10112/

15112

CHEMKIN Tutorials Manual

CHEMKIN Software
CK-TUT-10112-1112-UG-1
December 2011
Licensing:
For licensing information, please contact Reaction Design at (858) 550-1920 (USA) or [email protected].

Technical Support:
Reaction Design provides an allotment of technical support to its Licensees free of charge. To request technical support, please include your license
number along with input or output files, and any error messages pertaining to your question or problem. Requests may be directed in the following
manner: E-mail: [email protected], Fax: (858) 550-1925, Phone: (858) 550-1920.
Additional technical support hours may also be purchased. Please contact Reaction Design for the hourly rates.

Copyright:
Copyright 2011 Reaction Design. All rights reserved. No part of this book may be reproduced in any form or by any means without express written
permission from Reaction Design.

Trademark:
CHEMKIN and REACTION DESIGN are registered trademarks of Reaction Design in the United States and other countries.
AURORA, CONP, CHEMKIN-CFD, CHEMKIN, CRESLAF, ENERGICO, EQUIL, EQUILIB, FORT, KINETICS, MODEL FUELS CONSORTIUM, OPPDIF, OVEND,
PARAMETER STUDY FACILITY, PARTICLE TRACKING FEATURE, PASR, PLUG, PREMIX, REACTION WORKBENCH, SENKIN, SHOCK, SPIN, SURFACE CHEMKIN,
SURFTHERM, TRANSPORT, TWAFER, TWOPNT are all trademarks of Reaction Design or Sandia National Laboratories. All other trademarks are the
property of their respective holders.

Limitation of Warranty:
The software is provided as is by Reaction Design, without warranty of any kind including, without limitation, any warranty against infringement of
third party property rights, fitness or merchantability, or fitness for a particular purpose, even if Reaction Design has been informed of such purpose.
Furthermore, Reaction Design does not warrant, guarantee, or make any representations regarding the use or the results of the use, of the software or
documentation in terms of correctness, accuracy, reliability or otherwise. No agent of Reaction Design is authorized to alter or exceed the warranty
obligations of Reaction Design as set forth herein. Any liability of Reaction Design, its officers, agents or employees with respect to the software or the
performance thereof under any warranty, contract, negligence, strict liability, vicarious liability or other theory will be limited exclusively to product
replacement or, if replacement is inadequate as a remedy or in Reaction Designs opinion impractical, to a credit of amounts paid to Reaction Design
for the license of the software.

Literature Citation for CHEMKIN and CHEMKIN-PRO:

CHEMKIN 10112 should be cited as:
CHEMKIN 10112, Reaction Design: San Diego, 2011.

CHEMKIN-PRO 15112 should be cited as:

CHEMKIN-PRO 15112, Reaction Design: San Diego, 2011.
 CHEMKIN

Contents

Table of Contents
1 Introduction.................................................................................................................................................. 19

2 Combustion in Gas-phase Processes ....................................................................................................... 23

2.1 Equilibrium .....................................................................................................................................23
2.1.1 Adiabatic Flame Temperature ........................................................................................23
2.2 Using Equivalence Ratio ................................................................................................................24
2.2.1 Example: Propane in Air.................................................................................................25
2.2.2 Example: Stoichiometric Products..................................................................................27
2.3 Ignition, Flames and Flammability..................................................................................................28
2.3.1 Steady-state Gas-phase Combustion.............................................................................28
2.3.2 Autoignition for H2/Air .....................................................................................................30
2.3.3 Ignition-delay Times for Propane Autoignition................................................................33
2.3.4 Burner-stabilized Flame..................................................................................................38
2.3.5 NO Emission from High-pressure Flames with Gas Radiation [CHEMKIN-PRO
Only] .............................................................................................................................................41
2.3.6 Soot Formation in Radiating Opposed-flow Diffusion Flame [CHEMKIN-PRO
Only] .............................................................................................................................................47
2.3.7 Flame Speed of Stoichiometric Methane/Air Premixed Flame with Reaction
Path Analyzer [CHEMKIN-PRO Only].....................................................................................55
2.3.8 Parameter Study: Propane/Air Flame Speed as a Function of Equivalence
Ratio and Unburned Gas Temperature...................................................................................71
2.3.9 Hydrogen/Air Opposed-flow Flame ................................................................................81
2.3.10 Flame Extinction Analysis [CHEMKIN-PRO Only]..........................................................85
2.3.11 Stagnation Flame Analysis [CHEMKIN-PRO Only] ........................................................94
2.4 Internal Combustion Engine...........................................................................................................97
2.4.1 Homogeneous Charge Compression Ignition (HCCI) Engine ........................................97
2.4.2 Multi-zone HCCI Engine Simulation [CHEMKIN-PRO Only] ....................................... 103
2.5 Simulating a Shock-tube Experiment ........................................................................................... 110
2.5.1 Shock-heated Air (Shock).............................................................................................110
2.6 Combustion in Complex Flows..................................................................................................... 111
2.6.1 Gas Turbine Network.................................................................................................... 111
2.6.2 Jet Flame Network........................................................................................................ 114
2.6.3 Using Tear Streams to Estimate Initial Gas Composition in an HCCI Engine
with Exhaust Gas Recirculation (EGR) [CHEMKIN-PRO Only]........................................119
2.6.4 Partially Stirred Reactor for Methane/Air ...................................................................... 127
2.6.5 Side Inlet on a Plug Flow Reactor ................................................................................ 132
2.6.6 Co-flowing Non-premixed CH4/Air Flame ..................................................................... 135

CK-TUT-10112-1112-UG-1 3 2011 Reaction Design

CHEMKIN Table of Contents

2.7 Particle Tracking Feature [CHEMKIN-PRO Only] ....................................................................... 141

2.7.1 Soot Formation and Growth in a JSR/PFR Reactor [CHEMKIN-PRO Only] ............... 141
2.7.2 Soot Particles in Flame Simulators [CHEMKIN-PRO Only].......................................... 147
2.7.3 Sectional Method for Particle-Size Distribution with Pre-mixed Laminar Burner-
Stabilized Stagnation Flame [CHEMKIN-PRO Only].........................................................153
2.7.4 Simulating Particle-Size Distributions in a Burner-Stabilized Stagnation Flame .......... 159
2.7.5 Detailed Particle Aggregation in a Batch Reactor [CHEMKIN-PRO Only] .................. 162
2.8 Uncertainty Analysis [CHEMKIN-PRO Only]............................................................................... 170
2.8.1 Uncertainty Analysis of NOx Emissions [CHEMKIN-PRO Only] ................................. 170
2.9 Chemistry Sets............................................................................................................................. 173
2.9.1 Hydrogen/Air................................................................................................................. 173
2.9.2 Methane/Air .................................................................................................................. 174
2.9.3 NOx and CH4, C2H4, C2H6, C3H6, and C3H8................................................................................ 175
2.9.4 Propane/Air................................................................................................................... 175
2.9.5 Ethylene/Air Combustion and Soot Formation and Growth.......................................... 175
2.9.6 C2_NOx Mechanism .................................................................................................... 178

3 Catalytic Processes................................................................................................................................... 181

3.1 Catalytic Combustors, Converters and Aftertreatment................................................................. 181
3.1.1 Two-Stage Catalytic Combustor................................................................................... 181
3.1.2 Engine Exhaust Aftertreatment with a Transient Inlet Flow.......................................... 191
3.2 Parameter Studies ....................................................................................................................... 200
3.2.1 Parameter Study Facility for Surface Chemistry Analysis ............................................ 200
3.3 Chemistry Sets............................................................................................................................. 204
3.3.1 Methane Oxidation on Pt ..............................................................................................204
3.3.2 Pt/Rh Three-way Catalyst............................................................................................. 205

4 Materials Problems.................................................................................................................................... 207

4.1 Chemical Vapor Deposition.......................................................................................................... 207
4.1.1 Equilibrium Analysis of Chlorosilane CVD.................................................................... 208
4.1.2 PSR Analysis of Steady-state Thermal CVD................................................................ 210
4.1.3 Approximations for a Cylindrical Channel Flow ............................................................ 214
4.1.4 Deposition in a Rotating Disk Reactor.......................................................................... 219
4.1.5 Trichlorosilane CVD in Planar Channel Flow Reactor .................................................. 223
4.2 Atomic Layer Deposition (ALD).................................................................................................... 226
4.2.1 Time-dependent Simulations of ALD Process.............................................................. 227
4.3 Plasma Etching ............................................................................................................................ 233
4.3.1 Steady-state Chlorine Plasma ...................................................................................... 233
4.3.2 Spatial Chlorine Plasma PFR with Power Profile ......................................................... 237
4.3.3 Fluorocarbon Plasma Etching of Silicon Dioxide.......................................................... 241
4.4 Chemistry Sets............................................................................................................................. 246
4.4.1 Silicon Nitride CVD from Silicon Tetrafluoride and Ammonia ....................................... 246
4.4.2 Silicon Deposition from Silane...................................................................................... 247
4.4.3 Silicon Deposition from Trichlorosilane ........................................................................ 247
4.4.4 Alumina ALD................................................................................................................. 248
4.4.5 Chlorine Plasma ...........................................................................................................250
4.4.6 Fluorocarbon Plasma with SiO2 Etch Products ............................................................ 251

5 Chemical Mechanism Analysis ................................................................................................................ 253

5.1 Mechanism Analyzer.................................................................................................................... 253
5.1.1 Background Information ............................................................................................... 254
5.1.2 Reaction Mechanism for Diamond CVD....................................................................... 256

2011 Reaction Design 4 CK-TUT-10112-1112-UG-1

Table of Contents Tutorials Manual

Index ........................................................................................................................................................... 269

CK-TUT-10112-1112-UG-1 5 2011 Reaction Design

CHEMKIN Table of Contents

2011 Reaction Design 6 CK-TUT-10112-1112-UG-1

 CHEMKIN

Contents

List of Tables
1-1 Reactor Models Used in Sample Problems...................................................................................................................20

2-1 Reactant Mole Fractions Sum to 1.0 .............................................................................................................................25

2-2 Relative Moles Normalized So Mole Fractions Sum to 1.0 ...........................................................................................25

2-3 N2 as Added Diluent .....................................................................................................................................................26

2-4 N2 as Component of Oxidizer .......................................................................................................................................26

2-5 Determining Stoichiometric Products ............................................................................................................................27

2-6 Initial Temperatures and Ignition Times ......................................................................................................................36

2-7 Inlet velocities for pressures studied by Thomsen et al. ...............................................................................................45

2-8 Test Engine Specifications ............................................................................................................................................98

2-9 Composition of Initial Gas Mixture.................................................................................................................................99

2-10 Engine parameters of the validation case. ..................................................................................................................103

2-11 Natural gas composition used in the validation case. .................................................................................................103

2-12 Zone configuration used by the multi-zone model analysis for the validation case. ...................................................104

2-13 Components in Fuel and Oxidizer Streams.................................................................................................................133

2-14 Properties of the Co-flowing Jets ................................................................................................................................136

3-1 Excerpt of Data Representing Engine-out Test Measurements ..................................................................................193

CK-TUT-10112-1112-UG-1 7 2011 Reaction Design

CHEMKIN List of Tables

2011 Reaction Design 8 CK-TUT-10112-1112-UG-1

 CHEMKIN

Contents

List of Figures
2-1 Adiabatic Flame TemperaturesHydrogen/Air Mixture ................................................................................................24

2-2 Steady-state Gas-phase CombustionHydrogen/Air Temperatures ..........................................................................29

2-3 Steady-state Gas-phase CombustionMolar Conversions .........................................................................................30

2-4 Autoignition for Hydrogen/AirTemperature Profile .....................................................................................................32

2-5 Autoignition for Hydrogen/AirSpecies Composition Profiles ......................................................................................32

2-6 Autoignition for Hydrogen/AirSensitivity Coefficients ................................................................................................33

2-7 Temperature and OH Mole Fraction Profiles as a Function of Time .............................................................................36

2-8 Ignition Based on Varying Criteria .................................................................................................................................37

2-9 Ignition Times vs. Inverse of Temperature (Semi-log) ..................................................................................................37

2-10 Burner-stabilized FlameExperimental Gas Temperature Profile ...............................................................................40

2-11 Burner-stabilized FlameMole Fractions .....................................................................................................................41

2-12 Axial gas temperature profiles predicted with and without gas radiation heat loss as compared against
experimental temperature profile for the phii=0.6 CH4/O2/N2 flame at 14.6 atm. .......................................................46

2-13 Comparisons of measured and predicted NO mole fraction profiles as a function of axial distance
from the burner surface for the phi=0.6 CH4/O2/N2 flame at 14.6 atm. .......................................................................46

2-14 Measured and predicted NO concentrations behind the phi=0.6 CH4/O2/N2 flame versus pressure. ........................47

2-15 Predicted axial velocity profile of Flame 1 studied by Atreya et al. 13 (p. 48). The location of
(velocity) stagnation plane is indicated by the blue dash-dotted line. ...........................................................................51

2-16 Predicted gas temperature profile is compared against the experimental profile for Flame 1 studied by
Atreya et al. 13 (p. 48). .....................................................................................................................................................52

2-17 Comparisons of predicted and measured fuel (CH4) and oxidizer (O2) profiles for Flame 1 studied by
Atreya et al. 13 (p. 48). .....................................................................................................................................................52

2-18 Comparisons of predicted and measured profiles of H2 and CO for Flame 1 studied by Atreya et

CK-TUT-10112-1112-UG-1 9 2011 Reaction Design

CHEMKIN List of Figures

al. 13 (p. 48). ....................................................................................................................................................................53

2-19 Predicted and measured C2H2 and OH profiles of Flame 1 studied by Atreya et al. 13 (p. 48). The
OH peak indicates the flame front and the C2H2 peak marks the major soot growth region. ......................................53

2-20 Predicted pyrene (A4) profile showing soot inception mainly occurs in the region between the
stagnation plane and the diffusion flame in Flame 1 studied by Atreya et al. 13 (p. 48). .................................................54

2-21 Predicted soot volume fraction profile is compared against the experimental profile as a function of
distance from the fuel nozzle for Flame 1 studied by Atreya et al. 13 (p. 48). .................................................................54

2-22 Flame SpeedTemperature Profile Panel ..................................................................................................................56

2-23 C1_ Flame SpeedReactor Physical Properties ........................................................................................................56

2-24 Flame SpeedAxial Velocity vs. Distance ...................................................................................................................58

2-25 Flame SpeedTemperature vs. Distance ....................................................................................................................59

2-26 Flame SpeedOutput Calculation pane with Sensitivity Option. .................................................................................60

2-27 Flame SpeedAnalyze Results panel with Reaction Path Analyzer option. ................................................................61

2-28 Flame SpeedAltering the solution chosen, and y variable shown in the solution plot. ..............................................62

2-29 Flame SpeedSwitching to a hierarchical layout option. .............................................................................................63

2-30 Flame SpeedIncreased complexity in the diagram due to changing the solution point, and increasing
the number of allowed species. .....................................................................................................................................64

2-31 Flame SpeedAltering the zoom setting to see the contents of the diagram. .............................................................65

2-32 Flame SpeedDecomposition pathways starting with methane. .................................................................................66

2-33 Flame SpeedThe Rate of Production and Sensitivity charts of methane. .................................................................67

2-34 Flame SpeedSplitting a composite reaction pathway into one per reaction. .............................................................68

2-35 Flame SpeedTree layout method of the decomposition of methane. ........................................................................69

2-36 Flame SpeedSetting the hydroxide radical and oxygen radical as side species. The source was
switched from methane to oxygen, and the element filter was changed to O element. ................................................70

2-37 Flame SpeedColored reactions based on influence of side species. ........................................................................71

2-38 Panel for Setup of Parameter Studies ...........................................................................................................................73

2-39 Automatic Estimation of Starting Temperature Profile .................................................................................................74

2-40 Auto-Populate Option ...................................................................................................................................................76

2-41 Varying Each Parameter Independently .......................................................................................................................77

2-42 Parameter Study After Setup Is Complete ...................................................................................................................77

2011 Reaction Design 10 CK-TUT-10112-1112-UG-1

List of Figures Tutorials Manual

2-43 Post-Processor Options After Running Parameter Studies ..........................................................................................79

2-44 Cluster Endpoint vs. Parameter ..................................................................................................................................80

2-45 Flame-Speed Calculated Values as a Function of Equivalence Ratio ........................................................................81

2-46 Hydrogen/Air FlameTemperature vs. Axial Distance .................................................................................................83

2-47 Hydrogen/Air FlameMole Fractions ...........................................................................................................................83

2-48 Hydrogen/Air FlameTemperature Sensitivity to Reaction Rates ...............................................................................84

2-49 Hydrogen/Air FlameTemperature Sensitivity to Heats of Formation ........................................................................84

2-50 Icon for Extinction of Premixed or Opposed Flow Flame reactor ..................................................................................85

2-51 Extinction Basic reactor properties ............................................................................................................................87

2-52 Dynamic plotting of Temperature vs. extinction response, showing a turn at ~550 s-1 ................................................89

2-53 Output file opposed-flow_flame__extinction_ExtinctionResponse.csv .........................................................................90

2-54 Output file ExtinctionProfilesOut. ..................................................................................................................................91

2-55 Flame response curve showing extinction (turning) for premixed stoichiometric methane-air flame. The
inlet temperature is 296 K and ambient pressure is 1 atm. The calculated extinction strain rate is
550 /s. ...........................................................................................................................................................................92

2-56 Axial velocity and temperature profile for premixed stoichiometric methane-air flame at extinction point.
The global strain rate at extinction is 289 s-1 ...............................................................................................................93

2-57 Axial velocity and velocity gradient magnitude for premixed stoichiometric methane-air flame at
extinction point. Also shown are the points corresponding to the definitions of various extinction strain
rates ..............................................................................................................................................................................93

2-58 Icon for Premixed Laminar Burner-Stabilized Stagnation Flame model. .....................................................................94

2-59 Parameter Study set-up. ...............................................................................................................................................95

2-60 Temperature and velocity profiles for two inlet velocities ..............................................................................................96

2-61 Major species mole faction profiles for each of two simulation runs. ...........................................................................97

2-62 IC Engine Icon ...............................................................................................................................................................98

2-63 C1_ IC EngineReactor Physical Property for HCCI Woschni Heat Loss model ......................................................100

2-64 HCCI EngineEGR Temperature Comparison ..........................................................................................................101

2-65 HCCI EngineEGR Pressure Comparison ...............................................................................................................102

2-66 HCCI EngineEGR Heat Loss Comparison ..............................................................................................................102

2-67 Reactor model palette showing the multi-zone model icon in the CHEMKIN User Interface. .....................................104

CK-TUT-10112-1112-UG-1 11 2011 Reaction Design

CHEMKIN List of Figures

2-68 Diagram view of a multi-zone simulation project. ........................................................................................................104

2-69 Dialog for specifying the number of zones in the multi-zone simulation. ...................................................................105

2-70 Engine parameters in Reactor Physical Properties tab for multi-zone model. ...........................................................106

2-71 Zone Properties table. ...............................................................................................................................................106

2-72 Individual Zone Properties tab of the multi-zone model graphic user interface using a temperature
profile called zone5 ..................................................................................................................................................107

2-73 Temperature profiles predicted by the single-zone and the multi-zone models. The transition angle is
set to 3 degrees BTDC. ...............................................................................................................................................108

2-74 Comparison of number density distributions obtained with and without aggregation model.......................................109

2-75 Zone volume profiles predicted by the multi-zone models. ........................................................................................109

2-76 Nitrogen Atom Reaction in CHEMKIN Format ............................................................................................................110

2-77 Shock Tube ExperimentNO Mole Fraction ..............................................................................................................111

2-78 Gas Turbine NetworkSchematic .............................................................................................................................113

2-79 Gas Turbine NetworkDiagram View .......................................................................................................................113

2-80 Gas Turbine NetworkTemperature Comparisons ..................................................................................................114

2-81 Gas Turbine NetworkCO and NO Comparisons ....................................................................................................114

2-82 Jet Flame NetworkDiagram View ...........................................................................................................................115

2-83 Jet Flame NetworkTemperature Distribution ...........................................................................................................117

2-84 Jet Flame NetworkCO Distribution ..........................................................................................................................117

2-85 Jet Flame NetworkNO Distribution ..........................................................................................................................118

2-86 Jet Flame NetworkMole Fractions ...........................................................................................................................119

2-87 Diagram view of the EGR_network project. ................................................................................................................120

2-88 Reactor Properties panel of the EGR_Merge reactor. ................................................................................................121

2-89 Stream Properties panel of the fresh fuel-air mixture entering the EGR_Merge reactor . ..........................................121

2-90 Engine Properties panel of the HCCI engine model....................................................................................................122

2-91 Properties of the Heat_Loss reactor............................................................................................................................123

2-92 Initialization panel of the inlet stream to the Heat_Loss reactor. .................................................................................123

2-93 Properties of the inlet stream to the Heat_Loss reactor. .............................................................................................123

2-94 Split Factor panel for the Gas Splitter. ........................................................................................................................124

2011 Reaction Design 12 CK-TUT-10112-1112-UG-1

List of Figures Tutorials Manual

2-95 Tear Stream Controls panel. .......................................................................................................................................124

2-96 Popup reminder for the initialize Tear Streams panel. ..............................................................................................124

2-97 Panel for entering guess properties of the Tear Stream. ............................................................................................125

2-98 Detail view of the calculations in which clusters of the same Tear Stream are grouped together and
highlighted. ..................................................................................................................................................................125

2-99 Project Results ............................................................................................................................................................125

2-100 Solution selection panel. .............................................................................................................................................126

2-101 Comparison of EGR effect on in-cylinder pressure. Solid line: no EGR; Dash-dotted line: 30% EGR
by mass. ......................................................................................................................................................................126

2-102 Comparison of EGR effect on gas temperature. Solid line: no EGR; Dash-dotted line: 30% EGR by
mass. ...........................................................................................................................................................................126

2-103 Comparison of EGR effect on CO2 emission. Solid line: no EGR; Dash-dotted line: 30% EGR by
mass. ...........................................................................................................................................................................127

2-104 Comparison of EGR effect on NO emission. Solid line: no EGR; Dash-dotted line: 30% EGR by
mass. ...........................................................................................................................................................................127

2-105 PaSR Methane/AirTemperature Comparison ..........................................................................................................130

2-106 PaSR Methane/AirTemperature Variance Comparison ...........................................................................................131

2-107 PaSR Methane/AirCO Comparison .........................................................................................................................131

2-108 PaSR Methane/AirSelect the Import File Format.....................................................................................................132

2-109 PaSR Methane/AirPDF(T) Comparison ...................................................................................................................132

2-110 Illustration of the System to be Modeled. ....................................................................................................................133

2-111 Total Flow Growing Directly as a Result of Side Inlet Flow. .......................................................................................134

2-112 Peaking of NO Concentration Dependent on Temperature in a PFR with an Approximated Mixing

Region. .......................................................................................................................................................................135

2-113 Confined Co-flowing Annular Jet Configuration. .........................................................................................................136

2-114 Preferences Panel Showing the Display User Routine Options Is Enabled................................................................137

2-115 Get Initial Solution Profile from User Routine Check box. ...........................................................................................138

2-116 Visualization of Full Computational Domain Can Be Acquired by Selecting Proper Reflect Contour Data
Option in the CHEMKIN Post-Processor Panel...........................................................................................................139

2-117 Contour Plot Window Showing the Pull-down List Selections of the Insert Menu. .....................................................140

2-118 Colorbar Property Panel. .............................................................................................................................................140

CK-TUT-10112-1112-UG-1 13 2011 Reaction Design

CHEMKIN List of Figures

2-119 Final 3-D Temperature Contours of the Entire Physical Domain with Vertical Contour Legend
(Colorbar). ...................................................................................................................................................................141

2-120 A schematic of the JSR/PFR reactor configuration used by Marr27 ............................................................................142

2-121 Diagram View of the CHEMKIN Project Used to Simulate the JSR/PFR Experiment ................................................143

2-122 Specifying Particle Tracking Feature Parameters and Initial Conditions of Particle Size Moments in the
Reactor .......................................................................................................................................................................143

2-123 Tolerances for Particle Size Moments Can Be Given Explicitly in the Solver Window ...............................................144

2-124 Comparisons of Mole Fraction Profiles of Selected Gas Phase Species Inside the PFR for the 1630K
and F = 2.2 Case of the C2H4/O2/N2 JSR/PFR Experiment by Marr27. Symbols: data; Solid lines:
predictions with HACA and PAH condensation growth mechanisms ..........................................................................145

2-125 Comparisons of Soot Mass Concentration Profiles Inside the PFR for the 1630K and F = 2.2 Case of
the C2H4/O2/N2 JSR/PFR Experiment by Marr27. Symbols: data; Solid line: prediction with both HACA
and PAH condensation growth mechanisms; Dash-dot line: prediction with HACA growth mechanism
only .............................................................................................................................................................................146

2-126 The Particle Diameter Evolution Inside the PFR Predicted by the Present Soot Module for the 1630K
and F = 2.2 Case of the C2H4/O2/N2 JSR/PFR Experiment by Marr27 (see p. 141) ......................................................146

2-127 Distance vs temperature and particle volume fraction ..............................................................................................151

2-128 Distance vs species mole fraction and particle number density ................................................................................151

2-129 Distance vs Temperature and Particle volume fraction ..............................................................................................152

2-130 Distance vs species mole fraction and particle number density ................................................................................153

2-131 Selecting the Sectional Method on the Reactor Physical Properties tab. ..................................................................155

2-132 Appearance of the Dispersed Phase tab when the Sectional Method is selected. ....................................................155

2-133 Sectional Method plot of temperature and mole fraction of pyrene (A4). ...................................................................157

2-134 Particle-size distribution as a function of distance. .....................................................................................................158

2-135 PDF as a function of distance ....................................................................................................................................159

2-136 Particle-size distribution at the stagnation plane for three conditions imposed by the parameter study. ........................
161

2-137 Imposed temperature profile with the mole fraction of the precursor TiCl4. ................................................................165

2-138 Imposed temperature profile with the nucleation rate of the precursor TiCl4. .............................................................165

2-139 Complete Aggregation - Aggregate number density as a function of the volume equivalent section
diameter. .....................................................................................................................................................................166

2-140 Evolution of the primary particle diameter. ..................................................................................................................167

2-141 Number density distributions obtained without aggregation. ......................................................................................168

2011 Reaction Design 14 CK-TUT-10112-1112-UG-1

List of Figures Tutorials Manual

2-142 Collision diameter as a function of (volume equivalent) section diameter. .................................................................169

2-143 Comparison of the total particle number and surface area densities as a function of time with and
without aggregation model ..........................................................................................................................................170

2-144 Setting Up an Uncertainty Analysis for Heat Loss.......................................................................................................172

2-145 Setting Up the Uncertainty Distribution for Heat Loss .................................................................................................172

2-146 Probability Density Function of NO .............................................................................................................................173

3-1 Turbine Flow Capacity.................................................................................................................................................182

3-2 Two-Stage Catalytic CombustorDiagram View .......................................................................................................185

3-3 Catalytic Pre-combustor (C1_)Honeycomb Monolith, Catalyst sub-tab .................................................................185

3-4 Catalytic Pre-combustor (C1_)Honeycomb Monolith, Honeycomb sub-tab ...........................................................186

3-5 Homogeneous Stage Combustor (C3_)Reactor Physical Property ........................................................................186

3-6 Two Stage Catalytic CombustorExcess_Air_Dilution (Cluster 4) Output Results ...................................................187

3-7 Two Stage Catalytic CombustorTemperature Comparison .....................................................................................189

3-8 Two Stage Catalytic CombustorPressure Comparison ...........................................................................................189

3-9 Two Stage Catalytic CombustorCH4 Comparison ...................................................................................................190

3-10 Two Stage Catalytic CombustorCO Comparison ....................................................................................................190

3-11 Two Stage Catalytic CombustorNO Comparison ....................................................................................................191

3-12 Sample USRINLET Subroutine for User-Defined Transient Inlet Conditions..............................................................196

3-13 Engine Exhaust AftertreatmentMolar Conversion Rates .........................................................................................199

3-14 Engine Exhaust AftertreatmentMole Fractions ........................................................................................................199

3-15 Engine Exhaust AftertreatmentGas Temperatures ..................................................................................................200

3-16 Setting Up a Parameter Study for Sticking Coefficients vs. Predicted Temperature Profiles ....................................202

3-17 Varying Rate Constants of Reactions 1 and 6. ..........................................................................................................203

4-1 Chlorosilane CVDEquilibrium Calculations ..............................................................................................................209

4-2 Chlorosilane CVDMole Fractions.............................................................................................................................210

4-3 Steady-state Thermal CVDDeposition Rate vs. SiF4 Mole Fraction ........................................................................212

4-4 Steady-state Thermal CVDROP Comparison..........................................................................................................213

4-5 Steady-state Thermal CVDGrowth Rates ................................................................................................................213

4-6 Steady-state Thermal CVDSite Fractions ................................................................................................................214

CK-TUT-10112-1112-UG-1 15 2011 Reaction Design

CHEMKIN List of Figures

4-7 Cylindrical Channel FlowShear-flow Simulation of SiF4 Mole Fractions .................................................................216

4-8 Cylindrical Channel FlowSiF4 Mole Fractions Comparison .....................................................................................217

4-9 Cylindrical Channel FlowDeposition Rates Comparison .........................................................................................218

4-10 Cylindrical Channel FlowAxial Gas Velocities Comparison .....................................................................................218

4-11 Deposition in a Rotating DiskGas Velocity Components .........................................................................................220

4-12 Deposition in a Rotating DiskMole Fractions ...........................................................................................................221

4-13 Deposition in a Rotating DiskSi Atom Concentrations .............................................................................................222

4-14 Deposition in a Rotating DiskExperimental Data .....................................................................................................222

4-15 Trichlorosilane CVDGas Temperatures vs. Axial and Radial Distance ...................................................................224

4-16 Trichlorosilane CVDTrichlorosilane Mole Fraction .................................................................................................225

4-17 Trichlorosilane CVDSilicon Deposition Rate............................................................................................................225

4-18 Trichlorosilane CVDMole Fractions .........................................................................................................................226

4-19 Time-dependent ALD SimulationsPSR, Total Flow Rates vs. Time ........................................................................230

4-20 Time-dependent ALD SimulationsTMA Contour Plot ..............................................................................................230

4-21 Time-dependent ALD SimulationsO Mole Fraction .................................................................................................231

4-22 Time-dependent ALD SimulationsStagnation-flow Site Fractions ...........................................................................231

4-23 Time-dependent ALD SimulationsO Mole Fractions Comparison ...........................................................................232

4-24 Time-dependent ALD SimulationsDeposition Thickness Comparison ....................................................................232

4-25 Steady-state Chlorine PlasmaElectron Temperatures vs. Power ............................................................................235

4-26 Steady-state Chlorine PlasmaMole Fractions..........................................................................................................235

4-27 Steady-state Chlorine PlasmaCl and Cl2 Mole Fractions ........................................................................................236

4-28 Steady-state Chlorine PlasmaROP Analysis ...........................................................................................................237

4-29 Spatial Chlorine PlasmaPlasma Power vs. Distance...............................................................................................239

4-30 Spatial Chlorine PlasmaElectron Mole Fraction ......................................................................................................239

4-31 Spatial Chlorine PlasmaCl Mole Fraction ................................................................................................................240

4-32 Spatial Chlorine PlasmaCl+ and Cl2+ Mole Fractions .............................................................................................240

4-33 Spatial Chlorine PlasmaElectron Temperature .......................................................................................................241

4-34 Fluorocarbon Plasma Etching of Silicon DioxideSiO2 Etch Rates Variations..........................................................243

2011 Reaction Design 16 CK-TUT-10112-1112-UG-1

List of Figures Tutorials Manual

4-35 Fluorocarbon Plasma Etching of Silicon Dioxide10 Highest Mole Fractions ...........................................................244

4-36 Fluorocarbon Plasma Etching of Silicon DioxidePositive Ion Mole Fractions ..........................................................244

4-37 Fluorocarbon Plasma Etching of Silicon DioxideSurface Site Fractions .................................................................245

4-38 Fluorocarbon Plasma Etching of Silicon DioxideHighest ROP ................................................................................245

CK-TUT-10112-1112-UG-1 17 2011 Reaction Design

CHEMKIN List of Figures

2011 Reaction Design 18 CK-TUT-10112-1112-UG-1

 CHEMKIN

1 Introduction

The information provided in this manual describes both CHEMKIN and CHEMKIN-PRO.
Feature descriptions that apply only to CHEMKIN-PRO are highlighted with a
CHEMKIN-PRO Only tag in the heading and a icon in the margin.

The CHEMKIN software is designed for modeling many chemically reacting flow
configurations. This manual consists of tutorials that illustrate how to use the
CHEMKIN Reactor Models to address a variety of problems. The tutorials generally
represent realistic situations that might be encountered by practicing scientists or
engineers. They have been chosen to demonstrate the wide range of software
capabilities, and the different ways CHEMKIN can be used.

In this manual, we address three major categories of chemically-reacting flow

problems: Combustion problems are covered in Chapter 2, Catalysis in Chapter 3,
and Materials Processing in Chapter 4. Methods for analyzing chemical reaction
mechanisms are covered in Chapter 5. For use of the Reaction Path Analyzer, please
see the CHEMKIN Visualization Manual. In many cases, the same reactor models
are used for simulations in different categories of problems. Table 1-1 lists the
tutorials in this manual along with a cross-reference to the Reactor Models employed
in that tutorial.

Hint: Before working with the tutorials in this manual, we recommend that users first review
Getting Started with CHEMKIN Manual to become familiar with the operation of the
CHEMKIN Interface and the available Reactor Models.

CK-TUT-10112-1112-UG-1 19 2011 Reaction Design

CHEMKIN Chapter 1: Introduction

The files for the sample problems have been copied to a location specified by the user
during the CHEMKIN installation. On a Windows machine, the default location is
%userprofile%\chemkin\samples2010. On a UNIX machine, the default location is
$HOME/chemkin/samples2010. The sample-problem files have a uniform naming
convention based on the relevant Reactor Model name combined with a descriptor for
that specific system. For example, a project file called plasma_psr__chlorine.ckprj
uses the Plasma PSR Reactor Model, and involves chlorine chemistry. The data files
used for a given sample are located in subdirectories of the samples2010 directory.
For the example project mentioned above, the corresponding chemistry and profile
files (if used) would be located in the samples2010\plasma_psr\chlorine directory.
Sample problems involving multiple models or networks of reactors are treated as
separate groups.

In the interest of keeping computational times reasonable, the sample problems

generally do not include any of the larger chemical reaction mechanisms that are
often used in combustion and plasma research. Some of the mechanisms are
included only for the purposes of illustration, and should not be used for scientific
research or engineering projects without consulting the current scientific literature.

Table 1-1 Reactor Models Used in Sample Problems

Tutorial Reactor Models Used

2.1.1 Adiabatic Flame Temperature Equilibrium Calculation

2.3.1 Steady-state Gas-phase Combustion Perfectly Stirred Reactor

2.3.2 Autoignition for H2/Air Close Homogeneous Batch Reactor

2.3.3 Ignition-delay Times for Propane Autoignition Closed Homogeneous Batch Reactor

2.3.4 Burner-stabilized Flame Premixed Burner-stabilized Flame

2.3.5 NO Emission from High-pressure Flames with Premixed Burner-stabilized Flame

Gas Radiation [CHEMKIN-PRO Only] Radiation Model

2.3.6 Soot Formation in Radiating Opposed-flow Opposed Flow Flame

Diffusion Flame [CHEMKIN-PRO Only] Radiation Model

2.3.7 Flame Speed of Stoichiometric Methane/Air Premixed Flame-speed Calculation

Premixed Flame with Reaction Path Analyzer Reaction Path Analyzer
[CHEMKIN-PRO Only]

2.3.8 Parameter Study: Propane/Air Flame Speed as a Premixed Flame-speed Calculation

Function of Equivalence Ratio and Unburned Gas
Temperature

2.3.9 Hydrogen/Air Opposed-flow Flame Opposed-flow Flame

2.3.10 Flame Extinction Analysis [CHEMKIN-PRO Only] Extinction of Premixed or Opposed

Flow Flame

2011 Reaction Design 20 CK-TUT-10112-1112-UG-1

Chapter 1: Introduction Tutorials Manual

Table 1-1 Reactor Models Used in Sample Problems (Continued)

Tutorial Reactor Models Used

2.3.11 Stagnation Flame Analysis [CHEMKIN-PRO Only] Pre-mixed Stagnation Flame

2.4.1 Homogeneous Charge Compression Ignition Internal Combustion Engine

(HCCI) Engine

2.4.2 Multi-zone HCCI Engine Simulation Multi-Zone Engine Simulator

[CHEMKIN-PRO Only]

2.5.1 Shock-heated Air (Shock) Normal Incident Shock

2.6.1 Gas Turbine Network Perfectly Stirred Reactor

Plug Flow Reactor

2.6.2 Jet Flame Network Perfectly Stirred Reactor

2.6.3 Using Tear Streams to Estimate Initial Gas Internal Combustion Engine
Composition in an HCCI Engine with Exhaust Gas Perfectly Stirred Reactor
Recirculation (EGR) [CHEMKIN-PRO Only]

2.6.4 Partially Stirred Reactor for Methane/Air Partially Stirred Reactor

2.6.5 Side Inlet on a Plug Flow Reactor Plug Flow Reactor

2.6.6 Co-flowing Non-premixed CH4/Air Flame Cylindrical Shear Flow Reactor

2.7.1 Soot Formation and Growth in a JSR/PFR Jet-Stirred Reactor

Reactor [CHEMKIN-PRO Only] Plug Flow Reactor

2.7.2 Soot Particles in Flame Simulators Pre-mixed Flame-speed Calculation

[CHEMKIN-PRO Only] Opposed-flow Flame

2.7.3 Sectional Method for Particle-Size Distribution Pre-mixed Laminar Bumer-Stabilized

with Pre-mixed Laminar Burner-Stabilized Stagnation Flame
Stagnation Flame [CHEMKIN-PRO Only]

2.7.4 Simulating Particle-Size Distributions in a Burner- Pre-mixed Laminar Bumer-Stabilized

Stabilized Stagnation Flame Stagnation Flame

2.7.5 Detailed Particle Aggregation in a Batch Reactor Constant Pressure and Given
[CHEMKIN-PRO Only] Temperature Batch Reactor

2.8.1 Uncertainty Analysis of NOx Emissions Perfectly Stirred Reactor

[CHEMKIN-PRO Only]

3.1.1 Two-Stage Catalytic Combustor Perfectly Stirred Reactor

Plug Flow Reactor
Honeycomb Reactor

3.1.2 Engine Exhaust Aftertreatment with a Transient Perfectly Stirred Reactor

Inlet Flow

3.2.1 Parameter Study Facility for Surface Chemistry Honeycomb Catalytic Reactor
Analysis

4.1.1 Equilibrium Analysis of Chlorosilane CVD Equilibrium

4.1.2 PSR Analysis of Steady-state Thermal CVD Perfectly Stirred Reactor

CK-TUT-10112-1112-UG-1 21 2011 Reaction Design

CHEMKIN Chapter 1: Introduction

Table 1-1 Reactor Models Used in Sample Problems (Continued)

Tutorial Reactor Models Used

4.1.3 Approximations for a Cylindrical Channel Flow Perfectly Stirred Reactor

Plug Flow Reactor
Cylindrical Shear Flow

4.1.4 Deposition in a Rotating Disk Reactor Rotating Disk CVD Reactor

4.1.5 Trichlorosilane CVD in Planar Channel Flow Planar Shear Flow

Reactor

4.2.1 Time-dependent Simulations of ALD Process Perfectly Stirred Reactor

Stagnation Flow CVD Reactor

4.3.1 Steady-state Chlorine Plasma Plasma Perfectly Stirred Reactor

4.3.2 Spatial Chlorine Plasma PFR with Power Profile Plasma Plug Flow Reactor

4.3.3 Fluorocarbon Plasma Etching of Silicon Dioxide Plasma Perfectly Stirred Reactor

5.1.2 Reaction Mechanism for Diamond CVD Mechanism Analyzer

2011 Reaction Design 22 CK-TUT-10112-1112-UG-1

 CHEMKIN

2 Combustion in Gas-phase Processes

2.1 Equilibrium
2.1.1 Adiabatic Flame Temperature

2.1.1.1 Project Description

This user tutorial presents the use of a gas-phase equilibrium calculation to determine
the adiabatic flame temperature for the hydrogen/air system. The adiabatic flame
temperature is a measure of the maximum temperature that could be reached by
combusting a particular gas mixture under a specific set of conditions. In a real
system which includes heat losses, chemical kinetic and/or mass transport limitations,
the flame temperature is likely to be lower than the adiabatic flame temperature.

2.1.1.2 Project Setup

The project file is called equilibrium__gas.ckprj. The data files used for this sample
are located in the samples2010\equilibrium\gas directory. This reactor diagram
consists of a single equilibrium reactor.

The equilibrium calculation only needs a list of species with their thermodynamic data.
There is no need for a reaction list. For this sample problem, the chemistry input file
includes only 3 elements: H, O and N; and 9 species: H2, H, O2, O, OH, HO2, H2O,
N2, and H2O2. It is important to include all likely radical species as well as stable
species in the product list so as to obtain an accurate flame temperature
prediction.For equilibrium calculations generally, it is better to include many
unimportant species than to leave out species that may turn out to be important.

CK-TUT-10112-1112-UG-1 23 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Setting up this problem first involves the C1_Equilibrium panel. The problem type
(constant pressure and enthalpy), initial temperature (300 K) and pressure (1 atm) are
entered on the Reactor Physical Properties tab. An estimated solution temperature of
2000 K is used to help ensure that the solution obtained is for an ignited gas rather
than the unburned state. The presence of an estimated solution temperature is often
unnecessary for equilibrium simulations but required when a trivial secondary solution
may exist. The starting composition is entered on the Reactant sub-tab of the
Reactant Species tab. The reactant mixture defines the initial state, which provides
the initial moles of chemical species and the initial energy of the system. The
Continuations panel is used to specify two additional simulations with increasing initial
temperatures.

2.1.1.3 Project Results

Figure 2-1 shows the equilibrium temperatures from these simulations, which
represent the adiabatic flame temperatures for a hydrogen/air mixture with a H/O ratio
of 2.0. The temperatures are on the order of ~2400 K, and thus clearly correspond to
the combusted gas. As expected, these adiabatic flame temperatures increase with
increasing initial gas temperature.

Figure 2-1 Adiabatic Flame TemperaturesHydrogen/Air Mixture

2.2 Using Equivalence Ratio

Many properties of combustion processes strongly depend on the stoichiometry of the
combustion mixture. The parameter used most frequently to describe the
stoichiometry of the mixture is the equivalence ratio, . CHEMKIN gives the user an
option of describing the initial mixture composition by either of two methods:

Listing all of the reactant species and their respective mole or mass fractions.

2011 Reaction Design 24 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Alternately, by identifying fuel, oxidizer and complete-combustion

(stoichiometric) product species, the mole fraction of any additional species
(such as an inert like Ar or N2) and the equivalence ratio.

Examples of the use of the equivalence ratio can be found in Section 2.3.1, Steady-
state Gas-phase Combustion, and Section 2.3.8, Parameter Study: Propane/Air
Flame Speed as a Function of Equivalence Ratio and Unburned Gas Temperature.

2.2.1 Example: Propane in Air

The stoichiometric equation for propane combustion in air is:

C3H8 + 5(O2 + 3.76N2) 3CO2 + 4H2O + 5 * 3.76N2

Here, the equivalence ratio, , is defined as the ratio of the actual fuel/oxidizer ratio to
the fuel/oxidizer ratio in the stoichiometric equation, as follows:
XC3 H8 X O2
= -----------------------------------------
- = 5 X C3 H8 XO2
X C3 H8 X O2 stoich
CHEMKIN is able to interpret all of the input formats as illustrated in Tables 2-1 to 2-4.
Tables 2-1 and 2-2 illustrate the use of mole fractions to specify mixture stoichiometry
and tables 2-3 and 2-4 illustrate the use of equivalence ratio.

Table 2-1 Reactant Mole Fractions Sum to 1.0

Name Mole Fraction

Reactant C 3H 8 0.04

Reactant O2 0.202

Reactant N2 0.758

Table 2-2 Relative Moles Normalized So Mole Fractions Sum to 1.0

Name Mole Fraction

Reactant C 3H 8 0.2

Reactant O2 1

Reactant N2 3.76

CK-TUT-10112-1112-UG-1 25 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Note that in Table 2-2, the relative moles given will be normalized such that the total
mole fractions sum to 1.0. However, it is often more convenient to enter relative
moles, as described in Tables 2-3 and 2-4, using the equivalence ratio option.

Table 2-3 N2 as Added Diluent

Equivalence ratio = 1

Name Mole Fraction

Fuel C 3H 8 1.0

Oxidizer O2 1.0

Added Species N2 0.758

Complete-combustion Product H 2O

Complete-combustion Product CO2

Table 2-4 N2 as Component of Oxidizer

Equivalence ratio = 1

Name Mole Fraction

Fuel C 3H 8 1.0

Oxidizer O2 1.0

Oxidizer N2 3.76

Complete-combustion Product H2O

Complete-combustion Product CO2

Complete-combustion Product N2

Note that while using equivalence ratio (Tables 2-3 and 2-4), N2 can be considered as
either an added diluent (Table 2-3), or a component of the oxidizer, i.e., air
(Table 2-4). If it is considered part of the oxidizer then it must be included in the
complete-combustion products list. Also, the oxidizer composition can be given in
relative moles or in mole fractions that sum to one (relative moles will subsequently be
normalized to sum to 1.0). If using the equivalence ratio input format (Tables 2-3 and
2-4), the mole fraction of fuel is relative to the total number of moles of fuel and the
mole fraction of oxidizer is relative to the total number of moles of oxidizer; they will
equal to unity unless more than one species is given for each field of fuel and/or
oxidizer. However, the added species mole fractions are relative to the total number
of moles of the reactants, thus the total mole fraction of added species should not sum
up to unity or higher.

2011 Reaction Design 26 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

In general, the equivalence ratio is a good parameter for quantification of mixture

stoichiometry for combustion of hydrocarbon fuels (composed only of hydrogen and
carbon atoms) in air or in oxygen. However, one has to be careful when applying to
describe fuel stoichiometry for oxygenates (oxygen is chemically bound to the fuel
molecule) as well as other non-traditional fuels and oxidizers. Because the definition
of the equivalence ratio does not properly account for the oxygen that might be
chemically bound in the fuel, it might not be a useful parameter for some fuels.1

When trying to determine the Complete-combustion or Stoichiometric Products for

such fuels, one must remember that a species can be a saturated Stoichiometric
Product if and only if the valence orbitals of all of its constituent atoms are filled. That
is, if the oxidation numbers of all of the atoms in a given product species sum up to
zero.

2.2.2 Example: Stoichiometric Products

Table 2-5 Determining Stoichiometric Products

Element Oxidation number

C +4

H +1

O -2

N 0

Ar 0

The oxidation numbers of the atoms in CO2 sum up to (+4) + 2 * (-2) = 0 and in H2O
they sum up to (-2) + 2 * (+1) = 0 as well, thus CO2 and H2O are Stoichiometric
Products. However, oxidation numbers of the atoms in O2 do not sum up to zero
(-2 + -2 = -4), thus O2 can not be entered as a Stoichiometric Product. 2,3

1. C. J. Mueller, M. P. B. Musculus, L. M. Pickett, W. J. Pitz, C. K. Westbrook. The Ox-

ygen Ratio: A Fuel-Independent Measure of Mixture Stoichiometry, 30th International
Symposium on Combustion (2004).
2. D. W. Oxtoby, H. P. Gillis, and N. H. Nachtrieb, Principles of Modern Chemistry,
Thompson Learning, Inc. (2002).
3. S. S. Zumdahl, Chemistry. D.C. Heath and Company, Lexington, Massachusetts
(1989).

CK-TUT-10112-1112-UG-1 27 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.3 Ignition, Flames and Flammability

2.3.1 Steady-state Gas-phase Combustion

2.3.1.1 Project Description

This user tutorial presents the simulation of the steady-state combustion of a mixture
of hydrogen, nitrogen, and oxygen in a perfectly-stirred reactor at atmospheric
pressure. This project uses the chemistry set for hydrogen combustion, described in
Section 2.9.1. This project demonstrates the use of equivalence ratio for specifying
the starting gas mixture, as well as the use of a continuation to alter the equivalence
ratio.

2.3.1.2 Project Setup

The project file is called psr__gas.ckprj. The data files used for this sample are
located in the samples2010\psr\gas directory. This reactor diagram contains one
gas inlet, one perfectly stirred reactor model, and an outlet.

Many of the important parameters for this simulation are input on the C1_PSR panel.
On the Reactor Physical Properties tab, the Problem Type is first set to Solve Gas
Energy Equation, and the Steady State Solver is chosen. The residence time of the
gas in the PSR (0.03 milliseconds), the estimated gas temperature (1700 K), system
pressure (1 atm) and volume (67.4 cm3) are also set on this panel. No value is input
for the Heat Loss, so the system will be treated as adiabatic. The Species-specific
Properties tab provides an input field for an estimate of the gas composition to help
the solver converge on a solution. In the current case, no solution estimates are
supplied for the species fractions, so an equilibrium calculation will be performed at
1700 K with the reactant mixture to determine the initial estimates for them. These
initial estimates are used as the starting point for the iterations that converge to the
steady-state conditions.

Parameters pertaining to the incoming gas are input on the C1_Inlet1 panel. In this
case, the Stream Property Data tab only has the value of 298 K for the inlet
temperature. The gas residence time and reactor volume are input on the Reactor
Physical Properties tab of the C1_PSR panel, such that including a flow rate here
would over-specify the problem. The Equivalence Ratio box is checked at the top of
the Species-specific Properties tab of the C1_Inlet1 panel, and a value for the fuel/air
equivalence ratio of 1.0 is supplied. As defined on the Fuel Mixture sub-tab, the fuel is
composed of 80% H2 and 20% N2. Likewise, on the Oxidizer Mixture sub-tab, the
oxidizer is defined as 79% N2 and 21% O2 (air). Use of the equivalence-ratio form of
input requires that the user specify the products of complete combustion for the Fuel

2011 Reaction Design 28 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

and Oxidizer specified. This is defined on the Complete-Combustion Products sub-

tab. In this case, the product species are H2O and N2, where N2 is included because
it is part of the fuel and oxidizer mixtures. Note that all of the elements contained in
the fuel and oxidizer species must also appear in the product species.

For this example, all inputs except Relative Tolerance are left at default on the Basic
tab of the Solver panel. We use the default settings, in which the application first
solves a fixed-temperature problem and then uses the results of this solution as the
initial guess to solve the full problem including the energy equation. On the Advanced
tab of the Solver panel, the minimum bonds of species fractions have been specified
as smaller than the default, to aid in convergence to a physical solution.

On the Continuations panel, 8 additional simulations are specified where the

equivalence ratio is gradually changed. There are no inputs on the Output Control
panels for this problem, since default output options will be used.

2.3.1.3 Project Results

Figure 2-2 shows the steady-state temperatures for the combusting
hydrogen/air/nitrogen mixture. In this case, the temperature peaks at a fuel/air
equivalence ratio of about 1.20. As shown by the molar conversions in Figure 2-3,
neither the fuel nor the oxidizer is completely consumed in this combustor, as a result
of the PSR residence time.

To get the plot in Figure 2-3, be sure to select the molar_conversion in the Select Results
panel when the Post-Processor is first launched.

Figure 2-2 Steady-state Gas-phase CombustionHydrogen/Air Temperatures

CK-TUT-10112-1112-UG-1 29 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-3 Steady-state Gas-phase CombustionMolar Conversions

2.3.2 Autoignition for H2/Air

2.3.2.1 Project Description

This user tutorial presents a transient simulation of the spontaneous ignition of a
stoichiometric hydrogen/air mixture at constant pressure. This project uses the
chemistry set for hydrogen combustion described in Section 2.9.1. Here we are
interested in determining the ignition time under a specified set of initial pressure and
temperature conditions, assuming no heat loss to the environment (adiabatic
conditions). In addition, we would like to determine which reactions contribute most to
the CHEMKIN results, using sensitivity analysis. The system is a closed or batch
process, so there is no flow of mass into or out of the reactor.

2.3.2.2 Project Setup

The project file is called closed_homogeneous__transient.ckprj. The data files
used for this sample are located in the
samples2010\closed_homogeneous\transient directory. This reactor diagram
contains only one closed homogeneous reactor.

The Closed_Homogeneous (C1) group of panels become active after running the Pre-
Processing step. On the Reactor Physical Properties tab of the C1_Closed
Homogeneous panel, first the problem type is selected as Constrain Pressure And
Solve Energy Equation (the default). The end time of the simulation is set to 0.0002
sec. The initial temperature (1000 K) is then input, along with the pressure (1 atm). A
volume is not specified as it is not important for the results of this simulation, so the
default value of 1 cm3 will be used for the initial volume. Since this is a closed
homogeneous system, the results in terms of species fraction and temperature will be
the same, regardless of the volume value. If surface chemistry were included, the

2011 Reaction Design 30 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

volume-to-surface ratio would be important, but in this case it is gas only. On the
Reactant Species sub-tab, the starting gas mixture is given in relative moles as 2.0
H2, 1.0 O2, and 3.76 N2. Writing it this way makes it easy to see that the O2/N2 ratio
matches the composition of air, and that the fuel/air ratio is stoichiometric (two H2 per
one O2). The Normalize button will set the mole fractions so that they sum to one,
although this is optional, since the normalization will also occur automatically within
the program.

The text output file lists a value for the ignition time near the end of the file. The
criterion for Ignition Delay is specified using a Temperature Delta of 400 K in the
Ignition Delay sub-panel of the Output Control panel. This means that ignition will be
registered when the temperature reaches a value of 400 K above the initial
temperature. This panel shows other possible definitions of ignition time.

On the Output Control panel, the checkbox for All A-factor Sensitivity has been
marked. This results in sensitivities being calculated for all species and all reactions,
saved in the XML Solution File, and printed in the output file. This option should be
used with care, as the computation time and solution-file sizes increase as a higher
power of the size of the reaction mechanism. Except in the case of a very small
reaction mechanism, such as the one being used in the sample problem, it is better to
use the Species Sensitivity and ROP panel to request that these quantities be output
only for a few species of highest interest. Including a value for the Threshold for
Species Sensitivity that is higher than the default value of 0.001 will also help keep the
amount of information to a manageable level.

There are no Continuations used for this problem.

2.3.2.3 Project Results

Figure 2-4 shows the temperature profile as a function of time for this problem. At the
end of this simulation, the temperature is still rising; if it is run much longer, the
temperature increases another ~300 K, nearing the adiabatic flame temperature.
Although not shown, the volume in this constant-pressure system shows a
corresponding increase at ignition. The text output file lists a value for the ignition time
of 1.7263E-04 sec, where ignition is defined as the time at which the gas reaches a
temperature of 1400 K. Figure 2-5 shows a close-up of species mole fractions as a
function of time. Note that zooming in on the x-axis shows the expected increase in
radical species and the products at ignition, along with a decrease in hydrogen and
oxygen reactants.

CK-TUT-10112-1112-UG-1 31 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-4 Autoignition for Hydrogen/AirTemperature Profile

Figure 2-5 Autoignition for Hydrogen/AirSpecies Composition Profiles

Figure 2-6 shows normalized sensitivity coefficients as a function of time for the four
reactions that have the largest effect on the gas temperature. As one might expect,
the largest sensitivity occurs near the time of ignition, when the most rapid change in
temperature is taking place. The results also show that the dominant reaction for
determining the temperature during ignition is the exothermic radical-recombination
reaction #11: O + OH O2 + H. The sensitivity coefficient for this reaction is
positive, indicating that increasing the rate of this reaction will lead to a higher
temperature (more heat production). In contrast, the sensitivity coefficient for
reaction #1 is large and negative, indicating that increasing the rate of this reaction
will lead to a lower temperature (less heat production).

2011 Reaction Design 32 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-6 Autoignition for Hydrogen/AirSensitivity Coefficients

2.3.3 Ignition-delay Times for Propane Autoignition

2.3.3.1 Project Description

This user tutorial presents an ignition-delay time calculation for the homogeneous,
isobaric, adiabatic combustion of propane in air. The ignition times are computed
using two distinct definitions. Discrepancies between results are presented.

Premixed combustion technology is well established in the gas turbine industry. One
of the major concerns of such technology, however, is avoiding the autoignition
phenomenon to protect combustor components as well as to limit levels of pollutant
emissions. Numerical predictions of the ignition time can be very useful in
understanding the autoignition parameters, which can be important for the automotive
and turbine industries. Chemical kineticists can also benefit from predicting ignition
times. Work in validation and testing of detailed mechanisms as well as reduction of
detailed mechanisms often requires analyzing ignition times. One of the better-known
validation techniques for detailed chemical kinetic mechanisms consists of comparing
computational predictions of the ignition-delay times to shock-tube experiments.4, 5, 6
Such comparisons can provide a good understanding of the underlying chemistry,
since the 0-D computations are free from transport effects.

4. Reaction and Ignition Delay Times in Oxidation of Propane, B.F. Mayers and E.R.
Bartle, AIAA Journal, V.7, No10, p.1862
5. Validation of Detailed Reaction Mechanisms for Detonation Simulation, E. Schultz
and J. Sheperd, Explosion Dynamics Laboratory Report FM99-5, California Institute of
Technology, Pasadena, CA, February 8,2000
6. A Small Detailed Chemical-Kinetic Mechanism for Hydrocarbon Combustion, M.V.
Petrova and F.A. Williams, Combustion and Flame, Volume 144, Issue 3, February 2006, p.
526

CK-TUT-10112-1112-UG-1 33 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

There are various ways of defining the ignition time, experimentally as well as
computationally, for combustion applications. For example, it is often defined as the
time at which either the maximum or onset of certain species concentrations is
reached, the time at which a specified rate of increase of temperature occurs, the time
at which luminous radiant output from the system is first observed, etc. The reported
experimental data can vary greatly, depending on which definition was used in the
experiments5, 7. Thus, it is often useful to select which ignition-delay time definition
should be used in numerical computations. CHEMKIN allows the user such flexibility.
For example, in CHEMKIN's closed homogeneous batch reactor, the ignition time can
be defined to be the time during which the maximum amount of heat is released
during a combustion process (as indicated by the inflection point in the temperature
profile), as well as the time corresponding to the maximum of a certain species
concentration chosen by the user. CHEMKIN further allows users to input a specific
definition of the ignition time via the Ignition Criterion User Routine.

2.3.3.2 Project Setup

The project file is called closed_homogeneous__ignition_delay.ckprj. The data
files used for this sample are located in the
samples2010\closed_homogeneous\transient\ignition_delay directory. This
sample uses the mechanism and thermodynamic data from the University of
California, San Diego6, 8. This reactor diagram contains only one closed
homogeneous reactor.

Open the project file. After running the Pre-Processor step, on the Reactor Physical
Properties tab of the C1_Closed Homogeneous panel, the problem type is selected
as Constrain Pressure And Solve Energy Equation (the default). The initial
temperature (1200 K) is then input, along with the pressure (1 atm). A volume is not
specified as it is not important for the results of this simulation, so the default value of
1 cm3 will be used for the initial volume. Since this is an isobaric closed homogeneous
system, the results in terms of species fraction and temperature will be the same,
regardless of the volume value. If surface chemistry were included, the volume-to-
surface ratio would be important, but in this case only gas-phase chemistry is present.
On the Reactant Species sub-tab, the starting gas mixture is given as 0.02 C3H8,
0.05 O2, and 0.93 Ar. These rather dilute conditions are representative of shock-tube
experimental conditions.

7. Combustion Theory, Second Edition, F.A. Williams, Addison-Wesley Publishing Com-

pany, The Advanced Book Program, Redwood City, CA, 1985.
8. http://www-mae.ucsd.edu/~combustion/cermech/

2011 Reaction Design 34 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

The end time is specified on the Reactor Physical Properties panel. It is important to
check the resulting output file after the run is complete to make sure that the End
Time is large enough to allow for ignition to occur. If no ignition time is provided at the
end of the output file, ignition has not yet occurred and the End Time of the simulation
should be adjusted.

On the Output Control tab, on the Ignition Delay sub-tab, the Temperature Inflection
Point box is checked as well as Species Maximum Fraction, for the computational
ignition time criteria. Species Maximum Fraction is set to the OH species. This means
the ignition time will be computed based on the maximum of the OH concentration.
The user can choose any species from the pull-down menu. The ignition time will also
be computed at the point where the rate of change of temperature with respect to time
is the largest (Temperature Inflection Point criteria). The user can choose any or all of
the definitions of ignition times provided by CHEMKIN.

It is of interest to run the same problem, but vary the initial temperature in order to
demonstrate ignition time dependence on temperature as well as its dependence on
the chosen ignition time criteria. For this purpose CHEMKIN's Parameter Study
Facility is used. The initial Temperature is changed over a range of 1200 - 2600 K.
Please refer to the CHEMKIN Advanced Analyses Manual for guidance in setting up a
Parameter Study.

Once the project is run, the ignition times (based on peak OH concentration and
temperature inflection point) for the three runs are printed in the text output files and
also stored in the solution files, stored in the
closed_homogeneous__ignition_delay_<date>_<time> folder, contained in your
working directory. You can look at the output files by clicking on Click to View
Results under Run Calculations, using the Display Detail option.

2.3.3.3 Project Results

On the Run Calculations panel, click on Display Detail and then on Click to View
Results to view the output files corresponding to each temperature. Table 2-6 lists the
initial temperatures as well as ignition times based on temperature inflection and OH
concentration maximum for some of the runs obtained from the bottom of the output
files. The ignition times based on the two criteria differ more and more with increasing
initial temperature. Looking at temperature and OH mole fraction profiles will give the
user better insight as to how the ignition times are obtained.

CK-TUT-10112-1112-UG-1 35 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Table 2-6 Initial Temperatures and Ignition Times

Ignition time [sec]

T [K] T inflection OH max

1200 1.69E-02 1.71E-02

1600 1.85E-04 2.11E-04

2600 1.95E-06 9.03E-06

Figure 2-7 shows OH mole fraction and temperature profiles as a function of time, for
the 1200-K case, obtained with CHEMKIN's Graphical Post-Processor. The solid line
is the temperature profile and the dashed line is the OH mole fraction. There is an
obvious spike in OH mole fraction at the time of ignition, which in this lower-
temperature case corresponds very nicely to the maximum rate of change of the
temperature (the inflection point). This is also a good example of how one should be
careful when interpreting numerical results for ignition times. In this example,
CHEMKIN produces two numbers for ignition times for the OH mole fraction peak.
Only one of these numbers represents the ignition time.

Figure 2-7 Temperature and OH Mole Fraction Profiles as a Function of Time

In Figure 2-8, there is an obvious difference (about a factor of four) between the
ignition time values as defined by the two chosen criteria. This plot demonstrates why
one should be mindful of ignition time definitions used in computations. The solid line
is the temperature profile and the dashed line is the OH mole fraction.

2011 Reaction Design 36 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-8 Ignition Based on Varying Criteria

Finally the ignition times vs. the inverse of temperature, are shown in Figure 2-9 on a
semi-log plot. Here the difference between the two ignition time definitions is obvious.

Figure 2-9 Ignition Times vs. Inverse of Temperature (Semi-log)

CK-TUT-10112-1112-UG-1 37 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.3.4 Burner-stabilized Flame

2.3.4.1 Project Description

This user tutorial presents a simulation of a burner-stabilized laminar premixed flame
of hydrogen and oxygen at low pressure. This project uses the chemistry set for
hydrogen combustion described in Section 2.9.1. Burner-stabilized laminar premixed
flames are often used to study chemical kinetics in a combustion environment. Such
flames are effectively one-dimensional and can be made very steady, facilitating
detailed experimental measurements of temperature and species profiles. Also,
laminar flame speed is often used to characterize the combustion of various fuel-
oxidizer combinations. Therefore, the ability to model chemical kinetics and transport
processes in these flames is critical to interpreting flame experiments and to
understanding the combustion process itself. Examples of the use of flame modeling
to interpret experimental observations and to verify combustion chemistry and
pollution formation can be found in Miller, et al.9

2.3.4.2 Project Setup

The project file is called pre-mixed_burner__burner_stabilized.ckprj. The data files
used for this sample are located in the
samples2010\pre-mixed_burner\burner_stabilized directory. This reactor model is
simple and contains a gas inlet, a premixed-burner reactor, and an outlet.

On the Reactor Physical Properties tab of the C1_ Pre-Mixed Burner panel, the
problem type is selected as Fix Gas Temperature because a measured temperature
profile is used rather than computing the gas temperatures from the energy equation.
For these laminar flames, the gas temperatures are often obtained from experiment
rather than by solving an energy conservation equation. This is because there can be
significant heat losses to the external environment, which are unknown or difficult to
model. For cases where the heat losses are known or negligible, the user can solve a
burner-stabilized flame problem in which the temperatures are determined from the
energy conservation equation. Even if the energy equation is to be solved for the
temperatures, the iteration converges more reliably if the species profiles are first
computed using a fixed temperature profile. In any case, the user needs to input an
estimate of the temperature profile. For this example, a temperature profile called
pre-mixed_burner__burner_stabilized_TPRO.ckprf is input on the Reactor Physical

9. J. A. Miller, R. E. Mitchell, M. D. Smooke, and R. J. Kee, in Proceedings of the Nineteenth

Symposium (International) on Combustion, The Combustion Institute, Pittsburgh, Pennsylva-
nia, 1982, p. 181. J. A. Miller, M. D. Smooke, R. M. Green, and R. J. Kee, Combustion Science
and Technology 34:149 (1983). J. A. Miller, M. C. Branch, W. J. McLean, D. W. Chandler, M.
D. Smooke, and R. J. Kee, in Proceedings of the Twentieth Symposium (International) on Com-
bustion, The Combustion Institute, Pittsburgh, Pennsylvania, 1985, p. 673.

2011 Reaction Design 38 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Properties tab and only the species transport equations are solved using the
temperature as a constraint. The system pressure (25 Torr) is also input on this panel,
along with the choice of Mixture-averaged Transport and the use of the Correction
Velocity Formalism.

The Ending Axial Position (10 cm) for the simulation is input on the Grid Properties
tab of the C1_ Pre-Mixed Burner panel, along with a number of parameters
concerning the gridding of the problem. A few of the grid parameters have been
changed from the default values. The pre-mixed burner reactor model has adaptive
gridding. The initial simulations are therefore done on a very coarse mesh that may
have as few as five or six points. After obtaining a solution on the coarse mesh, new
mesh points are added in regions where the solution or its gradients change rapidly.
The initial guess for the solution on the finer mesh is obtained by interpolating the
coarse mesh solution. This procedure continues until no new mesh points are needed
to resolve the solution to the degree specified by the user.

The simulation also needs a starting estimate from which to begin its iteration. This
estimate is given in terms of a reaction zone in which the reactants change from their
unreacted values (the unburned composition) to the products. Intermediate species
are assumed to have a Gaussian profile that peaks in the center of the reaction zone
with the width such that the profile is at 1/10 of its peak value at the edges of the
reaction zone. The user provides estimates for the location and thickness of this
reaction zone on the Initial Grid Properties tab of the C1_ Pre-Mixed Burner panel.

Starting estimates for the gas composition in various parts of the flame are input on
the Species-specific Properties tab of the C1_ Pre-Mixed Burner panel. The species
on the Intermediate Fraction tab are generally short-lived radical species, or species
that are expected to be present throughout the flame. The species on the Product
Fraction sub-tab are those expected to be present in the fully burned state. If no
product species estimates are given, an equilibrium composition will be used for the
product estimate. Within the reaction zone the model uses straight lines between the
initial and final values for both the reactants and products. On the hot side, the
product species are flat at the estimated product values. Note that any given species
can be both a reactant and a product species. For example, the nitrogen in an air
flame will be both a reactant and a product, while the excess fuel in a rich flame will
also be both a reactant and a product.

CK-TUT-10112-1112-UG-1 39 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Parameters pertaining to the incoming gas are input on the C1_Inlet1 panel. The input
mass flow rate (4.6 mg cm-2sec-1), which corresponds to an experimental value, is
the only input on the Stream Properties Data tab. The composition of the fuel-rich
input gas (28% H2, 9% O2 and 63% Ar) is input on the Species-specific Data tab of
the C1_Inlet1 panel. The Solver panel has some inputs where the default settings are
being replaced. There are no additional inputs on the Output Control panel and no
Continuations are used for this project.

2.3.4.3 Project Results

Figure 2-10 shows the experimental gas temperature profile as a function of distance
imposed on the simulation. The gas composition in Figure 2-11 exhibits the expected
behavior of the primary combustion species as a function of distance above the
burner, with most of the oxygen reacting away within the first 2 cm. At the larger
distances, some of the H atoms recombine to form molecular hydrogen. This would
not have been significant in a fuel-poor flame, but does occur in this fuel-rich situation.
An inspection of the text file shows that the simulation now has 33 grid points, a
significant increase from the initial six. The grid is more dense at the lower distances,
as needed to resolve the more rapid changes in chemical composition and
temperature occurring in that region.

Figure 2-10 Burner-stabilized FlameExperimental Gas Temperature Profile

2011 Reaction Design 40 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-11 Burner-stabilized FlameMole Fractions

2.3.5 NO Emission from High-pressure Flames with Gas

Radiation [CHEMKIN-PRO Only]

2.3.5.1 Project Description

This tutorial presents a simulation of a burner-stabilized, lean premixed
methane/oxygen/nitrogen flame at high pressure. This project uses the chemistry set
for the C2_NOx mechanism described in Section 2.9.6. C2_NOx Mechanism. The
flame configuration represented in this sample is frequently featured in studies of
premixed flame structure and chemistry, because of its simple geometry. Simulations
are typically carried out with a fixed temperature profile obtained from the
experiments, which minimizes the impact of temperature uncertainties on reaction
rates. When a reliable temperature profile is not available experimentally, however,
we have to resort to solving the energy equation, either by neglecting heat loss or by
estimating the heat loss from the flame to the surroundings. The radiation-model
option is a good way to include the first-order effect of heat loss, without having to
make extra measurements. This project uses the radiation heat-transfer model option
to estimate the radiation heat loss to the environment from gas products in the flame
and post-flame region. The optically-thin limit assumed by the radiation model for both
gas and dispersed phases is described in the CHEMKIN Theory Manual. This project
assumes that CO2, H2O, CO, and CH4 are the only gas species emitting thermal

CK-TUT-10112-1112-UG-1 41 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

radiation. Emissivity of individual gas species is expressed as a temperature

polynomial and is provided as an optional input of the thermodynamic data. . For this
project, the emissivity information is already included in the C2_NOx chemistry set
installed with CHEMKIN.

The premixed radiating flames modeled in this project are from the CH4/O2/N2 flames
studied by Thomsen et al.10. The equivalence ratio of the unburned mixture is 0.6,
with an N2-O2 ratio of 2.2 (this is more oxygen-rich relative to air, where the molar N2-
O2 ratio of air is 3.76). The pressure of the experiments is 14.6 atm. The goal of this
project is to demonstrate the impact of gas radiation on temperature prediction at high
pressure and, consequently, on the NO emission predictions.

2.3.5.2 Project Setup

The project file is called pre-mixed_burner_GasRadiation.ckprj. The data files used
for this sample are located in the samples2010\pre-mixed_burner\GasRadiation
directory. This sample project consists of a gas inlet, a premixed-burner reactor, and
an outlet. Since this is a burner-stabilized flame, the project setup follows similar steps
as the burner-stabilized flame project described in Section 2.3.4; please refer to this
sample for more information about general set-up instructions. For this tutorial, only
the key inputs that are different for this project, i.e., for specifying gas radiation, are
discussed.

On the Reactor Physical Properties tab of the C1_Pre-Mixed Burner panel, the
problem type is selected as Solve Gas Energy Equation because the gas
temperature will be computed from the energy equation with radiation heat loss from
the gas mixture. This project utilizes the feature for automatic estimation of the
temperature profile, such that user input of an estimated temperature profile is not
necessary. The unburned gas temperature, however, is a required input and it is set to
360 K for all flames simulated here. Pre-processing of the chemistry set determines
whether thermal absorption coefficients are available for any of the gas-phase
species in the mechanism. When such coefficients are found during pre-processing,
the Include Gas Radiation option will appear on the panel and will be activated by
default. The Ambient Temperature needed by the radiation heat loss calculation is
estimated to be 1500 K For this experiment, the flame is enclosed in a pressure
chamber and there is no active cooling of the wall. For this reason, the temperature
to which the flame radiates should be much higher than typical ambient temperatures
of open flames. That estimation was made by matching the experimental temperature
gradient of an atmospheric flame. If no ambient temperature is given, the default

10. D. D. Thomsen, F. F. Kuligowski, and N. M. Laurendeau, Combustion and Flame

119:307-318 (1999).

2011 Reaction Design 42 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

value is 298 K. The gas radiation calculation can be turned off entirely by clearing the
check mark in the Include Gas Radiation box. It is also possible to exclude certain
gas species from the thermal radiation calculation. All gas species for which
absorption coefficient data are available in the thermodynamic data file are listed
under the Radiating Gas-Phase Species tab on the Species-specific Properties panel.
By deselecting a species on the list, its contribution to radiation heat loss will be
omitted.

The Pre-mixed Burner-Stabilized Flame model enforces a zero-gradient condition for

all variables at the hot boundary by default. While the zero-gradient boundary
condition is appropriate for adiabatic flames and for prediction species participating
only in the combustion process (i.e., not in emissions formation), it can lead to
uncertainties for simulations involving heat loss and slow-forming pollutant species,
such as NO. Fortunately, the post-flame region of a burner-stabilized premixed flame
behaves almost like a plug flow. In this way, any inaccuracy at the downstream hot
boundary is unlikely to propagate very far upstream. Since measurement is usually
taken in the vicinity of the flame, the impact of inaccurate boundary treatment can be
minimized by increasing the size of the computational domain, moving the boundary
constraint away from the flame region. In the experimental facility modeled, the are
located within 1 cm from the burner surface. A series of test simulations with various
end points indicates that a computational domain greater than 3 cm is sufficient to
eliminate impact of the boundary constraint on solution profiles within 1 cm from the
burner surface. An ending axial position of 10 cm is used in the simulations and this
information is entered on the Basic tab under the Grid Properties panel.n and this
information is entered on the Basic tab under the Grid Properties panel.

2.3.5.3 Project Results

The experimental NO concentration is reported in the units of ppm @ 15% O2 wet,
which is not a readily available quantity in the CHEMKIN Post-processor. To convert
the raw PPM value to these units, the solution must be exported from the Post-
processor to a text file so that an external application, such as Excel, could be used to
convert NO mole fraction to the correct unit.

The NO concentration corrected to 15% O2 wet-basis, XNO,15%O2wet, can be computed

from NO mole fraction, using the formulation given by Eq. (15.8) in Turns combustion
textbook11:

N mix
X NO ,15%O2 wet X NO
N mix ,15%O2 wet

CK-TUT-10112-1112-UG-1 43 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

XNO is the predicted NO mole fraction. Nmix and Nmix,15%O2wet, respectively, are the
total number of moles in original and corrected mixtures and are defined by
Eq. (15.9 a) as11 (p. 44)

2 X O2
N mix 1 3.2
1 3.2 X O2

and

2 0.15
N mix ,15%O2 wet 1 3.2 14.23
1 3.2 0.15
with x = 1 and y = 4 for CH4. Note that the constant 4.76 in Turns formulation is
changed to 3.2 in the above equations because the N2-O2 ratio in this case is 2.2
instead of 3.76 for air.

Figure 2-12 compares gas temperature predictions with and without radiation heat
loss to experimental data near the burner surface. The two temperature profiles
obtained by the models stay very close to each other and only start to deviate at the
flame zone where major radiating species such as CO2 and H2O are formed by
combustion. The temperature predicted by the adiabatic model, although showing
good agreement with the data, remains almost constant behind the flame. In contrast,
both experiment and radiation model indicate that the gas temperature decreases
gradually in the post-flame region. The axial temperature profile obtained by the
radiation model also agrees well with the measurement, but slightly over-predict the
heat loss rate, such that the predicted temperature drops faster than observed with
the experimental data.

Measured and predicted profiles of NO concentration corrected to 15% O2 wet are

presented in Figure 2-13. The adiabatic model (neglecting radiation) significantly
over-predicts NO formation rate and NO level in the post-flame region. Without any
temperature decrease in the post-flame region, the adiabatic model finds the NO level
increasing constantly and fails to capture the eventual slowdown of NO formation
indicated by the experimental profile. The radiation model, predicting a cooler
temperature behind the flame, yields an NO profile that is in excellent agreement with
the measurement.

11. S. R. Turns, An Introduction to Combustion: Concepts and Applications, 2nd Ed., Mc-
Graw-Hill, New York, p. 556, 2000.

2011 Reaction Design 44 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

The substantial difference in NO solution profiles between the adiabatic and the
radiation models suggests that thermal radiation in the post-flame region is critical for
predicting accurate NO emission from high pressure premixed flames. A follow-up
rate-of-production or sensitivity analysis can reveal the NO formation pathways that
are affected by the cooler post-flame temperature due to gas radiation.

This project can be easily expanded to study the pressure effect on NO emission from
this CH4/O2/N2 premixed flame by setting up a parameter study with respect to
pressure (Reactor Properties panel) and inlet velocity (Inlet Stream Properties Panel).
Table 2-7 lists inlet velocities for all pressures investigated by Thomsen et al.10 (p. 42)
in the phi=0.6 flame experiments.

Table 2-7 Inlet velocities for pressures studied by Thomsen et al.

Volumetric Flow
Pressure (atm) Inlet Velocity (cm/sec)
Rate (slpm)

1.00 3.50 13.0517

3.05 6.20 7.5804

6.10 9.10 5.5630

9.15 10.95 4.4626

11.90 12.75 3.9954

14.60 14.50 3.7035

NO emission from the phi=0.6 CH4/O2/N2 flame at different pressures are presented
in Figure 2-14. Here, the model with radiation continues to show very good
agreement with data, over the range of pressures. NO emission values from
measurements and predictions are taken at 0.3 cm from the burner surface except for
the atmospheric pressure ones, which are at 0.7 cm from the burner surface. The
predicted NO emissions shown in Figure 2-14 are obtained by interpolation from the
NO solution profiles.

Alternatively, there are two ways to obtain the computed NO value at the exact location where
measurement is taken: you can provide an initial grid containing the measurement location
(0.3 cm) rather than using the automatic grid option; or you can utilize the continuation option
by setting the end point at the measurement location (0.3 cm) in the first run, and then extend
the end point to its final location (10 cm) in the continuation run.

CK-TUT-10112-1112-UG-1 45 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-12 Axial gas temperature profiles predicted with and without gas radiation heat loss as compared against
experimental temperature profile for the phii=0.6 CH4/O2/N2 flame at 14.6 atm.

Figure 2-13 Comparisons of measured and predicted NO mole fraction profiles as a function of axial distance from
the burner surface for the phi=0.6 CH4/O2/N2 flame at 14.6 atm.

2011 Reaction Design 46 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-14 Measured and predicted NO concentrations behind the phi=0.6 CH4/O2/N2 flame versus pressure.

2.3.6 Soot Formation in Radiating Opposed-flow Diffusion Flame [CHEMKIN-PRO

Only]
This tutorial presents a simulation of a sooting opposed-flow diffusion flame at
atmospheric pressure, including the effects of heat loss from radiating soot particles.
This project uses the chemistry set for Ethylene/Air Combustion and Soot Formation
and Growth as described in Section 2.9.5. Since the fuel stream is diluted with helium,
the gas-phase mechanism is modified by replacing all instances of argon with helium.
Although this is an approximation, the uncertainties introduced are likely small for the
purposes of this project. Opposed-flow flames have been gathering popularity
recently as targets for studying soot formation and oxidation, because they offer
simple flow fields with no wall interference and can facilitate both non-premixed
(diffusion) and premixed flames. Two classes of sooting diffusion flames can be
established in the opposed-flow configuration12. When the flame is located on the
oxidizer side of the stagnation plane, soot particles formed on the fuel side of the
flame will be carried away from the flame and oxidizer, and a soot formation (SF)
flame is created, i.e., soot oxidation is absent. If the flame is instead located on the
fuel side of the stagnation plane, soot particles will be pushed towards the flame and
oxidizer and a soot formation/oxidation (SFO) flame is produced.

12. J. Y. Hwang and S. H. Chung, Combustion and Flame 125:752-762 (2001).

CK-TUT-10112-1112-UG-1 47 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

This project simulates an SF flame (flame 1) investigated by Atreya et al.13 The

simulation starts from scratch, with no estimated profiles necessary for either gas
phase or dispersed phase (soot particle). It is also possible to initiate the calculation
from a converged gas-phase-only solution of the same conditions. By adding a
surface reaction input file containing soot inception, growth and oxidation, a gas-only
simulation could be expanded to predict sooting characteristics of the system. The
restart arrangement provides a convenient and efficient method for soot mechanism
development because it can be easily adopted into parametric studies of soot
formation/growth rate coefficients.

For this project, the radiation heat transfer model is employed to compute the
radiation heat loss to the environment from both gas and dispersed phases. The
optically-thin limit assumed by the radiation model for both gas and dispersed phases
is described in the CHEMKIN Theory Manual.

2.3.6.1 Project Setup

The project file is called opposed-flow_flame__gas_soot_radiation.ckprj. The data
files used for this sample are located in the samples2010\opposed-
flow_flame\gas_soot_radiation directory. The diagram for this sample project
consists of two gas inlets, an opposed-flow flame reactor, and an outlet. Please refer
to the Hydrogen/Air Opposed-flow Flame tutorial in Section 2.3.9 for general
instructions on setting up an opposed-flow flame simulation. Similarly, procedures
regarding soot-particle simulations and the gas radiation model option are available in
tutorials for Soot Particles in Flame Simulators [CHEMKIN-PRO Only] (Section 2.7.2)
and NO Emission from High-pressure Flames with Gas Radiation [CHEMKIN-PRO
Only] (Section 2.3.5), respectively. This project focuses on the additional inputs
required and results obtained for modeling radiation heat transfer from soot particles
within a sooting flame.

The restart file must be specified in the IniSource2 panel of Initialization Data Sets.
Use the browse button to select the Solution File as XMLrestart_opposed-
flow_flame__gas_only.zip in the working directory. This solution file was obtained by
running this project without including the surface reaction mechanism in the chemistry
set. The gas-only solution contains 99 gas species and 196 grid points. It is possible
to redistribute the grid points in the solution during the restart; however, this option is
not used here.

13. A. Atreya, C. Zhang, H. K. Kim, T. Shamim, and J. Suh, Proceedings of the Combustion
Institute 26:2181-2189 (1996).

2011 Reaction Design 48 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

On the Reactor Physical Properties tab of the C1_Opposed-flow Flame panel, the
Problem Type is selected as Solve Gas Energy Equation, because the gas
temperature will be computed from the energy equation with radiation heat loss from
both the gas and the soot particles. Because absorption coefficient data for CO2,
H2O, CO, and CH4 are given in the thermodynamic data file in the chemistry set used
in this project, the gas radiation model is activated automatically. The default value of
298 K is used for the Ambient Temperature.

On the Dispersed Phase::CARBON tab, the Scaling Factor for Moments under the
Basic tab is set to 106. This factor controls the magnitude of the moment terms in the
equations being solved, which helps to improve convergence behavior when
considering particle populations that may be very larger or very small. Reduce this
scaling factor if more soot is expected in the system, and increase it if less soot is
expected. The Aggregation Model is turned off in this project to allow focus on the
radiation effects. The Particle Radiation Model is defined under the tab bearing the
same name. Unlike the gas radiation model, the particle radiation model is off by
default. Check the box next to Include Particle Radiation to activate particle
radiation in the simulation. The total particle radiation input parameter is 700 (K-1-m-1).
This is a semi-empirical parameter that is described in the CHEMKIN Input Manual.

The simulation uses a segregated solver in which gas-phase conservation equations

and those of the dispersed phase are solved separately from the particle equations,
with iteration to resolve the equation coupling. When the dispersed phase, i.e., soot,
is solved, the gas phase properties such as species mass fractions and gas
temperature are kept constant, and vice versa. On the Basic tab of the Solver panel, a
Relaxation Factor of 0.7 is used to speed up the convergence rate of the segregated
solver. During the first few iterations between the gas phase and the dispersed phase,
the dispersed phase calculation tend to over-estimate surface consumption rates of
some PAH species, which could cause mass fractions of these PAH species to
become slightly negative in ensuing gas-phase calculation. Therefore, to keep the
gas phase calculation from failing, especially for heavily sooting flames, the lower-
bounds value for gas species mass fractions must be increased (relaxing the solver
constraint) from the default. The Minimum Bounds on Species Fractions can be
found under the Advanced tab and is set to -0.01 in this project. The negative mass
fraction issue will be corrected in the following iterations as the dispersed-phase
solution settles down. Since the gas-species lower-bounds value is very relaxed, it is
important to make sure there is no large negative gas species mass fraction in the
final solution. It is also necessary to increase the number of pseudo time steps
allowed by the steady-state solver. When the dispersed phase is solved for the first
time, the steady-state solver might need to take many small time steps to grow the

CK-TUT-10112-1112-UG-1 49 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

particle size moments toward values that are more easily tracked by the steady-state
solver. Consequently, the Maximum Number of Pseudo Time Stepping Operation
Allowed is increased from 100 to 3000 to prevent the solver from failing due to this
limitation.

2.3.6.2 Project Results

Atreya et al. 13 (p. 48) reported that the stagnation plane, where axial velocity is zero,
and the diffusion flame of Flame 1 in their study are located at 1.23 cm and at 1.93 cm
from the fuel side, respectively. Soot particles are observed experimentally between
the stagnation plane and the flame. According to the definition of Hwang and Chung12
(p. 47), this Flame 1 is a typical SF flame. Once the soot particles are created, they will

be pushed away from the flame and towards the stagnation plane and from the
diffusion flame in this case. Once the soot particles are created, they will be pushed
away from the flame by convection, such that no soot oxidation will occur.

The axial velocity profile shown in Figure 2-15 indicates the stagnation plane
predicted by the model is at 1.236 cm from the fuel side. This good agreement is a
result of adjusting the fuel inlet velocity gradient in the model. Matching the stagnation
plane is important, because it ensures that the strain rate in the model is comparable
to the one in the experiment, and inlet-velocity gradients are often very difficult to
know accurately. From the computed velocity profile, the strain rate is 8.7 sec-1 (1/2
velocity gradient at the stagnation plane) while the actual value is 5.6 sec-1. While
this is ~40% difference, that is considered relatively good agreement (i.e., same order
of magnitude), since the strain rate is a particularly difficult and sensitive
measurement.

The predicted and measured gas temperature profiles are presented in Figure 2-16.
The temperature on the oxidizer side of the flame agrees well with the experimental
profile, but the mode over-predicts the temperature on the fuel side of the flame.
Since all soot particles settle between the flame and the stagnation plane, it is
plausible that the particle emissivity used by the radiation model is too low. The
predicted peak temperature location is also slightly off towards the fuel side at
1.89 cm.

Figure 2-17 to 2-20 compare computed and measured profiles of some major and
radical species. In general, the model does well in predicting major species in the
system. The OH peak in Figure 2-19 reveals the location and the thickness of the
diffusion flame. The Acetylene (C2H2) profile in Figure 2-19 and the pyrene (A4)
profile in Figure 2-20 show the regions where soot inception and mass growth take
place, respectively.

2011 Reaction Design 50 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-21 compares the predicted soot volume fraction (cm3/cm3) against the
experimental data as a function of distance from the fuel nozzle. The peak soot
volume fraction predicted by the model is about the same as the measurement and
the location and the width of the sooting zone agree well with the data. The model
also captures salient features of the soot volume fraction profile: a sharp drop in soot
volume at the fuel side of the sooting zone, marking the particle stagnation plane, and
a gradual decay in soot volume on the opposite side. The soot volume peak coincides
with the particle stagnation plane and is on the fuel side of the peak soot inception
location, represented by the pyrene peak in Figure 2-20. The O2 and the OH profiles
in Figure 2-17 and Figure 2-19 suggest that there is no significant O2 or OH
penetration into the sooting zone and that soot oxidation is negligible in this flame.

Figure 2-15 Predicted axial velocity profile of Flame 1 studied by Atreya et al. 13 (p. 48). The location of (velocity)
stagnation plane is indicated by the blue dash-dotted line.

CK-TUT-10112-1112-UG-1 51 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-16 Predicted gas temperature profile is compared against the experimental profile for Flame 1 studied by
Atreya et al. 13 (p. 48).

Figure 2-17 Comparisons of predicted and measured fuel (CH4) and oxidizer (O2) profiles for Flame 1 studied by
Atreya et al. 13 (p. 48).

2011 Reaction Design 52 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-18 Comparisons of predicted and measured profiles of H2 and CO for Flame 1 studied by Atreya et al. 13 (p.
48)
.

Figure 2-19 Predicted and measured C2H2 and OH profiles of Flame 1 studied by Atreya et al. 13 (p. 48). The OH peak
indicates the flame front and the C2H2 peak marks the major soot growth region.

CK-TUT-10112-1112-UG-1 53 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-20 Predicted pyrene (A4) profile showing soot inception mainly occurs in the region between the stagnation
plane and the diffusion flame in Flame 1 studied by Atreya et al. 13 (p. 48).

Figure 2-21 Predicted soot volume fraction profile is compared against the experimental profile as a function of
distance from the fuel nozzle for Flame 1 studied by Atreya et al. 13 (p. 48).

2011 Reaction Design 54 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

2.3.7 Flame Speed of Stoichiometric Methane/Air Premixed Flame with Reaction

Path Analyzer [CHEMKIN-PRO Only]

2.3.7.1 Problem Description

In this tutorial we seek to determine the laminar flame speed and structure of an
adiabatic, atmospheric-pressure, freely propagating, stoichiometric methane-air gas
mixture, and also use the Reaction Path Analyzer to acquire a visual representation of
the connecting reactions that form or deplete chemical species.

To assist in the solution of the freely propagating flame problem, we follow one
popular strategy for assuring quick convergence to an accurate solution. This method
uses continuations to successively refine the domain and grid of the solution until a
desired accuracy and grid-independence is achieved.

2.3.7.2 Problem Setup

The project file for this tutorial is called flame_speed__freely_propagating.ckprj
and is located in the samples2010 directory. Since we are not interested in NOx
formation here, we use a skeletal methane-air combustion mechanism to speed up
the calculation. As long as it contains all essential steps for methane oxidation under
the desired conditions, the skeletal mechanism should yield adequate results while
still providing the essential high-temperature chemistry of methane-air oxidation.

The Flame-speed Calculator simulates a freely propagating flame, in which the point
of reference is a fixed position on the flame. In this coordinate system, the flame-
speed is defined as the inlet velocity (velocity of unburned gas moving towards the
flame) that allows the flame to stay in a fixed location, which is an eigenvalue of the
solution method (see the CHEMKIN Theory Manual for details). To set up the flame-
speed calculation, we need to specify the properties of the fresh gas mixture.
Composition of the unburned methane-air mixture is input on the Species-specific
Properties tab of the C1_Inlet1 panel and the initial guess of the mass flow rate on the
Stream Properties Data tab. The unburned gas temperature is specified in the
Reactor Physical Properties tab of the C1_Flame Speed panel. The pressure and an
initial, coarse temperature profile are also input on the Reactor Physical Properties
tab of the C1_ Flame Speed panel (see Figure 2-22 and Figure 2-23).

CK-TUT-10112-1112-UG-1 55 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-22 Flame SpeedTemperature Profile Panel

Figure 2-23 C1_ Flame SpeedReactor Physical Properties

2011 Reaction Design 56 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

First, we establish a fixed-flame coordinate system by explicitly constraining the gas

temperature to stay at the initial fixed value at one grid point in the computational
domain. The fixed temperature grid-point should be unique and must lie between the
unburned temperature and the expected adiabatic flame temperature for the gas
mixture. The temperature value to be fixed is specified on the Reactor Physical
Properties tab, as the optional user-defined Temperature Constraint If no value is
given for this input, a default value of the average of the unburned temperature and
the adiabatic flame temperature is used. An initial guess is required for the
temperature profile and for the inlet velocity of the unburned gas (i.e., initial guess of
the flame speed that will be calculated). The temperature values used in the profile,
except the first point, are estimates, and so only a very rough estimate of the
temperature profile is required.

The convergence rate normally is not very sensitive to the initial guess of mass flow
rate but a good mass flow rate guess can be very helpful when the equivalence-ratio
is close to the flammability limit. Note that the mixing zone width is larger than the
initial domain. This is fine as these parameters have relatively little physical meaning,
but we find that more spread-out guesses are often more likely to lead to convergence
than narrow ones.

To compute an accurate flame speed, it is important to have the boundaries

sufficiently far from the flame itself so that there is negligible diffusion of heat and
mass through the boundary. In other words we need to have the domain sufficiently
large that the flame is "freely' propagating, unconstrained by the boundaries of the
domain. However, in the solution strategy described here, we start with an initial run
on a very coarse grid that contains only a few grid points and a computational domain
just wide enough to encompass the flame, and then use continuations to gradually
expand the domain until we have a domain-independent solution.

Once we obtain an initial solution on this coarse grid, we expand the domain while
reducing/tightening the parameters that control the degree to which the solution
gradient and curvature is resolved. We repeat this process a couple of times until the
temperature and species slopes at the boundaries are close to zero and both gradient
and curvature controls are at least 0.5 or less. The parameters varied during
continuations runs are defined in the Continuations panel of the CHEMKIN Interface.
The continuation runs start from the solution of the previous run and so rapidly
converge to the refined solution.

CK-TUT-10112-1112-UG-1 57 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.3.7.3 Project Results

Solutions from the last continuation (solution number 3) are shown in Figure 2-24 and
Figure 2-25. The fact that the burned gas temperature of 2234 K (Figure 2-25) is
within 3 K of the published adiabatic flame temperature for these conditions is
indicative of the accuracy of the chemistry mechanism used in this simulation. The
laminar flame speed by definition is the relative speed between the unburned gas
mixture and the flame front. Since the coordinate system is fixed to the flame, all
velocity solutions are actually relative velocities with respect to the flame front.
Accordingly, the flame speed should be the velocity solution at the point where
temperature and composition are the same as the unburned gas mixture. It is
important to check the gradients of gas temperature and major species to make sure
those values are nearly zero at both boundaries. If there is a non-zero gradient at one
of the boundaries, we would need to extend the domain farther to ensure that the
assumed adiabatic and zero-diffusive-flux conditions are met. As Figure 2-24 and
Figure 2-25 show, however, the species and temperature gradients at both
boundaries are sufficiently small that there is no appreciable loss of mass or energy
through the boundaries. The predicted flame speed is therefore 41.01 cm/sec, which
corresponds to the minimum temperature in Figure 2-24. This value is also reported
in the text output file and is available as a result for post-processing of a flame-speed-
based parameter study. The laminar flame speed measured by Egolfopolous et al.14
for the stoichiometric methane-air flame at one atmosphere is 36.53 cm/sec. The
discrepancy may be due to the simplified kinetics used in the simulation or to potential
non-adiabatic conditions in the experimental measurement.

Figure 2-24 Flame SpeedAxial Velocity vs. Distance

14. F.N. Egolfopolous et al., Proceedings of Combustion Institute vol. 23 p. 471 (1988).

2011 Reaction Design 58 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-25 Flame SpeedTemperature vs. Distance

2.3.7.4 Reaction Path Analysis

Applying the CHEMKIN Reaction Path Analyzer (RPA) described in the CHEMKIN
Visualization Manual to the results provides a visual representation of the connecting
reactions that form or deplete chemical species. Employing the RPA, the
decomposition pathways at different sections of the methane/air flame-speed
calculation become evident.

To see the sensitivities on methane, the appropriate output calculation must be added
to this sample calculation, as shown in Figure 2-26.

CK-TUT-10112-1112-UG-1 59 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-26 Flame SpeedOutput Calculation pane with Sensitivity Option.

Begin by selecting Analyze Reaction Paths from the Analyze Results panel, as
shown in Figure 2-27. In this case, methane was explicitly added in the sensitivities
calculations on the Species Sensitivity tab of the Output Control panel. After adding
the sensitivity, Run Calculations must be double-clicked in the project tree to open
the Analyze Results panel with the Analyze Reaction Paths option and Begin
Analysis button.

2011 Reaction Design 60 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-27 Flame SpeedAnalyze Results panel with Reaction Path Analyzer option.

Hint: Calculate sensitivities only for the species that most interest you: this decreases the size
of the solution and consequently speeds up the RPA.

CK-TUT-10112-1112-UG-1 61 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

The RPA will then generate the diagram for the initial point in the premixed solution.
Clicking in the variable solution plot selects another point in the solution file. Since
this project has two continuations with ever-increasing grid resolution, the third
solution is chosen to display, as seen in Figure 2-28. The Y variable is also chosen to
be temperature, so we can determine where the flame region begins.

Figure 2-28 Flame SpeedAltering the solution chosen, and y variable shown in the solution plot.

On Windows architecture, if the Graphviz option has been selected in the installer,
then a Hierarchical layout option is available in the Preferences tab. Selecting the
Hierarchical option will change the way that the species are laid out in the diagram.

2011 Reaction Design 62 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-29 Flame SpeedSwitching to a hierarchical layout option.

The solution point is altered by entering the value 0.18 into the Distance text field, and
the number of species is increased to 15. The complexity of the diagram increases
due to the high temperature activation of various reaction channels, as well as an
increased number of species allowed in the diagram. Consequently, only a small
section of the diagram is currently visible, as seen in Figure 2-30.

CK-TUT-10112-1112-UG-1 63 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Tip: Clicking left and right within the solution plot will decrement and increment the position
within the solution. Changing the point will require a reread of the solution file.

Figure 2-30 Flame SpeedIncreased complexity in the diagram due to changing the solution point, and increasing
the number of allowed species.

2011 Reaction Design 64 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

A simple method to see the entire diagram is to decrement the zoom setting found
under the second tab, labeled Preferences, as shown in Figure 2-31.

Figure 2-31 Flame SpeedAltering the zoom setting to see the contents of the diagram.

CK-TUT-10112-1112-UG-1 65 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Back near the colder side of the flame, the diagram is altered to focus on the
decomposition pathway of methane, and the number of species is reduced back to
ten. This results in a simpler diagram that follows the major pathways flowing out
from methane, as shown in Figure 2-32. Additionally, a C element filter is applied,
restricting the kind of species displayed in the diagram to those containing Carbon.

Figure 2-32 Flame SpeedDecomposition pathways starting with methane.

2011 Reaction Design 66 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Clicking on the methane species in this diagram creates two plots, as shown in
Figure 2-33. The first plot describes the Rates of Production of all reactions that
influence the composition of methane at this solution point. The second plot shows
the Sensitivity of the selected species, in relation to the rates of the reactions within
the mechanism.

Figure 2-33 Flame SpeedThe Rate of Production and Sensitivity charts of methane.

CK-TUT-10112-1112-UG-1 67 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Each connection can be split into constituent reactions. This way each reaction is
represented on the screen, explicitly illustrating the contributions, as shown in
Figure 2-34.

Figure 2-34 Flame SpeedSplitting a composite reaction pathway into one per reaction.

A relative rate analysis can be accomplished while each reaction is split. This kind of
analysis will scale the width of exiting arrows relative to the total of all reactions
depleting a species. This scaling helps to determine which reaction has the largest
influence on the removal of a species locally. The absolute rate analysis will scale
every reaction pathway relative to a single global value.

2011 Reaction Design 68 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

The default layout method, the tree format, is available, which graphically illustrates
decomposition as the downward flow of a species. A selected start species is chosen
and the other species are laid out in generations from the selected root, as shown in
Figure 2-35.

Figure 2-35 Flame SpeedTree layout method of the decomposition of methane.

Tip: On the Preferences panel, using the Straight line option compresses the diagram and
makes it easier to manage.

A second way to decrease the complexity of a diagram is to, for example, remove one
branch of a bimolecular reaction product. This can be done by specifying one of the
reaction products as a side species. A side species is one that may be important to a
reaction, but for some reason should not be drawn explicitly in the diagram. For

CK-TUT-10112-1112-UG-1 69 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

example, in Figure 2-36 the hydroxide radical and oxygen radical are chosen as side
species. Consequently, the hydroxide radical will not be drawn in the diagram. The
other two default side species (H and HO2) were removed from the side species list
for this example.

Tip: If a species appears isolated from the others on the diagram, try decreasing the Relative
Cutoff Fraction.

Figure 2-36 Flame SpeedSetting the hydroxide radical and oxygen radical as side species. The source was
switched from methane to oxygen, and the element filter was changed to O element.

2011 Reaction Design 70 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Colors can help distinguish the effect of a side species on the reactions in a diagram.
For example, in Figure 2-37, after setting the color preferences, any reaction with the
hydroxide radical is colored red, and any reaction with the oxygen radical is colored
green.

Figure 2-37 Flame SpeedColored reactions based on influence of side species.

The coloring provides an easy way to distinguish the effects that one species may
have on the formation or depletion of another.

These diagrams graphically show the decomposition pathways of the fuel, and the
primary formation pathways of the products in the flame speed calculation.

2.3.8 Parameter Study: Propane/Air Flame Speed as a Function of Equivalence

Ratio and Unburned Gas Temperature

2.3.8.1 Project Description

This tutorial illustrates using the Parameter Study Facility to calculate laminar flame
speeds of propane-air mixtures over a wide range of equivalence ratios and unburned
gas temperatures at atmospheric pressure. The tutorial also illustrates some features
in the CHEMKIN flame speed calculator model for easy setup.

CK-TUT-10112-1112-UG-1 71 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

The behavior of flames in fuel-lean and fuel-rich systems at a range of unburned gas
temperatures are of interest in several applications, such as engine combustion. In
this tutorial, the flame speeds of propane-air mixtures are calculated for the following
conditions:

1. Equivalence ratios of 0.6 1.4.

2. Initial temperatures of 300 K and 700 K.
3. Pressure of 1 atm

Equivalence ratio and unburned gas temperature are set as parameter studies. A total
of 18 parameter-study cases are run to cover the range of operating conditions.

Parameter studies can be used for several purposes:

To study the impact of varying inputs of operating conditions such as pressure

or stream properties such as concentrations.
To analyze the sensitivity of output to reaction rate parameters and transport
properties, and to analyze the impact of uncertain parameters.
The problem uses the chemistry set described in Section 2.9. The project file is
named flame_speed__parameter_study.ckprj. The data files for this project are
located in the samples2010\flame_speed\parametric_study directory.

2.3.8.2 Project Setup

Most of the inputs are conceptually similar to the tutorial on flame speed calculation of
methane/air flame in Section 2.3.7. CHEMKIN provides useful defaults for several
inputs, including the automated generation of a guess temperature profile at which to
start the calculations. We will use this option, along with an initial grid based on this
automated temperature profile. We will also use the default values for all the solver
parameters and the inlet velocity guess value.

On the Reactor Physical Properties panel, the pressure is specified as 1 atm. A

parameter study is set up for the Unburned Gas Temperature. The nominal value of
the Unburned Gas Temperature is specified as 300 K, and a parameter study is set up
with 2 values of the unburned gas temperature, 300 K and 700 K, as shown in
Figure 2-38. To set up the parameter study based on unburned temperature, the
adjacent icon (see Figure 2-39) is selected, and the window shown in Figure 2-38
appears. The parameter study can be populated by entering the Start and End value
for the equivalence ratio, and selecting one of 4 options from the dropdown list for the
appropriate distribution of values between the beginning and end value specified, as
shown in Figure 2-38.

2011 Reaction Design 72 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-38 Panel for Setup of Parameter Studies

To modify one of the values, double-click it.

To delete a value, select it and press the Delete button.
To add a value, press the Populate button.
Once the setup is filled out, click the Done button to complete the setup. After the
specification of the inputs for the parameter study are complete, the parameter will
appear in blue and be underlined, as shown in Figure 2-39. This indicates that the
setup of the parameter study is complete for that parameter.

In the Reactor Physical Properties panel, we have used the automatic estimation of
the starting temperature profile, as shown below. This guess temperature profile can
sometimes have a large influence on convergence. The automatic estimation of the
temperature profile feature has been added for easier model setup, and this

CK-TUT-10112-1112-UG-1 73 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

estimated temperature profile provides a good starting guess and aids convergence.
If the automated estimation of the starting temperature profile option is chosen, the
optional user-defined temperature constraint input is grayed out, and CHEMKIN
calculates this value.

Figure 2-39 Automatic Estimation of Starting Temperature Profile

Parameter Study Icon

Mixture-averaged transport properties are used in this tutorial, with correction velocity
formulation.

Under the Grid Properties panel, the domain-sizerelated inputs and the grid-
resolution inputs are specified. As with the freely propagating tutorial in Section 2.3.7,
we will start the problem on a small domain and a coarse grid, and subsequently use
Continuations to expand the domain and refine the grid, which improves
convergence. A starting small domain of 0.3 cm is specified in the Grid Properties
panel. Subsequently, in the Continuations panel, the domain is expanded to 12 cms.

2011 Reaction Design 74 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

(starting axial position is expanded to -2 cm, and ending axial position to 10 cms).
This ensures that the gradients of gas temperature and major species are nearly zero
at both boundaries, so that the requirement of adiabatic and zero-diffusive-flux
conditions at the boundaries are met.

The model uses a non-uniform grid that is successively and automatically adapted
based on solution gradients and curvatures determined on an initially coarse grid.
Relative gradient and curvature parameters that determine the extent to which the
solution will be refined are model inputs. The relative gradient and curvature
parameters for the grid refinement are both set in the Grid Properties panel to 0.9.
Subsequently, in the Continuations panel, these values are both lowered to 0.1, to
obtain a finer grid. To allow sufficient grid points for a well refined grid, we set the
maximum number of grid points to 300.

In the Grid Properties panel, the estimated center position and zone width are left to
program defaults. These inputs are grayed out when the automated estimation of the
temperature profile option is chosen.

In the Species-specific properties panel, Intermediate fractions and Product fractions

of species can be specified to provide a starting guess of species profiles. These are
optional inputs, but typically help in convergence. The CHEMKIN model incorporates
features that make these specifications easier. For the Intermediate fractions, there is
an Auto-Populate button that can be used, as shown below. Based on the
mechanism used in the problem, the Auto-Populate gives a reasonable guess for the
intermediate fractions. If no product species guess values are specified, equilibrium
values are calculated by CHEMKIN and used for the product fraction initial guesses.

CK-TUT-10112-1112-UG-1 75 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-40 Auto-Populate Option

On the Inlet panel, the inlet velocity guess value is left to default. In this tutorial, the
reactant concentrations are specified in terms of equivalence ratio. The equivalence
ratio is a measure of the fuel/oxidizer ratio, as compared with the stoichiometric value
(explained further in Section 2.2). In this tutorial, the fuel consists of pure propane,
while the oxidizer consists of air. For specifying the oxidizer as air, the Auto-Populate
Air button under the Oxidizer Mixture panel can be used. The stoichiometric products
from combustion must be specified; once the Fuel Mixture and the Oxidizer Mixture
have been specified, the Auto-Populate button can be used for specifying the
Complete-Combustion Products, and in this way CHEMKIN automatically determines
the complete-combustion products. The nominal value of the equivalence ratio has
been set to 1.0. Now, a parameter study varying the equivalence ratio can be set up.
Nine values have been used to provide a range of equivalence ratios of 0.6 1.4. We
would like to create a parameter study to set up equivalence ratios of 0.6 1.4 for
each of the unburned gas temperature values of 300 K and 700 K. To achieve this, we
click on the next to the equivalence ratio, and then select the Vary each
parameter independently so that only one parameter varies on each run option,
as shown in Figure 2-41.

2011 Reaction Design 76 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-41 Varying Each Parameter Independently

Once this parameter study is set up, there will be 18 parameter study runs, varying
the equivalence ratio and unburned temperature, as shown below.

Figure 2-42 Parameter Study After Setup Is Complete

CK-TUT-10112-1112-UG-1 77 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

All Solver panel inputs are left at defaults. There are several solver input options that
CHEMKIN provides for the advanced user that can be used when convergence is
more difficult.

Continuations have been set to expand the initial domain and refine the grid, as
explained above.

No other inputs need to be explicitly provided. For several model inputs, the default
values provided in the CHEMKIN Interface have been used.

Once the setup is done, to run the parameter study, select the Run Calculations panel
from the project tree, and click on Begin. While the model is running, the Progress
Monitor and the message area show the status of the runs. Results can be post-
processed after the runs are finished (using the Analyze Results node of the project
tree).

2.3.8.3 Project Results

After running the parameter studies, several additional post-processing options are
available. By selecting the Line plot from Plot Type in the Post-Processor, options
appear for accessing the calculated results as shown in Figure 2-43. The first and
second options are exclusive to parameter studies.

2011 Reaction Design 78 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-43 Post-Processor Options After Running Parameter Studies

2.3.8.3.1 Cluster endpoint vs parameter

In this tutorial, we have used parameter studies to cover a range of operating
conditions, and continuations to refine and expand the grid for each parameter study.
Since we have used continuations for each parameter study, there are multiple
solutions available for each parameter study run. Consequently, there would be
multiple flame speed values for each parameter study run. Since we have used
continuations only for refining the grid, we are interested in only the flame speed
value of the final continuation. To access this value in the post-processor, we can use
the The second option in the Plot Set (under Plot Type as Line plot), i.e., Cluster1C1
end point vs parameter, and one of the values here is the Flame Speed end point.

CK-TUT-10112-1112-UG-1 79 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

We can now plot the flame speed values as a function of equivalence ratio. Since
there are two unburned gas temperatures, it would be good to separate the flame
speed vs. equivalence ratio plots into two series. This is done through the Use Plot
Grouping Variable option as shown in Figure 2-44, and choosing the unburned gas
temperature.

Figure 2-44 Cluster Endpoint vs. Parameter

The flame speeds can now be plotted as shown below.

2011 Reaction Design 80 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-45 Flame-Speed Calculated Values as a Function of Equivalence Ratio

2.3.9 Hydrogen/Air Opposed-flow Flame

2.3.9.1 Project Description

This user tutorial presents a simulation of opposed flow diffusion flame of hydrogen
and air at low pressure. This project uses the chemistry set for hydrogen combustion
described in Section 2.9.1. The opposed-flow geometry makes an attractive
experimental configuration, because the flames are flat, allowing for detailed study of
the flame chemistry and structure. The two or three-dimensional flow is reduced
mathematically to one dimension. This problem uses cylindrical coordinates where
one stream contains fuel and the other oxidizer. It also demonstrates the use of
sensitivity analysis for reaction rates and species heats of formation. The latter
analysis is useful for evaluating the effects of thermochemical parameters that may
have been estimated, or have high uncertainties.

2.3.9.2 Project Setup

The project file is called opposed-flow_flame__h2_air.ckprj. The data files used for
this sample are located in the samples2010\opposed-flow_flame\h2_air directory.
This reactor diagram contains two gas inlets, one Opposed-flow Flame Reactor, and
an outlet.

CK-TUT-10112-1112-UG-1 81 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

On the Reactor Physical Properties tab of the C1_ Opposed-flow Flame panel, the
problem type of Solve Gas Energy Equation is selected. Here, the use of Mixture-
averaged Transport properties is also selected, as is the choice of a plateau-shaped,
rather than a linear, profile for the starting guess used in the simulation. An optional
value for the maximum temperature to be used in the initial profile is provided to help
convergence.

On the Grid Properties tab of the C1_ Opposed-flow Flame panel, the use of
cylindrical geometry is selected for this problem, and the axial length of the simulation
(2 cm) is input. In this reactor model, the fuel always enters the system at the origin,
and the oxidizer inlet is located at the Ending Axial Position. The opposed-flow reactor
model uses adaptive gridding, and in this case, the spacing of the 14 initial grid points
have been specified by use of a profile file, opposed-flow_flame__h2_air.ckprf.
There are 4 optional parameters on this panel that provide input for the adaptive
gridding, two of which have values that we have input to override the defaults. The
simulation also needs a starting estimate of the solution from which to begin its
iteration, and the Estimated Center Position and Estimated (reaction) Zone Width
help specify that. The gas composition giving the expected combustion products that
are input on the Product Fraction sub-tab of the Species-specific Properties tab on the
C1_ Opposed-flow Flame panel are also part of the initial guess.

The gas inlet panels are named to reflect their function. The inlet gas velocities
(100 cm sec-1) are input on the Stream-specific Data tabs of the Fuel and Oxidizer
panels, along with inlet gas temperatures (300 K). The inlet gas compositions, pure
hydrogen and pure air, are input on the Reactant Fraction sub-tab of the Species-
specific Properties tab of the Fuel and Oxidizer panels, respectively.

On the Solver panel, there are a number of inputs on the Basic and Advanced tabs to
override the default values and assist convergence. On the Output Control tab of the
Output Control panel, boxes are checked to request that sensitivity calculations be
done for all variables with respect to both reaction-rate A factors and species heats of
formation. On the Species Specific Data tab, three species are listed as being Output
Species. There are no inputs on the Cluster Properties or Continuations panels for
this project.

2.3.9.3 Project Results

Figure 2-46 shows the gas temperature from the simulation as a function of axial
distance. The flame is located on the fuel side of the stagnation plane, which is a
result of using hydrogen as the fuel. Most fuels require more air than fuel by mass, so
the diffusion flame usually sits on the oxidizer side of the stagnation plane. In a
stoichiometric mixture, the fuel usually diffuses through the stagnation plane to

2011 Reaction Design 82 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

establish the flame. For H2 however, more fuel is required than air. The mole fractions
in Figure 2-47 for the major species show that the flame sits on the fuel side of the
stagnation plane in this case. An inspection of the text file shows that the simulation
now has 45 grid points, a significant increase from the initial 14 grid points.

Figure 2-46 Hydrogen/Air FlameTemperature vs. Axial Distance

Figure 2-47 Hydrogen/Air FlameMole Fractions

Figure 2-48 shows the normalized sensitivity coefficients as a function of distance for
the five reactions with the largest temperature sensitivities. Reaction #1,
H + O2 + M = HO2 + M, has both the most positive and most negative values.
Increasing the rate of this reaction would increase the temperature on the fuel side,

CK-TUT-10112-1112-UG-1 83 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

and decrease it at the oxidizer side of the flame. Figure 2-49 shows the sensitivity
coefficients as a function of distance for the five species with the largest temperature
sensitivities for heats of formation. The temperature is most sensitive to the heat of
formation of H2O, which makes sense as this is the primary product species.

Figure 2-48 Hydrogen/Air FlameTemperature Sensitivity to Reaction Rates

Figure 2-49 Hydrogen/Air FlameTemperature Sensitivity to Heats of Formation

2011 Reaction Design 84 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

2.3.10 Flame Extinction Analysis [CHEMKIN-PRO Only]

2.3.10.1 Project Description

This tutorial describes how to use the reactor model named Extinction of Premixed or
Opposed Flow Flame, which is available under the Flame Simulators (see
Figure 2-50). Flame extinction studies are fundamental to flame characterization and
response analysis. Moreover, such studies also provide data for validation of reaction
mechanisms. In a typical flame extinction experiment, premixed fuel-air mixture is
supplied from one nozzle and an inert gas for the other. The sample presented here
simulates one such case for a premixed methane-air mixture.

This extinction simulator can read an existing Opposed-Flow Flame model solution
file and start the extinction analysis or can originate an opposed-flow flame problem
and carry it out to extinction analysis.

Figure 2-50 Icon for Extinction of Premixed or Opposed Flow Flame reactor

As discussed in the CHEMKIN Theory Manual, extinction analysis is conducted using

successive flame-controlling continuations. The flame-control is achieved by using
internal boundary conditions on temperature. (See Section 13.6 in the CHEMKIN
Theory Manual.) In this tutorial, the one-point temperature control method is used.
The one-point control method is more appropriate when a premixed mixture is
supplied from one nozzle and an inert gas from the other; as in this tutorial example.
On the other hand, when fuel and oxidizer are supplied from each nozzle separately
(i.e., a diffusion flame), two-point control can be considered more flame-controlling.
Mathematically, however, the problem is well-posed with either technique although
numerical difficulties can appear when one-point control is used for the diffusion
flame.

When setting up an extinction analysis, it is important to remember the following one-

point and two-point temperature control conventions. With one-point temperature
control, the internal boundary condition on temperature is set for 0 < x < xflame,
where x is the spatial coordinate and xflame is the flame (maximum temperature)
location. For two-point control, the simulator chooses a temperature control point on
either side of the flame. By default, inlet 1 is located at x = 0. If you name the inlets
as Fuel and Oxidizer, the Fuel inlet is treated as inlet 1 and is located at x = 0. For
all inlet pairs, inlet 1 is at x = 0, as in the default case. As described in
Section 2.3.10.4, the simulator reports the extinction strain rate assuming that the
Fuel (or premixed fuel-air) inlet is located at x = 0.

CK-TUT-10112-1112-UG-1 85 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.3.10.2 Project Set-up

The project file is named opposed-flow_flame__extinction.ckprj and is located in
the samples2010 directory. Extinction analysis is conducted for a premixed
stoichiometric methane-air flame at 1 atm pressure.

A reduced (17 species) methane-air mechanism is used in this tutorial; the chemistry
set is described in Section 2.9.2.2. After successfully preprocessing the mechanism,
the extinction problem can be set up. Note that, unlike an Opposed-flow Flame
reactor model, a fixed-temperature problem is meaningless for extinction and thus the
energy equation is always solved. In addition to the three tabs that appear for the
Opposed-flow Flame reactor, the Reactor Properties tab has one extra sub-tab
named Basic. In the following discussion, the set-up of the options pertaining to the
nominal Opposed-flow Flame model is described first, followed by the details of
controls on the Basic sub-tab.

The controls for the nominal opposed-flow flame model are on the Opposed Flow
Model sub-tab of the Physical Properties tab. In this tutorial, pressure is set to 1 atm
while all other controls are left at their default values. Ending Axial Position is set to
1.4 cm in the Grid Properties tab. The adaptive mesh control options, for both the
solution gradient and curvature, are set to 0.1 while the maximum number of grid
points allowed is set to 1000.

The stoichiometric methane-air mixture is assigned to inlet 1 (named MethaneAir)

while pure nitrogen (N2) is assigned to inlet 2 (named Nitrogen). The inlet gas
temperature for both inlets is set to 296 K. The inlet velocities for both nozzles are set
to 80 cm/s. Note that the first opposed-flow flame solution is thus calculated for these
parameters.

Figure 2-51 illustrates the controls set on the Basic tab. As mentioned earlier, the
one-point option is chosen for the extinction method because a premixed mixture is
used. The number of extinction steps is effectively the number of successive
opposed-flow flame simulations to be conducted in search of the extinction turning
point. Although its default value is 100, in this tutorial lesson the number of steps is
set to 450. Typically, the number of steps can be of the order of 500 and this value
depends not only on the temperature step size used but also on all the other controls
on the Basic tab. The step interval for saving option dictates solution saving
frequency. This control can be useful for restarting a new extinction analysis from a
failed run or continuing from a previous simulation, if the extinction (turning) point is
not reached within the initial value of the number of steps. Also since extinction
analysis is computationally intensive, a new run (for example, for different
stoichiometry) can be restarted from an intermediate solution from a previous run.

2011 Reaction Design 86 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

The nozzle velocities can be constrained for the one-point control method (see
Section 13.6 in the CHEMKIN Theory Manual). In this case, we choose momentum
constraint, which gives the stagnation plane roughly at the middle of the computation
domain. For two-point control, such constraint on the nozzle inlet velocities cannot
be enforced and, if specified, is ignored by the simulator. The maximum and
minimum temperature fractions are set to 0.9 and 0.5 whereas temperature steps of
10 K are used. At any point during the calculations, if the maximum temperature
obtained is less than that specified by the minimum flame temperature, the
calculations stop declaring that the flame existence criterion is violated. The default
value of this control is 1500 K. This control is useful to avoid taking too many steps
beyond the extinction point.

Figure 2-51 Extinction Basic reactor properties

The input parameters maximum temperature fraction, minimum temperature fraction,

and temperature step size control the location and magnitude of internal boundary
condition on temperature. Because this tutorial simulation is not a restart from an old
solution, the simulator first computes a nominal flame for the given input parameters.
(For a restart, it uses the solution supplied.) Then it finds the location at which the
following statement is satisfied:

T = maximum temperature fraction * (Current maximum temperature Inlet

temperature)

CK-TUT-10112-1112-UG-1 87 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

The temperature at this location is then successively decreased by the user-specified

temperature-step-size until it reaches T = minimum temperature fraction * (Current
maximum temperature Inlet temperature). A new location is then selected using the
maximum temperature fraction. This process is repeated until the desired number of
steps is reached or until the flame is effectively extinguished as dictated by the
minimum flame temperature control.

As an example, suppose that the initial flame solution has a maximum temperature of
2100 K for the inlet temperature of 300 K and the maximum and minimum
temperature fractions specified are 0.8 and 0.5, respectively. The simulator would
then find the first location between x = 0.0 and x (at Tmax) such that x(T) >= 300 +
0.8*(2100-300) >= 1740 K. (For two-point control, such a location is found on either
side of the flame.) We will label this location xc1. The internal boundary condition
(T = 1740 K) is imposed at this point. After a flame solution for this new problem is
calculated, the new boundary condition at this location will be T = 1740 K -
temperature step size, and a new flame solution is computed. This process is
repeated until the temperature at the control location is reduced to T = 300 +
0.5*(2100-300) = 1200 K. After a flame solution corresponding to this temperature is
obtained, a new location (xc2) is computed using the maximum temperature fraction
and the flame temperature calculated for the problem in which T = 1200 K is imposed
at xc1. The process is then repeated for xc2.

2.3.10.3 Project Results

The text output file generated by an extinction project contains the output from each
opposed-flow flame simulation conducted along the path to the extinction point.
Typically, size of the output text-file is large since all the solutions are written to it. The
solution file contains corresponding solutions saved according to the specified saving
frequency. The output file corresponding to extinction analysis is opposed-
flow_flame__extinction_ExtinctionResponse.csv. As the path to extinction is
calculated, it is dynamically displayed in a plot that is monitoring the content of this
text output file, as shown in Figure 2-52. The display of this plot is optional, and can
be turned off in User Preferences by clicking on Show Plot Monitor. This file can be
opened with a simple text editor. While a more detailed overview of the general
contents of this file is given in the next section, the following observations can be
made from the contents of this file generated by the tutorial problem. For the initial
solution at nozzle velocities of 80 cm/s, the maximum temperature is about 2066 K.
The simulator then searches for the location of the control temperature computed as
296 + 0.9*(2066-296) = 1889 K. Since there is no grid-point at which the temperature
exactly matches this value, the first grid point with x > 0 at which temperature exceeds
this value is chosen. The temperature at this grid point is approximately 1908 K. The

2011 Reaction Design 88 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

continuation process then decreases the temperature at the control point in steps of
10 K until the control temperature is equal to or less than 296 + 0.5*(2066-
296) = 1181 K. This happens at the 74th step; at this point the flame temperature is
about 2027 K.

Figure 2-52 Dynamic plotting of Temperature vs. extinction response, showing a turn at ~550 s-1

As described earlier, the simulator then finds a new control location such that the
temperature there is 296 + 0.9*(2027-296) = 1854 K. The fuel and oxidizer nozzle
velocities continuously increase to the value in the neighborhood of 200 cm s-1. The
turning point is reached when the MethaneAir nozzle velocity reaches about
203 cm s-1 (after 425 steps). The global strain rate at the turning point is
(203 + 201)/1.4 = 289 s-1. The extinction strain rate (see the definition in
Section 2.3.10.4) is 550 s-1.

2.3.10.4 Post-processing
The user input parameter step interval for saving controls the number of solutions
saved. In this tutorial, it is set to 10. Thus, there are 450/10 = 45 total solutions that
are saved in the solution file. Just as in the case of the nominal opposed-flow flame
solution file, the data for all variables and corresponding derived quantities such as
heat release, rate of production, etc., is available for all these solutions. However,
from a design point of view, only the data at or near the extinction point are important.

CK-TUT-10112-1112-UG-1 89 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

The extinction simulator creates two raw-data output files named opposed-
flow_flame__extinction_ExtinctionResponse.csv and opposed-
flow_flame__extinction_ExtinctionProfiles.out. Figure 2-53 and Figure 2-54 show
snapshots of these files. The former contains the data such as control temperature,
pressure eigenvalue, global strain rate, extinction strain rate, and maximum
temperature at each solution point. (The definitions of the strain rates are given later
in this section). As mentioned in the previous section, a plot of the extinction curve is
dynamically displayed using the data from this csv file. By default it shows Tmax vs
KExt (see the definitions below). However, data in any two columns can be plotted
against each other using the drop-down boxes at the bottom of the plot. The velocity
and temperature profiles at each solution point are stored in the opposed-
flow_flame__extinction_ExtinctionProfiles.out with the global strain rate
identifying each solution.

Figure 2-53 Output file opposed-flow_flame__extinction_ExtinctionResponse.csv

2011 Reaction Design 90 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-54 Output file ExtinctionProfilesOut.

The file opposed-flow_flame__extinction_ExtinctionResponse.csv prints four

strain rates. These are:

1. KGlobal = Global strain rate = (Fuel velocity Oxidizer velocity) / Separation

distance.
2. KExt = Extinction strain rate = Maximum axial velocity gradient on the fuel
side just before the flame.
3. KFuel = Maximum strain rate on fuel side.
4. KOxid = Maximum strain rate on the oxidizer side.
In these definitions, the fuel side refers to x = 0, i.e., left of the stagnation plane and
the oxidizer side refers to x = xmax, i.e., right side of the stagnation plane.

The dynamic plot monitor does not close after the simulation ends and so continues to
provide the information about the simulation. In addition, the Analyze Results panel
provides an Analyze Extinction Results radio button that is selected by default.
Clicking the Next Step button after selecting this radio button will allow you to access
contents of the csv file through the Post-Processor. Alternatively, the file opposed-
flow_flame__extinction_ExtinctionResponse.csv can be imported to a plotting utility

CK-TUT-10112-1112-UG-1 91 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

such as Microsoft Excel, employing the fixed-width format. Figure 2-55 shows the
plot of flame temperature vs. extinction strain rate. It can be seen that the computed
value of the extinction strain rate is about 550 s-1. The global strain rate at this point is
about 289 s-1.

Figure 2-55 Flame response curve showing extinction (turning) for premixed stoichiometric methane-air flame. The
inlet temperature is 296 K and ambient pressure is 1 atm. The calculated extinction strain rate is 550 /s.

Solution point number 425 corresponds to the turning point. Since the solution saving
frequency is 10, solution numbers 42 and 43 are the closest saved solutions.
However, the file opposed-flow_flame__extinction_ExtinctionProfiles.out can be used
to get the velocity and temperature profiles at the turning point. For this purpose, for
the current case you can search for the character string Global StrainRate [/s] = 289
and retrieve the appropriate block. Each block contains grid point number, x location
(cm), axial velocity (cm/s), and temperature (K). Plotted in Figure 2-56 are the
velocity and temperature profiles at extinction point obtained by using the data block
for global strain rate of 289.42 s-1 (use fixed-width format in Excel when loading this
data block) while Figure 2-57 shows the velocity profile along with local strain rates
obtained using first-order finite difference. The locations corresponding to various
strain rates are also shown in Figure 2-57.

2011 Reaction Design 92 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-56 Axial velocity and temperature profile for premixed stoichiometric methane-air flame at extinction point.
The global strain rate at extinction is 289 s-1

Figure 2-57 Axial velocity and velocity gradient magnitude for premixed stoichiometric methane-air flame at
extinction point. Also shown are the points corresponding to the definitions of various extinction strain
rates

CK-TUT-10112-1112-UG-1 93 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.3.11 Stagnation Flame Analysis [CHEMKIN-PRO Only]

2.3.11.1 Project Description

This tutorial describes how to use the Pre-mixed Stagnation Flame Model, which is
available under the Flame Simulators in the model palette (see Figure 2-58).
Stagnation flames are used in a variety of practical applications such as flame-based
chemical vapor deposition, polymer film processing, etc. In such processes, the
flame-generated radicals are utilized to alter the properties of the surface.

Figure 2-58 Icon for Premixed Laminar Burner-Stabilized Stagnation Flame model.

2.3.11.2 Project Set-up

The project file for the burner-stabilized stagnation flame simulation is named pre-
mixed_stagnation__MethaneAir.ckprj and is located in the samples2010 directory.
Because this is a new sample, it is possible that your "local" samples directory does
not have this sample project. If this is the case, please copy the project file named
pre-mixed_stagnation__MethaneAir.ckprj and the corresponding directory named
pre-mixed_stagnation from your installation location to the local samples directory.
(On Windows operating systems, the local samples directory is usually
C:\Documents and Settings\user\chemkin\samples2010. The installation
directory would be C:\Program Files\Reaction\15101_win64\samples2010 for
Win64 and C:\Program Files (x86)\Reaction\15101_pc\samples2010 for Win32.)

The simulation is conducted for a premix stoichiometric methane-air flame at 1 atm

pressure impinging on a wall at 800 K. The separation distance between the burner
and the wall is 1 cm. A cylindrical coordinate system is used.

A reduced (17 species) methane-air mechanism is used in this tutorial; the chemistry
set is described in (in Section 2.9.2.2). The computational domain is set to span 0.0 -
1.0 cm. The stoichiometric methane-air mixture is assigned to inlet 1 (named
MethaneAir). The inlet gas temperature is set to 300 K and the ambient pressure is 1
atm. The wall temperature is set to 800 K. For the initial temperature profile, the
guess value for maximum temperature is set to 2200 K. A mixture-averaged
formulation is used for the transport model.

2011 Reaction Design 94 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

A parameter study is set up for the inlet velocity from the burner. Inlet velocity values
of 30 cm/s and 60 cm/s are chosen. Figure 2-59 shows relevant panels from setting
up the Interface.

Figure 2-59 Parameter Study set-up.

2.3.11.3 Project Results

The laminar flame speed of a stoichiometric methane-air flame at 1 atm pressure and
300 K unburned gas temperature is approximately 38 cm/s. Note that the two inlet
velocities chosen in the parameter study lie on either side of this flame speed.

Figure 2-60 shows temperature and velocity profiles obtained for the two inlet
velocities. Run 1 corresponds to an inlet velocity of 60 cm/s while run 2 corresponds
to 30 cm/s. The flame blows off the burner for the first case and crawls upstream for
the second.

CK-TUT-10112-1112-UG-1 95 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-60 Temperature and velocity profiles for two inlet velocities

This behavior is expected, since the laminar flame speed is lower than the inlet
velocity of run 1, and higher than that of run 2. The presence of a wall creates a
boundary layer next to it and thus provides a mechanism for the flame in run 1 to
stabilize. At the wall, the bulk velocity has to vanish. The flow thus slows down (and
spreads). It can be seen that the minimum velocity just upstream of the flame for run
1 is almost equal to the laminar flame speed of 38 cm/s. As the temperature
increases, the density decreases and therefore velocity increases. Since the wall is
cooler than the flame, the gas temperature then drops down to the wall temperature.
Further increase in the inlet velocity will push the flame towards the wall. At some
critical velocity (strain rate), it will be impossible to establish a flame.

For run 2, the flame propagates upstream, because the laminar flame speed is higher
than the inlet velocity. However, in doing so there is also heat loss to the burner as its
temperature is fixed. The net result is a sharp temperature rise exactly at the lip of the
burner.

Shown in Figure 2-61 are the major species mole fraction profiles for the two
simulation runs. It can be seen that the flame structure itself is quite similar, with a
shift in the axial location. The mole fraction gradients vanish at the wall since thermal
diffusion is not included

2011 Reaction Design 96 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-61 Major species mole faction profiles for each of two simulation runs.

2.4 Internal Combustion Engine

2.4.1 Homogeneous Charge Compression Ignition (HCCI) Engine

2.4.1.1 Problem Description

Homogeneous Charge Compression Ignition, or HCCI, is an advanced combustion
technology for internal combustion engines. The main appeal of the HCCI engine is
the great potential for lowering emissions and improving fuel economy.

For this problem, we have some data from a single-cylinder HCCI test engine and
would like to use the single-zone IC engine model to simulate the same HCCI
process. We would also like to find out how sensitive the solutions are to the assumed
heat loss at the cylinder wall.

CK-TUT-10112-1112-UG-1 97 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

The HCCI engine for this example runs on natural gas, a mixture of CH4, C2H6 and
C3H8, with exhaust gas recirculation (EGR). The use of EGR helps in more reliable
ignition under a wider range of conditions. Additionally, CO2 from the exhaust gas can
keep combustion temperatures low due to its relatively large heat capacity.
Specifications of the test engine that are related to our model setup are given in
Table 2-8.

Table 2-8 Test Engine Specifications

Parameter Setting

Compression ratio 16.5

Cylinder clearance volume 103.3 cm3

Engine speed 1000 rpm

Connecting rod to crank radius ratio 3.714286

Cylinder bore diameter 12.065 cm

2.4.1.2 Problem Setup

CHEMKIN has a pre-defined model for HCCI engine simulations. The model is called
the Closed Internal Combustion Engine Simulator and we can select it either by drag-
and-dropping to the Diagram View or by double-clicking on the reactor icon in the
Models Palette (see Figure 2-62).

Figure 2-62 IC Engine Icon

The next step is to define the chemistry set, or combustion mechanism, for our HCCI
simulation. Since methane is the main component of natural gas, the mechanism we
have selected is the GRI Mech 3.0 to describe the combustion process and used
methane as the surrogate fuel. After pre-processing the mechanism data, we specify
the engine parameters in the Reactor Properties panel.

The IC Engine model is appropriate for a closed system, representing the time
between intake-valve closure and exhaust-valve opening in the engine cycle. The
start time (or start crank angle) therefore represents the time of intake-valve closure.
As a convention, engine events are expressed in crank rotation angle relative to the
top dead center (TDC). The intake valve close (IVC) time of our test engine is
142 degrees (crank angle) before TDC (BTDC). Since the GUI requires input as the
crank angle after TDC, we should set our simulation starting crank angle to 142

2011 Reaction Design 98 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

degrees (ATDC). We let the simulation run for 257 crank angle degrees to 115
degrees (ATDC). The gas mixture pressure and temperature at IVC are 107911 Pa
(or 1.065 atm) and 447 K, respectively. The composition of the initial gas mixture is a
combination of natural gas, air, and EGR gas and is given in Table 2-9.

Table 2-9 Composition of Initial Gas Mixture

Species Mole Fraction

CH4 0.0350

C2H6 0.0018

C3H8 0.0012

O2 0.1824

CO2 0.0326

H2O 0.0609

N2 0.6861

There are two different approaches to defining the heat loss through the cylinder wall.
One approach assumes the cylinder is adiabatic and the other considers heat loss
through the cylinder wall.

The project file for this HCCI engine simulation problem is called
ic_engine__hcci_heat_loss_methane.ckprj and is located in the samples2010
directory. This project file actually contains two sub-projects: the
ic_engine__hcci_adiabatic project assumes the cylinder is adiabatic and the other
project, ic_engine__hcci_heat_loss_woschni, considers heat loss through the
cylinder wall. By default CHEMKIN will append the new project to an existing project
file if the name of the new project is different from the ones already saved in the file.
Therefore, we can group similar projects into the same project file by saving those
projects one by one to the same project file name.

Parameters that describe heat transfer between the gas mixture inside the cylinder
and the cylinder wall are specified on the Reactor Physical Properties panel. For non-
adiabatic conditions, we have several ways to describe the heat loss to the wall: a
constant heat transfer rate (where a positive value corresponds to for heat loss from
the gas to environment), a piecewise-linear heat-transfer rate-profile, a user-defined
subroutine, or a heat-transfer correlation specifically designed for engine cylinders.
Here we choose the heat-transfer correlation for our HCCI problem. We also apply
the Woschni correction15 to get better estimates of gas velocity inside the cylinder.
The parameters used in the heat-transfer correlation are shown in Figure 2-63.

CK-TUT-10112-1112-UG-1 99 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-63 C1_ IC EngineReactor Physical Property for HCCI Woschni Heat Loss model

2.4.1.3 Project Results

After successfully running both adiabatic and heat-transfer-correlation IC Engine
projects, we can launch the CHEMKIN Post-Processor from the Analyze Results
panel of one the projects. We can load the solutions from the other project by using
the File > Open Solution File option of the Post-Processor Control Panel.

15. J. B. Heywood, Internal Combustion Engines Fundamentals, McGraw-Hill Science/En-

gineering/Math, New York, 1988.

2011 Reaction Design 100 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

The temperature solutions are shown in Figure 2-64. Again, a crank-rotation angle of
0 degrees corresponds to TDC. We notice that the measured gas temperature before
ignition is higher than the one predicted by the adiabatic model. If the measurement
was done correctly, this could mean that a small portion of fuel starts burning before
TDC and our model fails to capture this phenomenon. The ignition time predicted by
the adiabatic model is about 5 degrees earlier than those obtained by experiment and
by the non-adiabatic model. The temperature solution from the non-adiabatic model
generally agrees with the measurement. Both models predict stronger ignition in the
cylinder as indicated by the sharp increases in the temperature profiles. The weaker
ignition shown by the experimental data is likely due to temperature variation inside
the real cylinder. Although gas composition is homogeneous throughout the cylinder,
gas temperature in the core region can be different from that in the boundary layer.
The temperature difference can be large if heat loss at the cylinder wall is large. If the
gas mixture of the hot core region ignites first, the mean temperature inside the
cylinder will not jump as sharply as predicted by the models.

Figure 2-64 HCCI EngineEGR Temperature Comparison

CK-TUT-10112-1112-UG-1 101 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-65 gives the comparisons of the measurement and the predictions. In
general, the profiles show similar trends as observed in the temperature profiles
except that the pressure results are less sensitive to heat loss.

Figure 2-65 HCCI EngineEGR Pressure Comparison

We can examine the heat-release rate profile for the timing and the magnitude of heat
generated by combustion. Since the temperature profile predicted by the non-
adiabatic model is in good agreement with the measurement, we can determine how
much thermal energy is dissipated to the environment, i.e., the heat loss rate, during
the combustion/expansion period. The heat-release rate and the heat loss rate
profiles predicted by the non-adiabatic model are shown in Figure 2-66.Note that the
heat-release profile usually contains narrow spikes. If we want to calculate the total
heat-release from the heat release rate profile, we must make sure the profile has
enough time resolution to reduce numerical error.

Figure 2-66 HCCI EngineEGR Heat Loss Comparison

2011 Reaction Design 102 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

2.4.2 Multi-zone HCCI Engine Simulation [CHEMKIN-PRO Only]

2.4.2.1 Project Description

This multi-zone model example is based on the 2-bar boost HCCI combustion
analysis described in the 2000 SAE paper by Aceves et al.16 The supercharged
HCCI engine is running on natural gas at very low equivalence ratio (= 0.26). Engine
parameters needed to set up the multi-zone simulation are given in Table 2-10.
Composition of the natural gas is provided by Aceves et al.16 and is listed in
Table 2-11.

Table 2-10 Engine parameters of the validation case.

Displacement volume 1600 cm3

Bore 12.065 cm

Stroke 14 cm

Connecting rod length 26 cm

Engine speed 1000 rpm

Compression ratio 21

Table 2-11 Natural gas composition used in the validation case.

component CH4 C2H6 C3H8 N2 CO

Vol % 91.1 4.7 3.1 0.6 0.5

A 10-zone model is employed in the validation. Configuration and properties of the

zones are summarized in Table 2-12. The zone mass distribution recommended by
Aceves et al.17 rather than the one used by the original HCT-multi-zone simulation is
used. The distribution of zone surface-area fraction is somewhat arbitrary but, in
general, it follows the zone mass distribution. Zone temperature profiles are provided
by Aceves et al. and will be used to test the numerical stability of CHEMKIN multi-
zone model when the energy equation is switched on at the transition angle (see
Section 8.4.2 of the CHEMKIN Theory Manual). The transition angle in the present
multi-zone simulation is set at 3 degrees crank angle BTDC, the same as in the
original simulation by Aceves et al.

16. Aceves, S. M., D. L. Flowers, et al. (2000). A multi-zone model for prediction of HCCI
combustion and emissions. SAE Technical Paper 2000-01-0327.
17. Aceves, M. S., J. Martinez-Frias, et al. (2001). A Decoupled Model of Detailed Fluid
Mechanics Followed by Detailed Chemical Kinetics for Prediction of Iso-Octane HCCI
Combustion. SAE Technical Paper 2001-01-3612.

CK-TUT-10112-1112-UG-1 103 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Table 2-12 Zone configuration used by the multi-zone model analysis for the validation case.

Zone # 1 2 3 4 5 6 7 8 9 10

Region Boundary
Crevice Core
Layer

Mass % 2 1 1 1 2 5 10 18 25 35

Area % 5 5 5 10 10 15 15 15 10 10

2.4.2.2 Project Setup

The project file is called ic_engine__multizone.ckprj and data files specifying zone
temperature profiles are located in the samples2010\ic_engine\multizone directory.
For convenience, the GRI 3.0 mechanism is used to model the oxidization of natural
gas.

The icon for the multi-zone model can be found in the Models palette under Closed
0-D Reactors (see Figure 2-67).

Figure 2-67 Reactor model palette showing the multi-zone model icon in the CHEMKIN User Interface.

Multi-zone model

Figure 2-68 shows the diagram view of a multi-zone simulation project. Since the
multi-zone model is a closed reactor, the project does not require any inlet or flow
connection.

Figure 2-68 Diagram view of a multi-zone simulation project.

When a new multi-zone project is created, a dialog prompts for the number of zones
to be used in the simulation, as shown in Figure 2-69. This information is needed
once and cannot be changed after the project is created. To use a different number of
zones for the same engine simulation, a new multi-zone project must be created.

2011 Reaction Design 104 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-69 Dialog for specifying the number of zones in the multi-zone simulation.

The User Interface of the CHEMKIN multi-zone model is very similar to its single-zone
counterpart. Engine parameters such as compression ratio and parameters for the
Woschni wall heat transfer correlation are specified in the Reactor Physical Properties
panel, Figure 2-70. The multi-zone model has additional tabs in the Reactor Physical
Properties panel for configuring individual zones. Individual zone properties such as
zone mass fraction and initial zone temperature can be provided either in the Zone
Properties Table, Figure 2-71, or in the individual Zone tabs, Figure 2-72. If the
hybrid solution approach is employed, the transition time or crank angle can be
assigned by the Energy Calculation Switch in the Reactor Physical Properties
panel. Figure 2-70 shows that the energy equation will be switched on at 357 degrees
crank angle (i.e., 3 degrees BTDC).

To create a smaller solution file for the multi-zone model, go to the Output Control panel and
select the Store zone-average values only option before running.

CK-TUT-10112-1112-UG-1 105 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-70 Engine parameters in Reactor Physical Properties tab for multi-zone model.

Figure 2-71 Zone Properties table.

2011 Reaction Design 106 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-72 Individual Zone Properties tab of the multi-zone model graphic user interface using a temperature profile
called zone5

2.4.2.3 Project Results

Temperature profiles of individual zones predicted by the CHEMKIN multi-zone
simulation are shown in Figure 2-73 together with the average temperature and
temperature prediction by the single-zone model. The transition angle is set to 3
degrees BTDC as given by Aceves, et al.16 (see p. 103) The model obtains zone
temperatures for corresponding zone temperature profiles before the transition angle
is reached. The option to use the temperature profile as a constraint allows the
CHEMKIN multi-zone model to take advantage of the zone temperature history that is
computed by CFD when heat release from chemical kinetics is not significant. At the
transition angle, the model will stop utilizing the temperature profiles and begin to
solve zone temperature based on the energy balance of the zone. The Woschni wall
heat transfer correlation will be used once the energy equation is turned on. The
smooth profiles indicate the temperature governing the equation switch does not give
rise to any numerical problem. However, the transition angle seems to be somewhat
late in this case, as zone temperatures deviate rapidly after the switch. In a later
publication, Aceves et al.18 suggested that a better transition angle would be about
15.5 degrees before the TDC.

Figure 2-74 compares measured and predicted cylinder pressure profiles.

(Experimental data used in the graph for comparison are for illustration only and are
not provided in the samples directory.) The CHEMKIN multi-zone model prediction
overall shows good agreement with the experimental data. The single-zone
prediction, also shown in Figure 2-73, can match the experimental pressure curve
very well up to the ignition point by properly adjusting wall heat transfer parameters.

18.Aceves, S. M., D. L. Flowers, et al. (2005). Analysis of Premixed Charge Compression

Ignition Combustion with a Sequential Fluid Mechanics-Multizone Chemical Kinetics
Model. SAE Technical Paper 2005-01-0115.

CK-TUT-10112-1112-UG-1 107 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

The multi-zone model is able to yield good predictions of maximum cylinder pressure
and the timing of the pressure peak, while the single-zone model significantly over-
predicts the maximum pressure and the predicted peak timing is a bit too early
compared to the data.

Figure 2-73 Temperature profiles predicted by the single-zone and the multi-zone models. The transition angle is set
to 3 degrees BTDC.

2011 Reaction Design 108 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-74 Comparison of number density distributions obtained with and without aggregation model.

Figure 2-75 Zone volume profiles predicted by the multi-zone models.

Average

Average

CK-TUT-10112-1112-UG-1 109 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.5 Simulating a Shock-tube Experiment

Mechanism development often involves analyzing experimental data to understand
the chemical reactions and extract rate parameters. Shock tube experiments are often
used to obtain chemical kinetic data at high temperatures, which is especially relevant
to combustion modeling.

2.5.1 Shock-heated Air (Shock)

2.5.1.1 Problem Description

Shock tube experiments are commonly used to study reaction paths and to measure
reaction rates at elevated temperatures. We can apply the Normal Shock Reactor
Model to validate the reaction mechanism or kinetic parameters derived from such
experiments.

In this tutorial, we want to reproduce one of the shock tube experiments done by
Camac and Feinberg.19 Camac and Feinberg measured the production rates of nitric
oxide (NO) in shock-heated air over the temperature range of 2300 K to 6000 K. They
also assembled a reaction mechanism with kinetic parameters derived from their
experimental results. The reaction N2 + M = N + N + M in their mechanism has a
different temperature dependency when the third body is a nitrogen atom (N). To
properly incorporate different temperature dependencies for different third bodies, we
exclude N from participating as a third body in the original reaction, i.e., the effective
third body efficiency for N is set to zero. And we add a new reaction
N2 + N = N + N + N to explicitly address the different temperature dependence of
nitrogen atom as the third body. Figure 2-76 shows these two reactions in CHEMKIN
format.

Figure 2-76 Nitrogen Atom Reaction in CHEMKIN Format

N2+M=N+N+M 1.92E17 -0.5 224900.

N2/2.5/ N/0/
N2+N=N+N+N 4.1E22 -1.5 224900.

2.5.1.2 Problem Setup

Setting up a shock tube model requires information from the corresponding
experiment. In addition to the conditions of the initial (unshocked) gas mixture, we will
need to provide information on the diameter of the shock tube, the viscosity of the gas
at 300 K, and the velocity of the incident shock. If we do not know the shock velocity

19. M. Camac and R.M. Feinberg, Proceedings of Combustion Institute, vol. 11, p. 137-
145 (1967).

2011 Reaction Design 110 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

from the experiment, we can estimate it by using the Equilibrium Reactor Model with
the Chapmen-Jouguet detonation option. The shock tube diameter and the gas
viscosity at 300 K are only required when the boundary-layer correction is used in the
shock simulation.

The project file, incident_shock__normal_air.ckprj, is stored in the samples2010

directory and the air dissociation mechanism by Camac and Feinberg is located in the
associated working directory.

2.5.1.3 Project Results

The NO mole fraction behind the incident shock is shown in Figure 2-77 as a function
of time. The NO mole fraction profile rapidly rises to a peak value then gradually falls
back to its equilibrium level. Reasons for the greater-than-equilibrium peak NO
concentration can be found in the paper by Camac and Feinberg and references
therein. The predicted peak NO mole fraction is 0.04609 and is in good agreement
with the measured and the computed data by Camac and Feinberg.

Figure 2-77 Shock Tube ExperimentNO Mole Fraction

2.6 Combustion in Complex Flows

2.6.1 Gas Turbine Network

2.6.1.1 Problem description

This tutorial utilizes a PSR-PFR network to predict NOx emission from a gas turbine
combustor.

CK-TUT-10112-1112-UG-1 111 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.6.1.2 Problem setup

A PSR network or a hybrid PSR-PFR network is commonly used to simulate mixing
and flow characteristics of a gas turbine combustor.20,21 This reactor network
approach can greatly reduce the computational burden yet provide reasonable
solutions for a complicated combustion process. However, constructing such a
reactor network is rather empirical. Slight changes in combustor operating conditions
often lead to a new reactor network configuration with a different number of reactors
and connectivity. To speed up the time-consuming trial-and-error process, the
CHEMKIN Interface provides a Diagram View that facilitates building and modification
of a reactor network. A gas turbine reactor network is shown in Figure 2-79. Typically,
a gas turbine reactor network consists of a flame/ignition zone, a recirculation zone,
and a post-flame zone. However, depending on how the fuel and oxidizer are
delivered and the complexity of the flow field, additional reactors and inlets may be
needed to properly represent the combustor. The reactor network shown in
Figure 2-79 has two reactor network clusters. The first cluster represents the region
around the flame in the combustor and the second cluster uses a single PFR for the
post-flame region between the flame and turbine inlet. We set the first PSR as the
mixing zone because the fuel stream is partially premixed. A flame zone PSR is
directly connected to the mixing zone and is followed by a recirculation zone for back
mixing of hot combusted gas. The solutions of the through flow from the flame zone
(the last reactor of cluster number 1) are automatically fed to the post-flame zone (the
second cluster) as indicated by the gray line. Figure 2-78 shows the reactor
configuration that will be modelled.

20. A. Bhargava et al., J of Engineering for Gas Turbines and Power 122:405-411 (2000).
21. T. Rutar and P.C. Malte, J of Engineering for Gas Turbines and Power 124:776-783 (2002).

2011 Reaction Design 112 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-78 Gas Turbine NetworkSchematic

The project file, reactor_network__gas_turbine.ckprj, can be found in the

samples2010 directory. The reaction mechanism for methane-air combustion is the
GRI Mech 3.0, as described in section Section 2.9.2.

Figure 2-79 Gas Turbine NetworkDiagram View

2.6.1.3 Project Results

The solutions from the reactor network are shown in Figure 2-80 and Figure 2-81. The
temperature solutions in Figure 2-80 indicate that the gas mixture ignites in the third
reactor (flame zone) and the temperature continues to increase in the second reactor
(recirculation zone) where part of the recycled CO is consumed. Gas temperature
increases slightly in the post-flame region (the second cluster) as remaining CO is
converted to CO2 in the hot flue gas (see Figure 2-81).

The NO profiles are also given in Figure 2-81. As we expected, NO level continues to
rise in the post-flame region mainly due to the thermal NO formation.

CK-TUT-10112-1112-UG-1 113 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-80 Gas Turbine NetworkTemperature Comparisons

Figure 2-81 Gas Turbine NetworkCO and NO Comparisons

2.6.2 Jet Flame Network

2.6.2.1 Problem Description

This user tutorial describes the mechanics of constructing a PSR (perfectly stirred
reactor) cluster to represent a non-premixed jet flame.

Often, chemical kinetics is either omitted entirely or greatly reduced to make CFD
simulations possible for chemically reactive flow systems. When pollutant emissions
are to be predicted, the assumption of local chemical equilibrium is not appropriate
and the utilization of detailed reaction mechanism is needed. For example, the
characteristic chemical time scale of NO is much larger than that of a typical flame
species and is compatible to the characteristic time scale, or residence time, of the

2011 Reaction Design 114 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

flow system. Consequently, the local NO concentration level is far from its equilibrium
value and depends on the chemical state, the age, and the history of the gas mixture.
For simple flow fields, the exit concentration of the species of interest can be obtained
by simply integrating its production rates along streaklines (or streaktubes). In this
case, the detailed reaction mechanism is used to calculate production rates according
to local chemical states and can be a component of the post-processing utility.
However, this approach is not suitable for complex flow fields as strong mixing actions
and re-circulations make tracking streaklines difficult. Building a reactor network from
cold CFD solutions is a plausible approach under this situation as it can utilize the
detailed reaction mechanism while preserving some key fluid dynamic features that
are important to emission predictions such as the residence time. Discussions on how
to identify the reactors and their connectivity are beyond the scope of this tutorial.
General guidelines on deriving reactor networks from CFD solutions can be found in
papers done by Bhargava et al.20 (see p. 112) and Faravelli et al.22

2.6.2.2 Problem Setup

For this problem, we assume that a seven-reactor network is derived from local gas
composition and temperature solutions of a CH4-air diffusion jet flame simulation. The
residence times of these reactors and the connectivity and mass flow rates among
them are also obtained from the velocity solutions of the CFD simulation. The PSR
network representing the diffusion jet flame is given in Figure 2-82.

Figure 2-82 Jet Flame NetworkDiagram View

We use GRI Mech 3.0, as described in section Section 2.9.2, for the gas phase
reactions. Because we use PSR's in the network, no transport data is needed. The
chemistry set file for this project, reactor_network__jet_flame.cks, is located in the
working directory, samples2010\reactor_network\jet_flame. Since our reactor
network model does not consider molecular transport processes, i.e., diffusion and
thermal conduction, we cannot resolve the fine structure of the diffusion flame. The

22. Faravelli et al., Computers and Chemical Engineering 25:613-618 (2001).

CK-TUT-10112-1112-UG-1 115 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

chemical state of each reactor in our model will be controlled by chemical kinetics,
residence time, and compositions of gas mixtures entering the reactor, rather than by
chemical kinetics and molecular diffusion of reactants as in a fully resolved diffusion
flame.

The input parameters for properties of reactors and inlet streams can be found in the
project file, reactor_network__jet_flame.ckprj, in the samples2010 directory. Since
reactor and stream inputs are straightforward and are already described in previous
tutorials, we will focus only on the new Recycling panel here.

The Recycling panel allows us to direct a certain fraction of exit mass flow from one
reactor to other reactors in the network. The recycling fraction must be a non-negative
number and the sum of recycling fractions from a reactor must equal to one to
conserve mass.

If all exit mass flow from reactor n is going to reactor n + 1, we do not need to provide
a value for the through-flow stream, i.e., the default through-flow recycling fraction
is 1. For instance, the recycling stream Jet_Edge_2 to Developed_Zone_1 on the
Recycling panel is actually a through-flow stream from reactor C1_R3 to reactor
C1_R4 (see Figure 2-82). We can leave the text box blank and CHEMKIN will use the
default value of 1 for the recycling fraction. For other recycling streams, there is no
default value and a recycling fraction (between zero and one) must be provided in the
text box.

2.6.2.3 Project Results

The solution plots of a reactor network are difficult to interpret because the reactor
number does not necessarily correspond to its actual location in the flow field. In our
case, reactor C1_R6 actually represents the outer edge of the fuel jet and should be
physically located right next to the fuel jet nozzle in an upstream region. On the other
hand, reactors C1_R5 and C1_R7 are the flame zone and the post-flame region,
respectively, and are both located downstream from reactor C1_R6. Therefore, when
we look at the solution plots as a function of the reactor number, we should ignore the
sudden changes corresponding to reactor C1_R6.

The predicted temperature distribution is shown in Figure 2-83. The adiabatic flame
temperature of the stoichiometric CH4-air mixture is also given on the plot for
reference. We can see that reactors C1_R3, C1_R4, C1_R5, and C1_R7 have
temperatures close to the adiabatic flame temperature, which is a good indication for
the flame zone and post-flame region. The reactor temperature dropping when
coming from reactor C1_R5 to reactor C1_R7 is appropriate, since C1_R7 is in the

2011 Reaction Design 116 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

post-flame region. We can further confirm this speculation by checking the solutions
of other variables. We also think the jet flame is slightly lifted from the fuel jet nozzle
because the outer edge of the fuel jet (reactor C1_R6) stays at a relatively low
temperature.

Figure 2-83 Jet Flame NetworkTemperature Distribution

Figure 2-84 gives the CO mole fraction distribution among the reactors. Again, the CO
mole fraction at the adiabatic flame condition is provided to show where the
combustion zone ends. The CO profile shows a spike starting from reactor C1_R3 to
reactor C1_R7. We ignore reactor #6 because it does not connect either to reactor
C1_R5 or reactor C1_R7. This CO spike resembles the one we observe in a typical
flame zone and makes us believe that the flame zone ends before reaching reactor
C1_R7. The predicted CO mole fraction in reactor C1_R7 is also lower than its
adiabatic flame value and is consistent with our observation for the post-flame region.

Figure 2-84 Jet Flame NetworkCO Distribution

CK-TUT-10112-1112-UG-1 117 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

One of the advantages of using a reactor network is the ability to predict NO formation
without running a full CFD calculation with detailed chemistry. The NO profile in
Figure 2-85 shows that NO starts to form in the flame zone (reactors C1_R3, C1_R4,
and C1_R5) and continues to rise in the hot post-flame region (reactor C1_R7). We
can find out which NO formation mechanism is responsible for the NO increase in the
high temperature region. In the absence of fuel nitrogen, NO in our jet flame can be
formed via prompt NO and thermal NO mechanisms.23 Since the prompt NO
mechanism is characterized by the existence of radicals such as CH and HCN, we
can plot the profiles of these radicals to find out the region where thermal NO is the
dominant NO formation mechanism. From the profiles in Figure 2-86, we can see that
all prompt-NOrelated radicals disappear in the post-flame region (reactor C1_R7).
so we can conclude that thermal NO is the main NO formation mechanism in the hot
post-flame region. The reactor model also shows that, unlike gas temperature and CO
mole fraction, the exit NO mole fraction is far below its equilibrium value. This is in
accord with the fact that the characteristic chemical time scale of NO is greater than
the fluid mechanical time scale of our jet-flame system.

Figure 2-85 Jet Flame NetworkNO Distribution

23. Miller and Bowman, Prog. Energy Combust. Sci., 15:287-338 (1989).

2011 Reaction Design 118 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-86 Jet Flame NetworkMole Fractions

2.6.3 Using Tear Streams to Estimate Initial Gas Composition in an HCCI Engine
with Exhaust Gas Recirculation (EGR) [CHEMKIN-PRO Only]

2.6.3.1 Problem Description

This HCCI_EGR project simulates a homogeneous charged compression ignition
(HCCI) engine with exhaust gas recirculation (EGR). One of the strategies to control
ignition in an HCCI engine is to charge the cylinder with a mixture of exhaust gas and
fresh fuel-air mixture. The introduction of exhaust gas impacts ignition in several
ways. Thermodynamically, the added exhaust gas could increase the initial gas
temperature and affect gas temperature during compression by modifying mixture-
specific heat capacity. The presence of exhaust gas would also influence ignition
kinetics by dilution and by alternative reaction pathways such as NOx-mutual
sensitization.

Since the composition and temperature of recycled exhaust gas mixture are part of
the solution, they are not known prior to running the simulation. This project utilizes a
tear-stream reactor network to estimate exhaust gas properties by running the
simulation iteratively. Ideally, this iterative scheme would reach a steady-state, that
is, initial gas properties inside the cylinder do not show significant cycle-over-cycle
variations. Moreover, this kind of project could be employed to build correlations
between inlet gas properties, engine operating conditions, EGR ratio, and initial gas

CK-TUT-10112-1112-UG-1 119 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

properties inside the cylinder and to map a stable HCCI engine operation regime in
terms of EGR ratio, RPM, fuel equivalence ratio, and fuel type. The single-zone
Internal-Combustion engine model in the EGR network can be replaced by the Multi-
zone Engine Model if desired.

2.6.3.2 Problem Setup

The project is called ic_engine__EGR_network.ckprj and the combustion
mechanism is GRI 3.0 methane oxidation mechanism. This EGR_network project
contains three major components or clusters, as shown in Figure 2-87. The HCCI
Engine cluster (C1) represents the HCCI engine cylinder of which the initial condition
at intake valve close (IVC) is not well known due to external EGR. The Heat_Loss
(C3) cluster is a simplified model of the collection of pipes and valves that circulates
exhaust gas back to the intake manifold. This project assumes 30% of exhaust gas
mass gets recycled. Chemical reactions and heat transfer are allowed in the EGR
pipe system. Cluster EGR_Merge (C1) simulates the mixing of the fresh fuel-air
mixture and the recycled exhaust gas in the intake manifold. The green dash line in
Figure 2-87 represents the recycle, or tear stream of this EGR_network. The
simulation is converged (or the EGR_network system is running at steady-state)
when variations of properties of this tear stream satisfy a given criterion. The grey
dashes lines indicate that the solution from the source reactor is used as the initial or
inlet condition of the target reactor. The Gas_Splitter simply distributes the outgoing
mass flow rate without considering any chemical or heat transfer effects.

Figure 2-87 Diagram view of the EGR_network project.

The EGR_Merge reactor facilitates the merge between the fresh fuel-air mixture and
recycled exhaust gas. In this project, an open PSR model is used so that chemical
reactions would take place as soon as the gases are mixed; otherwise, a Gas Mixer
could be employed here. The Reactor Properties panel of the EGR_Merge reactor is
shown in Figure 2-88. Note that the reactor pressure should be the same as the in-
cylinder pressure of the HCCI engine at IVC.

2011 Reaction Design 120 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-88 Reactor Properties panel of the EGR_Merge reactor.

Figure 2-89 shows the inlet properties panel of the fresh fuel-air stream. Temperature,
composition, and flow rate of the fresh fuel-air mixture are specified here.

Figure 2-89 Stream Properties panel of the fresh fuel-air mixture entering the EGR_Merge reactor .

Parameters and operating conditions of the HCCI engine are entered in the
corresponding Reactor Properties panel (Figure 2-90). Initial gas temperature,
pressure, and composition are not needed here as they will be obtained directly from
the solution of the EGR_Merge reactor.

CK-TUT-10112-1112-UG-1 121 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-90 Engine Properties panel of the HCCI engine model.

The main purpose of the Heat Loss reactor is to provide a way to cool down the hot
exhaust gas, so it is important to provide suitable values for residence time, heat
transfer coefficient, internal surface area, and ambient temperature (Figure 2-91). The
solution from the HCCI engine simulation will supply the temperature and composition
of the inlet mixture. By default, values of the last time point in the solution file will be
used. Since the HCCI engine is a closed model, i.e., it has no outlet mass flow rate,
the mass flow rate of the inlet must be provided and the Override Velocity with User
Supplied Data should be checked, as shown in Figure 2-92. The value of mass flow
rate must be chosen carefully so that the mass flow rate ratio of the recycled exhaust
gas and the fresh fuel-air mixture gives the correct EGR ratio. In this case, the
Heat_Loss reactor has a mass flow rate of 10 g/sec (see Figure 2-93) and the split
ratio is 0.3 (see Figure 2-94), and therefore the mass flow rate of the recycled exhaust
gas is 3 g/sec. Since the mass flow rate of fresh fuel-air mixture to the EGR_Merge
reactor given in Figure 2-89 is 7 g/sec, the mass ratio of the fresh fuel-air mixture and
exhaust gas is 7:3 or 30% EGR by mass. Double-click on the Splitter node on the
project tree to activate the Split Factor panel, and then enter the split factor
(Figure 2-94).

2011 Reaction Design 122 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-91 Properties of the Heat_Loss reactor.

Figure 2-92 Initialization panel of the inlet stream to the Heat_Loss reactor.

Figure 2-93 Properties of the inlet stream to the Heat_Loss reactor.

CK-TUT-10112-1112-UG-1 123 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-94 Split Factor panel for the Gas Splitter.

Convergence of this iterative scheme is determined by properties of the tear stream.

Convergence criteria can be supplied in the Tear Stream panel, which is activated by
double-clicking on the Tear Stream Controls node on the Project tree. Figure 2-95
shows the Tear Stream Controls panel along with the default values. It is possible to
speed up the convergence process of the iterative scheme by providing hot initial
guess values for the Tear Stream. The initial guesses are entered on the Initialize
Tear Streams tab of the Tear Stream Controls panel.

Note that the Initialize Tear Streams functionality is optional. However, when it is applied,
guess values of all stream properties must be provided. Every time the Initialize Tear Streams
tab is clicked, a reminder will show up (see Figure 2-96). Figure 2-97 shows the Tear Stream
Properties panel where guess values could be given.

Figure 2-95 Tear Stream Controls panel.

Figure 2-96 Popup reminder for the initialize Tear Streams panel.

2011 Reaction Design 124 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-97 Panel for entering guess properties of the Tear Stream.

After the project is set up, it is run like other CHEMKIN projects. The only difference is
that, in the Detail View, Clusters belonging to the same tear stream will be grouped
together and highlighted, as shown in Figure 2-98. When a project containing a tear
stream is run, the Progress Monitor will provide text messages and two graphical bars
indicating progress of the current iteration step and overall convergence.

Figure 2-98 Detail view of the calculations in which clusters of the same Tear Stream are grouped together and
highlighted.

Figure 2-99 Project Results

The simulation results can be processed by clicking on the Analyze Results node on
the project tree. Since the main interest of this project is the solution of the HCCI
engine, select IC_Engine (C2) in the Solution to View box and click the Begin
Analysis button as shown in Figure 2-100. This will open the Graphical Post-
Processor for plotting and/or exporting results from the HCCI engine simulation.

CK-TUT-10112-1112-UG-1 125 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-100 Solution selection panel.

Plots comparing steady state results of the current HCCI Engine simulation (30%
EGR) and those without EGR are given in Figure 2-101 to Figure 2-104.

Figure 2-101 Comparison of EGR effect on in-cylinder pressure. Solid line: no EGR; Dash-dotted line: 30% EGR by
mass.

Figure 2-102 Comparison of EGR effect on gas temperature. Solid line: no EGR; Dash-dotted line: 30% EGR by
mass.

2011 Reaction Design 126 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-103 Comparison of EGR effect on CO2 emission. Solid line: no EGR; Dash-dotted line: 30% EGR by mass.

Figure 2-104 Comparison of EGR effect on NO emission. Solid line: no EGR; Dash-dotted line: 30% EGR by mass.

2.6.4 Partially Stirred Reactor for Methane/Air

2.6.4.1 Problem description

In this tutorial, we want to explore effects of imperfect mixing on the exit flow condition
from a premixed gas turbine combustor.

When chemical kinetics is the limiting factor of the reacting system under
investigation, we would model the reactor as a perfectly stirred reactor (PSR). The
perfect mixing of reactants and products inside a PSR is usually accomplished by
using mixers or multiple-jet injections. However, a gas-turbine combustor normally

CK-TUT-10112-1112-UG-1 127 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

does not have these mixing mechanisms and has to rely on fluid motions, i.e., large-
scale eddies and turbulence, to provide the necessary mixing actions. Local
turbulence is particularly important as it promotes micro-scale mixing among the gas
species.

Although it is adequate in many cases to treat the gas-turbine combustor as a PSR,24

this PSR approach does not always yield proper predictions for the combustor outlet
condition. One of the factors that can cause the failure of a PSR approach is the
interaction between the turbulence (micro-mixing) and chemical kinetics. If the
turbulence is too weak to provide fast mixing among the gas species, the micro-
mixing process will interfere with the chemical kinetics. In some cases, when the
reactants (non-premixed cases) or the reactants and the products (premixed cases)
fail to mix microscopically before they are blown out of the combustor, no combustion
zone can be established inside the combustor. The partially stirred reactor (PaSR)
model is a tool that can be used to assess the extent of turbulence-kinetics interaction
in a gas-turbine combustor or to provide information on how turbulence intensity will
affect the combustor.

2.6.4.2 Problem Setup

The project file for this sample problem is located in the samples2010 directory and is
called pasr__ch4_air.ckprj. A skeletal methane/air combustion mechanism is used
to speed the calculation, since we are interested in knowing whether combustion can
be sustained inside the combustor by turbulence mixing.

Molecular mixing is important to this problem despite the fact that the fuel and air are
premixed before entering the combustor. For premixed problems, good and fast
mixing between the fresh reactants and the burned products is required to anchor the
combustion zone inside the combustor. To provide a good starting point for the back
mixing, we need to initialize the PaSR with the burned state. This is similar to starting
a gas-turbine combustor with a pilot flame. We can use the equilibrium model or the
steady state PSR model to obtain the burned state of the premixed fuel-air mixture.
Effects of the initial condition on the solutions will be minimal once the simulation time
passes the residence time of the combustor.

There are several model parameters that are unique to the PaSR model. First, we
need to specify how the PaSR will treat chemical reactions. In the current case, we
want to use finite rates defined by the reaction mechanism. Secondly, we have to
select a Monte Carlo mixing model and define parameters for the mixing model. The
choice of mixing model depends on how we envision the mixing process will behave

24. T. Rutar and P.C. Malte, J of Engineering for Gas Turbines and Power, 124:776-783
(2002).

2011 Reaction Design 128 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

in the combustor. In this case, we choose the modified Curl's model because we think
turbulence eddies in our combustor have a relatively large size variation. The mixing
time on our panel (0.0001 sec) is actually the mechanical time scale of the turbulence,
or the large eddy turnover time. The factor for mixing models entry is the scaling
factor between the mechanical and scalar time scales and is usually set to 2. We also
need to specify the time step size for the Monte Carlo simulation. The time step size
should be no greater than the mixing time scale. Finally, define the size of our PaSR
simulation. The solution time profiles will be smoother if we use more statistical
events, or particles, in the simulation. However, the run time and memory requirement
also increase with the number of statistical events in the ensemble.

There are several solvers available for the PaSR model. Usually, we want to pick
between the default DDASPK solver and the DVODE solver with backward
differencing. The DDASPK solver is more reliable but more time consuming. Here we
use the DVODE solver because the chemistry is simple.

In addition to providing the time profiles of the mean and root-mean-squared (rms)
values of scalar variables, the PaSR model can generate probability density functions
(pdf) of scalars in separate output data files. The pdf profile shows the instantaneous
distribution of a scalar at the end of the simulation time. The shape of a pdf profile is a
result of the turbulence-chemistry interaction. The pdf will become a delta function (a
spike) if the PaSR behaves closely to a PSR. We can also use the pdf's to find out all
possible states inside the combustor. Here we would like to examine the pdf of gas
temperature at the end of the simulation time. We want the temperature pdf to have a
resolution of 100 intervals, or number of bins. The Output Probability Distribution of
Scalar tab of the Output Control panel provides the setup parameters for this pdf
output, which is set with T = 100. The temperature pdf profile will be saved to a file
called pdf_T.plt and we can import this file to the Post-Processor later.

We want to compare solutions from two cases with different mixing time scales so we
can find out how turbulence intensity would affect our premixed combustor. The first
case is a relatively strong mixing case with a mixing time of 0.1 msec. The weak
mixing case will have a mixing time of 1 msec. After we finish running the first case,
we have to rename the solution files, XMLdata.zip and pdf_T.plt, to XMLdata_0.1.zip
and pdf_T_0.1.plt, respectively, so they will not be overwritten by the second run. To
switch to the weak mixing case, we can simply go back to the Reactor Properties
panel, C1_ PaSR, and increase the mixing time to 0.001 sec. We can then go to the
Run Calculations panel and run the second case.

CK-TUT-10112-1112-UG-1 129 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.6.4.3 Project Results

The mean and rms gas temperature are shown in Figure 2-105 and Figure 2-106 for
mixing times of 0.1 msec and 1 msec. The mean temperature profile of the strong
mixing case (mixing time = 0.1 msec) indicates that a combustion zone is established
inside the combustor and the outlet temperature is around 1800 K. On the other hand,
the mean outlet temperature of the weak mixing case continues to drop because poor
mixing between the fresh gas and the burned products fails to stabilize a combustion
zone. The rms temperature profile of the weak turbulence case has a higher peak
value than that of the strong mixing case. This large temperature variation indicates
that part of the gas mixture inside the combustor does not burn in the weak mixing
case. As most of the initial burned products get pushed out of the combustor, the
statistics contain more and more non-burned events and the temperature variation
starts to decrease consistently. Statistics of the strong mixing case are mainly made
up of burning states so that the temperature variation is relatively small and stable.
Similar trends are also shown by the mean CO mole fraction profiles in Figure 2-107.

Figure 2-105 PaSR Methane/AirTemperature Comparison

2011 Reaction Design 130 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-106 PaSR Methane/AirTemperature Variance Comparison

Figure 2-107 PaSR Methane/AirCO Comparison

Since we requested the temperature pdf to be saved to a data file, we can view the
pdf profile by importing it into the Post-Processor. To import the pdf profile generated
by the PaSR, we use the Import command in the File menu of the CHEMKIN Post-
Processor. After we select the data file, pdf_T_0.1.plt, we should instruct the Post-
Processor to skip the first two lines of the text and to read the third line as column
titles as shown in Figure 2-108. We also need to set the column delimiter to space.
We can repeat the same procedures for the second file containing the temperature
pdf of the weak mixing case. Once we have both pdf's imported to the Post-
Processor, we can plot these two profiles together for easy comparisons. As shown in
Figure 2-109, the pdf of strong mixing case peaks around 2000 K with a small tail
over temperatures slightly lower than the peak value. We can conclude that a stable
combustion zone is established in the strong mixing combustor. The weak mixing
case, however, has two peaks in its temperature pdf profile. A smaller peak is located

CK-TUT-10112-1112-UG-1 131 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

near 2000 K suggesting there is still some initially burned gas mixture left in the
combustor. The large peak of the temperature pdf indicates the combustor most likely
has a temperature of 500 K which is the temperature of the inlet gas mixture.
Although there might be limited chemical reactions, the slow mixing process cannot
sustain a combustion zone inside the combustor.

Figure 2-108 PaSR Methane/AirSelect the Import File Format

Figure 2-109 PaSR Methane/AirPDF(T) Comparison

2.6.5 Side Inlet on a Plug Flow Reactor

2.6.5.1 Project Description

This tutorial illustrates the use of the side inlet flow into a plug flow reactor, to
approximate the effect of a mixing region on product composition. The chemistry used
by this project is the Lawrence Livermore National Laboratory NOx mechanism25.
Ammonia as a product of a nitrogen fuel contributes to the production of NO in post-
flame oxidation. This model predicts how a mixing region affects the production of NO
in a high-temperature oxidizing region that is post-flame.

2011 Reaction Design 132 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

2.6.5.2 Project Setup

The project file is called pfr__side_inlet_param_study.ckprj. All data files are found
in the samples2010\pfr\side_inlet_parameter_study directory. The simulated
model is based on a Lawrence Berkeley National Laboratory experimental setup26.
Their report includes a description of a concentric tube, with the ammonia stream in
the center, and the oxidizing stream around the edges, as illustrated in Figure 2-110.

Figure 2-110 Illustration of the System to be Modeled.

The two streams have the following composition:

Table 2-13 Components in Fuel and Oxidizer Streams

Component Stream a (fuel) Stream b (oxidizer)

N2 0.974 0.88

CH4 2.E-3 0

NH3 6.E-4 0

O2 0 0.08

H 2O 0 0.04

25. Marinov, N. M., Pitz, W. J., Westbrook, C. K., Hori, M., and Matsunaga, N. An Ex-
perimental and Kinetic Calculation of the Promotion Effect of Hydrocarbons on the NO-
NO2 Conversion in a Flow Reactor, Proceedings of the Combustion Institute, Volume 27,
pp. 389-396, 1998. (UCRL-JC-129372).
26. Grcar, J. F., Glarborg, P., Bell, J.B., Day, M. S., Loren, A., Jensen, A. D. Effects of Mix-
ing on Ammonia Oxidation in Combustion Environments at Intermediate Temperatures,
Lawrence Berkeley National Laboratory report LBNL-54187.

CK-TUT-10112-1112-UG-1 133 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

A length of 3 cm in the mixing region is an estimate based on CFD calculations26 (see

p. 133)
. The temperature of the gas was varied between the range of 1000 to 1500 K in
a parameter study. The diameter profile and the flow rate profile vary in the 3-cm
mixing region. The fuel and oxidizer streams have a 1:1 flow ratio, with their sum
being a value of 2 l/min, which is 16.667 cm3/s for each stream. The oxidizer stream
volumetric flow rate is brought up from zero to 16.6667 over the 3-cm mixing region.

2.6.5.3 Project Results

Figure 2-111 shows the increase in mass attributed to the effect of the side inlet flow
when the maximum X value in the Post-processor is constrained to 3.0 cm (see the
CHEMKIN Visualization Manual).

Figure 2-111 Total Flow Growing Directly as a Result of Side Inlet Flow.

At the exit, the predicted NO concentration has a peak near 1250 K, similar to
experimental results26.

To create the following plot in the CHEMKIN Post-Processor, select PFRC1 End Point
vs. Parameter for the Plot Set and then select Mole Fraction NO End Point for the Y axis.

2011 Reaction Design 134 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-112 Peaking of NO Concentration Dependent on Temperature in a PFR with an Approximated Mixing Region.

Mixing causes a drop in the peak of NO and a lowering shift in the corresponding
temperature, similar to the results found in the report. The NO peak shift is attributed
to the locations of the reactions in the mixing zone. At temperatures higher than
1300 K, reaction occurs quickly when the oxidizing stream is in low concentrations,
thereby decreasing the amount of NO. When the temperature is in the 1200-1300-K
range, the reactions occur later in the mixing zone, a location with more oxidizer and
consequently more NO. At temperatures lower than 1200 K, the formation of NO is
not favored, so the NO concentration drops off. This peak behavior is illustrated in
Figure 2-112.

2.6.6 Co-flowing Non-premixed CH4/Air Flame

2.6.6.1 Problem Description

By default, the cylindrical shear flow reactor model uses the uniform inlet and initial
profile for all variables except axial velocity. The default inlet velocity profile is
considered to be fully developed, i.e., parabolic. In this example, we will establish a
co-annular flow inlet condition by using the user routine option to override the default
inlet profiles. Properties of both jets are assumed to be uniform when they enter the
reactor so there is a jump in the inlet profiles at the jet interface. A non-premixed
flame will be established downstream as fuel and air are mixed due to diffusion. The
co-flowing annular jet configuration is shown in Figure 2-113 and properties of the
inner (fuel) and the outer (air) jets are given in Table 2-14. The outer wall is adiabatic.

CK-TUT-10112-1112-UG-1 135 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-113 Confined Co-flowing Annular Jet Configuration.

Table 2-14 Properties of the Co-flowing Jets

Inner Jet
Outer Jet (Air)
(Fuel)

Radius (cm) 0.8 4.0

Velocity (cm/sec) 10 25

Temperature (K) 600 1000

H2 Mass Fraction 0.05 0

CH4 Mass Fraction 0.45 0

O2 Mass Fraction 0 0.2329

N2 Mass Fraction 0.5 0.7671

2.6.6.2 Problem Setup

The project file, cylindrical_shear_flow__profile.ckprj, is located in the
samples2010 directory. The FORTRAN subroutine used to set up inlet profiles is
called CRUPROF and can be found inside the user routine file creslaf_user_routines.f
in the user_routines directory. The GRI Mech 3.0, described in Section 2.9.2, is used
for the gas phase combustion chemistry. No surface chemistry mechanism is needed
because the outer wall is chemically inert.

The user routine options are not shown in the graphical User Interface by default. To
make these options available from the User Interface, the Display User Routine
Options box in the Preferences panel must be checked. Figure 2-114 shows a typical
Preferences panel and the Display User Routine Options is located near the bottom of
the panel.

2011 Reaction Design 136 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-114 Preferences Panel Showing the Display User Routine Options Is Enabled.

Once the Display User Routine Options is enabled, the Get Solution Profile from User
Routine option will become available when the Reactor Physical Property panel is
opened. The check box for the Get Solution Profile from User Routine option is
located near the bottom of the Reactor Physical Property panel as shown in
Figure 2-115. Once the box is checked, the reactor model will use the initial profiles
defined by the CRUPROF subroutine (in
chemkin<version>_xxx\user_routines\creslaf_user_routines.f) instead of the
default ones. The other way to activate the CRUPROF subroutine is by including the
keyword UPROF in the reactor model input file cylindrical_shear_flow__profile.inp. In
addition to the user routine option, other reactor parameters such as the number of
grid points in the radial direction, reactor pressure, reactor radius (radius of the outer
pipe wall), and ending axial position, must be specified in this panel.

The inlet velocity and temperature are entered in the Stream Property Data panel.
These two inlet parameters are required even though their values will be overridden
by the user routine later. Note that the cylindrical shear flow reactor model will set the
outer wall temperature to the inlet stream temperature when the wall is not chemically
active (no surface chemistry). If the inlet temperature given in the Stream Property
Data panel is different from that of the outer jet (as defined in the user profile routine
CRUPROF), a thermal boundary layer will be developed next to the outer wall.

CK-TUT-10112-1112-UG-1 137 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-115 Get Initial Solution Profile from User Routine Check box.

2.6.6.3 Project Results

After the simulation has completed successfully, contour plots of solution variables
can be visualized in the CHEMKIN Post-Processor. Since this is an axisymmetric
problem, the reactor model only solves the top half of the flow domain. To visualize
the solutions of the entire physical domain, we can have the CHEMKIN Post-
Processor reconstruct contours in the bottom half of the physical domain by
mirroring the contours with respect to the x-axis. This operation is activated by
selecting About the X-Axis under the Reflect Contour Data option, as shown in
Figure 2-116.

2011 Reaction Design 138 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-116 Visualization of Full Computational Domain Can Be Acquired by Selecting Proper Reflect Contour Data
Option in the CHEMKIN Post-Processor Panel.

Once the 3-D contour plot of the full computational domain is displayed in a new plot
window, we can add a colorbar to label the contours. We first select the 3-D contour
plot to be labeled by right-clicking the mouse over the plot. Then, we click on the
Insert button on the menu bar to make the pull-down list available (see Figure 2-117).
Select Colorbar to insert a contour level legend. By default, the colorbar is horizontal
and is shown at the middle lower area of the plot. We can customize the colorbar by
adjusting the colorbar properties. Double-clicking on the colorbar will open the
Colorbar property panel. We will change Orientation from Horizontal to Vertical, set
the Text show to True, and type Temperature (K) in the text field next to the Title.
Figure 2-118 shows the final look of the Colorbar property panel after all the changes.
The contour plot will be updated as changes in the Colorbar property panel are made
so we do not need to close the property panel. Finally, we can re-position the colorbar
on the plot by dragging it with the right mouse button pressed. The finished 3-D
temperature contour plot is shown in Figure 2-119. We can print the contour plot
directly to a printer or copy-and-paste it to a document.

CK-TUT-10112-1112-UG-1 139 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-117 Contour Plot Window Showing the Pull-down List Selections of the Insert Menu.

Figure 2-118 Colorbar Property Panel.

2011 Reaction Design 140 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-119 Final 3-D Temperature Contours of the Entire Physical Domain with Vertical Contour Legend (Colorbar).

2.7 Particle Tracking Feature [CHEMKIN-PRO Only]

2.7.1 Soot Formation and Growth in a JSR/PFR Reactor [CHEMKIN-PRO Only]
The JSR/PFR system developed at MIT27 provides a good platform for kinetic studies
of soot formation and growth because, under the JSR and PFR conditions, the
influence of mass diffusion on gas phase species profiles is minimized. The JSR
serves as the pre-heat and flame zones of a premixed flame and the PFR is used to
simulate the postflame region. From the prospect of model simulation, the JSR/PFR
implementation greatly reduces the complexity of the numerical process as well as
the run time. A schematic of the JSR/PFR configuration is shown in Figure 2-120.

As an example to assess the performance of the Particle Tracking Feature, modified

PSR and PFR models are employed to simulate one of the premixed C2H4/O2/N2
experiments by Marr27. The experimental data include mole fractions of major gas-
phase species and PAH's and soot mass concentration at various locations inside the
PFR. Therefore, this data set is useful not only to validate the Particle Tracking
Feature but to provide insights on the kinetics of nucleation and soot growth surface
reactions used to simulate the experiment.

27. J.A. Marr, PhD. Thesis, Dept. of Chemical Engineering, MIT (1993).

CK-TUT-10112-1112-UG-1 141 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-120 A schematic of the JSR/PFR reactor configuration used by Marr27

2.7.1.1 Problem Setup

The CHEMKIN project file for this tutorial is named
reactor_network__soot_JSRPFR.ckprj and is located in the samples2010
directory. The reaction mechanisms for ethylene/air combustion and soot formation
and growth are described in Section 2.9.

The JSR/PFR experiment shown in Figure 2-120 can be modeled by one PSR and
two PFRs in series. The first PSR is for the upstream (or flame zone) JSR, the
following PFR is to model the transition piece between JSR and PFR in the
experimental setup, and the last PFR is for the postflame PFR where measurement
was performed. The main purpose of the transition PFR is to allow the JSR exhaust to
cool down from 1630K to 1620K before entering the test section. The diagram view of
this three-reactor network is given in Figure 2-121. Since the Particle Tracking
Feature is activated by special keywords in surface reaction mechanism, all soot
simulations will need both gas phase and surface chemistry input files. Once the
chemistry files have been pre-processed, the Dispersed Phase tab will appear in the
Reactor Physical Properties panel. Most parameters for the Particle Tracking Feature
can be assigned in this Dispersed Phase tab. Initial conditions of the particle size
moments can also be specified here. The initial size moments can be constructed
from particle number density alone. Additional particle size information such as
particle mass density or particle volume fraction can also be specified. Because the
dispersed phase does not exist on the reactor wall, the surface area fraction of the
particle material must be set to zero in the Material-specific Data tab. Here, we set
Carbon to 0.0 and Wall to 1.0. (Parameters for all materials can be specified on the
Reactor Physical Properties tab.)

2011 Reaction Design 142 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

If particles exist in the inlet streams, their size moments can be provided in the
Dispersed Phase tab of the Stream Properties panel.

The values of high particle-size moments can become very large so sometimes the
absolute tolerance suitable for species mass fractions might not work well for those
high moments. Therefore, the Particle Tracking Feature allows the tolerances for size
moments solutions to be given explicitly for the steady state PSR model. The
tolerances for particle size moments can be specified in the Solver window, as shown
in Figure 2-123.

Once all model parameters, initial/guess conditions and inlet stream properties are
set, the JSR/PFR simulation can be launched from the Run Calculations window like
any other CHEMKIN project.

Figure 2-121 Diagram View of the CHEMKIN Project Used to Simulate the JSR/PFR Experiment

Figure 2-122 Specifying Particle Tracking Feature Parameters and Initial Conditions of Particle Size Moments in the
Reactor

CK-TUT-10112-1112-UG-1 143 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-123 Tolerances for Particle Size Moments Can Be Given Explicitly in the Solver Window

2.7.1.2 Project Results

Two simulations, one with both H-abstraction-C2H2-addition (HACA) and PAH
condensation growth reactions and the other with HACA growth reactions only, are
performed so that contributions of either soot mass growth mechanism can be
identified. When the number of variables available for plotting exceeds a preset limit,
the CHEMKIN post-processor presents an option to filter the variables to a smaller
number. For example, A1 represents benzene, A2 represents naphthalene, A1C2H
represents phenyl acetylene, A4 represent pyrene, and so on.

Results of the TJSR = 1630K and = 2.2 case are presented in Figure 2-124. Use
PFR(C3) for analysis. As can be seen from these figures, predictions obtained by the
Particle Tracking Feature are in good agreements with experimental data. The
Particle Tracking Feature in general slightly underpredicts gas-phase species. There
are many factors that can contribute to the discrepancies shown in the figures. For
example, Marr did not provide details composition and temperature of the inlet gas
mixture to the JSR. Since the temperature of the JSR is maintained by adjusting the
N2 fraction in the inlet gas stream, uncertainties in inlet condition, reactor heat loss,
and reactor residence time will surely affect the simulation results in the PFR section
behind the JSR.

Comparison of the predicted and measured soot mass concentration profiles in the
PFR is presented in Figure 2-125. While the HACA-only mechanism shows an
excellent agreement with the data at the PFR inlet, the slope of the soot mass profile
predicted by the HACA-only mechanism (dash-dot line) is much smaller than that of
the experimental profile. This is an indication that the HACA growth mechanism alone
gives a too-slow soot mass growth rate in the post-flame region. Since the present
soot model underpredicts C2H2 mole fraction in the PFR (Figure 2-124), it is possible
that the lower C2H2 concentration leads to lower HACA soot mass growth rate. It is

2011 Reaction Design 144 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

also possible that another growth mechanism, possibly PAH condensation, might
contribute equally to soot mass growth under this condition. The soot mass growth
rate predicted by the HACA + PAH mechanism (solid line), on the other hand, shows
a much better agreement to the experimental data that the HACA only mechanism
does. Since the sticking coefficients of all the PAH considered here are within the
range suggested by Marr27, the PAH contribution to soot mass growth should be
reasonably predicted by the model. However, the HACA + PAH mechanism does
overpredict the soot mass density at the PFR inlet. Note that experimental data
indicate that soot mass density increases by about 4 x 10-8 gm/cm3 for the first
5 mini-seconds in the PFR. Since the residence time in the JSR is about 5 mini-
seconds and the temperature in JSR is only 10K higher than that of PFR, the soot
mass density at the PFR inlet should be higher than the measured value. Of course,
this assessment is based on the assumption that soot particles start to grow once
they are created inside the JSR.

The evolution of average soot particle diameter inside the PFR is shown in
Figure 2-126. The soot particle diameter increase along the plug flow reactor due to
particle coagulation and mass growth. Note that the average particle diameter actually
drops a little near the PFR entrance. This signifies that soot nucleation is still
occurring as the gas mixture entering the PFR.

Figure 2-124 Comparisons of Mole Fraction Profiles of Selected Gas Phase Species Inside the PFR for the 1630K and
= 2.2 Case of the C2H4/O2/N2 JSR/PFR Experiment by Marr27. Symbols: data; Solid lines: predictions
with HACA and PAH condensation growth mechanisms

CK-TUT-10112-1112-UG-1 145 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-125 Comparisons of Soot Mass Concentration Profiles Inside the PFR for the 1630K and = 2.2 Case of the
C2H4/O2/N2 JSR/PFR Experiment by Marr27. Symbols: data; Solid line: prediction with both HACA and
PAH condensation growth mechanisms; Dash-dot line: prediction with HACA growth mechanism only

Figure 2-126 The Particle Diameter Evolution Inside the PFR Predicted by the Present Soot Module for the 1630K and
= 2.2 Case of the C2H4/O2/N2 JSR/PFR Experiment by Marr27 (see p. 141)

2011 Reaction Design 146 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

2.7.2 Soot Particles in Flame Simulators [CHEMKIN-PRO Only]

2.7.2.1 Project Description

This tutorial presents the use of Particle Tracking in the Flame-speed Calculator and
in the Opposed-flow Flame Simulator. The segregated solution technique described
in Section 18.9 of the CHEMKIN Theory Manual is used for flame/particle simulations.
In this tutorial, however, note that the aggregation model is turned off for this example.

Simulations of flames with particle formation and growth can be CPU-intensive,

particularly for sooting hydrocarbon flames. This is due to the relatively large gas-
phase mechanisms required to include soot-formation precursors, as well as the
added computation of tracking the particle formation, growth, and size distribution.
Since most of the computational cost is usually due to the gas-phase reacting-flow
solution, a common technique is to solve the gas-phase-only flame simulation first,
and then perform the flame-plus-soot simulation by "restarting" from the solved gas-
only flame. This approach is recommended for the following types of particle
simulations:

1. When the gas-phase mechanism is large.

2. When simulations are being used in parametric analysis that might involve
changes to just the surface (soot) chemistry.
3. When difficulties in convergence are encountered for the coupled problem and
some experimentation is required with solver parameters.
In general, the default solver settings are satisfactory for most cases. However,
convergence problems can occur. For example, convergence can be particularly
difficult in cases when the particle-flame interaction is very strong. In general,
restarting from the gas-phase solution improves the convergence behavior for the
coupled solution and is therefore recommended. In this tutorial we will demonstrate
that technique.

2.7.2.2 Project Set-up

The project files for soot simulation with the Flame-speed Calculator and Opposed-
flow Diffusion Flames are named as flame_speed__soot_particles.ckprj and
opposed-flow_flame__soot_particles.ckprj, and are located in the samples2010
directory. Both flame simulations use ethylene as the fuel and air as the oxidizer. The
gas-phase and surface reaction mechanisms described in Section 2.9 are used. In
these simulations, only the HACA (hydrogen abstraction, carbon addition) part of the
mechanism is used in the surface chemistry input.

CK-TUT-10112-1112-UG-1 147 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

In the Flame-speed Calculator (flame_speed__soot_particles.ckprj) project, we

consider a freely propagating ethylene (C2H4) - air flame with equivalence ratio of 3.0.
The computation domain is set between -2 to 5 cm. The inlet gas temperature is set
to 300 K and the ambient pressure is 1 atm. An initial temperature profile is selected
with fixed temperature of 1200 K located at x = 0, thus effectively anchoring the flame
at x = 0.

For the Opposed-flow Flame simulation (opposed-

flow_flame__soot_particles.ckprj) project, pure ethylene is on the fuel side and air
on the oxidizer side with a separation distance of 1 cm. The velocity from each nozzle
is set to 10 cm/s thus giving a global strain-rate of 20 s-1. The gas temperature from
both inlets is set to 300 K and ambient pressure is 1 atm.

Both simulations use mixture averaged transport with correction velocity formulation.
For the soot particles, we solve for first three moments.

2.7.2.2.1 Controls on the Dispersed Phase Tab

After the chemistry sets have been pre-processed, the Dispersed Phase tab appears
in the Reactor Physical Properties panel. This tab contains two sub-tabs: (a) Basic,
and (b) Aggregation model. The former can be used to set up all the basic level
controls such as number of moments, a switch for including thermophoresis, while the
latter is exclusively for the aggregation parameters.

Basic Sub-tab

The particle moments and surface species concentrations span many orders of
magnitude. For a typical flame particle simulation, these values when expressed in
the units of number/unit volume can be between 0 to 1016. The usual choice for units
of mole/cm3 is often too large and can create problems in ensuring "positivity" for
these values. For example, in the double-precision representation of the numbers,
near-zero negative values can be on the order of 10 to 20. If the particle moments
and surface species concentrations are in "moles/cm3", the zero-th moment can be
-10 to -20 at a point in the solution domain. While such a number is clearly of an
appropriate order of precision, it means that there are 1000 (= zero-th
moment*Avogadro number) particles at that location. To alleviate this difficulty, the
two controls, Scaling Factor for Moments and Scaling Factor for Surface
Species, can be used. A value of 109 for these control parameters means CHEMKIN
internally uses "nano-moles/cm3" as the unit for moments and surface species
concentration. The default values are 1012. The choice of the scaling factor should
be based on some estimate of the maximum value for number density of the particles.
In typical flame simulations, this is about 109 to 1014, leading to the choice of default.

2011 Reaction Design 148 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

For the opposed-flow sample, a non-default value of 1e+15 is used for both scaling
factors. This choice is motivated by the small amount of soot formed in this flame.
Other controls, such as number of moments, coagulation regime, etc., are as
explained in Section 2.7.

Aggregation Sub-tab

The aggregation-related parameters are described in the CHEMKIN Input Manual and
Section 18.8 of the CHEMKIN Theory Manual.

2.7.2.2.2 Solver Options

The segregated solution scheme used in flame-particle simulations checks iteration
convergence by monitoring the differences in the gas-phase mole fractions. The
corresponding convergence criterion is given by the parameter Absolute Tolerance
in Segregated Scheme on the Basic tab of the Solver options panel. Specifically,
after updating the particle source terms the gas-phase equations are solved and the
resulting mole fractions are compared against those obtained from the previous
iteration. If the maximum absolute change in the mole fraction is below the
convergence parameter, the iterations stop and the solution is complete; otherwise
new particle source terms are computed and the procedure is repeated.

Advanced Solver Options Tab

Similar to the small negative "floor" value specified for the gas-phase species mole
fractions, there are two options available for setting bounds: Minimum Bounds on
Surface Species Concentration and Minimum Bounds on Particle Moments.
The default value for both these controls is zero. While it is acceptable to set the
former to a small negative number, the particle moments should not be set to a
negative number. This restriction is not only due to the physical constraint but also
due to the fact that a logarithmic interpolation scheme is used when computing the
fractional moments. The lower bound on the zero-th moment, i.e., number density, is
fixed at zero. The bound control thus applies only to higher moments. Also note that
these values are applied to the "scaled" variables, i.e., to the actual solution variable
and not to the mole/cm3 or number/cm3 values. For example, if the scale factor is
specified as 1.0E+09, then for the value of -1.0E-06 for the lower bound, the search
can explore a solution space with -10-6 nanomoles.

The maximum number of iterations performed by the segregated scheme is

controlled by the parameter Maximum Number of Iterations in Segregated
Scheme. The default value of this parameter is 100.

CK-TUT-10112-1112-UG-1 149 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.7.2.3 Monitoring Convergence

A text output file named "SgConvergence.out" is written in the output directory. The
data in this file consists of 4 columns: (1) iteration number, (2) absolute value of the
maximum difference in the mole fraction, (3) the gas-phase species index for which
the maximum difference happens, and (4) the grid-point number where this happens.

This output can be useful in a variety of ways; for example to detect which species is
most sensitive to particle interactions or to monitor if the convergence is stalling.
While the former can shed some light on the reaction mechanism, the latter typically
happens when the grid is not sufficiently fine and/or the absolute convergence
tolerance on the gas-phase species is not tight enough. A tight tolerance is required;
for example, in the opposed-flame example in this tutorial, the absolute tolerance is
set to 1.e-12.

2.7.2.4 Project Results

As mentioned earlier, the surface mechanism used in these simulations is the HACA
soot-growth mechanism and thus soot growth by reactions with aromatic species is
not considered. For the Flame-speed simulation, soot oxidation by hydroxyl radical
(OH) is included. Following the general guidelines in these examples, a parametric
analysis can be conducted to understand and evaluate various soot mechanisms and
individual reactions.

2.7.2.4.1 Flame-speed Calculator with Particle Formation and Oxidation Simulation

Figure 2-127 shows particle volume fraction and temperature as a function of distance
for the ethylene-air premixed flame (phi = 3.0). It can be seen that while the flame is
located at around x = 0.0 cm, the particle nucleation and growth continues in the post-
flame region. Considering the slope of the volume fraction profile at the downstream
side, the computation domain may be extended further so that the gradient vanishes
smoothly. Figure 2-128 shows, along with the particle number density, the mole
fractions of the gas-phase precursor, pyrene (A4), the species that contributes to
growth, acetylene (C2H2), species that reacts with particle surface (H), and species
that consumes the particle (OH). Since, as per the reaction mechanism, A4 is
responsible for soot nucleation, the shapes of particle number density profile and
pyrene profile are qualitatively similar. Acetylene, in addition to coagulation,
increases the size of soot particles by attacking the 'open sites' on the particle
surface. It is also interesting to note that A4 and OH mole fractions are of the same
order of magnitude in the post-flame region. Using the rates of production and
reaction progress variables, further analysis can be conducted to evaluate
competition between the soot nucleation and oxidation reactions.

2011 Reaction Design 150 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-127 Distance vs temperature and particle volume fraction

Figure 2-128 Distance vs species mole fraction and particle number density

2.7.2.4.2 Opposed-flow Diffusion Flame Simulation with Soot Formation Simulation

In general, for soot studies, the opposed flow diffusion flames can be classified in two
types: (i) Soot formation (SF) and (ii) Soot formation-oxidation (SFO). Depending on
the composition of reactants in the fuel and oxidizer nozzles, the flame may be
located on the either side of the stagnation plane. When the flame is on the oxidizer
side, as in this example, the soot particles that nucleate and grow are convected away
from the flame towards the stagnation plane and thus cannot oxidize. The HACA
mechanism alone is therefore a reasonable model for the soot surface reactions in
these flames.

CK-TUT-10112-1112-UG-1 151 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-129 shows computed profiles of particle volume fraction and temperature.
As expected, the flame is situated on the oxidizer side of the stagnation plane at
about 6 mm from the fuel inlet whereas the particle number density shows a peak at
about 4 mm from the fuel inlet. The temperature of the "sooting zone" is about
1200 K. As seen from Figure 2-130, the particle number density practically vanishes
at the flame location and the profile is mono-modal. (In contrast, a bimodal behavior
is expected from SFO flames.) Although the OH radical peaks away from the particle
number density peak as expected, thus having minimal effect on soot oxidation, the
predicted soot volume fraction is still quite small. This suggests that HACA
mechanism alone may not be sufficient to capture soot growth in this case and one
may have to use a PAH condensation model and/or soot formation from acetylene.

Figure 2-129 Distance vs Temperature and Particle volume fraction

2011 Reaction Design 152 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-130 Distance vs species mole fraction and particle number density

2.7.2.5 Concluding Remarks

The purpose of these examples is to demonstrate techniques for applying Particle
Tracking Feature to simulate soot formation in pre-mixed and opposed-flow flames.
The area of soot nucleation, growth, and destruction is still the subject of research,
such that the mechanisms and values of the rate parameters have a lot of uncertainty.
The Particle Tracking Feature can be used to perform systematic parametric analysis
to better understand soot formation pathways and also to provide engineering
estimates of related quantities.

2.7.3 Sectional Method for Particle-Size Distribution with Pre-mixed Laminar

Burner-Stabilized Stagnation Flame [CHEMKIN-PRO Only]

2.7.3.1 Project Description

As shown in Figure 2-131, in the cases of CHEMKIN-PRO reactor models that
support the Sectional Method and the Particle Tracking feature, radio buttons on the
Reactor Properties tab provide a choice between the Moments Method and Sectional
Method. Selecting the Sectional Method means that the segregated solver computes
the solution by sequentially solving the gas-phase equations followed by the sectional
equations along with the surface species.

CK-TUT-10112-1112-UG-1 153 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.7.3.2 Project Setup

The project file is named pre-mixed_stagnation__soot_particles.ckprj and is
located in the samples2010 directory. The corresponding input and chemistry set
files are located in the samples2010\pre-mixed_stagnation\soot_particles
directory. The project is set up to represent roughly the flame facility at the University
of Southern California (USC)28, where extensive soot measurements are being
conducted. For this tutorial, as in Section 2.7.2, only the HACA (H-Abstraction-
Carbon-Addition) part of the surface reaction mechanism (described in Section 2.9) is
used. The fuel is ethylene, the equivalence ratio is about 2.067 and fuel-oxygen
mixture is diluted with argon. The burner inlet velocity is set to 15 cm/s at an inlet
temperature of 400 K. The stagnation plane (wall) temperature is set to 500 K and is
located at 1 cm from the inlet. The ambient pressure is set to 1 atm for the
atmospheric flame facility. The simulation uses mixture-averaged transport, which is
a good approximation for most hydrocarbon fuels. The adaptive grid control
parameters for gradient and curvature are set to 0.2 and 0.4, respectively. This
setting gives reasonable resolution while keeping the computational cost small.

2.7.3.2.1 Selecting the Sectional Method

As shown in Figure 2-131, in the cases of simulators CHEMKIN-PRO reactor models
that support the Sectional Method and the Particle Tracking module, radio buttons on
the Reactor Properties Tab provide the choice between the Moments Method and
Sectional Method. Selecting the Sectional Method currently means: (a) both the
Moments Method and Sectional Method are used, and (b) the segregated solver
computes the solution by sequentially solving the gas-phase, the moments, and the
Sectional Method. The coupling between the gas-phase and the particles is through
the Moments Method equations.

28. Wang, Hai, Personal communication, 2010.

2011 Reaction Design 154 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-131 Selecting the Sectional Method on the Reactor Physical Properties tab.

The appearance of the Dispersed Phase tab is controlled by the existence of a

DISPERSED phase material in the surface chemistry file of the chemistry set defined
for the project. When the Sectional Method is selected, this panel appears as shown
in Figure 2-132.

Figure 2-132 Appearance of the Dispersed Phase tab when the Sectional Method is selected.

CK-TUT-10112-1112-UG-1 155 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.7.3.2.2 Description of Input for the Sectional Method

The Sectional Method has three mandatory inputs, which control the range of particle
sizes used by the simulation. These inputs are (a) number of sections, (b) spacing
factor, and (c) section minimum size. The spacing factor is a geometric control of the
relative sizes of the sections or "bins" used for tracking particle units (monomers). For
example, the representative particle size in the (n+1)th section is the spacing factor
times the representative particle size in the nth section.

The sections consider particle units or monomers. For example, for soot defined by
carbon atoms the possible "particles" may be defined to start from a single carbon
atom in the bulk phase even if the nucleation reaction(s) have a higher inception class
(i.e., the nucleated particle has more than one carbon atom). In the soot-formation
mechanism used in this tutorial, there is one nucleation reaction and it yields particles
with 32 atoms (monomers) in it. The spacing factor of 2 then means that the
nucleated particles go in the 6th section. If the nucleated particles were then allowed
to be etched away (by including reducing reactions in the mechanism), then sections
1 through 5 could have non-zero particle number density. Since there are no reducing
reactions in the mechanism used in this tutorial, the minimum section size can be set
to 6, but for general demonstration purposes it is set to unity. Note that the minimum
size must be set to unity whenever there are bulk-etching reactions in the particle
surface chemistry mechanism.

For a given simulation, the setting of these parameters is influenced by the size range
of particles expected to form and/or permitted by the model approximations For
example, for the particle models implemented in the Particle Tracking module, the
diameter of the largest particle formed should not be greater than about 1 ; which is
the resolution with which the gas-phase solution of the reacting flow is typically
computed. For this tutorial, the number of sections used is 30, which, with a spacing
factor of 2 and 0.2765 nm for the diameter of a single C(B) atom, gives, means that
the representative particle diameter in the 30th section is approximately equal to
225 nm. While this is sufficient for the present simulation, the computed particle
number density in the largest section must be checked after the simulation to verify
that it vanishes smoothly at each grid point in the computational domain.

The input for scaling factors, coagulation regime and efficiency is kept at its default
values. The scaling factors for moment and surface species are set to 1E+14.
Particle thermophoresis is excluded in this tutorial. When it is included, the input for
the particle materials thermal conductivity should also be specified.

2011 Reaction Design 156 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

2.7.3.3 Project Results

When the Sectional Method is included in the simulation, the following variables
become available during post-processing: (a) the number density for each section,
(b) the average properties predicted using the sectional data, and (c) the particle-size
distribution function (PDF) at each grid-point. The experimental data on particles is
usually in the form of a PDF and the form (dN/dLogD) reported by the Sectional
Method matches that reported in the literature. To distinguish between the average
properties computed using the Moments Method used alone and Sectional Method,
the latter results include "total_(section)" in the name. For example,
total_(section)_number_density gives the total particle number density calculated by
the Sectional Method.

As mentioned above, this project allows no particles in the first 5 sections since there
is no oxidation reaction in the surface mechanism. These values are filtered out by
the Post-processor, as with all zero-value arrays.

Figure 2-133 is a plot of temperature and mole fraction of pyrene (A4), which is the
nucleating species, as a function of distance. Since the inlet velocity is about a factor
of 10 smaller than the laminar burning speed for the mixture at the inlet conditions, the
flame is attached to the burner. The presence of a wall creates a colder boundary
layer near that wall. It can also be seen that the mole fraction of A4 has a negative
slope in the region near the wall.

Figure 2-133 Sectional Method plot of temperature and mole fraction of pyrene (A4).

CK-TUT-10112-1112-UG-1 157 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-134 and 2-135 show the particle-size distribution and the PDF as a function
of distance. This plot is created by selecting the sectional data vs. section property
for the Plot Set textbox in the Post-processor control panel and then choosing
section diameter for the X-axis. The nucleation reaction in the mechanism used
here creates particles of diameter about 0.87 nm. Since smaller particles are not
present, the curves in Figure 2-134 and 2-135 start somewhat abruptly at the
corresponding point. The presence of the cold wall (distance = 1 cm) enhances
particle production, as seen by an increase in the number density of smaller particles
near the wall.

Figure 2-134 Particle-size distribution as a function of distance.

2011 Reaction Design 158 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-135 PDF as a function of distance

2.7.4 Simulating Particle-Size Distributions in a Burner-Stabilized Stagnation

Flame
When simulating particle-size information in flames in CHEMKIN-PRO, you have a
choice of using the Moment Method, which provides statistical information about the
size distribution and mean particle properties, or the Sectional Method, which predicts
the size distribution directly. These particle simulation options are part of the Particle
Tracking module that is included with CHEMKIN-PRO. In describing the formation
and growth of particles in flames, it is often important to include the effects of particle
aggregation. In this tutorial we apply the use of particle aggregation with the
Sectional Model method for simulating soot-particle size distributions in a flame. The
flame considered is a burner-stabilized premixed flame that impinges on a stagnation
surface; this is modeled using the Burner-Stabilized Stagnation Flame reactor model.

For flame simulations using the Sectional Model, the aggregation model employed is
the so-called Simple Aggregation Model, which is described in detail in the CHEMKIN
Theory Manual. This form of the aggregation model is particularly useful when the
data for characteristic fusion time is not well known, which is the case for many
particle materials of practical interest; most notably for carbon-based soot formed in
flames and combustion engines. For such systems, it is often easier to define a
limiting size for the primary particles in the system. This limiting size (diameter) is

CK-TUT-10112-1112-UG-1 159 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

dictated by the competition between collisions among the aggregates and

fusion/coalescence of individual aggregates that lead to spherical particles. When
collisions happen too quickly or fusion takes long time, a given aggregate cannot
coalesce and the primary particle size appears to reach a limit.

In the flame systems, then, the aggregates are modeled as either pure aggregates
consisting of primary particles of a fixed (limited) size or as completely coalesced
spheres with sizes less than or equal to the limiting size. In this tutorial, we describe
the setup and analysis of a flame with particles, considering this simplified
representation of particle aggregation for sooting flame conditions.

2.7.4.1 Project Setup

The project set-up is almost identical to the previous one (Section 2.7.3) except that
the following changes are made: The project file is named pre-
mixed_stagnation__soot_aggregation.ckprj (it is located in the samples2010
directory and the corresponding input and chemistry set files are located in the
samples2010\pre-mixed_stagnation\soot_particles directory); the inlet velocity,
the separation distance, and inlet equivalence ratio are increased to 25 cm/s, b cm,
and 2.2, respectively. For these conditions, the particle size distribution (PSD) shows
much more pronounced bimodality than that shown by the PSD obtained for the
conditions from the project in Section 2.7.3. The effect of using the simple
aggregation model is hence illustrated better with the new set of conditions used in
this tutorial.

Since larger particles may form at these simulation conditions, the number of sections
used is increased to 45. In addition, the transition regime is selected for collisions.
The corresponding collision kernel (CHEMKINTheory Manual Equation 18-131) is
applicable over the entire range of particle Knudsen number.

The Aggregation properties are entered on a sub-panel within the "Dispersed

Phase::CARBON" tab in the reactor properties node of the project tree. On the
Aggregation sub-panel, the Simple Model option is selected. The two user-input
parameters are: the fixed (limiting) primary particle diameter and the fractal dimension
of the aggregate. The fractal dimension of 3.0 implies a spherical particle. If the fixed
primary particle diameter were set to a value larger than the largest particle that can
form in the system, the simple aggregation model would yield the same results as if
the aggregation model were turned off. In this way, we can look at the limiting case
and compare this to other cases to see the effects of aggregation.

2011 Reaction Design 160 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

We set up a parameter study to explore the effects of the aggregation model. The
fixed primary particle diameter and fractal dimension are set to 500 nm and 3.0,
respectively, for the last run. For the first two runs, the fractal dimension is set to 1.8,
indicating cluster-cluster aggregation, while the fixed primary particle diameter is
varied between 30 nm and 25 nm.

2.7.4.2 Project Results

Figure 2-136 shows the particle number density distributions obtained at the
stagnation plane as a function of the section (volume-averaged) diameter for all three
parameter studies. In the figure legend, dp indicates the fixed primary particle
diameter in nanometers and Df indicates the fractal dimension. The effect of
aggregation can be easily seen. As mentioned above, the complete instantaneous
coalescence to spherical particles is mimicked by the last run (the dotted line).
Comparing this to the other runs, we can see that the smaller particles are
scavenged by the larger ones more effectively when the aggregation model is
active. This can be attributed to the higher collision diameter of aggregates. The
scavenging effect increases as the limiting primary particle size is decreased.

Figure 2-136 Particle-size distribution at the stagnation plane for three conditions imposed by the parameter study.

CK-TUT-10112-1112-UG-1 161 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.7.5 Detailed Particle Aggregation in a Batch Reactor [CHEMKIN-PRO Only]

2.7.5.1 Project Description

This tutorial describes how to use the particle-aggregation model, when using the
Sectional Method in the Particle Tracking feature for computing the particle size
distributions in a batch reactor. As discussed in the CHEMKIN Theory Manual, the
Sectional Method in the CHEMKIN-PRO Particle Tracking feature offers two ways
(models) to incorporate the effect of aggregation on particle size distribution for 0-D
and plug-flow reactor models in CHEMKIN-PRO. These options are the "simple" and
"complete" aggregation model. The "complete" model requires knowledge of the
characteristic fusion time for the particle material system being studied. In this
tutorial, we simulate formation of titania (TiO2) particles by oxidation of titanium
tetrachloride (TiCl4). Since the characteristic fusion time data is available for TiO2, the
complete aggregation model is used.

2.7.5.2 Project Setup

The project file is named closed_homogeneous__particle_aggregation.ckprj and
is located in the samples2010 directory. The corresponding input and chemistry set
files are located in the samples2010\closed_homogeneous\particle_aggregation
directory.

This tutorial models a constant-pressure and given-temperature batch reactor. The

pressure is set at 1 atm and a piecewise temperature profile as a function of time is
imposed on the closed system. The end time for the simulation is 1 s. The initial gas-
phase reactant mixture contains pure oxygen (O2) as the oxidizer and titanium
tetrachloride (TiCl4) as the precursor. For the TiCl4/O2 system, it is assumed that the
titania particles are nucleated by single-step oxidation of TiCl4. The rate of this
reaction is assumed to be only dependent on the amount of precursor. The initial
mixture consists of 10 moles of O2 and 10-4 moles of TiCl4.

2.7.5.2.1 Setting Sectional Method Inputs

For CHEMKIN reactor models that support the Particle Tracking feature and the
Sectional Method, radio buttons on the Reactor Properties tab provide a choice
between the Moments Method and Sectional Method. This choice is available when
the preprocessed chemistry set contains DISPERSED phase material in the surface
reaction mechanism. Based on the choice of particle tracking method selected, the
Dispersed Phase tab contains appropriate input controls.

2011 Reaction Design 162 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

2.7.5.2.2 Selecting the Number of Sections and Section Spacing

The number of sections used in the Sectional Method should be large enough that
vanishingly small number densities are obtained in the last sections, i.e., in the
sections with the largest particles. The number of sections is hence decided based on
an estimate of the size of the largest particle that can be formed in a given the system.
While the exact value is not known a priori, most reacting particle-flow mistures of
interest do not invlove particles of size greater than a micron in diameter. In this
tutorial, 40 sections are used with a section-spacing factor of 2. The section minimum
size is set to unity since the nucleation reaction produces monomers of titania. Since
the diameter of the titania monomer is about 0.4 nm, 40 sections with a spacing factor
of 2 gives aggregates of about a 3 micron volume-equivalent spherical diameter. This
is sufficient for this system.

As sub-micron-size particles are typically expected for most reacting flow particulate
systems, the free-molecular collision regime is usually appropriate to model particle
collisions. This is also true for the system considered in this tutorial. The
corresponding collision kernel for the free-molecular regime has a simple
mathematical form, as given by Equation 18-129 in the CHEMKINTheory Manual.
However, for illustrative purposes, we select the transition collision regime for this
tutorial since it is applicable to both free molecular and continuum regimes and in
some cases you may not know the regime in advance of the simulation. The collision
kernel for the transition regime is given by Equation 18-131in the CHEMKINTheory
Manual. Since the collision kernel expression for the transition regime needs the
viscosity of ambient gas, we provide viscosity parameters for the power law fit for
oxygen. Note that the computational cost of using the transition regime collision
kernel is generally higher than when using the kernels of either free-molecular or
continuum regime. (Note that the gas viscosity is also required to specify the collision
kernel for the continuum regime, as given by Equation 18-130 in the
CHEMKINTheory Manual.)

2.7.5.2.3 Setting Aggregation Model Inputs

The aggregation sub-tab in the Dispersed Phase panel provides the option to turn ON
the aggregation model (OFF is the default). When turned on, you can select either the
Simple or Complete aggregation formulation. For this tutorial, the complete
aggregation model is selected. The characteristic fusion time for titania is given by29

29. Xiong, Y., and Pratsinis S. E., Formation of Agglomerate Particles by Coagulation
and Sintering, Journal of Aerosol Science and Technology, 24: 301 (1993).

CK-TUT-10112-1112-UG-1 163 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Equation 2-1

3700 4
8.3 1016 T exp ( )d p
T
In Equation 2-1, the primary particle diameter (dp) is in units of cm while the
characteristic fusion time () is in s. As indicated by Equation 18-92 from the
CHEMKINTheory Manual, it can be noted that the characteristic fusion time indicates
the time required to reduce by 63% the excess surface area of an aggregate over that
of an equal mass (volume) spherical particle.

The input parameter threshold to include the fusion effect performs the task of limiting
the value of fusion rate. Aggregates with small primary particle diameter coalesce
very rapidly and the characteristic fusion times can be on the order of microseconds
or smaller. Resolving aggregation of such small particles is unnecessary and can
force very small integration time steps. The default value of this threshold parameter
is 1 microsecond, which is adequate for most cases. For this tutorial, by setting
equal to 1 microsecond in Equation 2-1, it can be seen that aggregates with primary
particle diameters of less than 1.5 nm will coalesce by 63% within 1 microsecond at
the highest temperature (1200K), where fusion is the fastest. Considering that the
diameter of the titania monomer is about 0.4 nm, particles of 1.5 nm are indeed small
(would fall within the first 5 to 8 sections with a spacing factor of 2). Thus the default
value is satisfactory.

The fractal dimension of aggregates is set to 1.8, indicating cluster-cluster

aggregation.

2.7.5.3 Project Results

Figure 2-137 and Figure 2-138 show the imposed temperature profile along with the
mole fraction of the precursor (TiCl4) and the nucleation rate of TiO2 monomers as a
function of time, respectively. As expected, due to the first-order dependence of he
nucleation reaction on the concentration of TiCl4, the decay of the precursor is
exponential once a critical temperature is reached. The nucleation rate peaks at
around 0.6 s. We note that nucleation is active throughout the entire duration of the
simulation.

2011 Reaction Design 164 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-137 Imposed temperature profile with the mole fraction of the precursor TiCl4.

Figure 2-138 Imposed temperature profile with the nucleation rate of the precursor TiCl4.

CK-TUT-10112-1112-UG-1 165 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Shown in Figure 2-139 is the aggregate number density as a function of the volume-
equivalent section diameter. At t = 0.4 s, the particle size distribution is almost
unimodal. As the nucleation slows down the bimodal shape of the distribution
becomes evident as seen from curves corresponding to t > 0.6 s.

Figure 2-139 Complete Aggregation - Aggregate number density as a function of the volume equivalent section
diameter.

Plotted in Figure 2-140 is the evolution of primary particle diameter. The primary
particle diameter is computed using Equation 18-96 from the CHEMKINTheory
Manual. At t = 0.4 s all aggregates are small: volume equivalent diameter is less than
5 nm. Consequently, all are almost completely coalesced spheres. Thus, the number
of primary particles per aggregate is close to unity and the primary particle diameter is
the same as the volume equivalent diameter. (In the plot, the series for t = 0.4 is
shown as points only in order to distinguish it from the data points of other series.)

2011 Reaction Design 166 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-140 Evolution of the primary particle diameter.

Since the rate of fusion is quite high for aggregates with small primary particles, this
condition (i.e., completely coalesced aggregates) stays true at all times for small
aggregates. For larger aggregates, the primary particle diameter clearly shows a
limiting value. This suggests that the rate of fusion cannot keep up with the rate of
collisions between the aggregates. (This fact, that a limiting primary particle size can
exist, illustrates the motivation behind a simple aggregation model. Please see the
CHEMKINTheory Manual, Simple Aggregation Model in Chapter 18.)

To see the effect of aggregation, the simulation can be repeated with the aggregation
model turned off. Shown in Figure 2-142 are the number density distributions
obtained without aggregation. As expected, the aggregate distribution does not differ
much between instantaneous coalescence (aggregation model turned-off) and finite
rate fusion. Divergence in the responses of the two models occurs as nucleation
slows down and/or larger particles are formed. The instantaneous coalescence shows
a gradual shift to a self-similar distribution through a bimodal distribution.

CK-TUT-10112-1112-UG-1 167 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-141 Number density distributions obtained without aggregation.

For a finite rate of fusion as given by Equation 2-1, small particles are depleted much
more quickly owing to the larger (than volume-averaged) collision diameters of bigger
aggregates. This scavenging effect is clearly seen at t = 1 s. As the nucleation rate
slows down, the number density of particles in certain (volume equivalent) diameter
ranges start to vanish. Since the nucleation is not completely stopped, particles of the
smallest size are still produced in significant enough numbers and the distribution
may eventually develop a gap/discontinuity.

Figure 2-142 shows the collision diameter as a function of (volume equivalent) section
diameter. It can be seen that the ratio of collision diameter to volume-equivalent
diameter is greater than unity; approaching an order of magnitude for aggregates with
100 nm and higher volume-equivalent diameter.

2011 Reaction Design 168 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

Figure 2-142 Collision diameter as a function of (volume equivalent) section diameter.

Shown in Figure 2-143 is the comparison of the total particle number and surface area
densities as a function of time with and without aggregation model. As expected,
there is very little difference when nucleation is more dominant (t < 0.6 s). However,
the number density predicted with finite rate of fusion (aggregation model turned on)
is a factor of 3 smaller and the surface area is a factor of 2 than that predicted when
the aggregation model is not used. Thus, when reliable data for fusion time is
available, the complete aggregation model with the Particle Tracking feature can be
used to guide design in particle production processes.

CK-TUT-10112-1112-UG-1 169 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-143 Comparison of the total particle number and surface area densities as a function of time with and
without aggregation model

2.8 Uncertainty Analysis [CHEMKIN-PRO Only]

2.8.1 Uncertainty Analysis of NOx Emissions [CHEMKIN-PRO Only]
Emission studies have become a vital part of the design and analysis of gas-phase
combustion processes, because of the increasingly stringent requirement imposed by
government regulations. The emission standard of NOx by combustion processes is
approaching the ppm (parts per million) level. However, the combustion processes in
general demonstrate variations in operating conditions due to the very nature of these
processes. These variations can induce uncertainties in the level of NOx emission. It
is therefore necessary to account for these uncertainties during the design and
analysis of the combustion processes to make sure the emission level is still within
the regulatory requirements. Uncertainty analysis can be performed to study the
effect of variations in the operating conditions on the NOx emission level.

2011 Reaction Design 170 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

In this tutorial, we apply an uncertainty analysis to investigate the effects of heat loss
and equivalence ratio for the NOx emission level of a perfectly stirred reactor. The
associated project file is called psr__NOx_uncertainty_analysis.ckprj. The
chemistry-set files used for this sample problem are located in the
psr/NOx_uncertainty_analysis folder of your CHEMKIN samples2010 directory.
For the gas-phase kinetics, we employ the GRI gas-phase mechanism and
thermodynamic data for methane combustion. This mechanism contains NOx
chemistry in methane combustion.

We choose to study the effect of the variations in heat loss from the wall of the reactor
and the equivalence ratio (i.e., fuel/air ratio) for the NOx emission of a perfectly stirred
reactor. These operating conditions are selected because (1) the exact heat loss
from the wall of the reactor/combustor is often unknown in the combustion industry (2)
non-perfect mixing in the combustion processes often lead to pockets of very fuel-rich
or fuel-lean reactant mixtures, thus making the overall reactant fuel/air ratio uncertain.

An uncertainty analysis has already been set up for heat loss and equivalence ratio in
the psr__NOx_uncertainty_analysis project. To see how this is done, double-click
the C1_PSR node on the project tree to open the Reactor Physical Properties panel,
as shown in Figure 2-144. Click on the Setup Uncertainty Analysis button (i.e. with
icon ) next to the Heat Loss parameter of the reactor to view the uncertainty
analysis setup, as shown in Figure 2-145. We are using the normal (Gaussian)
probability density function to represent the uncertainty in the heat loss because of
the lack of more specific information available. The mean value of the distribution is
the nominal value of heat loss in the reactor and the standard deviation is assumed to
1.0 cal/sec or 20% of the mean value for this analysis. You can view the uncertainty
setup of the equivalence ratio by double-clicking the C1_Inlet1 node on the project
tree and then clicking on the Species-specific Properties tab. We are using the
normal distribution for the equivalence ratio too but we are assuming a much smaller
standard deviation in the analysis. We use the final mole fraction of NO in the reactor
as the uncertain output in the NOx emission study.

CK-TUT-10112-1112-UG-1 171 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

Figure 2-144 Setting Up an Uncertainty Analysis for Heat Loss

Figure 2-145 Setting Up the Uncertainty Distribution for Heat Loss

Run the uncertainty analysis by selecting all of the runs in the Running Uncertainty
Analysis panel and clicking the Run Selected button. Once they are done, click the
Do Analysis button to perform the uncertainty analysis for the NOx emission level and
perform the variance analysis to determine the contribution to the uncertainty in NOx
emission from the variations in heat loss and equivalence ratio. A text file is displayed
to show the summary of the analysis. From Section 2, Error Analysis, you can see
that the accuracy of the uncertainty model is high in the predictions of NOx emission
level. From Section 3, Variance Analysis, you can see that the uncertainty in the NOx
emission level is quite moderate, since the standard deviation of the NOx emission
level is about 10% of the mean value. The more interesting results from the variance

2011 Reaction Design 172 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

analysis is that the smaller variation in the equivalence ratio has much more impact
on the uncertainty of the NOx emission level than the much larger variation in the heat
loss. This indicates that the variation in the equivalence ratio is more important to
maintain the required level of NOx emission in this perfectly stirred reactor.

Further analysis of the NOx emission can be achieved by examining the probability
density function (pdf) of NO. Click the View PDF button and select to view the pdf of
NO, as shown in Figure 2-146. The pdf of NO shows the distribution of the NO
emission given the variations in the heat loss and equivalence ratio of the reactor.
From the pdf, we can conclude that the most likely emission level of NO is between
2.3E-5 (23 ppm) and 3.2E-5 (32 ppm). If the emission requirement is 2.0E-5 (20 ppm)
(i.e., limit 1), the likelihood of the NO emission to be below this level is minimal, since
most portion of the pdf curve lies above the level of limit 1. If the emission
requirement is 3.5E-5 (35 ppm) (i.e. limit 2), it is almost certain that the NO emission
will satisfy this requirement, regardless of the variations in the heat loss and
equivalence ratio of the reactor.

Figure 2-146 Probability Density Function of NO

2.9 Chemistry Sets

In this section we provide a brief description of and references for the chemistry
mechanisms employed in Chapter 2.

2.9.1 Hydrogen/Air
The chemistry sets used to describe the combustion of hydrogen in air are provided
for the purposes of illustration only. These are slight variations of early work done at
Sandia National Laboratories, and are out-of-date. The gas-phase kinetics input files
are relatively simple. They generally contain 3 elements H, O and N or Ar, and 9 gas-

CK-TUT-10112-1112-UG-1 173 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

phase species: H2, H, O2, O, OH, H2O, HO2, H2O2 and N2 or Ar, with about 18-20
reactions. Nitrogen is generally present as N2 only, where it does not participate in
any chemical reactions (i.e., NO formation is not included). Pressure dependencies
for reaction rates are not explicitly treated, although there are enhanced collision
efficiencies for some reactions.

2.9.2 Methane/Air
Two somewhat different chemistry sets are used in the user tutorials to describe the
combustion of methane in air or oxygen. One is the full version of GRImech 3.0,30
which includes the reactions leading to NOx formation. The other is a smaller reaction
mechanism, which is used for the purposes of illustration in cases where reducing
computation time is helpful.

2.9.2.1 GRImech 3.0

GRImech version 3.0 is a relatively well-tested reaction mechanism that was
developed under the auspices of the Gas Research Institute. This reaction
mechanism, consisting of gas-phase chemistry, thermodynamic, and transport data
files, is provided as one of the default chemistry sets in CHEMKIN in the system data
folder. A detailed description of the development and extensive validation of this
mechanism is available.

The gas-phase kinetics input file contains 5 elements, C, H, O, N and Ar, 53 chemical
species, and 325 reactions. The reaction mechanism describes the combustion of
methane and smaller species such as hydrogen, as well as the formation of nitrogen
oxide pollutants. Most of the reactions are reversible; only a small number of
irreversible reactions are included. Many of the reactions have explicit descriptions of
their pressure-dependencies, using the Troe formulation and including enhanced
collision efficiencies for particular species.

2.9.2.2 Reduced Mechanism

This smaller reaction mechanism is from early work at Sandia National Laboratories
and is given for purposes of illustration only. The gas-phase kinetics input file contains
4 elements, C, H, O, and N, 17 chemical species, and 58 reversible reactions. The
reaction mechanism describes the combustion of methane and smaller species such
as hydrogen, but nitrogen is present as N2 only, and does not participate in any
chemical reactions. There is no explicit treatment of pressure-dependencies for
reaction rates, although there are enhanced collision efficiencies for some reactions.

30. http://www.me.berkeley.edu/gri-mech/.

2011 Reaction Design 174 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

2.9.3 NOx and CH4, C2H4, C2H6, C3H6, and C3H8

This reaction mechanism is made available by Lawrence-Livermore National Labs.
The gas-phase kinetics input file contains 5 elements, Ar, C, H, O, and N, 126
chemical species, and 638 reversible reactions. The chemical kinetic mechanism was
validated to describe the promotion effect of hydrocarbons (methane, ethane, ethene,
propene and propane) on NO to NO2 conversion in an atmospheric flow reactor. The
NO level was 20 ppm and the hydrocarbon level was 50 ppm. The flow reactor
temperature ranged from 600 to 1100 K.

2.9.4 Propane/Air
This mechanism is the result of work at the Center for Energy Research (CER),
University of California, San Diego. It consists of 46 species and 235 reactions. The
elements constituting the species are N, H, C, O, Ar, and He. The thermodynamic and
transport data in this chemistry set are included from the same source. All reactions
are reversible, and some of the reactions include pressure-dependencies on the rate
constant using the Troe formulation (see Equation 3-27 in the CHEMKIN Theory
Manual). Enhanced collision efficiencies are used for some reactions. The references
for the reaction rate parameters as well as for thermodynamic and transport data can
be obtained from the CER website:
http://www-mae.ucsd.edu/~combustion/cermech/

2.9.5 Ethylene/Air Combustion and Soot Formation and Growth

2.9.5.1 Combustion Mechanism and Soot Nucleation Reaction

The C2H4-air combustion mechanism of Appel et al.31 is provided with this chemistry
set. This reaction mechanism consists of 101 species and 543 reactions which
include PAH growth reactions up to pyrene, A4.

According to Frenklach and coworkers31,32, soot particles are created by the

dimerization of pyrene molecules and to represent this, a single nucleation reaction is
included in this chemistry set. The kinetic parameters of the soot nucleation reaction
were derived by matching the nucleation-only soot mass concentration prediction to
measurements33, and are shown as follows

31. J. Appel, H. Bockhorn, and M. Frenklach, Combust. and Flame, 121:122-136 (2000).
32. M. Frenklach and H. Wang, in Soot Formation in Combustion: Mechanisms and Models,
H. Bockhorn (Ed.), Springer-Verlag, pp. 165-192 (1994).
33. C.-P. Chou, D. Hodgson, M. Petrova, and E. Meeks, "Modeling Soot Growth and Ac-
tivity with Heterogeneous Kinetics and Method of Moments," 5th U.S. Combustion Meet-
ing, Combustion Institute, San Diego, CA, 2007.

CK-TUT-10112-1112-UG-1 175 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

(S1) 2A4=>32C(B)+20H(S)+28.72(S) 9.0E09 0.5 0.0

The A-factor computed from the collision frequency between A4 molecules is of the
order of 1012 and is much larger than the one used in the above nucleation reaction,
which suggests that about one in 1000 collisions results in a nucleation event.

2.9.5.2 Soot Mass Growth Reactions

One of the advantages of the Particle Tracking module is that soot mass growth and
oxidation reactions can be provided as regular surface reactions. For example, the H-
Abstraction-C2H2-Addition (HACA) soot growth sequence proposed by Frenklach and
coworkers31,32 can be given as:

(S2) H+H(S)=>(S)+H2 4.20E13 0.0 13000.0

(S3) H2+(S)=>H(S)+H 3.90E12 0.0 9320.0
(S4) H+(S)=>H(S) 2.00E13 0.0 0.0
(S5) H(S)+OH=>H2O+(S) 1.00E10 0.734 1430.0
(S6) H2O+(S)=>OH+H(S) 3.68E08 1.139 17100.0
(S7) C2H2+(S)=>H(S)+2C(B)+H 8.00E07 1.56 3800.0

Each sweep of the HACA growth sequence will increase the soot particles by two
classes.

Marr27 (see p. 141) suggested that PAH condensation on the soot particle surface can
have significant contribution to soot mass growth in a post-flame zone. He also found
from his experimental study that the collision efficiency of PAH condensation is of the
order of 0.1. Since a description of how PAH species interact with a soot particle is not
readily available, the PAH condensation reactions used in the simulation are
estimated. Only condensations of major PAH species in the gas mechanism are
considered. The reaction orders are determined by fitting the experimental data while
keeping the sticking coefficients to be on the order of 0.1. Basically, the PAH
condensation reactions used here are designed to increase soot particle mass in two
ways: they grow the soot particle by the addition of PAH species and they remove
some H atoms (H(S)) on the soot particle surface so that the more effective soot
growth reaction, C2H2 addition (S7), can proceed at a greater rate.

2011 Reaction Design 176 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

(S8) A1+6H(S)=>6C(B)+6(S)+6H2 0.2 0.0 0.0

FORD/H(S) 1.0/
FORD/(S) 1.0/
DCOL/2.46E-8/
STICK

(S9) A1C2H+6H(S)=>8C(B)+6(S)+6H2 0.21 0.0 0.0

FORD/H(S) 2.0/
DCOL/2.46E-8/
STICK

(S10) A2+16H(S)=>10C(B)+16(S)+12H2 0.1 0.0 0.0

FORD/H(S) 2.0/
DCOL/4.92E-8/
STICK

(S11) A2R5+16H(S)=>10C(B)+16(S)+11H2+C2H20.1 0.0 0.0

FORD/H(S) 2.0/
DCOL/4.92E-8/
STICK

(S12) A3+20H(S)=>14C(B)+20(S)+15H2 0.1 0.0 0.0

FORD/H(S) 2.0/
DCOL/4.92E-8/
STICK

(S13) A4+20H(S)=>16C(B)+20(S)+15H2 0.1 0.0 0.0

FORD/H(S) 2.0/
DCOL/4.92E-8/
STICK

Expressing the soot growth sequence as a chain of surface reactions has other
advantages. As part of the surface mechanism, it is easy to check for the
conservation of elements and to perform sensitivity analysis on these reactions.
Furthermore, as information about certain reactions, such as PAH condensation,
becomes available from surface-science experiments, utilization of such information
would be straightforward.

CK-TUT-10112-1112-UG-1 177 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2.9.5.3 Soot Oxidation Reaction

Once soot particles are created in the flame zone, they start interacting with the
surrounding gas mixture and with one another. Therefore, if oxidizers such as O, OH,
and O2 are available in the local gas mixture, the soot particles are also subjected to
oxidation. Neoh et al.34 found that the most effective soot oxidizer in the flame zone is
OH and determined that the collision efficiency (or sticking coefficient) for the OH
oxidation reaction is 0.20. For simplicity, the current mechanism only includes soot
oxidation due to OH using this rate:

(S14) OH+(S)+C(B)=>CO+H(S) 0.20 0.0 8000.0

STICK

2.9.6 C2_NOx Mechanism

The C2_NOx mechanism was developed at Reaction Design. This reaction
mechanism, consisting of gas-phase chemistry, thermodynamic, and transport data
files, is provided as one of the default chemistry sets in CHEMKIN-PRO in the system
data folder. It describes oxidation of hydrogen, methane, and ethane over a broad
range of temperature and pressure and includes NOx chemistry.

The mechanism is based on a literature mechanism for oxidation of model-fuel

components35. Many reactions in the base mechanism have been updated based on
most recent kinetic studies. Reaction sub-mechanisms have been updated based on
recent studies on hydrogen36 and methane37 oxidation. The chemically activated
reaction of ethyl radical with molecular oxygen has also been updated based on Naik
and Dean38. The base high-temperature NOx chemistry is from the GRI-mech 3.0
mechanism (see Section 2.9.2.1). Recently, researchers have identified the mutual
sensitization effect of NOx on hydrocarbon oxidation at lower temperatures. The

34. K.G. Neoh, J.B. Howard, and A.F. Sarofim, in Particulate Carbon Formation During
Combustion, D.C. Siegla and G.W. Smith (Eds.), Plenum Publishing Corp., pp. 261-282
(1981).
35. Naik, C.V., et al., Detailed Chemical Kinetic Modeling of Surrogate Fuels for Gasoline and
Application to an HCCI Engine. Society of Automotive Engineers, 2005. SAE 2005-01-3741.
36. O'Conaire, M., et al., A Comprehensive Modeling Study of Hydrogen Oxidation. Int. J.
Chem. Kinet., 2004. 36: p. 603-622.
37. Petersen, E.L., et al., Methane/Propane Oxidation at High Pressures: Experimental and De-
tailed Chemical Kinetic Modeling. Proceedings of the Combustion Institute, 2007. 31: p. 447-
454.
38. Naik, C.V. and A.M. Dean, Detailed Kinetic Modeling of Ethane Oxidation. Combustion
and Flame, 2006. 145: p. 16-37.

2011 Reaction Design 178 CK-TUT-10112-1112-UG-1

Chapter 2: Combustion in Gas-phase Processes Tutorials Manual

current mechanism includes the most recent pressure-dependant NOx-hydrocarbon

reactions from Rasmussen39. Reactions pertaining to NOx chemistry in the
intermediate temperature region has also been updated based on the review of
Dagaut, Glarborg et al.40

The final pressure-dependant mechanism contains 99 chemical species and 693

elementary reactions that describe hydrocarbon oxidation chemistry as well as NOx
chemistry over broad range of temperatures and pressures. Most of the reactions are
reversible and many of the reactions have explicit descriptions of their pressure-
dependencies, using the TROE and PLOG formulation and including enhanced
collision efficiencies for particular species.

39. Rasmussen, C.L., A.E. Rasmussen, and P. Glarborg, Sensitizing Effects of NOx on CH4
Oxidation at High Pressure. Comb. Flame, 2008. 154(3): p. 529-545.
40. Dagaut, P., P. Glarborg, and M.U. Alzueta, The Oxidation of Hydrogen Cyanide and Re-
lated Chemistry. Progress in Energy and Combustion Science, 2008. 34: p. 1-46.

CK-TUT-10112-1112-UG-1 179 2011 Reaction Design

CHEMKIN Chapter 2: Combustion in Gas-phase Processes

2011 Reaction Design 180 CK-TUT-10112-1112-UG-1

 CHEMKIN Tutorials Manual 10112

3 Catalytic Processes

3.1 Catalytic Combustors, Converters and Aftertreatment

3.1.1 Two-Stage Catalytic Combustor

3.1.1.1 Problem Description

For this problem, we need to design a virtually zero-NOx combustor for a sub-scale
test unit of our micro-turbine system. This new combustor must also match several
gas turbine inlet specifications to optimize the overall system performance under full-
load, steady-state operation. A typical gas turbine flow capacity chart is shown in
Figure 3-1. The combustor-related design specifications normally include total mass
flow rate of the gas turbine system and the turbine rotor inlet temperature, TRIT.

CK-TUT-10112-1112-UG-1 181 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

Figure 3-1 Turbine Flow Capacity

The basic rule for minimizing NOx emissions from a gas turbine combustor is to keep
the gas temperature low by operating the combustor under ultra-lean conditions.
However, if the fuel-air ratio gets too low, the combustor will run into flame stability
problems. We decide to work around the flame stability issue with the implementation
of a catalytic combustor. A catalytic combustor produces essentially no NOx and can
convert very lean fuel-air mixtures at relatively low temperatures. There are some
disadvantages of a catalytic combustor, though. For instance, the honeycomb
monolith introduces a large pressure drop and the thermal mass of the honeycomb
material can slow down the catalytic combustor's response to changes in operating
conditions. The precious metals used as catalysts are usually very expensive (Pt
$870/oz, Pd $230/oz) and, to prolong the lifetime of the catalyst, the maximum
operating temperature is much lower for catalytic combustors than for homogeneous
(gas-phase only) combustors. To raise the catalytic combustor exit gas temperature
(< 1200 K) to the desired TRIT (~1475 K), a second stage combustor must be added
and it has to be a homogeneous combustor. We expect that almost all NOx emission
from this two-stage combustor system will come from the homogeneous combustor.
Fortunately, the gas mixture entering the homogeneous combustor is already at an
elevated temperature, we can try to push the fuel-air ratio as lean as possible to
minimize NOx generation without getting into flame stability issues. If the exit
temperature of the second combustor becomes too high for the turbine rotor, a third
stage can be added to cool the gas down with excess air.

2011 Reaction Design 182 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

3.1.1.2 Problem Setup

The CHEMKIN project file for this tutorial problem is called
reactor_network__two_stage_catalytic_combustor.ckprj located in the default
sample directory. We are going to use methane as the main fuel in our micro-turbine
combustor system, so we choose GRI Mech 3.0, as described in section
Section 2.9.2, to handle the gas phase combustion and, for the catalytic combustion,
the surface reaction mechanism developed by Deutschmann et al.41 for methane
oxidation on platinum catalyst, as described in section Section 3.3.1, is employed.
The chemistry set, reactor_network__two_stage_catalytic_combustor.cks, which
includes the gas phase (GRI methane oxidation mechanism) and surface
(Deutschmanns methane/platinum catalytic oxidation mechanism) reactions can be
found in the working directory samples2010\reactor_network\
two_stage_catalytic_combustor. Since we will use only the perfectly stirred reactor
(PSR) and plug flow reactor (PFR) in our model, we do not need any transport data.

Before building a model for our combustor system, we need to find out all the
important parameters of the system. The total mass flow rate of our sub-scale micro-
turbine system has to match the gas-turbine design point, which, in our case, is
980 g/sec. Since this designed flow rate is too large to be handled by a single
combustor, we divide the flow evenly into 6 identical combustor units in parallel and
merge them before entering the gas turbine. We set the mass flow rate of each
catalytic combustor to 127 g/sec so that there is about 21.5% (or 35 g/sec) excess air
per combustor for liner cooling and downstream dilution. Methane and compressed
air are pre-mixed before entering the catalytic combustor. We keep the fuel-air
mixture very lean so it will not ignite before reaching the catalytic combustor. The gas
temperature and pressure at the inlet of the catalytic combustor are 715 K and
3.75 atm, respectively. The inlet gas temperature is higher than that of a
homogeneous combustor. We have to use a higher inlet gas temperature to ensure a
light-off (or surface ignition) on the catalyst surface. Note that we could lower the inlet
temperature if a palladium-based catalyst were used or if a small amount of hydrogen
were added to the fuel. However, at this point, we consider methane as the only fuel
component.

The catalytic combustor is 10 cm in diameter and 10 cm long. It consists of a metal

outer liner providing structure support and a honeycomb monolith core. The catalyst is
coated on all internal surfaces of the monolith. The cell density of the honeycomb
monolith we selected is 400 cpsi (cells per square-inch) and the cell wall thickness is
0.18 mm. The pressure drop across the monolith is determined to be 0.064 psi/cm for
the given flow rate. 5.2 grams of platinum are wash-coated onto the internal surfaces

41. Deutschmann et al., Proceedings of Combustion Institute, 26:1747-1754 (1996).

CK-TUT-10112-1112-UG-1 183 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

of the honeycomb monolith. The metal surface area is 189 m2/g and the metal
dispersion is 70%. Based on the inlet stream properties, mass flow rate, and the
geometry of the honeycomb monolith, we find that the inlet velocity of the catalytic
combustor should be 1200 cm/sec.

Since we want to make the catalyst last longer to reduce operating costs, we need to
keep the catalyst surface below its maximum operating temperature. Theoretically the
maximum surface temperature in the catalytic combustor should not exceed the
adiabatic flame temperature of the inlet gas mixture. In other words, we can
determine the maximum equivalence ratio of the fuel-air mixture entering the catalytic
combustor by comparing the adiabatic flame temperatures against the maximum
catalyst operating temperature. The adiabatic flame temperature of a given
equivalence ratio can be easily obtained by using CHEMKIN's Equilibrium Reactor
Model. Of course, the minimum equivalence ratio of the inlet fuel-air mixture is the
one below which no light-off is observed on the catalyst surface. Accordingly, we set
the equivalence ratio to 0.185.

With all the basic information defined, we are ready to build a simple reactor network
model for our two-stage combustor system. We choose the Honeycomb Monolith
Reactor Model to represent the first-stage catalytic combustor and a Plug Flow
Reactor Model for the second-stage homogeneous combustor. Since all of the initial
fuel is expected to be consumed by the catalytic combustor, we have to inject
additional fuel to the second-stage homogeneous combustor. To achieve this, we
need to add a gas mixer between the Honeycomb Monolith Reactor and the Plug
Flow Reactor in our reactor network model. A fourth reactor, which can be either a
PSR or a PFR, is added after the PFR (the homogeneous combustor) to simulate the
post-flame flow in transition to the gas turbine and to allow the introduction of excess
air to cool down the flue gas if needed. Figure 3-2 shows the diagram of our
combustor system model that comprises four reactor clusters. We will run these
clusters in sequence.

Temperatures of both the additional fuel and the excess air are assumed to be the
same as the inlet temperature of the catalytic combustor. The mass flow rate of
excess air is the difference between the design flow rate of the gas turbine and the
exit mass flow rate of the homogeneous combustor. The amount of fuel added to the
homogeneous combustor should be able to raise the gas temperature high enough so
that, after excess air dilution, the gas temperature can still meet the required TRIT.
After a few iterations, we find the additional fuel mass flow rate is 1.5 g/sec.

2011 Reaction Design 184 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

Figure 3-2 Two-Stage Catalytic CombustorDiagram View

Now we can further set up our combustor system model by providing proper
information on each input panel. The catalyst properties and the honeycomb monolith
geometry are entered on two special sub-tabs (please see Figure 3-3 and Figure 3-4).
To find these two special sub-tabs, go to the Honeycomb Monolith Reactors Catalytic
Pre-combustor (C1_) panel and click on the Honeycomb Monolith tab.

Figure 3-3 Catalytic Pre-combustor (C1_)Honeycomb Monolith, Catalyst sub-tab

CK-TUT-10112-1112-UG-1 185 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

Figure 3-4 Catalytic Pre-combustor (C1_)Honeycomb Monolith, Honeycomb sub-tab

Since there is no surface reaction in the homogeneous combustor and the post-flame
mixer, we need to turn the surface chemistry calculations off in our reactor models.
There are several methods to make the models ignore the surface chemistry. Here we
choose to set the surface reaction rate multiplier to 0.0, using the parameter at the
bottom of the Reactor Physical Property tab as shown in Figure 3-5.

Figure 3-5 Homogeneous Stage Combustor (C3_)Reactor Physical Property

2011 Reaction Design 186 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

3.1.1.3 Project Results

We have to run the reactor clusters in order because calculations of later clusters
require information from the solution of the previous cluster. After finishing all four
cluster runs, we want to see if the predicted mass flow rate and gas temperature at
the exit of the last (fourth) reactor cluster match the gas turbine design point targets.
The predicted mass flow rate of each combustor system, which is 1/6 of the total
mass flow rate, and the turbine rotor inlet temperature can be found from the output
file from the last cluster run (see Figure 3-4).

Figure 3-6 Two Stage Catalytic CombustorExcess_Air_Dilution (Cluster 4) Output Results

TWOPNT: FINAL SOLUTION:

========================================================================================================================
PSPRNT: Printing of current solution from TWOPNT:
========================================================================================================================

RESIDENCE TIME 2.0000E-03 SEC

MASS FLOW RATE 1.6344E+02 GM/SEC
PRESSURE 3.706 ATM
MASS DENSITY 8.9763E-04 GM/CM^3
VOLUME 364.2 CM^3
TOTAL MASS 0.3269 G
TOTAL SURFACE AREA 0.000 CM^2
TOTAL SURFACE TO VOLUME RATIO 0.000 CM-1
GAS CHEM HEAT PRODUCTION 2.4785E-02 CAL/S/CM^3
TEMPERATURE (INLET: C4_Inlet1) 715.0000 K
TEMPERATURE (INLET: homogeneous_stage_combustor_(C3)_to_excess_air_dilution_(C4_R1)) 1604.8328 K
TEMPERATURE 1431.7571 K
SURF TEMP, CATALYST 1431.7571 K (same as gas temp)
HEAT LOSS, CATALYST 0.000 CAL/SEC
SURF CHEM HEAT PRODUCTION,
CATALYST 0.0000E+00 CAL/S/CM^2

EXIT GAS PHASE MOLE FRACTIONS

H2 = 2.4345E-07 H = 6.4644E-09 O = 1.2175E-06

O2 = 0.1415 OH = 4.3863E-05 H2O = 6.1973E-02
HO2 = 1.7455E-07 H2O2 = 2.1735E-08 C = 2.7918E-29
CH = 6.7819E-26 CH2 = 1.8560E-21 CH2(S) = 2.5248E-23
CH3 = 3.7230E-20 CH4 = 8.2935E-20 CO = 6.7122E-07
CO2 = 3.0997E-02 HCO = 1.5349E-17 CH2O = 5.8356E-19
CH2OH = 4.2927E-24 CH3O = 1.6348E-23 CH3OH = 3.0972E-20
C2H = 1.2566E-22 C2H2 = 1.3749E-17 C2H3 = 5.2738E-26
C2H4 = 4.7943E-25 C2H5 = -4.6945E-30 C2H6 = -3.4532E-28
HCCO = 1.4424E-21 CH2CO = 1.4861E-18 HCCOH = 3.2210E-15
N = 1.6160E-15 NH = 1.6167E-15 NH2 = 1.1761E-14
NH3 = 1.9500E-13 NNH = 2.5377E-14 NO = 2.1761E-07
NO2 = 2.4198E-09 N2O = 2.8742E-07 HNO = 5.3606E-14
CN = 1.0374E-19 HCN = 3.2462E-14 H2CN = 1.6896E-23
HCNN = 8.5484E-28 HCNO = 3.0537E-12 HOCN = 7.9801E-13
HNCO = 1.0132E-12 NCO = 4.8094E-15 N2 = 0.7655
AR = 0.000 C3H7 = 9.2007E-37 C3H8 = 1.0914E-37
CH2CHO = 9.6703E-26 CH3CHO = 3.8498E-28

Volatile Organic Compounds (ppm): 4.9039E-06

Unburned Hydrocarbons (ppm): 2.7620E-11

CO (ppmvd): 0.7156
NO (ppmvd): 0.2320
NOx (ppmvd): 0.5410
CO (ppmvd 15% O2): 0.7256
NO (ppmvd 15% O2): 0.2352
NOx (ppmvd 15% O2): 0.5486

CK-TUT-10112-1112-UG-1 187 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

SURFACE SITE FRACTIONS IN SURFACE PHASE, PT_SURFACE

Site density = 2.7063E-09 mole/cm^2
Standard State Site density = 2.7063E-09 mole/cm^2
Rate of change of site density = 0.000 mole/(cm^2*sec)

PT(S) = 9.0909E-02 H(S) = 9.0909E-02 H2O(S) = 9.0909E-02

OH(S) = 9.0909E-02 CO(S) = 9.0909E-02 CO2(S) = 9.0909E-02
CH3(S) = 9.0909E-02 CH2(S)s = 9.0909E-02 CH(S) = 9.0909E-02
C(S) = 9.0909E-02 O(S) = 9.0909E-02

BULK PHASE MOLE FRACTIONS AND ACTIVITIES IN BULK PHASE, PT_BULK

Linear growth rate of this bulk phase = 0.000 cm/sec
Total growth rate of this bulk phase = 0.000 gm/sec
Density of the bulk phase = -1.000 gm/cm^3
Average molecular weight of bulk phase = 195.1 gm/mole

Species Name Mole_frac Activity Density ------------------------Growth Rate-----------------------

(gm/cm^3) mole/(cm^2*sec) gm/(cm^2*sec) cm/sec (microns/hr)

PT(B) = 1.000 1.000 -1.000 0.000 0.000 0.000 0.000

========================================================================================================================

The predicted mass flow rate for each combustor is 163.4 g/sec so the total mass flow
rate is 980.4 g/sec (= 6*163.4). The predicted exit gas temperature, TRIT, is 1432 K.
Both values are very close to the targets. Since our goal is a zero-NOx combustor, we
want to find out NO, NO2, and N2O emissions from our new combustor. The solution
shows the mole fractions of NO, NO2, and N2O are 0.22 ppm, 0.003 ppm, and
0.27 ppm, respectively. All these concentrations are below 1 ppm and are not
detectable by instruments. Before we can say a job well done, we need to check on
CO and UHC (unburned hydrocarbon) emissions as well. Sometimes CO and UHC
concentrations increase when we try to minimize NOx formation. Our model indicates
our combustor has sub-ppm CO emission (~0.7 ppm) and essentially no UHC.

We are also interested in knowing how the gas temperature varies inside the
combustor system and whether the maximum temperature inside the catalytic
combustor exceeds its safe operating temperature. We can use the CHEMKIN Post-
Processor to obtain profiles along the two-stage combustor for quick visual
confirmation. We only need to load solutions of the first (catalytic combustor) and the
third (homogeneous combustor) clusters into the CHEMKIN Post-Processor because
the other two clusters yield a single solution point each. The axial profiles of gas
temperature, pressure, and mole fractions for CH4, CO and NO are shown in the
following figures.

2011 Reaction Design 188 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

Figure 3-7 Two Stage Catalytic CombustorTemperature Comparison

Figure 3-8 Two Stage Catalytic CombustorPressure Comparison

CK-TUT-10112-1112-UG-1 189 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

Figure 3-9 Two Stage Catalytic CombustorCH4 Comparison

Figure 3-10 Two Stage Catalytic CombustorCO Comparison

2011 Reaction Design 190 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

Figure 3-11 Two Stage Catalytic CombustorNO Comparison

The temperature profile (Figure 3-7) indicates that the gas temperature is increased in
two steps. The catalytic combustor has a lower operating temperature and only raises
the gas temperature to about 1100 K. The homogeneous combustor, which can
operate at higher temperatures, further raises the gas temperature to more than
1600 K before the excess air cools the gas mixture down to the target TRIT at about
1450 K. The CO profile (Figure 3-10) has a spike inside the homogeneous combustor
corresponding to the gas-phase ignition and all CO generated is later consumed in
the post-flame region. The model also predicts that NOx is formed after gas-phase
ignition and, unlike CO, its concentration continues to grow in the post flame region.

Note that this project only addresses the limitation posted by chemical kinetics. Mass
transport, i.e., diffusion and turbulence mixing, can become the limiting factor in
determining the performance of this gas combustor system. At high pressure, species
transport between the bulk gas and the active surface can be the rate-limiting factor
for the catalytic combustor. Poor molecular diffusion at high pressure could also affect
the homogeneous combustor. When the mixing between injected fuel and oxygen in
the hot gas slows down, the ignition distance becomes longer and could even cause
flame-out in the homogeneous combustor. The tutorial in Section 2.6.4 addresses
some of these issues.

3.1.2 Engine Exhaust Aftertreatment with a Transient Inlet Flow

3.1.2.1 Project Description

This user tutorial demonstrates how you can employ a user-defined subroutine to
read in transient engine-out data to set dynamic inlet conditions for an engine-exhaust
aftertreatment simulation. In this sample, the engine-out conditions as a function of
time are stored in a tab-delimited text file. A user-editable subroutine is provided that

CK-TUT-10112-1112-UG-1 191 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

reads the contents of this file and extracts from it the time-dependent inlet
composition, temperature, and instantaneous flow rate into a transient perfectly
stirred reactor model. The reactor approximates a 3-way catalytic converter, designed
to convert NOx, CO, and un-burned hydrocarbons (UHCs) through catalytic surface
reactions on a platinum/rhodium catalyst. For this sample, gas-phase chemistry is
neglected due to the dominance of the surface conversion reactions. This sample
user routine is already compiled and linked into the standard installation, such that it
can be run from the CHEMKIN Interface without the need of a FORTRAN compiler.
Modifying the behavior of the sample user routine will require access to an a
compatible FORTRAN compiler on the computer where CHEMKIN is installed. With
the pre-compiled routine, however, you can change the numerical values of the
engine-out data that is read by the user routine.

3.1.2.2 Project Setup

The project file is called psr__aftertreatment.ckprj. Catalytic conversions of NOx,
CO, and UHCs are modelled by the Pt/Rh three-way catalyst surface mechanism
described in Section 3.3.2. The data files used for this sample are located in the
samples2010\psr\aftertreatment directory. This reactor diagram contains only one
gas inlet and one perfectly-stirred reactor model. The sample user subroutine that is
used to define the inlet is provided in the standard CHEMKIN installation location, in
the user_routines sub-directory. The FORTRAN subroutine is named USRINLET
and is included in the file aurora_user_routines.f. The subroutine is discussed in more
detail below.

Before running this sample problem from the CHEMKIN Interface, the Display User Routines
box on the General tab of the User Preferences panel must be checked. Open the
Preferences panel from the Edit menu.

3.1.2.2.1 Reactor and Inlet Settings

On the Reactor Physical Properties tab of the C1_PSR panel, first the problem type is
selected as Solve Gas Energy Equation while selecting the Transient Solver. In this
case, the reactor volume is the void volume in the converter honeycomb or porous
media, estimated here as 1400 cm3 for the entire length of the converter. Similarly,
the reactive (internal) surface area within the converter is estimated based on the void
geometry to be 59000 cm2. No residence time is input because the volume and flow
rates are specified. The initial temperature (296.15 K) is input, along with the pressure
(1 atm), and the fact that the system is treated as adiabatic (no heat loss).

2011 Reaction Design 192 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

On the Species-specific Properties tab of the C1_ PSR panel, the initial conditions of
the converter system, prior to the exhaust gas flowing through the system, are
assumed to be air and are input on the Initial Gas Fraction sub-tab. The surface site
fractions are estimated based on understanding of the surface conditions or on initial
tests of the mechanism. A user-supplied initial guess can sometimes aid
convergence; if it is not specified CHEMKIN will assume a uniform distribution of
sites. In this case, we specify O(S) as the dominant site at the start for the platinum
portion of the catalyst and CO(S1) for the rhodium portion. These site species are set
to 1.0 on the Surface Fraction sub-tab to provide the initial conditions for the catalyst
surface.

The gas inlet has been given a name that reflects its function, engineout. On the
Stream Properties Data tab of the engineout panel, the Use Inlet User Routine option
is selected, which indicates that information about the gas inlet composition, flow rate
and temperatures should be obtained from the user inlet routine. The user routine
pre-packaged with CHEMKIN reads a text file. The name of this file, engineout.txt, is
set in the FORTRAN subroutine. The inlet conditions, some of which are shown in
Table 3-1, are representative of measurements that might be taken during an engine
test, where engine load and therefore exhaust flow rates and composition vary as a
function of time. In this case, mole fractions for only a few species are provided (CO,
NO, UHCs, and O2). We will assume that the balance of the gas can be represented
by N2 and that C3H6 will chemically represent the UHCs. On this panel, the Use
SCCM for User Inlet Routine box is also checked to indicate that flow rates are
given in volumetric flow units, rather than mass flow units.

On the Basic tab of the Solver panel, the end time of the simulation is set to 100 sec.
Although the input data file extends to longer times, this is sufficient for this
demonstration. The maximum Solver Time Step is set to 1 msec, the Time Interval for
Printing the solution to the text output file is set to 10 sec, and the Time Interval for
Saving Data is set to 1 sec. There are no inputs on the Output Control or
Continuations panels for this problem.

Table 3-1 Excerpt of Data Representing Engine-out Test Measurements

Time(s) InletT(C) Flow(SLM) C3xx(ppm) CO(ppm) CO2(ppm) NOx(ppm) O2(ppm)

0 23.9 87.08 924 4341 22764 38 127856

0.5 25 85.49 1238 23557 49535 127 77187

1 20.1 27.15 1369 36212 66499 126 50914

1.5 22.6 81.89 1267 37984 80793 118 28314

2 39.1 81.2 1176 29212 95735 92 14359

CK-TUT-10112-1112-UG-1 193 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

Table 3-1 Excerpt of Data Representing Engine-out Test Measurements (Continued)

Time(s) InletT(C) Flow(SLM) C3xx(ppm) CO(ppm) CO2(ppm) NOx(ppm) O2(ppm)

2.5 66.2 83.17 1016 18119 104492 89 9577

3 88.2 86 999 10391 110148 108 7744

3.5 111.3 87.08 912 5786 111276 101 13012

4 125.4 85.42 877 3663 112346 111 8645

4.5 141.2 82.53 843 2532 111864 104 11867

5 163 81.1 808 1967 111803 110 16508

5.5 175.2 82.43 681 1619 111685 101 12255

6 189.8 85.28 756 2224 110433 119 15728

6.5 208.9 87.06 825 6022 108765 207 7865

7 218.2 86.12 825 11532 107414 366 12835

7.5 230.6 83.34 732 9837 105707 440 20696

8 244.3 81.27 857 5259 102752 446 24211

8.5 251.6 193 957 2823 100132 381 28629

9 272.1 374.28 991 2942 99228 354 28534

9.5 297.4 479.67 864 2355 101356 331 28906

10 313.6 483.61 851 1514 103822 309 29511

10.5 333.2 431.87 796 1343 105570 276 23807

11 347 388.16 788 1251 106826 241 26600

11.5 349.6 344.94 868 910 106112 207 26409

3.1.2.2.2 Details of the User Routine

The subroutine for the user-defined inlet is reproduced in Figure 3-12. The data in
Table 3-1 has units of parts-per-million for species composition, standard liters per
minute for flow rate, and degrees Celsius for temperature, which are not the standard
units used by the CHEMKIN software. The user routine must therefore perform
necessary conversions before providing the data to the PSR program executable. It is
also responsible for interpolating between data points for each time value needed
during the time integration. At each time step during the transient simulation, the PSR
program will call USRINLET. Parameters needed by the USRINLET subroutine are
passed in through integer and real workspace arrays called IINWRK and RINWRK,
respectively. The comments at the top of the code explain what parameters are stored
in these arrays.

2011 Reaction Design 194 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

The sample USRINLET code in Figure 3-12 reads in data from a file on the first call,
as determined when the time is equal to the starting time of the simulation. In this
section, the routine uses the CHEMKIN library routine CKCOMP to find the location of a
given species name in the gas-phase species array (KNAMES). Conversion from
Celsius to Kelvin and from parts-per-million to mole fraction are performed as the data
is read and stored on the first call. On subsequent calls, the data is accessed from
memory (stored in a FORTRAN common block), and values are interpolated for the
specified simulation time. The interpolation uses the CHEMKIN GAS-PHASE KINETICS
library routine CKBSEC to linearly interpolate between the data points. Before
returning, the routine converts the flow rate from standard liters per minute (slm) to
standard cubic centimeters per minute (sccm). Note that the routine could have been
written to return mass flow rate in units of g/sec. In this case, however, CHEMKIN
Interface inputs tell the program to expect user-defined flow rates in units of sccm.

CK-TUT-10112-1112-UG-1 195 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

Figure 3-12 Sample USRINLET Subroutine for User-Defined Transient Inlet Conditions

SUBROUTINE USRINLET (LIUIN, IINWRK, LRUIN, RINWRK, INAME, KNAMES,

1 FLRT, TINL, TEIN, XIN)
!DEC$ IF DEFINED (DLLEXPORT)
!DEC$ ATTRIBUTES DLLEXPORT :: USRINLET
!DEC$ ENDIF
C This is a USER SUBROUTINE for defining Inlet properties
C as an arbitrary function of time.
C Use of this subroutine is controlled by the PSR program
C keyword, USRIN, as described in the CHEMKIN Input Manual.
C
C The subroutine USRINLET is used to supply values of
C FLRT - Real scalar, Mass flow rate in g/s
C (or if SCCM are the preferred units,
C use keyword LFPSC in addition to USRIN)
C TINL - Real scalar, Inlet temperature (K)
C TEIN - Real scalar, Inlet electron temperature (K)
C XIN(*) - Real array, Gas-phase reactant composition of
C the inlet (mole fraction);
C The length of this array is *exactly* KKGAS,
C the number of gas-phase species in the problem.
C and the user may set the Integer flag
C IINWRK(1) - if 0, inlet parameter setting was successful,
C else, there was a problem, so discontinue calculations
C
C Given the following data:
C
C LIUIN - Integer scalar, length of some Integer workspace
C IINWRK(*) - Integer workspace array, containing
C IINWRK(2), LOUT - if positive, the unit number of an open output file
C IINWRK(3), IPSR - PSR index number
C IINWRK(4), IINL - Index number of an Inlet of IPSR
C IINWRK(5), KKGAS- Gas-phase species count
C IINWRK(6), LENRGY-If 0, then inlet temperature TINL is fixed,
C else TINL must be supplied here
C IINWRK(7), LENRGE-If 0, then inlet electron temperature TEIN is fixed,
C else TEIN must be supplied here
C
C LRUIN - Integer scalar, length of some Real workspace
C RINWRK(*) - Real workspace array, containing
C RINWRK(1), TSTART - Initial time (sec.) of calculation
C RINWRK(2), TIM - Current time (sec.)
C
C INAME - Character-string Inlet name
C KNAMES- Character-string array, Gas-phase species names
C*****************************************************************
IMPLICIT DOUBLE PRECISION (A-H, O-Z), INTEGER (I-N)
include 'user_routines_interface.inc'
PARAMETER (PATM = 1.01325D6,TSTD=273.15,ZERO=0.0D0)
C Variables passed in from calling routine
DIMENSION IINWRK(LIUIN), RINWRK(LRUIN)
CHARACTER*16 INAME, KNAMES(*)
C Arrays returned by this user routine
DIMENSION XIN(*)
C Local variables:
CHARACTER*80 MYFILE
PARAMETER (MYFILE = 'engineout.txt',LUNIT=33,
1 MXPTS=10000,MXSPEC=6)
C Local storage space for variables
CHARACTER*16 MYSPEC(MXSPEC)
CHARACTER*80 HEADER
LOGICAL LENRGY,LENRGE
DIMENSION PPM(MXSPEC)
COMMON/USRINL1/ MAPSP(MXSPEC),NPTS
COMMON/USRINL2/ TIMEPT(MXPTS),SLMPT(MXPTS),XINPT(MXPTS,MXSPEC),
1 TPT(MXPTS)
EXTERNAL CKUFIRST
SAVE IFIRST
DATA IFIRST/0/
C
C set error flag
IINWRK(1) = 0
C set local variables from workspace data provided
LOUT = IINWRK(2)
C acknowledge that USRINLET has been applied
IF (IFIRST .EQ. 0) CALL CKUFIRST (IFIRST, LOUT, 'USRINLET')
C

2011 Reaction Design 196 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

IPSR = IINWRK(3)
IINL = IINWRK(4)
KKGAS= IINWRK(5)
LENRGY=IINWRK(6).GT.0
LENRGE=IINWRK(7).GT.0
C
TSTART=RINWRK(1)
TIME =RINWRK(2)
C
C Initialize returned variables
FLRT = 0.0
TINL = 298.D0
TEIN = 298.D0
DO K = 1, KKGAS
XIN(K) = 0.0D0
ENDDO
C
C First time in, read in all the points so we don't have to do
C IO on each call. Interpolate from saved points thereafter
IF (TIME .EQ. TSTART) THEN
C Open and read the time-date file
C Store points in arrays for access/interpolation at later times
IOS = 0
OPEN(LUNIT, FILE=MYFILE, FORM='FORMATTED', STATUS='OLD',
1 IOSTAT=IOS)
C Check that file open was successful, if not return with error
IF (IOS .NE. 0) GO TO 1000
C
C Map input species to CHEMKIN names and find indices
MYSPEC(1) = 'C3H6'
MYSPEC(2) = 'CO'
MYSPEC(3) = 'CO2'
MYSPEC(4) = 'NO'
MYSPEC(5) = 'O2'
MYSPEC(6) = 'N2'
DO MYK = 1, MXSPEC
CALL CKCOMP(MYSPEC(MYK),KNAMES,KKGAS,INDX)
IF (INDX .GT. 0 .AND. INDX .LE. KKGAS) THEN
MAPSP(MYK) = INDX
ENDIF
ENDDO
C Read in the arrays of available information
C In this case file format is as follows:
C Time(s),T(C),Flrt(SLM), C3H6(ppm),CO(ppm),CO2(ppm),NO(ppm),O2(ppm)
READ(LUNIT, '(A)', END=800, ERR=1000) HEADER
NPTS = 0
DO I = 1, MXPTS
READ(LUNIT, *, END=800, ERR=1000)
1 TIMEPT(I),TCELS,SLM,(PPM(K),K=1,MXSPEC-1)
NPTS = NPTS + 1
TPT(I) = TCELS + 273.15D0
SLMPT(I) = SLM
XSUM = 0.0D0
DO MYK = 1, MXSPEC-1
XINPT(I,MYK) = PPM(MYK) * 1.D-6
XSUM = XSUM + XINPT(I,MYK)
ENDDO
C Set fraction of N2 = 1 minus the sum of others
XREM = 1.0D0-XSUM
XINPT(I,MXSPEC) = MAX(XREM,ZERO)
C Note: if XSUM > 0.0, AURORA will normalize so that sum = 1
ENDDO
800 CONTINUE
IF (NPTS .EQ. 0) GO TO 1000
CLOSE(LUNIT)
ENDIF
C
C Interpolate data for input time and perform units conversions
XSUM = 0.0
DO MYK = 1, MXSPEC-1
XIN(MAPSP(MYK)) = CKBSEC(NPTS,TIME,TIMEPT,XINPT(1,MYK))
XSUM = XSUM + XIN(MAPSP(MYK))
ENDDO
XREM = 1.0-XSUM
XIN(MAPSP(MXSPEC)) = MAX(XREM, ZERO)
SLM = CKBSEC(NPTS,TIME,TIMEPT,SLMPT)
C Convert from SLM to SCCM
FLRT = SLM * 1000.D0
IF (LENRGY) THEN
C Set the gas inlet temperature
TINL = CKBSEC(NPTS,TIME,TIMEPT,TPT)

CK-TUT-10112-1112-UG-1 197 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

ENDIF
IF (LENRGE) THEN
C Set the electron inlet temperature
TEIN = TINL
ENDIF
RETURN
C
1000 CONTINUE
IF (IOS .NE. 0) THEN
WRITE (LOUT, *) ' ERROR...OPEN failure on inlet data file'
CALL CKWARN(2)
ELSE
WRITE (LOUT, *) ' ERROR...READ failure on inlet data file'
CALL CKWARN(2)
CLOSE (LUNIT)
ENDIF
IINWRK(1) = 1
C
RETURN
END

3.1.2.3 Project Results

Figure 3-13 shows molar conversions for C3H6, CO, NO as a function of time. Note
that "molar conversion" variables must be selected on the Select Post-Processing
Variables panel in order for molar conversion variables to be available for plotting. The
effectiveness of conversion can also be viewed by comparing the inlet mole fraction to
the outlet mole fractions, as shown in Figure 3-14 for C3H6 or by looking at the
calculated conversion efficiencies directly. These figures clearly show that C3H6 and
CO are converted more effectively than NO under these conditions. Early in the
simulation, the conversion rates are low for all of these species. Although at t = 0, the
calculated molar conversion is 100%, this is simply due to setting the initial conditions
in the reactor to pure air, which determines the initial exit flow. At certain times the
calculated conversion rates for CO and NO actually go negative. This results from the
fact that CO and NO can be formed on the surface and thus can be produced as the
state of the surface changes. Figure 3-15 shows inlet and exit gas temperatures as a
function of time. Temperatures show that the gas heats up relative to the inlet gas due
to exothermic surface reactions. The catalyst does not become effective until it
reaches a temperature above about 600 K, which is consistent with the work reported
by Chatterjee et al.42

42. D. Chatterjee, O. Deutschmann and J. Warnatz, Detailed Surface Reaction Mecha-

nism in a Three-way Catalyst, Faraday Discussions, 119:371-384 (2001).

2011 Reaction Design 198 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

Figure 3-13 Engine Exhaust AftertreatmentMolar Conversion Rates

Figure 3-14 Engine Exhaust AftertreatmentMole Fractions

CK-TUT-10112-1112-UG-1 199 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

Figure 3-15 Engine Exhaust AftertreatmentGas Temperatures

3.2 Parameter Studies

3.2.1 Parameter Study Facility for Surface Chemistry Analysis
The science of surface chemistry has undergone rapid development in recent years,
and there is a trend towards use of more detailed microkinetic models, especially in
the field of catalysis. However, compared with gas-phase combustion problems,
surface reaction mechanisms often contain larger uncertainties in reaction-rate
coefficients that must often be estimated by semi-empirical methods. For this reason,
and because there are variabilities in the catalyst properties, it is often necessary to
adjust kinetic parameters to model a specific catalytic system. Thus it is of interest to
see how uncertainties that exist in surface mechanisms affect predicted modeling
results. Performing a parameter study, which varies reaction-rate parameters for key
reactions in the system, is one way to analyze these effects.

In this tutorial, we apply a parameter study to investigate the effects of one reaction
rate coefficient for a simulation involving catalytic oxidation of methane on a platinum
catalyst. The associated project file is called
honeycomb_monolith__reaction_rate_param_study.ckprj. The chemistry-set files
used for this sample problem are located in the
honeycomb_monolith/reaction_rate_param_study folder of your CHEMKIN
samples2010 directory. For the gas-phase kinetics, we employ the GRI gas-phase
mechanism and thermodynamic data for methane combustion. For the catalytic
surface-chemistry that describes the oxidation of methane on a platinum catalyst, we

2011 Reaction Design 200 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

use the surface chemistry mechanism reported by Chou, et al.43 In this examples
reactor model, we focus on the first stage (Honeycomb Catalytic Reactor) of the Two-
stage Catalytic Combustor Sample
(reactor_network__two_stage_catalytic_combustor.ckprj).

3.2.1.1 Determining Influential Reactions with Sensitivity Analysis

We first intend to study those surface reactions that will be the most influential to our
output parameters of interest for the conditions in this sample. Since this is a catalytic
combustion case, where the catalytic stage serves to preheat the gas, the reactor
temperature is of interest. We therefore first run a sensitivity analysis in order to
determine the sensitivity of Temperature to the Arrhenius pre-exponential factors (A-
factors) of the various reaction-rate constants. In order to accomplish this, first open
the honeycomb_monolith__reaction_rate_param_study project and pre-process
the chemistry set. Then look at the nominal case, which is set up in the Reactor and
Inlet panels and which can be run from the Run Calculations node on the project tree.

Follow the steps below to run the sensitivity analysis.

1. Double-click on the Output Control in the Open Projects tree.

2. On the Output Control tabs Basic sub-tab, select Temperature A-factor
Sensitivity.
3. Open the Run Calculations panel, notice that Parameter Study is selected,
and click the Begin button.
4. Answer the prompt about number of parallel runs and click OK. Note that the
number of runs for parallel can be changed later by editing the user
Preferences (see Getting Started with CHEMKIN).
5. Once the Parameter Study has completed successfully, the Analyze Results
panel displays. Select to Plot Results and click the Next Step button.
6. In the Post-Processor control panel, set Plot Type to Line Plot and use the
pull-down menu to select File_1: Honeycomb MonolithC1:sensitivity for
solution no 1.
7. Select Distance for the X Variable and select as Y Variables all of the surface
reaction sensitivities that are listed. (To select multiple variables, hold down
the Ctrl key while clicking each variable.)
8. Click on Display Plot.

43. C.-P. Chou, J.-Y. Chen, G. H. Evans and W. S. Winters, Combustion Science and Tech-
nology, 150:27 (2000).

CK-TUT-10112-1112-UG-1 201 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

From the plot results, you should see that Surface reactions #1,
(O2 + 2PT(*) O(*) + O(*)) and #6, (CH4 + O(*) + PT(*) CH3(*) + OH(*)), are
particularly important in determining the resulting temperature. In the case of both of
these reactions, the A-factors are actually sticking coefficients as indicated by the
keyword STICK in the surface-chemistry file.

It is now of interest to vary the sticking coefficients for these rates one by one, and see
how such variations will affect predicted temperature profiles in the reactor. A
Parameter Study has already been set up to do this in the
honeycomb_monolith__reaction_rate_param_study project. To see how this is
done, go to the Mechanism Parameters tab in the Pre-processing panel. On the
Reaction sub-tab, select COMBUST (which is the name of our surface material, so
we're selecting reactions related to this material only) from the pull-down menu of the
Select Reaction Set and click on the Search Reactions button. The results are
reported in the drop-down list at the bottom of the panel. The first reaction to show up
in the pull-down window under the Search Reactions button is highlighted in blue,
indicating that a parameter study has been set up for this reaction rate constant. Click
on the Setup Parameter Study button next to the Pre-exponential Factor A value
for this reaction to view the Parameter Study Setup, as shown in Figure 3-16.

Figure 3-16 Setting Up a Parameter Study for Sticking Coefficients vs. Predicted Temperature Profiles

2011 Reaction Design 202 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

3.2.1.2 Viewing Results with the Post-Processor

Run the Parameter Study by selecting all of the runs in the Run Parameter Study
window and clicking the Post-Process button. Once in the Post-Processor control
panel, select Line Plot for Plot Type. Under Plot Set, select Solution no 1. Ctrl-click
all of the temperature profiles for each of the Parameter Study runs as the Y
Variables. Click on Display Plot.

Figure 3-17 demonstrates the results of the Parameter Study displayed by the
CHEMKIN Post-Processor. The results clearly indicate that if we keep the sticking
coefficient of reaction 1 constant, and increase the sticking coefficient for reaction 6,
the ignition occurs much sooner and the maximum combustion temperature is
reached almost as soon as the reactants enter the catalytic reactor. This makes
sense, because reaction 6, CH4 + O(*) + PT(*) CH3 (*) + OH(*), is a branching
reaction that liberates radicals CH3 and OH, thus rapidly increasing the radical pool
that in turn is responsible for catalytic light-off. On the other hand, increasing the
sticking probability of the O2 + 2PT(*) O(*) + O(*) reaction delays ignition probably
because it is competing with the adsorption of CH4 on open sites.

Figure 3-17 Varying Rate Constants of Reactions 1 and 6.

CK-TUT-10112-1112-UG-1 203 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

3.3 Chemistry Sets

In this section, we discuss briefly the chemistry sets used in Chapter 3 and provide
references for further study.

3.3.1 Methane Oxidation on Pt

Two projects in this chapter involve methane oxidation on a platinum surface. This
demonstrates that the gas-phase mechanism does not need to align to a particular
surface mechanism by implementing two different surface mechanisms to describe
the same surface process methane oxidation on a platinum (Pt) surface.

The chemistry set of the two-stage catalytic combustor project (Section 3.1.1) uses a
slightly altered version of the surface mechanism published by Deutschmann et al.41
(see p. 183) in 1996 and put into CHEMKIN format by L. Raja. A different surface

mechanism developed by Chou et al.43 (see p. 201) in 2000 is employed by the

Parameter Study Facility for the surface chemistry analysis project (Section 3.2.1).

If gas-phase reactions are expected to be important, any reaction mechanism

describing the gas-phase kinetics for methane oxidation chemistry can be used, but
Pt needs to be added to the element list. In both projects, an altered version of the
GRImech 3.0 is provided as an example. In the case of a simulation where it is known
that gas-phase chemistry is not important, a simple gas-phase chemistry input file can
be used that only defines the elements and species that are included in the SURFACE
KINETICS input file.

The SURFACE KINETICS input file for Duetschmann's mechanism describes chemistry
occurring on one material called CATALYST that has one site type called
PT_SURFACE. The PT(B) bulk material is defined, but does not participate in any
reactions, as neither etching or deposition of platinum occurs in this system. The
surface site can be occupied by any of 11 surface species, where PT(S) is the open
platinum site, and the other surface species represent H, C or O atoms, CO, CO2 or
H2O molecules, OH, CH3, CH2, or CH radicals adsorbed on the platinum surface
atom. Note that GRImech has already used the symbol CH2(S) to represent gas-
phase methylene radicals in the singlet electronic state, such that adsorbed
methylene radicals have been given the symbol CH2(S)s. Thermochemical data for
surface species are included in the surface kinetics input file, and have been defined
using PT(S) as the reference point. There are 22 surface reactions, some reversible
and some irreversible. These reactions include simple and dissociative adsorption

2011 Reaction Design 204 CK-TUT-10112-1112-UG-1

Chapter 3: Catalytic Processes Tutorials Manual

reactions, as well as simple and associative desorption reactions, plus reactions

between adsorbed species. Some reactions are described in terms of sticking
coefficients, a few reactions have reaction-order overrides, and in two cases, the
activation energy for an adsorption reaction varies with surface coverage.

The surface mechanism published by Chou et al. in 2000 describes the oxidation of
methane on the surface of a supported platinum catalyst under fuel-lean conditions.
The surface reactions take place on the PT_POLY surface phase (site type) of the
material. Bulk species PT(B) and surface species PT(*) represent the platinum atom
covered by a surface species and the bare surface platinum atom (or open surface
site), respectively. There are 9 other surface species that can exist on the platinum
surface (H, C, O, CO, H2O, OH, CH3, CH2, and CH), and their thermodynamic data
are provided in the SURFACE KINETICS input file. Since definitions of these surface
species (given as part of their thermochemical data) do not contain a platinum atom,
the PT(B) bulk species in Chou's mechanism must participate in some of the surface
reactions so that these reactions are balanced. The conservation of platinum atom is
important because, as a catalyst, platinum should not be created or consumed by the
surface reactions. Chou's surface mechanism consists of 23 irreversible reactions.
These reactions describe simple and dissociative adsorption of reactants, simple and
associative desorption of products, and reactions between adsorbed species. Some
adsorption rates are given in terms of sticking coefficients.

3.3.2 Pt/Rh Three-way Catalyst

This chemistry set describes several processes that occur on a platinum/rhodium
three-way catalyst. These processes are: the oxidation of unburned bicarbonates
(represented y C3H6) on Pt, the reduction of NO on Pt, the reduction of NO on Rh,
and the oxidation of CO on Rh. This reaction mechanism is based on the published
work of Chatterjee, et al.44 in 2001. This reaction mechanism was developed for a
surface made of 75% Pt and 25% Rh, and should be considered valid only for this
composition. In particular, the coverage dependent activation energies have been
scaled for that Pt/Rh ratio.

Gas-phase kinetics is ignored in this chemistry set. The GAS-PHASE KINETICS input file,
therefore, includes 6 elements: O, H, C, N, Rh and Pt; 9 species: O2, C3H6, H2, H2O,
CO2, CO, NO, NO2, and N2; and no reactions.

44. D. Chatterjee, O. Deutschmann and J. Warnatz, Detailed Surface Reaction Mecha-

nism in a Three-way Catalyst, Faraday Discussions, 119:371-384 (2001).

CK-TUT-10112-1112-UG-1 205 2011 Reaction Design

CHEMKIN Tutorials Manual 10112 Chapter 3: Catalytic Processes

The SURFACE KINETICS input file describes chemistry occurring on one material called
3WAYCATALYST that has two site types each representing one of the metals, called
PLATINUM and RHODIUM. The PLATINUM site can be occupied by any of 18 surface
species, of which Pt(S) is the open site, and one of which, the C3H6(S) species,
occupies two sites. The RHODIUM site has simpler chemistry with only 5 surface
species, of which Rh(S1) is the open site. Thermochemical data are provided for
some of the surface species, but others have placeholder values. All surface
reactions are irreversible, such that none of the thermochemical data are used to
obtain rates for reverse reactions. The PLATINUM site in this reaction mechanism has
a site density of 2.04E-9 moles cm-2, which is close to the number of 2.717E-
9 moles cm-2 that is obtained from the density and molecular weight of solid
platinum. This indicates that the creators of this mechanism chose to describe some
of the effects of the high-surface-area nature of the catalyst material by increasing the
site density, rather than by increasing the effective surface area for chemistry. This
aspect should be kept in mind in applying this mechanism to other systems.

The oxidation of C3H6 (or UHCs) on platinum is described in 47 irreversible reactions

that include simple and dissociative adsorptions, simple and associative desorptions,
plus reactions between adsorbed species. The latter include decomposition of
adsorbed hydrocarbon species by hydrogen transfer to Pt(S) species, as well as
various oxidation reactions for adsorbed hydrocarbon species ranging from a global
description of C3H5(S) oxidation to more step-wise oxidation reactions for other
hydrocarbon fragments. NO reduction on platinum is described in 5 reactions, while
NO reduction and CO oxidation on rhodium are described in another 9 reactions,
each involving adsorptions, desorptions and reactions among adsorbed species.
Some reactions are described in terms of sticking coefficients, a few reactions have
reaction-order overrides, and in several cases, the activation energy for a reaction
varies quite strongly with the extent of coverage of the surface by one or more
species.

2011 Reaction Design 206 CK-TUT-10112-1112-UG-1

 CHEMKIN

4 Materials Problems

4.1 Chemical Vapor Deposition

A number of CHEMKIN Reactor Models can be used for CVD simulations. Thermal
CVD processes generally involve furnaces or heated surfaces. Simulations of CVD
processes, therefore, often treat the temperature of the reactor or deposition surfaces
a fixed, experimentally determined input parameter, rather than a quantity that is
obtained by solving an energy equation. The surface temperature is usually assumed
to be equal to the temperature of the adjacent gas, except for processes performed at
very low pressures. Many CVD processes also have fairly long run times compared
with startup and end-of-process transients, such that a steady-state simulation is
often a good representation of a CVD process.

Generally, a low-dimensional simulation, such as an Equilibrium calculation or a PSR

simulation, will be used to assess the chemistry, and possibly to simplify a reaction
mechanism before using it in higher-dimensional simulations. A multiple-phase
equilibrium calculation can be used to determine the maximum possible deposition
rate for a certain gas mixture, pressure and temperature. In a PSR simulation of a
CVD process, the assumption is made that the chemical kinetics are rate limiting
rather than mass transport effects or the inlet reagent supply rate. Higher dimensional
simulations allow evaluation of the relative effects of chemical kinetics and mass
transport. The reactor models in the CHEMKIN software can be used to simulate a
number of reactor geometries used for CVD. The Plug Flow Reactor and Cylindrical
Shear Flow Reactor Models can be used to model tube-furnace CVD systems, while
the Planar Shear Flow Reactor can be used to model horizontal flow CVD systems.
The Stagnation Flow Reactor Model can simulate a vertical showerhead system,
while the Rotating Disk Reactor Model is for reactors in which high-speed rotation of
the plate dominates the gas flow field. There are also specialized models for the
thermal analysis and modeling of the multi-wafer batch low-pressure CVD furnaces
used in the fabrication of microelectronic devices.

CK-TUT-10112-1112-UG-1 207 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

4.1.1 Equilibrium Analysis of Chlorosilane CVD

4.1.1.1 Project Description

This user tutorial presents a multi-phase equilibrium calculation of the Si-Cl-H system.
A constant pressure and temperature equilibrium calculation gives the maximum
amount of solid product that could be formed in the absence of kinetic or transport
limitations. This can be a valuable screening tool in choosing operating conditions for
a CVD process. In this case, adding HCl to a mixture of SiCl3H, H2, and solid silicon
alters the Si/Cl ratio in the system enough to change the equilibrium composition from
one that would result in deposition of more solid Si, to one that would result in etching.

4.1.1.2 Project Setup

The project file is called equilibrium__siclh_cvd.ckprj. The data files used for this
sample are located in the samples2010\equilibrium\siclh_cvd directory. This
diagram contains only one Equilibrium calculator.

The equilibrium calculation only needs a list of species with their thermodynamic data;
a reaction list is not needed. It is important to include all likely radical species as well
as the desired and undesired product species in the calculation. It is generally better
to include some unimportant species than to leave out ones that turns out to be
important. The chem.inp file includes 3 elements and 22 gas-phase species and no
reactions. This file contains more species that the trichlorosilane CVD chemistry set
described in Section 4.4.3. In the interest of completeness, a number of species such
as SiH4 or atomic Cl were added that are not expected to be important at standard
CVD conditions, but that might be more important under different conditions. The
surf.inp file includes only solid silicon in the bulk (condensed) phase.

Setting up this problem first involves the C1_ Equilibrium panel. The problem type,
temperature (1400 K) and pressure (1 atm) are entered on the Reactor Physical
Property tab. The starting composition is entered on the Reactant Species tab.
Starting with a molecular mixture will give the same equilibrium composition as
starting with elemental Si, H, and Cl with the same Si/Cl and H/Cl ratios. The
Continuations panel is used to specify three additional simulations where increasing
amounts of HCl are added, replacing some of the hydrogen carrier gas. All
components of the starting mixture have been re-entered, because the composition
must be entered as a set.

2011 Reaction Design 208 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

4.1.1.3 Project Results

Figure 4-1 shows the initial and equilibrium amounts of solid silicon for an initial
mixture of 10% (by volume) SiCl3H in hydrogen with increasing amounts of HCl at
1 atm total pressure and 1400 K. With no added HCl, the equilibrium mole fraction of
solid silicon is larger than the initial amount, indicating that this gas mixture is likely to
result in the deposition of silicon. As the HCl content increases, the mole fraction of
Si(B) expected at chemical equilibrium decreases below the initial mole fraction. This
initial gas mixture is thus expected to result in etching, rather than deposition of Si
under these conditions. This primarily results from the decreasing Si/Cl ratio in the
system. As shown in Figure 4-2, the most prevalent gas-phase species at equilibrium
are the hydrogen carrier gas and various chlorinated silicon species. Increasing the
relative amount of Cl in the system favors the formation of these gas-phase species.

Figure 4-1 Chlorosilane CVDEquilibrium Calculations

CK-TUT-10112-1112-UG-1 209 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-2 Chlorosilane CVDMole Fractions

4.1.2 PSR Analysis of Steady-state Thermal CVD

4.1.2.1 Project Description

This user tutorial presents a model for the CVD of silicon nitride in a steady-state
PSR, using the chemistry set described in Section 4.4.1. The process operates at a
low pressure of 1.8 Torr, and a high temperature of 1440 C, which makes it
reasonable to approximate this system as a PSR. In the PSR model, the CVD reactor
is described using a volume, surface area, and gas flow rate. This is a fixed-
temperature simulation that uses continuations to see the effects of changing the
SiF4/NH3 ratio in the input gas. It also uses sensitivity and rate-of-production (ROP)
options to analyze the chemistry occurring in the system.

4.1.2.2 Project Setup

The project file is called psr__cvd.ckprj. The data files used for this sample are
located in the samples2010\psr\cvd directory. This reactor diagram contains only
one inlet, one Perfectly Stirred Reactor, and an output Product icon.

The inlet flow rate (11300 sccm, standard cubic centimeters per minute) is input on
the Stream Property Data tab of the Inlet Stream panel. The inlet gas composition is
input on the Species-specific Property tab of the Inlet Stream panel.

2011 Reaction Design 210 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

The constant reactor temperature (1440 C), pressure (2.368E-3 atm), volume, and
internal surface area are input on the Reactor Physical Property tab of the (C1_) PSR
panel. In this sample problem, no estimated gas mole fractions are input on the Initial
Gas Fraction sub-tab of the Species-specific tab of the (C1_) PSR panel. This is not a
restart problem, so the absence of a solution estimate for the gas-phase species
causes the equilibrium composition to be calculated at 1713 K and used as a starting
point for the iterations. However, solution estimates for the surface site concentrations
and bulk phase activities are input on the corresponding sub-tabs of the Species-
specific tab of the (C1_) PSR panel. The Bulk-phase-specific Data tab of the
(C1_) PSR panel does not have input values because this is a deposition rather than
an etching system.

The Output Control tab of the Output Control panel specifies that a threshold of 0.001
should be used for filtering A-factor sensitivities, while a threshold of 0.01 should be
used for ROP. These are the default values in determining what gets printed in the
output file and saved in the XML Solution File. The check box for indicating that
growth rate sensitivities should be calculated and saved is on this tab, as are similar
boxes to choose that sensitivities or ROP for all species should be calculated and
stored. These latter options should be used with care for large mechanisms, as they
can result in very large output and solution files. The Species Sensitivity and ROP tab
allows the user to specify that sensitivities should be calculated and saved for only
particular species, in this case HF, SiF4 and NH3. Also on this table are choices for
printing ROP information for particular species in the output file. The Continuations
panel has input specifying four more simulations to be done with increasing SiF4/NH3
ratios.

CK-TUT-10112-1112-UG-1 211 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

4.1.2.3 Project Results

Figure 4-3 shows that the predicted deposition rate increases with increasing
SiF4/NH3 ratio, as expected under these conditions of excess ammonia. Although this
particular plot shows the growth rate of Si(D), choosing to plot the growth rate of N(D),
BULK1 or BULK2 would give essentially the same result.

Figure 4-3 Steady-state Thermal CVDDeposition Rate vs. SiF4 Mole Fraction

In studying a CVD system, it is often useful to know the relative importance of gas-
phase and surface reactions. For this system, the model shows that very little gas-
phase decomposition occurs under these conditions. Figure 4-4 shows the total rates
of production (ROPs) for SiF4 and NH3 due to gas-phase and surface reactions. In
these cases, the negative numbers indicate that these are really loss rates rather than
production rates. However, these results clearly show that surface reactions (rather
than gas reactions) dominate the decomposition of these starting materials.
Sensitivity analysis gives complementary information to the ROP analysis. Figure 4-5
shows the silicon nitride growth rate is most sensitive to surface reaction #2, which is
the reaction of SiF4 at the surface. Figure 4-6 shows that the NH_NH2(S) site fraction
is much larger than the site fractions of the other surface species.

Include variables from other plot sets by checking the box on the Post-Processor Control
Panel:

2011 Reaction Design 212 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

Figure 4-4 Steady-state Thermal CVDROP Comparison

Figure 4-5 Steady-state Thermal CVDGrowth Rates

CK-TUT-10112-1112-UG-1 213 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-6 Steady-state Thermal CVDSite Fractions

4.1.3 Approximations for a Cylindrical Channel Flow

4.1.3.1 Project Description

This user tutorial explores three ways of modeling steady-state CVD in a cylindrical
flow reactor, and compares the results of the levels of approximation. This project
uses the silicon nitride CVD chemistry set described in Section 4.4.1. The three
approximations used are, in order of increasing simplification, Cylindrical Shear-layer
Flow Reactor, Plug Flow Reactor, and a series of Perfectly Stirred Reactors.

4.1.3.2 Project Setup

This project file is an example of a project that contains multiple sub-projects. It is
called multiple_models__channel_flow_approximations.ckprj. The data files
used for this sample are located in the
samples2010\multiple_models\channel_flow_approximations directory. The first
reactor diagram contains one inlet and a network of 10 sequential PSRs. The other
two reactor diagrams in this project are simple ones that contain only one inlet and
either a Plug Flow or a Cylindrical Shear-Flow Reactor Model.

4.1.3.2.1 Cylindrical Shear-Flow Reactor

The properties of the inlet gas are described on the R1_IN1 panel. The inlet gas
temperature and mass flow rate are input on the Stream Property Data tab. The mass
flow rate is the maximum gas-phase velocity at the inlet. For this problem, which is in
cylindrical coordinates, the average velocity equals one half of the maximum velocity
of the assumed parabolic velocity profile. The composition of the inlet gas is input on
the Species-specific Property tab of the R1_IN1 panel. This example uses mole
fraction, but the user may choose to input these values in mass fractions instead.

2011 Reaction Design 214 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

Parameters describing the reactor geometry and wall temperature are entered on the
Reactor Physical Property tab of the C1_ Cylindrical Shear Flow panel. The
temperatures for the wall are input on this tab. In this case, the inlet gas temperature
is equal to the wall temperature, so there is no need to provide transitioning
parameters. The pressure, grid parameters, as well as the use of multicomponent
diffusion and thermal diffusion (the Soret effect) are specified on this tab. To be
consistent in our approximation to the Plug Flow as well as to the multiple PSR
examples, the Boundary Layer Thickness has been set to 0.01 cm, to provide an
initially flat velocity profile. The Species-specific Data tab allows the specification of
initial guesses for the gas composition adjacent to the surface, which is not used in
this example, as well as estimated values for the surface site fractions and bulk
activities, which are provided. A good initial guess for these values is very helpful in
attaining convergence.

No values are entered on the Solver panel; the default tolerance values are used. The
Output Control panel has been used to specify that the text output file should have
solution data printed every 1 cm.

4.1.3.2.2 Plug Flow Reactor

The properties of the inlet gas are described on the R1_IN1 panel. The inlet mass
flow rate is input on the Stream Property Data tab. The composition of the inlet gas is
input on the Species-specific Property tab of the R1_IN1 panel. This example uses
mole fractions, but the user may choose to input these values in mass fractions
instead.

Parameters describing the reactor geometry and gas temperature are entered on the
Reactor Physical Property tab of the C1_ PFR panel. The problem type and pressure
are also specified on this tab. The Species-specific Data tab allows the specification of
initial guesses for the surface site fractions and bulk activities. No values are entered
on the Solver panel, except to specify that the text output file should have solution
data printed every 1 cm.

4.1.3.2.3 Perfectly-stirred Reactor Network

As for the other sub-projects, the properties of the inlet gas for the series PSR project
are described on the R1_IN1 panel. The mass flow rate is input on the Stream
Property Data tab. The composition of the inlet gas is input on the Species-specific
Property tab of the R1_IN1 panel.

This sub-project contains a series of 10 identical PSR reactors, so in this case, the
C1_Rx PSR (where x = 1 - 10) panels have no entries. Instead, the properties of
these reactors are input on the C1 Cluster Properties panel. Cluster properties will
apply to all PSRs in the cluster unless overridden in the individual reactor panels.

CK-TUT-10112-1112-UG-1 215 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Parameters describing the problem type, reactor geometry, pressure, and gas
temperature are entered on the Property for All Reactors tab of the Cluster Properties
panel. The Species-specific Data for All Reactors tab allows the specification of initial
guesses for the gas composition, which are not used in this example, as well as
estimated values for the surface site fractions and bulk activities, which are used. A
good initial guess for these values is very helpful in attaining convergence. There are
no entries on the Solver, Output Control, or Continuations panels; all the default
values are used.

4.1.3.3 Project Results

Figure 4-7 shows a contour plot of the SiF4 mole fractions as a function of radial and
axial position in the shear flow simulation. There is a minor degree of radial non-
uniformity for this reactant species, as a result of chemical reactions consuming SiF4
at the wall. Figure 4-8 shows SiF4 mole fraction as a function of axial distance for all
three simulations. The mole fraction for the centerline is higher than that for the
upper-wall in the shear-flow simulation, as expected, and these results bracket the
results from the plug-flow and series-PSR simulations. Although not shown, the area-
averaged value for the SiF4 mole fraction from the shear-flow simulation agrees well
with the results from the other two simulations.

Figure 4-7 Cylindrical Channel FlowShear-flow Simulation of SiF4 Mole Fractions

2011 Reaction Design 216 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

Figure 4-8 Cylindrical Channel FlowSiF4 Mole Fractions Comparison

Figure 4-9 shows the silicon nitride deposition rates from these three simulations. The
predictions from the plug-flow and series-PSR simulations are nearly identical, with
the shear-flow simulations results being somewhat lower. Although there is some
differences in the axial gas velocities shown in Figure 4-10 for the shear-flow and plug
flow simulations, the difference in deposition rates mostly results from the lower
reactant concentrations at the surface in the higher-dimensional simulations. The
series of 10 PSR reactors and the Plug Flow simulations represent comparable levels
of approximation to the channel flow. As the Plug Flow simulation actually runs faster,
it would be the recommended reactor model to use in developing chemical
mechanisms, or for exploring general trends for these conditions. However, the
series-PSR approach has the advantage that it can approximate transient flow, as
well as steady-state flow, in a channel. The deposition rate is affected by mass-
transfer in this case, so the higher-dimensionality shear-flow simulations would be
used for final mechanism adjustment and reactor design simulations.

CK-TUT-10112-1112-UG-1 217 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-9 Cylindrical Channel FlowDeposition Rates Comparison

Figure 4-10 Cylindrical Channel FlowAxial Gas Velocities Comparison

2011 Reaction Design 218 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

4.1.4 Deposition in a Rotating Disk Reactor

4.1.4.1 Project Description

This user tutorial presents a model for the CVD of silicon from silane in a steady-state
rotating disk reactor using the chemistry set described in Section 4.4.2. This is a
fixed-surface temperature simulation that represents the experimental rotating-disk
reactor used by Ho, Coltrin and Breiland,47 (p. 247) with the conditions corresponding
to Figure 4 in that paper. The inlet gas is a dilute mixture of silane in helium. The use
of a helium carrier gas, rather than hydrogen, favors the gas-phase decomposition
reactions of silane. For this case, a surface temperature of 923 K (650 C) is used.
This example has two continuations, where the silane partial pressure is increased.

4.1.4.2 Project Setup

The project file is called rotating_disk__sih4_cvd.ckprj. The data files used for this
sample are located in the samples2010\rotating_disk\sih4_cvd directory. This
reactor diagram contains only one inlet and one rotating-disk CVD reactor.

The properties of the inlet gas are described on the C1_Inlet panel. The inlet gas
temperature is input on the Stream Property Data tab. An inlet gas velocity is usually
not entered for a steady-state rotating disk simulation, as it is calculated from the spin
rate, pressure, and gas-properties, but the user may override this value on the
Reactor Physical Property tab, Basic sub-tab, if desired. The composition of the inlet
gas is input on the Species-specific Property tab of the C1_Inlet panel. Note that the
mole fractions do not add up to one, as the partial pressures (in Torr) have actually
been entered. This is permitted, as the program will normalize the gas composition
internally if the user does not do so in the CHEMKIN Interface. The Continuations
panel is used to input the new compositions of the reactant gas mixture for the second
and third simulations in this sample project.

Parameters describing the reactor conditions are entered on the Reactor Physical
Property tab, Basic sub-tab of the C1_Rotating Disk panel. The choice of a steady-
state simulation solving the gas energy equation is entered here, as well as the use of
multicomponent diffusion and thermal diffusion (the Soret effect). The temperature for
the deposition surface (923 K), pressure (200 Torr), and disk rotation rate (450 rpm)
are input on this tab. The other sub-tabs on the Reactor Physical Property tab are for
options that are not used in this example. The Initial Grid Property tab of the Reactor
Physical Property tab allows specification of the locations of the deposition surface
(x = 0, default) and the end axial location where the gas enters (6.2 cm). Grid
parameters are specified on this tab. This reactor model includes adaptive gridding,
and the use of a relatively-low value for the initial number of grid points is
recommended. The Species-specific Data tab allows the specification of initial

CK-TUT-10112-1112-UG-1 219 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

guesses for the gas composition at the inlet, adjacent to the surface, or maximum
mole fractions for intermediate species, none of which are used in this example.
However, estimated values for the surface site fractions and bulk activity are provided.
A good initial guess for these values can be very helpful in attaining convergence.

4.1.4.3 Project Results

Figure 4-11 shows the axial, radial and circumferential velocity components for the
rotating disk reactor as a function of height above the surface. The deposition surface
is at the origin, and the gas enters at x = 6.2 cm with an axial velocity of -12.4 cm/sec
and zero radial and circumferential velocity components. As the gas approaches the
rotating surface, the axial velocity initially increases slightly then decreases while the
radial and circumferential velocity components increase. At the surface, the axial and
radial velocity components are zero, and the circumferential velocity matches that of
the disk (in radians/sec), as expected.

Figure 4-11 Deposition in a Rotating DiskGas Velocity Components

Figure 4-12 shows predicted mole fractions for the various silicon hydrogen species
as a function of distance above the surface (the helium carrier gas is not included),
again for the first simulation in the project. The composition at the grid point with
largest x value is constrained to that of the inlet gas (silane and helium only). The
other grid points show varying amounts of product and reactive intermediate species
that are formed by gas-phase and surface reactions. Si atoms are present in very low
amounts (mole fractions ~10-12), but can easily be detected by laser-induced
fluorescence techniques, so they are kept in the mechanism. You can choose
concentration units for species composition in the Species/Variables tab of the Select
Results to Load panel when the Post-Processor is first launched. This yields the
results shown in Figure 4-13, illustrating that Si atom concentrations increase with
increasing silane concentration, as expected. The profiles of the Si atom

2011 Reaction Design 220 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

concentrations are shown as a function of distance above the surface for different
starting silane partial pressures: #1 = 0.11 Torr, #2 = 0.34 Torr, #3 = 0.67 Torr. The
curves in this figure suggest that the profiles might also be changing shape. This is
confirmed in Figure 4-14, which was made by exporting the simulation results,
normalizing and plotting experimental results from Ho, Coltrin and Breiland (see
Figure 4 of referenced paper)47 (p. 247) in third party software. It shows that: a) Si atom
profiles are experimentally observed to be narrower for higher silane concentrations,
and b) this reaction mechanism reproduces this observation. Comparisons between
the model and this experimental data set shown that the original reaction proposed for
Si atom formation, the collisionally-induced decomposition of SiH2, could not account
for the experimental observations. Two other reactions, H3SiSiH Si + SiH4 and
Si + Si2H6 H3SiSiH + SiH2, are instead the primary reactions involving Si atoms.

Figure 4-12 Deposition in a Rotating DiskMole Fractions

CK-TUT-10112-1112-UG-1 221 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-13 Deposition in a Rotating DiskSi Atom Concentrations

Figure 4-14 Deposition in a Rotating DiskExperimental Data

2011 Reaction Design 222 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

4.1.5 Trichlorosilane CVD in Planar Channel Flow Reactor

4.1.5.1 Project Description

This user tutorial presents a model for the CVD of silicon in a steady-state planar
shear-layer flow reactor using the chemistry set described in Section 4.4.3. The
process operates at atmospheric pressure, and a relatively high temperature
(1398 K). This is a fixed-temperature simulation that represents a horizontal cross-
flow reactor of the type used to deposit epitaxial silicon layers. In this case, the upper
wall temperature is held at a temperature (773 K), which is significantly lower than the
deposition substrate, but higher than the inlet gas temperature (623 K).

4.1.5.2 Project Setup

The project file is called planar_shear_flow__tcs_cvd.ckprj. The data files used for
this sample are located in the samples2010\planar_shear_flow\tcs_cvd directory.
The reactor diagram contains one inlet, one planar shear-flow reactor, and an outlet.

The properties of the inlet gas are described on the C1_Inlet panel. The inlet gas
temperature and inlet gas velocity are input on the Stream Property Data tab. The
Axial Velocity should be the maximum gas-phase velocity at the inlet. For this
problem. which is in cartesian coordinates, the average velocity equals two-thirds of
the maximum velocity of the parabolic velocity profile. The composition of the inlet gas
is input on the Species-specific Property tab of the C1_Inlet panel.

Parameters describing the reactor geometry and wall temperatures are entered on
the Reactor Physical Property tab of the C1_ Planar Shear Flow panel. The
temperatures for the upper wall and deposition surface (lower wall) are input on this
tab, as well as an optional parameter specifying the distance over which the wall
temperatures are smoothly transitioned from the inlet gas temperature to the desired
wall temperature. The pressure, grid parameters, as well as the use of
multicomponent diffusion and thermal diffusion (the Soret effect) are specified on this
tab. The Species-specific Data tab allows the specification of initial guesses for the
gas composition adjacent to the surface, which is not used in this example, as well as
estimated values for the surface site fractions and bulk activities, which are provided.
A good initial guess for these values is very helpful in attaining convergence.

On the Basic tab of the Solver panel, the text output file has been specified to have
solution data printed every 1 cm along the channel.

CK-TUT-10112-1112-UG-1 223 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

4.1.5.3 Project Results

The gas temperatures in Figure 4-15 show that the gas heats up substantially near
the deposition surface, with the hot zone expanding with axial distance, as expected.
The inlet gas temperature is lower than the upper wall temperature, such that the
coolest gas lies in a region near to, but below the top wall. This cooler gas region is
reflected in the trichlorosilane mole fraction contours shown in Figure 4-16. Depletion
as a result of chemical reaction causes the low SiCl3H mole fractions near the lower
wall, but thermal diffusion causes the heavier gas to move away from the upper wall.
A simulation run without thermal diffusion gives uniform SiCl3H mole fractions in the
upper part of the reactor. Figure 4-17 shows how the deposition rate of solid silicon
varies as a function of axial distance. Initially, the deposition rate is quite high, but it
drops rapidly as the SiCl3H near the surface is depleted, indicating that the deposition
process is transport limited in this system. Figure 4-18 shows gas-phase mole
fractions near the lower surface as a function of axial distance. The extent of SiCl3H
depletion and HCl formation is notably larger than the formation of the other silicon-
chlorine species (SiCl2, SiCl2H2 and SiCl4), which indicates that deposition of solid
silicon is the dominant reaction pathway.

Figure 4-15 Trichlorosilane CVDGas Temperatures vs. Axial and Radial Distance

2011 Reaction Design 224 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

Figure 4-16 Trichlorosilane CVDTrichlorosilane Mole Fraction

Figure 4-17 Trichlorosilane CVDSilicon Deposition Rate

CK-TUT-10112-1112-UG-1 225 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-18 Trichlorosilane CVDMole Fractions

4.2 Atomic Layer Deposition (ALD)

The CHEMKIN software includes several reactor models that can be used for
simulating ALD processes. ALD is a technique used to deposit thin films of solid
materials in a very controlled manner, and differs from CVD primarily in that it is a
transient process with the deposition surface being exposed to pulses of alternating
gases. Ideally, the deposition chemistry in ALD is self-limiting, with growth occurring in
a layer-by-layer manner and the deposition thickness being controlled only by the
number of cycles. ALD is also called: ALE, Atomic Layer Epitaxy, NLD, Nano Layer
Deposition, ALCVD, Atomic Layer Chemical Vapor Deposition, and AVD, Atomic
Vapor Deposition. There are also plasma enhanced variations such as PENLD,
Plasma Enhanced Nano Layer Deposition. It is a relatively new technology, having
made the transition from the research lab to production in the last decade.

The transient models within the CHEMKIN software are useful for optimizing pulse
sequences and thus minimizing cycle times in ALD. The major advantage of ALD over
CVD is the improved control over the deposition process and more conformal
deposition. The inherently lower deposition rates, however, lead to longer process
times and higher costs. During a pulse, it is important that enough molecules react
with all parts of the substrate to be coated. But many of the precursor materials are
expensive. Thus one of the process optimization goals is to reduce the amount of
precursor that flows through the reactor but does not react at the surface. Considering
the effects of finite-rate kinetics for surface reactions can be an important part of such
an optimization, as reactions do not always behave in an ideal manner.

2011 Reaction Design 226 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

4.2.1 Time-dependent Simulations of ALD Process

4.2.1.1 Project Description

This user tutorial demonstrates two ways of simulating a time-dependent ALD
process. The transient Perfectly Stirred Reactor Model runs relatively quickly and is
useful for developing and initial testing of the chemical reaction mechanism. The
transient stagnation flow model more realistically simulates a production-scale
shower-head ALD reactor by including mass-transport effects. The alumina ALD
chemistry described in Section 4.4.4 is used. The process operates at a pressure of
1 Torr, and a relatively low surface temperature (compared to CVD) of 450 C (723 K).
This sample demonstrates four cycles of the flow sequence including: metalorganic
precursor in argon; argon purge; ozone in oxygen and argon; and argon purge. The
four cycles are sufficient, in this case, to characterize the process.

4.2.1.2 Project Setup

This project file is an example that contains multiple reactor models or sub-projects. It
is called multiple_models__atomic_layer_deposition.ckprj. The data files used for
this sample are located in the
samples2010\multiple_models\atomic_layer_deposition directory. The two
reactor diagrams in this project each contain three gas inlets and either a Perfectly
Stirred Reactor or a Stagnation-flow CVD Reactor Model.

In this project, the names of the gas inlets, and their corresponding input panels, have
been changed away from the default names of R1_IN1, etc., to METORG, OXIDIZER
and PURGE, which reflect the function of the gas flowing through the inlets. The
listings of flow rate as a function of time for each gas-flow inlet are saved as time-
dependent profile files:

SCCMPRO_METORG.ckprf
SCCMPRO_OXIDIZER.ckprf
SCCMPRO_PURGE.ckprf

In this case, the profiles of total flow rates for each gas inlet are in volumetric flow
units of sccm, but flow profiles can also be specified in mass flow units. The PSR and
stagnation-flow models use the same flow-profile files.

Although both simulations run for the same length of process time, the stagnation
flow simulation takes significantly more compute time to run, due to the inclusion of
mass-transport effects.

CK-TUT-10112-1112-UG-1 227 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

4.2.1.2.1 Transient PSR

The names of the flow-profile files are designated on the Stream Property Data tab of
each Gas Inlet panel. If a profile file already exists, it can be selected by browsing or
by the pull-down menu, if it has already been accessed. The Edit button opens a
panel that allows profile files to be edited or created. In this case, inlet temperatures
are not specified on the Gas Inlet panel, as this is a fixed-gas temperature simulation.
The composition of each reactant gas inlet is specified on the Species-specific
Property panel of the corresponding Gas Inlet panel.

The choice of fixed gas temperature for this simulation is made with the Problem-type
drop-down list on the Reactor Physical Property tab of the C1_ PSR panel. The gas
temperature, pressure, volume and surface area (which are generally representative
of ALD reactors) are also input here. In this case, the gas temperature for the PSR
simulation was chosen to reproduce the degree of ozone decomposition to O atoms
and match the deposition rate observed in the higher dimensional simulation. The
surface temperature is different from the gas temperature, and it is also specified on
the Reactor Physical Properties tab. On the Species-specific Data tab, the starting
gas composition of pure argon is specified on the Initial Gas Fraction sub-tab, the
starting surface composition of complete O(S) coverage is specified on the Surface
Fraction sub-tab, and an activity of 1.0 specified for AL2O3(B) on the Bulk Activity
sub-tab. These correspond to a reasonable starting condition where the substrate
might have an initial oxide coating, and the system was purged with argon after
loading. No entries are made on the other parts of this panel.

The end time of the transient simulation is specified on the Reactor Physical
Properties panel. The solver maximum and minimum step time is on the Advanced
tab of the Solver panel. Intervals for saving and for printing solution data are given on
the Output Control panel. No entries are required on the SOlver, Output Control or
Continuations panels for this problem.

4.2.1.2.2 Transient Stagnation Flow Reactor

The names of the flow-profile files are designated on the Stream-property panel of
each Gas Inlet panel. In this case, the gas energy equation is being solved, so an inlet
temperature of 150 C (423 K) is a required input for each gas inlet. The composition
of each reactant gas inlet is specified on the Species-specific Property tab of the
corresponding Gas Inlet panel.

Choices to solve a transient problem, include the gas energy equation, use
multicomponent diffusion but not the Soret effect, etc., are input in the Reactor
Physical Property tab, Basic sub-tab of the C1_ Stagnation Flow panel. The process
pressure and surface temperature are input on this sub-tab, along with the name of

2011 Reaction Design 228 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

file containing the initial guess for the gas temperatures as a function of height above
the disk, and a cross sectional flow area used for translating volumetric flow rates to
linear flow velocities. The end time of the transient simulation is also specified on the
Basic sub-tab. There are no entries on the other sub-tabs of the Reactor Physical
Property tab.

The Initial Grid Property tab of the C1_ Stagnation Flow panel allows the input of the
number of points in the grid of distance above the substrate, along with the ending
axial distance. In transient simulations, the grid currently does not adapt as it does in
steady-state stagnation flow simulations. Thus, a reasonably dense initial grid should
be specified, and the adaptive gridding parameters are ignored. Note that the surface
is defined as being at a coordinate value of x = 0, and the maximum distance of
1.2 cm is the location of the gas inlets (showerhead). In other words, the solution is
given as a function of distance from the surface.

On the Species-specific Data tab of the C1_ Stagnation Flow panel, the starting gas
composition of pure argon is specified on the Initial Gas Fraction sub-tab, the starting
surface composition of complete O(S) coverage is specified on the Surface Fraction
sub-tab and an activity of 1.0 specified for AL2O3(B) on the Bulk Activity sub-tab.
These are the same as were used in the PSR simulation, and correspond to a
reasonable starting condition where the substrate might have an initial oxide coating,
and the system was purged with argon after loading. No entries are made on the
other parts of this panel.

The interval for printing data to the output file (on the Output Control panel) and some
tolerance parameters that have been relaxed from the default values (on the Basic
tab of the Solver panel). The Advanced tab of the Solver panel contains a time-step
specification and a solver parameter that have been altered from the default values.
No entries are made on the Continuations panel for this problem.

4.2.1.3 Project Results

Figure 4-19 shows the gas pulses used in this example for the metal-organic and
oxidizer gas inlets. The pure-argon purge gas pulses are not shown. The metal-
organic gas (trimethyl aluminum; TMA) pulses are considerably shorter than the
oxidizer gas pulses, but appear to be sufficient. The contour plot of one TMA pulse
from the stagnation-flow simulation in Figure 4-20 clearly shows that the TMA is
consumed at the surface only at the beginning of the pulse. By the end of the pulse,
all the TMA flowing into the reactor is also flowing out again.

CK-TUT-10112-1112-UG-1 229 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-19 Time-dependent ALD SimulationsPSR, Total Flow Rates vs. Time

Figure 4-20 Time-dependent ALD SimulationsTMA Contour Plot

The chemistry that occurs during the oxidizer pulses is more complex. The contour
plot of O atom mole fractions in Figure 4-21 show how the O atoms are formed by
gas-phase decomposition of ozone in the hotter regions of the gas, and then react
away at the surface. The site fractions in Figure 4-22 show that the methylated
surface species, ALME2(S) and ALMEOALME(S), are not completely converted to
O(S) during the oxidizer pulse. This probably results from the fact that the oxidation

2011 Reaction Design 230 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

occurs as two sequential steps and the kinetics are limiting the process. The O atom
mole fractions in Figure 4-23 shows that the O atoms are not being depleted at the
surface in the stagnation flow simulations, which suggests a kinetic limitation, possibly
resulting from the default reaction orders in this simple mechanism. The incomplete
oxidation of the methylated surface species in turn leads to less than unity O(S)
coverage at the beginning of the TMA pulse, which in turn leads to less efficient use of
TMA and some notable differences between the first and subsequent pulses.

Figure 4-21 Time-dependent ALD SimulationsO Mole Fraction

Figure 4-22 Time-dependent ALD SimulationsStagnation-flow Site Fractions

CK-TUT-10112-1112-UG-1 231 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-23 Time-dependent ALD SimulationsO Mole Fractions Comparison

The O atom mole fractions in Figure 4-23 from the stagnation flow simulations agree
well with those from the PSR simulations. Within the limits of this simple deposition
chemistry, this directly leads to the good agreement for deposition thickness shown in
Figure 4-24. The agreement between the two models, in this case, results from the
fitting of the gas-phase temperature in the PSR simulation to reproduce the O atom
mole fractions in the higher-dimensional simulation. This kind of a calibration allows
the faster-running PSR simulation to be used to quickly explore a wide range of pulse
sequences and other experimental conditions of interest, before returning to higher-
dimensional simulations for more careful study.

Figure 4-24 Time-dependent ALD SimulationsDeposition Thickness Comparison

2011 Reaction Design 232 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

4.3 Plasma Etching

The CHEMKIN software includes a number of special features for modeling plasmas,
most of which are demonstrated in these examples. For treating non-equilibrium
plasmas, the plasma models allow you to specify different temperatures for the
neutral gas, electrons, ions, and surfaces. A gas-phase reaction rate can be
designated to depend on the electron temperature rather than the default neutral gas
temperature. An electron-impact reaction can also be designated to involve a
specified energy loss per collision. This option is used to describe excitation energy
losses for electrons in solving the electron energy equation, without explicitly defining
each vibrationally or electronically excited-state as a separate species within the
corresponding thermodynamic data. For surface reactions, ion-assisted reactions can
have yields that depend on the ion energy. Such reactions can also be designated as
Bohm reactions, which means that the ion flux is limited by the Bohm velocity of the
ions, rather than by the ion's thermal speed. The implementation of the Bohm criterion
is described in detail in Section 8.5.3 of the CHEMKIN Theory Manual. This factor has
been introduced to account for the fact that the ion interaction with the surface will be
subject to transport limitations, such that a gradient near the walls in the ion density
will occur. A zero-dimensional model cannot capture this effect, such that a Bohm-flux
correction model is used.

4.3.1 Steady-state Chlorine Plasma

4.3.1.1 Project Description

This user tutorial provides an example of modeling a low-pressure plasma reactor as
a steady-state Perfectly Stirred Reactor, using the chlorine plasma chemistry set
described in Figure 4.4.5. The plasma pressure is 5 mtorr and pure chlorine gas flows
into the reactor at 35 sccm. The energy equation and the electron energy equation
are solved, and a heat-transfer correlation is used to account for heat loss through the
wall of the reactor.

4.3.1.2 Project Setup

The project file is called plasma_psr__chlorine.ckprj. The data files used for this
sample are located in the samples2010\plasma_psr__chlorine directory. This
reactor diagram contains one gas inlet, one plasma PSR, and one outlet.

CK-TUT-10112-1112-UG-1 233 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

The reactant gas mixture, which is pure Cl2 in this case, is input on the Species-
specific Property tab of the C1_Inlet1 panel. The initial guess for the steady-state gas
composition, which is input on the Initial Gas Fraction subtab of the Species-specific
Data tab of the C1_ Plasma PSR panel, is quite important. A good initial guess (one
that is close to the steady-state solution) will result in fast convergence, while a poor
initial guess can at worst lead to failure of the simulation.

The Reactor Physical Property tab of the C1_ Plasma PSR panel is where problem
type and reactor parameters such as pressure, temperatures, volume, area, heat
loss, and plasma power are entered. Solving the electron energy equation requires
that an initial guess for the electron temperature be input on this tab, as well as an
inlet electron temperature on the Stream Property tab of the C1_Inlet1 panel. The
latter has no impact on the solution unless there are electrons in the inlet gas mixture.
The Sheath Loss parameter describes the ion energy gained crossing the sheath and
may be specified differently for each material in the system. On the Solver panel,
skipping the intermediate fixed-temperature solution is usually more robust for plasma
simulations than trying to solve it first.

4.3.1.3 Project Results

Figure 4-25 shows the electron temperature as a function of power for this chlorine
plasma. The electron temperature shows only small changes over the large variation
in plasma power, since most of the power is transferred into ionization and
dissociation rather than electron heating. However, as the plasma power drops, the
electron temperature rises as the plasma nears extinction, since there are fewer
electrons and electron-driven events. The mole fractions of the electrons and positive
ions, increase steadily with increasing plasma power, as shown in Figure 4-26, but the
chlorine negative ions show a maximum at intermediate powers. Figure 4-27 shows
that the net dissociation of molecular chlorine to atomic chlorine steadily increases
with increasing plasma power, as expected.

2011 Reaction Design 234 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

Figure 4-25 Steady-state Chlorine PlasmaElectron Temperatures vs. Power

Figure 4-26 Steady-state Chlorine PlasmaMole Fractions

CK-TUT-10112-1112-UG-1 235 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-27 Steady-state Chlorine PlasmaCl and Cl2 Mole Fractions

Figure 4-28 shows a rate-of-production analysis for the input gas Cl2. (To create this
plot with plasma power deposition as the x-axis, select Include variables from other
plot sets in the CHEMKIN Visualizer. For details, see the Control Tabs section in
Chapter 3 of the CHEMKIN Visualization Manual.) In this case, there are clear major
and minor pathways. The primary reaction consuming molecular chlorine is gas-
phase reaction #3, which is the electron-impact dissociation to two Cl atoms. Gas
reactions #5 and 2, the electron-impact ionization and dissociative attachment
reactions of Cl2, respectively, are minor channels for Cl2 consumption. The primary
reaction producing Cl2 is surface reaction #4, the reaction between gas-phase and
adsorbed Cl atoms producing molecular chlorine in the gas phase. Gas reaction #18,
the neutralization reaction between Cl- and Cl2+, and surface reaction #2, the
neutralization of Cl2+ on the wall, are minor channels for Cl2 production.

2011 Reaction Design 236 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

Figure 4-28 Steady-state Chlorine PlasmaROP Analysis

4.3.2 Spatial Chlorine Plasma PFR with Power Profile

4.3.2.1 Project Description

This user tutorial presents an example of a Plasma Plug Flow Reactor with a specified
power versus distance profile. This feature might be used to model an experimental
apparatus where the inductive coils providing the power are spaced unevenly.
Alternatively, this could be used to model a system where there is a lower-power
region where a probe is used that cannot operate in a high-plasma density region, or
where a measurement window is located.

This example is for a plasma at a total pressure of one torr with pure chlorine gas
flowing into the reactor at 35 cm3/s, using the chlorine plasma chemistry set
described in Section 4.4.5. The electron energy equation is solved, and there is
consideration of heat loss through the wall of the reactor. The area used for surface
chemistry and heat loss corresponds to the physical walls of the tubular reactor.

4.3.2.2 Project Setup

The project file is called plasma_pfr__cl_power_profile.ckprj. The data files used
for this sample are located in the samples2010\plasma_pfr\cl_power_profile directory.
This reactor diagram contains a gas inlet, a Plasma Plug Flow Reactor, and an outlet.

CK-TUT-10112-1112-UG-1 237 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

The reactant gas mixture is input on the Species-specific Property tab of the
C1_Inlet1 panel. Although the input gas is actually pure Cl2, in order to have the
plasma light in the simulation, we have modified the inlet gas to include a small
amount of electrons and ions in the inlet gas.

The Reactor Physical Property tab of the C1_ Plasma PFR panel is where problem
type and reactor parameters such as pressure, temperatures, geometry, heat loss,
and plasma power are entered. The value for the electron temperature input on the
Reactor Physical Property tab of the C1_ Plasma PFR panel is used as the inlet
electron temperature. The plasma power profile can be entered by using the Profile
tool or by selecting an existing profile file on the Reactor Physical Property tab of the
C1_ panel. This is an alternative to entering a constant value for the Plasma Power
Deposition parameter. The profile is entered as a list of pairs of numbers giving the
power as a function of position. Values at intermediate points are straight-line
interpolations of the two nearest specified values. Unphysically abrupt changes in the
power profile can cause convergence problems in the simulation unless solver time
steps are set accordingly. Note that the power deposition for a plug-flow reactor profile
must be specified in per-distance units, as it will be integrated over the channel
distance.

As a simulation that marches forward from the inlet state, this sample problem has no
initial guess for the gas composition, but Surface Fractions are provided on the
Species-specific Data tab of the C1_ Plasma PFR panel. These site fractions provide
initial estimates for the surface state of the channel inlet. An initial pseudo-steady
calculation will be performed to determine consistent surface state based on these
estimates and the inlet gas composition, prior to the channel integration. The Sheath
Loss parameter on the Material-specific Data tab of the C1_ Plasma PFR panel
describes the ion energy gained crossing the sheath.

4.3.2.3 Project Results

Figure 4-29 shows the profile of electrical power deposited into the plasma as a
function of distance down the Plug Flow Reactor. As expected, the mole fractions of
the electrons and atomic chlorine, shown in Figure 4-30 and Figure 4-31, respectively,
generally follow the plasma power profile, with differing degrees of non-linearity. This
is consistent with the fact that the Cl is created in the plasma by electron-impact
dissociation of molecular chlorine. The mole fractions for the Cl+ and Cl2+ ions, shown
in Figure 4-32 also follow the plasma power profile.

2011 Reaction Design 238 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

Figure 4-29 Spatial Chlorine PlasmaPlasma Power vs. Distance

Figure 4-30 Spatial Chlorine PlasmaElectron Mole Fraction

CK-TUT-10112-1112-UG-1 239 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-31 Spatial Chlorine PlasmaCl Mole Fraction

Figure 4-32 Spatial Chlorine PlasmaCl+ and Cl2+ Mole Fractions

The electron temperature rises somewhat in the lower-power regions of the reactor
tube. This is consistent with the inverse trend between electron densities and electron
temperatures seen at low power densities in Section 4.3.1.

2011 Reaction Design 240 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

Figure 4-33 Spatial Chlorine PlasmaElectron Temperature

4.3.3 Fluorocarbon Plasma Etching of Silicon Dioxide

4.3.3.1 Project Description

This user tutorial is an example of modeling a low-pressure plasma reactor as a
steady-state Perfectly Stirred Reactor, using the fluorocarbon plasma chemistry set
described in Section 4.4.6. This example is representative of a high-density plasma,
with a plasma power of 3000 Watts and a pressure of 10 mtorr. Pure
hexafluoroethane is the inlet gas. The energy equation and the electron energy
equation are solved, and a heat-transfer correlation is used to account for heat loss
through the wall of the reactor. This example illustrates the use of multiple materials,
each with a different set of surface reactions, as well as the use of continuations with
varying flow rate and ion impact energy.

4.3.3.2 Project Setup

The project file is called plasma_psr__c2f6_etch.ckprj. The data files used for this
sample are located in the samples2010\plasma_psr__c2f6_etch directory. This
reactor diagram contains a gas inlet and a plasma PSR.

The reactant gas mixture, which is pure C2F6 in this case, is input on the Species-
specific Property tab of the C1_Inlet1 panel. The gas flow rate (30 sccm), inlet gas
temperature and inlet electron temperature are input on the Stream Property tab of
the R1_IN1 panel. The latter is required, but has little impact on the solution unless
there are electrons in the input gas mixture.

CK-TUT-10112-1112-UG-1 241 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

The C1_ Plasma PSR panel is where most of the parameters are input. The Reactor
Physical Property tab allows selection of problem type, as well as places to enter
parameters such as the plasma power, pressure (10 mTorr), volume, internal area,
heat loss parameters, cross-section for electron momentum loss, and initial guesses
for the electron, neutral, and ion temperatures. The Species-specific Data tab allows
input of initial guesses for the steady-state gas composition, surface site fractions,
and bulk activities, on the corresponding sub-tabs, as well as species-specific values
for electron-momentum loss cross-sections. A good initial guess for the gas and
surface compositions (one that is close to the steady-state solution) will result in fast
convergence, while a poor initial guess can lead to failure of the simulation. The Bulk-
phase -specific Data tab is used to indicate that all the bulk materials are being
etched, which affects some of the equations. The Material-specific Properties tab is
for specifying properties like surface temperature (if different from the neutral gas
temperature), ion energy or bias power, Bohm factor, sheath energy loss factor, area
fraction, or heat loss, either for all materials, or on a per-material basis. For this non-
parameter-study problem, the bottom part of the panel is used to set up values that
are dependent on the material. Material-independent properties can be set on the
Reactor Physical Properties panel. In this example, the WAFER is held at a
temperature near room temperature (consistent with active cooling), while the
TOPWALL is held at an elevated temperature (consistent with active heating), while
the SIDEWALL temperature is allowed to float with the gas temperature (the default).
A relatively high ion energy (in electron Volts) is specified for the WAFER, reflecting
that this material substrate has an applied electrical bias, separate from the main
power, for this system. The low ion energy for the SIDEWALL is consistent with the
plasma self-bias, rather than active biasing of this surface. The TOPWALL has no ion
energy specified, as no ion-energy dependent reactions occur on this material.

On the Solver panel, a number of parameters on both the Basic and Advanced tabs
have been altered from the defaults in order to help get a good solution. Skipping the
intermediate fixed-temperature solution is the default for plasma simulations and is
recommended. The Output Control panel allows specification of whether sensitivities
should be calculated and printed in the output and solution files, and whether ROPs
should be printed and saved. In this case, sensitivities for temperature, growth (etch)
rate, F atoms and SiF4 will be calculated. ROPs for 6 species will be included. The
Continuations panel specifies two continuations. First, the ion energy for the wafer is
changed from 200 to 300 V, and the flow rate is increased from 30 to 50 sccm.
Second, the flow rate is returned to the initial lower value, while leaving the ion energy
at the higher value. In this example, the plasma power is entered on both continuation
panels, although it is not changed. This is not necessary, but will not cause any
problems and may serve as a useful reminder.

2011 Reaction Design 242 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

4.3.3.3 Project Results

In Figure 4-34, the SiO2 etch rates for variations of C2F6 flow rate and Wafer ion
energy are: 1) 30 sccm, 200 V. 2) 50 sccm, 300 V. 3) 30 sccm, 300 V. This shows that
oxide etch rates (negative growth rates) on the Wafer are much larger than those on
the Sidewall, in accord with the higher ion energy for that material. Comparing
solutions 1 and 3 shows that increasing the Wafer ion energy increases the etch rate,
as expected. But comparing solutions 2 and 3 shows that increasing the reactant flow
rate also increases the etch rate, indicating that the etching system may be reagent-
supply limited. This is consistent with the results shown in Figure 4-35, which shows
the 10 neutral species with the highest mole fractions. C2F6, the reactant species,
does not appear, as it is mostly decomposed to smaller fragments in the plasma. The
species with the highest mole fraction, F atoms, is a reactant fragment that can
contribute to etching, as are CF and CF3, which are also present in relatively high
concentrations. These species have higher mole fractions in solution 2, which has the
highest flow rate, whereas etch-product species, such as CO and SiF3, have lower
mole fractions at the shorter residence time. Figure 4-36 shows that CF+ is the most
prevalent positive ion, with F+, CF2+ and CF3+ about a factor of 2 lower. There is less
F+ than CF+, even though there is more F than CF, and the major pathways producing
F+ and CF+ are ionization of F and CF, respectively. This reflects a higher rate for the
CF ionization reaction than the F ionization reaction. Although not shown, the electron
is the most prevalent negatively charged species, with CF3- roughly a factor of 6
lower. Negative ions stay in the body of the plasma and do not participate in reactions
at the surface due to the presence of an electronegative sheath.

Figure 4-34 Fluorocarbon Plasma Etching of Silicon DioxideSiO2 Etch Rates Variations

CK-TUT-10112-1112-UG-1 243 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

Figure 4-35 Fluorocarbon Plasma Etching of Silicon Dioxide10 Highest Mole Fractions

Figure 4-36 Fluorocarbon Plasma Etching of Silicon DioxidePositive Ion Mole Fractions

The surface site fractions shown in Figure 4-37 show minor changes with conditions.
Comparing solutions 1 and 3 shows that increasing the Wafer ion energy increases
the fraction of open sites and decreases the fraction of fluorocarbon-covered sites.
This is consistent with the increased ion energy causing an increase in yield for the
ion-enhanced etching reactions. Comparing solutions 3 and 2 shows that increasing
the C2F6 flow rate decreases the fraction of open sites and increases the fraction of
fluorocarbon-covered sites. This is consistent with an increased supply of
fluorocarbon radicals that can react with the open sites. Figure 4-38 shows the 5
reactions with the largest contribution to the loss of silicon dioxide from the wafer. In
order of decreasing importance, surface reactions 13, 14, 11, and 12 are etching of

2011 Reaction Design 244 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

WSIO2_CF2(S) sites assisted by CF+, F+, CF2+, and CF3+, respectively. Reaction 5 is
the F+ ion-assisted etching reaction of the WSIO2_F2(S) site. These five reactions
are responsible for most of the etching, but a number of other reactions contribute to
the total etch rate.

Figure 4-37 Fluorocarbon Plasma Etching of Silicon DioxideSurface Site Fractions

Figure 4-38 Fluorocarbon Plasma Etching of Silicon DioxideHighest ROP

CK-TUT-10112-1112-UG-1 245 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

4.4 Chemistry Sets

In this section, we describe in more detail the chemistry sets employed in the tutorials
described in Chapter 4.

4.4.1 Silicon Nitride CVD from Silicon Tetrafluoride and Ammonia

This chemistry set describes the deposition of silicon nitride (Si3N4) from a mixture of
SiF4 and NH3.45 This mechanism demonstrates one way of describing the deposition
of a compound solid, but it should generally be considered as illustrative only and not
as a source of kinetic data for Si3N4 deposition.

The chem.inp file includes 4 elements, 17 gas-phase species and 33 reactions. The
first 27 of these reactions describe the pyrolysis of ammonia, and are taken from
published models for ammonia oxidation and De-NOx.46 The other 6 reactions
describe SiF4 decomposition and likely cross reactions between Si and N-containing
species. Rate parameters were not available in the literature for these reactions, so
the mechanism uses estimates based on bond strengths and analogies with related
gas-phase reactions.

The surf.inp file defines six surface species in the Si3N4 surface phase and the Si3N4
solid, which is defined in terms of two bulk phases, SI(D) and N(D). The surface
species all occupy more than one site, and have placeholder thermodynamic data.
Six irreversible surface reactions are also defined in this file. The surface reactions
are lumped reactions, meaning that they each represent several elementary steps
that have been combined into one reaction. Rate parameters for these surface
reactions were determined by fitting the model results to experimental deposition-rate
data. The activation energies for the surface reactions are all zero, which reflects the
limitations of the experimental data set used to derive the rate parameters. Although it
is not immediately obvious, the surface species and reactions were designed to
produce a bulk-phase stoichiometry of 3 Si to 4 N. The densities of bulk species has
been defined to force the linear deposition rates of either Si(D) or N(D) match the
experimentally observed deposition rate for Si3N4. These reactions have zero for
activation energies because they were fit to an experimental data set taken at one
temperature. These rates would therefore not be likely to be valid at other
temperatures.

45. M. E. Coltrin, P. Ho, Sandia National Laboratories, private communication.

46. J. A. Miller, M. D. Smooke, R. M. Green, and R. J. Kee, Comb. Sci. Technol. 34:149
(1983).

2011 Reaction Design 246 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

4.4.2 Silicon Deposition from Silane

This chemistry set described the CVD of silicon from silane. This CVD system has
been the subject of numerous studies, facilitated by the simplicity of the precursor
molecule. This reaction mechanism was published by Coltrin and coworkers at
Sandia National Laboratories in the last47 of a series of papers. This mechanism has
been validated with a variety of optical diagnostic measurements and deposition rate
data in a variety of experimental reactor geometries.

The chem.inp file includes 3 elements (helium is used as a carrier gas), 9 gas-phase
species, and 10 reversible gas-phase reactions. Many of the reactions are
unimolecular decomposition reactions with explicit treatment of pressure dependent
rate parameters. The most important reaction is the decomposition of SiH4 to SiH2
and H2. The other reactions include the reaction of SiH2 with SiH4 to form Si2H6,
which can then react again to form Si3H8, or decompose to H3SiSiH and H2. Other
reactions involve interconversion between various silicon-hydride species, plus the
formation of atomic silicon, which was one of the experimentally measured species.
This mechanism was originally much larger, but over time, it was substantially
reduced by the elimination of species and reactions deemed to be unimportant.

The surf.inp file includes two surface species, open Si sites and hydrogen covered
silicon sites, plus solid silicon as a bulk species. There are 8 surface reactions, all
written as irreversible reactions with placeholders for the thermodynamic data of the
surface species. These reactions include the two-site dissociative adsorption of
silane, disilane and trisilane, which form hydrogenated silicon sites, deposited bulk
silicon, and H2. Collisions of Si atoms, SiH2, H3SiSiH and H2SiSiH2 species with the
surface result in deposition of bulk silicon and formation of gas-phase H2 with unit
probability. Adsorbed hydrogen is removed from the surface by the associative
desorption of H2, where a coverage dependence parameter has been used to alter
the default second-order dependence to the experimentally observed first order.

4.4.3 Silicon Deposition from Trichlorosilane

This chemistry set describes the deposition of silicon from trichlorosilane in a
hydrogen carrier gas. This mechanism is built on a significant number of
independently published chemical kinetic parameters, and is described in more detail
in a publication by Ho, et al.48

47. Laser-Induced Fluorescence Measurements and Kinetic Analysis of Si Atom For-

mation in a Rotating Disk Chemical Deposition Reactor, P. Ho, M. E. Coltrin,
W. G. Breiland, J. Phys. Chem. 98: 10138 (1994), and references therein.

CK-TUT-10112-1112-UG-1 247 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

The chem.inp file includes 3 elements, 11 gas-phase species, and 9 reversible gas-
phase reactions. The gas-phase reactions are reversible decomposition reactions of
various chlorosilanes and chlorinated disilanes. These reactions cause the
conversion of some of the initial chlorosilane starting material to these other gas-
phase species, which can be significant because the less-chlorinated molecules have
higher surface reactivities. These reactions are written as unimolecular
decomposition reactions at their high-pressure limit, so this reaction mechanism
would tend to overstate the importance of gas-phase chemistry if it were used at lower
total pressures.

The surf.inp file has 3 surface species: open silicon surface sites, hydrogen-covered
sites, and chlorine-covered sites, plus solid silicon as a bulk phase. There are 10
surface reactions, all written as irreversible reactions with placeholders for the
thermodynamic data of the surface species. These reactions include the dissociative
adsorption of SiCl3H, SiCl2H2 and SiCl4, which result in the formation of deposited
silicon and hydrogenated/chlorinated silicon surface species. These are lumped
reactions to the extent that the initial adsorption event, plus the successive transfer of
H and Cl atoms from the initially adsorbed Si to other surface Si atoms, are all
described as one step. The physical interpretation of this lumping is that adsorption
of the gas-phase species is assumed to be slow compared to subsequent transfer of
H and Cl atoms on the surface. For a chemically balanced reaction, this is written as
involving 4 open sites, which by default would make the reaction fourth order in open
sites, so the coverage dependence option has been used to set the kinetics to a more
reasonable first-order dependence. Other reactions include the dissociative
adsorption and associative desorption of H2, HCl and SiCl2, plus the dissociative
adsorption of HSiCl. In many cases, the rate parameters are based on experimental
surface-science studies, in other cases they are the result of fitting a model to
experimental silicon deposition rate data from a specific CVD reactor.

4.4.4 Alumina ALD

This chemistry set describes the atomic layer deposition (ALD) of alumina from
trimethylaluminum (TMA) and ozone. This mechanism is deliberately simplistic for
illustration purposes only. It demonstrates one way of describing the ALD of alumina,
but it should generally be considered as illustrative only and not used as a source of
kinetic data for this process. This mechanism is designed to deposit stoichiometric

48. Chemical Kinetics for Modeling Silicon Epitaxy from Chlorosilane, P. Ho,
A. Balakrishna, J. M. Chacin, A. Thilderkvist, B. Haas, and P. B. Comita, in Fundamental
Gas-Phase and Surface Chemistry of Vapor-Phase Materials Synthesis, T. J. Mountziaris,
M. D. Allendorf, K. F. Jensen, R. K. Ulrich, M. R. Zachariah, and M. Meyyappan, Editors,
PV 98-23, p. 117-122, Proceedings of the 194th Meeting of the Electrochemical Society,
11/1-6/98, The Electrochemical Society Proceedings Series, Pennington, NJ (1999).

2011 Reaction Design 248 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

alumina, with three oxygen atoms being deposited for each two aluminum atoms. If
the goal of a simulation was to track the elemental make-up of a material with a wide
range of possible compositions, the choice of surface species and reactions would
have to be quite different.

The fact that the different chemicals are separately pulsed into the system, rather than
being mixed in the gas, prevents many possible gas-phase reactions from occurring
in this process. The gas-phase chemistry described in the chem.inp file has
5 elements: Al, C, H, O, and Ar; and 6 gas-phase species: AlMe3, O, O2, O3, C2H6
and Ar. There are only two gas phase reactions; the collisionally-induced
decomposition of ozone and the reaction of O atoms with ozone to form molecular
oxygen. The rate parameters are from Benson and Axworthy.49 This mechanism does
not include reactions for TMA decomposition because it was determined to be too
slow at the temperatures of interest. But if such reactions were included, they would
be listed in the same chemistry input file. The temporal separation of the gas mixtures
is accounted for by omitting reactions between gas-phase species that would not be
present in the reactor at the same time, such as reactions between TMA and O
atoms. A reaction mechanism for alumina CVD, in contrast, would need to include all
such reactions.

The surface chemistry described in the surf.inp file defines three surface species:
O(S), ALME2(S) and ALMEOALME(S); where the last species occupies two surface
sites, plus the bulk alumina AL2O3(B). There are only three surface reactions, each of
which represents several elementary surface processes. The first surface reaction
represents the dissociative adsorption of TMA on the oxygenated surface species
O(S) to form the ALME2(S), combined with the recombination of two methyl groups
and desorption of an ethane molecule. This reaction has been given a moderately
high sticking coefficient of 0.1, and is written in terms of a half ethane molecule in
order to have a balanced reaction that is first order in TMA. This reaction only occurs
in the presence of TMA, and will terminate when all of the O(S) surface species have
reacted. The second surface reaction describes gas-phase oxygen atoms reacting
with the two ALME2(S) to form the ALMEOALME(S) species and a gas-phase ethane
molecule. The third surface reaction describes an O atom reacting with the
ALMEOALME(S) species to deposit AL2O3(B), regenerate O(S), and form gas-phase
ethane, where fractional molecules are used to write balanced reactions. Oxygen
atoms are expected to be very reactive, so these reactions have been given high
sticking coefficients of 1.0. These reactions only occur in the presence of O atoms
generated from ozone decomposition, and will terminate when all the methylated

49. S. W. Benson and A. E. Axworthy, Jr., J. Chem. Phys., 26:1718 (1957).

CK-TUT-10112-1112-UG-1 249 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

surface species have reacted. All the surface reactions are irreversible. As a result,
the thermodynamic properties for surface species provide the elemental composition
of the surface species, but the polynomial fitting parameters are merely placeholder
values.

4.4.5 Chlorine Plasma

This chemistry set is for a pure chlorine plasma without surface etching reactions that
is based on work published by Meeks and coworkers.50, 51

The gas-phase chemistry in the chlorine plasma is relatively simple. The element list
contains 3 elements: E (the electron), Cl and Si. Si does not actively participate in any
chemical reactions, as it only appears in surf.inp in the composition of surface species
on the wall material, but it still needs to be included in the element list in the GAS-
PHASE KINETICS input file. The gas-phase species list contains seven species: Cl2, Cl,
Cl* (chlorine atoms in a metastable electronically excited state), Cl2+, Cl+, Cl- and E.
The gas-phase reactions include electron collisions with Cl2 leading to vibrational and
electronic excitation, dissociation, ionization, and dissociative attachment. Electron
reactions with Cl include electronic excitation into a number of excited states,
including Cl* formation, and ionization. The gas-phase reaction mechanism also
includes electron collisions with Cl- leading to electron detachment, electron collisions
with Cl* leading to ionization, and gas-phase neutralization of Cl- with Cl+ and Cl2+
ions. All the reactions are irreversible as is typical of non-thermal plasmas. In low-
pressure plasmas, ionization and dissociation are balanced primarily by surface
recombination reactions. The rates for the electron-impact reactions depend on the
electron energy, rather than the neutral gas temperature.

The surface mechanism for reactions occurring on the reactor wall is also fairly
simple. It only involves neutralization of Cl+ and Cl2+ with electrons (subject to the
Bohm criterion), de-excitation of Cl* and radical recombination reactions for Cl to Cl2.
The neutralization and de-excitation reactions are non-site specific, but the
recombination reactions are described in terms of open and Cl covered sites.
Although this example problem does not include surface etching reactions, surface
recombination and neutralization reactions can be quite important in determining the

50. Modeling of Plasma-Etch Processes Using Well Stirred Reactor Approximations

and Including Complex Gas-Phase and Surface-Reactions, E. Meeks and J. W. Shon, IEEE
Transactions On Plasma Science, 23(#4):539-549 (1995).
51. Effects of Atomic Chlorine Wall Recombination: Comparison of a Plasma Chemis-
try Model With Experiment, E. Meeks, J. W. Shon, Y. Ra, P. Jones, JVSTA 13(#6):2884-2889
(1995).

2011 Reaction Design 250 CK-TUT-10112-1112-UG-1

Chapter 4: Materials Problems Tutorials Manual

overall composition of these kinds of low-pressure plasmas. All the surface reactions
are irreversible. The thermodynamic properties for surface species, therefore, provide
the elemental composition of the surface species, but the polynomial fitting
parameters are considered placeholder values and are not used in the simulation.

4.4.6 Fluorocarbon Plasma with SiO2 Etch Products

This chemistry set is for a fluorocarbon plasma (C2F6) used for etching silicon dioxide
films, a process that is used in the fabrication of microelectronic and MEMS devices.
The chemistry set includes detailed surface reactions, including multiple materials
with different reaction sets. Although it is quite complex, this mechanism is actually a
small, early version of a more complete mechanism developed by Meeks and
coworkers for etching several fluorocarbon precursors. Detailed descriptions of the
full mechanisms, including information on the sources of rate parameters, can be
found in their publications.52, 53, 54

The gas-phase chemistry described in the chem.inp file involves 6 elements, 36

species, of which there are 9 positive ions, 3 negative ions, the electron, and 23
neutral species. There are 149 gas-phase reactions. Of these, 75 are electron-impact
reactions leading to dissociation, dissociative ionization, attachment, vibrational and
electronic excitation. Such reactions are included for the starting reactant C2F6,
dissociation fragments such as CF3 and CF2, as well as etching products such as O2
and SiF4. The electron-impact reactions are all irreversible, and the rates depend on
the electron energy, rather than the neutral gas temperature. There are 22 irreversible
reactions describing the neutralization between positive and negative ions in the gas
phase. Such reactions have high A-factors and are not energy dependent. The
remaining 46 reactions are reactions of neutral gas-phase species, mostly reversible,
some with explicitly specified reverse reaction rates. Many of the reactions include the
participation of new specific collision partners (M), and a number of the unimolecular
decomposition/bimolecular recombination reactions have detailed descriptions of the
pressure dependence of the rate parameters.

52. Modeling the Plasma Chemistry of C2F6 and CHF3 Etching of Silicon Dioxide, with
Comparisons to Etch Rate and Diagnostic Data, P. Ho, J. E. Johannes, R. J. Buss, and
E. Meeks, J. Vac. Sci. Technol. A 19:2344 (2001).
53. Plasma Modeling, E. Meeks and P. Ho, Chapter 3 in Advanced Plasma Process-
ing Technologies, edited by R. J. Shul and S. J. Pearton, Springer-Verlag, Heidelberg,
(2000).
54. Chemical Reaction Mechanisms for Modeling the Fluorocarbon Plasma Etch of Sil-
icon Oxide and Related Materials, P. Ho, J. E. Johannes, R. J. Buss and E. Meeks, Sandia
National Laboratories Technical Report No. SAND2001-1292.

CK-TUT-10112-1112-UG-1 251 2011 Reaction Design

CHEMKIN Chapter 4: Materials Problems

The surface chemistry described in the surf.inp file involves three materials with
different reaction sets. The different materials in the model correspond to different
physical parts of a plasma reactor that are all exposed to the plasma, but are affected
by the plasma in different ways. The material called SIDEWALL and the material
called WAFER are both silicon dioxide. The wafer is expected to have an applied
electrical bias, resulting in higher ion energies. It thus has a more extensive set of
reactions, especially ion-enhanced chemical reactions that result in etching. Although
both sets of reactions describe the chemical species in the plasma interacting with
silicon dioxide, the mechanism description needs to have unique names for the two
sets of materials, surface sites, surface species, and bulk species to allow for different
reaction sets. The material SIDEWALL has a surface site called GLASS with species
SIO2(S) and SIO2_F2(S), and a bulk phase QUARTZ with a bulk species SIO2(B).
The material WAFER has a surface site called AMOXIDE (for amorphous oxide) with
species WSIO2(S), WSIO2_F2(S), and WSIO2_CF2(S), and a bulk phase OXIDE
with a bulk species WSIO2(B). In the species names, the (S) is a convention often
used to indicate a surface species, the _F2 and the _CF2 indicate a surface silicon
oxide site with two F atoms or a C atom and two F atoms bonded to it, respectively,
and the (B) indicates a bulk species. All the surface reactions are irreversible. All ion-
surface reactions are subject to the Bohm flux criterion.

There are 31 surface reactions for the material WAFER. First there is a reaction
describing the spontaneous etching of silicon dioxide by F atoms, and a reaction
describing the adsorption reaction of F atoms with an open-site surface species to
form a fluorinated surface species. There are 5 reactions describing the ion-enhanced
etching of SiO2 from the fluorinated surface species, producing SiF4 and O2 as etch
products, and regenerating the open SIO2(S) species. These reactions have yields
(number of surface sites converted per ion) that depend on the ion energy, as well as
overrides of the default order of the reaction. More details about these features are
provided in the input manual. Next are reactions describing the adsorption of CFx
radicals to form the WSIO2_CF2(S) species, and 5 reactions describing the ion-
enhanced etching of SiO2 from the fluorocarbon-covered surface species, producing
SiF4 and CO as etch products, and regenerating the open SIO2(S) species. These
reactions also have ion energy dependent yields and overrides of the default order of
the reaction. There are 9 reactions describing ion-neutralization with electrons on the
surface, plus 7 reactions describing the direct sputtering of SiO2 by ions.

The reaction set for the material SIDEWALL is a 16 reaction subset of that for the
material WAFER, but with the appropriate species names. The material called
TOPWALL is defined as silicon with two surface species. However, the reactions
consist only of non-site specific neutralization reactions of positive ions with electrons.

2011 Reaction Design 252 CK-TUT-10112-1112-UG-1

 CHEMKIN

5 Chemical Mechanism Analysis

5.1 Mechanism Analyzer

In the CHEMKIN software, the specification of the chemical reaction rates and
thermochemical data is necessarily very compact and efficient. As a result, however,
the information that might be most useful to a user in developing or analyzing a
reaction mechanism is often not readily available because it is hidden in the terse
Pre-processor input files. One example is rate information about the reverse rate of a
reversible reaction, which is not easily determined by examining the chemistry input
or output files. Other examples would be the free energy or enthalpy change
associated with a particular reaction, or the relative rates of two reactions for a
specific set of conditions. Another category of information that could be useful to
extract from a CHEMKIN mechanism are simple measures of transport rates,
expressed in terms of dimensionless numbers, or quantities such as the pure species
viscosity, pure species thermal conductivity, or binary diffusion coefficient for various
gas-phase species in the mechanism.

The Mechanism Analyzer provides a means to obtain this type of information, without
requiring the user to do any programming. It presents, in tabular and graphical form,
detailed information about the temperature and pressure dependence of chemical
reaction rate constants and their reverse rate constants, reaction equilibrium
constants, reaction thermochemistry, chemical species thermochemistry and
transport properties. In general, the user will want to select only a few of the many
types of information to be calculated and output, but there is a great deal of flexibility
in specifying the desired output. The current version does not, however, handle all of
the special rate options available in GAS-PHASE KINETICS and SURFACE KINETICS.
Specifically, it does not consider: (1) user-provided rate routines, (2) rate-order

CK-TUT-10112-1112-UG-1 253 2011 Reaction Design

CHEMKIN Chapter 5: Chemical Mechanism Analysis

changes, (3) options associated with modeling plasma reactions, (4) the Chebyshev
polynomial option for describing pressure-dependent gas-phase reactions, the (5)
Landau-Teller rate formulation used for energy transfer processes, or (6) multiple
materials. If such options are included in a reaction mechanism, they will be ignored.

5.1.1 Background Information

5.1.1.1 Bath Gas and Carrier Gas

The concept of a bath gas is used throughout the Mechanism Analyzer. The
specification of a bath gas consists of a characteristic temperature, pressure, and
composition at which quantities are to be evaluated by default. Composition, here,
refers to the default composition for all phases defined in the mechanism. Reaction
rate information is evaluated at the bath gas conditions, unless it is tabulated as a
function of a system parameter, such as temperature. In this case, all other
parameters are fixed at the bath gas conditions in the table. The default temperature
for the bath gas is 298.15 K. The default pressure is 1 atmosphere, and the default
composition is an equimolar composition in each phase. All defaults can be
overridden by selections in the CHEMKIN Interface input panels.

The concept of a carrier gas is also used in the Mechanism Analyzer. Unless
overridden, the carrier gas is assumed to be the gas component having the largest
mole fraction. CHEMKIN calculates a single number for the characteristic time scale
of diffusion (to be compared with the characteristic time scale of reaction). To make
this comparison, the diffusion coefficient is calculated for a specified major species
in the carrier gas. For example, the diffusion of the major species CH4 in the carrier
gas H2 is used to calculate the characteristic diffusion time scale. Unless overridden,
the major species in the gas phase is assumed to be the gas component having the
second largest mole fraction.

5.1.1.2 Uniform-dimensional and Non-dimensional Reaction Rate Information

It is often useful to know, in some sense, which reactions in a mechanism are fast
and which are slow. It is difficult or misleading to simply compare rate constants,
which can have different units depending on the molecularity of the reaction. In order
to compare the rates of reactions in the mechanism, we define a quantity

Equation 5-1

Sn
* g sn
kf = kf G
n

2011 Reaction Design 254 CK-TUT-10112-1112-UG-1

Chapter 5: Chemical Mechanism Analysis Tutorials Manual

which we call the uniform-dimensional rate constant. Regardless of the order of

reaction, it will have units of mole cm-3 sec-1 for a gas-phase reaction or
mole cm-2 sec-1 for a surface reaction. In this expression, kf is the rate constant for
the forward reaction, [G] is the total concentration of gas-phase species determined at
the bath gas conditions (in mole cm-3), g is the sum of the stoichiometric coefficients
of all gas-phase species appearing as reactants in the reaction, [Sn] is the total site
density of surface phase n determined at the bath gas conditions, Sn is the sum of the
stoichiometric coefficients of all surface species in phase n participating as reactants
in this reaction.

Using the usual rate constant, one calculates the forward reaction rate as kf times the
product of the concentrations of the reactant species (in mole cm-3 for gas species,
or mole cm-2 for surface species) raised to the power of their stoichiometric
coefficients. With the uniform-dimensional rate constant, one calculates the same
*
reaction rate as k f multiplied by the (dimensionless) species mole fractions (gas-
phase reactants) or site fractions (surface species) raised to the power of their
stoichiometric coefficients. Thus, independent of the molecularity of the reaction, the
*
reaction rate is k f times quantities that have maximum values on the order of unity
(the mole and site fractions), and it is easier to compare one reaction to another.

*
The quantity k f just discussed can point out which reactions are fast relative to one
another. It can also be of interest to know if a reaction is fast relative to a competing
process like molecular transport. The Damkhler number for gas-phase reactions, Da,
allows for such a comparison.

Equation 5-2
*
kf
Da = ----------------------
-
2
DG L
In Equation 5-2, D is a diffusion coefficient and L is a characteristic length scale for
diffusion; for example, a boundary-layer thickness or a characteristic reactor
dimension. The Damkhler number is a dimensionless number that is a measure of
the relative importance of gas-phase kinetics versus molecular mass transport. If Da
is much greater than 1, then a reaction is fast relative to transport; if it is much less
than 1, then transport processes occur on a shorter time scale than kinetic processes.

CK-TUT-10112-1112-UG-1 255 2011 Reaction Design

CHEMKIN Chapter 5: Chemical Mechanism Analysis

One must supply a diffusion coefficient in Equation 5-2 to evaluate Da. To do this, the
Mechanism Analyzer requires the user to name a major species and a carrier gas
species. Through internal calls to the TRANSPORT Subroutine Library, CHEMKIN
evaluates the binary diffusion coefficient between these two species at the specified
bath temperature and pressure. The user may also specify the length scale L. The
default value for L is 1 cm.

For each gas-phase reaction, a report is generated of k* and Da for the forward and
reverse directions. In some cases, the input reaction mechanism specifies the
reaction to be irreversible. In these cases, the quantities k* and Da are still calculated
for the reverse direction, but the numbers are enclosed in square brackets [ ] to flag
these reactions as not being part of the mechanism.

A uniform-dimensional rate constant of Equation 5-1 is also calculated for surface

*
reactions. In this case, k f has units of mole cm-2 sec-1. Thus, one can make a
comparison between reaction rates for surface reactions. The surface Damkhler
number is defined to be

Equation 5-3
*
kf
Da = --------------------
-
DG L
The equation for the surface Damkhler number differs from the equation for the gas-
phase Damkhler number by a factor of the length scale L. As before, it provides a
measure of the relative speed of the surface reaction rate versus the molecular mass
transport rate.

5.1.2 Reaction Mechanism for Diamond CVD

5.1.2.1 Project Description

This user tutorial demonstrates the use of the Mechanism Analyzer model to extract a
wider variety of detailed thermodynamic, chemical kinetic, and transport data from a
reaction mechanism. The chemistry set nominally describes a diamond CVD process
and is a subset of a published mechanism,55 but one that has now been superseded
by the authors.56 The reactions in this example, especially the surface reactions,
have been contrived to demonstrate the capabilities of the Mechanism Analyzer and
should not be used as a source of kinetic data for diamond deposition.

55. Analysis of diamond growth in subatmospheric dc plasma-gun reactors. Michael

E. Coltrin and David S. Dandy, Journal of Applied Physics, 74(#9):5803 (1993).

2011 Reaction Design 256 CK-TUT-10112-1112-UG-1

Chapter 5: Chemical Mechanism Analysis Tutorials Manual

5.1.2.2 Project Setup

The project file is called mechanism_analyzer__diamond_cvd.ckprj. The data files
used for this sample are located in the
samples2010\mechanism_analyzer\diamond_cvd directory. This reactor diagram
contains only the Mechanism Analyzer. Section 5.1.2.3 illustrates which sections of
the output result from which choices on the input panels, so only brief usage
instructions are provided here.

The Output Control tab of the Output Control panel contains a number of selections
that determine the kinds of information that will be put into the output file and XML
Solution File (XMLdata.zip). At the top of the Output Control tab, there is a choice to
turn All Tables On or Off. Turning All Tables On will result in a very large output file, so
this is generally not recommended, except for the smallest reaction mechanisms. The
usage of these controls is as follows. To obtain All output of a certain type, turn it On
at the highest level. To limit the output of a certain type to specific species or
reactions, turn it Off at the higher level(s), then On for the class (gas, surface or bulk)
of species/reactions or by listing the specific species/reactions of interest.

As an illustration of how to use the levels of choices to specify data types, consider
the case of thermodynamic data tables. There are two types of data tables available:
summary tables which are given at the bath gas conditions, and detailed tables which
cover a range of temperatures. In this example, the Summary Thermo Tables are
turned Off in general, but then back On for Species, but not for Reactions. Turning All
Summary Thermo Tables On is equivalent to checking all three boxes, but is simpler.
In the case of the more detailed Species Thermo Tables, they are turned Off in
general, and On for the Bulk species only. In addition, on the Species-specific Data
panel, Species Thermo Tables are specified for two gas phase species, CH3 and CH4,
plus two surface species, CH2(S) and CH(S). The results of these selections are
shown in Section 5.1.2.3. For reaction information, most of the output types are
turned On at the general level, so there are no entries on the Reaction-specific Data
tab of the Output Control panel.

56. The authors now prefer the mechanism in: A simplified analytical model of dia-
mond growth in direct current arcjet reactors. David S. Dandy and Michael E. Coltrin,
Journal of Materials Research, 10(#8):1993 (1995). Private communication, M. E. Coltrin, Feb.
1998.

CK-TUT-10112-1112-UG-1 257 2011 Reaction Design

CHEMKIN Chapter 5: Chemical Mechanism Analysis

The C1_ Mechanism Analyzer panel allows specification of the physically based
parameters that affect the output file. The Reactor Physical Property tab allows the
user to override the default values for pressure and temperature of the bath gas, as
well as the ranges used for the thermodynamics and kinetics tables. The Species-
specific Data tab allows specification of the composition of the bath gas, as well as
overriding the default definitions for the carrier gas and major species. No entries are
made on the Cluster Properties, Solver, or Continuations panels for this problem.

5.1.2.3 Project Results

Although an XML Solution File is created and the graphical CHEMKIN Post-
Processor can be used for viewing the results, the output file is generally the more
useful form of output for this application, as the user is generally looking for values of
specific chemical or transport parameters. This section reproduces the Mechanism
Analyzer diagnostic output file, which has been augmented in a number of places by
comments in larger font. The information in the tables is generally self-explanatory, so
these comments general deal with the input controls that produce that section of
output.

The output file first echoes the input file created by the CHEMKIN Interface.

*****************************************************************
* CHEMKIN-PRO Release 15101 *
* SURFTHERM Application *
* REACTION MECHANISM ANALYSIS *
* Copyright(c) 1997-2009 Reaction Design. All Rights Reserved. *
*****************************************************************

WORKING SPACE REQUIREMENTS

PROVIDED REQUIRED
LOGICAL 54 54
INTEGER 1009 1009
REAL 2443 2443
CHARACTER*16 43 43
CHARACTER*64 10 10

Initializing CHEMKIN Gas-phase Library a component of CHEMKIN-PRO Release 15101, Build date: Aug 17, 2010
This and All Other CHEMKIN(R) Libraries are Copyright (c) 1997-2009 Reaction Design. All rights reserved.

LICENSE INFORMATION:

LicNum: 9555
Licensed to Reaction Design
Contact: A. User
Expiring: 03-feb-2011
Platform: win64

Initializing SURFACE CHEMKIN Library a component of CHEMKIN-PRO Release 15101, Build date: Aug 17, 2010
This and All Other CHEMKIN(R) Libraries are Copyright (c) 1997-2009 Reaction Design. All rights reserved.

Surface material name: MATERIAL1

Initializing TRANSPORT Library a component of CHEMKIN-PRO Release 15101, Build date: Aug 17, 2010
This and All Other CHEMKIN(R) Libraries are Copyright (c) 1997-2009 Reaction Design. All rights reserved.

2011 Reaction Design 258 CK-TUT-10112-1112-UG-1

Chapter 5: Chemical Mechanism Analysis Tutorials Manual

KEYWORD INPUT
NONE
GEN ALL
GRXN ALL
GTHB ALL
NDIM ALL
PFAL ALL
SCOV ALL
SRXN ALL
STCK ALL
TFAL ALL
THRM NONE
TRAN ALL
TSUM NONE
CARR H2
LSCL 1.3
MAJ CH4
PBTH 20.0
PHIG 700.0
PLOW 100.0
PNUM 3
TBTH 1100.0
TDEL 400.0
THIG 1200.0
TLOW 300.0
THRM BULK
THRM CH(S)
THRM CH2(S)
THRM CH3
THRM CH4
TSUM SPECIES
XBTH CH3 0.03
XBTH CH4 0.05
XBTH H 0.02
XBTH H2 0.9
END

The Turn On General Tables button in the Output Control tab gives the following general
information about the elements, species, phases, and reactions in the mechanism.

==========================================================================================
GENERAL INFORMATION CONCERNING THE CHEMKIN MECHANISM
==========================================================================================
Total number of elements declared = 2 ( H C )
Total number of species = 12
Total number of phases = 3
Total number of gas-phase reactions = 5
Total number of surface-phase reactions = 4
Universal gas constant = 8.31447E+07 erg/(mol-K)
Universal gas constant used for activation energies = 1.99 cal/(mol-K)
Pressure of one standard atmosphere = 1.01325E+06 dyne/cm^2

GAS PHASE (Always phase # 1, with the name, "GAS")

------------------------------------------------------------------------------------------
Number of species = 5
Number of surface reactions where the # of gas
products is different than the # of gas reactants = 2
Number of elements in the phase = 2 ( H C )
------------------------------------------------------------------------------------------

SURFACE PHASES
------------------------------------------------------------------------------------------
Phase Phase SS Site Density Species Element Site_changing Elements:
Number Name mole/cm^2 Count Count Surf_rxns
------------------------------------------------------------------------------------------
2 DIAMOND 5.2200E-09 6 2 0 ( H C )
Tot SS Site Dens 5.2200E-09
(used in sticking coefficient express)
------------------------------------------------------------------------------------------

BULK PHASES
------------------------------------------------------------------------------------------
Phase Phase Species Element Mole_changing Elements:
Number Name Count Count Surf_rxns
------------------------------------------------------------------------------------------
3 BULK1 1 1 1 ( C )
------------------------------------------------------------------------------------------

CK-TUT-10112-1112-UG-1 259 2011 Reaction Design

CHEMKIN Chapter 5: Chemical Mechanism Analysis

The Turn On General Tables button in the Output Control tab gives the following table of
conditions. The bath gas conditions are set in the Reactor Physical Property tab. The bath gas
composition, carrier gas and major species are set in the Species-specific Data tab.

==========================================================================================
SUMMARY OF SPECIES IN THE MECHANISM, with a DESCRIPTION OF BATH GAS COMPOSITION
==========================================================================================
Total pressure = 20.00 torr
Temperature (where needed) = 1100.00 K
Carrier Gas (used in diff. calcs) = H2
Major Gas Species (used in nondim calcs) = CH4
----------------------------------------------------------------------
Number Name Mole_fraction Concentration
----------------------------------------------------------------------
1 CH3 3.0000E-02 8.7471E-09 mol/cm^3
2 C2H6 0.000 0.000 mol/cm^3
3 CH4 5.0000E-02 1.4579E-08 mol/cm^3
4 H 2.0000E-02 5.8314E-09 mol/cm^3
5 H2 0.9000 2.6241E-07 mol/cm^3
6 CH(S) 0.1667 8.7000E-10 mol/cm^2
7 C(S,R) 0.1667 8.7000E-10 mol/cm^2
8 CH3(S) 0.1667 8.7000E-10 mol/cm^2
9 CH2(S) 0.1667 8.7000E-10 mol/cm^2
10 CH2(S,R) 0.1667 8.7000E-10 mol/cm^2
11 CH(S,R) 0.1667 8.7000E-10 mol/cm^2
12 D 1.000 (activity) unit1ess
----------------------------------------------------------------------

The Turn On Summary Tables for Species checkbox gives the following table of
thermochemical data.

==========================================================================================
SUMMARY OF STANDARD STATE THERMODYNAMIC FUNCTIONS FOR SPECIES AT BATH GAS CONDITIONS
==========================================================================================
Bath Gas Temperature = 1100.00K
------------------------------------------------------------------------------------------
Number Name H(298 K) H(T_bath) Cp(T_bath) S(T_bath) G(T_bath)
kcal/mol cal/(mol-K) kcal/mol
------------------------------------------------------------------------------------------
1 CH3 34.82 44.55 14.64 61.47 -23.06
2 C2H6 -20.67 -2.14 30.89 82.90 -93.33
3 CH4 -17.90 -7.00 18.07 60.84 -73.92
4 H 52.10 56.08 4.97 33.88 18.82
5 H2 0.00 5.67 7.32 40.39 -38.77
6 CH(S) 0.00 4.84 8.80 7.16 -3.04
7 C(S,R) 43.36 46.63 5.24 5.15 40.96
8 CH3(S) 17.30 26.17 16.04 16.57 7.95
9 CH2(S) -11.49 -4.63 12.69 10.13 -15.77
10 CH2(S,R) 50.66 57.52 12.69 10.13 46.38
11 CH(S,R) 31.88 36.71 8.80 7.16 28.84
12 D 0.45 3.72 5.24 5.15 -1.95
------------------------------------------------------------------------------------------

The Turn On General Tables button in the Output Control tab gives the following general
information about reactions in the mechanism.

SURFTHERM Mechanism Analyzer: Initializing gas reactions...

==========================================================================================
SHORT DESCRIPTION OF GAS-PHASE REACTIONS
==========================================================================================
---------------------------------------------------------------------------
Number Description Gas_Mole Gas_Mole
Change Reactants

2011 Reaction Design 260 CK-TUT-10112-1112-UG-1

Chapter 5: Chemical Mechanism Analysis Tutorials Manual

---------------------------------------------------------------------------
1 2CH3(+M)<=>C2H6(+M) -1.00 2.00
2 CH4+H<=>CH3+H2 0.00 2.00
3 CH3+H(+M)<=>CH4(+M) -1.00 2.00
4 2H+M<=>H2+M -1.00 2.00
5 2H+H2<=>2H2 -1.00 3.00
---------------------------------------------------------------------------

SURFTHERM Mechanism Analyzer: Initializing surface reactions...

==========================================================================================
SHORT DESCRIPTION OF SURFACE-PHASE REACTIONS
==========================================================================================
------------------------------------------------------------------------------------------
Number Description Gas Mole Surf Mole Bulk Mole Surf_Site
Change Change Change Change
------------------------------------------------------------------------------------------
1 CH(S)+H<=>C(S,R)+H2 0.00 0.00 0.00 0.00
2 C(S,R)+H=>CH(S) -1.00 0.00 0.00 0.00
3 C(S,R)+CH3<=>D+CH3(S) -1.00 0.00 1.00 0.00
4 CH2(S,R)+CH(S,R)<=>CH2(S)+CH(S) 0.00 0.00 0.00 0.00
------------------------------------------------------------------------------------------

The Turn On Uni-dimensional Rate Tables button in the Output Control panel gives the
following tables of uniform-dimensional rates that allow comparisons between reactions of
different orders. The length scale used for evaluating the relative importance of kinetics to
transport is set on the Reactor Physical Property tab.

====================================================================================================
NON-DIMENSIONAL GAS REACTION RATE CONSTANTS AT THE BATH GAS CONDITIONS
====================================================================================================
Total Pressure = 20.00 torr
Temperature = 1100.00 K
----------------------------------------------------------------------------------------------------
Number Description k_star k_star_rev Gas_Da_For Gas_Da_Rev
mol/(cm^3-s)
----------------------------------------------------------------------------------------------------
1 2CH3(+M)<=>C2H6(+M) 0.2703 4.2798E-09 6113. 9.6773E-05
2 CH4+H<=>CH3+H2 4.5456E-02 2.0959E-03 1028. 47.39
3 CH3+H(+M)<=>CH4(+M) 2.8128E-02 1.5344E-14 636.0 3.4694E-10
4 2H+M<=>H2+M 2.2528E-06 5.6662E-20 5.0939E-02 1.2812E-15
5 2H+H2<=>2H2 3.4125E-05 8.5829E-19 0.7716 1.9407E-14
----------------------------------------------------------------------------------------------------
NOTE ON THE ABSOLUTE NUMBERS IN THIS TABLE:
The rate constants (mol/(cm^3-s)) should be compared to rate of mass transport in order
to characterize their values as being fast or slow. The nondimensionalization of the mass transport
involves the following multiplicative factor, which also has the units of mol/(cm^3-s):
Total_Concentration * Diffusivity / Length_scale^2
Using the binary diffusion coefficient between H2
and CH4, the following factors are calculated at bath gas conditions:
Total Concentration = 2.9155E-07 mol/cm^3
Binary Diffusion Coefficient = 256.4 cm^2/s
Length scale = 1.300 cm
Therefore, the non-dimensionalization factor for gas reactions becomes:
Conc * Diff / Length^2 = 4.4226E-05 mol/(cm^3-s)
Note that this number is independent of pressure

====================================================================================================
NON-DIMENSIONAL SURFACE REACTION RATE CONSTANTS AT THE BATH GAS CONDITIONS
====================================================================================================
Total Pressure = 20.00 torr
Temperature = 1100.00 K
----------------------------------------------------------------------------------------------------
Number Description k_star k_star_rev Surf_Da_For Surf_Da_Rev
mol/(cm^2-s)
----------------------------------------------------------------------------------------------------
1 CH(S)+H<=>C(S,R)+H2 2.8714E-03 5.7498E-06 49.94 0.1000
2 C(S,R)+H=>CH(S) 1.2367E-02 [ 1.5531E-13] 215.1 [ 2.7014E-09]
3 C(S,R)+CH3<=>D+CH3(S) 3.5161E-03 54.64 61.16 9.5040E+05
4 CH2(S,R)+CH(S,R)<=>CH2(S)+CH(S) 654.8 1.3621E-16 1.1390E+07 2.3691E-12
----------------------------------------------------------------------------------------------------
[ ] indicates that this reaction is not in mechanism
NOTE ON THE ABSOLUTE NUMBERS IN THIS TABLE:
The rate constants (mol/(cm^2-s)) should be compared to rate of mass transport to the surface in order
to characterize their values as being fast or slow. The nondimensionalization of the mass transport
involves the following multiplicative factor, which also has the units of mol/(cm^2-s):
Total_Concentration * Diffusivity / Length_scale
Using the binary diffusion coefficient between H2

CK-TUT-10112-1112-UG-1 261 2011 Reaction Design

CHEMKIN Chapter 5: Chemical Mechanism Analysis

and CH4, the following factors are calculated at bath gas conditions:
Total Concentration = 2.9155E-07 mol/cm^3
Binary Diffusion Coefficient = 256.4 cm^2/s
Length scale = 1.300 cm
Therefore, the non-dimensionalization factor for surface reactions becomes:
Conc * Diff / Length = 5.7493E-05 mol/(cm^2-s)
Note that this number is independent of pressure

Listing these species on the Species-specific Data tab, combined with the Turn On Bulk
Species Thermo Tables checkbox in the Output Control panel gives the following tables of
thermodynamic data for two gas-phase species, two surface species and the one bulk
species. The Turn On Transport Tables button adds the table of Viscosity, Thermal
Conductivity and Binary Diffusion Coefficients to the tables for the gas-phase species. The
temperature intervals are set on the Reactor Physical Property tab.

====================================================================================================
THERMO TABLE FOR MOLECULE "CH3" IN PHASE "GAS"
====================================================================================================
Overall, this is the 1th species in the mechanism
It is the 1th species in phase GAS
Elemental Composition:
H: 3
C: 1
L-J Potential well depth = 144.0 K
L-J collision diameter = 3.800 Angstroms
Dipole Moment = 0.000 Debye
Polarizability = 0.000 Angstroms^3
Rotational Collision number at 298K = 0.000
This molecule is linear
Heat of Formation at 298 = 34.823 kcal/mol
Molecular Weight = 15.04 gm/mol
----------------------------------------------------------------------------------------------------
Temp H-H298 G-H298 Cp S Viscosity Therm_Cond Dif_Co_with_H2
K kcal/mol cal/(mol-K) gm/(cm-s) erg/(cm-s-K) cm^2/s
----------------------------------------------------------------------------------------------------
298.15 0.000 -13.828 9.21 46.38 1.0685E-04 3556. 28.00
300.00 0.017 -13.914 9.23 46.44 1.0739E-04 3581. 28.30
700.00 4.336 -34.452 12.20 55.41 2.0316E-04 9207. 119.0
1100.00 9.732 -57.882 14.64 61.47 2.7667E-04 1.4856E+04 252.8
----------------------------------------------------------------------------------------------------
[Pressure for binary diffusion coeff. calc. = 20.00 torr]

====================================================================================================
THERMO TABLE FOR MOLECULE "CH4" IN PHASE "GAS"
====================================================================================================
Overall, this is the 3th species in the mechanism
It is the 3th species in phase GAS
Elemental Composition:
H: 4
C: 1
L-J Potential well depth = 141.4 K
L-J collision diameter = 3.746 Angstroms
Dipole Moment = 0.000 Debye
Polarizability = 2.600 Angstroms^3
Rotational Collision number at 298K = 13.00
This molecule is non-linear
Heat of Formation at 298 = -17.900 kcal/mol
Molecular Weight = 16.04 gm/mol
----------------------------------------------------------------------------------------------------
Temp H-H298 G-H298 Cp S Viscosity Therm_Cond Dif_Co_with_H2
K kcal/mol cal/(mol-K) gm/(cm-s) erg/(cm-s-K) cm^2/s
----------------------------------------------------------------------------------------------------
298.15 0.000 -13.259 8.40 44.47 1.1427E-04 3406. 28.41
300.00 0.016 -13.341 8.43 44.52 1.1484E-04 3436. 28.71
700.00 4.464 -33.081 13.70 53.64 2.1673E-04 1.0920E+04 120.7
1100.00 10.904 -56.016 18.07 60.84 2.9497E-04 1.8790E+04 256.4
----------------------------------------------------------------------------------------------------
[Pressure for binary diffusion coeff. calc. = 20.00 torr]

====================================================================================================
THERMO TABLE FOR MOLECULE "CH(S)" IN PHASE "DIAMOND"
====================================================================================================
Overall, this is the 6th species in the mechanism
It is the 1th species in phase DIAMOND
Elemental Composition:

2011 Reaction Design 262 CK-TUT-10112-1112-UG-1

Chapter 5: Chemical Mechanism Analysis Tutorials Manual

H: 1
C: 1
Number of surface sites occupied by the species = 1
Heat of Formation at 298 = 0.000 kcal/mol
Molecular Weight = 13.02 gm/mol
-----------------------------------------------------------------
Temp H-H298 G-H298 Cp S
K kcal/mol cal/(mol-K)
-----------------------------------------------------------------
298.15 0.00 -0.106 1.45 0.357
300.00 2.701E-03 -0.107 1.47 0.366
700.00 1.70 -0.858 6.52 3.65
1100.00 4.84 -3.04 8.80 7.16
-----------------------------------------------------------------

====================================================================================================
THERMO TABLE FOR MOLECULE "CH2(S)" IN PHASE "DIAMOND"
====================================================================================================
Overall, this is the 9th species in the mechanism
It is the 4th species in phase DIAMOND
Elemental Composition:
H: 2
C: 1
Number of surface sites occupied by the species = 1
Heat of Formation at 298 = -11.490 kcal/mol
Molecular Weight = 14.03 gm/mol
-----------------------------------------------------------------
Temp H-H298 G-H298 Cp S
K kcal/mol cal/(mol-K)
-----------------------------------------------------------------
298.15 0.00 -0.152 2.04 0.511
300.00 3.809E-03 -0.153 2.08 0.524
700.00 2.38 -1.21 9.20 5.12
1100.00 6.86 -4.28 12.7 10.1
-----------------------------------------------------------------

====================================================================================================
THERMO TABLE FOR MOLECULE "D" IN PHASE "BULK1"
====================================================================================================
Overall, this is the 12th species in the mechanism
It is the 1th species in phase BULK1
Elemental Composition:
C: 1
Bulk Density = 3.515 gm/cm^3
Activity (bath gas dependent) = 1.000
Heat of Formation at 298 = 0.454 kcal/mol
Molecular Weight = 12.01 gm/mol
-----------------------------------------------------------------
Temp H-H298 G-H298 Cp S
K kcal/mol cal/(mol-K)
-----------------------------------------------------------------
298.15 0.00 -0.113 1.44 0.378
300.00 2.684E-03 -0.113 1.46 0.387
700.00 1.29 -0.762 4.48 2.93
1100.00 3.27 -2.40 5.24 5.15
-----------------------------------------------------------------

The Turn On Gas Reaction Tables button in the Output Control tab gives the following tables
of high-pressure rates constants as a function of temperature for the gas-phase reactions.
Reactions 1 and 3 have (+M), so for these reactions, the Turn On Pressure Tables button
adds a table of rate constants as a function of pressure at the bath-gas temperature, and the
Turn On Temperature Tables buttons add a table of rate constants as a function of
temperature at the bath-gas pressure. Reaction 4 has +M, so for this reaction, the Turn On
3rd Body Reaction Tables button adds a table where the effect of the third body is lumped
into the rate constants. The bath gas conditions and temperature/pressure intervals are set on
the Reactor Physical Property tab.

===================================================================================================================================
Gas Reaction # 1 2CH3(+M)<=>C2H6(+M)
===================================================================================================================================
Change in moles in the reaction = -1.00
This reaction does have third body effects.
1 enhanced third body efficiencies were input

CK-TUT-10112-1112-UG-1 263 2011 Reaction Design

CHEMKIN Chapter 5: Chemical Mechanism Analysis

Species "H2" enhanced third body efficiency for the reaction = 2.000
This is a reversible reaction, having 2.00 reactant species and 1.00 product species
k cm^3/(mol-s) = 9.0300E+16 T^( -1.180 ) exp( - 0.65 kcal/mol / RT)
Reaction has a pressure-dependent behavior with a 6 parameter Troe function form:
klow cm^6/(mol^2-s) = 3.1800E+41 T^( -7.030 ) exp( - 2.76 kcal/mol / RT)
a = 0.6041
T*** = 6927. K
T* = 132.0 K

HIGH PRESSURE GAS REACTION RATE CONSTANTS AS A FUNCTION OF TEMPERATURE

-----------------------------------------------------------------------------------------------------------------------------------
UnifDimensnl Rate_Const
Temp k Afact Ea DeltaG DeltaH DeltaS k_rev Afact_rev Ea_rev k_star k_star_rev
K cm^3/(mol-s) kcal/mol kcal/mol cal/(mol-K) sec-1 kcal/mol mol/(cm^3-s)
-----------------------------------------------------------------------------------------------------------------------------------
298.15 3.601E+13 3.337E+13 -0.05 -79.16 -90.32 -37.42 1.384E-49 7.552E+16 89.68 38.8 1.384E-55
300.00 3.600E+13 3.313E+13 -0.05 -79.09 -90.33 -37.45 3.519E-49 7.590E+16 89.68 38.2 3.496E-55
700.00 2.479E+13 1.219E+13 -0.99 -63.31 -91.52 -40.30 7.410E-12 5.018E+16 89.14 2.83 1.846E-18
1100.00 1.725E+13 7.151E+12 -1.93 -47.21 -91.25 -40.03 7.964E-02 1.633E+16 87.14 0.270 4.280E-09
-----------------------------------------------------------------------------------------------------------------------------------

PRESSURE-DEPENDENT BEHAVIOR AS A FUNCTION OF TEMPERATURE

Reaction # 1
Bath Gas Pressure = 20.00 torr
-----------------------------------------------------------------------------------------------------------------------------------
Temp k Afact Ea k/kinf klow Reduc_Pres FC EffConc k_rev Afact_rev Ea_rev
K cm^3/(mol-s) kcal/mol cm^6/(mol^2-s) mol/cm^3 sec-1 kcal/mol
-----------------------------------------------------------------------------------------------------------------------------------
298.15 3.350E+13 2.644E+13 -0.14 0.930 1.209E+22 686. 0.931 2.044E-06 1.287E-49 5.983E+16 89.59
300.00 3.345E+13 2.615E+13 -0.15 0.929 1.192E+22 672. 0.931 2.031E-06 3.270E-49 5.991E+16 89.59
700.00 1.348E+13 1.279E+12 -3.28 0.544 4.356E+20 15.3 0.579 8.705E-07 4.030E-12 5.264E+15 86.85
1100.00 3.181E+12 7.839E+10 -8.09 0.184 3.738E+19 1.20 0.338 5.539E-07 1.468E-02 1.790E+14 80.97
-----------------------------------------------------------------------------------------------------------------------------------

PRESSURE-DEPENDENT BEHAVIOR AS A FUNCTION OF PRESSURE

Reaction # 1
Bath Gas Temperature = 1100.00K
Low Pressure Limiting Reaction Rate = klow = 3.7385E+19 cm^6/(mol^2-s)
-----------------------------------------------------------------------------------------------------------------------------------
Pres k Afact Ea k/kinf klow*EffConc Reduc_Pres FC EffConc k_rev Afact_rev Ea_rev
torr cm^3/(mol-s) kcal/mol cm^3/(mol-s) mole/cm^3 sec-1 kcal/mol
-----------------------------------------------------------------------------------------------------------------------------------
700.00 1.151E+13 1.447E+12 -4.53 0.667 7.248E+14 42.0 0.683 1.939E-05 5.312E-02 3.306E+15 84.53
264.58 9.107E+12 5.942E+11 -5.97 0.528 2.739E+14 15.9 0.561 7.328E-06 4.203E-02 1.357E+15 83.09
100.00 6.451E+12 2.165E+11 -7.42 0.374 1.035E+14 6.00 0.436 2.770E-06 2.977E-02 4.944E+14 81.64
-----------------------------------------------------------------------------------------------------------------------------------

===================================================================================================================================
Gas Reaction # 2 CH4+H<=>CH3+H2
===================================================================================================================================
Change in moles in the reaction = 0.00
This reaction does not have any third body effects
This is a reversible reaction, having 2.00 reactant species and 2.00 product species
k cm^3/(mol-s) = 2.2000E+04 T^( 3.000 ) exp( - 8.75 kcal/mol / RT)

HIGH PRESSURE GAS REACTION RATE CONSTANTS AS A FUNCTION OF TEMPERATURE

-----------------------------------------------------------------------------------------------------------------------------------
UnifDimensnl Rate_Const
Temp k Afact Ea DeltaG DeltaH DeltaS k_rev Afact_rev Ea_rev k_star k_star_rev
K cm^3/(mol-s) kcal/mol kcal/mol cal/(mol-K) cm^3/(mol-s) kcal/mol mol/(cm^3-s)
-----------------------------------------------------------------------------------------------------------------------------------
298.15 2.249E+05 1.171E+13 10.53 -1.08 0.63 5.73 3.612E+04 6.546E+11 9.90 2.602E-07 4.179E-08
300.00 2.509E+05 1.193E+13 10.54 -1.09 0.63 5.75 4.004E+04 6.612E+11 9.91 2.868E-07 4.575E-08
700.00 1.399E+10 1.516E+14 12.92 -3.81 1.31 7.31 9.043E+08 3.820E+12 11.61 2.937E-03 1.898E-04
1100.00 5.348E+11 5.881E+14 15.31 -6.73 1.14 7.15 2.466E+10 1.613E+13 14.17 4.546E-02 2.096E-03
-----------------------------------------------------------------------------------------------------------------------------------

===================================================================================================================================
Gas Reaction # 3 CH3+H(+M)<=>CH4(+M)
===================================================================================================================================
Change in moles in the reaction = -1.00
This reaction does have third body effects.
1 enhanced third body efficiencies were input
Species "H2" enhanced third body efficiency for the reaction = 2.000
This is a reversible reaction, having 2.00 reactant species and 1.00 product species
k cm^3/(mol-s) = 6.0000E+16 T^( -1.000 ) exp( - 0.00 kcal/mol / RT)
Reaction has a pressure-dependent behavior with a Lindeman function form:
klow cm^6/(mol^2-s) = 8.0000E+26 T^( -3.000 ) exp( - 0.00 kcal/mol / RT)

HIGH PRESSURE GAS REACTION RATE CONSTANTS AS A FUNCTION OF TEMPERATURE

-----------------------------------------------------------------------------------------------------------------------------------
UnifDimensnl Rate_Const
Temp k Afact Ea DeltaG DeltaH DeltaS k_rev Afact_rev Ea_rev k_star k_star_rev
K cm^3/(mol-s) kcal/mol kcal/mol cal/(mol-K) sec-1 kcal/mol mol/(cm^3-s)
-----------------------------------------------------------------------------------------------------------------------------------
298.15 2.012E+14 7.403E+13 -0.59 -96.09 -104.82 -29.30 3.047E-61 2.820E+15 103.64 54.6 7.690E-68
300.00 2.000E+14 7.358E+13 -0.60 -96.03 -104.83 -29.34 8.961E-61 2.836E+15 103.64 52.9 2.216E-67
700.00 8.571E+13 3.153E+13 -1.39 -83.31 -106.69 -33.41 1.463E-17 4.041E+15 103.91 0.416 1.551E-25
1100.00 5.455E+13 2.007E+13 -2.19 -69.68 -107.63 -34.51 8.675E-06 2.845E+15 103.26 2.813E-02 1.534E-14
-----------------------------------------------------------------------------------------------------------------------------------

2011 Reaction Design 264 CK-TUT-10112-1112-UG-1

Chapter 5: Chemical Mechanism Analysis Tutorials Manual

PRESSURE-DEPENDENT BEHAVIOR AS A FUNCTION OF TEMPERATURE

Reaction # 3
Bath Gas Pressure = 20.00 torr
-----------------------------------------------------------------------------------------------------------------------------------
Temp k Afact Ea k/kinf klow Reduc_Pres FC EffConc k_rev Afact_rev Ea_rev
K cm^3/(mol-s) kcal/mol cm^6/(mol^2-s) mol/cm^3 sec-1 kcal/mol
-----------------------------------------------------------------------------------------------------------------------------------
298.15 4.722E+13 1.748E+12 -1.95 0.235 3.018E+19 0.307 1.00 2.044E-06 7.149E-62 6.659E+13 102.28
300.00 4.626E+13 1.696E+12 -1.97 0.231 2.963E+19 0.301 1.00 2.031E-06 2.073E-61 6.536E+13 102.27
700.00 1.983E+12 3.894E+10 -5.47 2.314E-02 2.332E+18 2.369E-02 1.00 8.705E-07 3.386E-19 4.990E+12 99.83
1100.00 3.309E+11 6.172E+09 -8.70 6.067E-03 6.011E+17 6.104E-03 1.00 5.539E-07 5.263E-08 8.751E+11 96.74
-----------------------------------------------------------------------------------------------------------------------------------

PRESSURE-DEPENDENT BEHAVIOR AS A FUNCTION OF PRESSURE

Reaction # 3
Bath Gas Temperature = 1100.00K
Low Pressure Limiting Reaction Rate = klow = 6.0105E+17 cm^6/(mol^2-s)
-----------------------------------------------------------------------------------------------------------------------------------
Pres k Afact Ea k/kinf klow*EffConc Reduc_Pres FC EffConc k_rev Afact_rev Ea_rev
torr cm^3/(mol-s) kcal/mol cm^3/(mol-s) mole/cm^3 sec-1 kcal/mol
-----------------------------------------------------------------------------------------------------------------------------------
700.00 9.602E+12 2.982E+11 -7.59 0.176 1.165E+13 0.214 1.00 1.939E-05 1.527E-06 4.228E+13 97.86
264.58 4.075E+12 9.340E+10 -8.25 7.471E-02 4.404E+12 8.075E-02 1.00 7.328E-06 6.481E-07 1.324E+13 97.19
100.00 1.615E+12 3.234E+10 -8.55 2.962E-02 1.665E+12 3.052E-02 1.00 2.770E-06 2.569E-07 4.584E+12 96.90
-----------------------------------------------------------------------------------------------------------------------------------

===================================================================================================================================
Gas Reaction # 4 2H+M<=>H2+M
===================================================================================================================================
Change in moles in the reaction = -1.00
This reaction does have third body effects.
1 enhanced third body efficiencies were input
Species "H2" enhanced third body efficiency for the reaction = 0.000
This is a reversible reaction, having 3.00 reactant species and 2.00 product species (including the third body)
k cm^6/(mol^2-s) = 1.0000E+18 T^( -1.000 ) exp( - 0.00 kcal/mol / RT)

HIGH PRESSURE GAS REACTION RATE CONSTANTS AS A FUNCTION OF TEMPERATURE

-----------------------------------------------------------------------------------------------------------------------------------
UnifDimensnl Rate_Const
Temp k Afact Ea DeltaG DeltaH DeltaS k_rev Afact_rev Ea_rev k_star k_star_rev
K cm^6/(mol^2-s) kcal/mol kcal/mol cal/(mol-K) cm^3/(mol-s) kcal/mol mol/(cm^3-s)
-----------------------------------------------------------------------------------------------------------------------------------
298.15 3.354E+15 1.234E+15 -0.59 -97.17 -104.20 -23.57 8.156E-61 2.627E+15 103.01 4.174E-04 9.437E-74
300.00 3.333E+15 1.226E+15 -0.60 -97.13 -104.20 -23.59 2.383E-60 2.619E+15 103.01 4.072E-04 2.723E-73
700.00 1.429E+15 5.255E+14 -1.39 -87.12 -105.38 -26.09 1.576E-17 1.698E+15 102.60 1.374E-05 3.308E-31
1100.00 9.091E+14 3.344E+14 -2.19 -76.40 -106.50 -27.36 6.666E-06 1.301E+15 102.13 2.253E-06 5.666E-20
-----------------------------------------------------------------------------------------------------------------------------------

ANALYSIS OF THIRD BODY REACTIONS: LUMPING [M] WITH RATE CNST

Reaction # 4
Bath Gas Pressure = 20.00 torr
-------------------------------------------------------------------------------------------------------------------------
Temp k Afact Ea Concentration C_eff k_rev Afact_rev Ea_rev
K cm^3/(mol-s) kcal/mol mol/cm^3 mol/cm^3 sec-1 kcal/mol
-------------------------------------------------------------------------------------------------------------------------
298.15 3.6077E+08 4.8825E+07 -1.18 1.0757E-06 1.0756E-07 8.7733E-68 1.0395E+08 102.42
300.00 3.5633E+08 4.8224E+07 -1.19 1.0691E-06 1.0690E-07 2.5474E-67 1.0301E+08 102.41
700.00 6.5449E+07 8.8575E+06 -2.78 4.5818E-07 4.5814E-08 7.2210E-25 2.8611E+07 101.21
1100.00 2.6504E+07 3.5869E+06 -4.37 2.9157E-07 2.9155E-08 1.9435E-13 1.3949E+07 99.94
-------------------------------------------------------------------------------------------------------------------------

===================================================================================================================================
Gas Reaction # 5 2H+H2<=>2H2
===================================================================================================================================
Change in moles in the reaction = -1.00
This reaction does not have any third body effects
This is a reversible reaction, having 3.00 reactant species and 2.00 product species
k cm^6/(mol^2-s) = 9.2000E+16 T^(-0.6000 ) exp( - 0.00 kcal/mol / RT)

HIGH PRESSURE GAS REACTION RATE CONSTANTS AS A FUNCTION OF TEMPERATURE

-----------------------------------------------------------------------------------------------------------------------------------
UnifDimensnl Rate_Const
Temp k Afact Ea DeltaG DeltaH DeltaS k_rev Afact_rev Ea_rev k_star k_star_rev
K cm^6/(mol^2-s) kcal/mol kcal/mol cal/(mol-K) cm^3/(mol-s) kcal/mol mol/(cm^3-s)
-----------------------------------------------------------------------------------------------------------------------------------
298.15 3.014E+15 1.654E+15 -0.36 -97.17 -104.20 -23.57 7.329E-61 3.522E+15 103.25 3.751E-03 8.480E-73
300.00 3.003E+15 1.648E+15 -0.36 -97.13 -104.20 -23.59 2.147E-60 3.520E+15 103.25 3.668E-03 2.453E-72
700.00 1.806E+15 9.912E+14 -0.83 -87.12 -105.38 -26.09 1.993E-17 3.202E+15 103.16 1.737E-04 4.182E-30
1100.00 1.377E+15 7.557E+14 -1.31 -76.40 -106.50 -27.36 1.010E-05 2.939E+15 103.00 3.412E-05 8.583E-19
-----------------------------------------------------------------------------------------------------------------------------------

CK-TUT-10112-1112-UG-1 265 2011 Reaction Design

CHEMKIN Chapter 5: Chemical Mechanism Analysis

The Turn On Surface Reaction Tables button in the Output Control tab gives the following
tables of information about surface reactions in the mechanism. The first two surface reactions
were input as sticking coefficients, and the third reaction can be described that way, so for
these reactions, the Turn On Sticking Coefficient Tables button adds tables of sticking
coefficients as a function of temperature. The first reaction also has a surface coverage
dependence, so the Turn On Surface Coverage Dependence Tables button adds a table of
rate constants that include the effect of the surface coverage at bath-gas conditions for that
reaction.

===================================================================================================================================
Surface Reaction # 1 CH(S)+H<=>C(S,R)+H2
===================================================================================================================================
Change in gas moles in the reaction = 0.00
Change in surface moles in the reaction = 0.00
Change in bulk moles in the reaction = 0.00

This reaction has 1 species whose surface coverage modify the rate constant
Each of these species has three parameters that multiplicatively modify the rate constant as follows:
k_prime = k * 10^(Z_k*nu_ki) * Z_k^mu_ki * exp[ - eps_ki*Z_k / RT ]
where
Z_k = Site Fraction of species k
Species = CH(S)
nu_ki = 0.1000
mu_ki = 0.000
eps_ki = 0.00 kcal/mol

This is a reversible surface reaction, having the following types of reactant species:
1.00 gas-phase species
1.00 surface-phase species
0.00 bulk-phase species

and the following types of product species:

1.00 gas-phase species
1.00 surface-phase species
0.00 bulk-phase species
The reaction rate constant was input via a sticking coefficient in the interpreter input file
Motz-Wise Correction factor is used
Sticking Coeff = MIN(1, 2.140 exp( - 7.30 kcal/mol / RT) )
It can be fit to the following general rate constant form:
k cm^3/(mol-s) = 5.3961E+11 T^( 0.6423 ) exp( - 7.18 kcal/mol / RT)
The reverse rate constant can be fit to the following form:
k(rev) cm^3/(mol-s) = 1.4817E+11 T^( 0.5045 ) exp( - 15.83 kcal/mol / RT)
The reverse rate constant can also be expressed in a sticking coefficient form:
Sticking Coeff(rev) = 0.3030 T^( 4.0223E-03) exp( - 15.83 kcal/mol / RT)

FORWARD AND REVERSE SURFACE REACTION RATE CONSTANTS

-----------------------------------------------------------------------------------------------------------------------------------
Bath Gas Dependent
UnifDimensnl Rate_Const
Temp k Afact Ea DeltaG DeltaH DeltaS k_rev Afact_rev Ea_rev k_star k_star_rev
K cm^3/(mol-s) kcal/mol kcal/mol cal/(mol-K) cm^3/(mol-s) kcal/mol mol/(cm^2-s)
-----------------------------------------------------------------------------------------------------------------------------------
298.15 1.143E+08 4.229E+13 7.60 -9.88 -8.73 3.84 6.56 6.115E+12 16.33 6.672E-07 3.827E-14
300.00 1.237E+08 4.242E+13 7.60 -9.89 -8.73 3.85 7.77 6.098E+12 16.33 7.176E-07 4.508E-14
700.00 2.078E+11 6.713E+13 8.04 -11.70 -8.33 4.82 4.619E+07 5.938E+12 16.36 5.166E-04 1.148E-07
1100.00 1.816E+12 9.631E+13 8.68 -13.58 -8.62 4.51 3.635E+09 9.939E+12 17.30 2.871E-03 5.750E-06
-----------------------------------------------------------------------------------------------------------------------------------

ANALYSIS OF FORWARD AND REVERSE COVERAGE DEPENDENT SURFACE RATE CONSTANTS AT BATH GAS CONDITIONS
-----------------------------------------------------------------------------------------------------------------------------------
Temp k_prime Afact Ea k Cov_fac(cgs) k_rev krev_prime Afact_rev Ea_rev
K cm^3/(mol-s) kcal/mol cm^3/(mol-s) cm^3/(mol-s cm^3/(mol-s) kcal/mol
-----------------------------------------------------------------------------------------------------------------------------------
298.15 1.188E+08 4.394E+13 7.60 1.143E+08 1.04 6.56 6.82 6.355E+12 16.33
300.00 1.286E+08 4.408E+13 7.60 1.237E+08 1.04 7.77 8.08 6.336E+12 16.33
700.00 2.160E+11 6.975E+13 8.04 2.078E+11 1.04 4.619E+07 4.800E+07 6.170E+12 16.36
1100.00 1.887E+12 1.001E+14 8.68 1.816E+12 1.04 3.635E+09 3.778E+09 1.033E+13 17.30
-----------------------------------------------------------------------------------------------------------------------------------

BREAKDOWN OF FORWARD REACTION'S STICKING COEFFICIENT

Surface site density divisor = 5.2200E-09 mol^1.00/cm^2.00
-------------------------------------------------------------------------------------------------------------------------
Temp Stck_Coeff Afact Ea Eff_Vel Vel_Corr Sden_Ratio k* k
K unitless kcal/mol cm/s cm/s cm^3/(mol-s)
-------------------------------------------------------------------------------------------------------------------------
298.15 9.5391E-06 2.140 7.30 6.2563E+04 1.000 1.000 0.5968 1.1433E+08
300.00 1.0292E-05 2.140 7.30 6.2757E+04 1.000 1.000 0.6459 1.2374E+08

2011 Reaction Design 266 CK-TUT-10112-1112-UG-1

Chapter 5: Chemical Mechanism Analysis Tutorials Manual

700.00 1.1254E-02 2.140 7.30 9.5863E+04 1.006 1.000 1085. 2.0785E+11

1100.00 7.5871E-02 2.140 7.30 1.2017E+05 1.039 1.000 9477. 1.8155E+12
-------------------------------------------------------------------------------------------------------------------------

BREAKDOWN OF REVERSE REACTION'S STICKING COEFFICIENT

Surface site density divisor = 5.2200E-09 mol^1.00/cm^2.00
-------------------------------------------------------------------------------------------------------------------------
Temp Stck_Coeff Afact Ea Eff_Vel Vel_Corr Sden_Ratio k* k
K unitless kcal/mol cm/s cm/s cm^3/(mol-s)
-------------------------------------------------------------------------------------------------------------------------
298.15 7.7387E-13 0.4377 16.03 4.4239E+04 1.000 1.000 3.4235E-08 6.558
300.00 9.1439E-13 0.4351 16.03 4.4376E+04 1.000 1.000 4.0577E-08 7.773
700.00 3.5571E-06 0.2773 15.67 6.7786E+04 1.000 1.000 0.2411 4.6192E+07
1100.00 2.2330E-04 0.3700 16.20 8.4974E+04 1.000 1.000 18.98 3.6355E+09
-------------------------------------------------------------------------------------------------------------------------

===================================================================================================================================
Surface Reaction # 2 C(S,R)+H=>CH(S)
===================================================================================================================================
Change in gas moles in the reaction =-1.00
Change in surface moles in the reaction = 0.00
Change in bulk moles in the reaction = 0.00

This is an irreversible surface reaction, having the following types of reactant species:
1.00 gas-phase species
1.00 surface-phase species
0.00 bulk-phase species

and the following types of product species:

0.00 gas-phase species
1.00 surface-phase species
0.00 bulk-phase species
The reaction rate constant was input via a sticking coefficient in the interpreter input file
Motz-Wise Correction factor is used
Sticking Coeff = MIN(1, 0.3000 exp( - 0.00 kcal/mol / RT) )
It can be fit to the following general rate constant form:
k cm^3/(mol-s) = 2.4498E+11 T^( 0.5000 ) exp( - 0.00 kcal/mol / RT)
Even though this reaction is IRREVERSIBLE, the reverse rate constant will also be analysed:
The reverse rate constant can be fit to the following form:
k(rev) 1/s = 9.0322E+10 T^( 1.096 ) exp( - 94.69 kcal/mol / RT)

FORWARD AND REVERSE SURFACE REACTION RATE CONSTANTS

-----------------------------------------------------------------------------------------------------------------------------------
(note: reverse rate constant is not in mechanism) Bath Gas Dependent
UnifDimensnl Rate_Const
Temp k Afact Ea DeltaG DeltaH DeltaS k_rev Afact_rev Ea_rev k_star k_star_rev
K cm^3/(mol-s) kcal/mol kcal/mol cal/(mol-K) 1/s kcal/mol mol/(cm^2-s)
-----------------------------------------------------------------------------------------------------------------------------------
298.15 4.230E+12 6.974E+12 0.30 -87.29 -95.46 -27.41 1.793E-56 1.027E+14 95.17 2.375E-02 9.361E-65
300.00 4.243E+12 6.996E+12 0.30 -87.24 -95.47 -27.44 4.829E-56 1.040E+14 95.17 2.368E-02 2.521E-64
700.00 6.482E+12 1.069E+13 0.70 -75.41 -97.05 -30.91 3.218E-16 3.902E+14 96.36 1.550E-02 1.680E-24
1100.00 8.125E+12 1.340E+13 1.09 -62.82 -97.88 -31.87 2.975E-05 5.048E+14 96.79 1.237E-02 1.553E-13
-----------------------------------------------------------------------------------------------------------------------------------

BREAKDOWN OF FORWARD REACTION'S STICKING COEFFICIENT

Surface site density divisor = 5.2200E-09 mol^1.00/cm^2.00
-------------------------------------------------------------------------------------------------------------------------
Temp Stck_Coeff Afact Ea Eff_Vel Vel_Corr Sden_Ratio k* k
K unitless kcal/mol cm/s cm/s cm^3/(mol-s)
-------------------------------------------------------------------------------------------------------------------------
298.15 0.3000 0.3000 0.00 6.2563E+04 1.176 1.000 2.2081E+04 4.2301E+12
300.00 0.3000 0.3000 0.00 6.2757E+04 1.176 1.000 2.2150E+04 4.2432E+12
700.00 0.3000 0.3000 0.00 9.5863E+04 1.176 1.000 3.3834E+04 6.4816E+12
1100.00 0.3000 0.3000 0.00 1.2017E+05 1.176 1.000 4.2413E+04 8.1252E+12
-------------------------------------------------------------------------------------------------------------------------

===================================================================================================================================
Surface Reaction # 3 C(S,R)+CH3<=>D+CH3(S)
===================================================================================================================================
Change in gas moles in the reaction =-1.00
Change in surface moles in the reaction = 0.00
Change in bulk moles in the reaction = 1.00

This is a reversible surface reaction, having the following types of reactant species:
1.00 gas-phase species
1.00 surface-phase species
0.00 bulk-phase species

and the following types of product species:

0.00 gas-phase species
1.00 surface-phase species
1.00 bulk-phase species
A sticking coefficient was not used though the forward reaction could have used one
k cm^3/(mol-s) = 4.0000E+12 exp( - 1.20 kcal/mol / RT)
The forward rate constant (with Motz-Wise correction) can be fit to the following sticking coefficient expression:
Sticking Coeff = 13.56 T^(-0.4622 ) exp( - 1.08 kcal/mol / RT)
The reverse rate constant can be fit to the following form:
k(rev) 1/s = 1.0000E+13 T^(-8.1728E-12) exp( - 15.00 kcal/mol / RT)
k(rev) 1/s = 1.0000E+13 exp( - 15.00 kcal/mol / RT)

CK-TUT-10112-1112-UG-1 267 2011 Reaction Design

CHEMKIN Chapter 5: Chemical Mechanism Analysis

FORWARD AND REVERSE SURFACE REACTION RATE CONSTANTS

-----------------------------------------------------------------------------------------------------------------------------------
Bath Gas Dependent
UnifDimensnl Rate_Const
Temp k Afact Ea DeltaG DeltaH DeltaS k_rev Afact_rev Ea_rev k_star k_star_rev
K cm^3/(mol-s) kcal/mol kcal/mol cal/(mol-K) 1/s kcal/mol mol/(cm^2-s)
-----------------------------------------------------------------------------------------------------------------------------------
298.15 5.278E+11 4.000E+12 1.20 -47.78 -60.43 -42.43 101. 1.000E+13 15.00 2.964E-03 5.280E-07
300.00 5.344E+11 4.000E+12 1.20 -47.70 -60.44 -42.46 118. 1.000E+13 15.00 2.982E-03 6.172E-07
700.00 1.688E+12 4.000E+12 1.20 -29.94 -61.58 -45.19 2.074E+08 1.000E+13 15.00 4.038E-03 1.08
1100.00 2.310E+12 4.000E+12 1.20 -11.90 -61.29 -44.90 1.047E+10 1.000E+13 15.00 3.516E-03 54.6
-----------------------------------------------------------------------------------------------------------------------------------

BREAKDOWN OF FORWARD REACTION'S STICKING COEFFICIENT

Surface site density divisor = 5.2200E-09 mol^1.00/cm^2.00
-------------------------------------------------------------------------------------------------------------------------
Temp Stck_Coeff Afact Ea Eff_Vel Vel_Corr Sden_Ratio k* k
K unitless kcal/mol cm/s cm/s cm^3/(mol-s)
-------------------------------------------------------------------------------------------------------------------------
298.15 0.1567 0.6393 0.83 1.6199E+04 1.085 1.000 2755. 5.2779E+11
300.00 0.1581 0.6368 0.83 1.6249E+04 1.086 1.000 2790. 5.3442E+11
700.00 0.3015 0.4103 0.43 2.4821E+04 1.178 1.000 8812. 1.6882E+12
1100.00 0.3247 0.3382 0.09 3.1115E+04 1.194 1.000 1.2059E+04 2.3102E+12
-------------------------------------------------------------------------------------------------------------------------

===================================================================================================================================
Surface Reaction # 4 CH2(S,R)+CH(S,R)<=>CH2(S)+CH(S)
===================================================================================================================================
Change in gas moles in the reaction = 0.00
Change in surface moles in the reaction = 0.00
Change in bulk moles in the reaction = 0.00

This is a reversible surface reaction, having the following types of reactant species:
0.00 gas-phase species
2.00 surface-phase species
0.00 bulk-phase species

and the following types of product species:

0.00 gas-phase species
2.00 surface-phase species
0.00 bulk-phase species
k cm^2/(mol-s) = 6.0000E+19 exp( - 2.00 kcal/mol / RT)
The reverse rate constant can be fit to the following form:
k(rev) cm^2/(mol-s) = 5.9991E+19 T^( 2.0901E-05) exp( - 96.02 kcal/mol / RT)

FORWARD AND REVERSE SURFACE REACTION RATE CONSTANTS

-----------------------------------------------------------------------------------------------------------------------------------
Bath Gas Dependent
UnifDimensnl Rate_Const
Temp k Afact Ea DeltaG DeltaH DeltaS k_rev Afact_rev Ea_rev k_star k_star_rev
K cm^2/(mol-s) kcal/mol kcal/mol cal/(mol-K) cm^2/(mol-s) kcal/mol mol/(cm^2-s)
-----------------------------------------------------------------------------------------------------------------------------------
298.15 2.052E+18 6.000E+19 2.00 -94.03 -94.03 0.00 2.436E-51 6.000E+19 96.03 55.9 6.637E-68
300.00 2.095E+18 6.000E+19 2.00 -94.03 -94.03 0.00 6.618E-51 6.000E+19 96.03 57.1 1.803E-67
700.00 1.425E+19 6.000E+19 2.00 -94.03 -94.03 0.00 6.257E-11 6.000E+19 96.03 388. 1.705E-27
1100.00 2.403E+19 6.000E+19 2.00 -94.03 -94.03 0.00 5.00 6.000E+19 96.03 655. 1.362E-16
-----------------------------------------------------------------------------------------------------------------------

2011 Reaction Design 268 CK-TUT-10112-1112-UG-1

 CHEMKIN

Index

Index
A ethylene/air combustion 175
fluorocarbon plasma 251
adiabatic flame temperature 23
GRImech 3.0 174
A-factors 201
hydrogen/air 173
aftertreatment 181
methane air 174
aftertreatment, engine exhaust 191
methane oxidation on Pt 204
aggregation 162
methane/air 174
simple model 159
methane/air reduced mechanism 174
ALD 226
propane/air 175
ALD processes 226
Pt/Rh three-way catalyst 205
alumina 248
reduced mechanism 174
alumina, ALD of 248
silicon deposition from silane 247
ammonia 246
silicon deposition from trichlorosilane 247
approximations, cylindrical channel flow 214
silicon nitride CVD 246
Arrhenius pre-exponential factors 201
CHEMKIN
atomic layer deposition 226, 248
format 110
autoignition for hydrogen/air 30
CHEMKIN-PRO
autoignition, ignition times for propane 33
reactor models 19
axisymmetric results visualization 138
chlorine plasma 250
spatial 237
B steady-state 233
bath gas 254, 255 chlorosilane CVD 208
burner-stablilized laminar premixed flame 38 closed homogeneous panel 30, 34
coefficients, normalized sensitivity 83
colorbar property panel 139
C combustion 142
C1_Inlet1 panel 40 combustion in complex flows 111
calculation, equilibrium 23, 208 combustion, hydrogen 28
carrier gas 254, 256 combustion, propane 33
catalytic combustors 181 combustion, steady-state gas-phase 28
catalytic converters 181 combustor, gas turbine 111
catalytic oxidation 200 combustor, homogeneous 186
Center for Energy Research 175 combustor, two stage catalytic 181
Chapmen-Jouguet detonation 111 continuations panel 24, 29, 208, 211, 216, 219, 242
chemical mechanism analysis 253 CRUPROF subroutine 137
chemical vapor deposition 207 Curls model, modified 129
chemistry sets 173, 204, 246 CVD 207
alumnia ALD 248 chlorosilane 208
chlorine plasma 250 diamond 256

CK-TUT-10112-1112-UG-1 269 2011 Reaction Design

CHEMKIN Index

steady-state thermal 210 methane/air premixed 55

trichlorosilane 223 methane-air 86
CVD reactor 210 non-premixed jet 114
cylindrical channel flow 214 opposed-flow diffusion 81
cylindrical shear flow panel 215 stagnation 94
cylindrical shear flow reactor 207, 214 Flame-speed Calculator
Particle Tracking Feature 147
flammability 28
D fluorocarbon plasma 251
Damkhler number 255, 256 etching 241
DDASPK solver 129 FORTRAN subroutine 192
decomposition pathway fusion 162
Reaction Path Analyzer 66
deposition in a rotating disk 219
detonation, Chapmen-Jouguet 111 G
diagnostic output file 258 gas composition
diamond CVD 256 estimating initial with EGR
DVODE solver 129 gas inlet panel 228
gas radiation for burner burner-stabilized flame 41
Gas Research Institute 174
E gas turbine combustor 111
EGR gas turbine network 111
estimating initial gas composition gas, bath 254, 255
EGR project gas, carrier 254, 256
results 125 gas-phase combustion, steady-state 28
element filter GRI 3.0 methane oxidation mechanism 120
Reaction Path Analyzer 66 GRImech 3.0 174
engine exhaust aftertreatment, transient flow 191
engineout 193
equilibrium analysis 26, 208 H
equilibrium calculation 23, 208 HCCI 97
equilibrium panel 24, 208 HCCI engine
equivalence ratio 23, 71 estimating initial gas composition
etching, fluorocarbon plasma 241 multi-zone 103
etching, plasma 233 with exhaust gas recirculation 119
exhaust gas recirculation. See EGR. HCCI_EGR project 119
extinction heat-transfer correlation 99
flames 85 homogeneous charge compression ignition 97
Extinction of Premixed or Opposed Flow Flame reactor homogeneous combustor 186
85 honeycomb monolith panel 185
honeycomb monolith reactor 184
hydrogen combustion 28
F hydrogen/air chemistry set 173
files hydrogen/air flame 81
diagnostic output 258 hydrogen/air system 23
XML solution 31, 211, 257, 258 hydrogen/air, autoignition for 30
flame speed 71
flame speed panel 55
flame speed, laminar 55 I
flames 28 IC engine panel 98
burner-stabilized premixed laminar 38 IC engine, single-zone 97
burner-stabilized with gas radiation 41 ic_engine__EGR_network.ckprj 120
extinction 85 ignition 28
hydrogen/air 81 ignition time, defining 34

2011 Reaction Design 270 CK-TUT-10112-1112-UG-1

Index Tutorials Manual

ignition times for propane autoignition 33 output control panel 29, 31, 82, 211, 215, 242, 257, 261,
internal combustion engine 97 262

J P
jet flame network 114 panel
JSR/PFR system , soot formation 141 reactor properties 129
panels
C1_Inlet1 40
L closed homogeneous 30, 34
laminar flame speed 55 continuation 24, 29, 208, 211, 216, 219, 242
low-pressure plasma reactor 233 cylindrical shear flow 215
equilibrium 24, 208
flame speed 55
M gas inlet 228
mass flow rate ratio 122 honeycomb monolith 185
mechanism analysis, chemical 253 IC engine 98
mechanism analyzer panel 258 mechanism analyzer 258
mechanism analyzer tool 253 opposed-flow flame 82
methane oxidation on Pt 204 output control 29, 31, 82, 211, 215, 242, 257, 261,
methane/air chemistry set 174 262
methane/air, partially stirred reactor for 127 PFR 215
modified Curl's model 129 planar shear flow 223
Monte Carlo mixing model 128 plasma PFR 238
multiple reactor models 227 plasma PSR 234, 242
multi-zone HCCI engine 103 preferences 136
multi-zone model 104 pre-mixed burner 38, 39
icon 104 PSR 28, 193, 211, 215, 228
solution file, reduce size 105 R1_IN 55
multi-zone project recycling 116
number of reactors 104 rotating disk 219
solver 29, 40, 82, 193, 215, 223, 229, 234, 242
N stagnation flow 228
panels (Post-Processor)
networks
colorbar property (Post-Processor) 139
gas turbine 111
Parameter Study Facility 71, 200
jet flame 114
partially stirred reactor 127
PSR 112, 115, 215
particle aggregation 162
PSR-PFR hybrid 112
Particle Tracking Feature
non-dimensional reaction rate 254
Flame-speed Calculator 147
non-premixed jet flame 114
Opposed-flow Flame Simulator 147
normalized sensitivity coefficients 83
Particle Tracking feature
NOx-mutual sensitization 119
aggregation 162
number, Damkhler 255, 256
particle-size distribution 153
PaSR 128
O perfectly stirred reactor 183, 210, 214, 227, 233
opposed-flow diffusion flame 81 PFR panel 215
opposed-flow flame planar channel flow 223
soot and radiation 47 planar shear flow panel 223
opposed-flow flame panel 82 planar shear flow reactor 207
Opposed-flow Flame Simulator plasma chlorine 233
Particle Tracking Feature 147 plasma etching 233
opposed-flow flames silicon dioxide 241
extinction 85 plasma PFR panel 238

CK-TUT-10112-1112-UG-1 271 2011 Reaction Design

CHEMKIN Index

plasma plug flow reactor 237 Y variable 62

plasma PSR panel 234, 242 zoom 65
plasma, fluorocarbon 251 reaction paths 59
plug flow reaction rates 254
side inlet 132 reactor network
plug flow reactor 183, 184, 207, 214 recirculation 119
Post 131 tear stream 119
power profile, with spatial chlorine plasma 237 reactor properties panel 129
preferences panel 136 recycling panel 116
Preferences tab reduced mechanism, methane/air 174
Reaction Path Analyzer 65 relative rate analysis
pre-mixed burner panel 38, 39 Reaction Path Analyzer 68
premixed flames ROP 210
extinction 85 rotating disk panel 219
stagnation 94 rotating disk reactor 207, 219
Pre-mixed Laminar Burner-Stabilized Stagnation RPA 59
Flame 153
probability density functions 129
Progress Monitor S
EGR_network project 125 sample problems, models used 20
propane sectional method
ignition times for autoignition 33 aggregation model 159
propane-air mixtures 71 sectional model 153
PSR analysis 210 segregated solution technique 147
PSR network 112, 115 Sensitivity
PSR panel 28, 193, 211, 215, 228 Reaction Path Analyzer 67
PSR-PFR network 111 sensitivity
Pt/Rh three-way catalyst 205 Reaction Path Analyzer 60
shock-heated air 110
shock-tube experiment 33, 110
R side inlet on plug flow reactor 132
R1_IN panel 55 silane 247
radiation silicon deposition from silane 247
opposed-flow flame with soot 47 silicon deposition from trichlorosilane 247
rate constant, uniform-dimensional 255 silicon dioxide 251
Rate of Production silicon nitride CVD 246
Reaction Path Analyzer 67 silicon tetrafluoride 246
rate-of-production analysis 236 single-zone IC engine 97
rate-of-production options 210 solution file
reaction mechanism for diamond CVD 256 Reaction Path Analyzer point selection 62
Reaction Path Analyzer 55, 59 zone average option for multi-zone 105
colors 71 solver panel 29, 40, 82, 193, 215, 223, 229, 234, 242
decomposition pathway 66 solvers
element filter 66 DDASPK 129
Rate of Production 67 DVODE 129
Reaction Path Analyzer 65 transient 192
reduce diagram complexity 69 soot
relative rate analysis 68 opposed-flow flame and radiation 47
Sensitivity 67 soot formation 141
solution file point selection 62 soot nucleation reaction 142
species, isolated 70 spatial chlorine plasma 237
species, number of 66 species, number of
split paths 68 Reaction Path Analyzer 66
tree layout 68 splits paths

2011 Reaction Design 272 CK-TUT-10112-1112-UG-1

Index Tutorials Manual

Reaction Path Analyzer 68 X

Splitter node, Project tree 122
XML solution file 31, 211, 257, 258
stagnation
XMLdata.zip 129, 257
flames 94
stagnation flow panel 228
stagnation flow reactor 207 Y
steady-state gas-phase combustion 28 Y variable
steady-state thermal CVD 210 Reaction Path Analyzer 62
stoichiometric hydrogen/air mixture 30
stoichiometric methane/air premixed flame 55
sub-projects 227 Z
subroutine library zone-average option for multi-zone solution file 105
TRANSPORT 256 zones, multiple 104
surface chemistry analysis 200 zoom
Reaction Path Analyzer 65
T
tear stream 120
Tear Stream Controls node, Project tree 124
tear stream reactor network
temperature, adiabatic flame 23
time-dependent simulations of ALD process 227
titania 162
titanium tetrachloride 162
TMA 248
transient inlet flow 191
transient solver 192
TRANSPORT subroutine library 256
tree layout
Reaction Path Analyzer 68
trichlorosilane CVD 223
trichlorosilane, silicon deposition from 247
trimethylaluminum 248
TRIT 181
turbine rotor inlet temperature 181
two stage catalytic combustor 181

U
unburnt gas temperature 71
uniform-dimensional rate constant 255
uniform-dimensional reaction rate 254
UPROF keyword 137
user-defined subroutine 191
USRINLET 192
USRINLET subroutine 194

V
vapor deposition, chemical 207

W
Woschni correction 99

CK-TUT-10112-1112-UG-1 273 2011 Reaction Design

CHEMKIN Index

2011 Reaction Design 274 CK-TUT-10112-1112-UG-1