ANSYS Mechanism Reduction Best Practices

ANSYS, Inc. Release 2020 R1

Southpointe January 2020
2600 ANSYS Drive
Canonsburg, PA 15317 ANSYS, Inc. and
[email protected] ANSYS Europe,
Ltd. are UL
http://www.ansys.com registered ISO
(T) 724-746-3304 9001: 2015
(F) 724-514-9494 companies.
Copyright and Trademark Information

© 2020 ANSYS, Inc. Unauthorized use, distribution or duplication is prohibited.

ANSYS, ANSYS Workbench, AUTODYN, CFX, FLUENT and any and all ANSYS, Inc. brand, product, service and feature
names, logos and slogans are registered trademarks or trademarks of ANSYS, Inc. or its subsidiaries located in the
United States or other countries. ICEM CFD is a trademark used by ANSYS, Inc. under license. CFX is a trademark
of Sony Corporation in Japan. All other brand, product, service and feature names or trademarks are the property
of their respective owners. FLEXlm and FLEXnet are trademarks of Flexera Software LLC.

Disclaimer Notice

THIS ANSYS SOFTWARE PRODUCT AND PROGRAM DOCUMENTATION INCLUDE TRADE SECRETS AND ARE CONFID-
ENTIAL AND PROPRIETARY PRODUCTS OF ANSYS, INC., ITS SUBSIDIARIES, OR LICENSORS. The software products
and documentation are furnished by ANSYS, Inc., its subsidiaries, or affiliates under a software license agreement
that contains provisions concerning non-disclosure, copying, length and nature of use, compliance with exporting
laws, warranties, disclaimers, limitations of liability, and remedies, and other provisions. The software products
and documentation may be used, disclosed, transferred, or copied only in accordance with the terms and conditions
of that software license agreement.

ANSYS, Inc. and ANSYS Europe, Ltd. are UL registered ISO 9001: 2015 companies.

U.S. Government Rights

For U.S. Government users, except as specifically granted by the ANSYS, Inc. software license agreement, the use,
duplication, or disclosure by the United States Government is subject to restrictions stated in the ANSYS, Inc.
software license agreement and FAR 12.212 (for non-DOD licenses).

Third-Party Software

See the legal information in the product help files for the complete Legal Notice for ANSYS proprietary software
and third-party software. If you are unable to access the Legal Notice, contact ANSYS, Inc.

Published in the U.S.A.

Table of Contents
1. Mechanism Reduction Best Practices Introduction ................................................................................ 1
1.1. Prerequisite Experience .................................................................................................................... 1
1.2. Introduction to Reaction Mechanism Reduction ................................................................................ 1
2. Best Practices for Mechanism Reduction in Engine Applications .......................................................... 3
2.1. Examples of Various Engine Applications ........................................................................................... 3
2.2. Condition Coverage During Mechanism Reduction ............................................................................ 4
2.3. Steps in Setting Up a Mechanism Reduction ...................................................................................... 4
2.4. Chemkin-Pro Project With Parameter Study for Basis of Reduction ..................................................... 5
2.5. Setting Up a Reduction Operation Project in Reaction Workbench ..................................................... 7
2.6. Continuing or Restarting a Reduction Run ...................................................................................... 10

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. iii
 Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
iv of ANSYS, Inc. and its subsidiaries and affiliates.
List of Figures
1.1. Mechanism reduction iterative process ................................................................................................... 1
2.1. Mechanism Reduction session with typical operations ............................................................................ 8

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. v
 Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
vi of ANSYS, Inc. and its subsidiaries and affiliates.
List of Tables
2.1. Recommended ranges for tolerances in reduced mechanisms ................................................................. 9
2.2. Recommended tolerances for soot targets in reduced mechanisms ......................................................... 9

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. vii
 Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
viii of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 1: Mechanism Reduction Best Practices Introduction
This manual describes the prerequisites, input requirements, appropriate scenarios, methodology, and
recommended settings for reducing mechanisms for engine applications when using ANSYS Chemkin-
Pro Reaction Workbench.

1.1. Prerequisite Experience

You should know how to accomplish these goals in Chemkin-Pro :

• How to set up a Closed Homogeneous Reactor project.

• How to set up a parameter study. Parameter Studies are explained in the Chemkin-Pro Advanced Analysis
Guide. We also recommend going through the mechanism reduction tutorial in the Reaction Workbench
Tutorial Guide to become familiar with the mechanism reduction facility in Chemkin-Pro Reaction Workbench.

1.2. Introduction to Reaction Mechanism Reduction

The mechanism reduction capability in Reaction Workbench relies on ANSYS Chemkin-Pro to compute
the quantities of interest, such as ignition time, flame speeds, and species concentrations. This requires
choosing and setting up an appropriate reactor model in Chemkin-Pro that most closely describes the
kinetics phenomena occurring in your engine and that is able to capture one or many predicted variables
of interest. Refer to the tutorial, Mechanism Reduction for Compression Ignition Conditions, in the Re-
action Workbench Tutorial Guide for more information. The figure Figure 1.1: Mechanism reduction iter-
ative process (p. 1) shows a schematic overview of the process.

Figure 1.1: Mechanism reduction iterative process

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 1
 Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
2 of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 2: Best Practices for Mechanism Reduction in Engine
Applications
These best practices for mechanism reduction are designed for combustion applications.

To accomplish an effective mechanism reduction for engine combustion, you first must know the fol-
lowing:

• What is the desired purpose of the reduced mechanism:

– Flame speed or ignition time?

– Which emissions quantities need to be captured accurately? NOx? CO? UHC? Soot?

• The model fuel composition:

– Typically, a surrogate for liquid fuels (see the Reaction Workbench tutorial on using the Surrogate Blend
Optimizer).

– May be the actual or an approximate (surrogate) composition for gaseous fuels.

• Engine operating conditions, including:

– Estimated ranges of temperatures, pressures, EGR levels, and equivalence ratios in the combustion
chamber.

– For IC engines, these are the estimated local conditions in the cylinder, prior to ignition (whether the ig-
nition is by compression, spark or knock event); these are not the peak pressures /temperatures achieved
by ignition. For flame propagation in spark engines, these are the conditions of the unburned mixture
prior to the spark event.

2.1. Examples of Various Engine Applications

1. Ignition in diesel engines, HCCI or other compression-ignition controlled engines: Combustion in compres-
sion ignition engines is governed by autoignition of fuels. Autoignition can be simulated using the Closed
Homogeneous Batch reactor model in Chemkin-Pro.

Note:

Batch reactor template: For ease of mechanism reduction setup, a batch reactor
template has been created that can be used for mechanism reduction. While setting
up a reactor model as part of mechanism reduction, the option to use the batch re-
actor template is presented. This template has pre-defined most of the reactor settings
so that you only need to define the operating conditions (temperature, pressure,
and species composition of the reactor). The batch reactor template is usually the
most convenient way to set up a mechanism reduction operation. On the other

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 3
Best Practices for Mechanism Reduction in Engine Applications

hand, if more flexibility is needed to set up the model, the template option can be
declined and a Chemkin-Pro model can be set up in its entirety.

2. Knock in spark-ignited engines: Knocking occurs due to autoignition of end gases after spark initiation.
Autoignition can be simulated using the Closed Homogeneous Batch reactor template in Chemkin-Pro.

3. Flame propagation in spark-ignited engines: Turbulent flame speeds in CFD are typically computed from
laminar flame speeds of the unburned fuel-air mixture. Laminar flame speeds could be simulated using
the Flame Speed Calculator model in Chemkin-Pro. Although the Flame Speed Calculator could be used,
the run times may be too long for the iterative mechanism reduction process when the master mechanism
contains 1000 or more species. We therefore recommend the alternative of using the Closed Homogeneous
reactor template for mechanism reductions for flame-speed applications. There is no loss in accuracy, because
the underlying kinetics that govern flame speeds are captured accurately using the Closed Homogeneous
reactor as long as a parameter study is used that includes the range of high temperatures expected in the
flames. Section Chemkin-Pro Project With Parameter Study for Basis of Reduction (p. 5) discusses the
appropriate range of conditions to employ in the Closed Homogeneous reactor when it is used for flame-
speed calculations.

4. Emissions: In any engine, all emissions, including NOx, CO, UHC, and soot, are the result of combustion.
High temperatures during post-combustion typically produce NOx. Low temperatures and improper mixing
lead to incomplete combustion and emissions of CO and UHC. High temperatures with fuel-rich zones lead
to soot production. Using an appropriate parameter study in Chemkin-Pro, all emissions can therefore be
included in a mechanism reduction. They can be included as targets with either the Closed Homogeneous
reactor or the Laminar Flame Speed Calculator. As discussed above, we recommend use of the Closed
Homogeneous Reactor to allow a more complete parameter study with many iterations during the mech-
anism reduction.

2.2. Condition Coverage During Mechanism Reduction

To prepare a reaction mechanism for engine simulations, it is important to ensure that the mechanism
is valid over the range of local conditions that exists inside the combustion chamber, especially at the
start of combustion phasing. It is equally important to make sure the reduced mechanism is also valid
under the range of conditions that might exist in the engine prior to autoignition or initiation of flame
or production of emissions species. To cover these ranges of conditions, use the Parameter Study Facility
in Chemkin-Pro after setting up the baseline reactor model. (See the Chemkin-Pro Advanced Analysis
Guide.)

2.3. Steps in Setting Up a Mechanism Reduction

Here is an overview of the steps for setting up a mechanism reduction. Following this is more detail
for each step.

1. Set up the Chemkin-Pro reactor model. For this step, the Batch reactor template can be chosen for the
most convenient way to set up the model.

a. Use the fuel composition of interest.

b. Consider diluent or EGR composition if it is used in the engine.

c. Set the parameter study to cover the range of initial local conditions.

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
4 of ANSYS, Inc. and its subsidiaries and affiliates.
 Chemkin-Pro Project With Parameter Study for Basis of Reduction

2. Set up the mechanism reduction project in Reaction Workbench.

a. Select the method for reduction.

b. Select the targets of interest and provide the acceptable tolerance levels.

The following sections describe the two main steps in more detail.

2.4. Chemkin-Pro Project With Parameter Study for Basis of Reduction

These characteristic reactor properties settings may be used in typical cases. For parameter studies of
equivalence ratio or temperature, spread out the points roughly equally to cover the range of values.

Closed Homogeneous Reactor Problem Type

If the Batch reactor template is chosen, the following settings will be set automatically. If the template
is not used and a Batch reactor model is set up from the beginning, the following inputs must be made:

• Constant Volume and Solve the Energy Equation.

• Adiabatic (heat loss = 0)

• End time = 1 s. This is arbitrary, but is meant to set the simulation time long enough to cover one revolution
of an engine crank and to cover the complete ignition event for the longest ignition-delay time of interest.

Operating Conditions:
Pressure:

Set the value to one that is within the range of pressures at TDC in the engine. For knock modeling,
the pressure used should be set closer to the peak in-cylinder pressure of the normal cycle. However,
the exact pressure is not required since the same kinetics will be captured when the pressure used is
within a factor of 2 – 3 of that expected in the engine application. For this reason, no parameter study
on pressure is needed—unless the range of application covers a wider range of pressures.

For diesel engines, a typical pressure of 40 atm is a good approximation. Similarly, for SI engines, a
pressure of 20 bar is usually good. For gas turbines, operating pressure should be used. For applications
involving a wide variation in pressures, two pressures covering the low and high limits in the application
can be set in a parameter study.

Temperature:

Set a nominal value and then set up the parameter study for Initial Temperature. This is an important
setting that varies with the use case.

1. Ignition in diesel engines: Typically, low-temperature kinetics are the most relevant in this case. The
range of temperatures in the parameter study should cover temperatures between 700 K to 1200 K.
It should be sufficient to select up to 4 temperature points in this range to represent the range of
local temperatures prior to ignition in the engine.

2. For knock in SI engines: Combine ranges of settings similar to those recommended in both (a) and
(c).

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 5
Best Practices for Mechanism Reduction in Engine Applications

3. Flame propagation, or SI engines: Even though the temperature of the unburned mixture is low,
only high-temperature kinetics are needed to represent the kinetics in the flame. A temperature range
of 1200 K–2100 K is necessary to capture the flame kinetics. Select approximately 4 temperature points
in this range.

4. Emissions: No additional range of temperatures is required to capture the emissions as they are in-
herently captured by the detailed reaction mechanism reduced for appropriate conditions in (a), (b),
or (c).

Use the Equivalence Ratio option:

1. Set up a parameter study on equivalence ratio to cover the local conditions in the engine. Since
the local equivalence ratio varies widely, it is important to cover the range of values for various use
cases:

—For Diesel and SI engines: Use the range from 0.5 to 1.5. The parameter study should be
adequate with 3 to 4 values in the parameter study. In some HCCI engines, the equivalence
ratio may be highly fuel-lean; in such cases add the equivalence ratio of 0.3 to the parameter
study.

—When PAH and soot emissions are among the targets, an equivalence ratio of 3 should be
added to the parameter study.

2. Fuel composition: Specify that the fuel composition is the same as the model fuel composition to
be used in the engine simulations.

3. Oxidizer and EGR: If you’re interested in only a small range of EGR, you can assume a median value
of EGR rate and then use the fresh air charge plus EGR mixture (that is, IVC conditions minus the fuel)
as the Oxidizer. If you want to consider EGR range, you can set up a parameter study with two extreme
values (for example, 0 and maximum EGR level).

a. Example using One Oxidizer+EGR composition: For 40% EGR, we would calculate an “Oxidizer”
composition in mass fractions, based on complete combustion of a typical H/C ratio for diesel:

• CO2 = 0.0615, H2O = 0.0225, O2 = 0.1645, N2 = 0.7512.

• To the above we can add a small amount of NO, for example, NO ~0.000005 – 0.000050, depend-
ing on emissions and EGR levels of the targeted engine.

b. Using a Parameter Study to vary EGR: To account for variations in the EGR in a Chemkin-Pro
project, it is easiest to enter the EGR composition on the Added Species tab on the Reactant Species
panel. Then, use pure air composition for the Oxidizer Mixture tab (use the auto-populate button).
A parameter study is then set on the EGR species by doing a parameter study on the mass fractions
of each of the Added Species. Typically just 2 extreme values (0 and maximum EGR) are needed.

Solver Parameters:
Accept defaults for everything.

Output:
With the Batch reactor template, the Temperature Inflection option is chosen (on the Output Control
panel > Ignition Delay tab) for the Ignition criteria. If setting up the model from the beginning, you
must specify this setting.

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
6 of ANSYS, Inc. and its subsidiaries and affiliates.
 Setting Up a Reduction Operation Project in Reaction Workbench

For typical applications, a parameter study with 4 temperature values and 4 equivalence-ratio values
will contain a total of 16 runs in the Chemkin-Pro project. This number will grow when wider ranges
of operation are represented by the parameter study. For example, it becomes 32 if you consider two
pressures or two EGR levels. When multiple uses of the reduced mechanism are expected, use a wider
range of conditions to cover all the specific use cases.

2.5. Setting Up a Reduction Operation Project in Reaction Workbench

The next step is to set up a mechanism reduction operation in Reaction Workbench. To set up a reduction
operation, select the CHEMKIN project from the previous step as both the Initial CHEMKIN Project and
Target CHEMKIN Project on the Session Control panel. The Target CHEMKIN Project should always
be the project that uses the master mechanism. While the first reduction operation may use the same
project as the Initial CHEMKIN Project, the Initial CHEMKIN Project will be different at later stages
in the reduction, as it will use the result from a previous reduction step.

Note:

Chemkin-Pro Project settings should always be identical for the initial and target projects.

The steps to set up a mechanism reduction operation are listed below.

Note:

To avoid long names when setting up the initial mechanism reduction operation, provide a
short name (for example, test) in the text box Reduction Root Name on the Session Control
panel.

3. Workbench Preferences: Check the box that signifies that we only consider the last ignition point during
reduction. The rationale is that you won’t get the final ignition right if the pre-ignition is wrong, so setting
the last will assure the others.

4. Methods: When a high level of mechanism reduction is required, using one or two reduction methods is
not sufficient to produce a very small mechanism even if the number of targets and their ranges are limited
or tolerances are relaxed. In such situations, it is necessary to use multiple reduction methods in an iterative
manner. As an example, it may take more than 20 iterations involving various reduction methods to reduce
a gasoline mechanism to a size smaller than 300 or 400 species, depending on the targets. When setting
up a new mechanism reduction project, the recommended sequence of reduction methods is populated
in the Session Control panel. The reduction sequence can be repeated depending on the need for further
reduction. In general, the recommended methods for reducing mechanisms are DRG, DRGEP, and DRGPFA.
Avoid the Species Sensitivity analysis method if the mechanism is larger than 500 species due to long run
times. After the initial reduction, the linear lumping method can be used. Once the mechanism gets smaller
than about 500 species, the FSSA method can be used. We find the following order of methods to work
effectively.

a. Use DRGEP and DRG several times at first. The DRGPFA method can be used in later iterations.

b. Use the lumping method. Repeat (a) after using lumping. Add the DRGPFA method when the mechanism
has fewer than 1000 species.

c. Use DRGEP+Sensitivity, DRG+sensitivity, and DRGPFA+sensitivity, when the mechanism is lesser than
1000 species. Repeat (a) after each sensitivity method.

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 7
Best Practices for Mechanism Reduction in Engine Applications

d. Repeat b and c several more times. This typically results in a total of 10+ iterations.

e. Use the FSSA method. Then repeat d (p. 8). This will result in a total of 20+ iterations.

An example of the sequence of methods is shown in the snapshot below. You can select any variation
of this sequence involving the methods, such as DRG, DRGEP, Lumping, DRG+Sensitivity,
DRGEP+Sensitivity, and FSSA.

Note:

We do not recommend using PCA and CSP methods for most applications. The PCA
method is computationally expensive and the reduction is no greater than that gained
through DRG or DRGEP. CSP does not maintain the skeletal nature of the mechanism and
so prevents further reduction. With PCA and CSP, a solution is not always guaranteed,
especially for large mechanisms with more than 100 species.

Figure 2.1: Mechanism Reduction session with typical operations

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
8 of ANSYS, Inc. and its subsidiaries and affiliates.
 Setting Up a Reduction Operation Project in Reaction Workbench

Targets and Tolerances: The ignition time is an adequate target for capturing all the important kinetics
pathways and for capturing combustion involving both ignition and flame propagation using the reduced
mechanism. Additional targets are required to capture emissions.

Table 2.1: Recommended ranges for tolerances in reduced mechanisms (p. 9) and Table 2.2: Recom-
mended tolerances for soot targets in reduced mechanisms (p. 9) below show some recommended
ranges for tolerances determined to maintain sufficient accuracy in reduced mechanisms. Typically,
more relaxed tolerances will result in a smaller reduced mechanism. Also, a narrower range of targets
and fewer targets will result in smaller mechanisms.

Table 2.1: Recommended ranges for tolerances in reduced mechanisms

Species Mole Fractions

Target Ignition CO NO NO2 C2H2 * C3H6 *
Time
Absolute 1.E-6 1.E-4 1.E-5 1.E-5 1.E-4 1.E-5 to
Tolerance 1.E-4
Relative 10% 10-30% 10-30% 10-30% 10-30% 10-30%
Tolerance
Comparison End point Maximum Maximum Maximum Maximum Maximum
setting
* Representative UHC species. Exclude if only NOx and CO are the emissions targets.

Table 2.2: Recommended tolerances for soot targets in reduced mechanisms (p. 9) shows the recom-
mended tolerances for soot targets.

Table 2.2: Recommended tolerances for soot targets in reduced mechanisms

Species from the Soot Model*

Target C2H2 C3H3 & C6H5 & C6H5CH3
C4H2** C6H6 (toluene), Naph
(naphthalene),
A2R5
(acenaphthalene),
A4 (pyrene),
coronene **
Absolute Tolerance 1.E-4 1.E-4 1.E-5 1.E-6
Relative Tolerance 10-30% 10-30% 10-30% 10-30%
Comparison setting Maximum Maximum Maximum Maximum
* Select all species that participate as reactants in the soot nucleation and growth steps. Additional species
may be required depending on the choice of the soot model. If soot is not of interest, do not include
these species as targets.

** Select these species only if the MFL soot-surface mechanism is used.

Tip:

Hovering over the Target Tolerance % displayed for each reduced mechanism on the Run
Reduction panel displays the Run number and Error for each target. Sometimes a target
may not be relevant for the condition that results in a larger error during reduction. For ex-

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 9
Best Practices for Mechanism Reduction in Engine Applications

ample, soot precursors may not be relevant for reactor temperatures below 800 K, or a
higher level of error tolerance may be acceptable for ignition time at low temperatures for
certain applications. In those cases, disable certain runs either by clearing the check box on
the Select Conditions panel, or by increasing the Relative Tolerance, without impacting
conditions that matter for the end application.

6. It is good to begin with smaller values of tolerances. After several iterations of reduction methods and
obtaining smaller mechanisms, the tolerances can be relaxed as needed to achieve your desired balance
between accuracy and mechanism size.

7. Use default values on the Select Species panel and on the Select Conditions panel.

8. At this point, run the reduction operation. Note that the entire reduction session takes a fairly large amount
of disk space.

9. Once the smallest reduced mechanism is obtained, consider relaxing the tolerances if further reduction is
needed, and repeat the above steps.

10. Once you have reduced the mechanism to the desired size, we recommend archiving the resulting reduced
chemistry set, so that it can be easily unarchived in ANSYS Forte, Energico or Chemkin-Pro.

Refer to Figure 2.1: Mechanism Reduction session with typical operations (p. 8) to understand the ef-
fectiveness of various methods and their computation costs.

2.6. Continuing or Restarting a Reduction Run

Sometimes you will want to restart a reduction process after it has stopped. Examine the completed
steps and resume following a productive step. You may need to change methods, reduce the number
of parameter studies, or adjust parameter settings to allow the runs to proceed further.

The steps for restarting from a specific point in an existing reduction session are:

1. Save the last successful Chemkin-Pro project from the last reduction operation with acceptable error.
Archive and save this Chemkin-Pro project in a separate directory.

2. Unarchive the project from 1 in the working directory. Change the name of the mechanism file to a
meaningful name.

3. Load the new mechanism reduction session and specify the original (master) Chemkin-Pro project as Target
Project and the newly created project in b (p. 7)as the Initial Project. Run the session.

4. In the Select Operations panel, remove all the reduction methods from the top to the one that failed.

5. Continue with steps 9 and 10, from the section Setting Up a Reduction Operation Project in Reaction
Workbench (p. 7), to complete the procedure.

Release 2020 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
10 of ANSYS, Inc. and its subsidiaries and affiliates.