air 30 Computational study was also performed at 1100 K, Computational study was also performed at 1100, 1200, 1500 and 1700 K, is the equivalence ratio is the equivalence ratio.

PAGE 122

122 CHAPTER 5 CONCLUSION Mass spectrometry technique was used to dete rmine the species concentration distribution inside the catalytic combustor at two selected locations. The experimental findings were validated using computational techniques that modeled heterogeneous ca talysis. Non-reacting and combustion tests were conducted and fuel rela ted parameters such as fuel mixtures were varied. Four different catalysts were used for eac h of these fuel types. Numerical investigation was conducted for each of these experimental ca ses. Additional numerical investigation was conducted by varying equivalence ratio, inlet temperature and s ite density of catalyst. The conclusions are summarized below. Fuel type: Methane Air mixture platinum smooth and platinum rough: Methane failed to ignite on both platinum smooth and rough catalysts below 1200K. This could be due to lower activation energy of surface reactions and hi gher bare sites on the surface of catalysts A and B where methane adsorption followed by reactions on the surface forming product-surface complex and desorption of the products is favored. Computational investigation carried out for methane air combustion on platinum surface consolidates this experimental observation. The model shows that almost all O atoms are strongly adsorbed on platinum surface and t hus provide less or no bare site for CH4 to adsorb below 1200K. Since adsorption of CH4 is the primary step in initiating methane combustion on platinum, a higher sticking probability of O atoms over CH4 suppresses any combustion. Since O atoms desorbs from the platinum surface at and above 1200K so that available sites for CH4 adsorption are thus created and hence first signs combustion do appear to occur at and above 1200K. At even higher temperatures more and mo re O atoms leave the surface creating more opportunity for CH4 adsorption and hence favoring more combustion. Gas phase reactions appear to be dominant at and above 1500K because there appears no adsorbed species on platinum at these temperatur es and whole of the surface is just platinum only.

PAGE 123

123 Methane HydrogenAir mixture on platinum smooth: C atoms have the highest sticking probability amongst all the constituent species. This would mean that the oxide of carbon fo rmation is a difficult proposition. The H atoms are loosely bound thereby sugges ting that formation of OH followed by H2O formation is relatively easier. No products in the gas phase appear to be formed below 700K at which the first signs of product in the form of H2O seem to be produced. Predominant product formation occurs at 900K and above when both C atoms and H atoms are loosely bound to the surface and they both co mpete to abstract O atoms to form CO and OH. However, combination of O and H atoms ap pears to dominate over CO abstracting an O atom and thus oxidation of OH to H2O is preferred over CO forming CO2. Simulations were performed by varying the site density of platinum. It was observed that changes in site density had no effect on the conve rsion of reactants as well as formation of products. This can be attributed to the fact that surface reac tions take place at the mass transport limits. Syngas Air mixture platinum smooth and platinum rough: Syngas ignited on the two catalysts. The conversi on of reactants and formation of products appears to be the highest when using platinum smooth catalyst. Platinum appears to sustain combustion over th e entire length of the catalyst. Numerical investigations performed using a five step ga s phase reaction mechanism in addition to the Deutschmanns surface reaction mechanism s uggest experimental and computational temperature within an error range of 5%-12%. The mass fraction is predicted with less than 8% error for H2O at temperatures above 700K. However there is an erro r of 45% in predicting CO2 formation. This can be attributed to the fact that for experimental conditi ons of a very rich fuel mixture, equivalence ratio of 2.28, the simulations show that out of the species comprising the syngas, CO has the highest sticking probability. CO remains adhered to platinum surface for as high as 900K. This inhibits O atoms being abstracted by CO and get oxidized to CO2. However, H, OH and O atoms are loosely bound and thus H2O is readily formed. This model predicts H2O accurately and seems to model combustion better at temperat ures at and above 900K. Numerical investigations performe d to study effects of equivalence ratios show that oxides of carbon are formed at fuel lean conditions. This is attributed to a lo wer surface coverage of CO is lower at leaner conditions and increase as the equivalence ratio increases i.e. as the mixture becomes fuel rich.

PAGE 124

124 At lean conditions CO is loosely bound on the surface and H atoms occupy the whole of platinum surface. Hence formation of oxides of carbon eventually forming CO2 is favored over formation of H2O. Reverse conditions set in as the mixture beco mes stoichiometric a nd richer. More of the surface would now be covered with CO and H atoms appear to be loosely bound. Hence more H2O is formed with less or no CO2. It is thus conclude d that production of H2O is preferred at fuel ri ch conditions and that of CO2 at fuel lean conditions. Simulations were performed by varying the site density of platinum. It was observed that changes in site density had no effect on the conve rsion of reactants as well as formation of products. This can be attributed to the fact that surface reac tions take place at the mass transport limits.

PAGE 125

125 APPENDIX A DETAILED COMPUTATIONAL SETUP Boundary layer meshing is done by first selecting the edge wherever a boundary layer is required. These edges would be those places wher e the gradient in velocity, temperature and mass fractions are significant. The edge select ed is then divided into n number of nodes followed by selecting the distance between the edge and the first row, call ed the first distance of the boundary layer. Since a boundary layer has a significant thickness, this thickness is modeled by selecting the number of rows that would constitute the overall thickness of the boundary layer. Finally we select a method of how the distance between one row to the other varies. In my selection, I have selected the distance between each layer of the boundary layer to grow exponentially by choosing a factor to be multiplied to the first distance to get the second row distance from the edge and so on. A typical boundary layer on a face is shown in Figure A-1. Once the boundary layer is created, the face ma de out of the edge, has a diminished surface area now, the decrease in area being attributed to the area occupied by the circular boundary layer. Then we mesh this face. Face mesh is generated by selecting the face wh ich has already being meshed for boundary layer or otherwise. In either case the next step was to select the number of interval counts on this face. If a boundary la yer exists on this face then this interval count would be distributed over the area left over after the boundary la yer is created, if no boundary layer exists on that face, then the interval count would be distributed ove r the entire area available. A selection of the geometry of the elements of the mesh has to be made too. The two options out of possible many options are Quad and Tri, whic h represent quadrilateral and trigonal shaped mesh elements respectively. Next we choose the pattern in which these elements would be swept

PAGE 126

126 across the entire face. Possible options are Map and Pave. The quad-pave option results in a structured grid shown in Figure A-2. Once a face is meshed, we proceed to mesh the volume. Volume mesh is generated by sweeping the face me shed previously across the entire volume that constitutes, as one of its faces, the one previous ly meshed. This was done using hex/wedge (suggesting hexahedr al/wedge shape of the mesh elements) and the pattern of sweeping was selected to be cooper option. Th e limitation to this kind of volume meshing is that, it is available only if the faces are meshed using quad-pave option, if they are not (as in some complicated geometry) then we use Tet /hybrid-Tgrid. This type of volume meshing results in an unstructured pattern of grids shown in Figure A-3. Boundary types used are mass_flow_inlets and vel ocity_inlets for th e inlet faces of the three computational domain. Th e inlet face of the first domain uses mass_flow_inlets as inlet type while the inlet faces for the rest of the three domains use velocity_inlets, the outlet faces of all the four domains ar e specified as pressure outlet and the curv ed surface wall of all the domains are specified the boundary type WA LL. Next we specify the Continuum type. Continuum type specifies whether a given volume is a fluid or a solid. The fluid option is chosen when there is a fluid flow inside a hollow volume, while solid option specifies the entire volume as a solid. Zone One is comprised of two 0.4 cm inner diamet er tubes. One supplies the fuel and the other supplies the oxidizer. The fuel feed line is 40 cms long while the oxidizer line has a length of 116 cms. The mesh generated to capture boundary la yer at the inlet face of the feed lines (both identical) is shown in Figure A-1. The first row distance is chosen to be 0.0001 m and the increment factor is 1.1. The number of rows sel ected is four. The face bearing the boundary layer is then meshed. A quad-pave option was chosen to mesh the faces. A total of 422 cells were

PAGE 127

127 generated for this face The volume was then meshed using hex/wedge-cooper option. Volume representing fuel flow was me shed into 834,128 nodes with 789,888 hexahedral elements whereas the volume representing the oxidizer flow was meshed into 1,742,720 with 1,641,158 hexahedral elements. The overall meshed volume is shown in Figure A-4. Both the feed lines were meshed exactly the sa me except for the fact that the fuel line was modeled vertically while the oxidizer line horizontally. Inlets and outlets for both the volumes were specified as mass_flow_inlet and pre ssure outlet boundary t ype respectively. The curved surface was specif ied WALL boundary type. Zone two shown in Figure A-5 is a region where the fuel an d oxidizer is getting mixed. The incoming feeds as well as the mixing portion are all 0.4 cm inner diameter tubes. The oxidizer enters the domain from the left whereas the fuel enters vert ically downwards. Since significant gradients in velocity, temperature and mass fractions are expected where the fuel and oxidizer start mixing, provisions for capturing these gradients are made by allowing boundary layer generation at the location of mixing show n in Figure A-6. This could be accomplished by subdividing zone two into three volumes (named as Volume 1, Volume 2 and Volume 3 and shown in [Figure A-7 (a), (b) and (c) shown in red]) and then generati ng boundary layers where the edges of these three volumes meet. Boundary layers are also made on the two inlet faces. Faces for all these three volumes as well as th e volumes themselves, were meshed using the same schemes adopted for zone one. Figur e A-8 shows the meshed zone two. The entrance faces and the pressu re outlet face have 1177 cell elements and 892210 hexahedral cells combining the three volumes. Oxidizer and fuel entrance faces were specified as velocity_inlet, mixture outlet face as press ure outlet and the curved surface as WALL

PAGE 128

128 boundary type respectively. The X di rection span is 5.23 cm, Y direction span is 12.5 cm and the Z direction span is 0.4 cm (inner diameter of the tube) Zone three shown in Figure A-9 is a region where the fuel-oxidizer mi xture after mixing over a length of 9.5 cm (mixing length in zone two), expands over a CROSS joint. The span length of the CROSS in the X-direction is 2.7 cm ; in the Y direction is 3.4 cm and has a 0.635 cm inner diameter. Zone three was subdivided into regions of three volumes, named as Volume 1, Volume 2 and Volume 4 shown in Figure A-10 (a), (b), (c) in red. Volume 2 models the catalyst holder. The incoming mixture after e xpanding over the CROSS hits the holder and turns right angles and continues flowing out of the third zone in to the fourth zone. Faces for all these three volumes as well as the volumes themselves were meshed using the same schemes adopted for zone one. The mixture entrance face is specif ied as velocity_inlet, mixture outlet face as pressure outlet and all curved surfaces as well as other so lid surfaces as WALL boundary type. Figure A-11 shows the meshed zone three. Volume 1 has 13,000 nodes with 10,800 hexahedral elements; Volume 2 has 73911 nodes w ith 70770 hexahedral elements and Volume 4 has 194,641 nodes with 917,544hexahedral elements, i.e. a total of 281,55 2 nodes and a total of 999,114 hexahedral elements. Zone four shown in Figure A-12 is a region where the mixture after leaving the CROSS, flows through the annular passage of the catalytic combustor and over the surface of the catalyst. Zone four was subdivided into regions of five volumes, namely, Volume 4, Volume 7, Volume 10, Volume 11 and Volume 13 shown in Figure A-13 (a ), (b), (c), (d) and (e) in red. Volume 4 models the annular passage (22.8 cm length in the X-direction and 0.9145 mm width in the Ydirection) through which the fuel-oxidizer mixtur e is allowed to pass, Volume 7 models the combustor wall, Volume 10 models the cooling air flow through the inside of the catalyst,

PAGE 129

129 Volume 11 models the catalyst, which is 22.8 cm long and Volume 13 models the entrance. The catalyst thickness is 0.01875 cm, and the combus tor thickness is 0.84455 cm. Zone four has a total of 110,751 nodes with a to tal of 232,041 hexahedral elements. Figure A-14 shows the meshed zone four.

PAGE 130

130 Figure A-1 Boundary layer on Face Figure A-2 structured grid

PAGE 131

131 Figure A-3 Unstructured grid Figure A-4 Face and Volume Mesh

PAGE 132

132 Figure A-5 Computationa l Model of Zone Two Figure A-6. Boundary Layer capturing mixing

PAGE 133

133 (A) (B) (C) Figure A-7 Three separate Volumes that makes Zone Two (A) Volume 1 (B) Volume 2 and (C) Volume 3

PAGE 134

134 Figure A-8 Meshed Zone two Figure A-9. Computationa l Model of Zone three

PAGE 135

135 (A) (B) (C) Figure A-10 Three Volumes th at make Zone Three (A) Volume 1 (B) Volume 2 and (C) Volume 4

PAGE 136

136 Figure A-11 Meshed Zone three Figure A-12 Computati onal Model of Zone three

PAGE 137

137 (A) (B) Figure A-13 Five separate Volumes that make Zone four (A) Volume 4 (B) Volume 7 (C) Volume 10 (E) Volume 11 and (E) Volume 13

PAGE 138

138 (C) (D) Figure A-13 Five separate Volumes that make Zone four (A) Volume 4 (B) Volume 7 (C) Volume 10 (E) Volume 11 and (E) Volume 13

PAGE 139

139 (E) Figure A-13 Five separate Volumes that make Zone four (a) Volume 4 (b) Volume 7 (c) Volume 10 (d) Volume 11 and (e) Volume 13 FigureA 14MeshedZonefour

PAGE 140

140 APPENDIX B GAMBIT PROGRAM FOR DESI GN AND MESH GENERATION / Journal File for GAMBIT 2.2.30, Database 2.2.14, ntx86 BH04110220 / Identifier "default_id636" / File opened for write Fri Jun 02 11:10:32 2006. / Journal File for GAMBIT 2.2.30, Database 2.2.14, ntx86 BH04110220 / Identifier "default_id3260" / File opened for write Tue May 16 14:44:35 2006. volume create height 255.6 radius1 3.302 ra dius3 3.302 offset 127.8 0 0 xaxis frustum volume create height 255.6 radius1 2.3785 radius3 2.3785 offset 127.8 0 0 \ xaxis frustum volume split "volume.1" volumes "volume.2" connected bientity volume create height 255.6 radius1 5.1435 radius3 5.1435 offset 127.8 0 0 \ xaxis frustum volume create height 255.6 radius1 2.1 ra dius3 2.1 offset 127.8 0 0 xaxis frustum /ERROR occurred in the next command! /ERROR occurred in the next command! volume move "volume.2" "volume. 3" "volume.4" offset -27 0 0 volume create height 255.6 radius1 3.302 ra dius3 3.302 offset 127.8 0 0 xaxis frustum volume create height 255.6 radius1 2.3875 radius3 2.3875 offset 127.8 0 0 \ xaxis frustum volume split "volume.1" volumes "volume.2" connected bientity volume create height 255.6 radius1 5.1435 radius3 5.1435 offset 127.8 0 0 \ xaxis frustum

PAGE 141

141 volume move "volume.3" offset -27 0 0 volume split "volume.3" volumes "volume.1" connected bientity undo /Undone to: volume split "volume.3" volum es "volume.1" connected bientity /Undone to: volume delete "volume.1" lowertopology undo /Undone to: volume move "volume.3" offset -27 0 0 /Undone to: volume split "volume.3" volum es "volume.1" connected bientity undo /Undone to: volume create height 255.6 radius1 5.1435 radius3 5.1435 offset 127.8 /Undone to: volume delete "volume.2" lowertopology volume split "volume.3" volumes "volume.1" connected bientity /ERROR occurred in the next command! /ERROR occurred in the next command! volume move "volume.2" offset -27 0 0 volume create height 228.6 radius1 8 ra dius3 8 offset 114.3 0 0 xaxis frustum volume split "volume.3" volumes "volume.5" connected bientity default set "GRAPHICS.GENERAL.CONNEC TIVITY_BASED_COLORING" numeric 1 save name \ "C:\\Documents and Settings\\saurav\\D esktop\\3Dtestsection\\without-cooling-air16thmay.dbs" / File closed at Tue May 16 15:14: 19 2006, 2.64 cpu second(s), 3009008 maximum memory.

PAGE 142

142 / Journal File for GAMBIT 2.2.30, Database 2.2.14, ntx86 BH04110220 / Identifier "without-cooling-air-16thmay" / File opened for append Tue May 16 15:28:18 2006. /ERROR occurred in the next command! identifier name \ "C:\\Documents and Settings\\saurav\\D esktop\\3Dtestsection\\without-cooling-air16thmay.dbs" \ old saveprevious volume split "volume.2" faces "face.11" connected save volume create height 23 radius1 3.683 radi us3 3.683 offset -11.5 0 0 xaxis frustum /ERROR occurred in the next command! volume move "volume.9" offset -4 0 0 volume create height 27 radius1 2.3875 radi us3 2.3875 offset -13.5 0 0 xaxis frustum volume unite volumes "volume.9" "volume.10" /ERROR occurred in the next command! volume delete "volume.8" lowertopology volume unite volumes "volume.2" "volume.9" undo /Undone to: volume unite volumes "volume.2" "volume.9" /Undone to: volume unite volumes "volume.2" "volume.9" undo /Undone to: volume delete "volume.8" lowertopology

PAGE 143

143 /Undone to: volume delete "volume.8" lowertopology /ERROR occurred in the next command! volume delete "volume.8" lowertopology save save name \ "C:\\Documents and Settings\\saurav\\D esktop\\3Dtestsection\\without-cooling-air16thmay-copy.dbs"

PAGE 144

144 LIST OF REFERENCES [1] Deutschmann, O., Schmidt, R., Behrendt, F., and Warnatz, J., Numerical Modeling of Catalytic Ignition, Proceedings of the Combustion Institute Pittsburg, Vol. 26, 1996, pp.1747-1754. [2] Mantzaras, Appel, C., and P. Benz, P ., Catalytic Combustion of Methane/Air Mixtures over Platinum: Homogeneous Ignition Distances in Channel Flow Configurations, Proceedings of the Combustion Institute Edinburg, Vol. 28, 2000, pp. 1349-1358. [3] Veser, G., and Schmidt, L.D., Ignition and Extinction in the Catalytic Oxidation of Hydrocarbons over Platinum, American Institute of Chem ical Engineering Journal Vol. 42 (4), 1996, pp. 1077-1087. [4] Krebs, W., Gruschka, U., Fielenbach, C., a nd Hoffman, S., CFD-analysis of reacting flow in an annular combustor, Progress in Computational Fluid Dynamics Vol 1, pp.104-116, 2001. [5] Hua, J., Meng Wu, M., and Kumar, K., Numerical Combustion of Hydrogen-air mixture in micro-scaled chambers Part II: CFD analysis for a micro-combustor, Chemical Engineering Science Vol 60, pp. 3507-3515. [6] Lefebvre, A. H., Gas Turbine Combustion Hemisphere Publis hing Corporation, 1983. [7] Control of Nitrogen Oxide emissions: Selective Catalytic Reduction (SCR), Report conducted jointly by The U.S Department of Energy and Southern Company Services, Inc Number 9, July 1997. [8] Maxwell, J.D., and A.L. Baldwin, A. L., Demonstration of Selective Catalytic Reduction (SCR) Technology for the Contro l of Nitrogen Oxid e (NOx) Emissions from High-Sulfur, Coal-Fired Boilers, First Annual Clean Coal Technology Conference Cleveland, OH, September 1992. [9] Rao, S. N., McIlvried, H. G., and Mann, A. N., Evaluation of NOx Removal Technologies, Selective Catalytic Reduction Vol 1. [10] Chorpening, B., Richards, A. G., Casleton, K. H., D emonstration of a Reheat Combustor for Power Production with CO2 Sequestration, Journal of Engineering for Gas Turbines and Power Vol-127, Issue-4, October 2005, pp. 740-747. [11] Woike, M., and Willis, B., Performance of a dry low-NOx Gas Turbine combustor design with a new fuel supply concept Journal of Engineering for Gas Turbines and Power Vol 124, Issue 4, October 2002, pp.771-775.

PAGE 145

145 [12] Institute for Interfacial Catalysis Pacific Northwest National Laboratory, (http://iic.pnl.gov ) [13] Melhado, E. M., Jacob Berzelius, the emergenc e of his chemical systems. [14] IUPAC Compendium of Chemical Terminology 2nd Edition, 1997. [15] Bond, G. C., Heterogeneous Catalysis: Principles and Applications Second Edition, Oxford University Press. [16] Boreskov, G. K., Heterogeneous Catalysis Nova Science Publishers, Inc., New York, 2003. [17] Masel, R. I., Chemical Kinetics and Catalysis John Wiley & Sons, Inc., Publication, 2001. [18] James. C., Chemical and Catalytic reaction engineering McGraw Hill 1976. [19] Smith, J. M. Chemical engineering ki netics 3rd edition. New York; London: McGraw-Hill, 1981. [20] Xu and Chuang, Kinetics of acetic ac id esterification over ion exchange catalysts, Canadian journal of chemical engineering Vol. 74, 1996, pp. 493 [21] Deutschmann, O., Maier, L. I., and Ri edel, U., Hydrogen Assisted Catalytic Combustion of Methane on Platinum, Catalysis Today Vol. 59, 2000, pp. 141-150. [22] Nakra, S., Green J. R., and Anderson, L. S., Thermal decomposition of JP-10 studied by micro-flowtube pyr olysis-mass spectrometry, Combustion and Flame, Vol. 144, Issue 4, March 2006, pp. 662-674. [23] Seery, D. J., and Zabielski, M. F., Com parisons between flame species measured by probe sampling and optical spectrometry techniques, Combustion and Flame Vol. 78, Issue 1, October 1989, pp. 169-177 [24] Korobeinichev, O. P., Kuibida, L. V ., Volkov E. N., and Shmakov, G., Mass spectrometric study of combustion and thermal decomposition of GAP, Combustion and Flame, Vol. 129, Issues 1-2, April 2002, pp. 136-150. [25] American Society for Mass Spectroscopy (http://www.asms.org ) [26] American Society for Chromatography, (www.chromatography-online.org [27] Ancia, R., Tiggelen, P. J. V., and Vandooren, J., Gas Chromatography as a complementary Analytic Technique to Molecular Beam Mass Spectrometry for Studying Flame Temperature, Combustion and Flame Vol. 116, pp. 307-309 (1999).

PAGE 146

146 [28] Eckbreth, A. C., Laser Diagnostics for Combustion Temperature and Species 2nd Edition, Gordon and Breach Publications. [29] EML 6934 Course Study material, Dr. D. Hahn, spring 2006, University of Florida, Gainesville, FL. [30] Barlow, R. S., and Carter, C. D., Raman /Rayleigh/LIF Measurements of Nitric Oxide Formation in Turbulent Hydrogen Jet Flames, Combustion and Flame Vol. 97, 1994, pp. 261-280. [31] Barlow, R. S., and Frank, J. H., S imultaneous Rayleigh, Raman, and LIF Measurements in Turbulent Premixed Methane-Air Flames, 27th International Symposium on Combustion, the Combustion Institute Pittsburg, PA, pp. 759-766. [32] Steffen, C. J. Jr., and Yungster, S., C omputational Analysis of the Combustion Process in an Axisymmetric, RBCC Flow path, NASA/TM-2001-210679 Paper. [33] Cline, M. C., Deur, J. M ., Micklow, G. J., Harper, M. R., and Kundu, P. K., Computation of the Flow Field in an Annular Gas Turbine Combustor, 29th Joint Propulsion and Exhibit, AIAA 93-2074 ., June 28-30, 1993. [34] GAMBIT users guide documentation Version 2.3.32 [35] FLUENT users guide Version 6.2.16 [36] Patankar, S. V., Numerical Heat Transfer and Fluid Flow Taylor and Francis Publications. [37] Dalla Betta, R.A., Schlatter, J.C., Nic kolas, S.G., Yee, D.K. and Shoji, T., ASME Paper No. 94-GT-260 (1994). [38] Schlegel, A., Buser, S., Benz P ., Bockhorn, H., and Mauss, F., Proceedings, 25th Symposium (International) on Comb ustion, the Combustion Institute Pittsburgh, PA, 1994, pp. 1019-1026. [39] Kraft, M., Fey, H., Bockhorn, H., Schleg el, A., and Chen, J. Y., A Numerical Study on the Influence of Mixing Intensity on NOx formation, International: Proceedings of the 3rd Workshop on M odeling of Chemical Reaction Systems 1997. [40] Deutschmann, O., Behrendt, F., and Wa rnatz, J., Modeling and Simulation of Heterogeneous Oxidation of Methane on a Platinum foil, Catalysis Today Vol. 21, pp. 461-470, 1994. [41] Otto Beeck, CatalysisA Challenge to the Physicist, Reviews of Modern Physics Vol. 17, No. 1, January, 1945.

PAGE 147

147 [42] Song, X., Williams, W. R., Sc hmidt, L. D., and Aris, R ., Bifurcation behavior in homogeneous-heterogeneous combustion: II. Computations for stagnation-point flow, Combustion and Flame Vol. 84, Issues 3-4, April 1991, pp. 292-311. [43] Williams, W. R., Stenzel, M. Song T. X. and Schmidt, L. D., Bifurcation behavior in homogeneous-het erogeneous combustion: I. Experimental results over platinum, Combustion and Flame Vol. 84, Issues 3-4, pp. 277-291, 1991. [44] Pfefferle, L. D., Griffin, T. A., Winter, M. D. R., and Dyer, M. J., The influence of catalytic activity on the ignition of boundary layer flows Part I: Hydroxyl radical measurements, Combustion and flame Vol. 76, 1989, pp. 325-338. [45] Markatou, P., Pfefferle L. D., and Sm ooke, M. D., A computational study of methane-air combustion over heated ca talytic and non-catalytic surfaces, Combustion and Flame Vol. 93, 1993, pp.185-201. [46] Hickman, D.A. and Schmidt, L.D., AIChE Journal Vol. 39, 1993, pp. 1164-1177. [47] Griffin, T. A., and Pfefferle, L.D., Gas Phase and Catalytic Ignition of Methane and Ethane in Air over Platinum, AIChE Journal Vol. 36, 1990, 861-870. [48] Veser, G. and Schmidt, L. D., Ignition and Extinction in the Catalytic Oxidation of Hydrocarbons over Platinum, AIChE Journal Vol. 42 (4), 1996, pp. 1077-1087. [49] F. Behrendt, F., Ignition and Extinc tion of Hydrogen-Air and Methane-Air Mixtures over Plati num and Palladium, ACS Symposium Series 1996, pp. 48-57. [50] Schwiedernoch, R. Experimental and numerical investigation of the ignition of methane combustion in a platinum-coated honeycomb monolith, Proceedings of Combustion Institute Saporo, Japan, Vol. 29, 2002, pp. 1005-1011. [51] Baulch, D. L., Evaluated Kineti c Data for Combustion Modeling, Journal of Physical Chemistry Vol. 21 (3), 1992, pp. 411-734. [52] Yong, S. S., Sung J. C., Sung, K. K., a nd Hyun, D. S., Numerical Studies of Catalytic Combustion in a Catalytically Stabilized Combustor, International Journal of Energy Research Vol. 24, 2000, pp. 1049-1064. [53] Reinke, M., Mantzaras, J., Schaeren, R., Bombach, R., Kreutner W., and Inauen, A., Homogeneous Combusti on in High Pressure Combus tion of Methane/Air over Platinum: Comparison of Measurements and detailed Numerical Predictions, Proceedings of the combustion Institute Vol. 29, 2002, pp. 1021-1029. [54] Nedunchezian, N., and Dhandapani, S., S tudy of Flame quench ing and near-wall combustion of lean burn fuel-air mixture in a catalytically activated spark-ignited lean burn engine, Combustion and Flame Vol. 144, 2006, pp. 407-409.

PAGE 148

148 [55] Heywood, J. B., Internal Combustion Engine Fundamentals McGraw-Hill, New York, 1989. [56] Aghalayam, P., A C1 mechanism fo r methane oxidation on platinum, Journal of Catalysis Vol. 213 (1), 2003, pp. 23-38. [57] Ihmann, K., Ihmann, Jentoft, F. C., Gesk e, M., Taha, Pelzer A. K., and Schlogl R., Molecular beam mass spectrometer e quipped with a catalyt ic wall reactor for insitu studies in high temperat ure catalysis research, Review of Scient ific Instruments -May 2006. Review of Scie ntific Instruments, Vol. 77, 2006 [58] Grunwaldt1, J. D., Hannemann, S., Schroe r, C. G., Baiker, A., D-mapping of the structure of a heterogeneous catalyst in side a catalytic micr o reactor during the partial oxidation of methane, Journal of Physical Chemistry B Vol. 110, 2006. [59] Yeh, T., Lee, H., Chu, K., and Wang C., Charecterization and catalytic combustion of methane over hexaaluminate, Materials Science and Engineering A Vol. 384, Issue 1-2, pp. 324-330. [60] Dogwiler, U., Mantzaras, J., Benz, P., Homogeneous Ignition of Methane/Air Mixtures over Platinum: Comparison of Measurements and Detailed Numerical Predictions, 27th International Symposium on Combustion Boulder, CO, August 1998.

PAGE 149

149 BIOGRAPHICAL SKETCH Saurav Pathak was born in Be ttiah, India, on December 17th, 1976, and raised in Kolkata, India. Saurav attended the Birla Institute of Technology, located in Mesra India, where he received a Bachelor of Engineering degree in me chanical engineering in 2000. Thereafter Saurav joined the Department of Mechan ical Engineering at the Ohio Un iversity in August 2001 and got a M. S. in Mechanical Engineering in May 2003. After 2003, he has attended the College of Engineering at the Louisiana State Univers ity, Baton Rouge, LA to pursue his Ph.D. in Mechanical Engineering. After staying there fo r two semesters, Saurav transferred to the University of Florida, Gainesville and starte d working as Ph.D. from December 2005.During this time he worked as a research assistant in the Mechanical and Aerospace Engineering Department on a full-time basis. His research interests include experimenta tion and computational catalytic combustion.