Ansys Chemkin Advanced Analysis Guide

ANSYS, Inc. Release 2025 R1

Southpointe January 2025
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

© 2025 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. About This Manual .................................................................................................................................. 1
2. Using the Parameter Study Facility ......................................................................................................... 3
2.1. Properties Available for Use in the Parameter Study Facility ................................................................ 3
2.2. Setting up a Parameter Study ............................................................................................................ 4
2.2.1. Opening a Parameter Study Dialog for Chemistry Properties ..................................................... 4
2.2.2. Opening a Parameter Study Dialog for Reactor or Stream Properties ......................................... 5
2.2.3. Opening a Parameter Study Dialog for Species-specific Properties ............................................. 5
2.2.4. Setting up a Single Parameter Variation .................................................................................... 6
2.2.5. Setting up Multiple Parameter Variations .................................................................................. 8
2.3. Running a Parameter Study ............................................................................................................. 11
2.4. Running and Post-processing from Batch Scripts ............................................................................. 14
2.4.1. Script Modifications for Running on a Cluster .......................................................................... 16
3. Using the Uncertainty Analysis Facility ................................................................................................ 17
3.1. Properties Available in the Uncertainty Analysis Facility ................................................................... 17
3.2. Setting up a Probability Distribution for an Input ............................................................................. 18
3.2.1. Opening an Uncertainty Analysis Dialog for Chemistry Properties ........................................... 18
3.2.2. Opening an Uncertainty Analysis Dialog for Reactor or Stream Properties ................................ 19
3.2.3. Opening an Uncertainty Analysis Dialog for Species-specific Properties ................................... 19
3.2.4. Conflicts With an Existing Parameter Study Setup ................................................................... 20
3.2.5. Setting up a Probability Distribution for an Uncertain Input Parameter .................................... 20
3.3. Running Uncertainty Analysis ......................................................................................................... 22
3.4. Revising An Uncertainty Analysis ..................................................................................................... 29

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. iii
 Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
iv of ANSYS, Inc. and its subsidiaries and affiliates.
List of Figures
2.1. closed_homogeneous__transient.ckprj - Pre-exponential Factor of O+H2=OH+H .................................... 5
2.2. closed_homogeneous__transient.ckprj - Reactant Fraction of O2 ............................................................ 6
2.3. closed_homogeneous__transient.ckprj - Parameter Study Dialog for Temperature ................................... 8
2.4. closed_homogeneous__transient.ckprj - Parameter Study Dialog for Pressure ....................................... 10
2.5. closed_homogeneous__transient.ckprj - Parameter Study Dialog for Pressure ....................................... 11
2.6. closed_homogeneous__transient.ckprj - Finished Running Parameter Study ......................................... 12
2.7. closed_homogeneous__transient.ckprj - Select Post-Processing Variables .............................................. 13
2.8. closed_homogeneous__transient.ckprj - Contour Plot for Mole Fraction H ............................................. 14
3.1. closed_homogeneous__transient.ckprj — Pre-exponential Factor of O + H2 = OH + H ............................ 19
3.2. closed_homogeneous__transient.ckprj — Reactant Fraction of O2 ........................................................ 20
3.3. Error Message about an Existing Parameter Study .................................................................................. 20
3.4. closed_homogeneous__transient.ckprj — Uncertainty Analysis Dialog for Temperature ........................ 21
3.5. Error Message about Standard Deviation Being Too Large ...................................................................... 22
3.6. closed_homogeneous__transient.ckprj — Uncertainty Analysis ............................................................ 22
3.7. closed_homogeneous__transient.ckprj — Select Uncertain Output Variables Panel ............................... 23
3.8. closed_homogeneous__transient.ckprj — Setup Uncertainty Analysis Panel ......................................... 24
3.9. closed_homogeneous__transient.ckprj — Final View of Setup Uncertainty Analysis Panel ...................... 25
3.10. closed_homogeneous__transient.ckprj — Running Uncertainty Analysis Panel .................................... 26
3.11. closed_homogeneous__transient.ckprj — Monitor Project Run Panel .................................................. 26
3.12. closed_homogeneous__transient.ckprj — Results of Variance Analysis and Error Analysis .................... 27
3.13. closed_homogeneous__transient.ckprj — Post-Processor Control Panel .............................................. 28
3.14. closed_homogeneous__transient.ckprj — PDF of OH Concentration ................................................... 28

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. v
 Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
vi of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 1: About This Manual
This manual describes advanced analysis features in Ansys Chemkin, including:

• Parameter Study

• Uncertainty Analysis

These advanced analysis features provide options for investigating the parameter variations or uncer-
tainties on Chemkin results. Using the Parameter Study Facility (p. 3) describes the Parameter Study
Facility, which allows easy automation of multiple runs to consider the effects of variable inputs to a
particular reactor model or Chemkin project. Using the Uncertainty Analysis Facility (p. 17) describes
the Uncertainty Analysis Facility, which performs multiple-input, multiple-output uncertainty analysis
with quantitative assessment of variant contributions.

You can also learn more about using these analytic features by:

• Reading Chemkin Getting Started Guide

• Working through the tutorials that feature them, available from the Ansys Help website.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 1
 Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
2 of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 2: Using the Parameter Study Facility
This chapter provides a description of the Ansys Chemkin Parameter Study Facility. A Parameter Study
can easily be used to define a large set of cases to run, where each case involves a change of one or a
few parameters. These cases can be set up in an efficient manner and you can monitor the progress of
the runs through the Ansys Chemkin Interface. A Solution Harvester parses solution files from all runs
to create a merged solution set, which can then be easily analyzed with the Chemkin Visualizer.

This chapter describes how to set up a Parameter Study of one or more parameters, how to run the
models within a Parameter Study, and how to visualize the solution results. The description here focuses
on the Parameter Study dialog within the Run Calculations node at the project level of the Project
Tree. We recommend that you first read through Using the Chemkin Interface in the Chemkin Getting
Started Guide and read Chemkin Visualization Manual so that you are familiar with using the Ansys
Chemkin Interface and the Chemkin Visualizer.

The Parameter Study option is represented by the icon in the Ansys Chemkin Interface. This icon
appears next to any parameter or property that can be defines as varying in a Parameter Study.

2.1. Properties Available for Use in the Parameter Study Facility

The following properties are available for use in the Parameter Studies Facility:

• Arrhenius reaction-rate parameters for gas-phase and surface reactions

• Auxiliary reaction-rate parameters (for example, third-body efficiencies and pressure fall-off parameters
for gas-phase reactions, coverage-dependent parameters for surface reactions, etc.)

• Site density for surface phases

• Surface bulk-phase mass density

• Species transport-property data

• Reactor Cluster properties

• Reactor properties, species-specific properties and material-specific data for all reactors in a network

• Reactor properties, species-specific properties and material-specific data for a single reactor

• Stream properties and species-specific data for an inlet stream

• Some Solver properties

A property can be used in the Parameter Study Facility if the Parameter Study icon is displayed next to
(to the right of ) the property in the input panel. When you place your mouse over the icon, it should
display Setup Parameter Study as the tool tip. If the icon is grayed-out instead of being shown in

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 3
Using the Parameter Study Facility

color or if it is not shown at all, the Parameter Study Facility is not enabled for that property for your
specific problem. For example, most of the solver parameters are not enabled for parameter study, since
they should have little impact on the solution and default values are usually recommended.

2.2. Setting up a Parameter Study

This section will describe the steps required to set up a Parameter Study. We will use the project
closed_homogeneous__transient.ckprj from the samples<release_number> directory to illustrate
the various options that are available for setting up a Parameter Study. These instructions assume that
you are already familiar with the Pre-processing and Reactor Properties panels described in Open a
Sample Project File in the Chemkin Getting Started Guide. The sub-sections Opening a Parameter Study
Dialog for Chemistry Properties (p. 4)—Opening a Parameter Study Dialog for Species-specific Proper-
ties (p. 5) below describe how to get to the parameter-variation dialog from the interface for different
types of input parameters. Setting up a Single Parameter Variation (p. 6) then describes how to fill in
the parameter variation desired for the selected input, once in the parameter-study dialog.

2.2.1. Opening a Parameter Study Dialog for Chemistry Properties

Once you have the Pre-Processing panel open and the chemistry set has been pre-processed, the
Mechanism Parameters tab will become activated. This tab will display the chemistry set and its asso-
ciated properties in the Ansys Chemkin Interface. It includes:

• A Species sub-tab that displays species thermodynamic data, bulk species mass density, and
transport data if available.

• A Reaction sub-tab that displays Arrhenius reaction-rate parameters and auxiliary reaction-rate
parameters for gas-phase and surface reactions.

• A Surface Phases sub-tab, if surface chemistry is present, that displays surface site-phase site
density.

The following steps describe how to open a Parameter Study dialog for the pre-exponential factor
of the gas-phase reaction O + H2 = OH + H:

1. Go to the Reaction sub-tab.

2. Click the Select Species pull-down and select O.

3. Click the Select Species pull-down again and select H2.

4. Click the Search Reactions button to search for all reactions containing these two species. The
results of the search will be displayed in the bottom half of the panel. If no reaction can be found
satisfying the search criteria, a message will be displayed instead.

5. Go to the search result area and click the pull-down list to get the reaction. You will see the
Arrhenius parameters of the reaction O + H2 =OH + H displayed in the panel, as shown in Fig-
ure 2.1: closed_homogeneous__transient.ckprj - Pre-exponential Factor of O+H2=OH+H (p. 5).

6. Click the Parameter Study icon to the right of the Pre-exponential factor A to open the Parameter
Study dialog.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
4 of ANSYS, Inc. and its subsidiaries and affiliates.
 Setting up a Parameter Study

Figure 2.1: closed_homogeneous__transient.ckprj - Pre-exponential Factor of O+H2=OH+H

2.2.2. Opening a Parameter Study Dialog for Reactor or Stream Properties

Choose the property on one of the Reactor or Inlet panels that you would like to vary. If the Parameter
Study icon next to the property value is active, click on the icon to open a Parameter Study dialog
for the desired property.

2.2.3. Opening a Parameter Study Dialog for Species-specific Properties

For species-specific properties of the Inlet or Reactor mixture, you need to select the species to be
included in the Parameter Study. The following steps illustrate how to open a Parameter Study dialog
for the Reaction Fraction of O2:

1. Go to the Species-specific Properties panel of the C1_Closed Homogeneous node.

2. Click the Parameter Study icon next to Reactant Fraction label.

3. Click the pull-down list in the Setup Data Analysis dialog and select O2, as shown in Fig-
ure 2.2: closed_homogeneous__transient.ckprj - Reactant Fraction of O2 (p. 6).

4. Click the OK button to open the Parameter Study dialog.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 5
Using the Parameter Study Facility

Figure 2.2: closed_homogeneous__transient.ckprj - Reactant Fraction of O2

2.2.4. Setting up a Single Parameter Variation

This section demonstrates the setup of a Parameter Study for the variation of initial Temperature for
the closed_homogeneous__transient project. The initial temperature is located on the Reactor
Physical Properties tab for the C1_Closed Homogeneous node on the Project Tree. On this panel,
click on the Parameter Study icon next to Temperature. We are going to set up a Parameter Study
including six runs for the temperature, where the temperature is varied from 1000 K to 1500 K, with
a 100-K increment. The Parameter-Study dialog is shown in Figure 2.3: closed_homogeneous__transi-
ent.ckprj - Parameter Study Dialog for Temperature (p. 8). There are four different approaches for
setting up this single-parameter variation, which provides flexibility to accommodate different types
of variations:

1. Manually Enter Each Point: Enter the number of Empty/Imported entries to be added as 6 and
click the Add Empty Entry button. This will create six empty cells in the table for the temperature.
Now you can go to each cell and type a temperature value from 1000 to 1500 or you could paste
an array of values (from a spreadsheet, for example) if you put your mouse in the top-column
entry point.

2. Import Values from a File: If you have an existing data file that contains temperature values, you
can use the Import button to import the values. Use the file browser to locate the data file. Once
the file is selected, all the values from the file will be loaded into the table. The data file must be
formatted as comma-delimited columns of values, with the first row holding comma-delimited
parameter names.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
6 of ANSYS, Inc. and its subsidiaries and affiliates.
 Setting up a Parameter Study

3. Use the Number of Values Option: Set the Start value as 1000.0 and the End value as 1500.0.
Select # of Values in the pull-down list and enter 6 as its value. Click the Populate button to
populate the table.

4. Use the Value Increment Option: Set the Start value as 1000.0 and the End value as 1500.0.
Select Increment in the pull-down list and enter 100 as its value. Click the Populate button to
populate the table.

Once you have populated an initial Parameter Study table, you can later modify or manipulate the
data entered with the following operations:

• Use one of the approaches listed above (step 1 (p. 6) to step 4 (p. 7)) to append new data to
the existing data table.

• Edit a data value by placing your cursor in the cell

• Sort the data from the lowest to the highest value by clicking the column header displaying Tem-
perature….

• Remove selected rows by highlighting a group of cells and clicking the Delete button.

• Export all the data in the table into a comma-delimited text file by clicking Export.

To close the dialog, click the Done button. You should see the color of the Temperature label change
from black to blue, indicating that a Parameter Study has been set up for this parameter.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 7
Using the Parameter Study Facility

Figure 2.3: closed_homogeneous__transient.ckprj - Parameter Study Dialog for Temperature

2.2.5. Setting up Multiple Parameter Variations

Next we will set up a Parameter Study in which both Temperature and Pressure are varied. With ex-
isting Parameter Study data for Temperature, as shown in Figure 2.3: closed_homogeneous__transi-
ent.ckprj - Parameter Study Dialog for Temperature (p. 8), the new data for Pressure variations will
need to interact with the existing data to form model runs. Note that each row in the parameter-
study table is always treated as a separate run. For multiple-parameter variations, there are two modes
of interactions, which can be selected using one of the two radio buttons in Figure 2.3: closed_homo-
geneous__transient.ckprj - Parameter Study Dialog for Temperature (p. 8). These options are described
below:

1. Superimpose all parameter variations so that more than one parameter varies on each run.
With this mode, the initial parameter-variation determines the number of column entries for the
entire Parameter Study and thus for the total number of runs. All parameter-variations must have
the same number of entries in each column in this case. Ansys Chemkin ensures this by one of
the following methods:

• If subsequent parameter-variations have a smaller number of entries than the initial parameter-
variation, Ansys Chemkin will require the user to do one of the following in order to obtain the
same number of entries as the initial parameter-variation: 1) extend the new data set using the
value chosen for the parameter increment; 2) repeat the entire data set; 3) repeat the last data
point, or, 4) extend the new set with empty entries which then can be filled out by the user
individually.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
8 of ANSYS, Inc. and its subsidiaries and affiliates.
 Setting up a Parameter Study

• If the subsequent parameter-variations have a larger number of entries than the initial parameter-
variation, Ansys Chemkin will require the user to do one of the following in order to obtain the
same number of entries as the initial parameter-variation: 1) extend existing data sets using the
last chosen increment; 2) repeat the set of entries for the existing data sets; 3) repeat the last
data point for existing data sets; 4) extend existing data sets with empty entries, or, 5) ignore
the extra data.

2. Vary each parameter independently so that only one parameter varies on each run. With this
mode, the number of rows in the parameter table and thus the number of runs, will equal the
number of data points for the initial parameter variable, multiplied by the number of data points
for each subsequent variable. For example, if the first parameter varied is temperature with six
data entries, and the second parameter varied is pressure with five data entries, there will be a
total of 30 runs to include all combinations of data points for the two parameters.

To use the Superimpose all parameter variations mode to vary pressure with five runs from 1 atm
to 2 atm, complete the following steps:

1. Click the Parameter Study icon for Pressure to open the Parameter Study dialog.

2. Set the start value as 1.0 and the end value as 2.0. Select # of Values in the pull-down list and
enter 5 as its value. Click the Populate button to populate the table.

3. Because we are superimposing all parameter variations, we need equal numbers of temperature
points and pressure points. The number of pressure points is less than the number of temperature
points, so we need to adjust the pressure points to match the number of temperature points. A
Populate Parameter Study dialog pops up with four options to adjust the pressure points. Click
the pull-down list and select Repeat the last data point. Click the OK button to populate the
table with pressure points.

As you can see in Figure 2.4: closed_homogeneous__transient.ckprj - Parameter Study Dialog for
Pressure (p. 10), the sixth pressure point repeats the fifth point.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 9
Using the Parameter Study Facility

Figure 2.4: closed_homogeneous__transient.ckprj - Parameter Study Dialog for Pressure

In the above example, we have specified five pressure points in step 2 (p. 9) and we need to extend
the pressure points in step 3 (p. 9) to match the number of temperature points. If we specify six
pressure points in step 2 (p. 9), the number of pressure points matches the number of temperature
points and no extra step (that is, step 3 (p. 9)) needs to be performed. If we specify seven pressure
points in step 2 (p. 9), the number of pressure points is greater than the number of temperature
points. We can either ignore the extra pressure points or extend the temperature points so that they
match the pressure points.

To use the Vary each parameter independently mode to vary pressure with five values from 1 atm
to 2 atm, complete the following steps:

1. Click the Parameter Study icon for Pressure to open the Parameter Study dialog if it is not open
yet.

2. The Pressure column in the table should be highlighted as shown in Figure 2.4: closed_homogen-
eous__transient.ckprj - Parameter Study Dialog for Pressure (p. 10). If there is no Pressure column,
go to step 3 (p. 10). You must delete all existing pressure points by selecting them and then
clicking the Delete button because this mode is only applicable if there are no existing pressure
points.

3. Click Vary each parameter independently, so that only one parameter varies on each run, to
select the mode.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
10 of ANSYS, Inc. and its subsidiaries and affiliates.
 Running a Parameter Study

4. Set the start value as 1.0 and the end value as 2.0. Select # of Values in the pull-down list and
enter 5 as its value. Click the Populate button to populate the table.

By varying the parameters independently, we end up with 30 runs total, which includes all combinations
of each parameter, as shown in Figure 2.5: closed_homogeneous__transient.ckprj - Parameter Study
Dialog for Pressure (p. 11).

Figure 2.5: closed_homogeneous__transient.ckprj - Parameter Study Dialog for Pressure

2.3. Running a Parameter Study

The Parameter Study will be performed through a series of runs, where a separate subdirectory for each
run will be created beneath the Project Working Directory. This Run Working DIrectory contains a copy
of all needed input files and the output files generated by the run. The Solution Harvester is invoked
during post-processing of the Parameter Study. The Harvester serves to parse the solution files generated
for each run and creates a merged solution set that can easily be analyzed using the visualization options
of the Post-Processor.

Here, we use the project closed_homogeneous__transient.ckprj as an example of running a Parameter

Study.

You can run the Parameter Study using the following steps:

1. Open the project closed_homogeneous__transient.ckprj and pre-process its chemistry set.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 11
Using the Parameter Study Facility

2. Follow the steps listed in Setting up a Single Parameter Variation (p. 6) and Setting up Multiple
Parameter Variations (p. 8) to set up parameter variations for temperature and pressure using the
Superimpose all parameter variations mode.

3. Follow the steps listed in Setting up a Parameter Study (p. 4) to set up parameter variations for
activation energy of the gas-phase reaction O + H2 = OH + H to have six points from 6290.0 to
6490.0 using the Superimpose all parameter variations mode.

4. Double-click the Run Calculations button.

5. If you want to see the details regarding the Parameter Study runs, click the Display Detail button
and the Parameter Study panel will show up in the interface.

6. Click the Begin button to run the Parameter Study Facility. It will create a new directory,
closed_homogeneous__transient_########_######, in your Project Working Directory,
where the #s in the name are replaced by a unique date- and time-based tag. Each run creates a
separate run directory named by the run number within this directory.

7. During the runs, a Monitor Project Run panel will open in the interface. This panel displays the current
progress of the parameter study. You can click the Interrupt Jobs button to skip the runs that have
not started yet. You can also open this panel at any time by double-clicking on the Monitor Project
Run node of the project tree.

8. When all the runs are complete, you will see the overall status displayed at the top of the Running
Parameter Study panel as well as the Run Status column. To view the diagnostic output file and log
file of each run, you can click the View Results… pull-down list in the Run Results column.

Figure 2.6: closed_homogeneous__transient.ckprj - Finished Running Parameter Study

9. Double-click the Analyze Results node in the project tree. An Analyze Results panel will open in the
interface.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
12 of ANSYS, Inc. and its subsidiaries and affiliates.
 Running a Parameter Study

10. Click the Begin Analysis button and a Select Post-Processing Variables panel, as shown in Fig-
ure 2.7: closed_homogeneous__transient.ckprj - Select Post-Processing Variables (p. 13), will open
to allow you to select solution parameters to plot in the Ansys Chemkin Post-Processor. You can
select solution sets (if available), select species and variables, apply various filters and options, and
set the units of measure, as discussed in Properties Available for Use in the Parameter Study Facil-
ity (p. 3).

Figure 2.7: closed_homogeneous__transient.ckprj - Select Post-Processing Variables

11. When you are done selecting variables and units, click the Process Solution Data button. A Monitor
Project Run panel will open to display the progress of post-processing jobs. After all the post-pro-
cessing jobs are finished, the data harvester will parse through the solution files over all runs to
create a merged solution set and the Ansys Chemkin Post-Processor will be invoked automatically
to allow you to graphically visualize all of the solution data. For example, you can create a contour
plot from the contour set of CKSoln_vs_parameter.csv using Pressure C1 Closed Homogeneous as
the X variable, Time as the Y variable, and Mole Fraction H as the Z variable, as seen in Fig-
ure 2.8: closed_homogeneous__transient.ckprj - Contour Plot for Mole Fraction H (p. 14).

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 13
Using the Parameter Study Facility

Figure 2.8: closed_homogeneous__transient.ckprj - Contour Plot for Mole Fraction H

2.4. Running and Post-processing from Batch Scripts

The Parameter Study can be run in batch mode, using scripts. The solution files can then be analyzed
using the visualization options of the Post-Processor or exported.

Prepare for the off-line run in the Ansys Chemkin interface:

1. Go to Edit > Preferences > Run/Output and select the Prepare Parameter Study as Off-line Jobs
option.

2. Open an existing project with a parameter study or set up a new parameter study.

3. Click Run Calculations and then click Display Details.

4. Select the desired parameter study runs and then click the Prepare Selected button at the bottom
of the panel. You will see that the run status has changed to "manual".

5. At this point you can shut down the Ansys Chemkin interface.

Continue working in the system files:

6. Go to the directories for your selected parameter study runs. This will be of the format <Project-
Name>_########_##### (see Step 6. in Running a Parameter Study (p. 11) for an example) which

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
14 of ANSYS, Inc. and its subsidiaries and affiliates.
 Running and Post-processing from Batch Scripts

will contain subdirectories of the format RUN_######## (for example RUN_000001, RUN_000002,
RUN_000003, etc.) Each of these subdirectories contain RUNJOB_CKRunProcessor.bat or
RUNJOB_CKRunProcessor.sh scripts.

The <ProjectName>_########_##### directory will also contain a script run_batch_para-

meter_study.bat (Windows) or run_batch_parameter_study.sh (Linux).

To run your batch parameter study, run this script as follows:

a. Open a command window session on your machine and cd to your Parameter Study run
folder.

b. To run your Parameter Study in batch mode from your session,

On Windows, type:

run_batch_parameter_study.bat

And on Linux:

./run_batch_parameter_study.sh

Note:

To ensure that your Ansys license packs are consumed effciently, you MUST use the
run_batch_parameter_study script to run a parameter study and not call the
RUNJOB scripts in the RUNxxxxx folders directly.

7. Return to the Ansys Chemkin interface. Go to Run Calculations > Show Details where you will see
that the status of runs that have finished is reported as "success".

8. Select any successful run and click the Prepare Post-Process button to open the Visualizer for
analysis.

To continue with post-processing outside of Ansys Chemkin, you can shut down the Chemkin interface
at this point.

• Go to the directories for the runs and you will see RUNJOB_GetSolutionData_***.bat or
RUNJOB_GetSolutionData_***.sh scripts.

Run these scripts to post-process the runs and generate corresponding comma-delimited .ckcsv
files for each run.

The command-delimited .ckcsv files allow you to analyze the solutions using your preferred software.

To continue with post-processing from Ansys Chemkin, follow these steps:

1. After all the jobs have finished, go back to the Ansys Chemkin User Interface's panel and click the
Prepare Post-Process button.

2. In the Select Post-Processing Variables panel, select what to post-process and click Process Solution
Data.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 15
Using the Parameter Study Facility

3. The Monitor Project Run panel appears.

4. When the Monitor panel appears with the message, Done preparing post-processing jobs as off-line
jobs, click Close on the panel.

5. Now go back to the Analysis Dir in your file browser and each RUN Dir will have a RUNJOB_Get-
SolutionData_***.bat or RUNJOB_GetSolutionData_***.sh file.

6. Run this file to manually post-process the solutions for each run.

2.4.1. Script Modifications for Running on a Cluster

You may wish to run your batch parameter on different machines on a cluster (provided your Ansys
install and run folders are accessible from all machines) and want to run subsets of your parameter
runs on different machines rather than all of them.

To do this, modify the last line of the run_batch_parameter_study script and replace the

-nruns <number of parameter studies>

argument with

-nstart <starting run number> -nend <ending run number>

to define the subset you wish to solve, starting from run <starting run number> to <ending run
number>.

For example,

scheduler -nstart 32 -nend 63 -nparallel 16 -delay 1

runs jobs RUN_000032 to RUN_000063 running 16 in parallel.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
16 of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 3: Using the Uncertainty Analysis Facility
This chapter provides a description of the Uncertainty Analysis Facility. An uncertainty analysis allows
you to study a large set of uncertain input parameters, with each parameter represented by a probab-
ility distribution. You can identify such input variable distributions and the characteristics of the variable
interactions, as well as monitor the progress of the analysis through the Ansys Chemkin Interface. A
Solution Harvester parses solution files over all runs to create a merged solution set, which can then
be easily analyzed with the Chemkin Post-Processor.

This chapter describes how to set up a probability distribution for an uncertain input parameter, how
to set up the uncertainty analysis, how to refine the analysis based on previous results, and how to
post-process the solutions. The description here focuses on the Uncertainty Analysis dialog and the Run
Uncertainty Analysis node at the project level of the Project Tree.

Note:

We recommend that you read through Chapters 2 and 3 of Chemkin Getting Started Guide
and the Chemkin Visualization Manual so that you are familiar with using the Ansys Chemkin
Interface and the Chemkin Post-Processor.

In the Ansys Chemkin Interface, the icon representing the Uncertainty Analysis Facility appears next
to any parameter or property that can be identified as an uncertain input in an Uncertainty Analysis.

3.1. Properties Available in the Uncertainty Analysis Facility

Uncertainty Analysis is available for most inputs in an Ansys Chemkin project. You can set up a probab-
ility distribution for the following properties:

• Arrhenius reaction-rate parameters for gas-phase and surface reactions

• Auxiliary reaction-rate parameters (for example, third-body efficiencies and pressure fall-off parameters
for gas-phase reactions, coverage-dependent parameters for surface reactions, etc.)

• Site density for surface phases

• Surface bulk-phase mass density

• Species transport-property data

• Cluster properties

• Reactor properties, species-specific properties and material-specific data for all reactors in a network

• Reactor properties, species-specific properties and material-specific data for a single reactor

• Stream properties and species-specific data for an inlet stream

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 17
Using the Uncertainty Analysis Facility

• Some solver properties

A property can be used in an uncertainty analysis if the Uncertainty Analysis icon is displayed next to
(to the right of ) the property in the input panel. When you place your mouse over the icon, it should
display Setup Uncertainty Analysis as the tool tip. If the icon is inactive instead of being shown in
color, the Uncertainty Analysis Facility is not enabled for that property for your current problem.

3.2. Setting up a Probability Distribution for an Input

This section describes the steps required to set up a probability distribution for an uncertain input
parameter. We will use the project closed_homogeneous__transient.ckprj from the samples<release_num-
ber> directory to illustrate the various options that are available. These instructions assume that you
are already familiar with the Parameter Study facility as described in Using the Parameter Study Facil-
ity (p. 3) and the Pre-processing and Reactor Properties panels described in Setting up a Parameter
Study (p. 4) in Chemkin Getting Started Guide. The sub-sections Opening an Uncertainty Analysis
Dialog for Chemistry Properties (p. 18)—Opening an Uncertainty Analysis Dialog for Species-specific
Properties (p. 19) below describe how to get to the uncertainty-analysis dialog from the interface for
different types of input parameters. Setting up a Probability Distribution for an Uncertain Input Para-
meter (p. 20) then describes how to supply the distribution characteristics for the selected input, once
in the uncertainty-analysis dialog and how to complete the remaining uncertainty-analysis set-up.

3.2.1. Opening an Uncertainty Analysis Dialog for Chemistry Properties

Once you have the Pre-Processing panel open and the chemistry set has been pre-processed, the
Mechanism Parameters tab becomes activated. This tab will display the chemistry set and its associated
properties in the Ansys Chemkin Interface. It includes:

• A Species sub-tab that displays species thermodynamic data, bulk species mass density, and
transport data, if available.

• A Reaction sub-tab that displays Arrhenius reaction-rate parameters and auxiliary reaction-rate
parameters for gas-phase and surface reactions.

• A Surface Phases sub-tab, if surface chemistry is present, that displays surface site density.

The following steps describe how to open an Uncertainty Analysis dialog for the pre-exponential
factor of the gas-phase reaction O + H2 = OH + H:

1. Go to the Reaction sub-tab.

2. Click the Select Species pull-down and select O.

3. Click the Select Species pull-down again and select H2.

4. Click the Search Reactions button to search for all reactions containing these two species. The
results of the search will be displayed in the bottom half of the panel. If no reaction can be found
satisfying the search criteria, a message will be displayed instead.

5. Go to the search result area and click the pull-down list to get the reaction. You will see the
Arrhenius parameters of the reaction O + H2 = OH + H displayed in the panel, as shown in Fig-
ure 3.1: closed_homogeneous__transient.ckprj — Pre-exponential Factor of O + H2 = OH + H (p. 19).

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
18 of ANSYS, Inc. and its subsidiaries and affiliates.
 Setting up a Probability Distribution for an Input

6. Click the Uncertainty Analysis icon to the right of the Pre-exponential factor A to open the Uncer-
tainty Analysis dialog.

Figure 3.1: closed_homogeneous__transient.ckprj — Pre-exponential Factor of O + H2 = OH +

H

3.2.2. Opening an Uncertainty Analysis Dialog for Reactor or Stream Proper-

ties
Choose the property on one of the Reactor or Inlet panels whose uncertainty you would like to analyze.
If the Uncertainty Analysis icon next to the property value is active, click on the icon to open an Un-
certainty Analysis dialog for that property.

3.2.3. Opening an Uncertainty Analysis Dialog for Species-specific Properties

For species-specific properties of the Inlet or Reactor mixture, you need to select the species to be
included in the Uncertainty Analysis. The following steps illustrate how to open an Uncertainty Ana-
lysis dialog for the Reaction Fraction of O2:

1. Go to the Species-specific Properties panel of the C1_Closed Homogeneous node.

2. Click the Uncertainty Analysis icon next to Reactant Fraction label.

3. Click the pull-down list in the Setup Data Analysis dialog and select O2, as shown in Fig-
ure 3.2: closed_homogeneous__transient.ckprj — Reactant Fraction of O2 (p. 20).

4. Click the OK button to open the Uncertainty Analysis dialog.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 19
Using the Uncertainty Analysis Facility

Figure 3.2: closed_homogeneous__transient.ckprj — Reactant Fraction of O2

3.2.4. Conflicts With an Existing Parameter Study Setup

If a Parameter Study has been set up in a project, you cannot set up an Uncertainty Analysis in the
same project until the Parameter Study is deleted. When you try to click the Uncertainty Analysis icon
to open an Uncertainty Analysis dialog, you will receive an error message, as displayed in Figure 3.3: Er-
ror Message about an Existing Parameter Study (p. 20). To delete the existing Parameter Study, click
any active Parameter Study icon to open a Parameter Study dialog, click the Select All button first
and then click the Delete button. Once the Parameter Study setup is deleted, you can click the Un-
certainty Analysis icon to open a dialog.

Figure 3.3: Error Message about an Existing Parameter Study

3.2.5. Setting up a Probability Distribution for an Uncertain Input Parameter

This section demonstrates the setup of probability distributions for initial Temperature and Pressure
for the closed_homogeneous__transient project. These are both located on the Reactor Physical
Properties tab for the C1_Closed Homogeneous node on the Project Tree. On this panel, click on the

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
20 of ANSYS, Inc. and its subsidiaries and affiliates.
 Setting up a Probability Distribution for an Input

Uncertainty Analysis icon next to Temperature. The dialog is shown in Figure 3.4: closed_homogen-
eous__transient.ckprj — Uncertainty Analysis Dialog for Temperature (p. 21). First, you need to choose
a probability density function that can closely represent the uncertain behavior of initial Temperature.
In case you do not have enough information to make a selection, choose Normal distribution. Next,
fill in the distribution parameters. In the case of Normal distribution, the Mean Value is the same as
the nominal value of the initial Temperature, so that you cannot set the Mean Value. Set the
Standard Deviation to 50.0. To close the dialog, click the Done button. The Temperature label’s
color changes from black to green, which indicates that an Uncertainty Analysis has been set up for
this parameter.

Figure 3.4: closed_homogeneous__transient.ckprj — Uncertainty Analysis Dialog for Temperature

Now click on the Uncertainty Analysis icon next to Pressure, use the Normal distribution and set the
Standard Deviation as 0.5. When you click the Done button, an error message pops up, as shown
in Figure 3.5: Error Message about Standard Deviation Being Too Large (p. 22), to tell you that the
standard deviation is too large and the distribution will generate negative sampling points for pressure,
which has to be positive. You must reduce the standard deviation until this error message no longer
appears. Reset the Standard Deviation to 0.1 instead. Click the Done button to close the dialog.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 21
Using the Uncertainty Analysis Facility

Figure 3.5: Error Message about Standard Deviation Being Too Large

3.3. Running Uncertainty Analysis

The Uncertainty Analysis will be performed through a series of runs corresponding to the set of colloc-
ation points generated based on the characteristics of the distributions of the uncertain input parameters.
A separate subdirectory for each run will be created beneath the Project Working Directory. This Run
Working Directory contains a copy of all needed input files and the output files generated by the run.
The Solution Harvester is automatically invoked to collect solution data for user-specified output variables.
A variance analysis is performed to study the contribution of each uncertain input parameter to the
uncertainty of each user-selected output parameter. An error analysis is performed to determine the
accuracy of the approximation model used in the uncertainty analysis. During post-processing of the
Uncertainty Analysis, the Harvester is invoked to parse the solution files generated for each run and
creates a merged solution set that can easily be analyzed using the visualization options of the Post-
Processor.

Here, we use the project closed_homogeneous__transient.ckprj as an example of running

an Uncertainty Analysis.

You can run the Uncertainty Analysis using the following steps:

1. Open the project closed_homogeneous__transient.ckprj and pre-process the chemistry

set.

2. Follow the steps listed in Opening an Uncertainty Analysis Dialog for Reactor or Stream Proper-
ties (p. 19) and Setting up a Probability Distribution for an Uncertain Input Parameter (p. 20) to set
up distributions for temperature and pressure.

3. Double-click the Run Calculations node in the project tree. A Run Calculations panel will pop up in
the interface, as shown in Figure 3.6: closed_homogeneous__transient.ckprj — Uncertainty Analys-
is (p. 22).

Figure 3.6: closed_homogeneous__transient.ckprj — Uncertainty Analysis

4. Click the Begin button. A Select Uncertain Output Variables panel will pop up in the interface, as
shown in Figure 3.7: closed_homogeneous__transient.ckprj — Select Uncertain Output Variables
Panel (p. 23). You can use this panel to select the solution variables of interest. Click the Clear All

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
22 of ANSYS, Inc. and its subsidiaries and affiliates.
 Running Uncertainty Analysis

button to deselect all variables and then select OH only. With this selection, we are going to study
the effect of the variation in input values of reactor temperature and pressure on the final OH con-
centration in the reactor. If you are not sure which output variables are of most interest to you, you
can just select all of them for now. You can always deselect uncertain output variables later on, as
shown in the next steps. Click the Proceed to Uncertainty Analysis button.

Figure 3.7: closed_homogeneous__transient.ckprj — Select Uncertain Output Variables Panel

5. A Setup Uncertainty Analysis panel will pop up in the interface, as shown in Figure 3.8: closed_ho-
mogeneous__transient.ckprj — Setup Uncertainty Analysis Panel (p. 24). This panel allows you to
set up the Polynomial Chaos Expansions (PCE) of uncertain output variables.

6. As discussed in step 4 (p. 22), you can deselect any uncertain output (response) variables in this
panel by unchecking the corresponding rows in the Response Variables table. You can also click the
Re-select Uncertain Output button to re-run the nominal case and re-select the uncertain output
variables. The second column of the Response Variables table is of particular importance here. As
you know, this project is a transient calculation of a closed homogeneous batch reactor. A transient
calculation will generate a vector of values for each output variable, one value per time point.
However, the collocation calculation can use only a single value of each output variable. Therefore
you must select which value or which type of combination of the values will be used in the calculation.
You can click on an entry in the second column to select the type of data from a drop-down list.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 23
Using the Uncertainty Analysis Facility

For OH concentration, we will use the default data type, that is, the nominal (last) value in the
transient solution.

Figure 3.8: closed_homogeneous__transient.ckprj — Setup Uncertainty Analysis Panel

7. You can use the rest of the panel to set up the Polynomial Chaos Expansion (PCE) of uncertain
output variable, OH concentration. You can select the highest order of polynomials used in the ap-
proximation of the uncertain output variable. Set the order to be 3. The Polynomial Cross Terms
table is optional. If you think there is strong inter-dependency between the uncertain input para-
meters, you can use the polynomial cross terms to represent the dependency. In this case, we want
a cross term that is first order on pressure and second order on temperature. Enter I1 I2 under the
Uncertain Input ID List column and enter 1 2 under the Exponent List column. Provide an ID for this
cross term by entering term1 under the Group ID column. The last sub-panel in the panel displays
the total number of model runs that are required to complete the uncertainty analysis of OH con-
centration based on the current setup. Figure 3.9: closed_homogeneous__transient.ckprj — Final
View of Setup Uncertainty Analysis Panel (p. 25) shows the final view of the panel. Click the Setup
Uncertainty Analysis button to proceed.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
24 of ANSYS, Inc. and its subsidiaries and affiliates.
 Running Uncertainty Analysis

Figure 3.9: closed_homogeneous__transient.ckprj — Final View of Setup Uncertainty Analysis

Panel

8. A Running Uncertainty Analysis panel will pop up in the interface, as shown in Figure 3.10: closed_ho-
mogeneous__transient.ckprj — Running Uncertainty Analysis Panel (p. 26). This panel looks similar
to the Run Parameter Study panel described in Section 3.3 of Chemkin Getting Started Guide. You
can use this panel to launch jobs and process solutions.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 25
Using the Uncertainty Analysis Facility

Figure 3.10: closed_homogeneous__transient.ckprj — Running Uncertainty Analysis Panel

9. Click the Select All button to select all runs. Click the Run Selected button to run all the jobs. It
will create a new directory, closed_homogeneous__transient_########_######, in your Project
Working Directory, with each directory date- and time-stamped. Each run creates a separate run
directory named by the run number in this closed_homogeneous__transient_########_######
directory. During the runs, a Monitor Project Run panel will pop up in the interface, as shown in
Figure 3.11: closed_homogeneous__transient.ckprj — Monitor Project Run Panel (p. 26). This panel
displays the current progress of the runs. If you want to stop early, you can click the Interrupt Jobs
button to skip the runs that have not started yet. You can open this panel at any time by double-
clicking on the Monitor Project Run node of the project tree. When all runs are done, close this
panel by clicking the Close button.

Figure 3.11: closed_homogeneous__transient.ckprj — Monitor Project Run Panel

10. When all the runs are complete, you will see the overall status displayed at the top of the Running
Uncertainty Analysis panel as well as the Run Status column. To view the diagnostic output file and
log file of each run, you can click the Click to View Results… pull-down list in the Run Results
column.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
26 of ANSYS, Inc. and its subsidiaries and affiliates.
 Running Uncertainty Analysis

11. Click the Do Analysis button to perform the collocation calculation based on simulation results,
the variance analysis, and the error analysis. A file editor will pop up in the interface displaying the
results of the analysis, as shown in Figure 3.12: closed_homogeneous__transient.ckprj — Results of
Variance Analysis and Error Analysis (p. 27). The results contain three sections.

• The first section displays the polynomial chaos expansion of the uncertain output variable, OH
concentration in the reactor, in term of the uncertain input variables, reactor pressure and tem-
perature.

• The second section displays the results of the error analysis and lists suggestions on how to improve
the accuracy.

• The third section displays the results of the variance analysis and it shows that the variation in
reactor pressure makes little contribution to the variation in OH concentration.

Figure 3.12: closed_homogeneous__transient.ckprj — Results of Variance Analysis and Error

Analysis

12. When the analysis is complete, the View P.D.F. button becomes active. Click the View P.D.F. button
to view the plots of distributions of uncertain input and output variables. A Ansys Chemkin Post-
Processor Control Panel will pop up as shown in Figure 3.13: closed_homogeneous__transient.ckprj
— Post-Processor Control Panel (p. 28). Click the Plot Set pull-down list to select pdf of Cluster1
(C1): OH. Click the Display Plot button to view the probability distribution of the OH concentration
in the reactor at the last time point, as shown in Figure 3.14: closed_homogeneous__transient.ckprj
— PDF of OH Concentration (p. 28).

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 27
Using the Uncertainty Analysis Facility

Figure 3.13: closed_homogeneous__transient.ckprj — Post-Processor Control Panel

Figure 3.14: closed_homogeneous__transient.ckprj — PDF of OH Concentration

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
28 of ANSYS, Inc. and its subsidiaries and affiliates.
 Revising An Uncertainty Analysis

13. You can click the Post-Process button to process and display the results from all the runs. It is de-
scribed in detail in Section 1.3 of Chemkin Getting Started Guide and the Chemkin Visualization
Manual.

14. If you want to revise the setup of the uncertainty analysis, such as adjusting the order of the poly-
nomial and adding/removing polynomial cross terms, you can click the Revise Uncertainty Setup
button at the bottom of the Running Uncertainty Analysis panel to go back to the Setup Uncertainty
Analysis panel displayed in Figure 3.9: closed_homogeneous__transient.ckprj — Final View of Setup
Uncertainty Analysis Panel (p. 25). Because of the nature of the collocation sampling points, different
uncertainty setups usually share some sampling points. The Uncertainty Analysis Facility will auto-
matically recycle the runs of these shared sampling points from a previous uncertainty setup, as
long as they are saved in the project, so that you do not need to re-run these jobs.

15. If you want to remove all the distribution setups for uncertain input parameters and discard all the
run results and analysis results of the uncertainty analysis, you can click the Clear Uncertainty Setup
button at the bottom of the Running Uncertainty Analysis panel. This allows you to set up a new
Uncertainty Analysis or a new Parameter Study from scratch.

3.4. Revising An Uncertainty Analysis

Performing an uncertainty analysis with the collocation method is an iterative process, because we need
to approximate the uncertain output variables in terms of the uncertain input variables. Due to the
complexity of the model, we normally do not have enough insight to set up an accurate approximation
in the beginning. Therefore we recommend that you always begin with the simplest approximation
and add complexity as you gain more insight into the model along the way. The simplest approximation
means that you will set the highest order of the polynomials to 2 on the Setup Uncertainty Analysis
panel. If you think there is strong inter-dependency between some uncertain input variables, you can
add polynomial cross terms involving these variables. Keep in mind that the combined order of all
variables in a cross term cannot exceed the highest order of the approximation.

After clicking the Do Analysis button on the Running Uncertainty Analysis panel, you will see the results
of both error analysis and variance analysis. The error analysis provides two measures of accuracy for
the approximation: relative sum-square-root error and index of agreement. A very small relative sum-
square-root error (that is, << 1.0) and a high index of agreement (that is, very close to 1.0) indicate
good accuracy. If either of these requirements is not satisfied, the accuracy is considered low. To improve
the accuracy, you can revise the uncertainty analysis setup in two ways:

• Increase the highest order of polynomial approximation by 1

• Add polynomial cross terms for uncertain input variables that are correlated or inter-dependent.

You should always consider adding polynomial cross terms before increasing the highest order of
polynomial approximation to be greater than 3. An order greater than 5 is not recommended unless it
has shown clear improvement in the accuracy.

The variance analysis will indicate which uncertain input variables have the least impact on the variation
of the uncertain output variables. However, the results of the variance analysis are valid only if the ap-
proximation has enough accuracy, as indicated by the results of the error analysis. With reasonable ac-
curacy, you can remove uncertain input variables from the analysis if they have shown little impact on
the variations of all uncertain output variables. This will reduce the dimensionality of the approximation
and reduce the number of model runs needed for each iteration.

Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 29
 Release 2025 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
30 of ANSYS, Inc. and its subsidiaries and affiliates.