Introducing CHEMKIN 4.
1
Ellen Meeks CHEMKIN in Combustion Workshop August 6, 2006
Outline
Brief Introduction What is CHEMKIN? CHEMKIN 4.1 released July 2006
All New Graphical Post-processing Parameter-study Facility Particle Tracking Module (optional) More Complex Reactor Networks New Chemistry Formulations Reactor Model Enhancements
Future projects underway
Introduction
CHEMKIN software allows focus on chemistry using engineering approximations
Plug-flow Reactor (PFR) Perfectly Stirred Reactor (PSR) Partially Stirred Reactor
T v T v
Shear-layer Channel-flow Reactor Opposed-flow Diffusion Flame
N H2 CH2 O2
4
Pre-mixed Flame
N 2 H2OCO CO O
Flame-speed Calculation
Equilibrium Calculator
The CHEMKIN approach provides efficient solutions with very short set-up time
Supports large chemical kinetic mechanisms
100s of species 1000s of reactions
Provides accurate information about reacting system
Trace species emissions
(parts per billion or smaller)
Accurate handling of kinetics time and length scales Complex dependency between major & minor species Dominant reaction paths Sensitivity of results to reaction parameters de facto standard for describing elementary kinetics
Computational time for single reactor ~ seconds-minutes
PC , Linux, UNIX
5
Detailed kinetics are critical for accurately modeling chemical systems
Exhaust-gas emissions Ignition phenomena Engine knock Flame-speed Flammability limits Catalytic lightoff Chemical conversion Fuel reforming Material deposition Plasma processing
6
Using CHEMKIN requires a chemistry set, reactor definition, and operating conditions
Choose a reactor model Define the geometry Specify operating conditions
Define or select a chemistry set Thermodynamic data Gas-phase reactions Surface reactions Transport-property data
Reactor models predict the varying state of the gas and/or surface
Thermodynamic state of the gas
Temperature / Enthalpy / Entropy Pressure / Density Rate of heat transfer to the environment
Chemical state of gas / surface / solid
Gas-phase composition Surface site coverage Instantaneous deposition or etch rates Species production rates Rate of progress of individual reactions
Flow state
Convective velocities Diffusive velocities Stefan velocity due to deposition
Predicted as a function of Time Distance Zone Parameter
Solution sensitivity to parameters
8
Reactor networks are easily constructed using a visual diagram tool
Customizable palette of reactor modules Drag and drop diagrams for reactor networks and series simulations
Reactor templates guide user inputs for problem-specific applications
Reactor Models include:
Non-reactive Gas Mixer Chemical and Phase Equilibrium Calculations Mechanism Analyzer Closed Internal Combustion Engine Simulator Closed Homogeneous Batch Reactor Closed Partially Stirred Reactor Closed Plasma Reactor Perfectly Stirred Reactor (PSR) Plasma PSR Partially Stirred Reactor (PaSR) Plug Flow Reactor Honeycomb Plug Flow Reactor Plasma Plug Flow Reactor Planar Shear Flow Reactor Cylindrical Shear Flow Reactor Premixed Laminar Burner-stabilized Flame Premixed Laminar Flame-speed Calculation Diffusion or Premixed Opposed-flow Flame Stagnation Flow CVD Reactor Rotating Disk CVD Reactor Normal Incident Shock
10
Descriptions of reactor configurations are through context-specific panels
Project tree with links to input panels
Input panels guide problem set-up and execution
11
Users can also create their own Applications through use of the CHEMKIN/API
Subroutine Libraries are called from user programs
C/C++ or Fortran interfaces Functions for determining chemical terms and properties
Thermodynamic properties Gas-phase kinetics Surface kinetics Transport properties
Numerical solvers for transient and steady-state
Detailed examples provided in standard install
C++ and Fortran programs and makefiles Command-line / Scripting examples Extracting data from XML solution data file
12
Summary of CHEMKIN Introduction
Chemistry simulation is increasingly a key to design CHEMKINs unique capabilities allow accurate kinetics
13
State-of-the art reaction-rate formulations Advanced surface chemistry treatment Elementary or global kinetics expressions Multicomponent transport properties Solution algorithms designed to dynamically adapt to stiff chemistry Fast set-up and run time Visual diagramming of reactor networks Extensive graphical presentation of results Problem-oriented reactor models Application-oriented tutorials and sample projects API for advanced programming and command-line operation
CHEMKIN provides flexible, easy-to-use interfaces
CHEMKIN 4.1 Release Important New Capabilities
Several all-new capabilities are available with the release of CHEMKIN 4.1
New Visualizer / Post-processor New Parameter Study Facility New, advanced Particle Tracking Module (option)
15
The new Graphical Post-processor takes CHEMKIN visualization to the next level
16
New plot types expand visualization options
3-D Contour Plots
Interactive rotation
Mirrored Plots
Axes of symmetry
Multiple Y-axes
17
Users have full interactive control over formatting and presentation
Plot format
Color, line type, thickness Symbol on/off, shape, size, color Axis line, label, text font, ticks Legend text font, spacing, outline Background color, transparency
Annotation insertions
Text boxes Drawing, shapes, arrows
Easy scaling of axes Sequential Undo (Cntrl-Z), for all operations!
18
Extensive save / reapply options reduce time needed for repetitive tasks
Save / Reapply plot-set selections
Selections can be saved on a project-by-project basis
Save and apply user-defined plot styles Record and apply macros for other operations
Mouse clicks and key strokes can be recorded
19
The Parameter-Study Facility automates a series of simulations
Choose one or more user-specified input(s) to vary
User-interface panel inputs Chemistry-set parameters
Variations are controlled by:
Tables of discrete values Specification of Min, Max, Increment Specification of Min, Max, # of runs Linear or Logarithmic interpolation
Available for multiple-cluster reactor networks
Allows system studies Shows effects of upstream variation on downstream results
Complements existing continuation capability
Each cluster model can contain continuations Each run starts from user-specified initial conditions and generates a separate solution file
20
Visualization of parameter-study results includes added parametric dimension
Solutions are harvested from multiple runs
User selects variables for post-processing Data is extracted out of multiple solution files 2-D plot sets automatically constructed
Vary engine compression ratio
21
Multiphase CHEMKIN: the new Particle Tracking Module (PTM)
Nucleation of particles from gas-phase species
e.g., poly-aromatic hydrocarbons (PAHs) for soot Rates controlled by collision frequencies
Growth by coagulation of particles
Collision-regime options
Growth and reduction by general surface reactions
Vapor-deposition type reactions Condensation of large species
Tracking of particles
Moment Method calculates moments of size distribution Provides particle mean diameter, mass & volume fractions Average surface state and total surface area
22
The PTM can be used with any mechanism for particle growth and reduction
Coagulation
Particle collision frequency
Nucleation reactions
Precursor collision frequency
R0 R1
G0 G1
S0 S1
Surface reactions
Oxidation Growth Condensation
DM r = Rr + Gr + S r Dt
Moment Equations
23
Particle nucleation and growth/reduction are described within the surface chemistry input
MATERIAL SOOT_PARTICLE DISPERSED END SITE/PolyC/ SDEN/3.341E-9/ H(se) open(se) ! se indicates edge or active site. END BULK C(B) /1.8/ END REACTIONS ! nucleation from PAH 2A4 => 32C(B)+20H(se)+28.72open(se) 6.0E10 0.5 0.0 NUCL ! HACA growth H+H(se) => open(se)+H2 4.2E13 0.0 13000.0 open(se)+H2 => H(se)+H 3.9E12 0.0 9320.0 open(se)+H => H(se) 2.0E13 0.0 0.0 H(se)+OH => H2O+open(se) 1.0E10 0.734 1430.0 H2O+open(se) => OH + H(se) 3.68E8 1.139 17100.0 open(se)+C2H2 => H(se)+2C(B)+H 8.0E7 1.56 3800.0 ! PAH condensation C6H6+open(se) => H(se)+6C(B)+5H 0.1 0.0 0.0 STICK FORD/open(se) 2.0/ DCOL/4.E-8/
Indicate a dispersed phase and activates the soot module Precursors for nucleation Inception particle class Particle bulk composition Nucleation reaction
Grow particle by 2 classes Grow particle by 6 classes Sticking coefficient Collision diameter for benzene
24
The PTM is available for use with All Homogeneous 0-D and Flow Reactors
Closed Homogeneous Reactors Perfectly Stirred Reactors Plug Flow Reactors Shear Layer Flow Reactors
25
We have applied the new PTM in soot particle formation studies
Reaction Mechanisms Tested: Ethylene combustion, which determines PAH precursors
From Appel, Bockhorn, Frenklach
Nucleation by single path
Particle Concentration (gm/cc)
MIT JSR/PFR Experiments with Ethylene (Marr, 1993)
2.5E-07 2.0E-07 1.5E-07 1.0E-07 5.0E-08 0.0E+00 0 5 10 15 PFR Residence Time (ms) 20 Expt: soot Expt: tar Expt: sum Simulation
Pyrene (A4) dimerization From Frenklach
HACA growth by C2H2 addition
From Frenklach and Wang
T=1620 K
PAH condensation Particle coagulation Soot oxidation
26
The model allows differentiation of contributions from mechanism components
HACA cannot account for observed growth PAH condensation important for growth Large PAHs are under-predicted by gas mechanism
10
0
2.5E-07
10-1
-2
H2
Soot Mass Concentration (gm/cm3) 2E-07
Data (Marr 1993) HACA+PAH HACA only
10 Mole Fraction
C2H 2
-3
10
1.5E-07
Benzene
-4
10
Phenyl Acetylene
-5
1E-07
10
Naphthalene Pyrene
5E-08
10
-6
T =1620 K P = 1 atm =2.2
0 5 10 PFR Residence Time (ms) 15 20
10
-7
10 PFR Residence Time (ms)
15
20
27
Users have full control over the particletracking simulation from the user interface
Dispersed phases are identified in the Surface Kinetics Input file User options include
Initial particle density Coagulation parameters Solver controls
28
Many other important enhancements are introduced in CHEMKIN 4.1 Some Highlights:
Expansion of reactor-networking capability New reaction-rate formulations Improved solver convergence Advanced ignition-delay calculations Entrainment flow-rate profile for Plug Flow Multiple materials option for Shear Layer Flow Improved consistency between input panels Easier-to-use diagramming and file browsing Ability to plot ignition delay & flame-speed in parameter studies
29
Enhancements to reactor networking add flexibility in describing complex systems
Splitter allows flow splits to multiple reactors Parallel streams enabled Stored file icon allows initialization from previously saved solution Visual emphasis on flow streams
30
New chemistry formulations make it easier to transcribe and estimate rates
Collision probability for gas-phase reactions
Rate based on collision frequency User specifies reaction probability between 0 and 1
New surface-reaction UNITS options
Facilitates direct use of published catalysis data Partial-pressure units for gas-phase species Site-fraction (coverage) units for surface species
Dependence of surface species heats of formation on local coverage
Facilitates estimation of catalysis rates using BondOrder-Conservation methods Improved thermodynamic consistency
31
Several options allow more accurate prediction of ignition delay
Choices for ignition criteria:
Specified temperature threshold for ignition Specified T above initial temperature Peak dT/dt (inflection point) Peak concentration of any species (e.g., OH) User-defined criteria in user subroutine
32
Propane / air ignition delay
Flow-rate profile option for PFRs allows gradual entrainment of gases
CHEMKIN Profile Tool
Additional inlets in PFR allow varying of flow rate vs. distance Example for CH4/NH3 oxidation
0.2% CH4, 0.06% NH3 in N2
8% O2, 4% H2O in N2
Temperature varied: 1000-1500K
Based on Experiments by Grcar,J.F., Glarborg, P., Bella, J.B., Day, M.S., Loren, A., Jensen, A.D. Proceedings of the Combustion Institute, 30 (2004)
33
Summary CHEMKIN 4.1 saves you time
Parameter study More robust solutions Save/restore & macros for plotting Advanced visualization Easily plot single values from parameter study of 1-D simulations (e.g., flame speed, ignition) Improved diagramming 64-bit LINUX now available for AMD/Intel chips
34
Projects underway for future releases
A new Uncertainty Module is under development for use with CHEMKIN
Currently in Alpha Testing Scheduled for release in Q1 2007 Plug-in module to CHEMKIN (4.1 or Later) Based on polynomial chaos expansion theory
10-1000 X faster than Monte Carlo approaches User Specifies
Variant Inputs
P(k1) k1 P(km) km
36
Uncertainty Module Predicts
CHEMKIN Model y1 = f (k1,k2,,km) y2 = f (k1,k2,,km) Variant Outputs
P(y1) y1 P(y2)
...
y2
Other major projects
Implementation of tear-stream algorithm for heterogeneous reactor networks Extraction of reactor network from CFD solutions Particle Tracking Module option added to Flame Simulations Advanced mechanism-analysis tools Further improvements in solver speed and robustness
37
Discussion
Two years ago, we launched the Teaching Forum on our website
How many have used accessed the information there? Is the site useful? Do you plan to contribute a problem to the site? What would you like to see improved or changed?
39
What technology enhancements would you find helpful to your work?
New reactor models? New reaction rate formulations? More user-routine capabilities? Extended reactor connectivity options? Optimization or other add-on capability? Other thoughts or comments?
40
What aspects of CHEMKIN do you find the most difficult?
Flame-code convergence? Finding / accessing chemistry data? Manipulating solution data? Linking to the API? Other thoughts / comments?
41
Of the recent enhancements, what do you find to be the most valuable?
?
42
CHEMKIN 4.1 Demo