Hyperstudy 2019 Tutorials

Altair HyperStudy 2019
Tutorials
altairhyperworks.com
Intellectual Property Rights Notice
Copyrights, Trademarks, Trade Secrets, Patents & Third Party Software Licenses
Altair HyperStudy 2019 Copyright 1999-2019
The Platform for Innovation™
Altair Engineering Inc. Copyright © 1986-2019. All Rights Reserved.
Note: Pre-release versions of Altair software are provided ‘as is’, without warranty of any
kind. Usage of pre-release versions is strictly limited to non-production purposes.
Altair HyperWorks™ - The Platform for Innovation™

Altair AcuConsole™ ©2006-2019
Altair AcuSolve™ ©1997-2019
Altair ElectroFlo™ ©1992-2019
Altair ESAComp™ ©1992-2019
Altair Feko™ ©1999-2014 Altair Development S.A. (Pty) Ltd.; ©2014-2019 Altair Engineering Inc.
Altair Flux™ ©1983-2019
Altair FluxMotor™ ©2017-2019
Altair HyperCrash™ ©2001-2019
Altair HyperGraph™ ©1995-2019
Altair HyperLife™ ©1990-2019
Altair HyperMesh™ ©1990-2019
Altair HyperStudy™ ©1999-2019
Altair HyperView™ ©1999-2019
Altair Virtual Wind Tunnel™ ©2012-2019
Altair HyperXtrude™ ©1999-2019
Altair MotionSolve™ ©2002-2019
Altair MotionView™ ©1993-2019
Altair Multiscale Designer™ ©2011-2019
Altair OptiStruct™ ©1996-2019
Altair Radioss™ ©1986-2019
Altair SimLab™ ©2004-2019
Altair SimSolid™ ©2015-2019
Altair nanoFluidX™ ©2013-2018 Fluidyna GmbH, © 2018-2019 Altair Engineering Inc.
Altair ultraFluidX™ ©2010-2018 Fluidyna GmbH, © 2018-2019 Altair Engineering Inc.

Altair WinProp™ ©2000-2019;
Altair ConnectMe™ ©2014-2019;
Plus other products from the Altair solidThinking Platform.
Altair Packaged Solution Offerings (PSOs)

Altair Automated Reporting Director™ ©2008-2019
Altair GeoMechanics Director™ ©2011-2019
Altair Impact Simulation Director™ ©2010-2019
Altair Model Mesher Director™ ©2010-2019
Altair NVH Director™ ©2010-2019
Altair Squeak and Rattle Director™ ©2012-2019
Altair Virtual Gauge Director™ ©2012-2019
Altair Weight Analytics™ ©2013-2019
Altair Weld Certification Director™ ©2014-2019
Altair Multi-Disciplinary Optimization Director™ ©2012-2019
Altair solidThinking - Where Innovation Begins™

Altair Inspire™ ©2009-2019 including Altair Inspire Motion and Altair Inspire Structures
Altair Inspire™ Extrude-Metal ©1996-2019 (formerly Click2Extrude®-Metal)

Altair Inspire™ Extrude-Polymer ©1996-2019 (formerly Click2Extrude®-Polymer)
Altair Inspire™ Cast ©2011-2019 (formerly Click2Cast®)
Altair Inspire™ Form ©1998-2019 (formerly Click2Form®)
Altair Inspire™ Mold ©2009-2019 (initial release-Q2 2019)
Altair Inspire™ Studio ©1993-2019 (formerly ‘Evolve’)
Altair Compose™ ©2007-2019 (formerly solidThinking Compose®)
Altair Activate™ ©1989-2019 (formerly solidThinking Activate®)
Altair Embed™ ©1989-2019 (formerly solidThinking Embed®)

• Altair Embed SE™ ©1989-2019 (formerly solidThinking Embed® SE)
• Altair Embed/Digital Power Designer ©2012-2019
Altair SimLab™ ©2004-2019
Altair 365™ ©1994-2019
Altair PBSWorks™ - Accelerating Innovation in the Cloud™

Altair PBS Professional™ ©1994-2019
Altair Control™ ©2008-2019; (formerly PBS Control)
Altair Access™ ©2008-2019; (formerly PBS Access)
Proprietary Information of Altair Engineering

Altair Accelerator™ ©1995-2019; (formerly NetworkComputer)
Altair Accelerator Plus™©1995-2019; (formerly WorkloadXelerator)
Altair FlowTracer™ ©1995-2019; (formerly FlowTracer)
Altair Allocator™ ©1995-2019; (formerly LicenseAllocator)
Altair Monitor™ ©1995-2019; (formerly LicenseMonitor)
Altair Hero™ ©1995-2019; (formerly HERO)
Altair Software Asset Optimization™ (SAO) ©2007-2019
Note:
Compute Manager™ ©2012-2017 is now part of Altair Access
Display Manager™ ©2013-2017 is now part of Altair Access
PBS Application Services™ ©2008-2017 is now part of Altair Access

PBS Analytics™ ©2008-2017 is now part of Altair Control
PBS Desktop™ ©2008-2012 is now part of Altair Access, specifically Altair Access
desktop, which also has Altair Access web and Altair Access mobile
e-Compute™ ©2000-2010 was replaced by “Compute Manager” which is now Altair

Access
Altair SmartWorks™ - Innovation Intelligence®

Altair SmartCore™ ©2011-2019
Altair SmartEdge™ ©2010-2019
Altair SmartSight™ ©2014-2019
Altair intellectual property rights are protected under U.S. and international laws and treaties.
Additionally, Altair software is protected under patent #6,859,792 and other patents pending. All other
marks are the property of their respective owners.
ALTAIR ENGINEERING INC. Proprietary and Confidential. Contains Trade Secret Information.
Not for use or disclosure outside of Altair and its licensed clients. Information contained in Altair
software shall not be decompiled, disassembled, “unlocked”, reverse translated, reverse engineered,
or publicly displayed or publicly performed in any manner. Usage of the software is only as explicitly
permitted in the end user software license agreement. Copyright notice does not imply publication.
Third party software licenses

AcuConsole contains material licensed from Intelligent Light (www.ilight.com) and used by permission.
Software Security Measures:
Altair Engineering Inc. and its subsidiaries and affiliates reserve the right to embed software security
mechanisms in the Software for the purpose of detecting the installation and/or use of illegal copies of
the Software. The Software may collect and transmit non-proprietary data about those illegal copies.
Data collected will not include any customer data created by or used in connection with the Software

and will not be provided to any third party, except as may be required by law or legal process or to
enforce our rights with respect to the use of any illegal copies of the Software. By using the Software,
each user consents to such detection and collection of data, as well as its transmission and use if an
illegal copy of the Software is detected. No steps may be taken to avoid or detect the purpose of any
such security mechanisms.

Technical Support
Altair provides comprehensive software support via web FAQs, tutorials, training classes, telephone and
e-mail.
Altair Support on the World Wide Web

The Altair web site is a valuable online companion to Altair software. Visit www.altairhyperworks.com
for tips and tricks, training course schedules, training/tutorial videos, and other useful information.
Altair Training Classes

Altair training courses provide a hands-on introduction to our products, focusing on overall functionality.
Courses are conducted at our main and regional offices or at your facility. If you are interested in
training at your facility, please contact your account manager for more details. If you do not know who
your account manager is, please send an e-mail to training@altair.com and your account manager will
contact you.
Telephone and E-mail

When contacting Altair support, please specify the product and version number you are using along with
a detailed description of the problem. Many times, it is very beneficial for the support engineer to know
what type of workstation, operating system, RAM, and graphics board you have, so please have that
information ready. If you send an e-mail, please specify the workstation type, operating system, RAM,
and graphics board information in the e-mail.
To contact an Altair support representative, reference the following table or the information available on
the HyperWorks website: www.altairhyperworks.com/ClientCenterHWSupportProduct.aspx.
Location Telephone E-mail
Australia 64.9.413.7981 anzsupport@altair.com
Brazil 55.11.3884.0414 br_support@altair.com
Canada 416.447.6463 support@altairengineering.ca
China 86.400.619.6186 support@altair.com.cn
France 33.1.4133.0992 francesupport@altair.com
Germany 49.7031.6208.22 hwsupport@altair.de
India 91.80.6629.4500 support@india.altair.com
1.800.425.0234 (toll free)
Israel israelsupport@altair.com
Italy 39.800.905.595 support@altairengineering.it

Location Telephone E-mail
Japan 81.3.6225.5830 support@altairjp.co.jp
Malaysia aseansupport@altair.com
Mexico 55.56.58.68.08 mx-support@altair.com
South Africa 27 21 8311500 support@altair.co.za
South Korea 82.70.4050.9200 support@altair.co.kr
Spain 34 910 810 080 support-spain@altair.com
Sweden 46.46.460.2828 support@altair.se
United Kingdom 01926.468.600 support@uk.altair.com
United States 248.614.2425 hwsupport@altair.com
For questions or comments about this help system, send an email to connect@altair.com.
In addition, the following countries have resellers for Altair Engineering: Colombia, Czech Republic,
Ecuador, Israel, Russia, Netherlands, Turkey, Poland, Singapore, Vietnam, Indonesia
Official offices with resellers: Canada, China, France, Germany, India, Malaysia, Italy, Japan, Korea,
Spain, Taiwan, United Kingdom, USA
See www.altair.com for complete contact information.

Altair HyperStudy 2019 Tutorials p.1
Contents
Introductory Studies ............................................................................... 3
HS-1010: Simple Study Setup ............................................................................ 4
HS-1015: Working with an Internal Math Model ................................................... 10
HS-1020: Working with a Parameterized File Model for Size Variables ..................... 14
HS-1021: Working with a Parameterized File Model for Shape Variables .................. 21
HS-1025: Working with a HyperMesh and HyperMorph Model ................................. 28
HS-1030: Parameterize a MotionView Model ........................................................ 34
HS-1035: Optimization Study Using an Excel Spreadsheet ..................................... 40
HS-1036: Setting Up an Excel Model with a Visual Basic Script ............................... 47
HS-1040: Minimization of Internal Rosenbrock Function ........................................ 50
HS-1050: Minimization of External Rosenbrock Function ........................................ 55
HS-1060: Linking Variables of a Model to Output Responses of Other Models ........... 60
HS-1070: Defining Discrete Size Variables with Conditional Linking for use in
HyperStudy .................................................................................................... 67
HS-1080: Setting Up an Operator Model ............................................................. 74
HS-1090: Defining Discrete Size Variables with the Lookup Model .......................... 82
Advanced Solvers .................................................................................. 90
HS-1506: Material Calibration with a Curve Difference Integral .............................. 91
HS-1515: Using HyperView Results Math Integration for Output Response Extraction
................................................................................................................... 103
HS-1535: Coupling HyperStudy, Morphing and AcuSolve (HyperStudy Job Launcher)
................................................................................................................... 108
HS-1540: Shape Optimization Study Using HyperMesh and ANSYS ....................... 119
HS-1545: DOE and Optimization Study with HyperMesh, HyperStudy, and ANSYS CFX
................................................................................................................... 129
HS-1550: Shape Optimization Study Using HyperMesh and Abaqus ...................... 155
HS-1560: Study Setup Using LS-DYNA Model Parameters in HyperMesh ................ 169
HS-1070: Defining Discrete Size Variables with Conditional Linking for use in
HyperStudy .................................................................................................. 178
HS-1580: DOE Study Using CATIA ................................................................... 185
HS-1590: DOE Study Using Madymo ................................................................ 194
HS-1600: Creating an Optimization Study with MADYMO/Workspace Objective Rating
and HyperStudy ............................................................................................ 200
HS-1605: Setting Up an ANSYS Workbench Model .............................................. 209
HS-1610: Setting Up a SimLab Model Using JavaScript ....................................... 211
HS-1615: Setting Up a FEKO Model .................................................................. 217
HS-1620: Settings Up a Flux Model .................................................................. 223
Approach Overview ............................................................................. 228
HS-1700: Simple DOE Study ........................................................................... 229
HS-1705: Simple Fit Study .............................................................................. 231
HS-1710: Simple Optimization Study ................................................................ 234
HS-1715: Simple Stochastic Study ................................................................... 237
HS-1810: Post Proceesing with HyperStudy ....................................................... 239
Design of Experiment (DOE) ............................................................... 251
HS-2000: DOE Method Comparison: Arm Model Study ........................................ 252
HS-2005: DOE Study of a Cantilever Beam Using Templex .................................. 267
HS-2200: Using Existing Design Data from an Excel Spreadsheet ......................... 274
HS-2201: Using Existing Design Data from an Excel Spreadsheet with a Lookup Model

................................................................................................................... 279
HS-2210: Principle Component Analysis of a Cantilever Ibeam ............................. 283
HS-2215: Extensible DOE Study of a Space Frame Structure with Input Variable
Constraints ................................................................................................... 290
Fit ....................................................................................................... 301
HS-3000: Fit Method Comparison: Approximation on the Arm Model ..................... 302
HS-3005: Exporting Fit Models to Excel............................................................. 309
HS-3010: Fuselage Sizing Trade-Off using Categorical Variables ........................... 315
HS-3015: Automated Fit from CSV Data ........................................................... 325
Optimization ....................................................................................... 331
HS-4000: Optimization Method Comparison: Arm Model Shape Optimization .......... 332
HS-4200: Material Calibration Using System Identification ................................... 343
HS-4205: Multi-Objective Optimization Study Using an Excel Spreadsheet ............. 356
HS-4210: Multi-Disciplinary Optimization Study ................................................. 358
HS-4215: Multi-Disciplinary Design Optimization Study ....................................... 365
HS-4220: Size Optimization Study on an Impact Simulation Using RADIOSS .......... 373
HS-4230: Optimization Study with Discrete Variables ......................................... 384
HS-4405: Reliability-Based Optimization Study on an Impact Simulation Using
RADIOSS ...................................................................................................... 391
HS-4415: Optimization Study of a Landing Beam Using Excel .............................. 399
HS-4420: Optimization Study of a Spherical Impactor ......................................... 404
HS-4425: Multi-Objective Shape Optimization Study ........................................... 424
HS-4230: Optimization Study with Discrete Variables ......................................... 426
HS-4450: Multi-Objective Optimization of a Cantilever Ibeam using an Inclusion Matrix
................................................................................................................... 433
Stochastic ........................................................................................... 441
HS-5000: Stochastic Method Comparison and Reliability- Based Design Optimization of
the Arm Model .............................................................................................. 442

Introductory Studies

HS-1010: Simple Study Setup
In this tutorial you will learn how to setup a study on simple functions defined using a Templex
template. The base input template defines two input variables, DV1 and DV2, labeled X and Y,
respectively. The objective of this study is to investigate the two random variables X, Y
forming the two functions X+Y and 1/X + 1/Y – 2.
This tutorial starts HyperStudy from HyperMesh Desktop > TextView. You can also start
HyperStudy from HyperView, MotionView or directly in standalone mode.
The sample base input template you will use in this tutorial can be found in <hst.zip>/HS-
1010/. Copy the tutorial file from this directory to your working directory.
Step 1: View the Base Input Template in HyperMesh Desktop
1. Start HyperMesh Desktop.

2. On the Client Selector toolbar, select TextView.
3. On the Standard toolbar, click .

4. In the Open Document dialog, open the Simple.tpl file. The text editor displays the
following Templex statements:
{parameter(DVAR1,"Area
1",.5,0.2,5)}
{parameter(DVAR2,"Area
2",.5,0.2,5)}
{RES = DVAR1 + DVAR2}
{CON = 1/DVAR1 + 1/DVAR2 - 2}
{RES}
{CON}
{DVAR1}
{DVAR2}

6. On the Text toolbar, click . The text editor evaluates the Templex statements, replaces
the parameters with their initial values, and displays the following results:
1
2
0.5
0.5
7. Start HyperStudy by clicking Applications > HyperStudy from the menu bar.
Step 2: Perform the Study Setup
1. To start a new study, click File > New from the menu bar, or click on the toolbar.
2. In the HyperStudy – Add dialog, enter a study name, select a location for the study, and
click OK.
3. Go to the Define Models step.
4. Add a Parameterized File model.
a. From the Directory, drag-and-drop the simple.tpl file into the work area.
b. In the Solver input file column, enter res. This is the name of the solver input file
HyperStudy writes during the evaluation.
c. In the Solver execution script column, select Templex (templex).
5. Click Import Variables. Two input variables are imported from the Simple.tpl
resource file.
6. Go to the Define Input Variables step.
7. Review the input variable's lower and upper bound ranges.

9. Optional. Access additional editing and visualization features from the context menu by
right-clicking anywhere in the work area.
10. Go to the Specifications step.

Step 3: Perform the Nominal Run
1. In the work area, set the Mode to Nominal Run.

2. Click Apply.
3. Go to the Evaluate step.
4. Click Evaluate Tasks. An approach/nom_1/ directory is created inside the study
directory. The approaches/nom_1/run 00001/m_1 directory contains the .res file, which
is the result of the nominal run.
5. Go to the Define Output Responses step.
Step 4: Create and Define Output Responses
In this step you will create two output responses.

1. Create output response 1.
a. From the Directory, drag-and-drop the .res file, located in
approaches/nom_1/run_00001/m_1, into the work area.
b. In the File Assistant dialog, set the Reading technology to Altair®

HyperWorks® and click Next.

d. Select Single item in a time series, then click Next.

e. Define the following options, then click Next.
• Set Type to unknown.
• Set Request to Block 1.
• Set Component to Column 1.
f. Optional. Enter labels for the data source and output response.
g. Set Expression to First Element. The expression changes to m_1_ds_1[0].

Note: Because there is only a single value in this data source, [0] is inserted after
m_1_ds_1, thereby choosing the first (and only) entry in the data source.
i. Click Finish. Output response 1 is added to the work area.

2. Create output response 2 by repeating step 1.
3. In the Expression field for Response 2, select the second value by changing the [0]
to [1] after m_1_ds_2.
4. Click Evaluate to extract the output response values.

5. Proceed to the desired study type: DOE, Optimization, of Stochastic study.
Last modified: v2017.2 (12.1156684)

HS-1015: Working with an Internal Math Model
In this tutorial you will learn how to setup a study on simple functions defined using an
Internal Math model. In this study you will set up a beverage can design of experiments to
see how input variables effect the output responses. The beverage can has two input
variables: Diameter and Height and two output responses: Material Cost and Volume.
1. Start HyperStudy.
click OK.
5. Add an Internal Math model.
a. Click Add Model.
b. In the Add - HyperStudy dialog, select Internal Math and click OK.
7. Create two input variables.
a. Click Add Input Variable twice.
b. In the work area, Label column, change the labels for the two input variables to
Diameter and Height.
c. Change both input variable's lower, initial, and upper bounds to the values indicated in
the image below.

2. Click Apply.
directory.

In this step you will create and define the output responses, Cost and Volume.
1. Create two output responses.
a. Click Add Output Response two times.
b. In the work area, Label column, change the labels for the output responses to Cost
and Volume.
2. In the Expression column, enter the following:
a. For Cost, enter 2*(pi*var_1^2/4)+var_2*pi*var_1.
b. For Volume, enter (pi*var_1^2/4)*var_2.
Step 4: Run a DOE Study
1. In the Explorer, right-click and select Add from the context menu.
2. In the Add - HyperStudy dialog, select DOE and click OK.
4. In the work area, set the Mode to Full Factorial.
5. Click Apply.
7. Click Evaluate Tasks.
8. Go to the Post-Processing step.
9. Click the Summary tab.
Note that there are two input variables with lower and upper bounds which result in 22
= 4 runs.

10. Click the Linear Effects tab.

The data collected in the Summary tab is used to calculate the linear effects of the
Diameter and Height input variables on the Cost and Volume output responses. A line is
drawn between the average value of the output response when the input variable is at its
lower bound and the average value of the output response when the input variable is at
its upper bound.
Effects computation of Diameter on Cost and Volume
The effects of the input variable Height on the output responses Cost and Volume are
computed in the same manner. By displaying both input variables and output responses in
the same plot, you can compare the effects.
The slope of the lines could be positive or negative. In this example, both effects have a

positive slope which indicates that increasing the input variable's values will also increase
the output responses.
Last modified: v2017.2 (9.1144596)

HS-1020: Working with a Parameterized File Model for Size

Variables
In this tutorial you will learn how to create a template file for size variables, and how to import
them to HyperStudy. The input variables are the thickness of each of the three components,
defined in the input deck via the PSHELL card. The thickness should be between 0.05 and
0.15; the initial thickness is 0.1 (shown below).
The sample base input template used in this tutorial can be found in <hst.zip>/HS-1020/.
Copy the tutorial files from this directory to your working directory.
Figure 1: Double Symmetric Plate Model
click OK.
5. Click Add Model.
6. In the Add - HyperStudy dialog, select Parameterized File and click OK.
7. In the Resource column, click .

8. In the Open File dialog, open the plate.fem file. A HyperStudy dialog opens, informing
you that there are no parameters found in the file.

10. Click Yes to parameterize the file. The plate.fem file opens in the Editor.
11. In the Find area, enter PSHELL.
12. Click three times. Three PSHELL cards containing the shell thickness of each component
in the fourth field are highlighted.
Each field is eight characters long. The numbers need to be replaced with formatted
Templex statements to mark an input variable.
13. Starting at 0.1, highlight the first eight fields for thickness.
Tip: Quickly highlight 8-character fields by pressing CTRL to activate the Selector
(set to 8 characters) and then clicking the value.
14. Right-click on the highlighted fields and select Create Parameter from the context menu.
15. In the Parameter - varname_1 dialog, Label field, enter t1.
16. In the Format field, enter %8.5f.
17. Click OK.

19. Click OK to close the Editor.

20. In the Save Template dialog, save the template file as plate.tpl. HyperStudy
automatically populates the Resource column with the plate.tpl file, and the Solver
input file column with the solver input file name plate.fem.
21. In the Solver execution script column, select OptiStruct (os).
22. Click Import Variables. One input variables is imported from the plate.tpl resource
file.
25. Go to the Specification step.

2. Click Apply.

In this step you will create two output responses: Mass and Displacement.
1. Create the Mass output response.
a. From the Directory, drag-and-drop the plate.out file, located in


c. Select Single item in a time series, then click Next.

d. Define the following options, then click Next.
• Set Type to Mass.
• Set Request to Mass.
• Set Component to Value.
e. Label the output response Mass.

f. Set Expression to First Element.

g. Click Finish. The Mass output response is displayed in the work area.
2. Create the Displacement output response.
a. From the Directory, drag-and-drop the plate.h3d file, located in
d. Define the following options, and then click Next.
• Set Subcase to Subcase 1 (Load).
• Set Type to Displacement (Grids).
• Set Request to N298.
• Set Component to MAG.
e. Label the output response Displacement.

h. Click Finish. The Displacement output response is added to the work area.
4. Proceed to the desired study type (DOE, Optimization, of Stochastic study).
Last modified: v2017.2 (12.1156684)

HS-1021: Working with a Parameterized File Model for

Shape Variables
In this tutorial you will learn how to create a template file for shape variables and how to
import them to HyperStudy. The input variables are three shape variables; xtrans, ytrans and
radius. Each of these shapes are created by perturbing the mesh in the corresponding
direction by 1 unit.
Figure 1: Double Symmetric Plate Model Figure 2: Double Symmetric Plate Model with Shape Vectors
Step 1: Exporting Shape Variables from HyperMesh

2. In the User Profiles dialog, set the user profile to OptiStruct.
3. From the menu bar, click File > Open > Model.
4. In the Open Model dialog, open the plate_with_shapes.hm file. A model appears in the
graphics area.
5. From the Analysis page, click optimization.

7. Click shape.
8. Go to the desvar subpanel. In this subpanel you will create three design variables
(XTrans, YTrans and Rad) for three shapes.
9. Create a design variable.
a. In the desvar= field, enter XTrans.
b. Click Shape=.
c. Select the shape, xtrans.
d. Click create.
10. Create two more design variables labeled YTrans and Rad. Select the shape ytrans for
design variable YTrans, and select the shape radius for design variable Rad.
11. Go to the export subpanel to export the shape variables.
12. Set analysis code to HyperStudy.
13. Set sub-code to OptiStruct.
14. Click export as.
15. In the Save As dialog, save the file as plate_with_shapes.shp.

16. Quit HyperMesh by clicking File > Exit from the menu bar.

Step 2: Create the Base Input Template in HyperStudy
1. Start HyperStudy
2. From the menu bar, click Tools > Editor. The Editor opens.
3. In the File field, open the plate_with_shapes.fem file.
4. In the editor, right-click and select Select Nodes > GRID from the context menu. All of
the GRID cards in the plate_with_shapes.fem file highlight.
5. Right-click on the highlighted cards and select Include Shape from the context menu.
6. In the Shape Template dialog, open the plate_with_shapes.optistruct.node.tpl
file.
7. Click Save.
8. In the Save Template dialog, save the file as plate_with_shapes.tpl.
9. Close the Editor.
In this step, you will import the design variables (known as input variables in HyperStudy)
created in Step 1: Exporting Shape Variables from HyperMesh to HyperStudy.
click OK.
a. From the Directory, drag-and-drop the plate_with_shapes.tpl file into the work
area.

c. In the Solver input file column, enter plate.fem. This is the name of the solver
input file HyperStudy writes during any evaluation.
d. In the Solver execution script column, select OptiStruct (os).
5. Click Import Variables. Three input variables are imported from the
plate_with_shapes.tpl resource file.
Step 4: Perform a Nominal Run or a System Bound Check
1. In the work area, set the Mode to Nominal Run or System Bound Check.
A Nominal Run performs one run, and sets the input variable's values to their initial
values.
A System Bound Check performs three runs, and sets all of the input variable's values to
their initial, lower bound and upper bound values.
2. Click Apply.




g. Click Finish. The Mass output response is added to the work area.

Last modified: v2017.2 (12.1156684)

HS-1025: Working with a HyperMesh and HyperMorph Model
In this tutorial you will learn how to import size and shape variables to HyperStudy from
HyperMesh. The input variables are three shape variables; xtrans, ytrans and radius. Each of
these shapes are created by perturbing the mesh in the corresponding directions by 1 unit.
Figure 1: Double Symmetric Plate Model Figure 2: Double Symmetric Plate Model with Shape Vectors
click OK.
5. Add a HyperMesh model.
a. From the Directory, drag-and-drop the plate_with_shapes.hm into the work area.

6. Click Import Variables.

7. In the Model Parameters dialog, select parameters to import into HyperStudy.
a. Select the thickness and shape variables. A total of 6 parameters should be selected.
b. Click Add.
c. Click OK.



2. Click Apply.



g. Click Finish. The Displacement output response is added to the work area.

Last modified: v2017.2 (12.1156684)

HS-1030: Parameterize a MotionView Model
In this tutorial you will learn how to use HyperStudy to perform an optimization with
MotionSolve. The input variable is the angle q (swing angle) of the pendulum. The output
response target is to achieve Y-velocity of 6m/s at the tip of the pendulum. At the end of this
tutorial, you will know how to:
• Use MotionView to start HyperStudy and create the input variables.
• Setup a study.
• Run a system identification optimization study.
The files used in this tutorial can be found in <hst.zip>/HS-1030/. Copy the tutorial files from
this directory to your working directory.
click OK.
5. Add a MotionView model.
a. From the Directory, drag-and-drop the MotionView (.mdl) file Pendulum.mdl into the
work area.

c. In the Solver input file column, enter m1.xml. This is the name of the solver input
file HyperStudy writes for any evaluation.

7. In the Model Parameter Tree dialog, select parameters to import into HyperStudy.
a. Expand SolverVariables > theta > value, and select lin (scalar value for the swing
angle).
b. Click Add.
c. Click OK.
9. In the work area, change the Lower Bound to 0 and the Upper Bound to 2.


2. Click Apply.
In this step you will create one output response.

1. From the Directory, drag-and-drop the m1.mrf file, located in
2. In the File Assistant dialog, set the Reading technology to Altair® HyperWorks®
and click Next.
3. Select Single item in a time series, then click Next.
4. Define the following options, and then click Next.
a. Set Type to Marker Velocity.
b. Set Request to REQ/70000002 tip velocity- (on Pendulum body).
c. Set Component to VY.

5. Optional. Enter labels for the data source and output response.
6. Set Expression to Maximum.
7. Click Finish. The output response is displayed in the work area.
8. Click Evaluate to extract the output response value.

Step 4: Run an Optimization Study
2. In the Add - HyperStudy dialog, select Optimization and click OK.
3. Go to the Select Input Variables step.
5. Go to the Select Output Responses step.
6. Apply an objective on Response 1.
a. In the Objectives column of Response 1, click .
b. In the pop-up window, define the following settings and click OK.
• Set Type to System Identification.
• For Target Value, enter 6.0.
7. Click Apply.
9. In the work area, set the Mode to Adaptive Response Surface Method (ARSM).
Note: Only the methods that are valid for the problem formulation are enabled.
10. Click Apply.
12. Click Evaluate Tasks to start the optimization.
Step 5: View the Iteration History of an Optimization Study
1. Click the Iteration History tab to view a table with the Optimization's iteration results.
The optimal design is highlighted in green.
2. Click the Evaluation Plot tab to compare all of the entities of the Optimization (input
variables, output responses, and objectives) against the iteration.
Use the Channel selector to select all of the input variables, output responses, and
objectives.

Step 6: Post-Processing of an Optimization Study
The Post-Processing step in an optimization approach offers additional tools to review the
results. Statistics, histograms, and scatter plots can be used to help compare and analyze
designs.
Click the Integrity tab to view a series of statistical measures on input variables and output
responses.
Last modified: v2017.2 (12.1156684)

HS-1035: Optimization Study Using an Excel Spreadsheet
In this tutorial, you will learn how to:

• Couple HyperStudy with a spreadsheet.
• Identify input variables and output responses.
• Set up an Optimization study.
• Analyze Optimization results.
Problem formulation:
• Find the cross-sectional dimension's width and height in mm.
• Minimize the beam volume such that the tip deflection < 0.53 mm.
The Excel spreadsheet used in this tutorial can be found in <hst.zip>/HS-1035/. Copy the file
from this directory to your working directory.
Step 1: Review the Excel Spreadsheet
1. In Excel, open the hst_tut_1035(1070)_spreadsheet.xls file.

2. Review the information, and locate the columns that contain the input variables and
output responses.
Note: When you create an Excel spreadsheet model, it is important that the
spreadsheet is formatted correctly. A variable's value and label can be formatted
in two consecutive rows or two consecutive columns. Variable labels should only
contain English characters, or a combination of English characters and numbers.
If a label is not created for a variable, HyperStudy will assign one by default.
click OK.
5. Add a Spreadsheet model by dragging-and-dropping the
hst_tut_1035(1070)_spreadsheet.xls file from the Directory into the work area.

The Resource, Solver input file, and Solver input arguments fields become
populated. The Solver input file field displays hst_input.hstp, this is the name of the
solver input file HyperStudy writes during an evaluation.
6. Click Import Variables. The hst_tut_1035(1070)_spreadsheet.xls spreadsheet opens.

7. Add input variables.
a. In the Excel - HyperStudy dialog, click Yes to begin selecting input variables.
b. In the spreadsheet, select the cells that contain the input variable's labels and values.

c. In the Excel - HyperStudy Input selector dialog, click OK.

d. Click Cancel to stop selecting input variables.
8. Add output responses.
a. In the Excel - HyperStudy dialog, click Yes to begin selecting output responses.
b. In the spreadsheet, select the cells that contain the output response's labels and
values.
c. In the Excel - HyperStudy Output selector dialog, click OK.

d. Click Cancel to stop selecting output responses. Two input variables and two output
responses are imported from the hst_tut_1070_spreadsheet.xls spreadsheet.

2. Click Apply.
directory. The approaches/nom_1/run 00001/m_1 directory contains the sse_output.csv
file, which is the result of the nominal run.

Review the output responses imported into the study. The output responses were extracted
from the hst_output.hstp file, which HyperStudy created for each run.
4. Apply a set range of +17% to both input variable's lower and upper bounds.
a. In the Lower Bound column of both input variables, click .

b. Under Set Range, in the Percent field, enter +17.
c. Click the +/- button.
d. Click Apply.

6. Add an objective to the Beam Volume (m^3) output response.
a. In the Objectives column of Beam Volume (m^3), click .
b. In the pop-up window, set Type to Minimize and click OK.
7. Add a constraint on the Deflection at the tip (mm) output response.

a. In the Constraints column of Deflection at the tip (mm), click .
b. In the pop-up window, define the following and click OK.
• Set Bound Type to <= (less than or equal to).
• For Bound Value, enter 0.53.

8. Click Apply.
11. Click Apply.
14. Optional.
• To stop the optimization, click .

• To monitor the progress of the Optimization, click the Task tab.
Step 6: View the Evaluation Data and Plot of an Optimization

Approach
Use the Evaluation Data and Evaluation Plot tabs to review the values of input variables,
output responses, objective functions, and constraints for all runs evaluated during the
optimization.

Step 7: View the Iteration History and Plot of an Optimization

Approach
Use the Iteration History and Iteration Plot tabs to review the values of input variables,
output responses, objective function, and constraints for each iteration during the
optimization.
1. Click the Iteration History tab to view the iteration history results in a table.
As this study was run with ARSM, you will see the same designs in both the Evaluation
Data and Iteration History. From the iteration table, you can see that iterations 1, 2, and
5 are displayed in a red font, which indicates that these iterations have a constraint
violation. The constraint that is violated, Constraint 1, is displayed in a red, bold font.
Iterations 3, 4, and 6-8 are feasible, but not optimal designs. The ninth iteration,
highlighted in green, indicates that this design is the optimal design.
2. Click the Iteration Plot tab to view the iteration history results in a plot. Use the
Channel selector to select Objective 1 and Constraint 1.

4. To view a single plot that contains the iteration history of Objective 1 and Constraint 1,
click . In the Objective plot, infeasible designs are identified with bigger markers. In the
Constraint plot, you can see these designs have higher displacement value than the
constraint bound of 0.53 and only the last three designs meet the constraint bound.
Last modified: v2017.2 (12.1156684)

HS-1036: Setting Up an Excel Model with a Visual Basic

Script

• Couple HyperStudy with a spreadsheet containing Visual Basic scripts.
• Identify input variables and output responses.
The Excel spreadsheet used in this tutorial can be found in <hst.zip>/HS-1036/. Copy the file
from this directory to your working directory.
Step 1: Examine the Excel File
1. In Excel, open the iBeam_withMacros.xls file. This spreadsheet performs calculation for a
top loaded cantivlevered beam with an “I” cross section. The input variables and output
responses are labeled for clarity.
Note: When you create an Excel spreadsheet model, it is important that it is formatted
correctly. A variable's value and label can be formatted in two consecutive rows
or two consecutive columns. Variable labels should only contain English
characters, or a combination of English characters and numbers. If a label is not
created for a variable, HyperStudy will assign one by default.
2. In the cell to the right of Web Thick, change the value.
3. To update the calculated output responses, click Run macro named PerformCalcs. The
formulas and operations to calculate the output response values are contained in a Visual
Basic Script macro called “Perform Calculations”. When a change is made to one of the
input variables, it is not reflected in the output responses until you run the script.
Note: To run the Visual Basic script macro, you must enable macros in Excel.
4. Save any changes you made to the spreadsheet.
click OK.
5. Add a Spreadsheet model.
a. From the Directory, drag-and-drop the iBeam_withMacro.xls file into the work
area. The Solver input file field displays hst_input.hstp, this is the name of the

c. In the Solver input arguments column, enter PerformCalcs. This is the name of
the Visual Basic script.
6. Click Import Variables. The iBeam_withMacro.xls spreadsheet opens.

c. In the Excel - HyperStudy Input selector dialog, click OK.

d. Click Cancel to stop selecting input variables.

b. In the spreadsheet, select the cells that contain the output response's labels and
values.

d. Click Cancel to stop selecting output responses. Four input variables and four output
responses are imported from the iBeam_withMacro.xls spreadsheet.
Step 3: Perform the System Bounds Check
1. In the work area, set the Mode to System Bounds Check.

2. Click Apply.
directory. The approaches/nom_1/run 00001/m_1 directory contains the
hst_output.hstp file, which is the result of the nominal run.
5. When the evaluation is complete, click the Evaluation Data tab.
6. Examine the run data to ensure that the output response values changed in each
evaluation.
Last modified: v2017.2 (12.1156684)

HS-1040: Minimization of Internal Rosenbrock Function
In this tutorial you will learn how to register a Compose/OML or python function in HyperStudy
using the Preference file (.mvw), and then use the registered function for output response
evaluation in the study. The Rosenbrock function is defined as a python or OML script and
registered in HyperStudy.
The example defines two input variables labeled x and y, respectively. The objective of the
optimization is to minimize f(x,y)= 100*(y-x^2)^2 + (1-x)^2. The range for x and y is set to
[-2 ; -2] , and the start point is [-1 ; -1].
Step 1: Define the Rosenbrock Function
Define the Rosenbrock Function with:
Compose:
1. Start Compose.
2. From the menu bar, click File > New > OML File.
3. In the editor, enter the following OML commands:
function f = ros_eval(x,y)
f = 100.0*(y-x^2)^2 + (1-x)^2
4. Form the menu bar, click File > Save As.

5. In the Save As dialog, navigate to your working directory and save the file as
rosenbrock_function.oml.
6. Quit Compose by clicking File > Exit from the menu bar.
Tip: Avoid the subsequent preference file steps when using the Register Function … feature
in Compose with OML functions. Highlight the function name, then right click to access
this feature on the context menu.
Or
Python:
1. In your favorite text editor, enter the following python commands:
def ros_eval(x, y):

return 100*(y-x*x)*(y-x*x) + (1-x)*(1-x)
2. Form the menu bar, click File > Save As.

rosenbrock_function.py.

Step 2: Add the Function to the Preference File
Add the OML function to the Preference file:

1. In a text editor, enter the following OML registration function:
*Id("HyperGraph v11.0")
*BeginDefaults()
*BeginPlotDefaults()
*SetOMLRootDir("C:/Program Files/Altair/2019/Compose2019")
*RegisterOMLFunction("ros_eval",”<path>/rosenbrock_function.oml",2)
*EndPlotDefaults()
*EndDefaults()
Note: Replace <path> with the actual location of the file, and modify the
*SetOMLRootDir statement to point to the appropriate location.
2. Save the file as a preference file with the name rosenbrock_prefs.mvw file.
3. Close the text editor. or
Add the python function to a Preference file:

1. In a text editor, enter the following pyton registration function:
*Id("HyperStudy v14.0")
*BeginDefaults()
*RegisterPythonFunction("ros_eval","<path>/rosenbrock_function.py",
2)
*EndPlotDefaults()
*EndDefaults()
2. Replace <path> with the actual location of the .py file.

3. Save the file as a preference file with the name rosenbrock_prefs.mvw file.
4. Close the text editor.

3. From the menu bar, click File > Use Preferences File.
4. In the HyperStudy - Set Preference File dialog, open the rosenbrock_prefs.mvw
file.
click OK.
a. Click Add Model.
b. In the Add - HyperStudy dialog, select Internal Math and click OK.
a. Click Add Input Variable twice.
b. In the work area, label the input variables X and Y.
c. Change both input variable's lower, initial and upper bounds to the values indicated in
the image below.

2. Click Apply.
directory.
1. Click Add Output Response.
2. In the Expression column of Response 1, click .

4. In the Expression Builder, click the Functions tab.

5. From the list of available functions, select ros_eval.
6. Click Insert Varname. The function ros_eval() appears in the Evaluate Expression
field.
7. Click the Input Variables tab.
8. In the work area, select the input variables X and Y.
9. Click Insert Varname. The input variables appear in the expression as
ros_eval(var_1, var_2).
10. Click Evaluate Expression. The expression ros_eval(var_1, var_2) changes to 404.
11. Click OK. This completes the study setup.
4. Add an objective to Response 1.
5. Click Apply.
8. Click Apply.

12. Optional. Click the Iteration Plot tab to monitor the progress of the optimization.
The iteration history shows a significant reduction in the objective value. The Rosenbrock
function has a global minimum that is difficult for any optimizer to find due to its flatness
in the area of the true optimum, and ARSM has not found the theoretical solution at
(x,y)=(1,1).
Last modified: v2017.2 (12.1156684)

HS-1050: Minimization of External Rosenbrock Function
In this tutorial, you will learn how to use Compose or Python within an Optimization study. The
example consists of optimizing a 2-dimensional Rosenbrock function. You will be using either
Compose or Python as the solvers for HyperStudy. This example defines two input variables,
labeled x and y, respectively. The objective of the optimization is to minimize f(x,y)=
100*(y-x^2)^2 + (1-x)^2. The range for x and y is set to [-2 ; -2], and the start point is
[-1 ; -1].
click OK.
3. Go to the Define models step.
a. From the Directory, drag-and-drop the appropriate .tpl file into the work area.
• If you are using Python, use the rosenbrock_py.tpl file.
• If you are using Compose, use the rosenbrock_oml.tpl file.
b. In the Solver input file column, enter a name for the solver input file HyperStudy
writes during any evaluation.
• If you are using Python, enter rosenbrock.py.
• If you are using Compose, enter rosenbrock.oml.
c. In the Solver execution script column, select either Python (py) or Compose
(oml) accordingly.
Note: If you are using Compose as part of the HyperWorks suite, then HyperStudy
should automatically point to the correct .bat file. If you have Compose as
separate installation, then during the Register Solver Script step you
should point to Compose_batch.bat.
d. If you are using Compose as the Solver execution script, in the Solver input
arguments column, enter -f infront ${file}.

Note: For solver scripts running on Linux, enter "-f ${file} -nobg" in the Solver
input arguments column to ensure that the Compose batch mode runs in
the foreground instead of the background.
5. Click Import Variables. Two input variables are imported from the .tpl file.
7. Change both input variable's lower, initial and upper bounds to the values indicated in the
image below.

2. Click Apply.
In this step you will create one output responses.

1. Create the output response.
a. From the Directory, drag-and-drop the rosenbrock.res file, located in the
approaches/nom_1/run 00001/m_1 directory, into the work area.

e. Optional. Enter labels for the data source and output response.
f. Set Expression to First Element. The expression changes to m_1_ds_1[0].

g. Click Finish. Output response 1 is added to the work area. 2. Click

Evaluate. The value for expression m_1_ds_1[0] should be 404.0.
6. Add an objective to Response 1.
7. Click Apply.
10. Click Apply.
13. Optional. Click the Iteration Plot tab to monitor the progress of the optimization.
The iteration history shows a significant reduction in the objective value. The Rosenbrock
function has a global minimum that is difficult for any optimizer to find due to its flatness
in the area of the true optimum, and ARSM has not found the theoretical solution at
(x,y)=(1,1).

Last modified: v2017.2 (12.1156684)

HS-1060: Linking Variables of a Model to Output Responses

of Other Models
In this tutorial you will learn how to link variables of a model to output responses of other
models. The input variables are the thickness of each of the three components, defined in the
input deck via the PSHELL card. The thickness should be between 0.05 and 0.15; the initial
thickness is 0.1 (shown below).
The sample base input template files used in this tutorial can be found in <hst.zip>/HS-
1060/. Copy the file from this directory to your working directory.
1. Start HyperStudy
click OK.
a. From the Directory, drag-and-drop the plate.tpl file into the work area.

6. Add a Spreadsheet model by dragging-and-dropping Cost.xls file from the Directory
into the work area.
populated. The Solver input file field displays hst_input.hstp, this is the name of the
Note: When you create an Excel spreadsheet model, it is important that you format it
correctly. A variable's value and label can be formatted in two consecutive rows
or two consecutive columns. Variable labels should only contain English
characters, or a combination of English characters and numbers. If you do not
create a label for a variable, HyperStudy will assign one by default.
7. Click Import Variables. The cost.xls spreadsheet opens.


d. In the Excel - HyperStudy Input selector dialog, click OK.

e. Click Cancel to stop selecting input variables.
b. In the spreadsheet, select the cells that contain the output response's label and value.

d. Click Cancel to stop selecting output responses. Two input variables and one output
response are imported from the cost.xls spreadsheet.
11. Review the input variable's upper and lower bound ranges.

2. Click Apply.




g. Click Finish. The Mass output response is displayed in the work area.

Step 4: Linking Mass_Excel Input Variable of Model 2 to Mass Output

Response of Model 1
1. In the Explorer, click Define Input Variables.

2. Click the Links tab.
3. In the Expression column of the input variable Mass_Excel, click .
4. In the Expression Builder, click the Output Responses tab.

5. Select the output response Mass.
6. Click Insert Varname. The expression m_1_r_1 appears in the Evaluate Expression
field.

7. Click OK. The input variable Mass_Excel of Model 2 is now linked to the output response
Mass of Model 1.

9. In the work area, set the Mode to System Bounds Check.
10. Click Apply.
13. Click the Evaluation Data tab.
14. Verify that the input variable Mass_Excel is equal to the output response Mass.
Last modified: v2017.2 (12.1156684)

HS-1070: Defining Discrete Size Variables with Conditional

Linking for use in HyperStudy
This tutorial outlines the procedure for defining discrete input variables that are conditionally
linked. In this tutorial, beam thickness and material model values are used as variables.
Material model values are linked to the selection of the material to be used. The variables you
will use in this tutorial include: Shell Thickness, Young's Modulus, Poisson's, Ratio, and
Density. The output responses you will use in this tutorial includes: Mass.
The beam.fem model file that you will use in this tutorial can be found in <hst.zip>/HS-
1070/. Copy the tutorial files from this directory to your working directory.
Step 1: Create the Base Import Template in HyperStudy
2. From the menu bar, click Tools > Editor. The HyperStudy - Editor opens.
3. In the File field, open the beam.fem file.
5. Click until you find the PSHELL card.
6. In the same line as PSHELL, highlight the value 0.002 for the PSHELL thickness as
indicated in the image below.
Note: In an OptiStruct deck, each field within a card is 8 characters long. Properly select
the value for the PSHELL thickness by selecting 0.002 and the three spaces that
follow.
8. In the Parameter - varname_1 dialog, Label field, enter Thickness.
9. Set the Upper bound to 0.0022, the Nominal to 0.0020, and the Lower bound to
0.0018.
10. Set the Format to %8.5f.
11. Click OK.

13. Add three more input variables to the template file from the MAT1 card using the
information provided in the table below.
Parameter Selection Selection Input Lower Bound Format

Notification Variable & Upper
Label Bound
Young's Line number: 2.1E+11 & Young Lower Bound: %8.1e

Modulus 8267 1 space 7E+10
Fields: 16 to 24 Upper Bound:
2.1E+11
Poisson's Line number: 0.3 + 5 Poisson Keep default %8.5f

Ratio 8267 spaces values
Fields: 32 to 40
Density Line number: 7820.0 + 2 Density Keep default %8.3f

8267 spaces values
Fields: 40 to 48
14. Click Save.

15. In the Save Template dialog, save the file as beam.tpl.
16. Close the HyperStudy - Editor dialog.

18. In a text editor, open the beam.tpl file.

19. Replace:
{parameter(varname_3, "Poisson", 0.30000, 0.27000,

0.33000)}
{parameter(varname_4, "Density", 7820.000, 7038.000,
8602.000)}
with:
{if (varname_2==2.1E+11)}
{varname_3=0.3}
{varname_4=7820}
{elseif (varname_2==7E+10)}
{varname_3=0.33}
{varname_4=2700}
{endif}
20. Save the beam.tpl file and close the text editor.
click OK.
a. From the Directory, drag-and-drop the beam.tpl file into the work area.
b. In the Solver input file column, enter beam.fem. This is the name of the solver


5. Click Import Variables. Two input variables are imported from the beam.tpl resource
file.
7. Click the Modes tab.
8. For the input variable Thickness, set Mode to Discrete.
9. In the Values column for the input variable Thickness, click . A pop-up window opens.
10. In the Number of Points field, enter 3.
11. Click Set.
12. Change the values in the table to the following: 0.002, 0.004, and 0.005.
13. Click OK.

15. For the input variable Young, set the Mode to Discrete.
16. In the Value column of the input variable Young, click . A pop-up window opens.
18. Click Set.
19. Change the values in the table to 7E+10 and 2.1E+11.
20. Click OK.

2. Click Apply.
4. Click Evaluate Tasks. The run is executed using OptiStruct, and all input files are written
to the study directory.
In this step you will create one output response, Mass.

1. From the Directory, drag-and-drop the beam.out file, located in
and click Next.
4. Define the following options, then click Next.

5. Label the output response Mass.

6. Set Expression to First Element.
Note: Since the mass is a scalar quantity, the only component of the data source is
index zero, therefore a scalar mass value is properly written in the Expression
field as m_1_ds_1[0]. Multi-axis vectors may be accessed through each individual
axis by using the corresponding index in the vector array.
The mass value can be accessed directly out of the beam.out file and does not
require any mathematical operations before being used as an output response for
your study. The Expression Builder is capable of performing a full complement
of mathematical functions on an arbitrary number of vector solutions in order to
build an output response.

7. Click Finish. The Mass output response is added to the work area.
Last modified: v2017.2 (12.1156684)

HS-1080: Setting Up an Operator Model
In this tutorial you will use the Operator model type to run a script that uses a combination of
HyperView and HVTrans to split the solver result file in multiple result files, one for each
component in the model. This tutorial uses a model which consists of a plate with a hole which
is loaded in plane. The design has three thickness variables; one for each zone. The output
responses of interest are the maximum stress in each of the three zones.
1. Start HyperStudy
click OK.
b. In the Solver input file column, enter plate.fem. This is the name of the solver
c. In the Solver execution script column, select OptiStruct (os).
6. Click Import Variables. Three input variables are imported from the plate.tpl
resource file.


2. Click Apply.
Step 3: Setup the Operator Model
This step requires the plate.h3d file generated in step 2, which is why you cannot setup the
operator model until after the nominal run.
1. Go back to the Define Models step.
2. Add an Operator model.
a. Click Add Model.
b. In the Add - HyperStudy dialog, select Operator and click OK.
c. Set the Solver execution script to Python (py).
d. In the Solver input arguments field, enter ${m_2.file_1} ${m_2.file_3}.
The input arguments are references to the model resources' varnames. The first
argument (m_2.file_1) is a reference to the model resource’s varname, and tells
python which script to run. The second argument (m_2.file_3) is the varname to the
target result file to split, and will be the first argument to the python script.
3. Define model dependencies.
a. Click Model Resources. The Model Resources dialog opens.
b. Define a model dependency that references the python script that will be used as the
solver script. This is a reference to a file that is not generated during a solver run,
therefore it is of type Normal. This file does not need to be in the run directory.
i. Verify that Model 2 (m2) is selected.
ii. Click Add Resource > Add Input Resource.
iii. In the Select File dialog, navigate to your working directory and open the
hv_resultsbyComp.py file.
iv. Set Operation to None.
c. Define a model dependency that references the tcl script that will be used to run
HyperView and Hvtrans in batch. This is a reference to a file that is called by the
python script and not by a solver, therefore it is of type Normal. This file is required to
be in the run directory.


ii. Click Add Resource > Add Input Resource.
iii. In the Select File dialog, navigate to your working directory and open the
hv_resultsbyComp.tcl file.
iv. Set Operation to Copy.
e. Define a model dependency that will be used as a link to the result file from the first
model. This file is the target file for the tcl script, and it is required to be in the run
directory. It is a file that changes for each run, therefore it is a linked file and it is
copied in the run directory. Note that the file can be moved.
ii. Click Add Resource > Add Link Resource.
iii. In the Select File dialog, navigate to the approaches/nom_1/run 00001/m_1
directory and open the plate.h3d file.
iv. Set Operation to Copy.
f. Click Close to exit the Model Resources dialog.
In this step you will perform the same steps as Step 2, except during this nominal run the
Operator model will also be run.

3. Click Apply.
4. In the HyperStudy dialog, click Yes to overwrite the run matrix.
7. In the HyperStudy dialog, click Yes to overwrite files.
Step 5: Define Output Responses
In this step you will create four output responses: maxStressPart2, maxStressPart4,
maxStress3, Volume.
2. Create the maxStressPart2 output response.
a. From the Directory, drag-and-drop the plate_2_shell.h3d file, located in
HyperWorks® (Hyper3D Reader) and click Next.
c. Select Multiple items at multiple time steps, then click Next.
• Set Subcase to Subcase 1(Load).
• Set Type to Element Stresses (2D & 3D) (2D).
• For Request, set Start to First Request and End to Last Request.
• For Components, select vonMises (Z1).
• For Timestep, select Custom and enter 0.

e. Label the output response maxStressPart2.

f. Set Expression to Maximum.

h. Click Finish. The maxStressPart2 output response is displayed in the work area.
3. Create the maxStressPart4 output response.
a. From the Directory, drag-and-drop the plate_4_patch2.h3d file, located in
c. Select Multiple items at multiple time steps (readsim), then click Next.
• For Components, select vonMises (Z1).
e. Label the output response maxStressPart4.
g. Click Finish. The maxStressPart4 output response is added to the work area.
4. Create the maxStress3 output response.
a. From the Directory, drag-and-drop the plate_3_patch1.h3d file, located in

d. Select Multiple items at multiple time steps (readsim), then click Next.
e. Define the following options, and then click Next.
• For Components, select vonMises (Z1) and vonMises (Z2).
f. Label the output response maxStress3.
g. Set Expression to Maximum.
h. Click Finish. The maxStress3 output response is added to the work area.
5. Create the Volume output response.
HyperWorks® (osmass.tpl) and click Next.
• Set Type to OptiStruct Analysis.
• Set Request to Out File.
• Set Component to Volume.
e. Label the output response Volume.
g. Click Finish. The Volume output response is added to the work area.
Step 6: Define the DOE
2. In the Add - HyperStudy dialog, select Doe and click OK.


5. In the work area, set the Mode to Modified Extensible Lattice Sequence.
6. Click Apply.
10. Click the Pareto Plot tab to plot the effects of variables on output responses in
hierarchical order (highest to lowest).
Each variable contributes nearly equally to volume. A positive hashing indicates that the
relationship is positive: as the variable increases, mass increases. For the three stress
output responses, the maximum stress in each zone is dominated by the thickness of that
zone.
Last modified: v2017.2 (12.1156684)

HS-1090: Defining Discrete Size Variables with the Lookup

Model
In this tutorial you will define discrete size input variables with the Lookup model. You will
establish links between the input variables imported from a parameterized file with the output
responses imported from a .csv file using the Lookup model.
Model Files
The files used in this tutorial can be found in <hst.zip>/HS-1090/. Copy the beam.fem file and
the material_prop.csv file to your working directory.
Exercise
Step 1: Perform the Study Set Up
2. In the HyperStudy – Add dialog, enter a study name, select a location for the study,
and click OK.
a. In the work area, click Add Model.
b. In the Add – HyperStudy dialog, select Parameterized File and click OK.
c. In the Resource column, click .

d. In the Open File dialog, navigate to your working directory and open the
beam.fem file.
e. In the HyperStudy dialog, click Yes to parameterize the file.
f. In the Editor – HyperStudy dialog, Find area, enter PSHELL and click until
you find the PSHELL card.
g. In the same line as PSHELL, highlight the value 0.002 for the PSHELL thickness.
Note: In an OptiStruct deck, each field within a card is 8 characters long.
Properly select the value for the PSHELL thickness by selecting 0.002 and
the three spaces that follow.

h. Right-click on the highlighted fields and select Create Parameter from the
context menu.
i. In the Parameter - varname_1 dialog define the parameter and click OK.
i. In the Label field, enter Thickness.
ii. For the Upper bound, enter 0.0022.
iii. For the Nominal value, enter 0.0020.
iv. For the Lower bound, enter 0.0018.
v. In the Format field, enter %-8.5f.
j. Add three more input variables to the template file from the MAT1 card using
the information provided in the table below.
Tip: Quickly highlight 8-character fields by pressing CTRL to activate the
Selector (set to 8 characters) and then clicking the value.
Parameter Selection Selection Input Lower Format

Notification Variable Bound &
Label Upper
Bound
Young's Line number: 2.1E+11 & Modulus Lower %-8.1E

Modulus 8267 1 space Bound:
1.9E+011
Fields: 16 to
24 Upper
Bound:
2.3E+011
Poisson's Line number: 0.3 + 5 Poisson Lower %-8.2f

Ratio 8267 spaces Bound:
0.27
Fields: 32 to
40 Upper
Bound:
0.33

Density Line number: 7820.0 + Density Lower %-8.2f

8267 2 spaces Bound:
7038
Fields: 40 to
48 Upper
Bound:
8602
k. Click Save.
l. In the Save Template dialog, save the file as beam.tpl.
m. Click OK to close the Editor – HyperStudy dialog.
The Resource column displays the beam.tpl file, and the Solver input file
column displays beam.fem, which is the name of the solver input file HyperStudy
n. In the Solver execution script column, select OptiStruct (os).
5. Add a Lookup model.
a. In the work area, click Add Model.
b. In the Add – HyperStudy dialog, select Lookup and click OK.

d. In the HyperStudy – Load model resource dialog, navigate to your working
directory and open the material_prop.csv file.
7. In the Import Variables dialog, Number of design variables field, enter 1 and then
click OK.

9. Review the four input variables that were imported from the beam.tpl file in the
Parameterized File model, and the one input variable that was imported from the
material_prop.csv file in the Lookup model.
Notice: The label of fifth input variable has the same label as the first column in the
material_prop.csv file, that is Material.

The Lookup model automatically populates the input variables based on the number
you provided, and you can now identify the material by strings.

2. Click Apply.
3. Go to the Evaluate step and click Evaluate Tasks.

1. Review the three output responses that were automatically added to the study from in
Lookup model, which correspond to the nominal values from the material_prop.csv
file.
Notice: The output response labels are the same as the labels in the
material_prop.csv file.
2. Click the Data Sources tab.

3. Click Add Data Source to add two data sources.
4. Define Data Source 4.
a. In the File field for Data Source 4, click (…).
b. In the Data Source Builder dialog, File field, navigate to the
approaches\nom_1\run 00001\m_1 directory inside your working directory and
open the beam.h3d file.

d. Set Tool to File Source.

e. Set Subcase to Subcase 1 (Modal).
f. Set Type, Request, and Component to Frequency.
g. Click OK.
5. Define Data Source 5.

a. In the File field for Data Source 5, click (…).
b. In the Data Source Builder dialog, File field, navigate to the
approaches\nom_1\run 00001\m_1 directory inside your working directory and
open the beam.h3d file.
c. Set Tool to Read Simulation.
d. Set Subcase to Subcase 2 (Static).
e. Set Type to Displacement (Grids).
f. For Request, set Start to First Request and End to Last Request.
g. For Components, select MAG.
h. Set Time to All.
i. Click OK.

6. Click the Define Output Responses tab.

7. Define the 1st_natural_freq output response.
a. Click Add Output Response.
b. In the Label field, enter 1st_natural_freq.
c. In the Expression field, click (…).
d. In the Expression Builder dialog, click the Data Sources tab.
e. From the list of data sources, click Data Source 4.
f. Next to Insert Varname, click and select First Element.
g. Click Insert Varname.

The Evaluate Expression field displays ds_4[0].
h. Click OK.
8. Define the Max_disp output response.
b. In the Label field, enter Max_disp.
c. In the Expression field, click (…).
d. In the Expression Builder dialog, click the Data Sources tab.
e. From the list of data sources, click Data Source 5.
f. Next to Insert Varname, click and select Maximum.
g. Click Insert Varname.
The Evaluate Expression field displays max(ds_5).
h. Click OK.

9. Click Evaluate to extract the response values.
Step 4: Link the Input Variables and Output Responses

In this step you will establish links between the input variables imported from the beam.tpl
file in the Parameterized File model with the output responses imported from the
material_prop.csv file in the Lookup model.
3. In the Expression field for Modulus, click (…).
4. In the Expression Builder, click the Output Responses tab.
5. From the list of output responses, select Modulus.
6. Click Insert Varname.
7. Click OK to close the Expression Builder.
The Modulus input variable is now linked to the Modulus output response.
8. Link the Poissons input variable to the Poissons output response and the Density input
variable to the Density output response.
9. Go to the Specifications step and click Apply.

10. In the HyperStudy dialog, click Yes to overwrite the run matrix.
11. Go to the Evaluate step and click Evaluate Tasks to re-evaluate the setup.

2. In the Add – HyperStudy dialog, select DOE and click OK.

5. Verify that the Mode is set to Modified Extensible Lattice Sequence (Mels) and
that the Number of Runs is set to 6.
6. Click Apply.
Notice: The output responses (material property numbers) from the .csv file are
linked to the input variables (material property set in the FEA deck), and are
now controlled in the categorical input variable Material.
Any number of material data can be added using a library, without requiring
you to explicitly create “if” conditions in a .tpl file. This is the advantage of
using Lookup model in this case.

Advanced Solvers

HS-1506: Material Calibration with a Curve Difference

Integral
In this tutorial you will learn how to register a Compose, Python, or Templex function in
HyperStudy. The objective of this tutorial is to find the RADIOSS material parameter values so
that the stress-strain curve of the tensile test simulation matches the tensile test experimental
curve. HS-4200: Material Calibration Using System Identification provides an alternative
method to setup this problem using system identification.
By the end of this tutorial, you will know how to:
• Create an input template from a RADIOSS file using the HyperStudy - Editor
• Register a Compose, Python, or Templex function
• Setup a study
• Run a system identification optimization study
The sample base input template used in this tutorial can be found in <hst.zip>/ HS-1506/.
Copy all of the files from this directory to your working directory. The .py and .oml file contain
a function that calculates the area between two curves.
Description of the Model
A quarter of a standard tensile test specimen is modeled using symmetry conditions. A

traction is applied to a specimen via an imposed velocity at the left-end.
The units are: mm, ms, g, N, MPa.
Geometry of the Tensile Specimen (One Quarter of the Specimen is Modeled)
Sections of Node Saved for Time History

The material to be characterized is a 6063 T7 Aluminum. It has an isotropic elasto-plastic

behavior which can be reproduced by a Johnson-Cook model without damage (RADIOSS Block
Law2), defined as follows:
In this study, the parameters a, b, n, σmax (maximum stress), and the Young modulus are
defined as input variables. The stress-strain curve obtained by the experimental test is shown
in the following image.
Engineering Stress Versus Engineering Strain Curve (Experimental Data)
For the simulation results, engineering strains will be obtained by dividing the displacement of
node 1 by the reference length (75 mm), and engineering stresses will be obtained by dividing
the force in section 1 by its initial surface (10.2 mm2).

Engineering Stress Versus Strain Curve (Simulation Results)
In this step, you can create the base input template in HyperStudy or use the base input
template in the study Directory.
3. In the File field, open the TENSILE_TEST_0000.rad file.
4. In the Find area, enter /MAT/PLAS_JOHNS/1.
5. Click . HyperStudy highlights /MAT/PLAS_JOHNS/1.
6. Select variable E by highlighting the first 20 fields in row 51.


8. In the Parameter-varname_1 dialog, Label field, enter E_Young.
9. Set the Lower Bound to 50000, the Initial Bound to 60400, and the Upper Bound to
70000.
11. Click OK.

12. Define four more variables using the information provided in the table below.
Variabl Label Lower Initial Upper Format

e Bound Value Bound
a a_PlasticityYieldStress 90 110 120 %20.5f
b b_HardeningCoeff 100 125 160 %20.5f
n n_HardeningExpo 0.1 0.2 0.3 %20.5f
sigmax Sigma_Max 250 280 290 %20.5f
13. Click Save.

14. In the Save Template dialog, save the file as TENSILE_TEST_0000.tpl.

Step 2: Register the Curve Difference Function
The function curve_difference calculates the integral of the absolute value of the difference
between two curves over the common domain of the supplied functions. This is a robust
function that tends to zero only as the two functions become equal.
Register the curve_difference function using:
Compose
1. Start Compose.
2. From the menu bar, click File > Open.
3. In the Open File dialog, open the file Curve_Difference.oml.
4. Highlight the function curve_difference(...).
5. Right-click on the highlighted text and select Register Function
from the context menu.
6. In the Register Function dialog, click OK.
7. Close Compose.
Python
1. In your favorite text editor, open the hst_pref_py.mvw file.

2. Replace <path> with the actual location of the curve_difference.py file.
*Id("HyperStudy v11.0")
*BeginDefaults()
*RegisterPythonFunction("curve_difference",
"<path>/curve_difference.py", 1)
*EndPlotDefaults()
*EndDefaults()
3. Save the changes you made to the preference file.

5. In HyperStudy, from the menu bar, click File > Set Preferences File.
6. In the HyperStudy - Set Preferences dialog, navigate to your working directory and
open the hst_pref_py.mvw file.

Templex
1. From the menu bar, click File > Use Preferences File.
2. In the HyperStudy - Set Preference File dialog, open the curve_difference.mvw
file.
click OK.
a. From the Directory, drag-and-drop the TENSILE_TEST_0000.tpl file into the work
area.
b. In the Solver input file column, enter TENSILE_TEST_0000.rad. This is the name of
the solver input file HyperStudy writes during any evaluation.
c. In the Solver execution script column, select RADIOSS (radioss).
d. Verify that the Solver input arguments field reads $file. This solver input
argument runs the Starter and the Engine of RADIOSS for the crash analysis, and
also prevents the creation of the .h3d result file from animation files.
Note: X is the number of CPUs to use for the simulation.

5. Define a model dependency.

a. Click Model Resources.
b. In the Model Resource dialog, click Add Resource > Add Input Resource.
c. In the Select File dialog, navigate to your working directory and open the
TENSILE_TEST_0001.rad file.
d. Set Operation to Copy.
e. Click Close.
6. Click Import Variables. Five input variables are imported from the
TENSILE_TEST_0000.tpl resource file.
Step 4: Perform the Base Run

2. Click Apply.
4. Click Evaluate Tasks. An approaches/nom_1/ directory is created inside the study
TENSILE_TESTT01 file, which consist of the time history results of the simulation.

In order to fit the RADIOSS stress-strain curve to the experimental data, you must compare
the two curves. In this step, you will use the curve_difference function.
1. Create the Area Between Two Curves output response.
b. In the work area, Label field, enter Area Between Two Curves.
2. Create a data source labeled Disp_sim.
a. From the Directory, drag-and-drop the TENSILE_TEST01 file, located in
• Set Type to None/Node 1.
• Set Request to 4 Node 1.
• Set Component to DX-X Displacement.
e. Under Creating a new Data Source, enter Disp_sim in the Label field.
f. Clear the Linked to a new Response checkbox.
g. Click Finish

3. Create a second data source labeled Force_sim by repeating step 2, except select the
following options:
• Set Type to Section/SECTION_2.
• Set Request to 2 section 1.
• Set Component to FT-Resultant Tangent Force.
4. Create a third data source labeled Strain_exp.
a. In the Expression column of the Area Between Two Curves output response, click
.
b. In the Expression Builder, click the Data Sources tab.
c. Click Add Data Source. A new data source is added to the work area.
d. In the Label field for the new data source, enter Strain_exp.
e. In the File field, click .

f. In the Data Source Builder, define the data source and click OK.
• In the File field, navigate to your working directory and open the
experiment.xy file.
• Set Tool to File Source.
• Set Type to Unknown.
• Set Request to Block1.

5. Create a fourth file source labeled Stress_Exp by repeating step 4, except select the
following options:
• Set Request to Block1.
6. Define the Area Between Two Curves output response.
a. In the Expression Builder, click the Functions tab.
b. From the list of available functions, select curve_difference.
c. Click Insert Varname. The function curve_difference()appears in the Evaluate
Expression field.

e. In the Evaluate Expression field, edit the expression so that it reads

curve_difference(m_1_ds_1/75,m_1_ds_2/10.2,ds_3,ds_4).
Note: In the expression, you will notice that the four data sources are entered
differently. The varname for the first two data sources created using the File
Assistant are labeled as m_1_d_#, whereas the varname for the last two
data sources created in the Expression Builder are labeled as ds_#.
The displacements and forces are read from the simulation, whereas from the
experiment you have strains and stresses. In order to convert the displacement and
forces to strains and stresses, you need to divide the displacements by the length
(75) and forces by the area (10.2).
7. Click OK to close the Expression Builder.

6. Apply an objective on the Area Between Two Curves output response.
a. In the Objectives column for Area Between Two Curves, click .

7. Click Apply.
10. Click Apply.
13. Optional. Click the Evaluation Plot tab to plot the optimization iteration history of the
objective.
14. Optional. Click the Iteration History tab to review the iteration history in a table.
Last modified: v2017.2 (9.1144596)

HS-1515: Using HyperView Results Math Integration for

Output Response Extraction

• Setup a study
• Add a model
• Add output responses using Results Math
This tutorial runs a simple DOE study to showcase the process to use HyperView Results Math
to extract output responses. This feature enables you to write an .xml file which queries the
solver input and output files and then extracts the output responses in HyperStudy. This route
queries the result faster and is considerable faster way to extract output responses of a group
of elements or nodes than “readsim” function.
This setup requires you to add a study and load the input file into HyperStudy. The input
variables for the DOE study are selected as factors. A nominal run is performed (with
OptiStruct as the solver), and the output responses for the DOE study are selected (in this
case: Max stress of two different parts).
click OK.
a. From the Directory, drag-and-drop the Beam_Parts_4k.tpl file into the work area.

Note: The xml file needs to be copied to each run directory. A new batch file will
copy the xml file to each directory, and then execute the OptiStruct job that
needs to be registered.
c. In the Solver input file column, enter Beam_Parts_4k.fem. This is the name of the
solver input file HyperStudy writes during any evaluation.

Beam_Parts_4k.xml file.
e. Click Close.
7. Click Import Variables. Two input variables are imported from the
Beam_Parts_4k.tpl file.

2. Click Apply.
directory.

In this step you will create two output responses: Part 22 and Part 24.
1. Create the Part 22 output response.
a. From the Directory, drag-and-drop the Beam_Parts_4k.xml file, located in
• Set Subcase to SUBCASE 1 = 1.33 deltaP.
• Set Type to Part Stress (Part).
• Set Request to Part 22.
• Set Component to VM Max.
e. Label the output response Part 22.


h. Click Finish. The Part 22 output response is added to the work area.
2. Create the Part 24 output response by repeating step 1. Change the Request to Part 24.

3. Click Evaluate Expressions to extract output response values.
4. Click OK. This complete the study setup.
Last modified: v2017.2 (12.1156684)

HS-1535: Coupling HyperStudy, Morphing and AcuSolve

(HyperStudy Job Launcher)
In this tutorial, you will learn how to perform a DOE study using HyperStudy and the
HyperStudy Job Launcher within HyperMesh. HyperMorph is used to parameterize the shape of
the design. The subject of the study is to analyze sensitivity of flow to changes in the shape
(bending) of a pipe.
After performing the baseline simulation, a DOE study will be executed to analyze the effect of
changes in pipe shape on the pressure drop between inlet and outlet. This is one of the many
types of studies that can be done using AcuSolve with HyperMesh and HyperStudy.
A DOE or optimization study starts from a baseline model. This would be a model that has
already been simulated with AcuSolve. For completeness, this tutorial also describes typical
steps followed during the initial or baseline AcuSolve simulation. To this end, we use the file
pipe.hm, a mesh created in HyperMesh.
• Parameterize the model using HyperMorph and HyperStudy
• Use the HyperStudy Job Launcher to couple AcuSolve and HyperStudy
• Set up and run a DOE study
this directory to your working directory. The tutorial directory includes the following files:
• pipe.hm - HyperMesh model of the pipe.
• run_acusolve.bat - Customizable execution script for AcuSolve (Windows). The batch
file needs to be adapted to the current directory structure.
Pipe model with handles for the shapes
Step 1: Load the Model in HyperMesh Desktop

2. In the User Profiles dialog, set the user profile to Engineering Solutions, CFD,
AcuSolve.

5. In the Open Model dialog, open the pipe.hm file. A finite element model appears in the
graphics area.
Step 2: AcuSolve Case Setup in HyperMesh
1. In the Model browser, right-click and select Create > Material from the context menu. A
new material opens in the Entity Editor.

2. In the Entity Editor:

a. For Name, enter Air_HM.
b. Set Card Image to FLUID.
3. Create a second material named Water_HM, and set the card image to FLUID.
4. In the Model browser, Component folder, click inflow. The Entity Editor opens and
displays the component's corresponding data.
5. For Type, select INFLOW.
6. Change the Type for the following components:
Component Type
outflow OUTFLOW
wall WALL
fluid FLUID
7. In the Model browser, Components folder, click fluid. The Entity Editor opens and
display's the component's corresponding data.
8. For Material, click Unspecified >> Material.

10. In the Select Material dialog, select Air_HM and then click OK.
11. In the Model browser, Components folder, click inflow. The Entity Editor opens and
display's the component's corresponding data.
12. For Inflow type, select Mass flux.
13. For Mass flux ## (kg/sec), enter 0.0003.
Step 3: Morphing
1. In the panel area, click HyperMorph > morph volumes.

2. Go to the create subpanel.
3. Activate the entity selector.

4. Click elems > all. All of the elements in the model are selected.
5. Click create. A new morph volume is created.

6. Go to the split/combine subpanel.
7. In the graphics area, double-click on the edge of the model that is marked by a green
cross in the image below:
8. Click split. The morph volume splits.

9. Continue splitting the morph volume so that it resembles the image below.
10. Go to the update edges subpanel.

11. Click the first arrow and select update ends.
12. Click the third arrow and select master-slave.
Note: This options allows you to link any two edges together with a “master-slave”
relationship between two morph volumes.

Settings for steps 3.11 and 3.12.
13. Define the master edge by clicking the first edge of the morph volume.
14. Define the slave edge by clicking the edge that touches the master edge.
15. Repeat steps 3.13 and 3.14 for all of the edges until your model resembles the image
below (look for yellow arrows).
16. Click return to go back to the HyperMorph panel.

17. In the panel area, click morph.
18. Go to the move handles subpanel.
19. Click the second arrow and select translate.
20. Under translate, click the arrow and select along vector.
21. Under along vector, set the orientation selector to y-axis.
22. In the dist = field, enter 0.010.
Settings for steps 3.19 through 3.22.

23. Activate the handles selector.

24. Select the four middle handles, highlighted in grey, in the image below:
25. Click morph. The grid is morphed.
26. Go to the save shape subpanel.

27. In the name= field, enter s1.
28. Switch the first toggle from as handle perturbations to as node perturbations.
29. Click save.
30. In the window that appears, click Yes.
31. Click undo all.

32. In the Model browser, right-click on the Shape folder and select Hide from the context
sensitive menu. The shape, s1 is hidden.
33. Right-click on the Shape folder again and select Show from the context sensitive menu.
The shape, sh1 reappears.

35. From the menu bar, click View > Browsers > HyperMesh > Utility.
36. In the Utility browser, click Disp.
37. Remove any temporary nodes in the model by clicking Clear Temp Nodes.
38. Create design variables by clicking Design Study > Define DV from the menu bar.
39. Go to the desvar subpanel.
40. In the desvar = field, enter bend.
41. Click shape =.
42. Link a shape with the shape design variable by clicking s1.
43. Click create. The design variable bend is created.
44. Click return.
Step 4: Perform the Study in the HyperStudy Job Launcher
1. On the CFD toolbar, click . The HyperStudy Job Launcher opens.

2. In the Study directory field, navigate to the location of your working directory.
Note: By default path will be the same directory in which the .hm file is saved. It is
recommended that you create a separate folder for your study directory so that
all of the files will be placed in that folder.
3. In the No. of processors field, enter 2.
4. In the Define Output Responses table, select the following to identify what the pressure
change will be at inflow due to shape changes.
a. Set Responses to Pressure.
b. Set Components to inflow.
5. To see how the pressure contours for each optimum design will look, set Output format
to H3D-Single file.

7. Click Launch.
8. In the dialog that opens asking if you would like to continue, click Yes.
9. In the dialog that opens informing you of the files that were created, click OK.
10. In the dialog that opens asking if you would like to continue, click Yes.
11. A nominal run is submitted, and acuProbe and acuTrail are launched to provide you with
updated information about the run. Once the run is finished, HyperStudy opens and with
the study Setup completed.
12. Go through each step to ensure that everything was properly defined.
13. In the Define models step, you will see that the resource file, solver input file, and solver
execution script have been defined.
14. In the Define Input Variables step, you will see that the input variable you defined in
HyperMesh Desktop has been imported.
15. In the Define Output Responses step, ensure that the output response is defined
correctly.

5. Click the Levels tab.
6. Change the number of levels to 5.
7. Click Apply.
8. Go to the Evaluate step. The table used to run the study appears, showing all of the runs
(1 - 5) to be executed.
10. To view the optimization results of the five runs in a table, click the Evaluation Data
tab. The table displays the results of the five runs for the output responses.
11. To plot the optimization results of the five runs, click the Evaluation Plot tab. Plot the
results of the inflow_pressure output response by selecting
inflow_pressure with the Channel selector..

12. The extreme left (bend = -1), middle (bend = 0) and extreme right (bend = 1) results
correspond to the following pipe shapes.
13. Click File > Save As from the menu bar to save the study.
14. In the HyperStudy - Select location dialog, navigate to your working directory and save
the study.
15. The results of the DOE can be visualized in HyperView. Load the corresponding *.h3d
file from the run folder into HyperView.

HS-1540: Shape Optimization Study Using HyperMesh and

ANSYS
In this tutorial, you will learn how to pshape parameterization using HyperStudy.
• erform a shape optimization started from inside HyperMesh using the direct link to
HyperStudy. The finite element solver is Ansys. HyperMorph is used to do the shape
parameterization. The objective is to minimize the maximum stress of a plate with a
hole. The solution can be expected to be some kind of ellipse. Hence, the input
variables are the half-axes of the hole.
• Do a Run a study from inside HyperMesh.
• Set-up a study.
• Run an optimization study.
plate.cdb - Original Ansys input file.
ansys.bat - Sample execution script for the educational version of

Ansys.
Double symmetric plate model.

Step 1: Setup the Model in HyperMesh

2. In the User Profiles dialog, set the user profile to Ansys.
3. From the menu bar, click File > Import > Solver Deck.
4. Set File type to Ansys.
5. In the File field, open the plate.cdb file.
6. Click Import. A finite element model appears in the graphics area.
Step 2: Do the Shape Parameterization in HyperMorph
1. From the Tool page, click HyperMorph.
2. Click domains.
3. Go to the create subpanel

5. Click the first arrow and select auto functions.
6. Click generate. HyperMesh generates the domains and handles that you will use to
manipulate the shape of the mesh and to generate shape perturbations for shape
optimization.
7. Click return.
8. In the panel area, click morph.
9. Go to the set biasing subpanel.
10. In the graphics area, click the two yellow handles located in the corner of the quarter
circle.

11. In the bias= field, enter 2.000.

12. Click update.
13. Go to the alter dimensions subpanel.
14. Define the radius as a shape by clicking the first arrow and selecting radius.
15. Click the bottom arrow, and select hold center.
16. In the graphics area, click the red edge of the hole.
17. In the radius= field, enter 20.0000.

18. Click morph. The first shape is generated.
20. In the name= field, enter sh1.
21. Click save.
22. Click undo to prepare for the generation of the next shape.
24. Click the second arrow and select translate.
25. Click the third arrow and select along xyz.

27. In the graphics area, click the lower yellow handle located in the corner of the quarter
circle.
28. In the x val= field, enter 10.0000.

29. Click morph. The second shape is generated.
32. Click save.
33. Click undo to prepare for the generation of the next shape.
35. In the graphics area, click the upper yellow handle located in the corner of the quarter
circle.

36. In the x val= field, enter 0.000.

37. In the y val= field, enter 10.0000.
38. Click morph. The third shape is generated.
41. Click save.
42. Click undo to restore the initial mesh.
43. Save the HyperMesh model by clicking File > Save As > Model from the menu bar.
44. In the Save Model As dialog, save the file as plateDV.hm.
45. Close HyperMesh Desktop.
Step 3: Register ANSYS as a Solver
2. From the menu bar, click Edit > Register Solver Script.
3. In the Register Solver Script dialog, click Add Solver Script.
4. In the Add - HyperStudy dialog, enter Ansys in the Label and Varname fields.
5. From the list of solver script types, select Generic.
6. Click OK.
7. In the Path column of the script Ansys, click .

8. In the Open dialog, open the ansys.bat file.

Note: The script ansys.bat is a sample of an execution script for Ansys on Windows.
Copy ansys.bat to your working directory to use it.
10. Click Save.
11. In the Save Preferences dialog, navigate to your working directory.
Note: On UNIX, the preference file can also be saved in your home directory or in the
working directory from which you launched HyperStudy.
12. In the File name field, enter a name for the new user preference file (example:
userprefs.mvw).
13. Click Save.
Note: Do not overwrite the system preferences file, which is located in
<install_directory>/hw by default.
14. Click Close.
Note: When you start a new HyperStudy session, you can load your preference file by
clicking File > Set Preference File from the menu bar, and then selecting the
preference file you just saved. HyperStudy reads the default preference file in the
installation directory, followed by the preference file that you specify. This will
ensure that all solvers, readers and import templates are available.
You can also append the current user preference file using Append, or you can
exit solver registration using Close. In the last case, the solver will only be
registered for the current study.
click OK.
a. From the Directory, drag-and-drop the HyperMesh (.hm) file plateDV.hm into the
work area.

b. In the Solver input file column, enter plate.cdb. This is the name of the solver
c. In the Solver execution script column, select Ansys.
d. In the Solver input arguments column, enter plate.out plate after $file.

a. Expand Shape, and select sh1.S, sh2.S, and sh3.S.
b. Verify that the Lower bound is -1.0 and the Upper bound is 1.0.
c. Click Add.
d. Click OK.

8. Review the input variable's lower, initial, and upper bounds.

2. Click Apply.


1. From the Directory, drag-and-drop the plate.rst file, located in the

and click Next.
3. Select Multiple items at multiple time steps (readsim), then click Next.
4. Define the following options, and then click Next.
a. Set Subcase to Step 1.
b. Set Type to Stress (2D).
c. Set Request to E132 - E423.
d. Set Component to vonMises.
5. Optional. Enter labels for the data source and output response.
6. Set Expression to Maximum.
7. Click Finish. Output response 1 is added to the work area.
8. Click Evaluate Expressions.
6. Click Add Objective.
7. In the Add - HyperStudy dialog, add one objective.
8. Define the objective.
a. Set Type to Minimize.
b. Set Apply On to Response 1 (r_1).

9. Click Apply.
12. Click Apply.
1. Click the Iteration History tab to review the Optimization results.

Note: The optimal design is highlighted in green.
2. Click the Iteration Plot tab to plot the optimization results.

HS-1545: DOE and Optimization Study with HyperMesh,

HyperStudy, and ANSYS CFX
In this tutorial, you will learn how to perform a DOE and Optimization study of a fan using
Altair HyperMesh®, Altair HyperStudy®, and ANSYS CFX®. The DOE results are used to create a
response surface approximation to the fan efficiency (objective function), followed by the
optimization of the blade shape using the same output response approximation.
This tutorial and the associated images were created on Windows, however all steps are
identical on UNIX systems, except for changes in directory paths.
fan_6_blades.hm HyperMesh file: (1/6) of the six-blade fan model.

-
HST_CFX.tcl - Customizable execution script for CFX (Windows and UNIX).
preferences_wis Preferences file to add wish as a solver to execute HST_CFX.tcl.

h.mvw -
model.pre - CFX-Pre command file. Used to generate a new model.def definition file
for CFX-Solve. This file can be customized to meet the specific needs of
other models.
CFX_options.txt Customizable command line options for CFX-Solve.

-
model.cse - CFX-Post session file. Used to extract the quantities of interest from CFX
results files (model.res). Fan efficiency is the model output
response used for optimization in this tutorial example.
In this tutorial, you will:

• Do a shape parameterization using HyperMorph and export model/shapes for HST /
CFX.
• Set-up a new study with HyperStudy.
• Run the DOE study with HyperStudy and CFX.

• Post process the DOE results and generate output response approximations with
HyperStudy.
• Perform a shape optimization based on the DOE results/output response
approximations.
The first step to carry out a successful DOE / optimization study is to have a sound CFX model
that converges reasonably well and produces meaningful simulation results. This is a baseline
model and would be a model that has already been simulated with CFX. We have generated
the files model.cfx and model.res; the latter file (baseline simulation results) can be used to
initialize each DOE run, thus reducing the total CPU/wall time necessary to reach convergence.
For completeness, this tutorial describes typical steps followed during an initial or baseline CFX
simulation for which a file created in HyperMesh, fan_6_blades.hm, will be used.
After performing the baseline simulation, a DOE study will be done to examine how fan
efficiency depends on changes in the blade geometry at constant RPM and flow rate. These
operating conditions (boundary conditions) are selected because CFX converges very quickly.
Efficiency will be measured as the ratio between the power transferred to the airflow and the
mechanical power consumed by the rotating blades. This is just one illustrative example
among the many types of studies that can be done using CFX with HyperMesh and
HyperStudy.
Step 1: Load the Model in HyperMesh

2. In the User Profiles dialog, select Engineering Solutions, CFD, General.
3. Click OK.

6. In the Open Model dialog, open the fan_6_blades.hm file. A finite element model
appears in the graphics area.
Note: This model's mesh contains 3D elements in the fluid_rot component, and 2D
elements (for boundary conditions) in the remaining components.
7. From the menu bar, click File > Export > Solver Deck.
8. Set File type to CFD.
9. Set Solver type to Fluent.
10. In the File field, save the file as model.cas.
11. Click Export.
12. In the first dialog that opens, asking if you would like to continue, click Yes.
13. In the second dialog that opens, informing you that the model.cas file was created, click
OK.
14. Exit HyperMesh.

Step 2: Run the Baseline Simulation with CFX
1. Start CFX Launcher.

2. Start CFX-Pre.
3. From the menu bar, click File > New Simulation.
4. In the Simulation Type area, click General.
5. Click OK.
6. In the dialog that opens, click OK.
7. From the menu bar, click File > Import Mesh.
8. From the File type list, select Fluent (*cas).
9. Navigate to your working directory and open the model.cas file.
Note: The default mesh units should not change “m.”
10. If a Warning dialog opens, click OK.
Note: This is fine, as it is referring to the way FLUENT CAS files define the meshes by
internal faces.
11. The mesh with all boundary conditions will be read and displayed as shown in the image
below.
Note: All boundary components have been read with exactly the same names used in
HyperMesh.

12. From the Principle 3D Regions list, right-click on fluid_rot and select Insert >
Domain from the context sensitive menu.
13. Under Domain Motion, change the Option to Rotating.

14. In the Angular Velocity field, enter 2000 rev min^-1.
15. Leave the default Rotation Axis of Global Z.
16. Click OK.
17. In the Simulation folder, double-click on Simulation > fluid_rot to open the
Domain: fluid_rot tab and inspect the associated values.
18. From the Principle 2D Regions list, right-click on Principal 2D Region, periodic_rot1
and select Insert > Domain Interface from the context sensitive menu.

20. In the Domain interface definition tab, define the rotational periodicity interface as
21. A new domain interface will be visible.
22. From the Principle 2D Regions list, right-click on wall_blades and select Insert >
Boundary > Wall from the context sensitive menu.
23. In the Boundary: wall_blades dialog, select the parameters indicated in the images
below.

25. Click OK.

26. Right-click on hub_rot and select Insert > Boundary > Wall from the context sensitive
menu.
27. In the Boundary: hub_rot dialog, select the parameters indicated in the images below.
28. Click OK.

29. Right-click on shroud_rot and select Insert > Boundary > Wall from the context
sensitive menu.
30. In the Boundary: shroud_rot dialog, select the parameters indicated in the images
below.

32. Click OK.

33. Right-click on inlet_rot and select Insert > Boundary > Inlet from the context
sensitive menu.
34. In the Boundary: inlet_rot dialog, select the parameters indicated in the images below.

35. Click OK.

36. Right-click on outlet_rot and select Insert > Boundary > Outlet from the context
sensitive menu.
37. In the Boundary: outlet_rot dialog, select the parameters indicated in the images
below.
Note: 0.1 kg/sec
38. Click OK. The boundary conditions to analyze the performance of this fan under constant
air flow rate and rotor RPM are now defined.
39. To save the CFD simulation model, click File > Save Case As from the menu bar.
40. Navigate to your working directory and save the file as model.cfx.
41. To write the solver file, click Tool > Solve > Start Solver and Monitor from the menu
bar.
42. In the Write Solver Input File dialog, navigate to your working directory and save the
file as model.def.
43. In the CFX-Solver Manager dialog, the solver will converge in approximately 45
iterations.

45. If the CFX-Solver Manager dialog appears, click Post Process Results; if the dialog
does not appear start CFD-Post and load the .res file created by the CFX Solver.
46. Review the user-customizable CFX-Post session file model.cse, which defines several
performance parameters and saves them to a file named data.dat.
47. To execute this session file, click Session > Play Session from the menu bar.

48. Navigate to your working directory and open the model.cse file. After a couple of
seconds, a new file named data.dat will be available in your working directory. The file
data.dat (shown below) has five columns, a description of each column value can be
found browsing model.cse. The rightmost column with value 0.796. is the fan efficiency
calculated as the ratio between the power transferred to the airflow (outlet – inlet) and
the mechanical power consumed by the rotating blades.
CFX-Solve saves results files with the name model_###.res, where ### represents a
version number. Very often it is advisable to initialize the simulations done in the DOE
with this baseline results file to decrease total run time. For this reason, it is advisable to
rename a good baseline results file model_###.res to model.res, and use this name in
file CFX_options.txt (inspect the contents of this file which depends on your solution
approach, hardware, etc). The file CFX_options.txt contains the command line options
that are passed to the CFX solver in each of the runs started by HyperStudy.
49. Exit all CFX tools (e.g. Pre, Solve, Post) so that the license is available for the HyperStudy
DOE/optimization runs. Now you can begin generating the blade’s optimization
shapes/parameterization in HyperMesh using this model.
1. Optional: If you closed your previous Engineering Solutions session with the
fan_6_blades.hm model, reopen it.
2. In the Model browser, Component folder, right-click on fluid_rot and select Isolate
Only from the context menu. All of the components are hidden in the graphics area,
except fluid_rot.
3. In the panel area, click HyperMorph.

4. Click domains. The Domains subpanel opens, from which you can create domains for
the shape parameterization.
Note: For this tutorial, you can use the domains that are generated automatically.

5. Go to the create subpanel.

6. Click the first arrow and select 3D domains from the list.
7. Switch the toggle from all elements to elems.
8. On the Standard Views toolbar, click . The graphics area displays the XZ Right Plane
View of the model.
9. To rotate the model view clockwise 90 degrees, click on the Standard Views
toolbar.
Note: The model's X-Axis should point upwards and the Z-Axis should be point to the
right, as indicated in the following image.
10. Select all of the elements on the left half of the model.
Tip: Use box selection to quickly select elements. To box select, press SHIFT while
left-clicking, then drag your mouse.

11. Click create. A 3D domain and handles are created in the model.
12. Go to the parameters subpanel.
13. In the handle size= field, enter 0.005.
14. Click return. It make take a couple of seconds to regenerate the data structure before
you are returned to the HyperMorph panel.
15. Click morph. The Morph subpanel opens, from which you can morph the shape of the
mesh.
Note: The objective of the following steps is to create two input variables. The first
input variable will change the blades angle, and the second input variable will
change the fan thickness.
17. Click the first arrow and select handles.
18. Select the four handles on the fan inlet face of the model.
19. Click the second arrow and select rotate.

21. Select the true rotation checkbox.

22. Under rotation axis, select z-axis.
23. To select the base point for rotation, click .

24. In the x=, y= and z= fields, enter 0.0.
25. Click return.
26. In the angle field, enter 5.0 (degrees).
30. In the name= field, enter sh_5deg.
31. Click save. The new shape appears in the Model browser.
32. To prepare for the generation of the next shape, click undo all.
34. Repeat steps 17 and 18 above.
35. Click the first arrow and select Translate.
36. Click the second arrow and select along xyz.
37. In the z val= field, enter 0.005.

39. Click morph. A second shape is generated.

42. In the name= field, enter sh_5mm.
43. Click save. The new shape appears in the Model browser
44. Click undo all.
45. Click return twice.
Step 4: Export the Shape Parameterization from HyperMesh to CFX
1. In the panel area, click shape.

3. Click the first arrow and select multiple desvars.
4. Click shapes.
5. Select the shapes sh_5deg and sh_5mm.
6. Click select.

8. Click create.
9. In the window that appears, asking if you would like to switch to non-linear options, click
No. An input variable is created for each shape selected.
10. Optional: If you would like to animate or visualize the shapes, click animate.
11. Optional: In the Deformed panel, click linear or modal to animate the shape variables in
the graphics area.
12. Optional: While the shape is animating, you can adjust the animation speed by moving
the slider as indicated in the following image.
13. To exit the panels, click return.

14. From the menu bar, click File > Save As > Model.
15. In the Save Model As dialog, save the file as fan_6_blades_2shapes.hm.
16. From the menu bar, click View > Browsers > HyperMesh > Utility.
17. In the Utility menu, click CFD I/O.
18. Under Export files for CFX, click Shapes.
19. In the dialog that appears, asking if you would like to continue, click Yes.
20. In the Open baseline ACSII case file dialog, open the model.cas file.
21. In the dialog that appears, alerting you that your files have finished writing, click OK.
22. In the dialog that appears, alerting you that the HyperStudy template file for Fluent has
been created, click OK. The following files are created:
model.shp Grid perturbation vector data read by file
model.fluent.node.tpl.
model.fluent.node.tpl Grid coordinates template.
model.tpl Fluent case template read by HyperStudy.

2. Register the WISH interpreter as a solver to run the CFX simulation.
a. From the menu bar, click File > Use Preferences File.
b. In the HyperStudy - Set Preference File dialog, open the preference_wish.mvw
file.
click OK.
Note: It is important that you select the study directory that contains all the files that
you generated earlier. These files include the CFX native files: model.cfx,
model.pre, model.cse, and (if you plan to customize your CFX solution with
command line arguments) CFX_options.txt. If you are restarting from an
existing .res file, make sure the model.res file is also in your study directory.
For advanced users: The CFX_options.txt file can be changed at any time while
HyperStudy is running. This file is copied to the execution directories (e.g. m_1) at
the time when the simulation in that particular directory is starts, not when
HyperStudy first writes the .cas file to that directory. Verify that the options are
correct, otherwise CFX-Solve will abort the run. A typical use could be to add a
command line option to run in parallel, use a lower priority, etc.
a. From the Directory, drag-and-drop the model.tpl file into the work area.
b. In the Solver input file column, enter model.cas. This is the name of the solver
c. In the Solver execution script column, select Wish (wish).
d. In the Solver input arguments column, enter HST_CFX.tcl before $file. This is the
name of the CFX execution script.
Append the absolute path to the beginning of HST_CFX.tcl. For example,
C:/test/HST_CFX.tcl; on UNIX: /users/local/HST_CFX.tcl or equivalent.

Optional. If you are running HyperStudy on Linux, enter -nobg after $file. For
example, /users/local/HST_CFX.tcl $file -nobg.
Save the HST_CFX.tcl file to a location that can be reused by all optimization studies.
The HST_CFX.tcl file contains specific information about the CFX installation directory, as
illustrated in the image below.
Note: The optional CFX_options.txt file contains command line options that are
passed to the CFX solver in each of the runs started by HyperStudy. Inspect the
contents of this file. The contents will vary depending on your solution approach,
hardware, and so on. Comment lines start with #, and the last
defined “CFX_OPTS =” statement is used. One of the main functions of this file is
to specify an initialization results file. The objective is to speed up the
convergence of each run by initializing the solution with the results from the
baseline run done before starting the DOE/optimization work. It is crucial to make
sure that the .res file specified exists, and that the path name is correct.
7. Rename your baseline results file mode_###.res to model.res so that the file name listed
after -ini exists.
8. Click Import Variables. Two input variables are imported from the model.tpl file.

2. Click Apply.


directory. This directory contains the model.cas file, which is the input mesh file
generated by HyperStudy, and the CFX_run.log file, which is a log of HyperStudy’s scripts
execution (read the contents of this file if there are any errors). The
approaches/nom_1/run 00001/m_1 directory contains the pre_script.bat.out,
solve_script.bat.out, and post_script.bat.out files, which are the output files from
the CFX’s Pre, Solve, and Post executables. These file will contain information
regarding any licensing or runtime problems that may have occurred. If everything is
correct, you will find the data.dat file specified as output in your model.cse file; this file
contains five columns as described previously.
In this step you will create three output responses: Efficiency, W_fluid, W_blades.
1. Create the Efficiency output response.
a. From the Directory, drag-and-drop the data.dat file, located in the
e. Label the output response Efficiency.
Note: This expression evaluates the fifth column of the first (and only) line of
file data.dat.
g. Click Finish. The Efficiency output response is added to the work area.
2. Repeat step 1 to create the output responses W_fluid and W_blades.
• For W_fluid, set Component to Column 3.
• For W_blades, set Component to Column 4.

4. Verify that the input variable's Lower Bound, Initial and Upper Bound values are -
1.0, 0.0, 1.0.

7. Click the Levels tab.
8. Change the number of levels for both input variables to 3.
9. Click Apply.
11. Click Evaluate Tasks to start solving the run matrix.
12. Go to the Post processing step.
13. Click the Summary tab to view a table of all of the output responses extracted from the
nine runs.
The lowest average inlet pressure occurs during run nine, where sh_5deg and sh_5mm
equal 1.0. These results make sense because the elbow has the maximum cross sectional
area. The next two lowest values occur during runs six and eight, when one of the shapes
has a value 1.0 and the other has a value of 0.0. In relative terms, the pressure drop for
runs six and eight are not much larger than run nine.

14. Click the Interaction tab to view the effects of the input variables on an output response.
Use the Channel selector to select the following output responses: sh_5deg and
sh_5mm on the Efficiency. The interactions between the output responses are plotted.
Step 9: Create an Approximation from the DOE Study
2. In Add - HyperStudy dialog, select Fit and click OK.
3. Go to the Select matrices step.
4. Click Add Matrix.
5. In the Add - HyperStudy dialog, add one matrix.

7. In the Matrix Source column, select Doe 1 (doe_1).

8. Click Import Matrix. The data from Doe 1 is imported to the current Fit approach as an
input matrix.
10. In the work area, set the Mode to Least Squares Regression (LSR).
11. In the Settings tab, change the Order to 3.
12. Optional. View the input matrix by clicking Edit > Run Matrix from the top, right corner
of the work area.
13. Click the Regression Terms tab.

14. Using the Channel selector, select the output response Efficiency.
15. In the Active column, clear the a6 and a7 checkboxes.

16. Clear the a6 and a7 checkboxes for the output responses W_fluid and W_blades.
Note: The number of sampling points is nine, but the 3rd order approximation model
terms are 10. To build an approximation model with good fit quality and limited
sampling points, you need to neglect the pure 3rd order terms, and keep the rest
as a compromise. Later on in this tutorial, you will check the residuals and
diagnostics of the fit quality for this decision.
17. Click Apply.
19. Click Evaluate Tasks. HyperStudy computes the regression surface.
21. Click the Residuals tab to review the residuals of fitting.
The maximum errors are small enough in a relative sense, otherwise using the
approximation to conduct an optimization will not be meaningful.
22. Click the Diagnostics tab to review diagnostic information, T-values, and Confidence
Intervals of the regression terms.
23. Click the Trade-Off 2D tab to plot the response surface.
Tune the input variables and change the curve by moving the sliders in the Value
column above the plot.
24. Click the Trade-Off 3D tab to review the response surface in a 3D plot.

Step 10: Run an Optimization Study Using the DOE Study Results
6. Click the Responses tab.
7. In the Active column, clear the W_fluid and W_blades checkboxes. The Optimization
will not have to rely on the CFX solution anymore.
8. Select the output response Efficiency by highlighting it in the work area.
9. Click Add to Objectives. HyperStudy creates an objective for the output response.
10. Click the Objectives tab.


a. Set Type to Maximize.
b. Set Evaluate From to Efficiency_LSR (r_1_fit_1).
Note: You are instructed to select the Efficiency_LSR (r_1_fit_1) option rather then
the Solver option, due to the fact that it will run much faster. CFD analyses
generally take a considerable amount of CPU time. If you use the Solver
option, a new CFX solution will need to be computed for every iteration in
the optimization process, which translates to long run times.
13. Click Apply.

15. In the work area, set the Mode to Global Response Search Method (GRSM).
16. Click Apply.
18. Click Evaluate Tasks to start the Optimization.
19. Click the Iteration Plot tab to plot the iteration history of the objectives, constraints,
input variables, and unused output responses.
20. Click the Iteration History tab to review the actual values of the input variables and the
associated values of all the output responses for every iteration in the optimization
process. The green rows represent the optimal designs throughout the iteration history.
The best design was found in iteration 4. No better design is found in the subsequent
iterations which indicates that this design is likely not a poor local minimum.
The main objective of this optimization study was to illustrate the steps involved in doing
a shape optimization based on the results of a Doe study and associated approximations.
The maximum efficiency was obtained optimizing a response surface

approximation. It is recommended that such predictions be verified by running a small

DOE study zooming in on a small sub-domain (of the whole design space) around the
optimum design found through the optimization of the global approximation. Doing so will
verify (with the CFX solver) that the optimized design actually delivers the performance
expected.
Concluding Remarks
This tutorial illustrated the main steps to perform a DOE study, to build an approximation from
the DOE results, and to perform a shape optimization study based on a Least Square
Regression approximation. HyperMesh was used to generate the mesh for CFX and the shape
variables used to modify the geometry of the fan model. HyperStudy and CFX were used to
perform the DOE and optimization studies. Relatively simple modeling assumptions were used
in this tutorial to minimize run times; however the steps followed in this tutorial are equally
applicable to CFD optimization studies of any size and complexity.

HS-1550: Shape Optimization Study Using HyperMesh and

Abaqus
This tutorial demonstrates how to perform a shape optimization using HyperMesh and
HyperStudy. You will be using the finite element solver Abaqus, and HyperMorph to do the
shape parameterization. This tutorial also demonstrates how to solve a problem when
HyperMesh and HyperStudy are running in Windows and the solver is on a UNIX platform.
link.hm Original HyperMesh file.
HS- Basic Solver Script Setup (ONLY for LOCAL Machine) –

1550_solverScr
ipt.py Run everything on local machine (Windows or Unix)
run_abaqus.sh (ONLY for REMOTE Machine) - Shell script which is

designed to run the abaqus solver in the UNIX machine.
ssh_remote.bat Advanced Solver Script Setup (ONLY for REMOTE

Machine) – Run HyperStudy on Windows with Study
Directory in Unix and the Solver is Running on Unix.
Sample execution script to run Abaqus on UNIX.

• Use HyperMorph to generate a shape variable.
• Run a study from inside HyperMesh.
• Perform a shape parameterization using HyperStudy.
• Set up a study.
• Write a script to run Abaqus on UNIX and register the script in the preference file.
• Run an optimization study.
The objective of this tutorial is to minimize the mass of a link that is connected to a shaft,
given a stress constraint of 200MPa. The input variables are defined by the outer shape.

Step 1: Load the Model in HyperMesh

2. In the User Profiles dialog, set the user profile to Abaqus, Standard2D.

5. In the Open Model dialog, open the link.hm file. A finite element model appears in the
graphics area.
1. From the Tool page, click HyperMorph.
2. Click domains. The Domains subpanel opens, from which you can create domains for
the shape parameterization.
3. Click the first arrow and select 2D domains.
4. Click the toggle and select all elements.
5. Click create. Domains and handles are generated and will be used to manipulate the
shape of the mesh and to generate shape perturbations required for the shape
optimization.

7. Click return.
8. From the HyperMorph panel, click morph. The Morph subpanel opens, from which you
can morph the shape of the mesh. The goal of the following steps is to create one input
variable for the outer edge of the link.
9. Click the second switch and select translate.
10. In the y val= field, enter -5.0.
11. In the graphics area, click the yellow handle located at the top-right corner of the link as

14. Click save shape.

15. In the name= field enter, sh1.
16. Click save.
Step 3: Basic Solver Script Setup (ONLY for LOCAL Machines) - Run
everything on local machines (Windows or Unix)
To run the solver on a local machine, use the supplied python script HS-
1550_solverScript.py, which calls the Abaqus solver using arguments specified that you
specify.
Change the path in the file to the Abaqus executable on your machine.
Any additional arguments, such as memory requests, can be modified in the script as well.
# import statements
import os
import platform
import subprocess
import sys
# user edits---------------------------------------------------------
----------
# set abaqus path

abaqus_exe = '/my/path/to/executable/abaqus.exe'
# set abaqus arguments

abaqus_arguments = 'job=' + sys.argv[1] + ' ' + 'memory=200Mb
interactive'

# set log file name logFile = 'logFile.txt'
# -------------------------------------------------------------------
----------
# open log file

f = open('logFile.txt', 'w') f.write('Hello world')
# get environment information plat = platform.system() hst_altair_home =

''
strEnvVal = os.getenv('HST_ALTAIR_HOME') if strEnvVal:
hst_altair_home = strEnvVal else:
f.write('%EXA_xx-e-env, Environment variable not defined (
HST_ALTAIR_HOME )')
abaqus_exe = os.path.normpath(abaqus_exe) lstCommands = [abaqus_exe,

abaqus_arguments]
# echo log information f.write('platform: ' + plat + '\n')

f.write('hst_altair_home: ' + hst_altair_home + '\n') f.write('abaqus
command: ' + abaqus_exe + '\n')
#execute the command

f.write('Running the command:\n' + ' '.join(lstCommands) + '\n') p1 =
subprocess.Popen(lstCommands, stdout=subprocess.PIPE ,
stderr=subprocess.PIPE)
# log the standard out and error

f.write('\nStandard out' + '\n' + p1.communicate()[0] + '\n')
f.write('\nStandard error' + '\n' + p1.communicate()[1] + '\n')
# close log file f.close()

(OPTIONAL - ONLY for REMOTE Machine): Advanced Solver Script

Setup – Run HyperStudy on Windows with Study Directory in Unix and
the Solver is Running on Unix.
In order to run commands on a remote Unix machine, a local program must be installed to
communicate the commands remotely. In this example, the program ssh is being used, but
other equivalent or better programs exist. In most cases, the program’s protocols require
authentication from a password. For this setup to work the environment needs to be
configured to work without an active password entry. This setup may require help from your
network administrators.
The ssh_remote.bat file is a sample (Windows) batch file to run a script on Unix
(run_abaqus.sh) from a HyperStudy session running on Windows.
ssh user_name@unix_machine “./run_abaqus.sh %1 %HST_APPROACH_VARNAM

% %HST_STEP_INDEX% %HST_APPROACH_MODELS%”
The batch file uses the ssh command to log onto the Unix machine and execute the solver on
the files created by HyperStudy. This script will be registered in HyperStudy as the solver
script.
To use the sample script, change the following to match your setup:
1. Change the generic parameter “unix_machine” to match the name of the remote Unix
machine on your network.
2. Change the generic parameter “user_name” to match your logon on the remote
machine.
Note: As mentioned above, this logon should have been

configured to work without a password between these
two machines.
The file run_abaqus.sh is a shell script which is designed to run the Abaqus solver on a UNIX
machine. This file should be placed in your HOME directory on the Unix machine.
#!/bin/sh
#constuct the model directory

approachName=$1
runNum=`printf %04i $2`
modelString=$3
array=(${modelString//:/ })
modelName=${array[0]}
myDir=~/${approachName}/run_${runNum}/${modelName}
#change the directory to model directory

cd $myDir
#change the format from windows to unix

dos2unix $1 $1

#edit this line to match your solver path and the appropriate
arguments
/my/path/to/solver/example/abaq631 job=$1 memory=200Mb interactive
Once there, perform the following:

1. Change the text formatting to be Unix compatible using the command dos2unix on the
file: dos2unix run_abaqus.sh.
2. Make sure the file has executable permission. Type chmod 755 run_abaqus.sh.
3. Edit the run_abaqus.sh file and modify the path to the executable and any other options
to this command as needed.
Step 4: Register Abaqus as a Solver
1. Start Hyperstudy.
4. In the Add - HyperStudy dialog, enter Abaqus in the Label and Varname fields.
5. For solver script type, select Generic.
6. Click OK.

7. In the Path column of the script Abaqus, click .

8. In the Open dialog, open the python.exe file.
Note: You can also copy and paste the same path and file from the Python Path field.
9. In the Argument column of the Abaqus script, click .

10. In the Open dialog, navigate to your working directory and open the HS-
1550_solverScript.py file.
11. In the top, right corner of the Register Solver Script dialog, click Export.
13. In the File name field, enter a label for the new user preference file (example:
userprefs.mvw).
14. Click Save.
15. Optional. ONLY for REMOTE Machines: Register Abaqus as a solver.

click OK.
Note: The study directory MUST be your home on the mapped UNIX machine.
a. From the Directory, drag-and-drop the HyperMesh (.hm) file link.hm into the work
area.
b. In the Solver input file column, enter link.inp. This is the name of the solver input
file HyperStudy writes during any evaluation.
c. In the Solver execution script column, select Abaqus.

e. In the Solver input arguments column enter, $filebasename.
Optional. In addition, you may need to edit the Abaqus environment file (ex:
<ABAQUS INSTALL>\v6.11\6.11-1\site\abaqus v6.env) to include: ask_delete=OFF
or
comment the line ask delete=on if any.
This is needed because Abaqus prompts you to overwrite the old files when re-
running the analysis. In order to eliminate the need for user interaction, you need to
command Abaqus not to ask this question and overwrite.
a. Expand Shape, and click sh1.S.
b. Verify that the Lower bound is 0.0 and the Upper bound is 1.0.
c. Click Add.
d. Click OK.

8. Review the design variable's lower bound, initial and upper bound range.


2. Click Apply.
In this step you will create two output responses: Mass and Max_Stress.
a. From the Directory, drag-and-drop the link.dat file, located in the
• Set Type to ABAQUS.dat.
• Set Request to TOTAL MASS.
• Set Component to MASS.
2. Create the Max_Stress output response.
a. From the Directory, drag-and-drop the link.obd file, located in the
• Set Subcase to Step-2.
• Set Type to S-Global-Stress components (PART-1-1).
• Set Request to E1 - E378.
• Set Component to vonMises.
e. Label the output response Max_Stress.


h. Click Finish. The Max_Stress output response is added to the work area.
6. Add an objective.
a. Click Add Objective.
b. In the Add - HyperStudy dialog, add one objective.
c. Define the objective.
• Set Type to Minimize.
• Set Apply On to Mass (r_1).
7. Add a constraint.
a. Click the Constraints tab.
b. Click Add Constraint.
c. In the Add - HyperStudy dialog, add one constraint.
d. Define the constraint.
• Set Apply On to Max_Stress ( r_2).
8. Click Apply.

11. Click Apply.

13. Click Evaluate Tasks to start the Optimization.
Step 9: View the Iteration History of the Optimization Study
1. Click the Iteration History tab to review the Optimization results. The optimal design in
highlighted green.
2. Click the Iteration Plot tab to plot the Optimization results. Use the Channel selector to
select Constraint 1 and Objective 1.

HS-1560: Study Setup Using LS-DYNA Model Parameters in

HyperMesh
This exercise outlines the procedure for setting up a study for a LS-DYNA model in HyperMesh.
This model has shape variables created using HyperMorph. In addition to the shape variables a
HyperMesh parameter is created for material property, E. These variables are then imported to
HyperStudy as input variables. For this study, you will start HyperStudy from within
HyperMesh.
Step 1: Importing a HyperMesh Parameter into HyperStudy

2. In the User Profiles dialog, set the user profile to LS-DYNA.
4. In the Open Model dialog, open the boxbeam_morphed_noDV.hm file. A model appears in
the graphics area.
5. In the Model browser, Material folder, click Material. The Entity Editor opens and
displays the material's corresponding data.
6. Right-click on E and select Create and Assign Parameter from the context menu.

8. In the Create Parameters dialog, change the Name to E and then click Close.
Step 2: Register LD-DYNA as a Solver
In order to use LS-DYNA as a solver, you need to register it in the preference file for
HyperStudy.


5. In the Add - HyperStudy dialog, enter Dyna in the Label and Varname fields.
6. From the list of solver script types, select Generic.
7. Click OK.
8. In the Path column of the script Dyna, click .

9. In the Open dialog, navigate to the local installation of the LS-DYNA solver, and then
have it point to the LS-DYNA solver executable without any spaces in the file path.
Note: If LS-DYNA is not installed locally, write a solver script to call it properly. For
more information on this process, refer to Solver Script Files.

10. Click Save.

12. In the File name field, enter a label for the new user preference file (example:
userprefs.mvw).
13. Click Save.
14. Click Close.
When you start a new HyperStudy session, load your preference file by clicking File > Set
Preference File from the menu bar. The default preferences file in the installation
directory will be read, followed by the preference file that you specify. This ensures that
all solvers, readers, and import templates are available.
Append the current user preference file by clicking Append, or exit solver registration by
clicking Close. In the last case, the solver will only be registered for the current study.
During this step, you will import the input variables that you created in Step 1: Importing a
HyperMesh Parameter into HyperStudy.
click OK.
a. From the Directory, drag-and-drop the boxbeam_morphed_noDV.hm file into the work
area.
b. In the Solver input file column, enter boxbeam_morphed.kas. This is the name of


d. In the Solver execution script column, select Dyna (Dyna).
e. In the Solver input arguments column, enter i= before $file.

a. Select the following parameters: E, shape1_h.S, shape1_w.S, and shape1_l.S.
b. Click Add.
c. Click OK


2. Click Apply.


5. Click Evaluate Tasks. The run is executed using LS-DYNA, and an approach/nom_1/
directory is created inside the study directory. The approaches/nom_1/run 00001/m_1
directory contains the glstat (for the strain energy), binout0000 (for the reaction
force), and d3hsp (for the structural mass) files, which are the result of the nominal run.

2. In the Add - HyperStudy dialog, add three output responses and label them Energy,
Force, and Mass.
3. In the Expression column of the output response Energy, click .
5. From the list of available functions, select max.
6. Click Insert Varname. The function max()appears in the Evaluate Expression field.
7. Click the File Sources tab.

8. Click Add File Source.
9. In the Add - HyperStudy dialog, add one Solver output file.

10. In the File column of Vector 1, click .

11. In the Vector Source dialog, navigate to the approaches/nom_1/run 00001/m_1
directory and open the glstat file.
12. From the Type, Request, and Component fields, select the options indicated in the
image below.
13. Click OK.

14. Click Insert Varname. The expression max(m_1_ds_1[0]) appears in the Evaluate
Expression field.
15. Remove [0] from the expression, so that it reads max(m_1_ds_1). This expression
produces the value of the energy value extracted from the nominal run.
16. Click OK.
17. Repeat steps 3 through 10 for the Force output response.
18. In the Vector Source dialog, navigate to the approaches/nom_1/run 00001/m_1
directory and open the binout0000 file.
19. From the Subcase, Type, Request, and Component fields, select the options indicated
in the image below.

20. Click OK.

21. Click Insert Varname. The expression max(v_2[0]) appears in the Evaluate
Expression field.
22. Remove [0] from the expression, so that it reads max(v_2). This expression produces the
value of the Force value extracted from the nominal run.
23. Click OK.
24. In the Expression column of the output response Mass, click .
25. In the Expression Builder, click the ASCII Extracts tab.
26. Click Add Extract Source.
27. In the Add - HyperStudy dialog, add one extract source.
28. In the File Path column of FileParser1, click .
29. In the Extract File dialog, navigate to the approaches/nom_1/run 00001/m_1
directory and open the d3hsp file.
30. To search for certain keywords within the d3hsp file, select the Keyword checkbox.
31. In the Keyword field, enter total mass of body.
32. Click Next. HyperStudy locates total mass of body in the file.
33. Highlight the value for total mass of body.
34. Right-click on the highlighted fields, and select Value from the context menu.

35. Click OK.

36. Click Insert Varname. The expression f_1[0] appears in the Evaluate Expression
field. This expression produces the value of the Mass value extracted from the nominal
run.
37. Click OK.
38. Click Evaluate Expressions to extract the output response values.
40. Click the Scatter 2D tab to view the values of the output responses, and to check that
your output responses have similar values.
Use the Channel selector to set the X Axis to Energy and the Y Axis to Force.
Note: The version and architecture of LS-DYNA that you are using may produce slightly
different output response values.

HS-1070: Defining Discrete Size Variables with Conditional

Linking for use in HyperStudy
This tutorial outlines the procedure for defining discrete input variables that are conditionally
linked. In this tutorial, beam thickness and material model values are used as variables.
Material model values are linked to the selection of the material to be used. The variables you
will use in this tutorial include: Shell Thickness, Young's Modulus, Poisson's, Ratio, and
Density. The output responses you will use in this tutorial includes: Mass.
The beam.fem model file that you will use in this tutorial can be found in <hst.zip>/HS-
1070/. Copy the tutorial files from this directory to your working directory.
2. From the menu bar, click Tools > Editor. The HyperStudy - Editor opens.
3. In the File field, open the beam.fem file.
5. Click until you find the PSHELL card.
6. In the same line as PSHELL, highlight the value 0.002 for the PSHELL thickness as
Note: In an OptiStruct deck, each field within a card is 8 characters long. Properly select
the value for the PSHELL thickness by selecting 0.002 and the three spaces that
follow.
8. In the Parameter - varname_1 dialog, Label field, enter Thickness.
9. Set the Upper bound to 0.0022, the Nominal to 0.0020, and the Lower bound to
0.0018.

12. Click OK.
13. Add three more input variables to the template file from the MAT1 card using the
information provided in the table below.
Parameter Selection Selection Input Lower Bound Format

Notification Variable & Upper
Label Bound
Young's Line number: 2.1E+11 & Young Lower Bound: %8.1e

Modulus 8267 1 space 7E+10
Fields: 16 to 24 Upper Bound:
2.1E+11
Poisson's Line number: 0.3 + 5 Poisson Keep default %8.5f

Ratio 8267 spaces values
Fields: 32 to 40
Density Line number: 7820.0 + 2 Density Keep default %8.3f

8267 spaces values
Fields: 40 to 48
14. Click Save.

15. In the Save Template dialog, save the file as beam.tpl.

17. Close the HyperStudy - Editor dialog.

18. In a text editor, open the beam.tpl file.
19. Replace:
{parameter(varname_3, "Poisson", 0.30000, 0.27000,

0.33000)}
{parameter(varname_4, "Density", 7820.000, 7038.000,
8602.000)}
with:
{if (varname_2==2.1E+11)}
{varname_3=0.3}
{varname_4=7820}
{elseif (varname_2==7E+10)}
{varname_3=0.33}
{varname_4=2700}
{endif}
20. Save the beam.tpl file and close the text editor.
click OK.

b. In the Solver input file column, enter beam.fem. This is the name of the solver input
5. Click Import Variables. Two input variables are imported from the beam.tpl resource
file.
8. For the input variable Thickness, set Mode to Discrete.
9. In the Values column for the input variable Thickness, click . A pop-up window opens.
11. Click Set.
12. Change the values in the table to the following: 0.002, 0.004, and 0.005.

13. Click OK.

14. For the input variable Young, set the Mode to Discrete.
15. In the Value column of the input variable Young, click . A pop-up window opens.
17. Click Set.
18. Change the values in the table to 7E+10 and 2.1E+11.
19. Click OK.

2. Click Apply.
4. Click Evaluate Tasks. The run is executed using OptiStruct, and all input files are written
to the study directory.
In this step you will create one output response, Mass.

1. From the Directory, drag-and-drop the beam.out file, located in
and click Next.
4. Define the following options, then click Next.

5. Label the output response Mass.

6. Set Expression to First Element.
Note: Since the mass is a scalar quantity, the only component of the data source is
index zero, therefore a scalar mass value is properly written in the Expression
field as m_1_ds_1[0]. Multi-axis vectors may be accessed through each individual
axis by using the corresponding index in the vector array.
The mass value can be accessed directly out of the beam.out file and does not
require any mathematical operations before being used as an output response for
your study. The Expression Builder is capable of performing a full complement
of mathematical functions on an arbitrary number of vector solutions in order to
build an output response.

7. Click Finish. The Mass output response is added to the work area.
Last modified: v2017.2 (12.1156684)

HS-1580: DOE Study Using CATIA
This tutorial demonstrates how to do a design of experiments study of an I-beam modeled in

CATIA. The overall beam width and height are the input variables; volume and maximum x-
displacement are the output responses.
In this tutorial you will create a macro in CATIA which will be used to update the CATIA model
in batch mode; mesh it using BatchMesher, and run the model in OptiStruct to solve for the
output responses of interest. You will perform a DOE study using HyperStudy to understand
the relationships between the input variables and output responses.
The files used in this tutorial can be found in <hst.zip>/HS-1580/. Copy the files from this
directory to your working directory. Modify the batch script (.bat) file for the current CATIA
install path.
There are two process environment variables used in the batch script (.bat) files:
• %STUDY_DIR_PATH%: Holds the study directory, where the study (.xml) file resides.
• %HST_APPROACH_MODEL_PATH%: Holds the full path of the current run directory on
Windows.
Step 1: Record a Macro in CATIA
Recording a macro in CATIA produces a .txt file that can be converted to a .tpl file and
used to perform a batch mesh operation in HyperStudy. The objective of the following steps is
to optimize the height and width on one side of this simple beam model.
1. Open CATIA and load the model, Beam.CATPart.
2. In the CATIA tree, expand Parameters to reveal Height and Width.
3. Optional. If Parameters is not visible in the tree, activate it by selecting the Parameters
checkbox in the Options dialog (the Options dialog can be accessed by clicking Tools >
Options > Infrastructure > Part infrastructure > Display from the menu bar).
Note: A macro is needed to drive the parameter modification.
4. From the menu bar, click Tools > Macro > Recording.
5. In the Record macro dialog, save the macro as Marco.catvbs.
6. Click Start.
7. Click Height.
8. In the Edit Parameter dialog, enter 100mm in the Height field.
9. Click OK.
10. In the Width field, enter 50mm.
11. Click Save.
12. Stop recording. CATIA creates a macro that is able to drive the geometry.
13. Close CATIA.

Step 2: Create The Base Input Template in HyperStudy
3. In the File field, open the CATIAHST.catvbs file.
4. Under the parameter Height, highlight the length1.Value, 100.000000.
6. In the Parameter-varname_1 dialog, enter Height in the Label field.
7. Set the Lower Bound to 50, the Initial to 100, and the Upper Bound to 200.
9. Click OK.
10. Under the parameter Width, highlight the length2.Value, 50.000000.

12. In the Parameter-varname_2 dialog, enter Width in the Label field.

14. Set the Lower Bound to 25, the Initial to 50, and the Upper Bound to 100.
16. Click OK.
17. Click Save.
18. In the Save Template dialog, save the file as 12.0Macro.tpl.
19. Click OK.
Step 3: Register a Solver Script
3. In the Add - HyperStudy dialog, accept the default settings and click OK.
4. In the Path column of Script 10, click .

5. In the Open dialog, open the script_win64.bat file.
6. Click Save.
8. In the File name field, enter a label for a new user preference file (example
userprefs.mvw).
9. Click Save.


11. Click Close.
12. From the menu bar, click File > Set Preferences.
13. In the HyperStudy - Set Preference File dialog, open the preferences file that you just
saved.
The default preferences file in the installation directory will be read, followed by the
preference file that you specify. This ensures that all solvers, readers, and import
templates are available.
Append the current user preferences file by clicking Append, or exit solver registration by
clicking Close. In the last case, the solver will only be registered for the current study.
click OK.
a. From the Directory, drag-and-drop the 12.0Macro.tpl file into the work area.
b. In the Solver input file column, enter CATIAHST.catvbs. This is the name of the
c. In the Solver execution script column, select Script 10 (scr_10).
5. Click Import Variables. Two input variables are imported from the 12.0Macro.tpl
file.



2. Click Apply.
directory. The approaches/nom_1/run 00001/m_1 directory contains the beam.out,
beam.oslog, and beam.h3d files, which are the result of the nominal run.

2. In the Add - HyperStudy dialog, add three output responses.
4. In the Expression Builder, click the File Sources tab.
5. Click Add File Source.
6. In the Add - HyperStudy dialog, add one solver output file.
7. In the File column of Vector 1, click .

8. In the Vector Source dialog, File field, navigate to the
approaches/nom_1/run 00001/m_1 directory and open the Beam.out file.

9. From the Type, Request, and Component fields, select the options indicated in the
image below.
10. Click OK.

11. Click Insert Varname. The expression m_1_ds_1[0] appears in the Evaluate
Expression field.
12. Click Evaluate Expression. The expression m_1_ds_1[0] changes to 2000000.
13. Click OK.

17. In the Add - HyperStudy dialog, add one extract source.
18. In the File Path column of FileParser1, click .
19. In the Extract file dialog, navigate to the approaches/nom_1/run 00001/m_1
directory and open the Beam.out file.
20. Select the Keyword checkbox and enter Total # of Grids (Structural).
21. Click Next. The phrase Total # of Grids (Structural) is highlighted.

23. Right-click on the highlighted fields and select Keyword from the context menu. The
keyword Total # of Grids (Structural) is created.
24. Highlight the value 4141 to the right of Total # of Grids (Structural).
25. Right-click on the highlighted fields and select Value from the context menu. This number
gives you the total number of grids.
Note: You will need this number because the number of grids will change for each
design as they are re-meshed, therefore the index for searching the maximum x-
displacement among all the grids.
26. Click OK.
27. Click Insert Varname. The expression f_1[0] appears in the Evaluate Expression
field.
f_1, thereby choosing the first (and only) entry in the data source.
28. Click Evaluate Expression. The expression f_1[0] changes to 4141.
29. Click OK.

32. From the list of functions, select max.
33. Click Insert Varname. The function max() appears in the Evaluate Expression field.

35. From the list of functions, select ResVector.

36. Click Insert Varname.
37. In the resvector - Builder dialog, File field, navigate to the
approaches/nom_1/run 00001/m_1 directory and open the Beam.h3d file.
38. From the Subcase, Type, Request, and Component fields, select the options indicated
in the image below.
37. Click OK. The expression changes to

max(resvector(getenv("HST_APPROACH_RUN_PATH") +
"/m_1/Beam.h3d",1,0,4140,3,3,0,0)).
38. In the expression, replace 4140 with r_2-1.
39. Click Evaluate Expression. The expression
max(resvector(getenv("HST_APPROACH_RUN_PATH") + "/m_1/Beam.h3d",1,0,r_2-
1,3,3,0,0)) changes to 0.16365569829940796.

40. Click OK.
4. In the work area, set the Mode to Fractional Factorial.
5. Click the Levels tab, and verify that the number of levels is 2.
6. Click Apply.
18. After the results are extracted, go to the Post processing step.

HS-1590: DOE Study Using Madymo
In this tutorial you will setup a DOE study using Madymo as a solver. Madymo simulation is a
crash occupant safety application. The input variables used in this tutorial are scale of force-
penetration curve, scale of recliner loading function, position of the head rest.
The objective of this tutorial is to study the effect of changing the dummy-seat contact
stiffness, seat recliner loading function, and the distance between the dummy head and the
head rest on the dummy's neck injury criteria.
directory to your working directory.
2. In the File field, open the a_hffoam_TEST.xml file.
3. In the Search area, enter GROUP_DEFINE.
4. Click . GROUP_DEFINE is highlighted.
5. Under the variable Load_Function_scale_factor, highlight the value 1.0.
7. In the Parameter - varname_1 dialog, Label field, enter
Load_Function_scale_factor.

8. Set the Lower Bound to 0.7, the Initial value to 1.0, and the Upper bound to 1.3.
10. Click OK.
11. Repeat steps 3 through 10 to create two more parameters using the information provided
in the table below.
Label Lower Initial Upper Format

Bound value Bound
Head_Rest_position 0.00 0.02 0.04 %8.5f
Force_Penetration_scale_facto 0.8 1.0 1.2 %8.5f

r
12. Click Save.

13. In the Save Template dialog, save the file as a_hffoam_TEST.tpl.
Step 2: Create the madymo_DOE.bat File
1. In a text editor, open a new file.

2. Enter the following text line: "C:\Program Files\Madymo\madymo_75\em64t-
win\bin\madymo75.exe" -i a_hffoam_TEST.xml
Note: This example assumes that you are using MADYMO V7.5 on a Microsoft Windows
platform.
3. Save the file as madymo_DOE.bat.

Step 3: Register Madymo as a Solver Script
3. In the Add - HyperStudy dialog, enter MADYMO_DOE in the Label and Varname fields.
5. Click OK.
6. In the Path column of the script MADYMO_DOE, click .

7. In the Open dialog, open the madymo_DOE.bat file.
8. Click Close.
click OK.

a. From the Directory, drag-and-drop the a_hffoam_TEST.tpl file into the work area.
b. In the Solver input file column, enter a_hffoam_TEST.xml;d_hyb350el_inc.xml.

This is the name of the solver input file HyperStudy writes during any evaluation.
c. In the Solver execution script column, select MADYMO_DOE (MADYMO_DOE).
6. Click Import Variables. Three input variables are imported from the
a_hffoam_TEST.tpl file.

2. Click Apply.
directory.

2. In the Add - HyperStudy dialog, add four output responses and label them NTF, NTE,
NCF, and NCE.

4. In the Expression column of the output response NTF, click .

7. In the Add - HyperStudy dialog, add one extract source labeled NTF_peak.
8. In the File Path column of NTF_peak, click .

directory and open the a_hffoam_TEST.peak file.
10. Select the Keyword checkbox and enter NTF_ing.
11. Click Next. HyperStudy locates NTF_inj in the file.
12. In the Offset field, enter 87.
13. In the Length field, enter 11.
14. Click OK.
15. Click Insert Varname. The expression NTF_peak[0] appears in the Evaluate
Expression field.
16. Click Evaluate Expression The expression NTF_peak[0] changes to 0.0220855.
17. Click OK.
18. Create and define extract sources for the output responses NTE, NCF, and NCE by
repeating steps 3 through 16, except change the labels and keywords as indicated in the
image below.

5. Click the Levels tab, and change the number of levels for all variables from 2 to 3.
6. Click Apply.
8. Click Evaluate Tasks. All 27 simulations are extracted.

HS-1600: Creating an Optimization Study with

MADYMO/Workspace Objective Rating and HyperStudy
In this tutorial, you will learn how to setup an optimization study that combines
MADYMO/Workspace Objective Rating together with Hyperstudy. At the end of this tutorial,
you will find the maximum correlation between a MADYMO model and the hardware test
results.
You will set up an Objective Rating using the following rating criteria: Sprague & Geers, CORA,
peak timing and value matching. The rating is subsequently used to match results from the
hardware testing of a pedestrian leg impactor with a vehicle front with a MADYMO model of
this test. You will use Hyperstudy to find the maximum correlation between the test and
simulation results. The simulation model contains the following input variables: scaling factors
for the loading characteristics for the bumper, spoiler, headlights, bonnet leading edge and
bonnet.
The files used in this tutorial can be found in <hst.zip>/HS-1600/.
Step 1: Setup the MADYMO/Workspace Objective Rating Matrix
1. Start the MADYMO/Workspace Objective Rating application.

2. From the menu bar, click Tools > Settings.
3. In the Settings dialog, select Criteria from the list of settings.
4. From the Criteria Used in the Project list, select Global Min Time, Global Min Value,
and Weighted Integrated Score.

6. Click . The application removes the selected criteria from the list of rating criteria to be
used in this project.
7. From the Available Criteria list, select CORA and Sprague and Geers.
8. Click . The application adds the selected criteria to the list of rating criteria to be used
in this project.
9. Click OK.
10. On the toolbar, click .

11. From the filter field, select All Files.

13. Open the ACC.csv file.

15. Open the V3_LowerLeg_form.lac file.
16. In the Tree View, select the ACC.csv file.
17. Drag the file it into Row 1, Column 1 of the Rating matrix and then drop it into the
Reference signal.
18. In the Tree View, select the Res.acceleration file.

19. Drag the file it into Row 1, Column 1 of the Rating matrix and then drop it into the
Simulation signal.
20. In cell 1, click the Cell Properties tab.

21. Expand Reference curve and then select the X axis.

22. Set the Dimension in file to Time.
23. Set the Offset to 0.0085 s.
24. Select the Y axis.

25. Set the Dimension in file to Acceleration.
26. Set the Units in file to g.
27. Set the Scale to -1.

29. In the Save dialog, navigate to your working directory and save the file as rating.obr.
30. Exit the Objective Rating application.
3. In the File field, open the V3_LowerLeg_form.xml file.
4. In the Search area, enter cnt_char_ell69_bumper.
5. Click . cnt_char_ell69_bumper is highlighted.

7. Under the variable cnt_char_ell69_bumper, highlight the value 1.0.
9. In the Parameter - varname_1 dialog, enter cnt_char_ell69_bumper in the Label
field.
10. Set the Lower Bound to 0.1, the Initial value to 1.0, and the Upper bound to 2.0.
12. Click OK.
13. Repeat steps 3 through 11 for the variables listed below. Change the labels to be the
name of the variable above the highlighted values.
• cnt_char_ell70_spoiler
• cnt_char_ell71_headlights
• cnt_char_ell72_bonnet_lead_edge
• cnt_char_ell73_bonnet
14. Click Save.
15. In the Save Template dialog, navigate to your working directory and save the file as
V3_LowerLeg_form.tpl.

Step 3: Create the madymo_optimization.bat File
1. In a text editor, open a new file.

2. Enter the following text line:
"C:\Program Files\Madymo\madymo_75\em64t-win\bin\madymo75.exe" -i
V3_LowerLeg_form.xml
"C:\Program Files\Madymo\Workspace_75\em64t-win\bin\ObjectiveRating.exe" -
-batch --input Rating.obr --output Rating.obr
Note: This example assumes that you are using MADYMO V7.5 on a Microsoft Windows
platform.
3. Save the file as madymo_optimization.bat.
Step 4: Register Madymo as a Solver Script
3. In the Add - HyperStudy dialog, enter MADYMO_optimization in the Label and
Varname fields.
5. Click OK.
6. In the Path column of the script MADYMO_optimization, click .

8. In the Open dialog, open the madymo_optimization.bat file.

9. Click Close.
click OK.
a. From the Directory, drag-and-drop the V3_LowerLeg_form.tpl file into the work
area.
b. In the Solver input file column, enter V3_LowerLeg_form.xml; Rating.obr;

ACC.csv. This is the name of the solver input files HyperStudy writes during any
evaluation.
c. In the Solver execution script column, select MADYMO_optimization
(MADYMO_optimization).
d. In the Solver input arguments column, add the extension .xml to the $file.
V3_LowerLeg_form.tpl file.


2. Click Apply.
directory.

2. In the Add - HyperStudy dialog, one output response and label it Rating.
3. In the Expression column of the output response Rating, click .
6. In the Add - HyperStudy dialog, add one extract source labeled OBR.
7. In the File Path column of OBR, click .

directory and open the Rating.obr file.
9. Select the Keyword checkbox and enter <TOTALS>.
10. Click Next. <TOTALS> is highlighted.
11. In the Offset field, enter 157.
12. In the Length field, enter 8.
13. Click OK.
14. Click Insert Varname. The expression OBR[0] appears in the Evaluate Expression
field.
15. Click Evaluate Expression. The expression OBR[0] changes to 0.47476600000000002.
16. Click OK.
Run an Optimization Study


b. Set Apply On to Rating (r_1).
10. Click Apply.

13. Click Apply.
16. Click the Evaluation Plot tab to monitor the progress of the optimization.

HS-1605: Setting Up an ANSYS Workbench Model
In this tutorial you will learn how to setup an Ansys workbench model in HyperStudy.
HyperStudy’s interaction with Workbench is based on interpreting Workbench inputs as input
variables and Workbench direct output as output responses. The inputs and direct outputs
must be identified in the Workbench project file (*wbpj).
There are no files associated with this tutorial. A valid copy of Ansys Workbench and a user
supplied properly constructed Workbench project file are required to execute this tutorial.
click OK.
4. Add a Workbench project file (*wbpj) by dragging-and-dropping it from the Directory
into the work area.
Data is inserted into the Solver input file, Solver execution script, and Solver input
arguments columns.
Note: The Message log reports the number of imported input variables and output
responses based on the existing inputs and direct outputs defined in your
Workbench project file.
6. Review the input variable's lower and upper bound ranges. Clear the checkboxes of any
variables that should not be considered for this study.
Step 2: Register Workbench Solver
1. From the menu bar, click Edit > Register Solver Script. The Register Solver Script
dialog opens.
2. In the Path column of the script Workbench, click .

3. In the Open dialog, select the appropriate path to the workbench executable on your
computer.
4. Optional. Click Save to save these settings to a preference file (*mvw) for future use.
5. Click Close.


2. Click Apply.
directory.
Step 4: Review Output Responses
1. Review the output responses extracted from Workbench project file (*wbpj) selected in
step 1.7.
2. Optional. Clear the checkboxes of any output responses that should not be considered for
this study.

HS-1610: Setting Up a SimLab Model Using JavaScript
In this tutorial you will set up a mesh refinement study using a SimLab model to investigate
the relationship between the SimLab mesh parameters and max Stress and max displacement.
Model Files
The files used in this tutorial can be found in <hst.zip>/HS-1610/. Copy these files to your
working directory.
• ConnectingRod.xmt_txt
• Conrod_js_script.js
• HST_CONROD_Params.xml
• HST_CONROD_Responses.xml
• Mesh_spec_param.xml
The model used in this tutorial is a Parasolid CAD file (ConnectingRod.xmt_txt) that contains a
connecting rod. The connecting rod is loaded at one end and constrained at the other.
Connecting rod mesh representation with loads and boundary conditions

Pre-Requisites
In this tutorial you are required to use Simlab v2017.2.1 and HyperStudy 2017.2.2.
HyperStudy v2017.2.2 does not require you to set up the environment variable
HW_HST_CMD_SIMLAB.
Before creating the parameters inside Simlab, pause the recording of javascript. This is a
known limitation of SimLab. The parameters are passed separately from the .xml file to the
.js file. If the parameter definition already exists in the .js file, then any changes in the
values will be overwritten.
Exercise
Step 1: Register the SimLab Solver Script

1. Open HyperStudy.
3. In the Register Solver Script – HyperStudy dialog, SimLab row, Path column, click
.
4. In the Open dialog, navigate to SimLab2017.2.1/bin/win64 and open the
SimLab.bat file.
5. Click OK to close the Register Solver Script – HyperStudy dialog.
Step 2: Perform the Study Set Up
and click OK.
4. Add a SimLab model.
a. Click Add Model.
b. In the Add – HyperStudy dialog, select SimLab and click OK.

e. In the HyperStudy – Load model resource dialog, navigate to HST_CONROD

and open the Conrod_js_script.js file.
Notice: The Solver Input Arguments field automatically displays with -

nographics -auto Conrod_js_script.js -param
HST_CONROD_Params.xml -response HST_CONROD_Responses.xml.

7. In the work area, Active column, clear the checkboxes for the FilletMeshSize and Load
input variables.
Note: In this tutorial you will only focus on the BodyMeshSize input variable.
8. For the BodyMeshSize input variable, change the Lower Bound to 2.0 and the Upper
Bound to 8.0.
Step 3: Perform the Sweep

1. In the work area, set the Mode to Sweep.
2. From the Settings, set the Number of Runs to 7.
3. Click Apply.
5. From Run Tasks, click Multi-Execution and enter 4 jobs.



The Max_Von_mises output responses was extracted from the Conrod_js_script.js file in
the SimLab model. In this step you will create an additional output response, Max_Disp.
1. From the Directory, drag-and-drop the Conrod.h3d file, located in
and click Next.
3. Select Multiple items at multiple time steps (readsim), then click Next.
4. Define the following settings and click Next.
a. Set Subcase to Subcase 1.
b. Set Type to Displacement (Grids).
c. For Request, set Start to First Request and End to Last Request.
d. For Components, select MAG.
e. Set Time to All.

5. Define the following settings and click Finish.

a. Set Data Set Dimensions to Single Data Source (1).
b. Clear the Create a combined Data Source checkbox.
c. Enable the Create individual Responses(1) checkbox and set it to
Maximum.
6. In the work area, Label the second output response Max_Disp.
Step 5: Extract Evaluation Data for all 7 Runs in the Sweep Study
2. From the Run Tasks tab, clear the Write Input Files and Execute Analysis
checkboxoes.

Step 6: Post-Processing
2. Click the Scatter tab.
4. From the Channel selector, set the X Axis to BodyMeshSize and the Y Axis to
Max_Von_Mises and Max_Disp.

The results of the scatter plot indicate that as the size of the mesh gets smaller
(moving along the x-axis to the left), displacement starts to converge. However, stress
does not converge. This behavior is typical in finite element models when displacement
converges before derived quantities such as stress. In this tutorial, the Max_Von_Mises
output response may not converge at all due to the location of the maximum stress in
the model (adjacent to the load application area), which can be seen by opening the
resulting file in HyperView.

HS-1615: Setting Up a FEKO Model
The purpose of this tutorial is to illustrate the information and steps required to setup a FEKO
model in HyperStudy.
The model used in this tutorial is a waveguide transmission line that is being fed with a coaxial
cable.
The effect of the cable’s pin position on input impedence is studied. When the impedence is
reduced, this leads to improved power transmission.
• waveguide_pin_feed_placement.cfx
• waveguide_pin_feed_placement.cfx_extract.lua
click OK.
5. Add a FEKO model by dragging-and-dropping the waveguide_pin_feed_placement.cfx
from the Directory into the work area.
populated.
6. Click Import Variables. Ten input variables are imported from the
waveguide_pin_feed_placement.cfx file.
7. Go to the Define Input Variables step, and review the input variables.


2. Click Apply.
directory.
In this step you will create two output responses.

1. Create output response 1.
a. From the Directory, drag-and-drop the hst_output.hstp file, located in
b. In the File Assistant dialog, click Next.


e. Define the following, then click Next.
• Set Type to Output.
• Set Request to z_Imag.
f. Optional. Enter labels for the data source and output response.
g. Set the output response Expression to Maximum.

i. Click Finish. Output response 1 is added to the work area.

2. Create output response 2 by repeating step 1. Except, set Request to z_Real.
Step 4: Run a DOE Sweep Study
2. In the Add - HyperStudy dialog, add a Doe.
4. This tutorial is only studying the effects from the variation of the pin position index,
therefore deactivate all input variables except n.


6. In the work area, set the Mode to Hammersley.
7. In the Settings tab, change the Number of runs to 21.
8. Click Apply.
10. Click Evaluate Tasks to execute all 21 runs.
Step 5: Run a Fit Study
2. In the Add - HyperStudy dialog, add a Fit.
6. Set Matrix Source to Doe 1 (doe_1).
7. Click Import Matrix.

9. In the work area, set the Mode to Radial Basis Function.
10. Click Apply.
14. Click the Trade-Off tab to plot the response surface.


HS-1620: Settings Up a Flux Model
An electromagnetic actuator is a device that converts an electric current into a mechanical

output. In this tutorial, you will be investigating the relationship between actuator dimensions
and the mechanical force output. The actuator is composed of an U magnetic core, mobile
magnetic part, and two coils supplied by an amp-turn number. The finite element model is
created and analyzed by Flux. The Flux model also contains the input variables and output
responses of interest.
Electromagnetic actuator and its Finite element model
Input variables
Note: This tutorial requires registration of the installed version of Flux Software with
HyperStudy. For details, see Flux Model.
click OK.
5. Add a Flux model by dragging-and-dropping the Example.F2G from the Directory into the
work area.

The Resource, Solver input file, Solver execution script, and Solver input arguments fields
become populated.
7. Click Import Variables. Ten input variables are imported from the Example.F2G file.
8. Go to the Define Input Variables step, and review the input variables.

2. Click Apply.


directory. The approaches/nom_1/run 00001/m_1 sub-directory contains the run files.
1. Review the output responses.

Step 4: Run a Hammersley Stochastic Study
2. In the Add - HyperStudy dialog, add a Stochastic.
3. In the Select Input Variables step, clear the Active checkbox for GAP.
4. For each input variable, adjust the lower and upper bounds, which will be used to
calculate statistical distribution settings, in this case the variance of normal distribution.
a. In the Initial column, click .
b. In the window, enter 1 in the percent field and click +/-.
c. Click OK.


7. In the Settings tab, verify that the Number of runs is 100.
Note: If you are using a laptop with a Win64 operating system, 32 GB ram, Intel Core
i7 CPU, it may take around 40 minutes to run 100 Flux simulations. To reduce the
run times, change the Number of Runs to a lower value. However, a high number
of runs is necessary for better statistical accuracy.
8. Click Apply.
12. Click the Scatter 2D tab to plot correlation values.
Correlation measures the strength and direction between associated variables. Correlation
coefficients can have a value from -1 to 1; -1 indicates a strong but negative correlation
and 1 indicates a strong and positive correlation.
The correlation values for variables AT, DEPTH, and TCORE with Force are 0.80, 0.43, and
0.39, which indicates that Force is correlated to AT and Force is somewhat correlated to
DEPTH and TCORE. These three correlations are positive, meaning that you should expect
to see an increase in Force corresponding to an increases in AT, DEPTH, and TCORE. You
can also expect to see no changes in Force corresponding to changes in other variables.
DEPTH and TCORE are somewhat correlated to FORCE, therefore you may see deviations
from these predicted behaviors.

You can visualize these correlations in the Scatter2D plot. In the plot for FORCE vs AT,
you can see the design cloud follows a nice pattern of increasing Force with increasing AT.
In the plot for WCOIL vs FORCE, you should not observe any relationship between the
two.

Approach Overview

HS-1700: Simple DOE Study
In this tutorial you will learn how to run a DOE study on simple functions defined using a
Templex template.
The base input template defines two input variables; DV1 and DV2, labeled X and Y,
respectively. The objective of the study is to investigate the two input variables X, Y forming
the two functions: X+Y and 1/X + 1/Y – 2.
Before running this tutorial, you must complete tutorial HS-1010: Simple Study Setup or you
can import the archive file HS-1010.hstx, available in <hst.zip>/HS-1700/.
5. Click the Levels tab, and change the number of levels from 2 to 3 to spread the levels
between the lower and upper bounds.
6. Click Apply.
8. Click Evaluate Tasks. The results of the evaluation display in the work area.

Step 2: Post Process the Results of the DOE Study
In this step you will review the effects and interaction between both input variables and output
responses.
1. Click the Linear Effects tab.
a. Above the Channel selector, click to plot the linear effects.

b. Using the Channel selector, select both input variables and output responses.
c. Review the effects of Area 1 and Area 2 on Response 1 and Response 2. You can
observe that the effects of Area 1 and Area 2 on Response 1 are the same
(proportional with a magnitude 4.8). From the second plot, you can observe that the
effects of Area 1 and Area 2 on Response 2 are also the same (inversely proportional
with a magnitude -4.8). For information on how to calculate the magnitude in DOE
refer to Post Processing DOE.
2. Click the Interactions tab.

a. Using the Channel selector, set Variable A to Area 1 and Variable B to Area 2.
b. Review the interactions between Area 1 and Area 2 on Response 1 and Response 2.
From both plots, you can observe that there is no interaction between Area 1 and
Area 2 for both Response 1 and Response 2.

HS-1705: Simple Fit Study
This tutorial demonstrates how to run a DOE study on simple functions defined using a
Templex template.
The base input template defines two input variables; DV1 and DV2, labeled X and Y,
respectively. The objective of the study is to investigate the two input variables X, Y forming
the two functions: X+Y and 1/X + 1/Y – 2.
Before running this tutorial, you must complete tutorial HS-1700: Simple DOE Study or you
can import the archive file HS-1700.hstx, available in <hst.zip>/HS-1705/.
Step 1: Run a Space Filling DOE Study
5. Click Apply.
7. Click Evaluate Tasks. The evaluation results display in the work area.
9. Click the Scatter tab to view a plot which illustrates the dependency between Area 2 and
Response 1 and Response 2.
a. Using the Channel selector, set the X Axis to Area 2 and the Y Axis to both
Response 1 and Response 2.

b. Compare the scatter plots to determine if the runs are distributed homogeneously
throughout the design space.

Step 2: Run a FIT Study
2. In the Add – HyperStudy dialog, select Fit and click OK.
3. Go to the Select Matrices step.
5. In the Add – HyperStudy dialog, add one matrix.
6. In the work area, set Matrix Source to Doe 2 (doe_2).

10. Click Apply.
14. Click the Residuals tab to review the residuals of both output responses.
The data in the table shows the differences in the actual values and the predictions from
the constructed Fit. The Percent Error column of Response_1 is numerically zero for all
six runs; whereas the Percent Error column of Response_2 is up to 35%. The LSR fitting
for Response_1 is acceptable, but the LSR fitting for Response_2 is rather large.
15. Click the Diagnostics tab to review the overall Fit quality.
Several measures are shown to indicate the relative quality of the Fit. The R-Square
value can be interpreted as the percentage of variance in the data that can be explained
by the Fit. For Response_1, the Fit captures 100% of the data variance; this makes sense
as Response_1 is actually a linear function so the first order regression matches the actual
data with no error. For Response_2, it is shown below that the Fit explains about 90% of
the variance.

16. With first order least squares, you have a Fit which explains most of the data’s variance,
but it still has a relatively high prediction error. Go back to the Specifications step and
try different methods until you find an acceptable fitting for both output responses.

HS-1710: Simple Optimization Study
This tutorial demonstrates how to optimize a simple function defined using a Templex
template. The base input template defines two input variables, DV1 and DV2, labeled X and Y,
respectively. The objective of the optimization is to minimize X + Y with the constraint 1/X +
1/Y – 2 < 0.
Before running this tutorial, you must complete tutorial HS-1010: Simple DOE Study (or HS-
1700: Simple DOE Study, HS-1705: Simple Fit Study) or you can import the archive file HS-
1705.hstx, available in <hst.zip>/HS-1710/.
b. Set Apply On to Response 1 (r_1).
9. Click the Constraint tab.

10. Click Add Constraint.
11. In the Add - HyperStudy dialog, add one constraint.
12. Define the constraint.
a. Set Apply On to Response 2 (r_2).
b. Set Bound Type to <= (greater than or equal to).
c. In the Bound Value column, enter 0.0.
13. Click Apply.


17. Click Apply.
20. Optional. Click the different tabs in the Evaluate step to monitor the progress of the
Optimization.
21. After the optimization has finished, review the optimization history first.
22. With many of the algorithms (SQP, GA, …), each iteration requires many evaluations.
• Evaluation (plots or tables) show all runs performed during the optimization.
• Iteration (plots or tables) show the optimization iterations.
The Iteration History tab uses color coding to indicate which design are feasible,
optimal, and violated.
• White background/black font indicates the design is feasible.
• White background/red font indicates the design is violated.
• White background/orange font indicates the design is acceptable, but at least one
constraint is near violated.
• Green background/white font indicates the design is optimal.
• Green background/orange font indicates the design is optimal, but at least one
constraint is near violated.
23. The Evaluation Plot tab displays charts for all of the entities in the optimization (input
variables, output responses, objective functions, constraints) against the iteration.
• Select entities to plot using the Channel selector in the left pane.
• Plot multiple entities in separate windows by clicking .

Step 2: Post-Processing of an Optimization Study
In the Post-Processing step of an Optimization approach, you can access additional tools to
review the results. Use the Integrity, Distribution, and Scatter tabs to compare and
analyze designs.
1. Click the Integrity tab to analyze statistics of the optimization study.

HS-1715: Simple Stochastic Study
This tutorial demonstrates how to run a stochastic study on simple functions defined using a
Templex template.
The base input template defines two input variables, DV1 and DV2, labeled X and Y,
respectively. The objective of the study is to investigate the two random variables X, Y
forming the two functions X+Y and 1/X + 1/Y – 2.
Before running this tutorial, you must complete tutorial Simple Optimization Study - HS- 1710
or you can import the archive file HS-1710.hstx, available in <hst.zip>/HS-1715/.
Step 1: Run a Stochastic Study
2. In the Add - HyperStudy dialog, select Stochastic and click OK.
4. In the work area, set the Mode to Latin Hypercube.
6. Click Apply.
8. Click Evaluate Tasks. HyperStudy executes 50 runs and extracts the results.
Step 2: Post-Process the Results of the Stochastic Study
In the Post processing step of a Stochastic approach, you can access additional tools to
review results.
1. Click the Distribution tab.
2. Using the Channel selector, select Response 1.

4. Click the Reliability tab.

5. Click Add Reliability.
6. In the Add - HyperStudy dialog, add two reliabilities.
7. Define reliability 1.
a. Set Response to Response1 (r_1).
b. Set Bound Type to <= (less than or equal to).
8. Define reliability 2.
a. Set Response to Response2 (r_2).
b. Set Bound Type to <= (less than or equal to).
9. Study the effects of bounds on the reliability by entering different values in the Bound
Value column.
• For reliability 1, enter 1.050. The reliability value of Response_1 to be less than
1.05 is 52%.
• For reliability 2, enter 1.830. The reliability value of Response_2 to be less than ... is
...%.

HS-1810: Post Proceesing with HyperStudy
This tutorial demonstrates how to post process studies with HyperStudy. We will use the
model described in the HyperStudy tutorial HS-4415 and create various appraoches (Design of
Experiment, Approximation, Optimization, Stochastic) to illustrate the rich variety of tools and
post processing methods offered by HyperStudy.
Before running this tutorial, you must complete tutorial HS-4415: Optimization Study of a
Landing Beam using Excel or you can import the archive file HS-4415.hstx, available in
<hst.zip>/HS-1810/.
7. In the Settings tab, set Resolution to III.
8. Click Apply.
11. While the DOE is in progress, click the Tasks tab to view the feedback on the results of
the evaluation.
12. During the execution of the DOE, you can monitor the evaluation of the 16 runs in either
the Evolution Plot or Evolution Data tabs.
Step 2: Post Processing of the Results of the DOE Study
1. Click the Summary tab to view all input variable and output response run data in a table.
Tip: Use the Sort and Find options in the right-click context menu to sort and search
data.

2. Click the Integrity tab to view statistical measures over the population (samples of the
DOE) for all of the input variables and output responses.
3. Click the Distribution tab to view a histogram of the DOE results.

Display the entities in one or multiple plots by clicking . Each plot presents the range of
the entity in abscissa (an input variable or an output response), the histogram (red bar
chart for frequency), probability distribution function (PDF in green) and cumulative
probability distribution function (CDF in blue).

4. Click the Scatter tab to plot the DOE results.

Use the Channel selector to select entities to plot. Select one entity for the X-Axis, and
select one or more entities for the Y-Axis.
5. Click the Scatter 3D tab to view DOE results in a scatter plot.

Only one input variable/output response can be selected for the X and Y axes, whereas
multiple input variables/output responses can be selected for the Z axis.
6. Click the Linear Effects tab to review the effect of an input variable on an output
response, ignoring the effects of other input variables.
a. Above the Channel selector, click to plot the linear effects. Use the Channel
selector to select the variables AA_w1 and AA_w2 and the output response Area
ACE.

c. Above the Channel selector, click to view the linear effects in a table.
Tip: From the Channel selector, use the Sort and Filter options in the right-click
context menu to sort and filter effects.
1. Repeat Step 1: Run a DOE Study to add a second DOE to the study. In the
Specifications step, set the Mode to Hammersley and change the Number of runs to
50.
2. Repeat Step 1: Run a DOE Study once more to add a third DOE to the study. In the
Specifications step, set the Mode to Latin HyperCube and change the Number of
runs to 15.

Step 4: Run a FIT Study
2. In the Add - HyperStudy dialog, select Fit and click OK.
5. In the Add - HyperStudy dialog, add three matrices.
6. Define the matrices by selecting the options indicated in the image below from the
Type and Matrix Source columns.

9. In the work area, set the Mode to Radial Basis Function (RBF).
10. Click Apply.
13. To review the values of the output responses and their approximations while the
evaluation is in progress, click the Evaluation Data and Evaluation Plot tabs.

Step 5: Post Processing of the Results of the FIT Study
1. Click the Residuals tab to identify errors for each design.

The error (and percentage) between the original output response and the approximation is
listed for each run of the input, cross-validation, or validation matrices.
2. Click the Diagnostics tab to assess the accuracy of a Fit. Different criteria is displayed for
the Input, Validation, and merged matrices.
Use the Channel selector to select a specific output response for which to review
diagnostics.
3. Click the Trade-Off tab to modify the values of input variables in order to see their effect
on the output response approximations.
Use the Channel selector to select the desired output responses to display in the
Outputs pane. Input variable controls are located in the in the top frame (Inputs).
Change each input variable by moving the slider in the first Value column, or by

entering a value into the second Value column. Set input variables to their initial,
minimum, or maximum values by moving the slider in the upper right-hand corner of the
Inputs frame.
4. In the Trade-Off tab plot variables and output responses in order to see the input
variables effect on the output response approximations.
Select input variables to plot by enabling its corresponding X Axis checkbox in the Inputs
pane. Use the Channel selector to select output responses to plot. The values for the
input variables which are not plotted are modified in the top frame (Inputs).
Move the sliders in the Value column to modify the other input variables, while studying
the output response throughout the design space.

6. Add an objective.
a. Click Add Objective.
b. In the Add - HyperStudy dialog, add one objective.
c. Define the objective.
• Set Type to Minimize.
• Set Apply On to Area ACE (r_10).
7. Add constraints.
b. Click Add Constraint.
c. In the Add - HyperStudy dialog, add nine constraints.
d. Define Constraint 1 through Constraint 9 by selecting the options indicated in the
image below from the Apply On, Bound Type, and Bound Value columns.
8. Click Apply.

11. Click Apply.

13. Click Evaluate Tasks to launch the Optimization.
14. Click the Evaluation Plot tab to plot variables and output responses across runs
(abscissa are run numbers, not iterations).
15. Click the Iteration Plot tab to plot variables and output responses against iterations.
When the constraint history is plotted, the constraint bounds can be marked with a datum
line. Use the Channel selector to select a constraint, then click (located above the
Channel selector) and select Bounds.

6. Click Apply.
7. Go to the Evaluate step
Step 8: Post Processing the Results of a Stochastic Study
During the Post Processing step of a Stochastic approach, you can access additional result
analysis tools.
1. Click the Integrity tab to access a series of statistical measures on input variables and
output responses.
2. Click the Distribution tab to view variable and output response data in a histogram.

4. Click the Scatter to view sampling patterns and possible correlations between output
responses or between input variables and output responses.
Use the Channel selector to select entities to plot. Select one entity for the X-Axis, and
select one or more entities for the Y-Axis.
5. Click the Reliability tab to compute the probability of failure (bound is violated) and

the reliability (bound is respected).

a. Click Add Reliability.
b. In the Add - HyperStudy dialog, add one reliability.
c. Define the reliability.
• Set Response to Area ACE (r_11).
HyperStudy computes the reliability and probability of failure in the Reliability and
Probability of Failure columns.

Design of Experiment (DOE)

HS-2000: DOE Method Comparison: Arm Model Study
In this tutorial you will:

• Set up a Full Factorial DOE
• Set up a Fractional Factorial DOE with Resolution III
• Set up a Fractional Factorial DOE with Resolution IV
• Set up a Fractional Factorial DOE with Resolution V
• Compare these methods in accuracy and efficiency
directory to your working directory. The tutorial directory includes the following files:
arm_model.tpl Template file
arm_model.optistruct.node.tpl Grid coordinates template
arm_model.shp Grid perturbation vector data for

arm_model.optistruct.node.tpl
Problem Definition
An arm model is used for this tutorial; the arm is clamped at one end and under an axial
loading on the other end. It has been meshed and modeled in HyperMesh and analyzed using
OptiStruct.
Concentrated load and boundary

conditions Shapes defined on the arm model

The design can change shape in nine different regions, as shown in the image above, and
three output responses are of interest as detailed below.
Shape input Length1: Lower Bound = -0.5, Initial Bound = 0.0, Upper Bound = 2.0
variables
Length2: Lower Bound = 0.0, Initial Bound = 0.0, Upper Bound = 2.0
Length3: Lower Bound = -1.0, Initial Bound = 0.0, Upper Bound = 1.0
Radius1: Lower Bound = -2.0, Initial Bound = 0.0, Upper Bound = 2.0
Height: Lower Bound = -1.0, Initial Bound = 0.0, Upper Bound = 1.0
Output Volume (Nominal design volume 1.7667e6 mm3)

Responses
Max Von Mises Stress (Current design value = 195.29 MPa)
Nodal displacement of the node where the force is applied (mm) (1.411
mm)
Shape variables are created using HyperMesh’s morphing technology HyperMorph. In this
tutorial you will start from an already morphed model. Moreover these shape variables are
already exported for HyperStudy and a template is created.
click OK.
a. From the Directory, drag-and-drop the arm_model.tpl file into the work area.

c. In the Solver input file column, enter crank.fem. This is the name of the solver
input file HyperStudy creates during any evaluation.
6. Click Import Variables. HyperStudy imports nine input variables from the
arm_model.tpl resource file.

2. Click Apply.
In this step you will create three output responses: Max_Disp, Max_Stress, and Volume.
1. Create the Max_Disp output response.

a. From the Directory, drag-and-drop the crank.h3d file, located in

• Set Subcase to Subcase 1 (SUBCASE1).
• For Request (Start - End), enter N27099 - N40946.
• For Components, select MAG.
e. Label the output response Max_Disp.

h. Click Finish. The Max_Disp output response is added to the work area.
a. From the Directory, drag-and-drop the crank.h3d file, located in
c. Select Multiple items at one time step (readsim), then click Next.
• Set Subcase to Subcase 1 (SUBCASE1).
• Set Type to Element Stresses (3D).
• For Request (Start-End), enter E38257 - E94809.
• For Components, select vonMises (2D & 3D).
g. Click Finish. The Max_Stress output response is added to the work area.
a. From the Directory, drag-and-drop the crank.out file, located in


HyperWorks® (osmass.tpl) and click Next.
• Set Type to OptiStruct Analysis.
• Set Request to Out File.
• Set Component to Volume.
5. To minimize the size of the result files in the study directory, go to the Define Output
Responses step, Data Sources tab, and clear the Retain checkbox for Max_Disp (Data
Source 1) and Max_Stress (Data Source 2) data sources.
6. Click Evaluate to extract the output response values of each expression.
Parameter Screening DOE Setup
In this section, you will study the results from different parameter screening DOEs. For each
screening, only the results obtained on the Max_stress output response will be reviewed using
Pareto plot, Interactions, and Ordination displays.
As you proceed, keep the following points in mind:
• A Full Factorial provides excellent resolution but typically requires an exceedingly large
number or runs.
• A Fractional Factorial with a Resolution III confounds main effects with interactions;
results could be inaccurate if interactions are significant.
• A Fractional Factorial with a Resolution IV main effects are not confounded, but
interactions are confounded with each other.
• A Fractional Factorial with a Resolution V main effects and interactions are not
confounded.
Step 4: Run a Full Factorial DOE Study
In this Full Factorial DOE, you will set the shape variables to two levels. Since Full Factorial
runs all combinations of input variable values, you will have no loss of information; however
you will have an expensive DOE. You will use this study as a reference to compare the loss

of accuracy in the conclusions to other DOEs that are less expensive.

5. Click the Levels tab, and verify that the number of levels for each input variable is 2.
Using nine shape variables with two levels gives a full factorial plan made of 512 runs
(2^9).
6. Click Apply.
8. Above the Run Tasks tab, click Multi-Execution and select the Multi-Execution
checkbox. In the field below, verify that 4 Jobs is selected.


12. Click the Pareto Plot tab to view Pareto plots obtained from the Full Factorial DOE (512
runs).
A Pareto Plot shows the ranked influence (highest to lowest) of the design variables on the
response.
You can see that the 5 lengths have the largest influence, and the 3 radii have the least
influence on Max_stress. The slope of the hashed lines (positive or negative) indicates a
positive or a negative effect of a variable on the response. length_4, length_2, length_5,
and length_3 have negative slopes, which indicates that if these variables increase,
Max_stress will decrease. On the other hand, length_1 has a positive slope, which
indicates that increasing length_1 increases Max_stress.
13. Click the Interactions tab to plot the interactions between input variables.
An interaction is the failure of one variable to produce the same effect on a response at
different levels of another variable. An interaction can be either positive or negative.
The Interaction plot displays parallel lines if there is no interaction. The bigger the
interaction is, the less parallel the lines will be.
Looking at the Max_stress output response, you can see that several of the interactions
are very small. You can still find instances with a true interaction, such as the interaction
of length_5 and length_4. The effect of the variable length_4 on the output response
Max_Stress is in the same direction regardless of the value of length_5, but the effect
(magnitude) is much more important when length_5 is larger.
Interaction between length_5 and length_4 on No interaction between length_1 and length_5 on
Max_Stress Max_Stress

14. Click the Ordination tab to see the Principal Component Analysis results.
Each input variable and output response in the biplot is represented by a line. Lines that
point in the same direction have strong correlations (positive or negative depending on
whether the lines point in the same or opposite directions). You can see that lengths 2, 3,
4, and 5 contribute the most to the volume value, and improve the structural performance
the most along with height and length_1.
Step 5: Run a Fractional Factorial DOE Study with Resolution III
In this step you will set up a Fractional Factorial DOE, with resolution III and the same levels
as the previous Full Factorial DOE. Use the nine input variables with two levels, without
turning on any interaction, leads to a 12 run Fractional Factorial plan.
3. In the Specifications step, set the Mode to Fractional Factorial.
4. In the Settings tab, set Resolution to III.

6. Click Apply.
10. Click the Pareto Plot tab to view Pareto plots on Max_stress obtained from the Fractional
Factorial III DOE (12 runs).
Compared to the Pareto plot from the Full Factorial, several differences can be observed.
• The rank of variables is not the same; for instance, length_4 is the most influential
input variable in the Full Factorial, whereas length_2 is the most influential input
variable in the Fractional Factorial III.
• The effect of length_5 has decreased, while the effects of the three radii has
significantly increased, especially the effects of radius_3 and radius_1.
Fractional Factorial III Full Factorial
11. Click the Interactions tab to plot the interactions between input variables.
Compare the interaction between length_1 and length_3 on Max_stress. Note that the
results provided from the Full Factorial and Fractional Factorial III are quite different. The
Full Factorial establishes the real interactions, while the Fractional factorial III detects
confounded interactions inaccurately.

Full Factorial Fractional Factorial III
12. Click the Ordination tab to see the Principal Component Analysis results obtained with
the Fractional factorial III.
In the Full Factorial, you can see that length_2, length_3, and length_5 are correlated to
the Volume output response. In the Fractional Factorial III, the relationship is not as clear.
In the Fractional Factorial III, the lines representing length_2, length_3, length_5, and
Volume are not as collinear as they are in the Full Factorial.
Full Factorial

Fractional Factorial III
Step 6: Run a Fractional Factorial DOE Study with Resolution IV
In this step you will set up a Fractional Factorial DOE with Resolution IV. Using the nine input
variables with two levels, without turning on any interaction, leads to a 24 run Fractional
Factorial plan.
5. In the Settings tab, set Resolution to IV.
6. Click Apply.
Factorial IV DOE (24 runs).
Compared to the results obtained from the Full Factorial, you can see that the places of
length_2 and length_4 have been inversed in the ranking, but the rest of the information
is quite similar.

Fractional Factorial IV Full Factorial
Step 7: Run a Fractional Factorial DOE Study with Resolution V
In this step you will set up a Fractional Factorial DOE with Resolution V. Using the nine input
variables with two levels, without turning on any interaction, leads to a 128 run Fractional
Factorial plan.
5. In the Settings tab, set Resolution to V.
6. Click Apply.
Factorial V DOE (128 runs).
Notice that the results obtained from both DOE methods are the same, but the number of
runs for the Fractional Factorial V is only 1/4th of the runs for the Full Factorial.
Fractional Factorial V Full Factorial
11. Click the Ordination tab to see the Principal Component Analysis results obtained from
the Fractional factorial V.

Notice that there are no differences between the results obtained from the Full Factorial
and the Fractional Factorial V.
Full Factorial
Fractional Factorial V

Comparison
In this tutorial, you focused on the Max_Stress output response, but the same steps should be
applied for all output responses. Conclusions on the screened output responses are
summarized for the four DOE methods in the table below. Results for the Full Factorial and
Fractional Factorial V are the same; however, the Fractional Factorial V required only 1/4th the
number of runs as the Full Factorial. You can also observe that as the resolution decreased,
the accuracy of the information gathered also decreased. For example, the conclusions from
the Fractional Factorial III are the least accurate.
Fractional Fractional Fractional

Full Factorial
Factorial V Factorial IV Factorial III
Number of
512 128 24 12
runs
Remove all Remove radii

Remove all radii Remove radius
Max_Stress radii and 1, 2 and height
and height 2 and height
height
Remove all Remove all Remove all

Max_Disp Remove all radii
radii radii radii
Remove all Remove all Remove all

Volume Remove all radii
radii radii radii
Remove all Remove radii 1 Remove radius

Conclusion Remove all radii
radii and 2 2
Last modified: v2017.2 (12.1156684)

HS-2005: DOE Study of a Cantilever Beam Using Templex
This exercise runs a simple DOE study on a cantilever beam to examine the influence of the
length, width, and height of the beam on its volume, maximum bending stress, and deflection.
We would like to know all of the main effects and all of the interactions between the beam's
dimensions, so full factorial is used for this DOE study. The base input Templex template,
beam.tpl, defines three input variables (L, B, and H) for the beam dimensions labeled as
length, width, and height, respectively (see figure below).
The sample base input template beam.tpl used in this tutorial can be in <hst.zip>/HS-
2005/ and copied to your working directory.
Cantilever Beam
{parameter(L,"Length",50,50,100)}
{parameter(B,"Width",2.5,2.5,5)}
{parameter(H,"Height",5,5,10)}
{P = 100}
{E = 2e11}
{M = P*L}
{c = H/2}
{I = (B*H^3)/12}
{MAX_STRESS = (M*c)/I}
{MAX_DISP = (P*L^3)/(3*E*I)}
{VOLUME = B*H*L}
{MAX_STRESS}
{MAX_DISP}
{VOLUME}
In this exercise, you will learn how to:

• Setup a study
• Add a model
• Add output responses
• Set up a DOE approach
• Perform DOE Runs

• Extract output responses for a DOE study

• Perform post processing and evaluate the results
• Observe the interactions between variables
• Add approximations for a regression equation
This setup requires you to add a study and load the input file into HyperStudy. The input
variables for the DOE study (length, width, and height) are selected as factors. A nominal run
is performed (with Templex as the solver), and the output responses for the DOE study are
selected (in this case: volume, max. stress, and max. displacement).
click OK.
b. In the Solver input file column, enter beam.tpl. This is the name of the solver input
c. In the Solver execution script column, select Templex (templex).
6. Click Import Variables. Three input variables are imported from the beam.tpl
resource file.


2. Click Apply.
directory.
In this step you will create three output responses: Max_Stress, Max_Disp, and Volume.
a. From the Directory, drag-and-drop the beam.tpl file, located in


Note: Because there is only a single value in this data source, HyperStudy
inserts [0] after m_1_ds_1, thereby choosing the first (and only) entry in the
data source.
g. Click Finish. The Max_Stress output response is added to the work area.
2. Repeat step 1 to create the Max_Disp and Volume output responses.
3. In the Expression field:
a. For Max_Disp, after m_1_r_2, change [0] to [1].
b. For Volume, after m_1_r_3, change [0] to [2].
4. Click Evaluate Expressions to extract the output response values of each expression.
The following steps illustrate how to set up a DOE approach through the selection of a DOE
matrix and input variables. Input variables can either be controlled or uncontrolled. In this
exercise, a full factorial DOE matrix is considered. This provides the main effect of the input
variables and their interactions with one another for selected output responses. The length,

width, and depth of the input variables are introduced as controlled and their effects are
studied for the following output responses: maximum stress, maximum displacement, and
volume.
5. Click Apply.
8. Monitor the progress of the evaluation by clicking the Evaluation Data and Evaluation
Plot tabs.
Step 5: Perform Post Processing and Evaluate the Results
1. Click the Linear Effects tab to review the main effect of each controlled parameter on a
selected output response.
Review the effects of a single input variable on an output response, or multiple input
variables on an output response.
As illustrated in the plot above, as the length increases from 50 to 100 units, the max
stress increases from 225 to 450. Increasing the width from 2.5 to 5 units decreases the
max stress by the same amount. The contributions for max stress from both length and
width are identical. Increasing the height from 5 to 10 units decreases the max

stress by a larger extent (from 530 to 130). It is clear that for the max stress output
response, height has a larger main effect than length or width.
Input Variable Max_Stress - Lower Max_Stress - Upper Bound

Bound
Length 225 450
Width 450 225
Height 540 135
2. Click the Distribution tab to review a histogram of Max Stress.
Step 6: Add Approximations for a Regression Equation
In the following steps, you will create approximations for the DOE study. A first order
regression equation for each output response (Max Stress, Max disp, and Volume) is
created based on the DOE study. Since the input variables are of two levels, only the first
order equation can be generated.
6. In the Matrix Source column, select DOE 1 (doe_1).
7. Click Import Matrix. The matrix from the DOE is imported into the Fit approach.

11. Click Apply.
15. Click the Residuals tab.
The Residuals table displays the difference between the output response value from the
solver and the output response value calculated from the regression equation. This table
can also be used to determine which runs are generating more errors in the regression
model. The % error column shows that the predicted output response is not close to the
observed value. The % error can be reduced if more levels are chosen for example, using
a three-level design in this exercise.
16. Click the Trade-Off tab to plot output responses vs. input variables.
a. In the Inputs pane, select the X Axis and Y Axis checkboxes for Length.
b. Using the Channel selector, selector Max_Stress_LSR.
c. Review the plot in the Outputs pane.

HS-2200: Using Existing Design Data from an Excel

Spreadsheet
This tutorial demonstrates how to perform an optimization study for an application where only
design data in a Microsoft Excel spreadsheet is available (i.e.: no simulation model exists).
The Excel spreadsheet study.xls used in this tutorial can be found in
<hst.zip>/HS-2200/ and copied to your working directory.
The objective of this tutorial is to create a fit (approximation) using the designs in the
spreadsheet and then to perform an optimization study using the fit.
The spreadsheet used here contains five columns. The first column contains the numbering of
the designs, the second and third columns contain the values of the two input variables for
each design, and the fourth and the fifth columns contain the results of a DOE study
previously run. Sixteen designs have been evaluated.
click OK.
a. Click Add Model.
b. In the Add - HyperStudy dialog, add one Internal Math model.
7. Click Add Input Variable.
8. In the Add - HyperStudy dialog, add two input variables.
9. Optional. Copy the input variable labels from the study.xls spreadsheet, and paste them
into the Labels column of the work area.
Note: When you paste the input variable labels into the work area, select Paste
transpose from the context menu.


2. Click Apply.

2. In the Add - HyperStudy dialog, add two output responses.
3. Optional. Copy the output response labels from the study.xls spreadsheet, and paste
them into the Label column in the work area.
Note: When you paste the input variable labels into the work area, select Paste
transpose from the context menu.

Step 4: Import the DOE Results from the Spreadsheet
4. In the work area, set the Mode to None.
5. Click Apply.
6. In the top, right of the work area, select Edit > Run Matrix.
7. In the Run Matrix dialog, click Add Run to add 16 runs to the matrix, as there are 16
runs in the study.xls spreadsheet.
8. Open the study.xls spreadsheet in Excel.
9. Copy all of the input variable and output response data for each run in the spreadsheet.
10. In the Run Matrix dialog, highlight all of the runs in the matrix.
11. Right-click on the highlighted runs, and select Paste from the context menu. HyperStudy
pastes the input variable and output response data that you copied from the study.xls
spreadsheet into the run matrix.

13. Click Apply.

14. Click OK.
Step 5: Create a Fit (Approximation)
7. In the work area, set the Mode to Moving Least Squares (MLSM).
8. In the Settings tab, change the Order to 3 + interactions.
9. Click Apply.
13. Click the Residuals tab to investigate the accuracy of your approximation. From the table
you can see that this approximation is not as good.


HS-2201: Using Existing Design Data from an Excel

Spreadsheet with a Lookup Model
This tutorial demonstrates how to perform an optimization study for an application where only
design data in a .csv file is available (that is no simulation model exists).
The objective of this tutorial is to create a fit (approximation) using the designs in the
study.csv file and then to perform an optimization study using the fit.
The first and second columns contain the values of the two input variables for each design,
and the third and fourth columns contain the results of a DOE study previously run. Sixteen
designs have been evaluated.
Model Files
The study.csv file used in this tutorial can be found in <hst.zip>/HS-2201/. Copy these files
to your working directory.
Exercise

and click OK.
5. Add a Lookup model.
a. Click Add Model.
b. In the Add - HyperStudy dialog, select Lookup and click OK.
c. In the work area, Resource column, click .
d. In the HyperStudy – Load model resource dialog, navigate to your working
directory and open the study.csv file.

8. In the Import Variables dialog, enter 2 in the Number of design variables field and
click OK.
9. Go to the Define Input Variables step and review the two input variables that were
imported from the study.csv file.

2. Click Apply.

Step 3: Review the Output Responses

1. Review the two output responses that were imported from the study.csv file.
Step 4: Import the DOE Results from the .csv File

4. In the work area, set the Mode to Run Matrix.
5. From the Settings tab, Matrix File field, click .
6. In the Open dialog, navigate to your working directory and open the study.csv file.
7. Click Apply.
Step 5: Create a Fit (Approximation)

3. Add matrix.
a. Click Add Matrix.
b. In the Add - HyperStudy dialog, add one matrix.

7. Verify that the Fit Type assigned to each output response is FAST – Fit Automatically
Selected by Training.
Note: By default, FAST automatically selects the best Fit type from all available
Fits. You can manually select the Fit types FAST can choose by highlighting
one or more responses in the work area and selecting Fits from the Settings
tab.
8. Click Apply.
11. Click the Diagnostics tab.
Resp1 has the best Fit using LSR with custom terms; the Regression Terms tab shows
that only linear terms are required. Resp2 has the best results using a MLSM. For
Resp2, the R-square value that is based on the Input Matrix shows that the model
accuracy is very good. The moderate value for the cross validation indicates that the
model may benefit from more data because its accuracy is quite dependent on the
complete set of point. Removing any points from the input set can significantly alter
the Fit’s predictions.

HS-2210: Principle Component Analysis of a Cantilever

Ibeam
and click OK.
a. From the Directory, drag-and-drop the ibeam.tpl file into the work area.
b. In the Solver input file column, enter ibeam.py. This is the name of the solver
input file Hype any evaluation.
c. In the Solver execution script column, select Python (py).
6. Click Import Variables. Four input variables are imported from the ibeam.tpl
resource file.


2. Click Apply.
directory. The approaches/nom_1/run 00001/m_1 sub-directory contains the
output.hstp file, which is the result of the nominal run, and will be used during the
Optimization.
1. Create the Iy output response for the y-axis moment of inertia.

a. From the Directory, drag-and-drop the output.hstp file, located in
HyperWorks® (HstReaderPdd) and click Next.
• Set Request to Iy.

e. Label the output response Iy.


h. Click Finish. The Iy output response is added to the work area.

2. Create four more output responses by repeating step 1, except change the component
assigned to each output response to the following.
Output Response Component
Volume Vol
IZ Iz
Displacement d
Frequency1 Freq

Step 4: Run a D-Optimal DOE Study
4. In the work area, set the Mode to D-Optimal.
5. In the Settings tab, change the Number of runs to 13, which is 2 more runs than the
minimum re
6. Click Apply.
9. Go to the Post-Processing step, and click the Ordination tab.
The biplot is interpreted by looking at the relationship between the lines that each
represent one inp responses show a strong negative correlation because they are
aligned, but pointing in opposite dire the input variable web thickness, indicated by the
orthogonality. Similar strong positive correlations and flange length.



HS-2215: Extensible DOE Study of a Space Frame Structure

with Input Variable Constraints
In this tutorial you will use a DOE to investigate the effects of the cross sectional
dimensions and joint stiffness of a truss structure’s volume and natural frequencies. The
tubular truss dimensions must be constrained, such that the inner radius is always less than
the outer radius. You will also use the extensible feature of the Modified Extensible Lattice
Sequence in a progressive set of steps to add additional runs to a DOE.
click OK.
a. From the Directory, drag-and-drop the tinker_toy.tpl file into the work area.
b. In the Solver input file column, enter tinker_toy.fem. This is the name of the
d. In the Solver input arguments column, after ${file}, enter -core in.
This option forces OptiStruct to run with maximum memory, which will make the
analysis run more quickly. The small size of the finite element model makes this
possible in this example.
6. Click Import Variables. Three input variables are imported from the tinker_toy.tpl
resource file.


9. Click the Constraints tab.
10. Add an input variable constraint.
a. Click Add Constraint.
b. In the Add - HyperStudy dialog, add one constraint.
c. Define the constraint.
Note: Use the Expression Builder to select input variables to append to the Left
Expression and Right Expression fields.
• For Left Expression, enter m_1_outer_diam.
• Set Comparison to >=.
• For Right Expression, enter m_1_inner_diam.

2. Click Apply.
directory.

a. From the Directory, drag-and-drop the tinker_toy.out file, located in

• Set Type to Volume.

• Set Request to Volume.
f. Label the output response Volume.


i. Click Finish. The Volume output response is added to the work area.
2. Create the Frequency_Vector data source, which will be used in the frequency output
responses.
a. From the Directory, drag-and-drop the tinker_toy.out file, located in
• Set Type to Frequency.
• Set Request (Start - End) to Mode 1 - Mode 3.
• For Components, select Value.

e. Clear the Linked to a new Response checkbox.

f. Label the data source Frequency_Vector.
g. Click Finish.

3. Add three output responses.

a. Click Add Output Response three times.
b. In the work area, label the output responses Frequency_1, Frequency_2, and
Frequency_3.
4. Define the Frequency_1 output responses.
a. In the Expression column of the output response Frequency_1, click .
c. In the work area, select Frequency_Vector.
d. From the Insert Varname drop-down menu, click First Element.
e. Click Insert Varname. The expression m_1_ds_2[0] appears in the Evaluate
Expression field.
f. Click OK.
5. Repeat step 4 to define Frequency_2 and Frequency_3, except change the value after
m_1_ds_2.
a. For Frequency_2, change [0] to [1].
b. For Frequency_3, change [0] to [2].
Step 4: Run a Modified Extensible Lattice Sequence DOE Study, with

4 Runs

6. In the Settings tab, change the Number of runs to 4, which is the minimum number of
runs for a multivariate effects calculation.
7. Click Apply.
10. Go to the Post-Processing step, and click the Scatter tab. Using the Channel
selector, set the X Axis to Outer_diameter and the Y Axis to Inner_diameter. All four
runs satisfy the constraint, which is inner_radius < outer_radius.
11. Click the Pareto Plot tab, and note which input variables contribute to which output
responses.
Above the Channel selector, click and verify Multivariate effects is selected.

Step 5: Extend DOE with 7 Additional Runs
In this step you will run a second Modified Extensible Lattice Sequence DOE study with 7 new
runs, and include the 4 runs from DOE1. This DOE will have a total of 11 runs, which is the
default suggested number of runs for a MELS DOE with three input variables.
which is the minimum suggested number of runs for three input variables.
5. In the Settings tab:
a. Change the Number of runs to 7.
b. Select the Use Inclusion Matrix checkbox.
6. Import run data from the DOE 1 using an Inclusion Matrix.
a. Click Edit Matrix > Inclusion Matrix from the top, right corner of the work area.

c. In the Edit Inclusion Matrix dialog, click Import Values.

d. In the Import Values dialog, select Approach Evaluation Data and click Next.
e. Set approach to DOE 1.
f. Click Next.
g. Click Finish.
h. Review the imported run data and click OK.
7. Click Apply.
10. Go to the Post-Processing step, and click the Scatter tab. Using the Channel
selector, set the X Axis to Outer_diameter and the Y Axis to Inner_diameter. Note
that all 11 runs still satisfy the constraint, which is inner_radius < outer_radius.
11. Click the Pareto Plot tab, and compare the results to the Pareto Plot from DOE 1. Note
that the magnitude and order of importance has changed in some cases.

Pareto Plot from DOE2 Pareto Plot from DOE1
Step 6: Extend DOE with 4 Additional Runs
In this step you will run a third Modified Extensible Lattice Sequence DOE study with 4 new
runs, and include the 11 runs from DOE 2. This DOE will have a total of 15 runs, which
exceeds the number of suggested runs.
5. In the Settings tab:
a. Change the Number of runs to 4.
b. Select the Use Inclusion Matrix checkbox.
6. Import run data from the DOE 2 using an Inclusion Matrix.
b. In the Edit Inclusion Matrix dialog, click Import Values.

c. In the Import Values dialog, select Approach Evaluation Data and click Next.

e. Set the approach to DOE 2.

f. Click Next.
g. Click Finish.
h. Review the imported run data and click OK.
7. Click Apply.
10. Go to the Post-Processing step, and click the Pareto Plot tab. Compare the results to
the Pareto Plots from DOE 2.
Note that the results are qualitatively the same, indicating that you will likely have enough
runs to draw solid conclusions.
Pareto Plot from DOE2

Pareto Plot from DOE3

Fit

HS-3000: Fit Method Comparison: Approximation on the Arm

Model
In this tutorial, you will create approximations for the output responses of the arm example
introduced in tutorial HS-2000: DOE Method Comparison: Arm Model Study, and review the
differences between different Fit methods.
Before running this tutorial, complete tutorial HS-2000: DOE Method Comparison: Arm Model
Study. You can also import the archive file HS-2000.hstx, available in
<hst.zip>/HS-3000/.
In HS-2000, you learned that instead of using the nine input variables, you could continue
additional studies just as effectively with six shapes since the others did not have a great
influence on the output responses. This will save computational effort.
In this tutorial, you will use the six shapes variables.
Length2: Lower Bound = 0.0, Initial Bound = 0.0, Upper Bound = 2.0
Height: Lower Bound = -1.0, Initial Bound = 0.0, Upper Bound = 1.0
You will begin this tutorial by creating a Modified Extensible Lattice Sequence (MELS) DOE.
MELS is a space filling DOE designed to equally spread out points in a space by minimizing
clumps and empty spaces. The minimal required number of points to create a second order
polynomial with N variables is 1.1*(N + 1)*(N + 2)/2. Using this matrix, you will then create
the following Fits for both output responses: Least Square Regression (LSR), Moving Least
Square (MLSM), HyperKriging (HK), and Radial Basis Function (RBF).
Step 1: Run MELS DOE Study
In order to create the approximations to be used as surrogate models, you must perform
specific DOEs that will serve as the input matrix. You will need to run a DOE suitable to be
used in response surface creation, such as MELS.
4. In the work area, Active column, clear the radius_1, radius_2 and radius_3
checkboxes.


6. In the work area, set the Mode to Modified Extensible Lattice Sequence (MELS).
7. In the Settings tab, verify that the Number of runs is set to 31.
8. Click Apply.
12. Click the Scatter tab to review a 2D scatter plot of the results from the MELS DOE.
The image below illustrates a typical sampling of the MELS DOE with 31 runs (length_1
vs. length_2).
Note: This visualization is a projection of 31 points distributed in 6 dimensions onto a 2
dimensional plane.
13. Optional. Create a second DOE with less number of runs to be used as a Validation
matrix in the Fit approach.

A Validation matrix provides information on the Fit’s prediction accuracy.

Note: You should not use MELS as a Validation matrix, as it will take the same first runs
from the MELS Input matrix due to its extensibility.
In this tutorial, you will use the Hammersley method to create the Validation matrix.
a. In the Explorer, right-click on the MELS DOE and select Copy from the context
menu.
b. In the Specifications step, set the Mode to Hammersley.
c. In the Settings tab, change the Number of runs to 12.
d. Click Apply.
e. In the Evaluate step, click Evaluate Tasks.
Step 2: Setting Up a Fit Approach
Using the 31 runs from the MELS DOE as an Input matrix and the 12 runs from the
Hammersley DOE as a Validation matrix, create the following fits: Least Square Regression
(LSR), Moving Least Square Method (MLSM), HyperKriging (HK), and Radial Basis Function
(RBF).
4. Click Add Matrix two times.
5. Define FitMatrix 1 and FitMatrix 2, by selecting the options indicated in the image below
from the Type and Matrix Source columns.

8. In the work area, set the Mode to the appropriate Fit method.
9. For the Least Sqaure Regressions (LSR) Fit, in the Settings tab, set Regression Model
to Interaction.
An Interaction regression model enables linear and cross terms to be considered in the
function f(x,y)=A+Bx+Cy+Dxy; where the first three terms are linear, and the last term is
a cross term between the variables.

10. Click Apply.

14. Click the Scatter tab to compare the original Max_Stress output response to the Fit
Max_Stress.
The scatter shows the Fit accuracy. The closer together the points are along the diagonal,
the better the fit. In the Max_Stress vs Max_Stress_LSR plot, you can see some dispersed
points, which indicates the Fit has some inaccuracy. In comparison, the points in the
Max_Stress vs Max_Stress_MLSM plot follow the diagonal more closely, which indicates it
provides better Fit accuracy on Max_Stress.
You will not compare HyperKriging and Radial Basis Function using scatter plots, because
the results will be misleading. HyperKriging and Radial Basis Function go through the
exact points by default, therefore the scatter plot comparing the original output response
vs. the Fit output response will produce a straight line. However, this does not necessarily
mean that the Fit has good predictive capability.
Max_stress and Max_stress_LSR comparison Max_stress and Max_stress_MLSM comparison
15. Click the Diagnostics tab to review the diagnostics of the Fit study.
The R-Square value measures how much of the variability of the response data around its
mean is captured. If the model perfectly predicts the known values, R-Square will have a
maximum possible value of 1.0.

Diagnostics for Max_Stress, LSR Diagnostics for Max_Stress, MLSM
The R-square value for an Input Matrix in HyperKriging and Radial Basis Function has no
meaning because the runs will always go through the exact data points, which will result
in a value of 1.0. Although the value is 1.0, this does not mean the Fit will be accurate. In
HyperKriging and Radial Basis Function, the only meaningful diagnostic values are for
Cross-validation Matrix and Validation Matrix.
Diagnostics for Max_Stress, HK Diagnostics for Max_Stress, RBF
16. Click the Residuals tab to review the Error (and Percent Error) between the original
output response and the Fit output response for each run of the Input and Validation
matrices.
Input Matrix Residuals on Max_Stress, LSR Validation Matrix Residuals on Max_Stress, LSR
The Input Matrix Residual errors are slightly smaller with Least Square Regression, than
they are with Moving Least Square Method, but the Validation Matrix Residual errors are
much smaller with Moving Least Square Method.

Input Matrix Residuals on Max_Stress, MLSM
Validation Matrix Residuals on Max_Stress, MLSM
The Input Matrix Residuals are meaningless for HyperKriging and Radial Basis Function, as
indicated in the Validation Matrix Residuals below.
Validation Matrix Residuals on Max_Stress, HK Validation Matrix Residuals on Max_Stress, RBF
Comparison
The max percent of errors for Input and Validation matrices are as shown below:
LSR (Interaction MLSM HK RBF

Regression model)
Max_Disp -1.33% -2.56% - -
Max_Stress -6.88% -9.73% - -

Input Matrix Residuals

LSR (Interaction MLSM HK RBF

Regression model)
Max_Disp 6.23% -3.33% 9.48% -2.80%
Max_Stress 31.74% -18.00% 49.37% 17.60%

Validation Matrix Residuals
It can be seen that the percent of errors for Max_Disp are smaller than Max_Stress. These
results indicate the Fit approach works well for Max_Disp, but is not very efficient for
Max_Stress.
These finding suggest that it is best to use the Fit model obtained from the MLSM for
Max_Disp. An output response such as Max_Stress is a global envelope of localized effects.
The nature of such an envelope type of output responses makes them difficult to capture
accurately with a Fit. In contrast, the Max_Disp output response is not influenced by localized
effects, therefore it is easier to use a Fit for such data. When proceeding in this situation, it is
recommended that you either increase the number of samples, which is not guaranteed to
improve the accuracy, or create a series of more localized output responses that would be
simpler functions of the input variables; for example, several output responses that each
capture the stress in specific regions. The image below highlights the areas of high stresses
from the runs in the Input matrix.
Last modified: v2017.2 (12.1156684)

HS-3005: Exporting Fit Models to Excel
In this tutorials, you will learn how to:

• Run a Design of Experiments (DOE)
• Build a Fit to approximate the output responses
• Export the Fit model to an Excel report
• Use Excel to predict output response values
Before starting this tutorial, you must add the HstAddinFit add-in to Excel. For instructions on
to install the HstAddinFit add-in, refer to Fit Excel Plug-In.
and click
OK.
a. Click Add Model.
b. In the Add - HyperStudy dialog, add one Internal Math model.
7. Click Add Input Variable.
8. In the Add - HyperStudy dialog, add two input variables.
9. Change the input variable's Lower Bounds, Initial, and Upper Bounds to the values

2. Click Apply.


directory.

2. In the Add - HyperStudy dialog, add one output response.
3. In the Expression column, enter sin(dv_1)*cos(dv_2).
4. Click Evaluate.
Step 4: Run a Hammersley DOE Study
5. In the Settings tab, Number of runs field, enter 50.
Note: The large number of runs relative to the number of input variables is chosen to
capture the highly non-linear nature of the output response function. This
model is simple to evaluate, therefore the computational cost of the evaluation
is not an important consideration in this example.
6. Click Apply.
Step 5: Run a Radial Basis Function Fit


7. Define the matrix.
a. Set Type to Input.
b. Set Matrix Source to Doe1 (doe_1).

10. In the work area, set the Mode to Radial Basis Function.
11. Click Apply.
13. Click Evaluate Tasks to evaluate the designs.
15. Click the Trade-Off tab to visualize the response surface as a function of two input
variables.
a. In the Inputs pane, select the X Axis checkbox for DV 1 and the Y Axis checkbox
for
DV 2.
b. In the Outputs pane, click and adjust the plotting resolution of the display to
include 25 samples.

d. Visually examine the plotted response surface to inspect the quality of the
approximation to the original sinusoidal function.
16. In the Trade-Off tab, interactively predict output response values as a function of the
input variables.
a. In the Inputs pane, clear the X Axis and Y Axis checkboxes.
b. In the Inputs pane, modify the values of each input variable by moving the slider
in the first Value column, or by entering values in the second Value column. The
predicted output response value in the Value column of the Outputs table is
adjusted.
Note: The shaded spark lines in the Value cell indicate the relative value of the
predicted

output response with respect to the minimum and maximum of the sample.
The marker at the bottom of the cell references the value of the predicted
output response at the nominal values of the input variables.
Step 6: Export an Excel Report for the Fit
1. Go to the Report step.

2. Select the HyperStudy Spreadsheet checkbox.
3. Click Create Report. An Excel report is generated and opened in Excel.
4. In the Excel report, click the Trade-Off tab.
Note: The structure and functionality of this tab is a reflection of the corresponding
Trade-Off ta values on the right-hand side, and the predicted output response
values are updated and d

5. To verify that the same values occur in the output response prediction columns for the
same set of HyperStudy and the Excel report.
See Also:;
HyperStudy Tutorials


HS-3010: Fuselage Sizing Trade-Off using Categorical

Variables
The purpose of this tutorial is to investigate the relative effect of the variable on the identified
output responses. Furthermore, this tutorial will demonstrate how to create a Fit in order to
investigate combinations of variables that were not explicitly simulated.
Three continuous variables and three categorical variables are used in this tutorial. The frames
can take five possible sections, and the stringers can each take from four available sections.
Continuous variables
• Thickness of floor
• Thickness of floor beams
• Skin thickness
Category variables
• Cross sections of the frames
• Stringers above the floor
• Stringers below the floor
This tutorial uses three load cases:

• Free-free normal modes case
• Simple bending case
• Simple torsional case

click OK.
a. From the Directory, drag-and-drop the fuselage.tpl file into the work area.
b. In the Solver input file column, enter fuselage.fem. This is the name of the solver
6. Click Import Variables. Six input variables are imported from the fuselage.tpl
resource file.
8. Review the input variable's lower and upper bounds ranges.

2. Click Apply.



a. From the Directory, drag-and-drop the fuselage.out file, located in
HyperWorks® (hgosfreq.exe) and click Next.

h. Click Finish. The Mass output response is added to the work area.
2. Create two more output responses by repeating step 1, except change the type, request,
and component assigned to each output response to the following.
Because this is a free-free analysis, Freq1 will be the seventh frequency in the list due to
the six rigid body modes (all near zero). Freq2 will be the eighth frequency in the list.
Output Type Request Component

Response
Freq1 Frequency Mode 7 Value
Freq2 Frequency Mode 8 Value
3. Create the Bending displacement output response, which will have a magnitude of node
8196 (loading point).
a. From the Directory, drag-and-drop the fuselage.h3d file, located in


e. Define the following options, and then click Next.
• Set Subcase to Subcase 2 (bending).
• Set Type to Displacement(Grids).
• For Request, apply a filter of 8196. Press Enter to accept the value entered in
the Filter field.
• Set Component to Mag.
f. Label the output response Bending displacement.


i. Click Finish. The Bending displacement output response is added to the work area.
4. Create the Torsional rotation output response, which will have a z-direction of node 8196
(loading point).
a. From the Directory, drag-and-drop the fuselage.h3d file, located in
• Set Subcase to Subcase 3 (torsion).
• Set Type to Rotation (Grids).
• For Request, apply a filter of 8196. Press Enter to accept the value entered in
the Filter field.
• Set Component to Z.
e. Label the response Torsional rotation.
g. Click Finish. The Torsional rotation output response is added to the work area.

i. In the Expression field for Torsional rotation, edit the expression to be

max(m_1_ds_5)*360/3.14.
This expression converts the rotation from radians to degrees.
Step 4: Run a Doe
4. In the work area, set the Mode to None.
5. Click Apply.
6. Import run data using the Run Matrix.
a. Click Edit Matrix > Run Matrix from the top, right corner of the work area.
b. In the Edit Data Summary dialog, remove any existing run data.
c. Click Import Values.
d. In the Import Values dialog, select Plain Text and click Next.
e. In the Source File field, navigate to the doe.csv file and click Next.
f. Click Finish.
g. Review the imported run data and click Apply.
h. Click OK.


8. Click Evaluate Tasks to run the Doe.
9. Go to the Post-Processing step, and click the Pareto Plot tab. Enable multi-plot and
select all of the output responses in the Channel selector. If not all of the input variables
are plotted, you may need to alter the number of displayed input variables from the
options menu.
The relative effect of a input variable can vary from output response to output response.
The most influential input variables when analyzing frequency output responses are Frame
Section and Stringer Section Upper. In contrast, the most influential input variables when
analyzing the two stiffness conditions are Skin thickness and Stringer Section Upper.
Some input variables can have no effect on output responses. Floor beam thickness has
minimal effect on any of the output responses, which indicates that you may want to
consider removing this input variable from the analysis.
In a Pareto plot, the effect of input variables on output responses does not measure
sensitivity but rather absolute change. Floor thickness has a major effect on Volume. This
effect is not a derivative, but a measure of the possible increase over the range of the
input variables (the range is the difference between the upper and lower bounds). The
floor has a large area and the thickness has very large bounds (+/-0.1 inches), therefore
it can make a dramatic impact on Volume as the input variables move through the
available space.

Step 5: Run a Least Sqaures Regression Fit
6. Define the matrix.
a. Set Type to Input.
b. Set Matrix Source to Doe1 (doe_1).

9. In the work area, set the Mode to Least Squares Regression.
10. Click Apply.
12. Click Evaluate Tasks to evaluate the designs.


15. Click the Diagnostics tab to assess the accuracy of the Fit. Select the Mass output
response in the Channel selector.
The R-Square and R-Squared Adjusted values for Mass are 1.00, which indicates the
model perfectly predicted the known values.
16. Review the diagnostics for the remaining output responses. Notice the R-Squared values
are still high, which indicates a high quality fitting of the data.
17. Click the ANOVA tab and review the Mean Squares Percent column to see the relative
importance of input variables. The results should be similar to the results noted in the
Pareto Chart tab of the Doe.
18. Click the Trade-Off tab to perform "what if" scenarios. In the Inputs pane, modify the
values of input variables to see their effect on the output response approximations in the
Output pane.
Last modified: v2017.2 (12.1156684)

HS-3015: Automated Fit from CSV Data

- Create a Lookup model to link to tabulated data in an external .csv file.
- Run a DOE of type Run Matrix to import the data in the lookup .csv file.
- Build a predictive model using FAST (Fit Automatically Selected by Training).
- Examine the results from a FAST Fit.
Model Files
Exercise
Step 1: Review the CSV Data

1. Open the FAST_data.csv file and review its contents.
Notice: The .csv file contains two variables (x and y) and three responses.

and click OK.
5. Add a Lookup model by dragging-and-dropping the FAST_data.csv file from the
Directory into the work area.

8. In the Import Variables dialog, enter 2 in the Number of design variables field and
click OK.
The input variables are expected in the first two columns, and the remaining columns
are interpreted as output responses.

10. Review the input variables.
Notice: The bounds of the input variables are based on the FAST_data.csv file’s
contents. The nominal values are set to the first entry in the .csv file.

2. Click Apply.
The execution searches the .csv for a row of matching input variable values, and
returns the corresponding output responses in the row. If no match is found, you will
receive execution errors.

5. Go to the Define Output Responses step and review the output responses.
Notice: One output response is named Highly Nonlinear and two are polynomials.
Step 4: Run a Run Matrix DOE

4. In the work area, set the Mode to Run Matrix.
5. From the Settings tab, Matrix File field, navigate to your working directory and select
the FAST_data.csv file.
6. Click Apply.
The DOE matrix populates with the input variable values from the FAST_data.csv
file.
Step 5: Create a FAST Fit

2. In the Add – HyperStudy dialog, select Fit and click OK.
5. In the work area, Matrix Source column, select DOE 1 (doe_1).

9. Verify that the Fit Type assigned to each output response is FAST – Fit Automatically
Selected by Training.
Note: By default, FAST automatically selects the best Fit type from all available Fits.
You can manually select the Fit types FAST can choose by highlighting one or
more responses in the work area and selecting Fits from the Settings tab.
10. Click Apply.

Note: The choices for the best available Fit vary for each output response, which
can cause these loops to be time consuming compared to when you select a
single specific Fit. The steps for each output response are mutually
exclusive, therefore you can use the Multi-Execution option to accelerate this
process.
Step 6: Post-Process the FAST Fit

1. Click the Diagnostics tab.
Notice: The Highly Nonlinear response uses RBF, while the other responses use LSR.
In each case, FAST selected the specifics to have the highest x- validation R-
square value. The R-Square can be interpreted as the % of the data’s
variance that can be explained by the model.
2. Click the Regression Terms sub-tab.

4. Compare Poly1 and Poly2 by selecting them individually in the work area.
Poly1 and Poly2 are using stepwise regression, which means that the coefficients of the
regression are reduced to a minimal set that sufficiently models the data. Poly1 uses
only x, whereas Poly2 uses x and y^2.
Poly1
Poly2
5. If required, copy the data from the Fit Type and Fit Specifics columns in the Diagnostics
table and paste it into the Fit Type and Fit Specifics columns in the Specification step.
This step explicitly sets the Fit specifications to the results determined from FAST; if
the Fit must be re-run, this step can save time because FAST does not need to search
for the best settings.

6. Click the Trade-Off Tab to plot all the functions and see the predicted versus the
known data points.
Note: In each case, the Fit model follows the data closely regardless of the
sinusoidal functions in the Highly Nonlinear response to the simple planar
data of the polynomial responses.

Optimization

HS-4000: Optimization Method Comparison: Arm Model

Shape Optimization
Continuing from Tutorial HS-3000: Fit Method Comparison: Approximation on the Arm Model,
you will perform an Optimization and compare different methods for efficiency and
effectiveness.
Before running this tutorial, complete tutorial HS-3000: Fit Method Comparison:
Approximation on the Arm Model. You can also import the archive file HS-3000.hstx, available
in <hst.zip>/HS-4000/.
The corresponding output response values for your baseline design (all shape variables set to
0.0) were:
• Volume = 1.77E+06 mm3
• Max_Disp = 1.41 mm
• Max_Stress = 195.29 MPa
In this tutorial, the Optimization objective is to reduce Volume, while respecting a constraint
on Max_Disp that should be less than 1.5 mm.
In HS-3000, you learned that it was difficult to accurately capture the Max_Stress function
using a Fit approximation. In the DOE analysis, you learned that most of the tested design
configurations for Max_Stress were below 300 MPa. For these reasons, you will not consider a
constraint on the Max_Stress function. Max_Stress values can be collected throughout the
Optimization when running the exact solver.
Single Objective, Deterministic Optimization Study
In this tutorial, you will be comparing six single-objective, deterministic Optimization studies.
You will be changing the number of shape variables used, as well as the Optimization method.
The following Optimization methods will be used during this tutorial:
• Adaptive Response Surface Method (ARSM)
• Global Response Search Method (GRSM)
• Sequential Quadratic Programming (SQP)
• Genetic Algorithm (GA)
Step 1: ARSM, Six Input Variables, Exact Solver
Using the conclusions of the different design of experiments, you will consider only six shape
variables for the Optimization and omit the three radii (which will be fixed at their nominal
values).


5. In the work area, Active column, clear the radius_1, radius_2 and radius_3 check
boxes.

7. Apply an objective on the Volume output response.
a. In the Objectives column for Volume, click .
8. Apply a constraint on the Max_Disp output response.

a. In the Constraints column for Max_Disp, click .
b. In the pop-up window, define the following settings and click OK.
• Set Type to Deterministic.
• Set Bound Type to <=.

9. Click Apply.
12. Click Apply.
15. Click the Iteration History tab to view the optimum solution, which is highlighted green
in the table.
Note that the optimal design for Max_Stress is equal to 215, which is lower than 300.
16. Click the Iteration Plot tab to review the results of the optimization in an iteration plot.
• Select the Objective (Volume) and Constraint (Max_Disp) functions to see their
variations during the Optimization process.

• Select all of the design variables to see their variations during the Optimization
process.
Note if any of the input variables meet their bounds in the optimal design. If any
input variable's values meet their bounds, this indicates that relaxing these
bounds may enable you to find better solutions. In the plots below, only length_2
and length_5 meet their lower bounds.

Step 2: ARSM, Nine Input Variables, Exact Solver
1. Run a single objective, deterministic Optimization study by repeating Step 1: ARSM, Six
Input Variables, Exact Solver.
a. In the Define Input Variables step, activate all input variables.
2. Click the Iteration History and Iteration Plot tabs to review the results of the
Optimization.
a. Select the Objective (Volume) and Constraint (Max_Disp) functions to see their
Step 3: GRSM, Six Input Variables, Exact Solver
a. In the Specifications step, set the Mode to Global Response Search Method
(GRSM).
2. Click the Iteration history tab to review the results of the Optimization in a table. Note
that the optimal solution is found at the 19th evaluation (from 50).

4. Click the Iteration Plot tab to review the results of the Optimization in an iteration plot.
Step 4: SQP, Six Input Variables, Exact Solver
a. In the Specifications step, set the Mode to Sequential Quadratic Programming
(SQP).

Step 5: SQP, Six Input Variables, RBF_MELS
a. In the Select Output Responses step, Responses tab, set Evaluate From to
Fit, RBF (fit_4) for Max_Disp and Volume.
b. Clear the checkbox in the Active column for Max_Stress.

c. In the Specifications step, set the Mode to Sequential Quadratic Programming
(SQP).

3. For Optimizations using a Fit, it is recommended that you perform a validation run of the
optimal solution.
a. Click on Iteration History tab.
b. Select the parameter values for the optimal solution, then right-click and select
Copy from the context menu.
c. Go to the study Setup, Specifications step and click Edit Matrix > Run Matrix
from the top-right corner of the work area.
d. In the Edit Data Summary dialog, click Add Run and paste the copied values.
e. For the three radii, enter 0.

g. For height, enter 1.
h. Click Apply, then click OK to close the dialog.

i. Go to the Evaluate step and clear the checkbox in the Active column for the
previous nominal run, then click Evaluate Tasks.
4. Click the Evaluate Data tab, and compare the Volume, Max_Disp, and Max_Stress values
to those founds by the Optimization.
Note that the values are very similar.
Step 6: GA, Six Input Variables, RBF_MELS
a. In the Select Output Responses step, Responses tab, set Evaluate From to
RBF (fit_4) for Max_Disp and Volume.
b. Clear the Active checkbox for Max_Stress.
c. In the Specifications step, set the Mode to Genetic Algorithm (GA).

3. As you did in the previous Optimization (SQP, 6 IV, RBF_MELS), it is suggested that you
perform a validation run to compare the values provided by the Fit and by the solver.
Optimization Methods Comparison
The number of evaluations and the optimum found using the different Optimization methods
are compared in the table below. ARSM and GRSM are both response surface based methods.
Even though ARSM is the default method for single objective problems, it can fail when a
global optima is required. In such cases, it is suggested that you use GRSM. The best volume
decrease was obtained using GRSM by means of 50 solver evaluations. When SQP was used
with a solver, a good solution was found at the expense of additional evaluations. Also, when
SQP is used with a solver, it is sensitive to the starting point. When GA was applied on a Fit, a
solution similar to GRSM was found. GA was not run with the exact solver because it requires
many evaluations. In conclusion, GRSM or GA on a fit are the most efficient methods to solve
the Optimization problem.
Optimization Method # of Evaluations Volume Objective
ARSM, 9 IVs, Exact Solver 14 1702450.0
ARSM, 6 IVs, Exact Solver 11 1703330.0
GRSM, 6 IVs, Exact Solver 50 (22th is the optimum) 1652830.0
SQP, 6 IVs, Exact Solver 179 1659730.0
SQP, 6 IVs, Fit - 1666990.6
GA, 6 IVs, Fit - 1665387.3

Reliability-Based Design Optimization Study
In this section of the tutorial, you will be searching for 95% reliability on the Optimization
constraint (Max_disp < 1.5 mm). You will use fitting functions as opposed to the exact solver
to evaluate the output responses. Among the approximations, you will use the RBF that was
created with the MELS DOE. As a result, you will be using the SORA method. You will continue
using the six important input variables, which will all follow a normal distribution with a
variance of 0.1.
This topic will be discussed in HS-5000: Stochastic Method Comparison and Reliability- Based
Design Optimization of the Arm Model.
Multi-Objective Optimization Study
In this section of the tutorial, you will be searching for the Pareto front that minimizes both
volume and maximum displacement. You will be using MOGA with a Fit to save time.
Note: If a Fit was not available, GRSM would be the suggested method to use in order to
solve a MOO problem. MOO problems require many evaluations, therefore GRSM is
more efficient than MOGA.
This topic will be discussed in HS-4425: Multi-Objective Shape Optimization Study.
Last modified: v2017.2 (12.1156684)

HS-4200: Material Calibration Using System Identification
The purpose of this tutorial is to introduce a method for characterizing parameters of a

RADIOSS material law used for modeling elasto-plastic material. The characterization of a
ductile aluminum alloy is studied. A RADIOSS simulation is performed to replicate an
experimental tensile test. The parameters of the material law are determined to fit the
experimental results.
HS-1506: Material Calibration with a Curve Difference Integral provides an alternative method
to setup this problem using a HyperMath or Python function to measure the difference
between two curves.
Description of the Model

A quarter of a standard tensile test specimen is modeled using symmetry conditions. A
traction is applied to a specimen via an imposed velocity at the left-end.
The units are: mm, ms, g, N, MPa.
Geometry of the Tensile Specimen (One Quarter of the Specimen is Modeled)
Sections of Node Saved for Time History
The material to be characterized is a 6063 T7 Aluminum: it has an isotropic elasto-plastic

behavior which can be reproduced by a Johnson-Cook model without damage (RADIOSS Block
Law2), defined as follows:

In this study we define, as input variables, the parameters a, b, n, σmax (maximum stress)
and the Young modulus. The stress-strain curve obtained by the experimental test is shown in
the following image.
Engineering Stress Versus Engineering Strain Curve (Experimental Data)
For the simulation results, engineering strains will be obtained by dividing the displacement of
node 1 by the reference length (75 mm), and engineering stresses will be obtained by dividing
the force in section 1 by its initial surface (10.2 mm2).

Engineering Stress Versus Strain Curve (Simulation Results)

• Create an input template from a RADIOSS data file using the HyperStudy Template
Editor
• Set up a study
• Run a system identification optimization study
Copy the files TENSILE_TEST_0000.rad, TENSILE_TEST_0001.rad, and exper.xy from this
In this step, you can create the base input template in HyperStudy or use the base input
template in the study directory.
3. In the File field, navigate to your working directory and open the
Note: RADIOSS uses fixed fields of 20 characters for properties.
4. In the Find area, enter /MAT/PLAS_JOHNS/1 and click . HyperStudy highlights
/MAT/PLAS_JOHNS/1 in the TENSILE_TEST_0000.rad file.

5. Select E by starting at the beginning of row 51 and highlighting the first 20 fields. Tip:
To assist you in selecting 20-character fields, press CTRL to activate the
Selector (set to 20 characters) and then click the value.
7. In the Parameter Varname_1 dialog, Label field, enter E_Young.
8. Change the Lower Bound to 50000, the Initial Bound to 60400, and the Upper Bound
to 70000.
10. Click OK.

12. Define four more variables using the information provided in the table below.
Note: Some of the initial values are different from the values in the original file.
Variable Label Lower Initial Upper Format

Bound Value Bound
a a_PlasticityYieldStress 90 110 120 %20.5f
b b_HardeningCoeff 100 125 160 %20.5f
n n_HardeningExpo 0.1 0.2 0.3 %20.5f
sigmax Sigma_Max 250 280 290 %20.5f
13. Click Save.

TENSILE_TEST_0000.tpl.
15. Close the Editor dialog.
click OK.
a. From the Directory, drag-and-drop the TENSILE_TEST_0000.tpl file into the work
area.

b. In the Solver input file column, enter TENSILE_TEST_0000.rad. This is the name of
c. In the Solver input file column, select RADIOSS (radioss).
d. Verify that the Solver input arguments field reads $(file). This argument runs the
Starter, and the Engine of RADIOSS for the crash analysis. It also prevents the
creation of the .h3d result file from animation files. X is the number of CPUs to use
for the simulation.

e. Click Close.
TENSILE_TEST_0000.tpl resource file.
8. Check the input variable's lower and upper bound ranges.


2. Click Apply.
TENSILE_TESTT01 file, which stores the time history results of the simulation.
In order to fit the RADIOSS stress-strain curve to the experimental data we need to find a way
to compare the two curves. In this tutorial we will only focus on three specific points on the
curve.
Since damage is not modeled with this law, the comparison is not needed after the necking
point.
• Difference between experimental stress and RADIOSS at Strain equal 0.02 (1)
• Difference between experimental strain and RADIOSS at Necking point (2)
• Difference between experimental stress and RADIOSS at Necking point (3)

1. Create a data source labeled Disp_sim.

a. From the Directory, drag-and-drop the TENSILE_TEST01 file, located in
• Set Type to None/Node 1.
• Set Request to 4 Node 1.
• Set Component to DX-X Displacement.
e. Under Creating a new Data Source, enter Disp_sim in the Label field.
f. Clear the Linked to a new Response checkbox.
g. Click Finish

2. Create a second data source labeled Force_sim by repeating step 1, except define the
following options:
• Set Type to Section/SECTION_2.
• Set Request to 2 section 1.
• Set Component to FT-Resultant Tangent Force.
3. Create three output responses.
a. Click Add Output Response three times.
b. In the work area, Label column, change the labels for the three output responses to
Radioss_Strain_0_2, Radioss_Stress_Necking, and Radioss_Strain_Necking.
4. Define the Radioss_Strain_0_2 output response.
a. In the Expression column of the output response Radioss_Strain_0_2, click .

b. Click the Function tab.
c. From the list of functions, select lininterp.

e. Click Insert Varname. The function lininterp(,,) appears in the Evaluate

Expression field.
f. In the Evaluate Expression field, enter (m_1_ds_1/75,m_1_ds_2/10.2,0.02) in the
lininterp function.
This expression computes the Stress with respect to the Strain, at Strain equals
0.02.
g. Click Evaluate Expression.

h. Click OK to close the Expression Builder.
5. Define the Radioss_Stress_Necking output response.
a. In the Expression column of the output response Radioss_Stress_Necking, click
.
c. From the list of data sources, select Force_sim.
d. From the Insert Varname drop-down, select Maximum.
e. Click Insert Varname. The function max(m_1_ds_2) appears in the Evaluate

Expression field.
f. In the Evaluate Expression field, edit the function so that it reads
max(m_1_ds_2)/10.2.

This is the maximum of the force (m_1_ds_2), which is trimmed between the min
strain and the strain at the max value of Force, divided by 10.2 (surface area) to
obtain the stress.
Note: In some similar curve fitting problems it may be required to trim the data
after the ultimate load, the expression to do this is:
max(m_1_ds_2[subrange(m_1_ds_1,min(m_1_ds_1),m_1_ds_1[indexofma
x(m_1_ds_2)])])/10.2.
g. Click Evaluate Expression.
h. Click OK to close the Expression Builder.
6. Define the Radioss_Strain_Necking output response.
a. In the Expression column of the output response Radioss_Strain_Necking, click
.
b. In the Evaluate Expression field, enter m_1_ds_1[indexofmax(m_1_ds_2)]/75.
This is the displacement (m_1_ds_1) at the max value of the force value, divided by
75 (reference length) to obtain strain.
Note: In some similar curve fitting problems it may be required to trim the data
after the ultimate load, the expression to do this is:
m_1_ds_1[maxindex(subrange(m_1_ds_1,min(m_1_ds_1),m_1_ds_1[indexof
max(m_1_ds_2)]))]/75.
c. Click Evaluate Expression.
d. Click OK to close the Expression Builder.


7. Apply an objective on the Radioss_Strain_0_2 output response.
a. In the Objectives column for Radioss_Strain_0_2, click .
• Set Type to System Identification.
• For Target Value, enter 141.00000.
8. Apply an objective on the Radioss_Stress_Necking and Radioss_Strain_Necking output

responses.
a. Radioss_Stress_Necking
• Type = System Identification
• Target Value = 148.00000
b. Radioss_Strain_Necking
• Type = System Identification
• Target Value = 0.0800000
9. Click Apply.
12. Click Apply.
15. Click the Iteration Plot tab.
Using the Channel selector, select the three objectives from and Objective Function.
Activate multi-plot to see the each channel in its own plot.
The first three selections are the actual values used in the system identification
optimization problem. Observe their objective history to see that their values indeed

approach their respective target values. The final plot is the scalar objective which is used
in the system identification problem; a normalized sum of the squares difference between
the actual and target objective values. Note that the value of this combined function has
been reduced through the optimization.
Last modified: v2017.2 (9.1144596)

HS-4205: Multi-Objective Optimization Study Using an Excel

Spreadsheet
This tutorial demonstrates how to perform a multi-objective optimization with HyperStudy.

Before running this tutorial, you must complete HS-1035: Optimization Study Using an Excel
Spreadsheet or you can import the archive file HS-1035.hstx, available in
<hst.zip>/HS-4205/.
In the initial optimization problem, as stated in tutorial HS-1035, the objective is to find the
cross-sectional dimensions (width and height) of a beam that minimize the beam volume while
keeping the tip deflection below 0.53 mm.
In this tutorial, both the beam volume and tip deflection will be considered as objective
functions to be minimized. The constraint in the tip deflection is not applied in an effort to
understand the trade-off associated with this condition Using a Global Response Search
Method (GRSM), a series of solutions (called non-dominated solutions), will be found. These
solutions form a Pareto front in which users can do trade-off analysis.
Step 1: Start from Tutorial HS-1035
2. Perform all steps in tutorial HS-1035.
Step 2: Add an Approach and Run a Multi-Objective Optimization

Study
4. For both input variables, change the lower, initial, and upper bounds to the values

6. Add two objectives.
a. Click the Objectives tab.
b. Click Add Objective.
c. In the Add - HyperStudy dialog, add two objectives.
d. Define Objective 1 and Objective 2 by selecting the options indicated in the image
below from the Type and Apply On columns.

7. Click Apply.
10. Click Apply.
13. Optional. HyperStudy also allows you to perform multi-objective Optimizations using a
Weighted Sum approach. Add and activate multiple objectives amongst the output
responses, set a weight respectively for each objective in the Weighted Sum field, and
select an algorithm (SQP, MFD, GA, ARSM) in the Specifications table. Proceed with the
optimization.
Step 3: Post-Process the Multi-Objective Optimization Results
1. Click the Optima tab.

2. Using the Channel selector, select Objective 1 for the X Axis and Objective 2 for the
Y Axis.
The Optima tab plots the trade-off between competing objectives. One objective cannot
improve without the other competing objective getting worse.

HS-4210: Multi-Disciplinary Optimization Study
This tutorial demonstrates how to perform a multi-disciplinary size optimization for two finite
element models defined for OptiStruct that have common input variables. The sample base
input templates plate1.tpl and plate2.tpl, can be found in <hst.zip>/HS-4210/ and
copied to your working directory.
The objective is to minimize the volume of the plate under a stress and a frequency constraint.
The input variables are the thickness of each of the three components, defined in the input
deck via the PSHELL card. The thickness should be between 0.05 and 0.15; the initial
thickness is 0.1 (shown below). The optimization type is size. To demonstrate the use of the
optimization tool in a multi-disciplinary optimization, two models are created.
One model is used for the stress analysis and one for the frequency analysis. Both models
must have the same input variables.
Figure 1. Double symmetric plate model.
click OK.
a. From the Directory, drag-and-drop the plate1.tpl file into the work area.

c. In the Solver input file column, enter plate1.fem. This is the name of the solver
a. From the Directory, drag-and-drop the plate2.tpl file into the work area.
b. In the Solver input file column, enter plate2.fem. This is the name of the solver
7. Click Import Variables. Six input variables are imported from the plate1.tpl and
plate2.tpl resource files.
11. In the Expression column of the input variable Property 21, click .
12. In the Expression Builder, click the Input Variables tab.
13. In the work area, select Property 11.
14. Click Insert Varname. The expression m_1_TH1 appears in the Evaluate expression
field.

15. Click OK. Property 21 of Model 2 is linked to Property 11 of Model 1.

16. Create two more links.
• Link Property 22 to Property 12.
• Link Property 23 to Property 13.

2. Click Apply.
directory. The approaches/nom_1/run 00001/m_1 and
approaches/nom_1/run 00001/m_2 sub-directories contain the plate2.out (for the
structural volume and frequency) and plate1.h3d (for the stresses) files, which are the
results of the nominal run, and will be using during the Optimization.
5. Go to the Define Output Response step.
In this step you will create three output responses: Volume, Stress43, and Frequency1.
1. Create the Volume output response, which represents the volume of the plate.
a. From the Directory, drag-and-drop the plate2.out file, located in the


f. Label the output response Volume.

g. Set Expression to First Element.
Note: Because there is only a single value in this data source, HyperStudy inserts a
[0] after m_1_ds_1, thereby choosing the first (and only) entry in the data
source.
h. Click Finish. The Volume output response is added to the work area.

2. Create the Stress43 output response, which represents the von Mises Stress of Element
43.
a. From the Directory, drag-and-drop the plate1.h3d file, located in the
approaches/nom_1/run 00001/m_1 directory, into the work area. This file contains
the analysis results, including stresses.
• Set Request to E43.
• Set Component to vonMises (Mid).
e. Label the output response Stress43.
g. Click Finish. The Stress43 output response is added to the work area.
3. Create the Frequency1 output response, which represents frequency results.
a. From the Directory, drag-and-drop the plate2.out file, located in the
b. In the File Assistant dialog, click Next.
• Set Request to Mode 1.
e. Label the output response Frequency1.
g. Click Finish. The Frequency1 output response is added to the work area.

b. In the pop-up dialog, set Type to Minimize and click OK.
7. Apply a constraint on the Stress43 output response.

a. In the Constraints column for Stress43, click .
b. In the pop-up dialog, define the following and click OK.
• For Bound Value, enter 22.
8. Apply a constraint on the Stress43 output response.
a. In the Constraints column for Frequency1, click .
• Set Bound Type to >=.
9. Click Apply.
12. Click Apply.


16. Click the Iteration Plot tab to plot the progress of the Optimization iteration.
Using the Channel selector, select Objective 1, Constraint 1, and Constraint 2. Above
the Channel selector, activate multiplot and enable the Bounds setting.
Over the course of the optimization, the objective is minimized and at the conclusion, the
constraints are satisfied. In the plots, the large markers indicate a design which has at
least one violated constraint and a small marker indicates a feasible design. At the optimal
design, the only active constraint is Constraint 1. In contrast, constraint 2 is not active at
the optimum; this indicates Constraint 2 does not have an influence on the result.
Last modified: v2017.2 (12.1156684)

HS-4215: Multi-Disciplinary Design Optimization Study
This tutorial demonstrates how to perform a multi-disciplinary design optimization study. The
disciplines used in this tutorial are structural performance and cost. Structural performance is
simulated using OptiStruct and Cost is simulated using Compose or Python. Optimization
parameters for both the simulations are identified in template files corresponding to each input
deck (tail.fem (OptiStruct) and tail.oml (Compose)/tail.py (Python).
The sample base input templates used in this tutorial can be found in <hst.zip>/HS-4215/
and copied to your working directory. The tutorial directory includes the following files:
tail_structure_optistruct.tpl Base input parameterized file

model 1.
tail_cost_compose.tpl Base input parameterized file

model 2.
tail_cost_python.tpl Base input parameterized file

model 2.
Horizontal tail plane model
It is assumed that the tail is cantilevered about its inboard section. Three loading scenarios
are considered; one where the tail experiences pressure loads of 0.25 psi on the bottom skin,
a second where the tail experiences a tip load of 400 lbs, and a third where the tail
experiences both the pressure load and tip load simultaneously. The applied loading is
represented in the following figure.

Loading experienced by horizontal tail plane
Problem Formulation for this study is as follows:

Input variables:
• glass fabric thickness at inboards; initial value = 0.1; lower bound = 0.01, upper bound
= 2.0
• glass fabric thickness at midspan; initial value = 0.1; lower bound = 0.01, upper bound
= 2.0
• glass fabric thickness at outboards; initial value = 0.1; lower bound = 0.01, upper
bound = 2.0
• core thickness at inboards; initial value = 0.1; lower bound = 0.01, upper bound = 2.0
• core fabric thickness at midspan; initial value = 0.1; lower bound = 0.01, upper bound
= 2.0
• core fabric thickness at outboards; initial value = 0.1; lower bound = 0.01, upper
bound = 2.0
• aluminum rib thickness; initial value = 0.1; lower bound = 0.01, upper bound = 2.0
Note: Both models have seven input variables; however values

of the input variables need to be consistent between the
two models. In order to obtain this, we will be linking
the two sets of input variables to each other.
Objective:
To minimize the cost.
Design constraints:
Maximum displacement must be less than its baseline value of 31.

click OK.
a. From the Directory, drag-and-drop the tail_structure_optistruct.tpl file into
the work area.
b. In the Solver input file column, enter tail.fem. This is the name of the solver input
6. Add a second Parameterized File model.
a. From the Directory, drag-and-drop the appropriate .tpl file into the work area.
• If you are using Python, use the tail_cost_python.tpl file.
• If you are using Compose, use the tail_cost_compose.tpl file.
b. In the Solver input file column, enter a name for the solver input file HyperStudy
• If you are using Python, enter tail.py.
• If you are using Compose, enter tail.oml.
c. In the Solver execution script column, select either Python (py) or Compose
(oml) accordingly.
d. If you are using Compose as the Solver execution script, in the Solver input
arguments column, enter -f before $file.
Note: If you are using Compose as part the HyperWorks suite, than HyperStudy
should automatically point to the correct .bat file. If you have Compose as a
separate installation, than during the Register Solver Script step you must
point to Compose_batch.bat.
7. Click Import Variables. Fourteen input variables are imported from the two .tpl
resource files.
11. In the Varname column, copy all of the independent variables (all variables from

Model_1).
13. In the Expression column of all of the dependent input variables (all variables from
Model_2), paste the independent variables.

2. Click Apply.
directory. The approaches/nom_1/run 00001/m_1 and
approaches/nom_1/run 00001/m_2 sub-directories contain the tail.h3d (for maximum
displacement) and cost.res (for cost) files, which are the result of the nominal run, and
will be used in the optimization.
In this step you will create two output responses: MaxDisp and Cost.
1. Create the MaxDisp output response.
a. From the Directory, drag-and-drop the tail.h3d file, located in the


• Set Subcase to Subcase 5 (Combo).
• For Request, set Start to First request and enter N4660, and set End to
Last request and enter N7528.
• For Component, select MAG.
e. Label the output response MaxDisp.


h. Click Finish. The MaxDisp output response is added to the work area.
2. Create the Cost output response.
a. From the Directory, drag-and-drop the cost.res file, located in the
e. Label the output response Cost.
g. Click Finish. The Cost output response is added to the work area.

6. Click the Scatter tab to compare Cost versus MaxDisp.
6. Apply an objective on the Cost output response.
a. In the Objectives column of Cost, click .
b. In the pop-up dialog, set Type to Minimize and click OK.
7. Apply a constraint on the MaxDisp output response.

a. In the Constraints column of MaxDisp, click .


8. Click Apply.
11. Click Apply.
Click the Iteration Plot tab to plot the progress of the Optimization iteration..
Using the Channel selector, select Objective_1 and Constraint_1. The evolution of the
objective function and constraint vs. iterations is 2D plotted. You can see that the cost of the
horizontal tail plane is reduced from 72715 to 67700 (7% reduction), while keeping the
structural performance the same.
Last modified: v2017.2 (12.1156684)

HS-4220: Size Optimization Study on an Impact Simulation

Using RADIOSS
This tutorial demonstrates how to perform a size optimization on a finite element model
defined for RADIOSS. The RADIOSS model shown in figure 1 is run using the RADIOSS Starter
and Engine. The sample base input files boxbeam1_0000.rad and boxbeam1_0001.rad can be
found in <hst.zip>/HS-4220/and copied to your working directly.
The objective is to minimize the mass of the beam under the following two constraints: the
internal energy must be more than 450, and the resulting reaction force must be less than
75. The input variables are the thicknesses of the four components defined in the input deck
boxbeam1._0000.rad via the /PROP/SHELL entries. They are combined into two input
variables. The thickness should be between 0.5 and 2.0; the initial thickness is 1.0. The
optimization type is size.
Figure 1. Boxbeam model, undeformed.
Figure 2. Boxbeam model, deformed, t = 2.001.

1. Start HyperStudy
3. In the File field, navigate to your working directory and open the boxbeam1_0000.rad
file.
4. In the Find area, enter /PROP/SHELL/1.
5. Click until you find /PROP/SHELL/1.
6. Highlight the field for thickness.
Tip: To assist you in selecting 20-character fields, press CTRL to activate the
Selector (set to 20 characters) and then click the value. HyperStudy highlights
20 fields.
8. In the Parameter - varname_1 dialog, Label field, enter Upper part.
9. Set the Lower Bound to 0.5, the Nominal to 1.0, and the Upper Bound to 2.0.
11. Click OK.

12. Find /PROP/SHELL/2 and highlight the field for thickness.

13. Assign it the same thickness as /PROP/SHELL/1 by right-clicking on the highlighted
fields and selecting Attach to > varname_1 from the context menu.
16. In the Parameter - varname_2 dialog, Label field, enter Lower part.
17. Set the Lower Bound to 0.5, the Nominal to 1.0, and the Upper Bound to 2.0.
19. Click OK.

21. Assign it the same thickness as /PROP/SHELL/3 by right-clicking on the highlighted
fields and selecting Attach to > varname_2 from the context menu.
22. Click Save.
boxbeam1.tpl.

Step 2: Optional. View the Base Input Template in TextView
1. Start HyperGraph.
2. On the Client Selector toolbar, select TextView.
3. From the menu bar, click File > Open > Document.
4. In the Open Document dialog, open the boxbeam1.tpl file. The text editor displays the
following input variables that are defined by Templex parameter statements:
{parameter(t1,"Upper part",1.0,0.5,2.0)}
{parameter(t2,"Lower part",1.0,0.5,2.0)}
5. On the Text toolbar, click .

6. In the Find dialog, Find field, enter /PROP/SHELL.
7. Click . The parameterized /PROP/SHELL cards, which reference the input variables,
highlight.

9. On the Text toolbar, click . The text editor evaluates the Templex statements, and
replaces the parameters with their initial values.
10. Repeat steps 2.5 through 2.7, and search for /PROP/SHELL again. You will find:
11. Close HyperGraph; you do not need to save the session.
1. Return to HyperStudy.
click OK.
a. From the Directory, drag-and-drop the boxbeam1.tpl file into the work area.

c. In the Solver input file column, enter boxbeam1_0000.rad. This is the name of the
d. In the Solver execution script column, select RADIOSS (radioss).

boxbeam1_0001.rad file.
e. Click Close.
7. Click Import Variables. Two input variables are imported from the boxbeam1.tpl
resource file.



2. Click Apply.
directory. The approaches/nom_1/run 00001/m_1 directory contains the result files.
1. Create the Energy output response, which is the internal energy of the model.
a. From the Directory, drag-and-drop the boxbeam1T01 file, located in
HyperWorks® (hgradioss++.exe) and click Next.
• Set Type to Global Variables.
• Set Request to Internal Energy.

e. Label the output response Energy.

g. Click Finish. The Energy output response is added to the work area.
2. Create the Force output response, which is the resultant reaction force in the Z- direction.
• Set Type to Rigid wall/Wall Force.
• Set Request to 1 RWALL 1.
• Set Component to FNZ-Z NORMAL FORCE.
e. Label the output response Force.
g. Click Finish. The Force output response is added to the work area.

• Set Type to Global Variables.

f. Label the output response Mass.
g. Set Expression to First Element.
h. Click Finish. The Mass output response is added to the work area.
4. Review the lower and upper bound ranges of the input variables.
6. Apply an objective to the Mass output response.
a. In the Objectives column for Mass, click .
7. Apply a constraint to the Energy output responses.

a. In the Constraints column for Energy, click .
• Set Bound Type to >=.
8. Apply a constraint to the Force output responses.

a. In the Constraints column for Energy, click .

10. Click Apply.

13. Click Apply.
1. Click the Iteration History tab to display data in a tabluar view. The optimal design is
highlighted green, the infeasible designs are shown with red text, and the violated
constraints are indicated in bold text.
2. Click the Iteration Plot tab to plot the iteration history of the study's objectives,
constraints, and input variables.
Using the Channel selector, select Objective 1, Constraint 1, and Constraint 2.
In the initial design, the design was infeasible as indicated by the large circular marker for
the first iteration. A view of the constraint plots shows that the second constraint was
violated in the initial design. Initially, the optimizer added some weight in order to satisfy
the design constraints. Notice that both constraints are near their bounds in the optimal
design.

Last modified: v2017.2 (12.1156684)

HS-4230: Optimization Study with Discrete Variables
In this tutorial you will learn how to use discrete variables. The sample base input template
plate.tpl can be found in <hst.zip>/HS-4230/and copied to your working directory.
The objective of this tutorial is to maximize the minimum frequency of the first five modes of a
plate. The input variables are the thickness of each of the three components, defined in the
thickness within the files is 0.1. The optimization type is size. Furthermore, optimum design
should have input variables from a discrete set of 0.05, 0.08, 0.11, and 0.14 for all three
thicknesses. By default, HyperStudy will add the values from the lower and upper bounds to
this set. Hence the resulting set is 0.05, 0.08, 0.11, 0.14, and 0.15. Delete any of these
values if needed.
.
click
OK.

b. In the Solver input file column, enter plate.fem. This is the name of the solver
7. Click Import Variables. Three input variables are imported from the plate.tpl file.
10. In the Mode column of all three input variables, select Discrete.
11. In the Values column of Property 21, click .
12. Click Step Size, enter 0.03, and click Set.

14. Click Apply.
15. Click OK.
16. Repeat steps 10 through 13 for Property 22 and Property 23.
2. Click Apply.

directory. The approaches/nom_1/run 00001/m_1 sub-directory contains the plate.out
file, which is the result of the nominal run, and will be used in the Optimization.

1. Create the Freq1 output response.


e. Label the output response Freq1.
g. Click Finish. The Freq1 output response is added to the work area.



b. Set Apply On to Volume (r_2).

a. Set Apply On to Freq1 (r_1).
b. Set Bound Type to >= (greater than or equal to).
c. For Bound Value, enter 32.
13. Click Apply.
16. Click Apply.
1. Click the Iteration History tab to monitor the progress of the Optimization iteration.

Step 6: Setup a DOE to Find the True Best Design
5. Click Apply.
10. Sort run data based on the Volume (which was to be minimized) by right-clicking on the
Volume column and selecting Sort down from the context menu. The lowest volume
design which satisfies the constraint (frequency > 32) is the same as that found by the
optimizer.
Note: The DOE took 125 solver calls to exhaust all combinations, whereas the
Optimization found it in 8 solver calls.

HS-4405: Reliability-Based Optimization Study on an Impact

Simulation Using RADIOSS
This tutorial demonstrates how to perform a reliability-based optimization on a finite element

model defined for RADIOSS.
Before running this tutorial, you must complete tutorial HS-4220: Size Optimization Study on
an Impact Simulation using RADIOSS or you can import the archive file HS-4220.hstx,
available in <hst.zip>/HS-4405/.
In the initial optimization problem, as stated in tutorial HS-4220, the objective is to minimize
the mass of the beam under the following two constraints: the internal energy must be more
than 450, and the resulting reaction force must be less than 75. The input variables are the
thicknesses of the four components defined in the input deck boxbeam1_0000.rad via the
/PROP/SHELL entries. They are combined into two input variables. The thickness should be
between 0.5 and 2.0; the initial thickness is 1.0.
The reliability is added in this study through the definition of uncertainties on the input
variables and probability targets for the constraints.
The thicknesses follow a normal distribution, with mean 1.0 and coefficient of variation 0.10.
The constraints are expressed as follows: Prob(internal energy > 450) > 0.98
Prob(reaction force < 75) > 0.98
This means: Taking into account possible variations created by the random parameters;
we want the 98th percentile of the reaction force distribution to be less than 75.
The ARSM-SORA optimization engine is used in this tutorial. SORA (Sequential Optimization
and Reliability Assessment) is an algorithm that makes it possible to manage random variables
and set reliability targets on constraints. ARSM-SORA takes advantage of the response surface
based approach to reduce the computational effort needed in such problems
Step 1: Start from Tutorial HS-4220
2. Perform all steps in tutorial HS-4220.
Step 2: Run a Reliability-Based Optimization Study
1. In the Explorer, right-click on Optimization 1 and select Copy Approach from the
context menu.
2. In the Copy - HyperStudy dialog, click OK. A copy of Optimization 1 opens in the
Explorer.
4. Review and edit the probabilistic properties by right-clicking in the Select Input
Variables table and selecting Columns > Show All from the context menu. All of the
columns available appear in the work area.

6. In the Distribution Role column of both input variables, select Design with Random.
8. Edit constraints.
b. Define both constraints.
• Set Type to Random.
• For CDF Limit, enter 98.00.
9. Click Apply.
11. In the work area, set the Mode to ARSM based SORA (SORA_ARSM).
12. Click Apply.
Step 3: View the Iteration History of the Optimization Study
When using the SORA or ARSM-SORA engines, additional information is displayed for each of
the Random type constraints. These are the percentile values (labeled _PV) and they are
related to the CDF Limits.
For a constraint of the form Prob(g > b) > R , the constraint is satisfied in the probabilistic
way if the Rth-Percentile value of g is greater than b. R stands for the target probability, and b
stands for the bound value.
For a constraint of the form Prob(g < b) > R, the constraint is satisfied in the probabilistic way
if the Rth-Percentile value of g is smaller than b.
When using SORA or ARSM-SORA, the history table displays, as the first iteration, the result of
the deterministic optimization (i.e. without taking the random parameters into account). The
following iterations are successive iterations made to satisfy the probabilistic constraints.
1. Click the Iteration History tab to review the SORA_ARSM history. Iteration 1 is the
outcome of the deterministic optimization. Iterations 2 to 6 summarize the probabilistic
steps. The two constraints match the probabilistic target: constraint_2_PV = 75.08. This
indicates that the 98th percentile value of constraint 2 (reaction force) satisfies the
75.0 upper bound you defined.


HS-4410: Readmac Delegate for Mode Tracking

In this tutorial you will learn how to use the Readmac delegate for mode tracking applications.
This will be set up as an optimization problem with the objective of maximizing the first
natural frequency of a simple rectangular section beam.
Bar cross-section of the beam element from hyperbeam
Model Files
Exercise

and click OK.


a. From the Directory, drag-and-drop the Bar_beam.tpl file into the work area.
b. In the Solver input file column, enter Bar_beam.fem.

This is the name of the solver input file HyperStudy writes during any
evaluation.
Two input variables are imported from the .tpl resource file.
8. Go to the Specifications step
Step 2: Perform the Nominal run

2. Click Apply.
An approaches /nom_1/ directory is created inside the study directory.
/nom_1/run 00001/m_1 sub-directory contains the Bar_beam.h3d which is the result
of the nominal run, and will be used for next approach.


In this step you will create three output responses that are based on elements of a single data
source from the Readmac delegate.
1. Click the Data Sources tab.
2. Click Add Data Source.
3. In the File column of the data source, click (…).
4. In the Data Source Builder dialog, define the data source and click OK.
a. In the File field, navigate to \nom_1\run 00001\m_1\ inside your working
directory and select Bar_beam.h3d.
b. Set Tool to Modal Assurance Criteria.
c. In the References section, File field, navigate to \nom_1\run 00001\m_1\
inside your working directory and select Bar_beam.h3d.
d. Set Subcase to Subcase 1 (loadstep1).
e. Set Type to Eigen Mode (Grids).
f. For Mode, select Reference mode index = 1.
g. In the Target section, leave the settings set to their default values. The default
Target settings adopt the same Type and Subcase as the
Reference settings, which enables HyperStudy to search all modes for the
best match in the reference file.
Note: The file source uses a relative position of the Bar_beam.h3d file, and updates
for successive run numbers. However, the reference file uses an absolute
path and points to the given file source throughout all the runs.

5. Click Evaluate.
The Value field for ds_1 displays a row vector of three elements.
6. Click the Define Output Responses tab.
7. Create and define output responses.
a. Click Add Output Responses to add three output responses.
b. Label the output responses Freq of best matched mode, MAC of matched mode,
and ID of matched mode.
c. In the Expression field, enter the following expressions for each output
response.
• Freq of best matched mode = ds_1[0]
• MAC of matched mode = ds_1[1]
• ID of matched mode = ds_1[2]
Step 4: Run an Optimization

2. In the Add – HyperStudy dialog, select Optimization and click OK.
4. Apply an objective on the Freq of best matched mode output response.
a. In the Objectives column for Freq of best matched mode, click .
b. In the pop-up window, set Type to Maximize and click OK.
5. Click Apply.

11. Click the Iteration History tab.

Notice: The frequency of best matched mode in the target file is maximized to
50.067 from 33.388. The readmac function that was running the background
was able to track the ID of the matched mode in the target
.h3d file. The mode ID in the converged solution switched from 1 to 2.

HS-4415: Optimization Study of a Landing Beam Using Excel
This tutorial demonstrates how to perform an optimization study in which the input variables
are entered and the output responses are calculated in a Microsoft Excel spreadsheet. The
Excel spreadsheet LandingBeamCalc_Public.xls file can be found in
<hst.zip>/HS-4415/ and copied to your working directory. To watch a demonstration video of
this tutorial, click here.
The objective is to find the cross-sectional dimensions of a tapering I- beam at its three
sections that minimize the total cross-sectional area while meeting the margin of safety
requirements for buckling, crippling, and combined bending and shear under ten loadcases.
The spreadsheet used here contains a page with the initial design and separate pages for
crippling, buckling, and combined bending and shear calculations.
Step 1: Create a Matrix Input that HyperStudy Can Evaluate
1. In Excel, open the LandingBeamCalc_Public.xls spreadsheet.

2. Review the information, and locate the columns that contain the input variables and
output responses.
Note: When creating a Spreadsheet model for HyperStudy on a Mac or Windows
platform, variable labels should only contains English characters, or a
combination of English characters and numbers.
click OK.
5. Add a Spreadsheet model by dragging-and-dropping the LandingBeamCalc_Public.xls
file into the work area.
The Solver input file column displays hst_input.hstp, this is the name of the solver
input file HyperStudy writes during any evaluation. The Solver execution script column
now displays SpreadSheet_HST.

6. Optional. If a firewall prompt dialog appears, click Allow.

7. Click Import Variables. The LandingBeamCalc_Public.xls spreadsheet opens.
8. In the Excel - HyperStudy dialog, click Yes to begin selecting input variables.
9. In the spreadsheet, select the cells AA_w1, AA_w2, and AA_w3 in Section AA, along
with their corresponding values.
10. In the Excel - HyperStudy Input Selector dialog, click OK.

11. Select the following:
• Section AA: AA_h1 and AA_h2
• Section CC: CC_w1, CC_w2, CC_w3
• Section CC: CC_h1, CC_h2
• Section EE: EE_w1, EE_w2, EE_w3
• Section EE: EE_h1, EE_h2

12. Click Cancel to stop selecting input variables from the spreadsheet.
13. In the Excel - HyperStudy dialog, click Yes to begin selecting output responses.
14. In the spreadsheet, select the cell AA_MS_BS in Section AA, along with its
corresponding value.
15. In the Excel - HyperStudy Output Selector dialog, click OK.

16. Select the following:
• Section AA: AA_MS_C
• Section AA: AA_MS_B
• Section CC: CC_MS_BS
• Section CC: CC_MS_C
• Section CC: CC_MS_B
• Section EE: EE_MS_BS
• Section EE: EE_MS_C
• Section EE: EE_MS_B
• Area ACE value from cell C10 (illustrated in the image below):

17. Click Cancel to stop selecting output responses from the spreadsheet. Fifteen input
variables and ten output responses are imported from the LandingBeamCalc_Public.xls
spreadsheet.

2. Click Apply.
sse_output.csv file, which is the results file of the nominal run.
Step 4: Add an Optimization Approach and a Run Optimization
b. Set Apply On to Area ACE (r_10).

11. In the Add - HyperStudy dialog, add nine constraints.
12. Define Constraint 1 through Constraint 9 by selecting the options indicated in the
image below from the Apply On, Bound Type, and Bound Value columns.

14. Click Apply.

16. In the work area, set the Mode to Sequential Quadratic Programming (SQP).
17. Click Apply.
19. Click Evaluate Tasks to launch the optimization.
Step 5: View the Iteration History of the Optimization
1. Click the Iteration Plot tab to monitor the evolution of the objective function and
constraints vs. the iterations.
2. Using the Channel selector, select Constraint 1 through Constraint 9. The Optimization
iteration history of the constraints is plotted.

HS-4420: Optimization Study of a Spherical Impactor
This tutorial demonstrates how to perform an advanced study that has both size and shape
input variables on a RADIOSS finite element model.
The sample base input template can be found in <hst.zip>/HS-4420/. Copy the files
impactor.hm, impactor_0000.rad, and impactor_0001.rad from this directory to your
working directory.
The steps taken in this tutorial demonstrate how to analyze the input variables in order to
identify the most important variables and how to do an Optimization. The objective of the
Optimization is to minimize the maximum acceleration of the impactor, while keeping
maximum displacement lower than 16 mm.
• Create a base input template from a RADIOSS input file using the HyperStudy Editor
• Set up a study
• Run a DOE study (screening DOE)
• Post process DOE results in order to define the most important variables and reduce
the number of variables (screening)
• Create a new DOE in order to create an approximation
• Create an approximation
• Run an Optimization study based on the approximation created
This model simulates the dynamic impact of a sphere with an initial velocity on a box. There
are eight variables: four size variables, which are four box thickness, and four shapes
variables.
Size variables

Shapes variables
Step 1: Export the Shape Parameterization from HyperMesh

2. In the User Profiles dialog, change the user profile to RADIOSS.
4. In the Open Model dialog, open the impactor.hm file. The impactor.hm database has the
RADIOSS analysis setup, and the shapes have already been created. You must export the
shapes variables so that they are included in the template file.
5. From the Tool page, click Shape.

7. Switch single desvars to multiple desvars.

9. In the initial value= field, enter 0.

10. In the lower bound= field, enter -1.
11. In the upper bound= field, enter 1.
12. Click the shapes selector.
13. Select all of the shapes.

14. Click select.
15. Click create. A shape design variable is created for each shape.
16. Optional.
• If you would like to animate or visualize the shapes, click animate.
• In the Deformed panel, click linear or modal to animate the shape variables in the
graphics area.
• While the shape is animating, you can adjust the animation speed by moving the
slider as indicated in the image below.
17. Go to the export subpanel.

18. For analysis code, select HyperStudy.
19. For sub-code, select Radioss51.
20. In the File field, enter impactor.shp.

21. Click export as.

impactor.shp. HyperMesh writes the following files:
impactor.radioss51.node.tpl Grid coordinates template.
impactor.shp Grid perturbation vector data read by

impactor.radioss51.node.tpl.
23. Exit HyperMesh Desktop.
2. From the menu bar, click Tools > Editor.
3. In the Editor dialog, File field, open the impactor_0000.rad file.
4. Right-click anywhere in the editor and select Select Nodes > /NODE from the context
menu. All of the /NODE cards in the impactor_0000.rad file highlight.
5. Right-click on the highlighted cards and select Include Shape from the context menu.
6. In the Shape Template dialog, open the impactor.radioss51.node.tpl file. The shape
variables are now created and the grid has been replaced by the parameter file (which
contains the grid parameterized by the shapes) exported during step 1.
7. Locate the shape variable prop_external_skin.
8. Select the thickness value for prop_external_skin, as indicated in the image below. In a

RADIOSS deck, each field within a card is 20 characters long.

Tip: To assist you in selecting 20-character fields, press CTRL to activate the
Selector (set to 20 characters) and then click the value. HyperStudy highlights
20 fields.
11. In the Parameter - varname_1 dialog, Label field, enter th_external_skin.
12. Change the Lower Bound to 1.0, the Nominal value to 1.0, and the Upper Bound to
2.0.
13. Change the Format to %20.5f.
14. Click OK.

15. Define three more input variables for thickness using the information provided in the table
below.
Input Variable Label Lower Nominal Upper Format

Bound Value Bound
prop_internal_skin th_internal_skin 1.0 1.0 2.0 %20.5f
prop_external_flange th_external_flange 1.0 1.0 2.0 %20.5f
prop_internal_flange th_internal_flange 1.0 1.0 2.0 %20.5f
16. Click Save.

impactor.tpl.
click OK.
a. From the Directory, drag-and-drop the impactor.tpl file into the work area.
b. In the Solver input file column, enter impactor_0000.rad. This is the name of

the starter input file HyperStudy creates from the parameterization, and the name of
the Engine file.
d. In the Solver execution script column, select RADIOSS (radioss).
e. In the Solver input arguments column, enter -nproc 4 after ${file}.

impactor_0001.rad file.
e. Click Close.
6. Click Import Variables. Eight input variables are imported from the impactor.tpl
resource file.
8. Review the lower and upper bound ranges of the input variables.

2. Click Apply.


directory. The the approaches/nom_1/run 00001/m_1 sub-directory contains the
impactorT01 file, which is the result of the nominal run.
In this study, we want to analyze the maximum acceleration and the maximum displacement
observed by the box. This study is a function of the time; we need to extract the maximum of
each output response vector over time.
1. Create a file source for time.
a. From the Directory, drag-and-drop the impactorT01 file, located in
• Set Type to Time.
• Set Request to Time.
• Set Component to Time.

e. Clear the Linked to a new Response checkbox.

f. Click Finish

2. Create a second file source for impactor acceleration along the Z axis. Repeat step 1,
except define the following options:
• Set Type to Node/TH_node_sphere.
• Set Request to 4206 rigid_sphere_4206.
• Set Component to AZ-Z Acceleration.
3. Create a third file source for impactor displacement along the Z axis. Repeat step 1,
except define the following options:
• Set Type to Node/TH_node_sphere.
• Set Request to 4206 rigid_sphere_4206.
• Set Component to DZ-Z Displacement.
You have finished creating all of the result vectors for the Max_Acceleration output response.
As you can see from the graph on the left-hand side below, it has some noise. To eliminate
the noise, you will use a filter and work on the filtered output response as seen from the graph
on the right-hand side below.

4. Add two output responses.

a. Click Add Output Response twice.
b. In the work area, change the labels for the output responses to Max_Acceleration and
Max_Displacement.
5. Define the Max_Acceleration output response.
a. In the Expression column of the output response Max_Acceleration, click .
b. In the Expression Builder, click the Functions tab.
c. From the list of functions, select saefilter. This function will apply a filter to the
acceleration vector.
d. Click Insert Varname. The function saefilter(,,) appears in the Evaluate
Expression field.
You can now add the time vector and the acceleration vector as arguments to the
function, with a class parameter of 180.
e. In the Evaluate Expression field, enter (m_1_ds_1,m_1_ds_2,180) in the
saefilter function.
f. To calculate the max of the expression, add the max function to the beginning of the
expression.
The expression should read: max(saefilter(m_1_ds_1,m_1_ds_2,180)).
g. To express the result in G, divide the max of the expression by 9810.
The expression should read: max(saefilter(m_1_ds_1,m_1_ds_2,180)/9810).

h. Click OK.
6. Define the Max_Displacement output response.
a. Optional. Plot the displacement with respect to the time, to obtain the curve
illustrated below:
b. In the Expression column of the output response Max_Displacement, click .

c. In the Expression Builder, click the Functions tab.
d. From the list of functions, select abs.
e. Click Insert Varname. The function abs()appears in the Evaluate Expression
field.
f. From the list of functions, select min.
g. Click Insert Varname. The expression should now read, abs(min()).
h. In the Evaluate Expression field, enter m_1_ds_3 in the min function.
i. Click OK.
7. Click Evaluate to extract the output response values of each expression.

Step 6: Run a Screening DOE Study
The model has 8 variables which may lead to high computation times for direct optimization or
even for creating a response surface.
You will reduce the number of actual input variables by running a screening experiment. A full
factorial experiment with 8 factors at 2 levels will require 28 (256) runs and with 3 levels, it
will increase to 6561 runs. You will try to screen out some input variables by first doing a
Fractional Factorial screening DOE.
5. In the Settings tab, set Resolution to IV.
Note: Resolution IV enables an estimate of main effects unconfounded by two-factor
interactions. It also enables an estimate of two-factor interaction effects, which
may be confounded with other two-factor interactions.
6. Verify that the Number of runs is set to 16.

7. Click Apply.
9. Click Evaluate Tasks to execute the run matrix and extract the output responses for all
of the runs. HyperStudy runs 16 simulations in Fractional Factorial mode, therefore the
evaluation will take some time.

Step 7: Post Process the Screening DOE Study
1. Click the Linear Effects tab to review the linear effects. Observe the main effect of the
input variables on both output responses.
2. Click the Pareto Plot tab, then use the Channel selector to select both of the output
responses. Observe that results.
Note: A linear effects plot and a pareto plot with the Linear effects option enabled
(shown below) provide the same information. However, with a pareto plot, you
can use a statistical measure (that is, the 80-20 rule) to decide which input
variables are more significant and which input variables can be neglected.

For this tutorial, you will use the 80/20 rule to eliminate input variables that are not
significant to the study. The 80/20 rule is a Pareto principle that proposes 80% of the
total effects comes from only 20% of the variables.
Note: You should also use other practices to eliminate input variables that you feel
should be taken in consideration.
For screening purpose, you can see which input variables contribute to 80% or more of
the given output response. In the images below you can see that:
• For Max_Acceleration, the input variables length_internal, th_internal_skin, and
th_external_skin contribute to 80% of the linear effect.
• For Max_Displacement, the input variables length_internal and th_internal_skin fall
under the 80/20 rule.
For n responses, you can list out the input variables that follow the 80/20 rule, and take
union of the sets. In this case, the input variables that follow the 80/20 rule include:
length_internal, th_internal_skin, and th_external_skin. This narrows your list to three
significant input variables.
Step 8: Run a DOE Study for Approximation
Since this optimization is based on response surfaces, a central composite experiment will be
used, which will create a 2nd order response surface.
4. In the Active column, keep only the three significant input variables (established in step
7) selected. Clear the corresponding checkboxes for all other input variables.


7. In the work area, set the Mode to Central Composite.
8. Click Apply.
10. Click Evaluate Tasks to execute the run matrix and to extract the output responses.
Step 9: Run a DOE Study for the Validation Matrix
Other points will be used to check the quality of the approximation. The points will be defined
by a new DOE. In this DOE study, a Latin Hypercube of 10 runs will be used.
1. Add a third Doe to the study by repeating Step 8: Run a DOE Study for
Approximation.
a. In the Specifications step, set the Mode to Latin HyperCube.
b. In the Settings tab, change the Number of runs to 10.
Step 10: Create an Approximation
4. Click Add Matrix twice.
5. Define FitMatrix1 and FitMatrix2 by selecting the options indicated in the image below.


9. Review the input variables and output responses. Only the length_internal,
th_internal_skin, and th_external_skin input variables should be active.
11. In the work area, set the Mode to Moving Least Squares (MLSM).
12. In the Settings tab, verify that Regression Model is set to Linear (default).
Note: It is advisable to start with lowest order and increment it in case model Residuals
and Diagnostics do not look feasible.
13. Keep all other parameters to default as well.
14. Click Apply.
18. To assess the accuracy of the regression equations, click the Residuals and
Diagnostics tab.
19. To review the output response curves and surfaces, click the Trade-Off tabs.
In the Trade-Off 3D tab, use the Channel selector to plot input variables and output
responses. The values for the input variables which are not plotted are modified in the top
frame (Inputs). Move the sliders in the Value column to modify the other input variables,
while studying the output response throughout the design space.
4. In the Active column, clear the checkboxes for all input variables except length_internal,
th_internal_skin and th_external_skin.
6. Apply an objective on the Max Acceleration output response.
a. In the Objectives column for Max Acceleration, click .
c. Set Evaluate From to Max_Acceleration MLSM (r_1_fit_1).

7. Apply a constraint on the Max Displacement output response.

a. In the Constraints column for Max Displacement, click .
c. Set Evaluate From to Max_Displacement MLSM (r_2_fit_1).
8. Click Apply.
10. In the work area, set the Mode to Genetic Algorithm (GA).
11. In the Settings tab, change the Constraint violation tol. to 0.0.
Note: Optimization algorithms are usually solved so that a small violation in the
constraint is acceptable in order to save time in resolving the variables’
precision beyond a physical significance. Setting the Constraint violation tol. to
zero will force HyperStudy to satisfy the constraint identically, which can increase
the run time of the optimization in some cases.
12. Click Apply.
14. Click Evaluate Tasks to optimize the design that minimizes the maximum acceleration
while keeping the displacement of node 35527 smaller than 16.
15. Click the Iteration Plot tab to monitor the Optimization iteration.

17. Click the Iteration History tab to review a table of each iteration. The iterations that do
not respect the constraint are displayed red, the optimal design is displayed green.
Step 12: Evaluate the Results of the Approximate Optimal Design
You have now determined the best design evaluated from the approximation. You can launch
a solver run with the following input variables to check if the solution found by the
approximation is close to the solver results.
1. In the Explorer, study Setup, go to the Define Input Variables step.
2. In the work area, change the Nominal values to the values defined in the table below:
Variable Name Optimum Value
th_external_skin 1.9216688
th_internal_skin 1.9999998
th_external_flange 1.00
th_internal_flange 1.00
Radius 0.00
length_external 0.00
width_external 0.00
length_internal -0.9999824


7. Compare the optimum solution evaluated from the approximation to the same design
evaluated by the solver.
Note: Due to the use of different solver versions, results may vary. In following
comparison, the solver and approximation are in very good agreement with < 1%
error.
Output Responses Approximation Solver
Max_Acceleration 12.246401 12.293556
Max_Displacement 15.999990 16.046738
Last modified: v2017.2 (12.1156684)

HS-4425: Multi-Objective Shape Optimization Study
Continuing from tutorial HS-4000: Optimization Method Comparison: Arm Model Shape
Optimization, you will perform a multi-objective Optimization study.
In this tutorial, you will be searching for the Pareto front that minimizes both volume and
maximum displacement. You will be using MOGA with a Fit to save time.
Note: If a Fit was not available, GRSM would be the suggested method to use in order to
solve a MOO problem. MOO problems require many evaluations, therefore GRSM is
more efficient than MOGA.
Before running this tutorial, complete tutorial Tutorial HS-4000: Optimization Method
Comparison: Arm Model Shape Optimization. You can also import the archive file HS-
4. In the work area, Active column, clear the radius_1, radius_2 and radius_3 check
boxes.
6. Apply an objective on the Volume and Max_Disp output responses.
a. In the Objectives column, click .
7. In the Evaluate From column for Max_Disp, Max_Stress, and Volume, select Fit, RBF
(fit_4).
8. Click Apply.
10. In the work area, set the Mode to Multi-Objective Genetic Algorithm (MOGA).
11. Note: Only the methods that are valid for the problem formulation are enabled.
12. Click Apply.
14. Click Evaluate Tasks. HyperStudy stops MOGA after 50 iterations, and performs a total
of 13317 analyses. The Pareto front of the last iteration contains 408 points.
The Pareto front of Objective 2 versus Objective 1 is displayed in the plot.

The goal of this study was to minimize both Volume (Objective 1) and Max_Disp
(Objective 2). The Pareto plot shows all of the non-dominated solutions. A non- dominated
solution is a solution which can no longer improve one objective without deteriorating
another. You can see that minimizing Objective 1 will increase Objective 2, and
minimizing Objective 2 will increase Objective 1. According to these results, you must
decide what would be the optimal solution. For instance, the Pareto plot may allow a
compromise solution to be selected somewhere in the middle.
Last modified: v2017.2 (12.1156684)

HS-4230: Optimization Study with Discrete Variables
In this tutorial you will learn how to use discrete variables. The sample base input template
plate.tpl can be found in <hst.zip>/HS-4230/and copied to your working directory.
The objective of this tutorial is to maximize the minimum frequency of the first five modes of a
plate. The input variables are the thickness of each of the three components, defined in the
thickness within the files is 0.1. The optimization type is size. Furthermore, optimum design
should have input variables from a discrete set of 0.05, 0.08, 0.11, and 0.14 for all three
thicknesses. By default, HyperStudy will add the values from the lower and upper bounds to
this set. Hence the resulting set is 0.05, 0.08, 0.11, 0.14, and 0.15. Delete any of these
values if needed.
.
click OK.

6. Click Import Variables. Three input variables are imported from the plate.tpl file.
9. In the Mode column of all three input variables, select Discrete.
10. In the Values column of Property 21, click .

11. Click Step Size, enter 0.03, and click Set.

13. Click Apply.

14. Click OK.
15. Repeat steps 10 through 13 for Property 22 and Property 23.

2. Click Apply.
directory. The approaches/nom_1/run 00001/m_1 sub-directory contains the plate.out
file, which is the result of the nominal run, and will be used in the Optimization.

1. Create the Freq1 output response.


e. Label the output response Freq1.

g. Click Finish. The Freq1 output response is added to the work area.
b. Set Apply On to Volume (r_2).


a. Set Apply On to Freq1 (r_1).
b. Set Bound Type to >= (greater than or equal to).
c. For Bound Value, enter 32.
13. Click Apply.

16. Click Apply.
1. Click the Iteration History tab to monitor the progress of the Optimization iteration.

Step 6: Setup a DOE to Find the True Best Design
5. Click Apply.
10. Sort run data based on the Volume (which was to be minimized) by right-clicking on the
Volume column and selecting Sort down from the context menu. The lowest volume
design which satisfies the constraint (frequency > 32) is the same as that found by the
optimizer.
Note: The DOE took 125 solver calls to exhaust all combinations, whereas the
Optimization found it in 8 solver calls.

HS-4450: Multi-Objective Optimization of a Cantilever Ibeam

using an Inclusion Matrix
The inclusion matrix feature passes an already existing set of data to the running process. In
this tutorial, the data created from a DOE is passed to an optimization problem which re- uses
the data. This promotes efficient design exploration practices: an optimization using a direct
solver call can still be done in combination with a DOE to study the system without any loss of
data. This example focuses on the competing objectives in the design of a cantilever ibeam.
click OK.
a. From the Directory, drag-and-drop the ibeam.tpl file into the work area.
b. In the Solver input file column, enter ibeam.py. This is the name of the solver input
c. In the Solver execution script column, select Python (py).
6. Click Import Variables. Four input variables are imported from the ibeam.tpl
resource file.

9. Review the input variable's lower and upper bounds ranges.


2. Click Apply.
output.hstp file, which is the result of the nominal run, and will be used during the
Optimization.
1. Create the Iy output response for the y-axis moment of inertia.

a. From the Directory, drag-and-drop the output.hstp file, located in
• Set Request to Iy.

e. Label the output response Iy.


g. Click Finish. The Iy output response is added to the work area.

2. Create four more output responses by repeating step 1, except change the Request
assigned to each output response to the following:
Output Response Request
Volume Vol
IZ Iz
Displacement d
Frequency1 Freq1
3. Click Evaluate to extract output response values.
Step 4: Run a Hammersley DOE Study
5. In the Settings tab, verify that the Number of runs is 17.
6. Click Apply.
10. Click the Pareto Plot tab.
Enable multi-plot and select all of the output responses from the Channel selector. In
the options menu, ensure that Linear effects is enabled.
A Pareto Plot shows the ranked influence of the input variables on the output response.
For example, for the y-axis moment of the inertia, height has the largest influence and
web thickness has the least. In contrast, for the z-axis moment of inertia, the flange
length and flange thickness are the most influential variables. The size of the bar

indicates the magnitude of the influence, and the hashed line’s slope indicates the sign of
the effect: positive or negative. For example, increasing the height will increase Iy, but it
will decrease displacement.

7. Apply an objective on the Iy output response.

a. In the Objectives column for Iy, click .
b. In the pop-up window, set Type to Maximize and click OK.
8. Click Apply.
11. Click the More tab and define the following settings:
• Set Points per Iteration to 4.
Set Use Inclusion Matrix to Without Initial.
GRSM performs a global search, therefore the initial values of the variables are not
important and do not have to be used within the optimization.
12. Import run data from the DOE using an Inclusion Matrix.
b. In the Edit Inclusion Matrix dialog, click Import Values.

c. In the Import Values dialog, select Approach evaluation data and click Next.

e. Set the approach to DOE 1 and click Next.
f. Click Finish.
g. Review the imported run data and click OK.
13. Click Apply.
Observe the non-dominated front of designs. These points represent the trade-off
between the objective of minimizing volume and maximizing the y-axis moment of inertia.
In the plot, it is evident that as the moment of inertia increases, the volume increases as
well. This curve represents the trade-off of the best available designs given the competing
objective requirements.
18. Click the Scatter tab to plot the objectives along the same axes shown in the Optima
plot.
This scatter plot shows all of the runs from the optimization. When comparing the scatter
and optima plots, note that the optima plot contains only a subset of runs which

are non-dominated. A dominated design is a design for which both objectives could be
improved. A non-dominated design is one in which one objective may only be improved at
the expense of another.
Last modified: v2017.2 (9.1144596)

Stochastic

HS-5000: Stochastic Method Comparison and Reliability-

Based Design Optimization of the Arm Model
Continuing from tutorial HS-4000: Optimization Method Comparison: Arm Model Shape
Optimization, you will perform Stochastic studies and a Reliability-Based Design Optimization
(RBDO) using the same fitting function. You will run a Stochastic study around the nominal
point.
Before running this tutorial, complete tutorial Tutorial HS-4000: Optimization Method
Comparison: Arm Model Shape Optimization. You can also import the archive file HS-
In the Stochastic study, you will be using a Hammersley distribution successively with 100,
300, and 1000 runs in order to compare the convergence of statistical results. You defined all
six input variables as random variables following a normal distribution.
In this step, you will check the robustness of the optimal solution found with GRSM.
4. In the Active column, clear the radius_1, radius_2 and radius_3 check boxes.
5. In the Nominal column, copy the parameter values at the optimal design.
For Stochastic studies, you must provide data about the standard variation σ (or variance
σ²) of parameters in order to take into account uncertainties. By default, σ² is computed
in HyperStudy using the range rule σ² = ((Upper Bound-Lower Bound)/4)² which is a
function of the input variable's bounds. If you do not have reliable data about the
standard deviation, the default σ² can be modified by changing the upper and lower
bounds of the parameters.
a. Go to the GRSM, 6 IV, Exact Solver Optimization, Evaluate step, Iteration History
tab, and copy the length_1, length_2, length_3, length_4, and length_5 parameter
values.

c. Go to the Stochastic study, Select Input Variables step, and select the Nominal
fields for length_1, length_2, length_3, length_4, and length_5. Right-click on the
selected fields, and select Paste transpose fro the context menu.
d. For height, change the Nominal value to 1.0.

6. Change the lower and upper bounds for every active input variable to match the
dimensions’ tolerances. For this tutorial, assume that the tolerances are within 0.05 mm.

a. For all active input variables, click in the Nominal field.

b. In the pop-up window, Value field, enter 0.05 and click +/-.
c. Click OK to accept the changes and close the pop-up window.

The Lower and Upper bounds for all active input variables should replicate the image
below.
8. Click the Distributions tab, and verify that Distribution is set to Normal Variance
for all active input variables.
Look at the columns A and B. Column A displays the nominal parameter values, and
column B displays the variance σ² which is computed using the standard deviation σ of a
parameter around its mean.
Variance is computed as follows: σ² = ((Upper Bound-Lower Bound)/4)². For instance,
the variance for the input variable length_1 is σ²=((-0.45+0.55)/4)²=6.25e-4.


10. Define Max_Disp, Volume and Max_Stress by selecting the options indicated in the image
below from the Evaluate From column.
Even though Max_stress was not used in the Optimization, you will use it in the Stochastic
study to check the reliability.

Hammersley sampling belongs to the category of quasi-Monte Carlo methods. This
method uses a quasi-random number generator, based on the Hammersley points, to
uniformly sample a unit hypercube.
13. Click Apply.
15. Click Evaluate Tasks to execute all the runs and extract the results.
16. To study the method convergence:
a. Go back to the Specifications step and change the Number of runs to 300 and
then 1000 in the Settings tab.
b. Go to the Evaluate step, and click Evaluate Tasks.

Step 2: Post Processing the Results of the Stochastic Study
1. Click the Integrity tab to review the statistics of the input variables and the output
responses around the nominal design.
Using the Channel selector, select the Health category to get a summarized view of
statistics and spot eventual, missing, or bad values. The Range column can be useful to
understand the spread of values in the data from the minimum to the maximum.
2. Click the Distribution tab to review a histogram of the Stochastic results. Using the
Channel selector, select Length_1.
The chart shows three pieces of information about the distribution of values for the
selected input variable. The Histogram uses the left axis, and represents the frequency
of runs yielding a sub-range of response values. The Probability density uses the right
axis, and indicates the relative likelihood of the input variable to take a particular value. A
higher value indicates that the values are more probable to occur. The Cumulative
distribution is another curve that uses the right axis. It is equal to the integral of the
Probability density. The value of the Cumulative distribution indicates what percentage of
the data falls below the value’s threshold. Note that the initial value of the Cumulative
distribution will always equal 0, and the final value of the CDF will always be 1.0. This is
because all of the data will reside between the upper and lower bounds.
3. Review the Probability density and the Cumulative distribution of Max_Disp. Compare the
distributions obtained with the different number of runs for Hammersley (100, 300, and
1000).
Note that the pattern of the distribution changes quite a bit from 100 to 300 runs, but
very little from 300 to 1000 runs.

Hammersley, 100 runs

4. Click the Integrity tab to review the statistics of the output responses around the
nominal design.
The column range can be useful to understand the spread of values in the data from the
minimum to the maximum.
5. Add a reliability.
a. Click the Reliability tab to estimate the Probability of failure for the output responses
(probability for an output response to violate a user selected bound).
b. Click Add Reliability.
c. Define the reliability. The Bound Values are those from the Optimization problem
definition (Max_Disp<=1.5).
• Set Response to Max_Disp(m_1_r_1).
The results show the convergence of the Hammersley method in function of the
number of runs. The convergence is enhanced by increasing the number of runs. As

the results obtained with 300 or 1000 runs are quite similar, you can assume that in
this case 300 runs are enough to get accurate results.
d. Study the effects of bounds on the reliability by entering different values in the
Bound Value columns.
e. Optional: Check the reliability on Max_Stress.
6. Click the Reliability Plot tab to see the relationship between the desired threshold and
the reliability of the system. The values reported in the Reliability column of the
Reliability tab can be observed on these reliability curves. On the Max_Disp reliability
curve, only 11.4% of the designs have a value above 1.5, which means 88.6% are below
1.5. A reliability of 95% can be reached by changing the threshold for Max_Disp to
1.5034.

Step 3: Reliability-Based Design Optimization Study (RBDO)
In this step, you will be searching for 95% reliability on the Optimization constraint (max_disp
< 1.5 mm). You will use fitting functions as opposed to the exact solver to evaluate the output
responses. Among the approximations, you will use the RBF that is created with MELS DOE. As
a result, you will use the SORA method. You will continue using the six important input
variables that all follow a normal distribution. You will start the Optimization from the
previously obtained optimum solution with GRSM.
4. In the Active column, clear the radius_1, radius_2 and radius_3 check boxes.
5. In the Nominal column, copy the parameter values at the optimal design.
a. Go to the GRSM, 6 IV, Exact Solver Optimization, Evaluate step, Iteration History
tab, and copy the length_1, length_2, length_3, length_4, and length_5 parameter
values.
b. Go to the Select Input Variables step of the Optimization study you just created,
and select the Nominal fields for length_1, length_2, length_3, length_4, and
length_5. Right-click on the selected fields, and select Paste transpose fro the
context menu.
c. For height, change the Nominal value to 1.0.
6. Click the Distributions tab, and set the Distribution Role to Design with Random
for all active input variables.
7. In the B column, enter 6.25e-04 for all active input variables.

9. Go to the Select Output Responses tab.

c. In the Evaluate From column for Volume, select Fit, RBF (fit_4).
11. Apply an objective on the Max_Disp output response.

a. In the Constraints column for Max_Disp, click .
• Set Type to Random.
• For CFD Limit, enter 95.0.
c. In the Evaluate From column for Max_Disp, select Fit, RBF (fit_4).
12. In the Evaluate From column for Max_Stress, select Fit, RBF (fit_4).
13. Click Apply.
15. In the work area, set the Mode to Sequential Optimization and Reliability
Assessment (SORA).

17. Click Apply.

20. Click the Iteration History tab to review the results of the Optimization in a table.
A deterministic optimum is found first. As you can see in iteration 1, the displacement is
at the constraint bound. This design is not 95% reliable as required in this probabilistic
Optimization study. The SORA will work to make sure the constraint satisfies the
probabilistic requirement. As seen in iteration 2, the design that meets 95% reliability is
the one that has a max displacement shifted away from the bound of 1.5.
Corresponding to this improvement, there is an increase in the objective value (volume).
21. As SORA uses the approximation method for the reliability calculation, it is recommended
that you make a validation Stochastic study after the Optimization.
a. In the Explorer, right-click on the Hammersley Stochastic study and select Copy
Approach from the context menu.
b. In the Copy dialog, Label field, enter Validation SORA and click OK.
22. Copy and paste the optimum solution from the SORA Optimization into the Nominal
column of the input variables for the Validation SORA Stochastic study.
a. Go to the SORA Optimization, Evaluate step, Iteration History tab, and copy the
length_1, length_2, length_3, length_4, and length_5 parameter values.
b. Go to the Select Input Variables step of the Validation SORA Stochastic study you
just created, and select the Nominal fields for length_1, length_2, length_3,
length_4, and length_5. Right-click on the selected fields, and select Paste
transpose fro the context menu.
23. Change the lower and upper bounds for every active input variable to match the
dimensions’ tolerances. For this tutorial, assume that the tolerances are within 0.05 mm.
a. For all active input variables, click in the Nominal field.
b. In the pop-up window, Value field, enter 0.05 and click +/-.
c. Click OK to accept the changes and close the pop-up window.
25. For Max_Stress, verify that Evaluate From is set to Fit, RBF (fit_4).
26. Go to the Specifications step and verify that the Mode is set to Hammersley with 1000
runs.
27. Go to the Evaluate step, and click Evaluate Tasks.
29. Add a reliability on Max_Disp.

a. Click the Reliability tab.

b. Click Add Reliability.
c. Define the reliability.
• Set Response to Max_Disp.
The result confirms that the optimum solution found with SORA is reliable at 95%.
Last modified: v2017.2 (12.1156684)

Hyperstudy 2019 Tutorials

Diunggah oleh

Informasi Dokumen

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Hyperstudy 2019 Tutorials

Diunggah oleh

Hak Cipta:

Format Tersedia

Altair HyperStudy 2019

Altair HyperStudy 2019 Copyright 1999-2019

The Platform for Innovation™

Altair Engineering Inc. Copyright © 1986-2019. All Rights Reserved.

Altair HyperWorks™ - The Platform for Innovation™

Altair AcuSolve™ ©1997-2019

Altair ElectroFlo™ ©1992-2019

Altair ESAComp™ ©1992-2019

Altair Flux™ ©1983-2019

Altair FluxMotor™ ©2017-2019

Altair HyperCrash™ ©2001-2019

Altair HyperGraph™ ©1995-2019

Altair HyperLife™ ©1990-2019

Altair HyperMesh™ ©1990-2019

Altair HyperStudy™ ©1999-2019

Altair HyperView™ ©1999-2019

Altair Virtual Wind Tunnel™ ©2012-2019

Altair HyperXtrude™ ©1999-2019

Altair MotionSolve™ ©2002-2019

Altair MotionView™ ©1993-2019

Altair Multiscale Designer™ ©2011-2019

Altair OptiStruct™ ©1996-2019

Altair Radioss™ ©1986-2019

Altair SimLab™ ©2004-2019

Altair SimSolid™ ©2015-2019

Altair nanoFluidX™ ©2013-2018 Fluidyna GmbH, © 2018-2019 Altair Engineering Inc.

Altair ultraFluidX™ ©2010-2018 Fluidyna GmbH, © 2018-2019 Altair Engineering Inc.

Altair ConnectMe™ ©2014-2019;

Plus other products from the Altair solidThinking Platform.

Altair Packaged Solution Offerings (PSOs)

Altair GeoMechanics Director™ ©2011-2019

Altair Impact Simulation Director™ ©2010-2019

Altair Model Mesher Director™ ©2010-2019

Altair NVH Director™ ©2010-2019

Altair Squeak and Rattle Director™ ©2012-2019

Altair Virtual Gauge Director™ ©2012-2019

Altair Weight Analytics™ ©2013-2019

Altair Weld Certification Director™ ©2014-2019

Altair Multi-Disciplinary Optimization Director™ ©2012-2019

Altair solidThinking - Where Innovation Begins™

Altair Inspire™ Extrude-Metal ©1996-2019 (formerly Click2Extrude®-Metal)

Altair Inspire™ Cast ©2011-2019 (formerly Click2Cast®)

Altair Inspire™ Form ©1998-2019 (formerly Click2Form®)

Altair Inspire™ Mold ©2009-2019 (initial release-Q2 2019)

Altair Inspire™ Studio ©1993-2019 (formerly ‘Evolve’)

Altair Compose™ ©2007-2019 (formerly solidThinking Compose®)

Altair Activate™ ©1989-2019 (formerly solidThinking Activate®)

Altair Embed™ ©1989-2019 (formerly solidThinking Embed®)

Altair SimLab™ ©2004-2019

Altair 365™ ©1994-2019

Altair PBSWorks™ - Accelerating Innovation in the Cloud™

Altair Control™ ©2008-2019; (formerly PBS Control)

Altair Access™ ©2008-2019; (formerly PBS Access)

Proprietary Information of Altair Engineering

Altair Accelerator Plus™©1995-2019; (formerly WorkloadXelerator)

Altair FlowTracer™ ©1995-2019; (formerly FlowTracer)

Altair Allocator™ ©1995-2019; (formerly LicenseAllocator)

Altair Monitor™ ©1995-2019; (formerly LicenseMonitor)

Altair Hero™ ©1995-2019; (formerly HERO)

Altair Software Asset Optimization™ (SAO) ©2007-2019

Display Manager™ ©2013-2017 is now part of Altair Access

PBS Application Services™ ©2008-2017 is now part of Altair Access

e-Compute™ ©2000-2010 was replaced by “Compute Manager” which is now Altair

Altair SmartWorks™ - Innovation Intelligence®