Additive User's Guide (Print and Science)

ANSYS, Inc. Release 2021 R1

Southpointe January 2021
2600 ANSYS Drive
Canonsburg, PA 15317 ANSYS, Inc. and
[email protected] ANSYS Europe,
Ltd. are UL
http://www.ansys.com registered ISO
(T) 724-746-3304 9001: 2015
(F) 724-514-9494 companies.
Copyright and Trademark Information

© 2021 ANSYS, Inc. Unauthorized use, distribution or duplication is prohibited.

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

Disclaimer Notice

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

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

U.S. Government Rights

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

Third-Party Software

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

Published in the U.S.A.

Table of Contents
Welcome to ANSYS Additive ....................................................................................................................... 7
Why Use ANSYS Additive? ....................................................................................................................... 7
The Simulation Process ........................................................................................................................... 8
Calibration ....................................................................................................................................... 9
Verification ....................................................................................................................................... 9
Understanding the Additive Interface ..................................................................................................... 11
Managing Your Simulations and Data .................................................................................................... 12
Where is My Data Stored? ............................................................................................................... 14
Increasing Memory Allocation ......................................................................................................... 17
A Note About Data Noise with Parallel Processing ............................................................................ 18
Application Settings ............................................................................................................................. 19
Choosing a Simulation Type ..................................................................................................................... 21
Assumed Strain - Uniform Isotropic ....................................................................................................... 22
Scan Pattern Strain - Anisotropic ........................................................................................................... 23
Thermal Strain - Anisotropic .................................................................................................................. 23
Theoretical Overview ............................................................................................................................... 25
Overview of the AP Mechanics Solver .................................................................................................... 25
Overview of the Thermal Solver ............................................................................................................ 29
References ...................................................................................................................................... 32
Understanding Machine Parameters ....................................................................................................... 33
Additive Print Simulations ....................................................................................................................... 35
Prepare and Import a Part .................................................................................................................. 37
Guidelines for Part Orientation and Resolution ................................................................................ 37
Importing a Part ............................................................................................................................. 38
The Simulation Domain ............................................................................................................ 39
Editing and Exporting a Part ..................................................................................................... 39
Importing Supports ........................................................................................................................ 40
Importing a Build File ...................................................................................................................... 43
Using Additive Industries Build Files .......................................................................................... 46
Using EOS Build Files ................................................................................................................ 46
Using HB3D Build Files .............................................................................................................. 47
Using Renishaw Build Files ........................................................................................................ 48
Using Sisma Build Files .............................................................................................................. 48
Using SLM Solutions Build Files ................................................................................................. 48
Using Trumpf Build Files ............................................................................................................ 48
Deleting Parts, Supports, and Build Files .......................................................................................... 48
Set Up a Simulation - Assumed Strain ................................................................................................ 51
Set Details ...................................................................................................................................... 53
Select Geometry ............................................................................................................................. 53
Voxel Size and Voxel Sample Rate .............................................................................................. 54
Voxelization at Part and Support Interfaces—Handling Intrusion (Penetration) .......................... 58
Hanging/Floating Voxels ........................................................................................................... 59
Known Issues and Limitations with Voxelization ........................................................................ 60
Define Support Options .................................................................................................................. 61
Using Automatic Supports ........................................................................................................ 63
Using Support .stl Files .............................................................................................................. 67
Using Support Groups .............................................................................................................. 68
Using Supports in Build Files ..................................................................................................... 69
Simulating Without Supports .................................................................................................... 69

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. iii
Additive User's Guide (Print and Science)

Select Material ................................................................................................................................ 70

Linear Elastic versus J2-Plasticity Stress Mode ............................................................................ 71
Strain Scaling Factor ................................................................................................................. 73
Customizing a Material ............................................................................................................. 74
Select Outputs ................................................................................................................................ 76
On-Plate Residual Stress/Distortion ........................................................................................... 77
Strain Components ............................................................................................................. 77
Distortion Compensated .stl File .......................................................................................... 77
Displacement After Cutoff ......................................................................................................... 78
Layer by Layer Stress/Distortion ................................................................................................ 82
Files for Transfer to ANSYS Mechanical ....................................................................................... 83
Blade Crash Detection .............................................................................................................. 83
High Strain Areas ...................................................................................................................... 84
Save/Export a Draft Simulation ....................................................................................................... 85
Run a Simulation - Assumed Strain .................................................................................................... 87
How Long Will My Simulation Take? ................................................................................................ 87
Common Error Messages During a Simulation ................................................................................. 88
Can I Run Multiple Simulations at Once? ......................................................................................... 89
Review Results of a Simulation - Assumed Strain .............................................................................. 91
Output Files .................................................................................................................................... 92
Using ANSYS Viewer to Review Results ............................................................................................ 95
Performing a Scan Pattern Simulation ............................................................................................... 99
Anisotropic Strain Coefficients ........................................................................................................ 99
Machine Parameters for a Scan Pattern Simulation ......................................................................... 100
Performing a Thermal Strain Simulation .......................................................................................... 103
Mesh Resolution Factor ................................................................................................................. 103
Select Material for a Thermal Strain Simulation .............................................................................. 104
Machine Parameters for a Thermal Strain Simulation ...................................................................... 104
Thermal Solver Assumptions and Limitations that Affect Results at Part Edges ............................... 106
Additive Science Simulations ................................................................................................................. 109
Performing a Single Bead Parametric Simulation ............................................................................ 111
Machine Configuration ................................................................................................................. 112
Geometry Configuration ............................................................................................................... 114
Reviewing Results ......................................................................................................................... 114
Performing a Porosity Parametric Simulation .................................................................................. 121
Machine Configuration ................................................................................................................. 123
Geometry Configuration ............................................................................................................... 125
Reviewing Results ......................................................................................................................... 125
Guidelines for Interpreting Porosity Results ................................................................................... 126
Performing a Thermal History Simulation ........................................................................................ 129
Performing a Microstructure Simulation ......................................................................................... 131
Using Microstructure Data to Evaluate Mechanical Properties ........................................................ 131
Material Configuration .................................................................................................................. 132
Microstructure Configuration ........................................................................................................ 132
Machine Configuration ................................................................................................................. 135
Reviewing Results ......................................................................................................................... 135
Microstructure Validation Process .................................................................................................. 146
Known Issues and Limitations ....................................................................................................... 151
References .................................................................................................................................... 153
Creating User Defined Materials ............................................................................................................ 155
Importing a User Defined Material ...................................................................................................... 155

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
iv of ANSYS, Inc. and its subsidiaries and affiliates.
 Additive User's Guide (Print and Science)

User Defined Material File Specifications ............................................................................................. 158

Material Configuration File ............................................................................................................ 158
Material Lookup Table ................................................................................................................... 160
Material Characteristic Width Lookup Table ................................................................................... 161
Glossary .................................................................................................................................................. 163

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. v
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
vi of ANSYS, Inc. and its subsidiaries and affiliates.
Welcome to ANSYS Additive
ANSYS Additive is simulation software dedicated to the field of metal additive manufacturing. Offerings
include:

• Additive Print – a tool for machine operators to perform quick simulations of parts to ensure
they will print successfully (included with Additive Print and Additive Suite licenses).

• Additive Science – an exploratory environment for scientists to determine the optimum process
parameters for their machines and materials (included with Additive Suite licenses).

This user's guide is intended for all users of ANSYS Additive. Explanations of capabilities available only
in Additive Science are clearly identified.

Platforms and Languages Supported

The Additive application is supported on the Windows 10 operating system.

Additive is compatible with computers set up to operate in the following languages:

• English

• French

• German

• Japanese

See the ANSYS, Inc. Installation Guide for Windows for information about installing the application.

Known Limitations
• Linux is not supported.

• Some system ports the Additive application uses are hard-coded and cannot be changed. See Install-
ation Troubleshooting for details.

Why Use ANSYS Additive?

Companies today are designing products that are sophisticated, complex, and finely-tuned to operate
in their working environments. With the use of CAE and FEA tools, designers are aggressively reducing
their product’s time-to-market, cost, and material consumption. ANSYS Additive harnesses the next
level of design improvement by providing tools to simulate a part’s behavior during the manufacturing
process for those using the latest techniques in additive manufacturing (3D printing). Already shown
to be a revolutionary technology with immense advantages over traditional manufacturing methods,

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 7
Welcome to ANSYS Additive

additive manufacturing brings unique challenges as well as opportunities for even more time and cost
savings.

The Additive application simulates the layer-by-layer build process of metal parts undergoing Laser
Powder Bed Fusion (LPBF), a type of additive manufacturing that uses a laser to melt or fuse material
powder together. As each layer is processed, the region under the laser experiences extremely intense,
local heat that cools rapidly and results in thermal distortion. The simulation method uses a layer by
layer accumulation of inherent strain to predict this distortion. As a user, you will gain critical insight
into the complex physics-based phenomena associated with this layering process.

Simulating the build process may be performed at various points in the overall design/manufacturing
process depending on your goals. Additive generates practical solutions to residual stress, distortion,
and build failure, enabling you to:

• Improve Product Design – In complicated, asymmetrical parts, shrinkage and distortion due to
rapid heating and cooling during the 3D printing process may result in components outside of
tolerances. Simulations of the build process show magnitudes and locations of part distortion.
Designers can quickly make design changes to assure part conformance without iterations of
trial and error builds.

• Inform Build Preparation – Instead of building several part prototypes in different orientations
on the build plate, simulations of these orientations reveal best orientation in a fraction of the
time and expense.

• Validate Build Preparation – Parts designed using powerful topology optimization tools result
in complex and intricate shapes that present difficult challenges on where to place supports re-
quired for build. Simulations in Additive include the generation of optimized support structures
using predicted residual stress accumulation as criteria for support placement and thickness.
Engineers can use the optimized support information in their production builds to reduce build
failures due to insufficient supports.

Simulations accurately predict part distortion during the build. Additive’s Distortion Compensation
feature takes that information a step further and automatically creates a distortion compensated geo-
metry, essentially reversing distortion effects. Engineers can use the compensated geometry file in their
production builds and be assured of a final part that conforms to design intent.

The Simulation Process

A simulation in Additive consists of four steps:

1. Prepare and Import a Part

2. Set Up a Simulation

3. Run a Simulation

4. Review Results of a Simulation

Depending on your simulation goals, you may need to run multiple iterations of this four-step process.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
8 of ANSYS, Inc. and its subsidiaries and affiliates.
 The Simulation Process

Calibration
Before beginning a simulation for the first time, you should run a series of calibrations to determine
input factors that take unique aspects of your machine and material into consideration. The calibration
procedures and parts are available online here. Note that the calibration procedure may continue to
evolve as we work with more machine partners and key customers.

Verification
Verification test cases for the Additive application are included in the ANSYS Verification Manual here.
These test cases compare simulation results with measurements taken of actual 3D printed parts.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 9
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
10 of ANSYS, Inc. and its subsidiaries and affiliates.
Understanding the Additive Interface
The Additive interface is quite simple and straightforward. It includes a simulation dashboard in the
main portion of the window and resource libraries on the left. Clicking on the ANSYS logo ( )
from anywhere in the program takes you back to the simulation dashboard.

Resource libraries are repositories for parts, build files, and materials. The first step in the simulation
process, prepare and import a part, involves importing a part into the Parts Library (or the Build File
Library if you have a build file).

The dashboard shows your most recent simulations organized from left to right as Draft Simulations,
Running Simulations, and Completed Simulations. This structure parallels the next steps in the simulation
process: set up a simulation, run a simulation, and review results of a simulation. Draft simulations are
simply saved simulation forms that have not been run as simulations yet.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 11
Understanding the Additive Interface

Simulation ID
A simulation ID is a unique identifier for each simulation. Using the simulation ID in a folder name is
convenient when storing all the output files related to that simulation. Also, when reporting a problem
or looking for clarification on a specific simulation, this is the number that should be included with a
support request. You will see the simulation ID in the Overview section of Running and Completed
Simulations and in the dashboard listing of Completed Simulations.

Managing Your Simulations and Data

At most points while using Additive, you have opportunities to Export, Save, Start/Restart, Duplicate,
Cancel, and/or Delete a simulation, depending on where you are in the program. The operations are
designed to work within a database paradigm, as described below.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
12 of ANSYS, Inc. and its subsidiaries and affiliates.
 Managing Your Simulations and Data

Save: Clicking the Save button under Draft Simulations saves your simulation form internally (but not
as a file on your computer) and you will see it listed under Draft Simulations. It is removed from Draft
Simulations when you start a simulation (that is, when it is no longer a “draft”). All your input options
are stored when you run a simulation so that you may see your options at any time when you click on
a simulation in the Running Simulations and Completed Simulation areas of the dashboard. Use Export
to save your simulation form to a file.

Start: Click Start under Draft Simulations to start execution of a simulation. At this point, the simulation
is removed from Draft Simulations and is shown under Running Simulations.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 13
Understanding the Additive Interface

Cancel: Clicking the Cancel button under Running Simulations stops the simulation. (It takes a moment
for the processes to stop.) After canceling a simulation, you’ll see it as canceled under the Completed
Simulations list and you can Restart the simulation again with the Restart button.

Restart: Click Restart under Completed Simulations to restart a canceled or interrupted simulation. (If
you need to change an input value, click on Duplicate, at which point you will need to start the simu-
lation from the beginning again.)

Duplicate: Clicking Duplicate from any point in the application makes a copy of the inputs of that
simulation and creates a Draft simulation of the same name with those inputs. If you are working in
Draft Simulations, a Save is required before you can Duplicate.

Export Simulation: Clicking Export Simulation brings up your file manager so you can save simulation
form inputs to an .aasp file. This proprietary file format contains simulation input data that can be im-
ported into ANSYS Additive. Your geometry selection (that is, the part, supports, and/or build file) and
simulation results are not included on the .aasp file. If you are working in Draft Simulations, a Save is
required before you can Export.

Import: Click Import on the dashboard to bring up the file manager and load an .aasp file. This action
will populate a new simulation form with saved inputs. Note that the part is not included in saved inputs
and will need to be added to the simulation form.

Delete: Clicking the Delete button from either Draft Simulations or Completed Simulations removes
the simulation from the application. The operation will delete all metadata and output files. Data will
be permanently deleted and is not recoverable.

Save Logs: The Save Logs button under Completed Simulations is needed only if you have a problem
with your simulation and you need to contact customer support for a resolution. Clicking Save Logs
brings up your file manager and allows you to write a zipped file containing files used for diagnostic
purposes. Contact us at the ANSYS Customer Portal.

Customize: Clicking Customize under the details of a material page will allow you to edit properties
to define your own custom material to store in the Materials Library. Your custom material will then
appear as an option when selecting a material for a simulation.

Archive: Clicking Archive under the custom material page will remove a customized material from the
Materials Library. Simulations using these materials will remain in your stored application data.

Where is My Data Stored?

Temporary Data Directory

During the normal operation of the application, some code components need a place to store data
temporarily. The data is stored in the temporary data directory. The default directory is C:\Users\[your
username]\AppData\Local\Temp (or %TEMP%). These files are cleaned up during the course of a
simulation.

Permanent Data Storage - Application Data Directory

Think of the Additive application as a database where all your simulation-related data is kept. As such,
you can have only one instance of the application (the database) open at a time. On your computer,
your data is stored in the application data directory. Stored data include your parts, supports, materials,

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
14 of ANSYS, Inc. and its subsidiaries and affiliates.
 Managing Your Simulations and Data

log files, and simulation data, including completed simulations. The default directory is C:\Users\[your
username]\AppData\Roaming\ansys-additive (or %APPDATA%\ansys-additive). This is usually a hidden
file but it can be viewed by typing %appdata% in the Windows File Explorer and then look for the
ansys-additive folder. (%appdata% is an environment variable determined by your operating system.)

Important:

Additive upgrades are “one-way,” meaning after you have upgraded to a new release you
cannot run a previous release without risking the corruption of the simulation database.
Upon installation and opening of a new version of ANSYS Additive, the simulation data
stored in the application data directory is modified to support new features and capabilities
found in the updated application. Additive does not support reverting to older versions
of the application data directory and using a prior version of Additive on upgraded applic-
ation data can, in some cases, corrupt the data, making it unusable for both the old and
new versions. We recommend that you uninstall previous versions of Additive before up-
grading to prevent these issues.

How to Change Data Directories

The temporary data directory and application data directory can be changed by going to Edit >
Settings. Click browse on each folder to change the directory where these files will be stored. Once
you click save, the application prompts you to restart. No automatic restart will occur. The changes
to the data directories will not take effect until after you've restarted. If you already have data (com-
pleted simulations, parts, supports, materials, etc.), it will not be copied over to the new location. It
will remain in the old directory, but you won't be able to access it from the application unless you
switch back to that directory under Settings and restart the application.

You can reset the directories to the default factory settings by clicking "Reset to Default."

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 15
Understanding the Additive Interface

How to Migrate Existing Application Data to a New Directory

1. Under Edit > Settings, take note of your current application data folder.

2. Change your application data folder. (The new folder should be empty.)

3. Close ANSYS Additive.

4. Navigate to the old application data folder.

5. Copy the entire contents of the folder (cassandraBinaries, cassandraData, database, logs, minio,
minioConfig).

6. Navigate to the new application data folder. (It should be empty.)

7. Paste the contents of the old folder into the new folder.

8. Start ANSYS Additive again. Your old data should be there.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
16 of ANSYS, Inc. and its subsidiaries and affiliates.
 Managing Your Simulations and Data

Known Limitations
ANSYS Additive does not support installation across a network so AppData must be local to your
computer. Do not specify a network drive.

Increasing Memory Allocation

On rare occasions, such as when you attempt to view extremely large support files, you may experience
a black screen that takes over the simulation window. This is an indication that you have exceeded
the memory. To recover from this, click View > Reload from the toolbar above the ANSYS logo at the
top of the application.

You can reduce the likelihood of experiencing black screen errors by increasing the maximum memory
consumption for the UI. Do this by changing the target path for the launch shortcut to "C:\Program
Files\ANSYS Inc\v202\AddPrint\ANSYS Additive 20.2.exe" --js-flags="--max-old-space-size=8192" (where
v202 and "20.2" are references to the particular release of ANSYS Additive, in this example, 2020 R2).

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 17
Understanding the Additive Interface

A Note About Data Noise with Parallel Processing

You may notice very slight differences in the stress and distortion outputs from one run to the next
when using more than one core. This is a due to a known limitation in shared-memory parallel pro-
cessing sometimes referred to as "noise." Operational randomness and numerical round-off inherent
to parallelism can cause slightly different results between runs on the same machine using the same
number of cores (p. 53) or different numbers of cores. This difference is often negligible. Assuming
we compute displacements up to the order of ~1 mm, you may expect the noise to be on the order
of ~1e-5 mm. Assuming stress values on the order of ~1 MPa, you may see differences on the order
of ~10 Pa. However, in some cases the difference is appreciable. This sort of behavior is most seen
when simulations are numerically unstable, that is, not converging. The more numerically unstable
the model is, the more likely the convergence pattern or results will differ as the number of cores
used in the simulation is changed.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
18 of ANSYS, Inc. and its subsidiaries and affiliates.
 Application Settings

If run with a single thread, the results between successive runs with identical inputs should produce
identical outputs.

Application Settings

In the menu bar under Edit > Settings you will find the following options:

• Directory Options - Directory locations for temporary and application data. See Where is My
Data Stored? (p. 14)

• Legacy Options - Click Show to make features from previous releases available for selection in
the application interface. Legacy options are hidden by default.

An example of this usage involves the Displacement after Cutoff output option. By default, the
solver used when simulating cutoff scenarios is the Mechanical APDL solver. Before Additive
Release 2020 R2, cutoff was simulated with the Mechanics solver. To make the Mechanics solver
option available in the UI under Displacement after Cutoff, you'll need to select Show for the
Legacy Options setting here in the Settings menu.

• MAPDL (for cutoff only) Parallelization Mode - For simulations using the Mechanical APDL
solver for cutoff mode (p. 78), choose between Shared-Memory Parallel (SMP - default) or Dis-
tributed Computing (DMP) for High Performance Computing. See Overview of Parallel Processing
in the Parallel Processing Guide for more information.

Like in most desktop applications, settings are "sticky" in that once you have set an option, it will remain
that way each time you open the application until you change the setting again.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 19
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
20 of ANSYS, Inc. and its subsidiaries and affiliates.
Choosing a Simulation Type
From the dashboard, click the New dropdown box to choose a new form corresponding to a simulation
type.

There are three types of simulations available in Additive Print: Assumed Strain (p. 22) simulation, Scan
Pattern (p. 23) simulation, and Thermal Strain (p. 23) simulation. The simulation types (sometimes called
strain modes) specify the different ways inherent strain is calculated as an input to the Mechanics
Solver. All three strain modes offer the same simulation output options.

There are four types of simulations available in Additive Science: Single Bead, Porosity, Thermal History,
and Microstructure (Thermal History simulation type is a Beta feature at this release). The goal of Additive
Science is to determine the best process-parameter combination to use for building your part, given a
LPBF machine and a material. You begin this exploration with a Single Bead Parametric simulation to
narrow the process-parameter combinations down to a smaller number of acceptable candidates based
upon melt pool dimensions. Typically you will then want to do a Porosity simulation using your chosen
parameters from the Single Bead simulation to determine the lack-of-fusion porosity associated with
those process parameters. Finally, Microstructure simulations reveal information about grain patterns
and may be compared to EBSD laboratory tests.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 21
Choosing a Simulation Type

Assumed Strain - Uniform Isotropic

Assumed strain mode is the fastest simulation type available. It assumes that a constant, isotropic strain
occurs at every location within a part as it is being built. The strain is equal to the Strain Scaling Factor
multiplied by yield strength and divided by elastic modulus:

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
22 of ANSYS, Inc. and its subsidiaries and affiliates.
 Thermal Strain - Anisotropic

The Strain Scaling Factor (SSF) is an important factor quantifying the variables unique to each build
scenario. It must be experimentally determined for each machine/material/strain/stress mode combination
of interest. Calibration procedures are available online here.

Except indirectly through the use of the experimentally determined calibration factor (SSF), an Assumed
Strain simulation does not use information about the machine or scan pattern.

See Set Up a Simulation - Assumed Strain (p. 51) for procedures.

Scan Pattern Strain - Anisotropic

This strain mode uses the same average strain magnitude as assumed uniform strain, but it subdivides
that strain into anisotropic components based on the local orientation of scan vectors within the part.
This strain mode requires the creation of scan vectors using user-provided scan settings or by reading
scan vectors from a machine's build file. This extra step results in a small, increased amount of simulation
time compared to assumed uniform strain. For parts where the scan pattern is randomized, scan pattern
and assumed strain should give a similar answer. For parts where the scan patterns are aligned, scan
pattern strain will result in a more accurate prediction. As in an Assumed Strain simulation, you will
need to calibrate for Strain Scaling Factor.

See Performing a Scan Pattern Simulation (p. 99) for procedures.

Thermal Strain - Anisotropic

This strain mode provides the highest degree of accuracy by predicting how thermal cycling affects
strain accumulation at each location within a part. A "thermal ratcheting" algorithm assigns a base strain
to each location within the part as it solidifies. Each time a location within the part is heated above a
temperature threshold (approximately 40% of its absolute melting temperature) an increase in strain
in that location occurs. If a location re-melts, the strain is reset to the base strain. The more times a
location is heated above the threshold without melting, the higher the strain accumulates. Once the
strain magnitude is calculated for each location within a part using the thermal ratcheting algorithm,
that strain is passed to the Mechanics Solver and applied as an anisotropic strain based upon both
local strain magnitude and local scan orientation. Because thermal strain requires a thermal prediction
for every scan vector, this strain mode requires a much longer computational time. As in Assumed Strain
and Scan Pattern simulations, you will need to calibrate for Strain Scaling Factor.

See Performing a Thermal Strain Simulation (p. 103) for procedures.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 23
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
24 of ANSYS, Inc. and its subsidiaries and affiliates.
Theoretical Overview
This chapter presents theoretical descriptions for customers who want to better understand how the
application uses input data to calculate output, and how to interpret various results.

The following topics are available:

Overview of the AP Mechanics Solver
Overview of the Thermal Solver

Overview of the AP Mechanics Solver

In the process of additive manufacturing, parts are produced by sintering/fusing material in a track-by-
track fashion compositing material horizontally into thin layers and then vertically into actual part
geometry.

Additive Print's Mechanics Solver (the AP Mechanics Solver, or simply the Mechanics Solver) simulates
this process of material consolidation into geometric shapes and the resulting distortions from intended
shape and induced internal residual stresses. The manufacturing process involves depositing very thin
tracks of material sequentially. But to simulate the micron level of detail relies upon a highly accurate
thermal solution and huge amount of computing resources. To get a reasonably accurate solution in a
faster manner the Mechanics Solver makes a set of simplifying assumptions.

Theory and Assumptions

The AP Mechanics Solver assumes that the entire powder layer is melted at once instead of in a track-
by-track fashion. This melted layer sits on top of an existing solid layer and experiences thermal
shrinkage during the solidification process. Cooling of the material is assumed to happen instantaneously.
Multiple micron-scale physical layers are combined into one layer for simulating this shrinkage process.
Each of these layers is meshed using the closest possible approximation of voxels, the size of which is
defined by the input parameter, Voxel Size (p. 54). This results in a voxelized representation of the ac-
tual geometry after all layers are meshed. Due to the thermal gradient across solidified layers and sub-
sequent variable shrinkage, the material experiences internal stresses that get locked in as residual
stresses.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 25
Theoretical Overview

For a given deposition layer the shrinkage of the material can be written in equation form as:

where is mechanical strain and is thermal strain.

Internal force in the material in this layer is generated as a resistance to thermal shrinkage:

where is the strain displacement matrix and is elastic modulus.

is obtained by assuming for the top layer to get . The actual value of and are
obtained after solving for global equilibrium. We can solve for equilibrium across the geometry using:

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
26 of ANSYS, Inc. and its subsidiaries and affiliates.
 Overview of the AP Mechanics Solver

where is the global stiffness matrix and is the global displacement vector. This is a simple
linear elastic formulation to obtain a fast solution. For the plastic solution, to obtain equilibrium we use
Newton-Raphson iterations to reach a converged equilibrium solution. For that we calculate a yield
function that is dependent on current stress and plastic strain:

where is the stress tensor, is the yield function, is the deviatoric stress tensor, is the equivalent
plastic strain, and is the hardening modulus.

At every Newton-Raphson iteration the yield function is calculated to check for convergence:

A residual is calculated based on the internal force and used for convergence:

An incremental contribution to the total displacements is calculated based on this residual:

The final solution is based upon this cumulative of the incremental displacements. The total and elastic
strains are calculated as:

The shrinkage strains are not applied to support material and omitting these strains reduces computation
time. Overhang features are assumed to be either connected to the part or supported. If they are not,
no strain will be applied to them until they are joined with the rest of the part. (See Hanging/Floating
Voxels (p. 59).) Nodes on the bottom of the part and support are fixed in place to represent the con-
nection to the base plate.

Strain Modes
The structural problem of the additive simulation is driven by thermal shrinkage ( ) that results in part
deformation and residual stresses. There are three strain modes that the Mechanics Solver can use to
drive it: assumed strain, scan pattern-based anisotropic strain, and thermal strain. The assumed strain
option applies isotropic inherent strain to each layer for simulation. This inherent strain can be obtained
from previously calibrated parts. See the Additive Calibration Guide. The scan pattern-based anisotropic
strain uses the scan paths for the laser and generates anisotropic directional dominant strains. Experi-
mental studies show that shrinkage strain is higher parallel to the scan direction than perpendicular to
the scan direction. The scan pattern strain mode can be used to apply this kind of anisotropic strain
based upon the scan vectors. Only fill scan patterns are used; contour scan patterns are ignored. The
strains for the parallel, perpendicular, and z-directions are determined by multiplying the inherent strain
by the anisotropic strain coefficient. Directional strains for each scan are aggregated within each indi-
vidual voxel used in the structural solution to give location dependent strains. The strains in the thermal
strain option are generated by the Thermal Solver (p. 29) as explained in the following section.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 27
Theoretical Overview

Coupling with Thermal Solver

The Thermal Solver produces a rich spatio-temporal temperature field. This solver follows the laser path
in small time intervals and simulates the thermal conduction through the solidifying part. The resulting
temperature field is available at micron scale. The Mechanics Solver is weakly coupled with the Thermal
Solver which means the temperature fields are used passively to simulate the mechanical response.
Based on the time history of temperature at a given location, the Thermal Solver uses a thermal
ratcheting method to modify the user-defined inherent strain and generate shrinkage strain values to
pass to the Mechanics Solver. The Mechanics Solver reads the shrinkage strain data and averages it over
the given mesh element and applies it as an unbalanced force boundary condition. The thermal
ratcheting algorithm used to generate thermal strain is discussed in the following section.

Thermal Strain – Thermal Ratcheting Algorithm

This strain mode provides the highest degree of accuracy by predicting how thermal cycling affects
strain accumulation at each location within a part. A "thermal ratcheting" algorithm assigns a base strain
to each location within the part as it solidifies. Each time a location within the part is heated above a
temperature threshold (approximately 40% of its absolute melting temperature) an increase in strain
in that location occurs. If a location re-melts, the strain is reset to the base strain. The more times a
location is heated above the threshold without melting, the higher the strain accumulates. Once the
strain magnitude is calculated for each location within a part using the thermal ratcheting algorithm,
that strain is passed to the Mechanics Solver and applied as an anisotropic strain based upon both
local strain magnitude and local scan orientation. Because thermal strain requires a thermal prediction
for every scan vector, this strain mode requires a much longer computational time. As in Assumed Strain
and Scan Pattern simulations, you will need to calibrate for Strain Scaling Factor.

Post-processing of Simulation Results

After the additive manufacturing process, distortion and residual stress that result in a part create various
problems and opportunities for improvement. Using simulation can inform your decision-making regard-
ing how to set-up, prevent failures, and even compensate the geometry for best printing results. The
Additive application's Mechanics Solver internally produces results that can be useful in various contexts.
For example, warping results are used to find curling of the part in the vertical direction and a potential
recoater blade crash (p. 83) is indicated in the blade crash result file. Relieving the part by cutting it
off from the base plate results in a release of some of the locked-in stresses and this rebalances the
part's internal forces. This results in a re-distortion of the part. The Mechanics Solver can simulate this
cutoff process and calculate the after cutoff distortion (p. 78). The final distortion of the part can be
used to invert the part to various magnitudes and rebuild to obtain a to-the-spec geometry. This kind
of compensated geometry (p. 77) can be produced by the solver. Residual stresses at the part and
support interface are used to generate an optimized support (p. 63) by changing the thickness of the
support walls or change the distance between support walls. When the part is being built it undergoes
a history of distortion and residual stresses. Some areas of the geometry may experience high stresses
in one of the layers despite having a low final stress after deposition of all layers. This could be a result
of rebalancing of stresses after further layers are deposited. This history of high stress might indicate
internal/external cracking (p. 84) and some other microscopic defects. The Mechanics Solver can save
the high-water mark of stresses and strains through the history and output for the entire geometry (p. 82)
at the end of simulation. These high stresses and strains can be used to inspect and possibly change
the geometry to mitigate failures.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
28 of ANSYS, Inc. and its subsidiaries and affiliates.
 Overview of the Thermal Solver

Relationship Between the AP Mechanics Solver and the MAPDL Solver

The AP Mechanics Solver is the structural solver included in the ANSYS Additive Application. Everything
written in this theoretical overview applies to the AP Mechanics Solver.

The Mechanical APDL (PCG) Solver is used in Workbench Additive. The FEA technology used by the AP
Mechanics Solver is very similar to that of the Mechanical APDL (PCG) Solver, except for a few notable
differences listed in the table below.

Additive Print Mechanics Solver Mechanical APDL Solver In Workbench Additive

The bottom of the build is automatically The bottom of the build is typically connected
constrained. to a base plate. The user applies a fixed
condition to the base plate.
Inherent strain is applied to the part only starting Inherent strain is applied to the part and
from the second layer. supports starting from the first layer.
Strain is not applied to unconnected Strain is applied to unconnected overhangs.
overhangs (p. 59), such as “stalactite” points, lines,
and surfaces (also known as islands), until the
voxels are connected to voxels that touch the base.
Nodal results displayed at the interface between Nodal results displayed at the interface between
support and part elements show results averaged support and part elements show results
from part elements only. averaged from both part and support elements.

Starting at Release 2020 R2, the Mechanical APDL Solver was implemented in the Additive application
as the solver just for the specific case of displacement after cutoff (p. 78), including both instantaneous
and directional cutoff options.

Overview of the Thermal Solver

The Thermal Solver simulates scan tracks on each deposit layer of the Laser Powder Bed Fusion (LPBF)
additive manufacturing process. The solver takes a multi-scale approach, leveraging periodic behavior
observed in LPBF to improve performance. At the fine scale, a transient FEM solution is used to determine
a temporally evolving temperature field. The fine-scale solution is then coarsened out to apply a peri-
odic heating solution to each deposit layer of the entire build, while still following the laser path.

The fine-scale solution is an FEM simulation of diffusion in a single scan track. The material properties
(such as conductivity, density, specific heat) are temperature-dependent and state-dependent for powder
and the bulk solid. Powder properties are a ratio of the bulk-solid properties, and the ratio may also be
temperature dependent. The powder is treated as a homogeneous absorbing scattering medium in
which a Gaussian laser model is applied as a volumetric heat source along the top of this track (see
Gusarov et. al. [1]). An absorption coefficient and extinction coefficient are used to model the energy
penetration for a given scan speed and laser power combination, as found from the material tuning
process.

The mesh in the FEM solution is not the same voxel mesh as used in the Mechanics Solver. Rather, it is
a Cartesian grid with uniform horizontal resolution and a different grid resolution in the build direction
determined dynamically based upon the deposit layer thickness. The horizontal mesh resolution is dif-
ferent for various simulation types. However, these fine-scale mesh resolutions are not user adjustable
so if a melt pool becomes too large and reaches the domain boundaries, edge effects (p. 31) may appear.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 29
Theoretical Overview

Going from the fine-scale single-track result to the part level, the horizontal mesh resolution is coarsened
by a Mesh Resolution Factor (MRF) (p. 103) to improve performance. The MRF is user adjustable for some
simulation types. Too large of an MRF may result in poor resolution of melt pools. At the part level,
each deposit layer starts out as all powder at the ambient heater temperature. The MFR-coarsened
heating solution is applied to each individual solution step within each scan track in the scan pattern.
Cooling is simulated during the inter-scan dwell time between tracks. The mesh nodes marked as powder
become melted once the temperature exceeds the material's solidus value. Once a node is melted, the
material state is no longer powder and this history is retained through successive layers.

The user-specified Baseplate Temperature is used as a fixed-temperature boundary condition at the

bottom of the build. In addition to the Gaussian laser heat-source model, the top boundary of the build
has a constant forced convection applied uniformly. Aside from the laser model, radiation effects are
neglected. The lateral boundary conditions are adiabatic while using a buffer region of powder extending
out from the part's bounding box to avoid any boundary effects.

The current approach has its limitations. As periodic behavior is a key assumption for performance, only
fill-type scans can be simulated well in terms of simulation runtime. The solver can predict lack-of-fusion
porosity via the powder-solid state tracking but does not predict balling or keyhole phenomena. Other
physical phenomena that are not explicitly modeled include: latent heat, surface tension effects, vapor-
ization, plasma, and spatter. As such, an experimentally determined "cap temperature" is found for each
alloy as part of the material tuning process. This cap temperature will offset some of the adverse effects
of the assumptions and limitations.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
30 of ANSYS, Inc. and its subsidiaries and affiliates.
 Overview of the Thermal Solver

Thermal Solver Default Settings

Simulations using the Thermal Solver have some input parameters that are not available to the user.
These variables include thermal material properties, mesh size, absorptivity, and laser penetration depth.
Thermal material properties are temperature dependent and have been set to match material produced
through the additive manufacturing process. Laser penetration depth and absorptivity have been tuned
to fit data from single bead experiments and vary according to the power and velocity used in the
simulation. Mesh size is set to 15 µm for Single Bead, Thermal Strain, Microstructure and Thermal History
simulations, and 25 um for Porosity simulations. These values have been identified as values that give
good results both in terms of accuracy and simulation speed.

Re-tuning of the ANSYS-supplied materials may or may not be performed each software release, depend-
ing on the extent of the solver updates. When material re-tuning is performed, small changes in results
from previous releases may be observed.

Guidelines for Resolving Warnings and Errors

As the model is not completely user adjustable, some errors may arise. This section discusses some
potential issues you may encounter.

• Edge effects in highly diffusive materials and/or high energy density scenarios

The fine-scale scan track domain width and depth have been set based on tuning of commonly used
build settings and are not user adjustable. Process-parameter inputs that provide a very large energy
density may result in edge effects due to the boundary conditions applied to the fine-scale single
track solution. Melt pool width and depth growth may be impeded by the adiabatic lateral boundaries
and constant temperature bottom boundary, respectively, under extreme conditions. That is, melt
pool dimensions may become artificially constrained. This presents in different ways depending on
the material's diffusivity.

– For highly diffusive materials, such as Aluminum alloys, if the melt pool width or depth ap-
proaches the domain boundary exceeding a set tolerance, the solver logs a warning or errors-
out with a message describing which dimension(s) triggered the failure.

– For less diffusive materials, the melt pool growth may not be sufficient to trigger the warn-
ing/error but may still be constrained by domain edge effects. In this scenario, the solution
continues. The way to discern the phenomena is to observe that the melt pool dimension(s)
does not change as the energy into the system continues to increase.
If either scenario is encountered, we recommend reducing the input energy density (any combination
of reducing the Laser Power, or increasing the Scan Speed or Laser Beam Diameter) or selecting a
material with a lower diffusivity.

Specifically, this behavior has been observed when the melt pool width and/or the melt pool depth
approach certain limits. For width, this is on the order of ∼600 microns, and becomes more severe
as it approaches ∼800 microns. Due to differences in sensitivity to melt pool depths for different
simulation types, the onset for this issue will vary from melt pool depths of ∼400 microns for Single
Bead, Porosity, and Microstructure simulations; to ∼200 microns for Thermal History (Beta) and ∼140
microns for Thermal Strain simulations. We do not expect this issue of edge effects to be present
under typical operating ranges for most simulation types, that is, using typical process parameters.
It is typically only when you "push the boundaries" of the operating window that you may experience
these edge effects. But we believe testing the boundaries is exactly what simulation is for, so we do
not necessarily discourage it. To serve as a potential alert that you may be pushing the boundaries,

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 31
Theoretical Overview

we include recommended ranges that are tighter than allowable ranges for the process parameters
that affect energy density most directly: Laser Power and Scan Speed. For example, the allowable
range for Laser Power is 50 to 700 Watts but a range of 50 to 500 Watts is shown in the UI as the
"recommended range." The following simulation types may be affected in the following ways:

– Thermal Strain: The thermal strain may be less accurate, affecting the quality of the resulting
stress and distortion. It should be noted that while Thermal Strain simulations use a shallower
mesh and are more likely to be influenced by this on a localized basis, the overall impact is
mitigated by the averaging that occurs when preparing to run a structural solution.

– Single Bead: The reported melt pool width and depth may be artificially capped at around
800 microns and 400 microns, respectively.

– Porosity: The resulting solid ratio may be under-predicted.

– Microstructure: The resulting melt pool width and depth may be under-predicted, and the
thermal gradients and cooling rates may be incorrect.

– Thermal History (Beta): The reported melt pool width and depth may be under-predicted, and
average temperature may be less accurate.

• Warnings related to Mesh Resolution Factor

The Mesh Resolution Factor (MRF) (p. 103) is user adjustable for some simulation types. Too large of
an MRF will lead to poor resolution of the melt pool. The Thermal Solver will emit a warning when
this occurs. A smaller MRF may resolve the melt pool and should resolve the warning, at the cost of
longer simulation times.

References
1. Gusarov, A.V., et. al., (2009) Model of Radiation and Heat Transfer in Laser-Powder Interaction
Zone at Selective Laser Melting, ASME Journal of Heat Transfer. Web.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
32 of ANSYS, Inc. and its subsidiaries and affiliates.
Understanding Machine Parameters
There are several standard machine parameters, also known as process parameters, used in Additive
simulations. Each simulation (except the Assumed Strain simulation) uses some, or all, of these parameters.
The machine parameters are shown in the following figure and are listed here for your reference.

When entering values for process parameters in the application interface, you may see a recommended
range message. Values in a recommended range represent the values that ANSYS has validated.

Laser Power (W)

The power setting for the laser in the machine. Must be between 50 and 700 Watts. Defaults to 195
Watts.

Scan Speed (mm/sec)

The speed at which the laser spot moves across the powder bed along a scan vector to melt material,
excluding jump speeds and ramp up and down speeds. Must be between 350 and 2500 mm/sec. Defaults
to 1000 mm/sec.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 33
Understanding Machine Parameters

Laser Beam Diameter (µm)

The width of the laser on the powder or substrate surface defined using the D4σ beam diameter
definition. Usually this value is provided by the machine manufacturer. Sometimes called laser spot
diameter. Must be between 20 and 140 µm. Defaults to 100 µm.

Starting Layer Angle (°)

The orientation of fill rasters on the first layer of the part. This is measured from the X axis, such that 0
degrees results in scan lines parallel to the X axis. The starting layer angle is commonly set to 57 degrees.
Must be between 0 and 180°. Defaults to 57°.

Layer Rotation Angle (°)

The angle at which the major scan vector orientation changes from layer to layer. This is commonly 67
degrees. Must be between 0 and 180°. Defaults to 67°.

Layer Thickness (µm)

The thickness of the powder layer coating that is applied with every pass of the recoater blade. We re-
commend that you use the actual thickness used for your machine and build material. Must be between
10 and 100 microns. Defaults to 50 microns.

Hatch Spacing (µm)

The distance between adjacent scan vectors when rastering back and forth with the laser. Hatch spacing
should allow for a slight overlap of scan vector tracks such that some of the material re-melts to ensure
full coverage of solid material. Must be between 10 and 1000 microns. Defaults to 100 microns.

Slicing Stripe Width (mm)

When using the stripe pattern for scan strategy, the geometry can be broken up into sections, called
stripes. The stripes are scanned sequentially to break up what would otherwise be very long continuous
scan vectors. Slicing Stripe Width is commonly set to 10 mm wide. Memory requirements for the thermal
solution will expand significantly as you increase the Slicing Stripe Width much beyond the default.
Must be between 1 and 100 mm. Defaults to 10 mm.

Baseplate Temperature ( °C )
The controlled temperature of the baseplate. Must be a real number between 20 and 500. Defaults to
80 °C.

Machine
A placeholder for future machine configurations. Defaults to Generic for now.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
34 of ANSYS, Inc. and its subsidiaries and affiliates.
 Additive Print Simulations
Prepare and Import a Part (p. 37)

Set Up a Simulation - Assumed Strain (p. 51)

Run a Simulation - Assumed Strain (p. 87)

Review Results of a Simulation - Assumed Strain (p. 91)

Performing a Scan Pattern Simulation (p. 99)

Performing a Thermal Strain Simulation (p. 103)

Prepare and Import a Part
A typical workflow begins with preparing a part for importing to the Additive application. From within
a CAD program, such as ANSYS SpaceClaim, you’ll need to export an .stl file. The standard file format
for most rapid prototyping and 3D printing applications, an .stl file (from stereolithography) is a tessellated
representation of a 3D object that consists of triangle elements that define the external surface of the
object. Rather than being a volume representation, surface normal vectors define the inside versus the
outside of the object.

Guidelines for Part Orientation and Resolution

There are some important considerations to be aware of before you export your geometry from the
CAD program:

• Part geometry files are in .stl format.

• Only one part can be simulated. There can be multiple bodies, but they have to be on the same
part file.

• Do not include supports in the part geometry file. You will have an opportunity to import supports
separately or have the Additive application create supports automatically for the part. If you
have designed supports along with the part, keep them as separate bodies and save them to a
separate file.

• Do not include a baseplate (build plate) in the part geometry file. The Additive application takes
the baseplate into consideration internally in the simulation and it never needs to be explicitly
modeled.

• Dimensions of the part must be in units of millimeters (mm). While .stl files are unitless, the Ad-
ditive application does not provide the ability to switch unit systems and Metric units of milli-
meters are assumed.

• Currently, build size is limited to one cubic meter, that is, 1000 millimeters in all directions (1000
x 1000 x 1000 mm). "Build size" includes the part and the supports. Be aware that the larger the
part, the greater the memory required for solution.

• The .stl file must have the part positioned in the orientation in which it will be printed.

• A part with its longest dimension in the Z direction will require the longest simulation time.
While the domain volume doesn’t change with part orientation, the number of voxel layers
changes, which means more calculations in the solver are being performed. Alternatively, a part
with the smallest dimension in the Z direction will have the shortest simulation time. (Note that
parts oriented with their longest dimension in the Z direction will similarly take longer to print
on the machine because more layers are required. Powder recoating/spreading time is an order
of magnitude higher than laser printing time.)

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 37
Prepare and Import a Part

• The time required to slice and voxelize an .stl file exponentially increases with the number of
triangles. Given two .stl files for the same geometry, the file with the smallest number of triangles
will be processed faster. (Slicing, as used here, refers to the internal process of dividing a part
into scan vectors according to the scan pattern input parameters that will be used in the 3D
build process. Voxelization refers to the dividing of a part into voxels, or elements, used in the
mathematical simulation. See Voxel Size (p. 54).)

Importing a Part

To import a part to the Parts Library, click the Parts button in the left panel of the dashboard
and then click Import Part. Part files are .stl files, either ascii or binary. While .stl files are unitless, dimen-
sions of the part in Additive are assumed to be in units of millimeters (mm).

The name, tags, and description fields allow you to identify the part in some way that makes logical
sense to you. These fields are searchable at any location in the program that allows for searching on
text fields. Tags should be at least three characters long. Be aware that the search function is case-
sensitive.

Click Save to begin the import.

It may take a few minutes to import a part, depending on the size of the file. The status indication in
the Parts Library shows “Processing” while the part is importing and “Available” when importing is
complete. A common practice is to import the part and then begin setting up your simulation. When
you get to the step of selecting geometry on the simulation form, the part is usually available for selec-
tion.

You will see your part in the geometry window when the part is available. Use your mouse buttons to
move the part around in the image preview; left button for spin, middle for zoom, and right button for
pan. Also shown is a triad legend to provide directional perspective.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
38 of ANSYS, Inc. and its subsidiaries and affiliates.
 Importing a Part

The Simulation Domain

When you import a part, the Additive application internally creates a bounding box around the part
using the outermost edges of its dimensions. This bounding box is known as the simulation domain.
The location of the minimum X, Y, and Z dimensions of this domain becomes the new origin (0, 0, 0)
for the purposes of voxelization (p. 54) and calculations. Furthermore, the application assumes the
X-Y plane at Z=0 is the top of the baseplate. Result items, such as displacements, will be output in
coordinates based on this translated origin.

The simulation domain will be adjusted again, with the origin translated to the minimum X, Y, and
Z dimensions, when you import supports (p. 40) for the part. Ultimately, when meshed with
voxels (p. 54), some voxels within the domain are part material, some are support material, and some
are powder.

In the geometry preview window, be careful to distinguish between the triad legend, which is shown
simply for directional perspective, and the new, translated origin of the simulation domain, which is
not shown. The triad legend is not the origin.

Editing and Exporting a Part

If you want to see details about a part that has already been imported, you may view it in the Parts
Library by clicking on it or searching on any text term used in the name, tags, or description of the
part. Click the part name to bring up a details page with information and an image preview of that
part.

Click Edit to edit the name, tags, or description. You cannot edit the features of the geometry itself.
(Note that if the original .stl file is changed, it does not affect the imported part. A new copy is made
inside the application that is not linked to the original file.)

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 39
Prepare and Import a Part

Click Export to export a saved part .stl file. At this time, supports that may be associated with the
part are not saved with the part when exported.

Importing Supports
Once a part is available after importing, you can import supports that will be associated with that part.
Click Import Support to bring up the import support window.

See Define Support Options (p. 61) for general information about how supports are used in the additive
manufacturing process. For now, it is important to know the following:

• Support files are in .stl format.

• Dimensions are assumed to be in millimeters (mm).

• The support must be oriented in the same 3D space as the part (aligned with the part in the X-Y
plane).

• Each support .stl file must be homogeneous with respect to its geometry, that is, either all volumeless
supports or all solid supports, within the file. See Support STL Type (p. 41).

• There is no limit on the number of support .stl files that can be imported for any given part.

• Create Support Groups (p. 42) to use multiple support .stl files for the same part in a simulation.

Lock Support to Part

Check this option to lock the position of the support with that of the part. This implies that you have
created the part and supports so that they are properly aligned with respect to each other in X, Y, and
Z. Then, regardless of how each geometry may have been rotated for viewing when it was imported,
the application assumes they are in alignment in 3D space.

If a support is a part-to-part only support (that is, no portion of the support structure touches the
baseplate), this check box must be checked to ensure proper alignment. If some segment of the support
does touch the baseplate and the support and part are aligned in X, Y, and Z, this feature will automat-
ically calculate the proper Minimum Support Height.

When this check box is checked, the Minimum Support Height input is disabled since that value will
be calculated automatically.

Minimum Support Height (mm)

Enter a Minimum Support Height when Lock Support to Part is unchecked. This is the height, in milli-
meters, that the part will be elevated off the baseplate when the support is properly positioned below
the part and the support is resting on the baseplate. In other words, the Minimum Support Height is
the distance between the baseplate and the lowest point on the part when it is properly aligned with
a given support. A value of 0 means that some point of the part will start printing directly on the
baseplate.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
40 of ANSYS, Inc. and its subsidiaries and affiliates.
 Importing Supports

It is important to be sure the supports and the part are properly aligned in the X and Y axes before
importing. The correct Minimum Support Height ensures that the part and support are properly aligned
in the Z axis at simulation time.

Known Limitation: When Supports Penetrate the Baseplate

The bounding box defining the simulation domain is expanded to account for the supports upon import,
and the origin of the domain is translated to the minimum X, Y, and Z location of the expanded domain.
The application assumes the X-Y plane at Z=0 of this translated origin is the top of the baseplate. This
is based on the assumption that the part and support are always built on top of the baseplate.

A known limitation occurs when a support is designed with the intent of penetrating into the baseplate,
where a small portion of the support is below the baseplate top surface. (Usually this is set by intrusion
or penetration settings in support generation tools.) In these situations, the support will still be translated
to the baseplate top surface, which can cause unexpected behavior in the simulation results, such as
the part or other support structure being disconnected from the baseplate. You may recognize this by
seeing unexpected gaps or unexpected result trends in some cases, but in other cases, where only some
elements are affected, this could be difficult to observe.

In general, the known limitation exists where a part and/or support are designed with a lowest point
being something other than Z=0. In the Additive application, because of the automatic translation de-
fining the domain, they are treated as if the lowest point is at Z=0. The workaround is to not add bottom
intrusion to supports, or to trim the bottom intrusion portion when preparing for simulation.

Support STL Type

Choose the type of support you are importing:

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 41
Prepare and Import a Part

Volumeless STL: These supports are usually single-bead width support walls such as lattice supports.
These structures do not need to be "watertight." Other names for this type include thin wall, vector,
and facet.

Solid STL: These supports are standard, watertight geometry bodies. Other names for this type include
thick wall, bulk, and volume.

Part and Support Preview

Click Save to initiate the import of the support. As is the case for importing the part, the status shows
"Processing" while the support is being imported, and "Available" when the support has been fully
processed. Once the support is Available, click the Preview check box to see the part and support to-
gether in the preview window.

To import another support for the same part, simply click Import Support again and go through the
same process to import the new file. Use the Preview check box for each support to confirm its location
with respect to the part. Note that you can preview only one support at a time using the Preview check
box.

Support Groups
To use multiple supports .stl files in a single simulation, you must first create a support group. You can
mix support .stl types, that is, volumeless supports and solid supports, in a support group. There is no
limit to the number of supports a support group can contain, but supports can be in the same group
only under these conditions:

• The Minimum Support Heights are the same, or

• The Minimum Support Height is zero and the support is locked to the part.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
42 of ANSYS, Inc. and its subsidiaries and affiliates.
 Importing a Build File

Importing a Build File

In the Additive Print application, we define a build file as a .zip file containing, at a minimum: one .stl
file for the part geometry and one machine-specific print file defining the scan vectors. Supported machine
manufacturers include Additive Industries, EOS, Renishaw, and SLM. ANSYS may add additional options
as we continue to work with more machine partners.

Machine-specific requirements are documented in the following sections but here are the general re-
quirements for build files:

• The build file is a .zip file. Do not nest the files to be zipped within a folder, as a folder structure
is not readable. Rather, zip the individual files together as shown in the following figure.

• Only one part can be included in a build file.

• The baseplate should not be included in the part geometry file.

• One or more supports may be included as separate .stl files. Support file names are required to
have a suffix indicating their support .stl type (p. 41), either volumeless ("*_vless.stl") or solid
("*_solid.stl"); for example, supports _at_holes_vless.stl, supports_at_overhang_vless.stl, sup-
ports_at_arch_solid.stl, etc. All the supports included in the build file will be used in your simu-
lation if you choose to use the build file supports.

• For the machine-specific print file:

– One laser head is assumed for the simulation. Multi-laser build files are not supported. If
you import a build file that includes multiple lasers, how it is handled in the application

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 43
Prepare and Import a Part

is machine-specific. In most cases the application either produces an error on import or

ignores the extra lasers.

– One set of process parameters is used in the simulation. If multiple parameter sets are
included in the build file, say different scan speeds and laser powers for the part hatches
versus the support hatches, the part hatch parameters will override. When running thermal
simulations, parameters under Machine Configuration (p. 104) in the simulation form
override.

– Only one part layer thickness is allowed.

– Only one support layer thickness is allowed, and it must be equal to, or a multiple of, the
part layer thickness.

– The scan sequence is always simulated from the inside out, that is, from hatch to contour
scans, regardless of how they are defined in the build file. The appropriate order is
maintained within the hatch area and within a contour, however. For example, if the build
file order is: contour line 1 → contour line 2 → hatch line 1 → hatch line 2, it will be
changed to be hatch line 1 → hatch line 2 → contour line 1 → contour line 2.

– Scan vectors marked as contour will not be simulated in simulations that perform a full
thermal solution (Thermal Strain and Thermal History simulation types). The definition of
contour/hatch is established by the software that creates the build file. If contour-like
scan vectors are marked as hatch they will be simulated and results may not be as expec-
ted.

Build files from ANSYS Additive Prep are automatically created with a file name of "ansys_addit-
ive_print.zip.

To import a .zip file to the Build Files Library, click the Build Files button in the left panel of the
dashboard and then click Import Build File. Click Choose File and navigate to the appropriate .zip file
on your computer. Provide a Name and Description (optional) and then click Build File Type to choose
which of the following approved machine type translators Additive will use when reading the data:

• Additive Industries (p. 46)

• EOS (p. 46)

• HB3D (p. 47)

• Renishaw (p. 48)

• Sisma (p. 48)

• SLM (p. 48)

• Trumpf (p. 48)

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
44 of ANSYS, Inc. and its subsidiaries and affiliates.
 Importing a Build File

Once you click Save, you will see a message "Valid file selected. Build file importing now..." The import
process may take several minutes, depending on the size of the file. During processing, a preview window
appears with "Availability: Processing" status. The geometry is ready for simulation when you see the
"Availability: Available" status.

If your build file includes supports, click the Preview Supports button to show the part and the supports
together in the preview window.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 45
Prepare and Import a Part

Note:

A build file may contain scan vectors that exceed the part geometry boundaries. Because
the Mechanics Solver uses voxels based upon the part geometry, strains resulting from the
portion of scan lines exceeding part boundaries may not be included in the stress calculation.
This depends upon the amount by which scan vectors extend beyond the part boundaries.
See the discussion of Voxel Sample Rate (p. 56).

Using Additive Industries Build Files

A build file for an Additive Industries machine should be a zip file containing:

• Part = *.stl file

• Support = *_vless.stl and/or *_solid.stl file(s), optional, but supports will not be simulated

• Two machine files are required, one *.daij file and one *.bin file.

Notes:

• Support scan vectors will be ignored and will not be simulated. An Additive Industries build file
with supports has not been tested.

• Only a stripe scan pattern has been tested.

Using EOS Build Files

A build file for an EOS machine should be a zip file containing:

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
46 of ANSYS, Inc. and its subsidiaries and affiliates.
 Importing a Build File

• Part = *.stl file

• Support = *_vless.stl and/or *_solid.stl file(s), optional

• Machine = *.openjz file. Additive Print uses version 2.8 of the EOS API.

Notes:

• You must have an EOSPRINT 2 license, in the form of a dongle or from a license server, from EOS
in order to import an EOS build file into Additive Print. If you are using a dongle, do not unplug
it until the import availability status changes to "Available." The Additive application uses the
EOS build processor only when importing. Once a build file is stored in the Additive Build Files
Library, you do not need to consume the license/dongle when you use that build file for a
simulation.

• EOS M100, M290 and M400 single-laser machines are supported.

• The only support scans that will be included in the simulation are supports in the same layers
as the part. An EOS build file does not include enough information to identify support scan
lines.

Known Issues and Limitations

• When preparing the .openjz file in EOSPRINT, start height and end height (<height> element)
should not be zero since it is never the position of any layer. The build file may be rejected upon
importing to the Additive application.

• If you stop CodeMeter service while importing an EOS build file into the Additive application, the
import availability status remains in the processing state indefinitely. You will need to restart
CodeMeter and reimport the build file.

• When attempting to import an older EOS build file, the Additive application may reject it since
version 2.8 of the EOS API is stricter in regards to the OpenJob format. If this happens, you will see
an error under the import availability status like this: Input file(s) was not loaded
successfully. OpenJz: EOS_ERR_OPENJOB_INVALID.

Using HB3D Build Files

A build file for an HB3D machine should be a zip file containing:

• Part = *.stl file

• Support = *_vless.stl and/or *_solid.stl file(s), optional

• Machine = *.h3d file. Additive Print uses version 1.0 of the *.h3d file specification.

Notes:

• An HB3D build file ignores the last layer when the distance from its Z coordinate to the Z max
is less than one layer thickness.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 47
Prepare and Import a Part

Using Renishaw Build Files

A build file for a Renishaw machine should be a zip file containing:

• Part = *.stl file

• Support = *_vless.stl and/or *_solid.stl file(s), optional

• Machine = *.mtt file. Additive Print uses version 1.06 of the *.mtt file specification.

Using Sisma Build Files

A build file for a Sisma machine should be a zip file containing:

• Part = *.stl file

• Support = *_vless.stl and/or *_solid.stl file(s), optional

• Machine = *.wza file. Additive Print uses version 3.0.9 of the *.wza file specification.

Using SLM Solutions Build Files

A build file for an SLM Solutions machine should be a zip file containing:

• Part = *.stl file

• Support = *_vless.stl and/or *_solid.stl file(s), optional

• Machine = *.slm file. Additive Print uses version 1.10 of the *.slm file specification.

Using Trumpf Build Files

A build file for an Trumpf machine should be a zip file containing:

• Part = *.stl file

• Support = *_vless.stl and/or *_solid.stl file(s), optional

• Machine = *.wza file. Additive Print uses version 3.0.9 of the *.wza file specification.

Deleting Parts, Supports, and Build Files

Parts, supports, and build files can all be deleted from their individual details pages. However, if the
part, support, or build file you want to delete has already been used in a simulation, you'll need to first
delete the simulation in which it is used. For convenience, a list is provided of the simulations that are
affected. The Delete button deletes all the affected simulations and the part, support, or build file.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
48 of ANSYS, Inc. and its subsidiaries and affiliates.
 Deleting Parts, Supports, and Build Files

Other considerations to be aware of include:

• Deletion is permanent and data is not recoverable.

• You cannot delete a part, support, or build file that is currently being used in a running simulation.

• Deleting a support does not automatically delete the associated part.

Deleting Support Groups

You can find a list of support groups for a particular part in that part's details page. Deleting a support
group results in slightly different behavior than that described for parts, supports, and build files:

• Click the Archive button next to a support group to delete it.

• You can archive support groups within completed and running simulations.

• You can restart a simulation even if you have archived the support group within it and the
support group will be intact.

• If you duplicate a simulation with an archived support group, the support group will not be in-
cluded in the duplicated simulation.

• If you delete a support that is included in a support group, it will be deleted from the support
group. If a support group has only one support left after a support deletion operation, the support
group will also be deleted.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 49
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
50 of ANSYS, Inc. and its subsidiaries and affiliates.
Set Up a Simulation - Assumed Strain
Set Details
Select Geometry
Define Support Options
Select Material
Select Outputs
Save/Export a Draft Simulation

Once you have added a part into the Parts Library (or a build file into the Build File Library), you are
ready to begin setting up a simulation. You will use a simulation form to specify the criteria necessary
for a simulation, including the part and its material and stress behavior, support options, and the desired
output options of your simulation.

We will begin by performing an Assumed Strain simulation. This is the simplest and fastest simulation
type. Most of the following steps described for an Assumed Strain simulation are also required for the
other simulation types.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 51
Set Up a Simulation - Assumed Strain

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
52 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Geometry

Set Details
Details include Simulation Title, Tags, Description, and Number of Cores.

Simulation Title, Tags, and Description

The title (required), tags, and description fields allow you to identify the simulation in some way that
makes logical sense to you. These fields are searchable at any location in the program that allows for
searching on text fields. Tags should be at least three characters long.

Number of Cores
To take advantage of High-Performance Computing, the Additive Desktop application allows you to
specify multiple processor cores. Depending on your Additive license, you may have up to 12 cores to
use. The default is 4.

See A Note About Data Noise with Parallel Processing (p. 18).

Select Geometry
You select a part for simulation by adding it to your simulation form. Regardless of whether you add a
part or a build file, it must have been imported first to the Parts Library or Build File Library, respectively.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 53
Set Up a Simulation - Assumed Strain

Voxel Size and Voxel Sample Rate

Voxel Size
Upon adding a part to the simulation form, you will see a preview of that part as well as a summary
of the part’s overall dimensions in millimeters in x, y, and z coordinates, a minimum Voxel Size recom-
mendation, and an estimate for memory usage.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
54 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Geometry

You will need to specify a Voxel Size and a Voxel Sample Rate to be used for your simulation. A voxel
is a hexahedral (cubic) element used in the finite element method. In the following figure of a voxel,
Voxel Size is the length of the yellow line.

When combined, voxels define the domain of the geometry. Minimum Voxel Size is the estimated
Voxel Size that can safely run without the simulation risking failure due to insufficient memory. This
is calculated automatically when the part is imported. Voxel size defaults to 0.5 mm.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 55
Set Up a Simulation - Assumed Strain

Voxel Sample Rate

The voxelization function divides the domain into voxels for simulation in the Mechanics Solver. The
technique employs subvoxels within each voxel to better represent geometry, in particular at edges
and curves. A user input parameter, Voxel Sample Rate , allows control over the number of subvoxels
to be used, ultimately controlling the trade-off between voxelization time and accuracy of geometry
representation. A higher number than the default of 5 may sometimes be necessary to adequately
capture fine supports and part features.

The following figure illustrates how subvoxels are used in defining the edges of geometries.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
56 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Geometry

How Do I Determine Voxel Size?

Currently, the calculated Minimum Voxel Size assumes a Minimum Overhang Angle of 45 degrees
and a Minimum Support Height of zero. (See Define Support Options (p. 61) for further details about
supports.) A smaller angle could result in fewer support voxels and a Minimum Support Height > 0
will include more support voxels. In both cases, memory requirements will change, and a larger Voxel
Size may be required for a successful simulation.

Generally, there should be at least four voxels through the thickness of the finest feature of interest.
This is accomplished by setting the Voxel Size to one-fourth the minimum feature dimension. It should
be noted however, that for a geometry with highly disproportionate overall dimensions compared
to its finest features, some accuracy may be sacrificed in the fine features to obtain a shorter run time
by applying the above rule to a thicker area of the part.

How Does Voxel Size Affect Run Time?

Decreasing Voxel Size by any factor can exponentially increase the solution run time by that factor
raised to the power of 4 (assuming a uniform part, like a cube).

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 57
Set Up a Simulation - Assumed Strain

As an example, if the starting Voxel Size takes 5 minutes of run time, and then the Voxel Size is reduced
by a factor of two, the run time could be expected to increase to approximately 5*(2) 4 = 80 minutes.

Currently, Voxel Size is limited to between 0.02 mm and 2 mm. However, due to memory requirements
of smaller Voxel Size simulations, this range of Voxel Size is not guaranteed to finish the simulation.

Voxelization at Part and Support Interfaces—Handling Intrusion (Penetra-

tion)
It is common practice when additively manufacturing parts for supports to intrude, or penetrate, into
part geometry. This ensures a good connection between the part and the support. The following
steps describe how the Additive application handles the interface between part and support when
an stl support file includes intrusion of support walls into part geometry:

1. The support is voxelized at subvoxel resolution, marking the state of support subvoxels as
support material.

2. The part is voxelized at subvoxel resolution, marking the state of part subvoxels as part ma-
terial, and overwriting any support subvoxels that happen to be in the way (that is, coincident).

3. For each voxel, the resulting part subvoxels and support subvoxels are tallied up.

4. The voxel's density = (# non-empty subvoxels) / (# total subvoxels).

5. The voxel's state is considered support material if the voxel contains only support subvoxels.
Otherwise, if there is at least one subvoxel marked as part, the voxel's state is considered
part material.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
58 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Geometry

So where there is overlap, there is no double-counting of density; the part material takes precedence
over the support material. This is relevant because, as discussed later, the support material's strength
is multiplied by a knockdown factor called Support Yield Strength Ratio (p. 67).

Hanging/Floating Voxels
We will discuss supports (p. 61) in more detail in the next section, but be aware of one scenario related
to geometry that you should try to avoid—having an overhanging "stalactite" point, line, or surface
that is unsupported. In the following figure, the middle section of the M-shaped geometry is not
supported. When the appropriate layer is reached and the laser travels the scan vectors for the bottom
of the M, an "island" of solid is formed surrounded by powder. This is undesirable for two reasons:

• When you attempt to print this part on an LPBF machine, the layers at the center of the M
will likely get wiped out by the recoater blade with each new pass because there is nothing
holding the solidified layer in place on the powder.

• In the numerical simulation, the unsupported area at the center of the M will result in what is
known as hanging and floating voxels, with no connection or support from voxels that touch
the base. The Mechanics Solver will accommodate for this by not applying strains to any
floating voxels until they are no longer floating, that is, until they are connected to other voxels
that are connected to the base. You will see a message in your logs indicating the number of
hanging/floating elements.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 59
Set Up a Simulation - Assumed Strain

Known Issues and Limitations with Voxelization

Occasionally, the voxelizer may produce extra rows of extremely low-density voxels. Depending on
the Voxel Sample Rate chosen, the density can range from 0.04 down to 0.001. They can occur on
both parts and supports. These extra voxels are due to the numerical error inherent in fixed-precision
floating-point numbers. It is difficult to predict if, or where, these extra voxels might occur, but they
are more easily seen on large, flat faces that are neither parallel nor perpendicular to the baseplate.
If these extra voxels occur and pose a potential problem to the simulation, try re-running voxelization
with a higher Voxel Sample Rate or a different Voxel Size.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
60 of ANSYS, Inc. and its subsidiaries and affiliates.
 Define Support Options

Define Support Options

We're going to take a moment here to summarize what we know about supports and then explain how
they are handled in the additive simulation. During Laser Powder Bed Fusion, as the laser passes over
each layer of metal powder it creates a melt pool similar to a welding process. The melt pool area cools
and is reheated again in the next laser pass. With each successive layer, the material underneath cools
and contracts. This process of heating and cooling, expanding and shrinking, results in strain, distortion,
and residual stress in the part that effectively act to lift the part off the baseplate. Therefore, support
structures are required to hold printed parts in place during fabrication.

These support structures are commonly thin “walls” printed along with the part that are fixed to the
baseplate and connect to the part at areas of the geometry that overhang the main body of the part.
The supports are printed of the same metal material as the part and must be cut or machined off upon
completion of the build. Too many supports, or support walls that are too thick, will require excessive
post-build time to remove. Too few supports, or support walls that are too thin, may not be strong

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 61
Set Up a Simulation - Assumed Strain

enough to hold a part in place and may result in cracks, excessive distortion, drooping between walls,
or breaks.

In an Additive simulation, be aware that the conditions defining inherent strain are not applied to
supports, they are applied only to the solid part material. Thus, stress is not accumulating in the supports
as the layer-by-layer addition of material is simulated until solid part material is reached, at which time
the part material will cause some stress to develop in the supports.

Support Yield Strength Ratio

The Support Yield Strength Ratio (SYSR) is a factor that is used in the simulation assumptions to assign
a strength to the support material as compared to the solid material. It is used as a knockdown factor
to modify the strength of the support material. It affects both yield strength and elastic modulus of the
support material. For example, a value of 1.0 results in a support strength equal to the solid material
while 0.5 is half the strength of the solid material.

The default SYSR value depends on which type of supports you use in your simulation, as described
next.

Support Type
The Additive application offers the following options for handling supports in a simulation:

• Automatic (p. 63) creation of supports (default)

• Using a single, user-imported support .stl file (p. 67)

• Using multiple user-imported support .stl files combined into a support group (p. 68)

• Using supports included in a build file (p. 69)

• Simulation without supports (p. 69)

Select which of these options you will use with the Simulate With Supports check box and the Support
Type drop-down selection, as shown here. Depending on the part you have selected, only the options
that are appropriate will be available to choose. So, for example, if you do not have a support group
associated with your selected part, the support group option is grayed out.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
62 of ANSYS, Inc. and its subsidiaries and affiliates.
 Define Support Options

Using Automatic Supports

By default, the Additive application creates supports automatically for your part. It is important to
understand how these automatic supports are implemented. The application simulates the build
process with an initial set of supports based upon geometry considerations only (that is, a user-specified
overhang angle), and then generates two new sets of optimized supports based upon the simulation
results. The initial supports, called Uniform Volumeless Supports, are thin, single-bead width support
walls placed uniformly underneath overhang areas defined by Minimum Overhang Angle. The max-
imum residual stresses that supports must withstand are predicted in the simulation (in the Mechanics
Solver). The optimized support structures are then automatically generated (in the support generation
module) based upon an algorithm that varies the support density to carry these maximum residual
stresses. Two sets of optimized supports are generated:

• Optimized Volumeless Supports are of a uniform wall thickness (single-bead width), but wall
spacing is varied such that more walls are placed in regions of higher residual stress and fewer
walls in regions of lower residual stress.

• Optimized Solid Supports are uniformly spaced walls with varying thicknesses such that
thicker walls are placed in regions of higher residual stress and thinner walls in regions of
lower residual stress.

On your simulation form, you will need to specify certain parameters that guide the support generation
process.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 63
Set Up a Simulation - Assumed Strain

Minimum Overhang Angle (°) - The overhang angle is measured from the powder bed surface (ho-
rizontal = 0 degrees) up to the surface of the part. Any point on the surface of the part having an
angle less than the Minimum Overhang Angle will be supported. The default Minimum Overhang
Angle is 45 degrees. Avoid using a value that is the same as the angle of the geometry of your part,
as it can cause asymmetric support structures due to finite rounding errors. For example, if your
geometry includes an overhanging feature of precisely 45 degrees, use 46 or 44 degrees for Minimum
Overhang Angle.

Supports will be created for overhang areas even in cases where the supports cannot reach the
baseplate because a portion of the part is in the way. In that case, supports will span part-surface to
part-surface. We call these part-to-part supports.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
64 of ANSYS, Inc. and its subsidiaries and affiliates.
 Define Support Options

Minimum Support Height (mm) - This is the height, in millimeters, that the part will be elevated off
the baseplate. For example, if you set a value of 3 mm then the part will be elevated such that the
lowest point on the part is at least 3 mm above the baseplate. This value should be set to allow for
an easy part cutoff from the baseplate while also considering how many voxels must be created to
add that additional height. (More voxel layers = more simulation time.) We recommend that this value
be set as low as is realistic for each simulation. The default value is 0. (See Simulating Without Sup-
ports (p. 69))

Support Factor of Safety - The Support Factor of Safety is a parameter that drives the strength of
the automatically generated optimized support structures. If you would like the supports to withstand
2x the expected load, then you would enter a 2 in this field and the predicted strength of the auto-
matically generated support structure would be double the predicted stress. The strength of the
support structure is driven by the number and thickness of support walls that are generated. The
default Support Factor of Safety is 1.

Support Yield Strength Ratio - The default SYSR (p. 62) value for the initial (geometry-based)
automatic supports is 0.4375. This default was determined by studies where the support strength for
default supports built on an EOS M270 machine were tested and compared to solid material built on
the same machine.

Volumeless Support Parameters

Wall Thickness (μm) - A parameter used for the optimized volumeless supports. It is the wall thickness
of the generated support walls.

Maximum Wall Distance (μm) - Maximum Wall Distance is a parameter used for the optimized
volumeless supports. It is the allowed maximum distance between two neighboring support walls.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 65
Set Up a Simulation - Assumed Strain

Regardless of the predicted stress level in the support structure, the walls in supported regions will
be spaced not more than this value. Too large of a wall distance might result in failures such as the
part breaking away from the support or the development of cracks in the support structure. When a
laser scans a relatively large area of powder where the support wall distance is too wide, cracking
might happen since powder has no strength to hold the solidified part in place. The excessive distortion
might cause blade and part collision. We recommend that Maximum Wall Distance should not exceed
2 mm when a volumeless (single bead) support wall is used.

Solid Support Parameters

Minimum and Maximum Wall Thickness (μm) - Minimum and Maximum Wall Thickness are para-
meters used for the optimized solid supports. Minimum Wall Thickness is the thinnest possible support
wall that the machine will build. Usually you will specify the thickness of a single bead scan. The default
value is 100 microns. The thickness of support walls will not exceed the Maximum Wall Thickness.
The default value is 1 mm.

Wall Distance (μm) - A parameter used for the optimized solid supports. It is the distance between
support walls.

The following figure shows the bevel gear example with all three sets of supports generated with
the automatic supports option using default settings. A close-up of the optimized volumeless supports
is also shown.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
66 of ANSYS, Inc. and its subsidiaries and affiliates.
 Define Support Options

Disabling Support Optimization

If you don't care about optimized supports, clear the Generate Optimized Supports check box to
disable the feature. This will result in shorter run times.

The time to perform the support optimization function is related to the number of triangles (facets)
in your geometry. The more triangles, the longer support optimization will take. This may be especially
noticeable if you choose the distortion compensation output option that produces a distortion com-
pensated stl file with more triangles than the original stl file.

Using Support .stl Files

Choose Support STL from the Support Type drop-down to use a support file that has already been
imported for the part (p. 40). (This menu option will not be available for selection if you have not
selected a part with associated .stl files.)

Support - Choose the desired support .stl file from the ones associated with your selected part. Use
support groups (p. 68) if you want to use more than one support .stl file in a simulation.

Support Yield Strength Ratio - The default SYSR (p. 62) value for .stl supports is 1. The voxelization
process will account for the thin nature of the support walls so an additional knockdown factor is
not needed.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 67
Set Up a Simulation - Assumed Strain

Using Support Groups

Choose Support Group from the Support Type drop-down to use multiple support .stl files that have
already been imported for the part and combined into a group (p. 42). (This menu option will not
be available for selection if you have not selected a part with an associated support group.)

Support Group - Choose the name of the support group from the ones associated with your selected
part.

Support Yield Strength Ratio - The default SYSR (p. 62) value for support groups is 1. At this release,
you can specify only one SYSR, even for support groups with mixed support .stl types (volumeless
and solid).

Volumeless STL Thickness (μm) - This parameter appears if one or more of the supports in the group
is a volumeless type. Enter the thickness of a single bead scan. The default value is 100 microns.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
68 of ANSYS, Inc. and its subsidiaries and affiliates.
 Define Support Options

Using Supports in Build Files

Choose Build File Supports from the Support Type drop-down to use the supports in your build file.
(This menu option will not be available if you have not selected a build file in Geometry Selection.)
All the supports included in the build file will be used in the simulation, even if they are in separate
.stl files within the .zip file.

Support Yield Strength Ratio - The default SYSR (p. 62) value for supports in build files is 1. At this
release, you can specify only one SYSR, even for build files with mixed support .stl types (volumeless
and solid).

Volumeless STL Thickness (μm) - This parameter appears if one or more of the supports in the build
file is a volumeless type. Enter the thickness of a single bead scan. The default value is 100 microns.

Simulating Without Supports

To simulate without supports, simply clear the Simulate With Supports check box on the simulation
form. However there are some subtle assumptions to be aware of, depending on the outputs that
you select. The scenarios are described in the following table.

Minimum
Simulate Displacement
Support
with After Cutoff
Height What Happens? Part
Supports (output
(support
(check box) selection)
option)
No supports generated
N/A (0) Bottom nodes are rigid
(no displacements)

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 69
Set Up a Simulation - Assumed Strain

Minimum
Simulate Displacement
Support
with After Cutoff
Height What Happens? Part
Supports (output
(support
(check box) selection)
option)
One layer of support voxels
added to simulate part after
cutoff
N/A (0) No other supports added
anywhere
Support Yield Strength Ratio
set to 1
Supports generated
below Minimum
Overhang Angle
0
Bottom nodes are rigid
(no displacements)
Supports generated below
Minimum Overhang Angle

One layer of support

0 voxels added to simulate
part after cutoff

Support Yield Strength

Ratio set by user
Supports generated below
Minimum Overhang Angle
Non-zero,
for 3mm of voxel layers
example, created
3 mm
Support Yield Strength
Ratio set by user

Select Material
You may choose from standard ANSYS pre-defined materials or you may customize your own material.
Upon selecting a material from the drop-down selection, the properties of Elastic Modulus (in GPa),
Poisson Ratio, and Yield Strength (in MPa) for that material automatically populate the fields on the
form. These values are for materials at room temperature. To see the other properties associated with
a given material, or to customize a material, you will need to bring up the Materials Library. See Cus-
tomizing a Material (p. 74).

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
70 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Material

Linear Elastic versus J2-Plasticity Stress Mode

Once you choose a material, you have the option of choosing material behavior in calculations of
stress that is either linear elastic or elastoplastic (exhibiting both elastic and plastic properties). The
elastoplastic calculations are based upon the J2 (von Mises) plasticity model.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 71
Set Up a Simulation - Assumed Strain

The stress mode option is associated with a material’s ductility, a measure of a material's ability to
undergo significant plastic deformation before rupture. The following figure shows stress-strain curves
for a typical metal material. After yield, for a given strain, A, in the plastic deformation region, notice
that the stress at point B (fully linear elastic) is higher than the stress at point C (elastoplastic). Stress
values differ depending on your assumptions about material behavior.

In the Additive application, an assumption of linear elastic behavior will result in higher stress values
for a given strain beyond the yield point of the material. This over-prediction may not be realistic for
parts with larger strains. The simulation will run faster, however, which may be beneficial if you care
about on-plate distortions only (because you will heat-treat the final part to relieve residual stress,
for example). It is important to note that while stress values beyond the elastic range will be artificially
high, on-plate distortion values will generally be correct using the linear elastic option. Therefore, using
linear elastic stress mode can be useful for analyzing distortion trends while the part is still on the
baseplate.

An assumption of elastoplastic behavior (using the J2-plasticity model) applies best to ductile mater-
ials, such as most metals. Currently, small deformation plasticity has been used in these models where
addition of elastic and plastic strains amount to total strain, since metals do not exhibit the large
deformations we see in polymers, for example. Von Mises stresses are used to reduce the stress levels
when strain values exceed elastic strain. Strain hardening algorithms are included in the stress calcu-
lations (see Hardening Factor (p. 73)).

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
72 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Material

The simulation will run longer with the J2-plasticity option, but this option is required if you want
accurate distortion after-cutoff results, or accurate indications of stresses and strains.

Load Stepping Type

If you select the J2-plasticity option, you'll need to specify a Load Stepping Type. To improve conver-
gence, a total load for each layer will be applied in incremental steps in one of the following ways:

• Dynamic Load Stepping (default): The full load will be applied initially, and the solver will it-
erate until equilibrium is achieved. If not achieved with the initial load, it will be halved and
repeated. If equilibrium is achieved, the next incremental step is applied at the current load
fraction until applying the full load, otherwise, it is halved again. A lower limit of 1/(200) load
fraction is enforced, after which the solution will terminate.

• Fixed Load Stepping: Fixed load stepping divides the load into a user-defined number of
load steps.

– Number of Load Steps: The number of increments that a given load will be divided into for
plasticity. A larger number of load steps will require more loading calculations, but total
time may or may not increase due to potential improvements in convergence. (Valid range
is between 1 and 200. Defaults to 8.)

Hardening Factor
If you select the J2-plasticity option, a material-specific strain Hardening Factor is used in stress calcu-
lations to provide further information about the material’s behavior in the plastic deformation region.
The Hardening Factor is used to calculate the slope of the stress-strain curve (Et) above the material’s
Yield stress:

This factor may be changed here or when customizing a material.

Note: Previous to Release 19.2, a hardening factor of 0.1 was used for all materials. Following the 19.2
update, the default materials each have their own hardening factor. Custom materials created by the user
prior to this change will use 0.0198 as the hardening factor. A consequence of this change is that simulations
run from Release 19.2 forward may have slightly different output values than those run with previous re-
leases. The magnitude of the difference depends on a variety of factors including part geometry and ori-
entation, material, scan pattern, laser power and whether supports are used in the simulation.

Strain Scaling Factor

The Strain Scaling Factor (SSF) is a calibration factor used to account for differences in machines that
you may use to improve the accuracy of your simulations. This value is a direct multiplier to the pre-
dicted strain values. Using a value of 1 will result in strain magnitudes as calculated by the solver.
Some material and geometry combinations result in bulging/expansion rather than shrinkage and so
a negative SSF is possible. Values between -1 and 1 will reduce displacement and stress while values
outside of that range will amplify them. Using a value of 0 will result in no strain and the final dis-
placement will match the input geometry. The default Strain Scaling Factor is 1.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 73
Set Up a Simulation - Assumed Strain

You should use the default SSF of 1 for your first simulation in Additive and then work from there to
calibrate the best value of SSF. Because these are single data point values and there can be variations
even from batch to batch of material, we recommend that you calibrate for SSF for each specific
machine and material combination. Calibration procedures are available online here.

It is not necessary to perform this calibration if you are conducting a trend analysis, that is, if you will
be examining the effects of variable changes on stress or distortion relative to each other.

Customizing a Material

Click the Materials button in the left panel of the dashboard to bring up the Materials Library.
There you will see a list of ANSYS pre-defined materials, as well as any of your own customized ma-
terials.

Throughout this User's Guide, we will distinguish between customized materials and user defined
materials as follows:

• Customized materials refers to those materials that have simple edits to one or more factors.
Typically you may want to edit an ANSYS pre-defined material to modify the Strain Scaling
Factor after you run a calibration process for that material, for example.

• User defined materials refers to your own new materials that have been tuned in a much more
elaborate procedure and that are typically temperature dependent. See Creating User Defined
Materials (p. 155) for more information on user defined materials.

Selecting any of these materials will bring up a panel of detailed properties for that material. Click
Customize to create a new material entry based upon one of these pre-defined materials. Note that
you cannot edit an ANSYS pre-defined material, but you may edit your own customized materials.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
74 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Material

When you select a custom material to use in your simulation, the custom values of the material
properties will automatically populate the appropriate fields in the simulation form.

To remove your customized materials from the Materials Library, bring up the details page for that
material and click archive. Simulations using these materials will remain in your stored application
data.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 75
Set Up a Simulation - Assumed Strain

Select Outputs
The powerful features of ANSYS Additive leading to the most useful insights for users are initiated by
simple output option check boxes. On a simulation form under Outputs, there are several options from
which to choose, depending on your simulation goals. These options may affect simulation run-time
but provide additional output files that will be available under Completed Simulations when the simu-
lation is complete. In some cases, additional inputs are required.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
76 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Outputs

On-Plate Residual Stress/Distortion

The basic output produced for all simulations (it appears grayed out because it is the default set of
results) contains a voxelized representation of the part with predicted residual stresses and displace-
ments at the end of the build while the part is still attached to the baseplate. Both end-state stresses
and maximum stress during the build are contained in this .vtk file. Results may be viewed directly
in ANSYS Viewer or exported. Additional options allow you to include strain components in the on-
plate results file and to generate a distortion compensated .stl file.

Strain Components
If you selected J2-plasticity stress mode, check the "Include strain in output files" check box if you
want to see the strain components, both elastic and plastic, in the on-plate .vtk file. This option
will increase both your memory consumption and disk space for the results file.

Distortion Compensated .stl File

The Additive application can predict the location and magnitude of distortion and then "reverse"
distort the original .stl file. Then when you build your part using the compensated geometry, the
result will be closer to the original design.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 77
Set Up a Simulation - Assumed Strain

On a simulation form under Outputs (either on-plate or after cutoff ), check the box for Distortion
compensated .stl file. Once selected, you have the option to specify scale factors. The Scale Factor
will change the magnitude of the distortion applied to the original .stl file. A Scale Factor of 1 (de-
fault) will create an .stl file with distortion compensated by the same magnitude as the simulated
results. A Scale Factor < 1 will compensate less than the simulation-predicted magnitude and a
value > 1 will compensate more than the simulation-predicted magnitude.

You may enter Multiple Scale Factors to determine which distortion compensation factor works
best for the part. Values must be between -5.0 and 5.0.

Once the simulation is complete, files labeled Compensated Geometry (and Compensated Geometry
(after cutoff )) can be found in the Output Files section under Completed Simulations. The Scale
Factor is used in the file name to distinguish results among multiple Scale Factors. Click the export
link to get the distortion compensated .stl file. A second output is also created named Geometry
with Distortion (and Geometry with Distortion (after cutoff )). This is a geometry representation in
.vtk format with displacement vectors with each vertex.

How Does Distortion Compensation Work?

The Additive application simulates spatial distortion that occurs during the build process. With
Distortion Compensation activated, each vertex in the Compensated Geometry .stl file is moved in
the opposite direction of the closest distortion vector (that is, new position = original position +
deformed (dX, dY, dZ) x -1). To ensure the .stl file has fidelity on par with the simulation, any triangle
with an area greater than an equilateral triangle with sides equal to the selected Voxel Size is split
into smaller triangles.

Since the Distortion Compensation feature is a simple linear assumption, it may not always be
correct. Parts can respond non-linearly. For example, if you use a Scale Factor of 1 (default) in a
simulation and then use the compensated file to rerun the simulation, you will see if the part is
predicted to distort to the correct shape. If not, you’ll know that the compensated .stl file was under-
compensated or overcompensated and you can adjust the Scale Factor up or down from there.
Generally, the Distortion Compensation feature tends to be an iterative process. In fact, our experience
shows good results using two iterations with Scale Factor = 0.5 for each simulation. Alternatively, a
good starting point for one iteration is a Scale Factor = 0.75.

Displacement After Cutoff

In addition to simulating the additive build process, you can simulate removal of the part and/or
supports by checking the Displacement After Cutoff output option. Cutoff Mode determines what
will be cut off whereas Cutoff Method determines how it will be cut off. The various option combina-
tions are illustrated at the end of this section (p. 79).

Simulation results will include a voxelized representation of the part with predicted displacements
after the cutoff occurs.

Cutoff Mode
Use the Cutoff Mode drop-down to select what elements of the model will be removed:

• Part and Support Cutoff - The part and supports are cut from the baseplate and cut from
each other at the part-support boundaries. If the part touches the baseplate, one layer of
support voxels is added below the part to avoid cutting the part.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
78 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Outputs

In the case where you have chosen to simulate without supports (that is, you have unchecked
Simulate with Supports in the Supports section of the simulation form) and you choose Dis-
placement After Cutoff/Part and Support Cutoff, one layer of support voxels will be added to
the base of the part to simulate part after cutoff so as to avoid cutting the part. (See Simulating
Without Supports (p. 69).)

• Support-only Cutoff - The support voxels are cut at the part-support boundaries, separating
the part from the support, but not the part from the baseplate. This option is available only
if the Minimum Support Height is less than the Voxel Size, otherwise it is disabled. This ensures
that there are part voxels touching the baseplate, thereby keeping the part attached to the
baseplate even after support removal.

Cutoff Method
Use the Cutoff Method drop-down to select how the cutoff will occur:

• Instantaneous (default) - This method uses the Mechanical APDL solver to simulate instantan-
eous cutoff of part and support, or support-only, depending on the Cutoff Mode.

• Directional (baseplate only) - This method uses the Mechanical APDL solver to simulate a
progressive cutoff at the baseplate, that is, at the bottom layer of elements only, in which you
specify the number of step increments and the direction for removal from the baseplate.

– Cutoff Direction - The direction in which the support and/or part will be progressively
removed. Choose either the +X, -X, +Y, or -Y direction. Refer to the triad legend for
directional perspective.

– Cutoff Steps - The number of incremental steps taken to complete the directional cutoff.

• Legacy - This method is visible only if you have Legacy Options = Show under Settings in the
menu bar (p. 19). Legacy uses the Mechanics solver to simulate cutoff scenarios rather than
the default Mechanical APDL solver.

Note that if you Restart an old simulation that used the Mechanics solver for cutoff (releases
prior to 2020 R2), the Legacy option will be set automatically and the Mechanics solver will
be used. Use the Duplicate button on the original completed simulation to bring up the sim-
ulation form if you want to change cutoff options.

Illustrations of the various option combinations are provided here.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 79
Set Up a Simulation - Assumed Strain

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
80 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Outputs

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 81
Set Up a Simulation - Assumed Strain

Layer by Layer Stress/Distortion

Choose this option for detailed voxel layer-by-voxel layer results to learn insights about the behavior
as the part is being built. Results are for the surface of the part, not the interior information. The
output consists of a layerwise .zip file containing a series of .vtk files, one for each voxel layer in your

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
82 of ANSYS, Inc. and its subsidiaries and affiliates.
 Select Outputs

part, as layerwise files are not written for support-only layers. Use these files to “animate” the build
process. View locations throughout the part of potential blade crashes and high strain areas that may
indicate cracks.

The disk space used for this output can get very large. To prevent filling up your disk, or loss of im-
portant layer data, consider managing storage using the maximum storage option.

Maximum storage used for layer by layer VTK files (GB)

The cumulative maximum storage for layer by layer .vtk files (before zipping). Once this limit is reached,
subsequent .vtk layer files will not be written, however the simulation will continue and output files
for other selected outputs will be written, as needed. The default is 20 GB.

Files for Transfer to ANSYS Mechanical

Check this option to write out files required by ANSYS Mechanical if you want an alternate method
to perform cutoff, or if you want to simulate a prescribed cutoff sequence in Mechanical. One zipped
file, sim#-mapdl.zip, is written and it includes the following individual files:

• Model.cdb: A file containing mesh, material, and named selection information. (In Mechanical,
named selections allow you to group similar geometry or meshing entities.)

• States.ist: A file containing Additive Print's on-plate end-state information (elastic strains, plastic
strains, and accumulated equivalent plastic strain). This will provide the initial stress state for
subsequent processing in Mechanical.

• Knockdowns.ist: A file containing element (voxel) density information.

• Displacements.csv: A file containing Additive Print's on-plate end-state displacements.

For details on how to use these files in ANSYS Mechanical, read our additional instructions here.

Blade Crash Detection

Considering the amount of distortion that is possible within a part being built, a scenario of concern
is a collision of the recoater blade and the distorted part. The Additive application provides a blade
crash detection feature that predicts, based primarily on part distortion, if and where such a collision
may occur. Physical blade crash is dependent on other factors, as well, such as scanning procedure,
temperatures, gas flows, part orientation, and materials, among others, so it is important to understand
that our calculation is a detection of potential blade crash.

On a simulation form under Outputs, check the box for Detect potential blade crash due to distortion.
Once selected, you have the option to specify a Threshold Scaling Factor and Layer Thickness. (Note
that for Scan Pattern and Thermal Strain simulations, the Layer Thickness parameter appears in the
Machine section of the simulation form.)

Once the simulation is complete, a file labeled Potential blade crash locations can be found in the
Output Files section under Completed Simulations. Click the export link to get the .csv file. Indications
of blade crash are also available on the On-plate stress/displacements .vtk and .avz files and the Lay-
erwise .vtk files.

Blade crash potential is calculated and each voxel is assigned a 0, 1, or 2 as follows:

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 83
Set Up a Simulation - Assumed Strain

• 0 (none predicted)

• 1 = warning/potential blade crash — when positive Z displacement is greater than (Layer

Thickness x Threshold).

• 2 = critical/likely blade crash — when positive Z displacement is greater than (Layer Thickness
x Threshold + ½ Layer Thickness). The additional ½ Layer Thickness is an approximation to
take into consideration the actual offset distance of the recoater blade needed to deposit a
given layer thickness.

Layer Thickness (10-100 μm)

Layer Thickness is the thickness of the powder layer coating that is applied with every pass of the
recoater blade. The default value is 50 microns. We recommend that you use the actual thickness
specific to your machine.

Threshold Scaling Factor

This value is used to modify blade crash calculations so that you can allow for flexibility in the recoater
blade.

Using the default value of 1 and a Layer Thickness value of 50 μm, any displacement in the positive
Z direction over 50 μm will be marked as a warning (potential blade crash) and any displacement
over 75 μm will be marked as a critical area (likely blade crash). In another example, if you know there
is not a lot of flexibility in the recoater blade, use a threshold value of 0.8. With a 50-micron Layer
Thickness, any displacement in the positive Z direction over 40 microns will be marked as a warning
and any area with displacement greater than 65 microns will be marked as critical.

High Strain Areas

When the strain in a part exceeds the elongation a material can withstand, a failure can occur resulting
in cracking throughout the part or supports. This is a common issue that can affect parts built with
additive manufacturing. The Additive application allows you to quickly look at the design and process
settings that would alleviate potential cracking. The High Strain Areas feature allows you to identify
regions of the part that may be prone to forming cracks during or after the build process by highlight-

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
84 of ANSYS, Inc. and its subsidiaries and affiliates.
 Save/Export a Draft Simulation

ing critical strain values. Required inputs for this output type include the Support Strain Threshold,
Part Strain Threshold, and Strain Warning Factor.

Strain threshold values should be entered as a percent for the support and part material respectively.
(Engineering strain can be calculated as the change in length divided by the original length. For ex-
ample, a 3.0" titanium bar that has been stretched to 3.3" is said to have experienced a tensile strain
of 0.1, or 10 percent.) The default values of 10 and 20 percent strain are simply sample values, and
you should adjust these values as needed for different materials or other factors that may affect the
total elongation of your build material. When a calculated strain exceeds these threshold values, it
will be labeled as “Critical.”

The Strain Warning Factor allows you to establish a “Warning” range, to identify further areas where
strain is approaching the critical range.

Support Strain Threshold (%)

Percentage strain in the supports above which strain will be considered critical. Defaults to 10%.

Part Strain Threshold (%)

Percentage strain in the part above which strain will be considered critical. Defaults to 20%

Strain Warning Factor

This value is multiplied by each of the strain thresholds above to define limits where strain is labeled
as a warning. Defaults to 0.8.

Using default values for all inputs, strains in the supports between 8 and 10 percent will be in the
warning range, while strains over 10 percent will be considered critical. Strains in the part between
16 and 20 percent will be in the warning range, while strains over 20 percent will be considered
critical.

The output of High Strain Areas can be viewed as part of the On-plate stress/displacement .vtk and
.avz files and the Layerwise .vtk files, as well as a High strain regions .csv file. When viewed through
the On-plate stress/displacement output, critical strain locations are given a value of 2 and warning
locations are given a value of 1. All other locations retain a value of 0 to show low risk. The High
strain regions .csv file consists of strain values for all points with strains at or above the warning
threshold. Information about each point includes the x, y, and z locations along with the strain value
and the deposit layer. The deposit layer represents the actual powder layer during a build. When using
the Assumed Strain analysis type, a layer thickness of 50 μm is used to identify the deposit layer of
each location.

Save/Export a Draft Simulation

You may want to save your simulation form periodically before starting the simulation. While it is not
saved to a specific file on your computer, it is saved internally, and you will see it listed under Draft
Simulations. It is removed from Draft Simulations when you start a simulation (that is, when it is no
longer a “draft”). All your input options are stored when you run a simulation so that you may see your
options at any time when you select a simulation in the Running Simulations and Completed Simulation
areas of the dashboard.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 85
Set Up a Simulation - Assumed Strain

To save your inputs to a file, use the Export button. (A Save action is required before you can Export.)
Exported files have an .aasp extension and may be imported using the Import button under Draft
Simulations. Exported files do not include the part.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
86 of ANSYS, Inc. and its subsidiaries and affiliates.
Run a Simulation - Assumed Strain
Start a simulation by clicking on the Start button at the bottom of your simulation form. You will imme-
diately see status activity in a convenient summary format.

How Long Will My Simulation Take?

ANSYS Additive is fast! For simple simulations, your run time will take minutes or even just seconds.
For more complicated geometries and simulation options, the program may run for many hours, or
even days, but you will see status activity in the log indicating progress throughout the simulation.
Keep in mind these general considerations:

• A simulation with J2-plasticity (stress mode) will take slightly longer than a simulation assuming
linear elastic stress behavior. Stress results will be more accurate. See Linear Elastic versus J2-
Plasticity Stress Mode (p. 71).

• A Scan Pattern simulation will take slightly longer than an Assumed Strain simulation. See Scan
Pattern Strain - Anisotropic (p. 23).

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 87
Run a Simulation - Assumed Strain

• A Thermal Strain simulation will take much longer than any other type of simulation, but it
provides the highest level of accuracy in results. See Thermal Strain - Anisotropic (p. 23).

• Too small of a voxel size can increase simulation time significantly. See How Does Voxel Size
Affect Run Time? (p. 57)

• A part oriented with the longest dimension in the Z direction will take longer during simulation
than parts oriented with their longest dimension in the plane of the baseplate. While the domain
volume doesn’t change with part orientation, the number of voxel layers changes. See Guidelines
for Part Orientation and Resolution (p. 37).

• Run time will get slower per voxel layer as the simulation proceeds through the voxel layers.
There are more calculations being performed for each new layer as the simulation progresses.

• Simulation time does not equal build time but it is related. If the time to build your part takes
a week, it is reasonable to expect the simulation to take a few days.

Common Error Messages During a Simulation

There may be occasions when a simulation fails and cannot continue. If that is the case, a warning
message will pop up on your desktop and you will get an error message in the activity log. The following
table lists a common error message related to the number of iterations in the simulation and recom-
mended actions.

Error Message:

Invalid state: Hit maximum iterations (12000) with a residual of 10071.14…

Cause Suggested Solution
Too coarse of a mesh, leading to poor Try reducing voxel size to improve connectivity of the
connectivity between elements (that mesh.
is, elements only sharing
corners/edges).
As the number of iterations required Try increasing voxel size to reduce the problem size.
to reach convergence scales
nonlinearly with the problem size,
this may be indicative of too large of
a problem.
May happen in regions of a part with May try adjusting support structure settings to avoid
large unsupported cantilevered large unsupported overhangs.
overhangs, resulting in excessive local
distortion.
Convergence is dependent on the May try modifying geometry to improve
level of interconnectedness of the interconnectedness or try simulating the part in a
mesh. Thin shell parts or parts different orientation.
connected only by slender members
may not be so well interconnected.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
88 of ANSYS, Inc. and its subsidiaries and affiliates.
 Can I Run Multiple Simulations at Once?

Can I Run Multiple Simulations at Once?

You can queue multiple simulations so that one will start immediately after the one before it completes.
This is extremely convenient if you want to line up a few simulations to run overnight, for example.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 89
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
90 of ANSYS, Inc. and its subsidiaries and affiliates.
Review Results of a Simulation - Assumed Strain
Look for the Success status indication to know your simulation has completed. In the Overview and
Logs sections, you will see beginning and ending time stamps and other useful information. Simulation
results are found in the Output Files section.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 91
Review Results of a Simulation - Assumed Strain

Output Files
Results from your simulation are viewable in ANSYS Viewer, or are exportable files (.avz files are ANSYS
Viewer files), or both, as described in the following tables.

Part Output Files Type Description

Solver Voxel Input* .vtk An output file with a voxelized representation of the part and
supports showing voxel densities. This file is output early in the
simulation, after voxelization but before solution begins. If you
specified automatic supports, the supports in this file are simply
a generalized density, not specifically one of the optimized
supports. If you imported support .stl files, this file shows your
imported supports. You can see how well the part and support
geometries are represented by the voxel mesh and to confirm
that your supports are aligned correctly to the part.

• Density: A value between 0 (empty) and 1 (full) for each voxel

indicating the material density ratio of the subvoxels after
voxelization.

• State: Either part, support, or interface. Interface state means

this voxel belongs to the part but has a support voxel below
it.

On-plate .vtk Voxelized part with predicted displacements and stresses (and
Stress/Displacement* and optionally, strains) at the end of the build (that is, end state) while
.avz the part is still attached to the baseplate.

• Displacement in mm: magnitude, x, y, and z components

• End-state stress in Pa: von Mises, xx, yy, zz, xy, yz, zx
components

• End-state elastic and plastic strain: xx, yy, zz, xy, yz, zx
components (if Include strain in output files output option is
chosen)

• Max stress during build in Pa: von Mises, and xx, yy, zz in
compression & tension for each

• Blade crash severity (if Detect potential blade crash output

option is chosen)

• High strain severity (if High Strain Areas output option is

chosen)

After Cutoff .vtk Voxelized part with predicted displacements after cutoff has
Displacement and occurred (either part and support cutoff or support-only cutoff,
.avz depending on the Cutoff Mode option).

• Displacement in mm: magnitude, x, y, and z components

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
92 of ANSYS, Inc. and its subsidiaries and affiliates.
 Output Files

Part Output Files Type Description

MAPDL After Cutoff .zip A zipped file containing the results file (.rst) and common database
file (.cdb) that can be used in ANSYS Mechanical to evaluate the results
of the cutoff.
Layerwise .vtk files .vtk A series of .vtk files that show voxelized representation of part layer
(zipped) by layer during the build. You will have as many .vtk files as voxel layers
in your part, as layerwise files are not written for support-only layers.
Use these files to animate the build process. View locations throughout
the part of potential blade crashes and high strain areas that may
indicate cracks.

For each voxel layer:

• Displacement in mm: magnitude, x, y, and z components

• End-state stress in Pa: von Mises, xx, yy, zz, xy, yz, zx
components

• Blade crash severity (if Detect potential blade crash output

option is chosen)

• High strain severity (if High Strain Areas output option is

chosen)

Geometry with .vtk Contains a representation of the original, undistorted part with
Distortion predicted displacements at the end of the build while part is still
attached to the baseplate. The geometry does not include the offset
for supports between the baseplate and the part.

• Displacement in mm: magnitude, x, y, and z components

Geometry with .vtk Contains a representation of the original, undistorted part with
Distortion (after Cutoff ) predicted displacements after cutoff has occurred (either part and
support cutoff or support-only cutoff, depending on the Cutoff Mode
option).

• Displacement in mm: magnitude, x, y, and z components

Positioned Part .stl Input geometry (non-compensated) positioned into its start location
and orientation, that is, offset to account for supports between the
baseplate and the part.
Compensated Geometry .stl Contains the distortion-compensated 3D surface representation
(tessellated triangles) of the part while the part is still attached to the
baseplate. The compensated geometry is placed flush with the
baseplate surface and does not include the offset for supports between
the baseplate and the part.

You may want to use this file as the part geometry for final
production builds.
Compensated Geometry .stl Contains the distortion-compensated 3D surface representation
(after Cutoff ) (tessellated triangles) of the part after cutoff has occurred (either part

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 93
Review Results of a Simulation - Assumed Strain

Part Output Files Type Description

and support cutoff or support-only cutoff, depending on the Cutoff
Mode option).

You may want to use this file as the part geometry for final
production builds.

* = Default output for all strain simulations

Supports Output Files Type Description

Supports .vtk Contains the voxelized representation of the support structure with
Stress/Displacement and predicted displacements and stresses at the end of the build (that is,
.avz end state) while the part is still attached to the baseplate.

• Displacement in mm: magnitude, x, y, and z components

• End-state stress in Pa: von Mises, xx, yy, zz, xy, yz, zx components

• Max stress during build in Pa: von Mises, and xx, yy, zz in
compression and tension for each

View this file together with On-plate stress/displacement to see full

build together (part and supports). Note that these are the default
uniform supports, not the optimized supports. Recall that stress is
not accumulating in the supports until they are affected by the
solid part material. See Define Support Options (p. 61).
Uniform Volumeless .stl Representation of uniform thin-walled supports used in the finite element
Support simulation before support optimization occurs.

View this file to see areas of the part that need supports based on
Minimum Overhang Angle specified on the simulation form.
Compares to typical third-party software that generates supports
on part areas that meet the overhang minimum.
Optimized Volumeless .stl Contains optimal volumeless support layout based on predicted stresses
Support and distortions that can be used to minimize risk of support failure.
Optimized Solid Support .stl Contains optimal solid support layout based on predicted stresses and
distortions that can be used to minimize risk of support failure.

.csv Files Type Description

Potential Blade Crash .csv Includes locations of all potential and likely blade crashes and the
Locations predicted magnitude of the Z displacement at those points. Note that
values of zero Z displacement indicate hanging/floating voxels (p. 59).

• Global x, y, z coordinates, and total Z displacement in mm

High Strain Regions .csv List of high strain warning areas during the build.

• Global x, y, z coordinates

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
94 of ANSYS, Inc. and its subsidiaries and affiliates.
 Using ANSYS Viewer to Review Results

Files for Transfer to

Type Description
ANSYS Mechanical
Sim#-mapdl .zip A zipped file containing the four files listed below. For details on
how to use these files in ANSYS Mechanical, read our additional
instructions here.
Model .cdb An output file with mesh, material, and named selection information
for the part and supports, if any. The .cdb format is an
ANSYS-specific coded database file that is used to import an external
model into Mechanical.

In Mechanical, named selections allow you to group similar geometry

or meshing entities. The following named selections are included
in the Model.cdb file and will appear in Mechanical's project tree:

• BUILD (External Model) – Element information for the part

• SUPPORT (External Model) – Element information for the supports,

if any

• Constraints (External Model) – The boundary conditions

representing the baseplate

State .ist Predicted strains at the end of the build (that is, end state) while
the part is still attached to the baseplate. This will provide the initial
stress state for subsequent processing in Mechanical.

• End-state elastic and plastic strain tensors (xx, yy, zz, xy, yz, zx
components)

• Accumulated equivalent plastic strain

Knockdowns .ist An output file of the voxel densities of the part and supports. (These
are the same densities as contained in the Solver Voxel Input file.)
They are imported into Mechanical as knockdown factors.

• Density: A value between 0 (empty) and 1 (full) for each voxel

indicating the material density ratio of the subvoxels after
voxelization.

Displacements .csv Predicted displacements at the end of the build (that is, end state)
while the part is still attached to the baseplate.

• Displacement in mm: magnitude, x, y, and z components

Using ANSYS Viewer to Review Results

ANSYS Viewer is an interactive 3D image viewer that is embedded in ANSYS Additive. It allows you to
easily visualize simulation results in 3D models.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 95
Review Results of a Simulation - Assumed Strain

To bring up ANSYS Viewer, click “View” next to any of the outputs with that link in the Output Files
section under Completed Simulations.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
96 of ANSYS, Inc. and its subsidiaries and affiliates.
 Using ANSYS Viewer to Review Results

You can click on different result items in the View Manager to see them displayed. Use your mouse
buttons to move the part around in the image preview; left button for spin, middle for zoom, and right
button for pan.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 97
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
98 of ANSYS, Inc. and its subsidiaries and affiliates.
Performing a Scan Pattern Simulation
A Scan Pattern simulation uses anisotropic strain calculations to improve upon the assumed strain
method. That is, a Scan Pattern simulation takes into account the fact that more strain develops in the
scan direction than perpendicular to it, as shown in the following figure. Anisotropic strain is rapidly
calculated for each powder layer based on the major orientation of the fill scan vectors or the specific
scan vector files if loaded through a build file. Then these individual layer strain values are collected
and averaged to the voxel size. The predicted strain is then used for a rapid mechanics analysis.

In addition to the standard inputs as described in Set Up a Simulation - Assumed Strain (p. 51), there
are two unique sets of input required for a Scan Pattern simulation. These are the anisotropic strain
coefficients of your material, and the type and process parameters of your 3D printing machine.

Anisotropic Strain Coefficients

The Material Configuration section of the simulation form will change to include anisotropic coefficients
if you have chosen a Scan Pattern or Thermal simulation type.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 99
Performing a Scan Pattern Simulation

Anisotropic strain coefficients are used to represent anisotropic strain behavior on coordinate systems
aligned with the local longitudinal, transverse, and depth scan directions. Positive values result in
compressive base strain (contraction), whereas negative values result in tensile strain (expansion). Default
values are shown in the following table.

Anisotropic strain coefficient (||) = 1.5 Longitudinal: The predicted strain in the direction that
the laser is scanning for the major fill rasters will be
multiplied by 1.5
Anisotropic strain coefficient (⊥) = 0.5 Transverse: The predicted strain orthogonal to the
direction that the laser is scanning for the major fill
rasters and in the plane of the surface of the build
plate will be multiplied by 0.5
Anisotropic strain coefficient (z) = 1 Depth: The predicted strain in the Z direction will be
multiplied by 1

Machine Parameters for a Scan Pattern Simulation

Your options for machine parameters in a Scan Pattern simulation depend on whether you have chosen
a part file (does not include scan pattern information) or a build file (does include scan pattern inform-
ation) in the Geometry Selection area of the simulation form.

If You Have Chosen a Part (in Geometry Selection)

Without a build file, the Additive application uses an internal slicing function (called the slicer) and as-
sumes a rotating stripe scan pattern with contours. As such, machine parameters required to establish
the scan pattern include Layer Thickness, Starting Layer Angle, and Layer Rotation Angle. See Under-
standing Machine Parameters (p. 33). You may exclude the contour scans in the simulation by clicking
on the Exclude check box.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
100 of ANSYS, Inc. and its subsidiaries and affiliates.
 Machine Parameters for a Scan Pattern Simulation

Note:

For parts with a gap in the Z dimension in which there should be no scans at all, the slicer
does add scans to those layers by taking the information from the last layer just before the
gap. This is not a common situation but it is possible (for example, when there are multiple
parts on the baseplate combined into one single .stl file). When this type of geometry is used
in a simulation that uses the generic machine to create the scan pattern (i.e., the slicer), those
empty layers will be filled with scan pattern generated using the information from the previous
non-empty layer. If the previous non-empty layer happens to be just a point, then all the
following gap layers will be empty as well. This can potentially cause problems such as incor-
rect simulation results. This limitation applies to any geometry in which its sliced layers have
no contour and end up empty.

If You Have Chosen a Build File (in Geometry Selection)

Currently, a Scan Pattern simulation supports build files with rotating stripe, checkerboard, and any
other scan patterns, with or without contours. The machine parameters will be obtained from the build
file. Your only option is whether to exclude contour scan vectors in the simulation.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 101
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
102 of ANSYS, Inc. and its subsidiaries and affiliates.
Performing a Thermal Strain Simulation
Thermal Strain simulations provide the highest degree of accuracy by predicting how thermal cycling
affects strain accumulation at each location within a part. It uses inherent strain but also implements
a thermal ratcheting algorithm to locally modify the inherent strain value.

In addition to the standard inputs as described in Set Up a Simulation - Assumed Strain (p. 51), you
will need to:

• Consider using a Mesh Resolution Factor, a scaling factor to control speed versus accuracy of
the simulation

• Select an ANSYS pre-defined, thermally validated material

• Define additional configuration parameters of your 3D printing machine

Mesh Resolution Factor

Because of the increase in the number of calculations at each point in a part, Thermal Strain simulations
take much longer than Assumed Strain or Scan Pattern simulations. Additionally, while the Voxel Size
and Voxel Sample Rate determine the mesh used in the Mechanics Solver, the Additive application uses
a much finer mesh in the Thermal Solver (p. 29). This mesh sizing is controlled internally. (See Thermal
Solver Default Settings (p. 31).) An option is available for simulations that use the Thermal Solver to
increase the speed of your simulation—or to increase the accuracy of the melt pool— by scaling the
mesh for the thermal strain portions of the calculation. The Mesh Resolution Factor (MRF) allows some
user control over mesh size. Valid inputs are integers between 1 and 12. We recommend you use the
default value of MRF = 5, in which case you should see comparable run times to the pre-2019 R1 release,
but the results are being generated at better than twice the resolution in X and Y.

MRF is inversely proportional to run time and fidelity. If the MRF is too low, the simulation will take a
long time to complete. If the MRF is too high, you will see a warning message, such as The Mesh
Resolution Factor (MRF) used is too coarse to accurately resolve the
maximum melt pool width. For more accurate results, adjust settings to
satisfy (MRF * 0.021 mm) < max melt pool width. This means the element size is too
large to fully capture the heat transfer phenomena in the melt pool. Either lower the MRF or increase
the energy density into the system to increase melt pool width.

Energy density is the energy being put into the material at the melt pool location per unit volume. The
most significant factors contributing to energy density are the laser power, scan speed, hatch spacing,
layer thickness, and the material's absorptivity. So for a given material, to increase the energy density
you can increase Laser Power, reduce Scan Speed, reduce Hatch Spacing, and/or reduce Layer Thickness.

Should you see a similar warning message regarding MRF in a simulation that does not have MRF as
an input parameter, such as a Microstructure simulation, adjust one or more of the machine parameters
mentioned above.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 103
Performing a Thermal Strain Simulation

Select Material for a Thermal Strain Simulation

ANSYS pre-defined materials are designed to capture the effect of a material’s chemical composition,
powder-to-liquid and liquid-to-solid state transitions, and high cooling rates. Nonlinear thermo-physical
properties such as thermal conductivity, density, and specific heat are essential for capturing solidification
phenomena in metal additive manufacturing. The materials that have had complete validation tests
performed and are recommended for Thermal Strain simulations are shown with a check-mark as
Available for Thermal Simulation in the Materials Library and in the details panel for a material. Only
those materials will be available for selection in the Material Configuration section of the Thermal Strain
simulation form.

Machine Parameters for a Thermal Strain Simulation

Your options for machine parameters in a Thermal Strain simulation depend on whether you have
chosen a part file (does not include scan pattern information) or a build file (does include scan pattern
information) in the Geometry Selection area of the simulation form.

If You Have Chosen a Part (in Geometry Selection)

Without a build file, a Thermal Strain simulation assumes a rotating stripe scan pattern. As such, machine
parameters required to establish the scan pattern include Layer Thickness, Starting Layer Angle, and
Layer Rotation Angle. Additional machine parameters required are Baseplate Temperature, Hatch Spacing,
Slicing Stripe Width, Laser Beam Diameter, Laser Power, and Scan Speed. See Understanding Machine
Parameters (p. 33). All of these machine parameters affect the temperatures at the nodes throughout
the simulation.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
104 of ANSYS, Inc. and its subsidiaries and affiliates.
 Machine Parameters for a Thermal Strain Simulation

If You Have Chosen a Build File (in Geometry Selection)

Currently, a Thermal Strain simulation supports build files with rotating stripe, checkerboard, and any
other scan patterns, with or without contours. However, the scan vectors marked as contour will not
be simulated. The definition of contour/hatch is established by the software that creates the build file.
If contour-like scan vectors are marked as hatch they will be simulated and results may not be as expec-
ted.

Important:

The values of Baseplate Temperature, Laser Beam Diameter, Laser Power, and Scan Speed
in the simulation form override the build file, so take care to set these parameters the way
you want them.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 105
Performing a Thermal Strain Simulation

Thermal Solver Assumptions and Limitations that Affect Results at Part

Edges
The following items affect results at the edges of parts (for both part files and build files):

• An assumption in our approach may result in scan lines that do not always extend to the edge
of the part boundary. The Thermal Solver uses a uniform time step size along the laser’s scan
path. When a scan line’s length is not evenly divisible by the distance traveled in a single time
step (Scan Speed * time step size), the last time step on the line is ignored. This should have a
relatively minimal impact on the overall solution, given the time step size is generally on the
order of 20 µs. This phenomenon may be observed when viewing Thermal History (coaxial average
sensor) results, as seen by the jagged edges in the following figure.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
106 of ANSYS, Inc. and its subsidiaries and affiliates.
 Thermal Solver Assumptions and Limitations that Affect Results at Part Edges

• In some cases, because of a discrepancy in the way that strain is calculated at the edges of parts
versus the assigned state (liquid or solid), you may see very slightly inaccurate stress/distortion
results for Thermal Strain simulations at the edges of parts. This is a known limitation at this
time.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 107
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
108 of ANSYS, Inc. and its subsidiaries and affiliates.
 Additive Science Simulations
Performing a Single Bead Parametric Simulation (p. 111)

Performing a Porosity Parametric Simulation (p. 121)

Performing a Thermal History Simulation (p. 129)

Performing a Microstructure Simulation (p. 131)

Creating User Defined Materials (p. 155)

Performing a Single Bead Parametric Simulation
We assume you are familiar with the basics of running ANSYS Additive, such as how to use the Additive
interface, start simulations, etc. If not, refer to Understanding the Additive Interface (p. 11) for that in-
formation. Now let's dive in to Additive Science.

A Single Bead Parametric simulation is a tool used to obtain information about the melt pool character-
istics of your selected material. It is set up to follow the industry practice of testing single beads on PBF
machines, in which the laser runs in one single scan line across powder. The goal is to determine the
geometry of the melt pool. Geometry refers to melt pool width, length, and depth, as shown in the
following figure. Using simulation is a good way for you to see how process parameters affect melt
pool characteristics without using a lot of expensive material and machine time.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 111
Performing a Single Bead Parametric Simulation

Machine Configuration
Enter values for the process parameters (p. 33) in this section of the simulation form. Inputs include
Baseplate Temperature (°C), Layer Thickness (µm), and Laser Beam Diameter (µm) as constant values
and Laser Power (W) and Scan Speed (mm/sec) as parametric variables. You can enter variable values
through either the incremental step tool, or manually, or both. (Note that each time you use the incre-
mental step tool it will overwrite the values you have already entered.)

Characteristic Width Calculation Mode

The Use Characteristic Width Calculation Mode check box should be checked only when you are creating
user defined materials (p. 155). When checked, certain parameters are automatically set and are not
available for input.

Bead Type
Unique to Single Bead simulations is the input option of Bead Type, an indication of how the bead is
deposited. This setting affects the calculation of material state, which in turn affects the material prop-
erties used in the solving process, as well as in the laser flux model. You may need to change from the
default if you are matching a particular experimental setup.

• Bead on powder layer (default): A single bead deposited on top of a layer of powder of Layer
Thickness.

• Bead on base plate: A single bead deposited directly on solid material.

On-Pad versus On-Plate Layer Thickness

Layer thickness input for a Single Bead simulation should represent the amount of material that is being
added to each layer of a build. To compare single bead experiments to the simulation you must make
sure that the Layer Thickness input field reflects this. When single bead experiments are deposited on
a pad built using the same material and layer thickness as the single bead, that layer thickness is appro-
priate to use for the simulation. We recommend this technique. By contrast, when single beads are built
on a plate that simply has powder spread on top of it, the simulation Layer Thickness must be adjusted
based on the powder packing density to account for the consolidation into solid. In this technique, the
Layer Thickness should be less than the machine's single bead thickness.

In general, Single Bead simulations show good and consistent data matches with experimental results
when the melt pool depths are 2 to 5 times the powder layer thickness. When the melt pool depth is
around the powder layer thickness or less, the experimental process is normally unstable (the powder
layer is not fully melted and/or balling occurs), and larger differences may occur when comparing sim-
ulation results with experimental results.

Throughout this and the following chapter, we will use an example to demonstrate just one of the
many ways you can use Additive Science to gain insights into your PBF process. This is hypothetical

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
112 of ANSYS, Inc. and its subsidiaries and affiliates.
 Machine Configuration

example where we have made assumptions about our process goals and chosen material that may not
be valid in other cases. We will run a Single Bead Parametric simulation using a generic material.

In this example, we entered Laser Powers starting at 50 and ending at 400 in increments of 50. After
clicking on the check-mark, the values of 50, 100, 150, 200, 250, 300, 350, and 400 are entered.

Continuing with this example, we entered Scan Speeds between 700 and 1300 in increments of 100.
This will result in 56 individual permutations in the simulation. That is, single bead scans will be simulated

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 113
Performing a Single Bead Parametric Simulation

at a Laser Power of 50 and a Scan Speed of 700, then another one for Laser Power = 50 and Scan Speed
= 800, another one at Laser Power = 50 and a Scan Speed = 900, and so on until every combination is
performed. This is a full factorial experiment, with each laser power matched with each scan speed.
There is a limit of 300 permutations in one simulation.

Geometry Configuration
The Geometry Configuration section of the simulation form is where you enter the Bead Length (in
mm). Valid input values are real numbers between 1 and 10. In our studies, we have seen that the melt
pool becomes stable (that is, reaches steady-state) well within 2 mm for all our validated ANSYS-defined
materials. It is seldom that you will need a Bead Length longer than 3 mm (default).

Reviewing Results
In the ANSYS Additive application, we track the instantaneous melt pool along the bead length and
then average the value of each dimension over the entire bead length. Output files are provided for
both the individual permutations showing the full progression along the bead length and a summary
file of the average and median melt pool length, reference depth, and reference width for each permuta-
tion. From the following figure, we see that the reference depth is the entire melt pool depth minus
the Layer Thickness, or the melt pool depth starting from the bottom of the first layer. Similarly, the
reference width is the width at the bottom of the first layer (the start of the substrate).

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
114 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

It is the median result of any particular dimension you should use when interpreting your data, rather
than the average result. The average will be skewed by the beginning of the bead when the melt pool
is not yet stable. Let's examine the results from our example.

Individual Melt Pool Results

Results from an individual permutation (that is, one combination of power and speed) are shown below.
A .csv file called L0_Meltpool.csv (for Layer 0) is output for each permutation. (You may find it useful
to open it in Microsoft Excel and save it as an Excel table with a bit of formatting as shown here.) The
timestep and distance along the bead (the mesh size) are determined internally. For this power/speed
combination, we can plot melt pool dimensions against distance along the bead (laserX column) to
observe when the melt pool has reached steady state. In our example, convergence is reached within
0.3 to 0.5 mm, as shown in the chart.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 115
Performing a Single Bead Parametric Simulation

Summary Results
The Single Bead Summary file summarizes the melt pool characteristics for all 56 permutations. The
parametric variables are listed in the left two columns; scan speed and laser power. We'll begin by
sorting the data by median melt pool reference depth to see how scan speed and laser power influence
melt pool depth. As a quick check of our data, for the last row with the deepest melt pool (0.182 mm
median reference depth), we would expect to see the highest power and slowest scan speed combination.
Indeed, our data shows the highest power (400 W) and slowest scan speed (700 mm/s) permutation
from our simulation.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
116 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 117
Performing a Single Bead Parametric Simulation

In our example simulation, we added columns in the summary table to calculate depth-to-width ratio
and length-to-width ratio using the simulation data. We made the following hypothetical assumptions
regarding the criteria for "good-candidate" power/speed combinations:

• We want a melt pool depth that reaches at least halfway through the third layer. A penetration
depth of about three layers through the thickness reduces porosity by remelting previous layers.
Since our layer thickness is 40 microns (0.04 mm), that means we want a melt pool depth of at
least 0.1 mm, which is a median melt pool reference depth of at least 0.06. The data that fall
outside of our acceptable criteria for melt pool reference depth are shown in the yellow shaded
area of the median melt pool reference depth column. These melt pools are not deep enough.

• We want a depth-to-width ratio below 0.95. The data that fall outside of our acceptable criteria
are shown in the blue shaded area of the depth-to-width ratio column. These melt pools are too
deep.

In fact, when the melt pool becomes too large and outside of the software's acceptable range,
the Additive application will error out with the following message..."INVALID STATE: The
melt pool has become too wide and deep to produce valid results
(current width = xxx microns, current depth = xxx microns). We re-
commend you reduce the energy input by reducing Laser Power and/or
increasing Scan Speed.

• We want a length-to-width ratio below 4.0. The data that fall outside of our acceptable criteria
are shown in the orange shaded area of the length-to-width ratio column. These melt pools
might be too long.

Data points that meet all the good-candidate criteria above (power/speed combinations that are not
in the yellow, blue, or orange shaded regions) are shown in the green shaded rows in the summary
table.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
118 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 119
Performing a Single Bead Parametric Simulation

The following is a chart of all the data from the summary table shown in a power/speed process map.
We can make the following observations:

• The eight good-candidate power/speed combinations are shown as green stars.

• Data points in yellow are the power/speed combinations that fall outside our acceptable criteria for
melt pool reference depth, indicating melt pools that are not deep enough. This makes sense given
that the yellow points are in a region of the map with high scan speeds and low laser power (that
is, low energy density), which could contribute to lack-of-fusion porosity between the layers.

• Data points in blue are the power/speed combinations that fall outside our acceptable criteria for
depth-to-width ratio, indicating melt pools that are too deep. This makes sense given that the blue
points are in a region of the map with low scan speeds and high laser power (that is, high energy
density), which could lead to keyhole formation.

• Data points in orange are the power/speed combinations that fall outside our acceptable criteria for
length-to-width ratio, indicating melt pools that may be too long. This is an area of the map with
the highest speeds and the highest powers, an area which has the potential for the generation of
spatter and for a beading effect known as balling.

Based on median width and length data from the table, melt pool sizes ( ) are shown for a few
sample points in the chart to show relative sizes of the melt pools. Note that these are not true scale.

We will examine the good-candidate combinations further in a Porosity Parametric simulation.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
120 of ANSYS, Inc. and its subsidiaries and affiliates.
Performing a Porosity Parametric Simulation
Following Single Bead Parametric simulations, typically you will want to expand your exploration to
include Porosity Parametric simulations. Porosity simulations are used to obtain information about the
level of porosity in additively manufactured material. They are set up to follow the industry practice of
building full 3D cubes (or cuboids) on PBF machines in which many layers are printed using your chosen
material and scan pattern.

Let us now revisit the power/speed process map. In our single bead example, we made some assumptions
about melt pool criteria that would yield good-candidate power/speed combinations based on hypo-
thetical experimental data. Without the benefit of experimental data, however, it becomes a wide open
exploration with many variables. ANSYS Additive solves for only lack-of-fusion porosity. Accurate simu-
lations of the other two porosity mechanisms (keyholing and balling) would require full fluids analyses
and would consider the effects of other physics within the melt pool. Still, our software can guide you
toward a more informed approach when choosing process parameters, and provide essential information
about lack-of-fusion porosity, in particular.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 121
Performing a Porosity Parametric Simulation

As in Single Bead Parametric simulations, up to 300 permutations are possible for Porosity Parametric
simulations. However, each permutation takes a much longer time to complete because, now, many
layers are being simulated in the cuboid geometry, not just a single bead. The general approach is to
select the best candidate process parameters as determined from the Single Bead Parametric simulation
and introduce a new variable, such as Hatch Spacing.

In our case study, we identified 8 target power/speed combinations to study further. If we introduce 5
hatch spacing values, we will have 40 porosity permutations.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
122 of ANSYS, Inc. and its subsidiaries and affiliates.
 Machine Configuration

Machine Configuration
Because many build layers are simulated in porosity simulations, you need to account for your particular
PBF machine and your chosen scan pattern. See Understanding Machine Parameters (p. 33) for descrip-
tions of these input parameters.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 123
Performing a Porosity Parametric Simulation

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
124 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

For our case study, note that we can't enter all our speeds, powers, and hatch spacings on one simulation
form because that would produce a full factorial simulation of 4 powers, 5 speeds, and 5 hatch spacings
for 100 total permutations. Instead, we ran 4 separate simulations, as shown here.

Simulation A Simulation B Simulation C Simulation D

Power = 250, Speed =
Power = 150, Speed = Power = 200, Speed = Power = 300, Speed =
1000, and 1100, and
800 900, 1000, and 1100 1100
1200
Hatch Spacing = 0.05 Hatch Spacing = 0.05 Hatch Spacing = 0.05 Hatch Spacing = 0.05
Hatch Spacing = 0.07 Hatch Spacing = 0.07 Hatch Spacing = 0.07 Hatch Spacing = 0.07
Hatch Spacing = 0.09 Hatch Spacing = 0.09 Hatch Spacing = 0.09 Hatch Spacing = 0.09
Hatch Spacing = 0.11 Hatch Spacing = 0.11 Hatch Spacing = 0.11 Hatch Spacing = 0.11
Hatch Spacing = 0.13 Hatch Spacing = 0.13 Hatch Spacing = 0.13 Hatch Spacing = 0.13
5 permutations 15 permutations 15 permutations 5 permutations

Geometry Configuration
Enter the Width, Length, and Height of the cuboid (in mm). Valid input values are real numbers between
1 and 10. In our studies, we have seen that the porosity pattern will reach steady state within a 3 x 3
x 3 mm cube (default) for all our validated ANSYS-defined materials. It is seldom that you will need a
cube larger than the default.

Reviewing Results
Lack-of-fusion porosity is identified in Porosity Parametric simulation results as values of solid ratio below
1. In other words, an ideal solid would have a solid ratio of 1. Anything that is not solid is a void, or
powder. The goal when additively manufacturing parts is to choose process parameters that optimize
our process. One guideline is to choose the fastest scan speed and the widest hatch spacing possible
while staying within our target processing region so as to avoid porosity. Depending upon what our
acceptable level of porosity may be, we can reduce our choices to a few viable candidates.

The following table shows porosity results for 25 of the 40 permutations. We have identified the best
candidates based on a target criteria of less than 0.5% porosity (shaded in green). The fastest build rate
of these candidates is the combination of 300 W power, 1100 mm/s scan speed, and 0.13 mm hatch
spacing.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 125
Performing a Porosity Parametric Simulation

Guidelines for Interpreting Porosity Results

Validating porosity results faces inherent challenges given the cost and effort required to obtain accurate
and statistically representative experimental datasets. As a result, some deviation between simulated
and measured data is to be expected. This is the case across all ranges of porosity, but particularly so
at relatively low porosity cases (  ∼1%), given both experimental error as well as practical challenges
of achieving sufficient simulation resolution without becoming computationally prohibitive. These
challenges may be reduced from an experimental side by gathering a sufficient number of high-quality
measurements to boost confidence in the magnitudes and trends expected.

It should be noted that there are various contributing factors to almost never achieving fully dense
parts experimentally. These include the following:

• Entrapped gases in the powder themselves can become trapped in the part as it is being built.
This is powder-supplier-dependent and puts a lower bound on achievable porosity with their
powder.

• Outgassing by contaminants may occur during melt pool formation. This is dependent on end-
user powder handling protocols and puts a lower bound on achievable porosity.

• Vaporization of the powdered material will occur during scanning. There will always be at least
a small amount of this gas vapor that becomes trapped in the melt pool during solidification,
which appears as porosity. This is alloy dependent and puts a lower bound on achievable
porosity.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
126 of ANSYS, Inc. and its subsidiaries and affiliates.
 Guidelines for Interpreting Porosity Results

• Large plasma plumes at high energy density result in keyhole formation. The shape of keyholes
make trapped vapor even more likely. As energy density increases, the amount of keyhole
porosity will increase.

We do not simulate any of these phenomena. When our lack-of-fusion predicted % porosity levels become
very low, these issues that we do not predict start to dominate and keep the actual part from becoming
100% dense even though the simulation indicates 100% density.

Finally, the validity of a Porosity simulation diverges more significantly from empirical data when
porosity is greater than 30%. This can be caused by high variance in empirical data at high levels of
porosity or from simulation resolution of melt pools in low energy density scenarios. We do not recom-
mend you simulate with settings that will generate porosity greater than 30%.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 127
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
128 of ANSYS, Inc. and its subsidiaries and affiliates.
Performing a Thermal History Simulation
In this simulation type, given a material, a part, and all the process parameters – melt pool dimensions
and thermal history are output for a particular cross-section of your part, simulating results from a
coaxial average sensor. This simulation type is a Beta feature at Release 2021 R1. Beta documentation,
if available, may be found here.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 129
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
130 of ANSYS, Inc. and its subsidiaries and affiliates.
Performing a Microstructure Simulation
For one of three validated materials (316L, AlSi10Mg, or Inconel 718), Microstructure simulations produce
grain orientation and grain boundary results of the material for three 2D planes (XY, XZ, and YZ) simu-
lating typical results of EBSD laboratory tests. Also, grain size distribution and orientation angle distri-
bution are output in graph form using a circle equivalence method and orientation map for all three
planes, respectively. Microstructure simulations are parametric so you can run several permutations to
determine which process parameters will affect grain size the most. A Specific Random Seed input
parameter may be used to instigate the same nucleation pattern for a given set of process conditions
so that comparison simulations are possible.

Using Microstructure Data to Evaluate Mechanical Properties

In a polycrystalline material, mechanical properties are controlled primarily by the material's inherent
microstructure characteristics—grain size, grain shape, grain orientation, and composition of materials.
These characteristics can change significantly with different processing methods and conditions, thereby
changing the mechanical properties. Due to the complex nature of the additive manufacturing process
where the material is subjected to repetitive heating cycles, the resulting microstructure is very inhomo-
geneous and different for different process parameters, creating variability in mechanical properties
(see Akram, et. al. [1]). Using a Microstructure Simulation in Additive Science, you can predict the changes
in microstructure—grain size, grain shape, and grain orientation—for a given set of process parameters.

Grain size can be used to predict the mechanical property of yield strength. The relationship between
yield strength and grain size is described mathematically by the Hall-Petch equation:

where is the yield strength, is a materials constant for the starting stress for dislocation motion,
is the strengthening coefficient (constant for a material), and is the diameter of a grain. This rela-
tionship shows that the grain size is inversely proportional to yield strength. The smaller the grain size,
the higher the yield strength, and vice versa. This relationship holds true for most materials and is a
widely accepted model to express the yield strength relationship with grain size. If you know the ma-
terial constants, you can use the grain size obtained from a Microstructure Simulation to predict the
yield strength of a material.

Grain orientation plays a major role in the directional nature of a part's mechanical properties. If the
grain size and orientation is uniform in all three orthogonal planes, this indicates the part's properties
are isotropic. If grain orientation is predominant in one of the planes, this indicates that the printed
part will be highly anisotropic. As-built parts printed by AM are usually anisotropic, meaning different
build orientations give different mechanical properties. The grain orientation data obtained from a Mi-
crostructure Simulation can provide insights into the grain orientation in the part produced using dif-
ferent process parameters such as laser power, speed, scan pattern, hatch spacing, powder layer thickness,
etc. Ultimately, the goal in AM simulation is to enable you to design a part based on your specific re-

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 131
Performing a Microstructure Simulation

quirements. For example, if your part requires anisotropy in one of the directions, you can design the
part by choosing the parameter set that results in the desired anisotropic microstructure.

Material Configuration
Materials validated for Microstructure simulations are 316L, AlSi10Mg, and Inconel 718. See Microstructure
Validation Process (p. 146).

Microstructure Configuration
A Microstructure Simulation is essentially a two-part solution. A thermal solution of a cuboid geometry
is performed first using the Thermal Solver (p. 29) to determine cooling rate, thermal gradient, and
melt pool size (width and depth) for the material given a set of process parameters. Then the Micro-
structure Solver is used on a smaller "cube" inscribed inside the cuboid to determine the microstructural
information. The cube is actually a set of three 2D planes with width and height equal to Sensor Dimen-
sion. A 1 mm buffer depth and a 0.25 mm buffer distance from all the edges of the cuboid in the XY
plane are required to ensure the microstructure is obtained with a thermal melt pool that has reached
steady-state. The thermal quantities of cooling rate, thermal gradient, melt pool width and melt pool
depth used by the Microstructure Solver are single, averaged values and are averaged from within the
microstructure cube only.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
132 of ANSYS, Inc. and its subsidiaries and affiliates.
 Microstructure Configuration

Geometry Configuration
Geometry Width (mm) - The width of the cuboid, in millimeters, for the thermal solution. Geometry
Width must be at least Sensor Dimension plus 0.5 mm (for a 0.25 mm buffer on each side). Valid input
values are positive real numbers between 1 and 10. Defaults to 1.5 mm.

Geometry Length (mm) - The length of the cuboid, in millimeters, for the thermal solution. Geometry
Length must be at least Sensor Dimension plus 0.5 mm (for a 0.25 mm buffer on each side). Valid input
values are positive real numbers between 1 and 10. Defaults to 1.5 mm.

Geometry Height (mm) - The height of the cuboid, in millimeters, for the thermal solution. Geometry
Height must be at least Sensor Dimension plus 1 mm (for a buffer above the sensor). Valid input values
are positive real numbers between 1.1 and 10. Defaults to 1.5 mm.

Sensor Dimension (mm) - The dimension (width and length, in millimeters) of the 2D planes for the
microstructure solution. The sensor point is the intersection of all three planes and is always 1 mm deep
into the cuboid. Sensor Dimension must be between 0.1 and 1.0. Defaults to 0.5 mm.

Thermal Solution Configuration

A full thermal solution is performed on the cuboid geometry by default but the Additive application
allows you to bypass the thermal solution if you know the material’s cooling rate, thermal gradient,
and melt pool characteristics. These items are provided as output on a completed Microstructure simu-
lation, specifically on an individual permutation's results page (shown here), for simulations that ran
with a thermal solution.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 133
Performing a Microstructure Simulation

Use Provided Thermal Parameters - Check this option to bypass the thermal solution and perform
the microstructure solution only. (The Laser Power and Baseplate Temperature fields under Machine
Configuration become disabled when you bypass the thermal solution.) Be aware that the Microstructure
simulation run time increases proportionally with melt pool width and depth. The larger the Melt Pool
Width, or the deeper the Melt Pool Depth, the longer the run time.

Cooling Rate (°K/sec) - Must be between 1x105 and 1x107. Defaults to 1x106.

Thermal Gradient ( °K/m) - Must be between 1x105 and 1x108. Defaults to 1x107.

Melt Pool Width (mm) - Must be between 0.075 and 0.8. Defaults to 0.15 mm.

Melt Pool Depth (mm) - Must be between 0.015 and 0.8. Defaults to 0.1 mm.

Random Seed
If desired, you can specify a specific random seed to be used as a starting point for the nucleation
pattern. Use an integer between 0 and 4294967295. This feature allows you to generate the same output

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
134 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

for the same inputs to check for repeatability. If you do not use a random seed (default), the solver will
assign a random number for the nucleation starting point and simulation results will not be repeatable.

Machine Configuration
Process parameters may be input with multiple values for each to set up parametric simulations. This
allows you to study which parameters affect grain size and morphology the most.

Machine - A placeholder for future machine configurations. Defaults to Generic for now.

Laser Beam Diameter (μm) - The width of the laser on the powder or substrate surface defined using
the D4σ beam diameter definition. Usually this value is provided by the machine manufacturer. Sometimes
called laser spot diameter. Must be between 20 and 140. Defaults to 100 μm.

Laser Power (W) - The power setting for the laser in the machine. Must be between 50 and 700. Defaults
to 195 Watts.

Scan Speed (mm/sec) - The speed at which the laser scans, excluding jump speeds and ramp up and
down speeds. Must be between 350 and 2500. The default value is 1000 mm/sec.

Baseplate Temperature (°C) - The controlled temperature of the baseplate. Must be between 20 and
500. The default value is 80°C.

Layer Thickness (μm) - The thickness of the powder layer coating that is applied with every pass of
the recoater blade. Must be between 10 and 100. The default value is 50 microns.

Hatch Spacing (μm) - The distance between adjacent scan vectors when rastering back and forth with
the laser. Hatch spacing should allow for a slight overlap of scan vector tracks such that some of the
material re-melts to ensure full coverage of solid material. Must be between 0.06 and 0.2 mm. The default
Hatch Spacing is 0.1 mm (100 microns).

Starting Layer Angle (°) - The angle at which the first layer will be scanned. It is measured from the X
axis, such that a value of 0° results in scan lines parallel to the X axis. Must be between 0 and 180°. The
starting layer angle is commonly set to 57° (default).

Layer Rotation Angle (°) - The angle at which the major scan vector orientation changes from layer
to layer. Must be between 0 and 180°. This is commonly 67° (default).

Reviewing Results
A Microstructure Simulation produces one summary result page (the parent simulation) showing the
set of permutations (child simulations) included within it. Selecting any of the permutations brings up
a new page of individual permutation results. These results include:

• The microstructure in XY, YZ, and XZ planes available in both .avz format that can be viewed
and exported, and as .vtk files that can be exported. Use these data to view grain orientation,
grain boundaries, and grain number.

• Grain size statistics available in both summary bar graph form that can be viewed and .csv files
that can be exported.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 135
Performing a Microstructure Simulation

Viewing Grain Orientation

The three primary result items in a Microstructure Simulation are grain orientation angle, grain bound-
aries, and grain number. Grain orientation represents grain growth direction with respect to the hori-
zontal axis of a 2D plane based on a moving heat source. Note that it does not represent crystal orientation
that is typically seen in an EBSD map. Grain orientation angles from an individual permutation are shown
in the following figure when viewed in the application (AVZ viewer). A unique grain number is assigned
to one continuous grain. A unique color is assigned for each orientation.

To view all three planes at once, use ANSYS EnSight, a tool for high performance graphics post-processing.
(Alternatively, you can use an open-source visualization application such as Paraview.) Within EnSight,
visualize the microstructure data by doing the following:

1. Choose File > Open to open any one of the three plane .vtk files. This will provide data to the
default “Case 1”, which can be renamed by right-clicking and choosing “Rename case…”.

2. Repeat File > Open for the next two planes, selecting “Keep currently loaded data” when the
“New case” dialog box appears. Change case name, if desired.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
136 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

3. Click and drag or use Ctrl + click to select all three data cases in the data panel.

4. Right-click and select Color by > Select variable.

5. Under Scalars, choose the desired variable (orientation, grain number, or grain boundaries).
Click OK.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 137
Performing a Microstructure Simulation

6. Changing the color palette may provide more contrast among individual grains, depending on
your selected variable. Right-click the legend and then select Palette > Edit palette.

7. In the Palette editor, under the Options tab, choose Banded for the type of color distribution.
Preset color maps can be chosen under the “Files” tab by double-clicking on the desired color
map. To view grain orientation, we recommend using a six-color color map. Click Restore to
activate your chosen color map.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
138 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

8. Use the pan, rotate and zoom controls to achieve the desired view. Right-click anywhere in the
viewport and under View, select any of the view directions to return to a specific plane.

If the sensor height is not evenly divisible by the layer thickness, the vertical planes (XZ and YZ) may
extend through the XY plane as shown in the next figure.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 139
Performing a Microstructure Simulation

Viewing Grain Boundaries

The purpose of the grain boundaries map is to see only the boundaries of the grains, similar to grain
scale images seen in optical microscopy. Internally, this two-color effect is achieved by setting a grain
boundary flag to 1 and the interior to 0.

As an example, using the EnSight instructions in the previous section:

1. Choose File > Open to open the XZ plane .vtk file. Rename Case 1 to XZ plane.

2. Select the XZ plane, right-click and select Color by > Select variable.

3. Under Scalars, choose grain boundaries. Click OK.

4. Right-click the legend and then select Palette > Edit palette.

5. For best viewing of the grain boundaries, we recommend using the “X Ray (2 colors)” option
in the Palette editor. In the Palette editor, under the “Files” tab, double-click the “X Ray (2
colors)” color map.

6. Right-click anywhere in the viewport and under View, select the -Y view to see the XZ plane.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
140 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

The same grain boundaries data are shown next in the AVZ viewer.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 141
Performing a Microstructure Simulation

Viewing Grain Size Statistics

For each grain, the area and the area fraction of the total area are calculated. To do this, the diameter
of an equivalent circle for each grain is calculated using the Circle Equivalence Method, as shown in
the following figure.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
142 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

Grain area and diameter values for all the grains are aggregated into a Grain Size Distribution bar chart
showing the area fraction of grains whose diameter falls in discrete ranges indicated on the chart. This
bar chart is displayed in the application when you click the Graph link as shown in the following figure.
While this particular bar chart is not exportable, the full data can be exported as a .csv file. The .csv file
does not contain the same data as the aggregate bar charts but contains full data for each grain with
area fraction and equivalent circle diameter such that you can produce similar graphs on your own. A
second bar chart shows a similar aggregate of Orientation Angle.

From the charts below, this particular result set shows the majority of grains have a diameter between
30 and 35 microns and are oriented between 80 and 90°.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 143
Performing a Microstructure Simulation

How Porosity is Revealed in Results

It is always wise to examine your results to be sure they make logical and intuitive sense and to enhance
your understanding of the microstructural phenomena occurring in the simulation. Instances of lack-

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
144 of ANSYS, Inc. and its subsidiaries and affiliates.
 Reviewing Results

of-fusion porosity, in particular, are important to discern and are, in fact, easy to recognize in microstruc-
ture results. A grain number of 0 indicates a location of unmelted material, or lack-of-fusion porosity.
Orientation and grain number maps will show gaps at these locations, as shown in the following figure.
This example was run as an exaggerated case to show lack-of-fusion porosity by defining inputs such
that the Melt Pool Depth was less than Layer Thickness, and Melt Pool Width was less than Hatch Spacing.

Evaluating Trends
When evaluating a material's microstructure, there is significant variability of results, even in tightly
controlled experimental lab tests. This is why our validation process (p. 146) includes printing multiple
test cubes, and why we recommend that you run at least three to five simulations using random
seed (p. 134) numbers and average the results.

Confidence can be found in evaluating trends in the data. In general, you should see these trends in
your Microstructure simulation results:

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 145
Performing a Microstructure Simulation

• The lower the energy density put into the system, the smaller the resulting grain size. Energy
density is the energy being put into the material at the melt pool location per unit volume. The
most significant factors contributing to energy density are the laser power, scan speed, hatch
spacing, layer thickness, and the material's absorptivity. So for a given material, to decrease the
energy density you can decrease Laser Power, increase Scan Speed, increase Hatch Spacing,
and/or increase Layer Thickness.

• If the cooling rate increases, grain sizes should decrease.

• If the average grain size is smaller, variation in grain size will also be smaller.

Microstructure Validation Process

Validation of the microstructure solver for each material is performed by comparing experimental and
simulation results for various sets of process parameters. Several cubes are printed using a Selective
Laser Melting method with various process parameters. After analyzing porosity data, samples are se-
lected that result in high density. These samples are prepared using the standard metallography pro-
cedure for EBSD (Electron Back Scattered Diffraction) analysis for grain size and grain orientation estim-
ation, which consists of grinding and polishing the surface of the cubes up to 0.05 µm surface finish.
The prepared samples are subjected to SEM (Scanning Electron Microscopy) equipped with an EBSD
detector to generate the grain orientation and grain size data. Simulation results are generated by
running the simulation end-to-end from thermal solution to microstructure solution using the same
process parameters as in the experiments.

Overlap Coefficient Analysis of Grain Sizes

To validate simulated microstructures against experimentally retrieved results, one essential aspect to
compare is grain size. Since the grain sizes of microstructures built under the SLM process show signi-
ficant variability, it may be insufficient to only compare the grain size average for validation purposes.
Another comparison technique is to represent grain size distributions as a normal/Gaussian distribution
and subsequently use the overlap coefficient to measure the overlap, or similarity, of grain size distri-
butions between simulated and experimental results.

The overlap coefficient is a coefficient ranging from 0 to 1 and can be conceptually understood as the
fraction of overlap shared between two distributions. A value of 1 indicates perfect similarity, or exactly
similar distributions, and a value of 0 indicates complete dissimilarity between two distributions.

Calculation Procedure

1. From experimental and simulated results, calculate the average, , and standard deviation, , of the
grain size:

Experimental →

Simulation →

Note: An assumption has been made that these grain sizes are distributed normally. This has been
observed empirically, but future improvements can be made to distinguish between two arbitrary
distributions.

2. Calculate the x-positions of the intersections between two distributions:

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
146 of ANSYS, Inc. and its subsidiaries and affiliates.
 Microstructure Validation Process

3. Calculate the overlap coefficient:

where D is the standard normal distribution function defined as:

Example of Validation Data

A representative example of validation data is provided here for the IN718 alloy. Similar work is done
for every material that is available for Microstructure simulations. Note that microstructure results may
change as solver refinements are made in subsequent releases.

For the IN718 alloy, seven samples were selected. The following graph shows the comparison between
experimental and simulated average grain size along with its variability values. The grain size obtained
from EBSD analysis was from one of the vertical planes of the cube, therefore, the simulation result is
compared with one vertical plane of the cube.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 147
Performing a Microstructure Simulation

Following is the corresponding plot of overlap coefficient for the seven samples.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
148 of ANSYS, Inc. and its subsidiaries and affiliates.
 Microstructure Validation Process

In addition to grain size validation, the grain morphology is also validated by comparing simulation
results with experimental results. Specifically, we examine how the grains are oriented in a sample frame
of reference. An example comparison of grain orientation is shown below. (The experimental results
image is shown upside down to ensure consistent viewpoints; they are both aligned with the build
direction.)

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 149
Performing a Microstructure Simulation

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
150 of ANSYS, Inc. and its subsidiaries and affiliates.
 Known Issues and Limitations

Known Issues and Limitations

• The simulated grain orientations do not represent crystal orientations that are typically seen in EBSD
maps. This is a limitation of the current 2D microstructure model. The simulation results are good for
comparing grain morphology, size and growth direction.

• When orientation output of a Microstructure simulation is viewed in the Additive application (that
is, using the AVZ viewer), the grain morphology tends to show no distinct formation. This is primarily
due to the limited color scheme and data extrapolation used in the viewer. A plot of the same output,
when visualized using external software, such as ANSYS EnSight or Paraview, tends to show the ex-
pected grain morphology. The following figures show the same results in ANSYS EnSight and with
the AVZ viewer. In the first figure you can see long centerline columnar grains (marked with black
arrows) that are equally separated by hatch spacing and tend to grow throughout the height of the
build, along with small grains sandwiched between these long centerline columnar grains. These are
typically seen in IN718 or 316L alloy when printed using a 0-90° scan pattern. This information is
difficult or impossible to see when viewed using the AVZ viewer. Therefore, for best viewing purposes,
we recommend using external software that allows color scheme changes, such as ANSYS EnSight
or Paraview, to visualize the correct grain morphology.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 151
Performing a Microstructure Simulation

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
152 of ANSYS, Inc. and its subsidiaries and affiliates.
 References

• Certain process-parameter combinations will cause the Thermal Solver to emit a warning regarding
the Mesh Resolution Factor (p. 103) (MRF) being too coarse to accurately resolve the maximum melt
pool width. This means the element size is too large to fully capture the heat transfer phenomena
in the melt pool. If this happens, increase the energy density into the system to increase melt pool
width. (Currently, for Microstructure simulations, the MRF is set at 2 and you cannot change this
value.)

References
1. Akram, Javed, et. al., (2018) Understanding Grain Evolution in Additive Manufacturing Through
Modeling. Vol 21. 255-268. Additive Manufacturing. Web.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 153
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
154 of ANSYS, Inc. and its subsidiaries and affiliates.
Creating User Defined Materials
Thermal simulations in Additive Science use a simple volumetric heat source model that allows for rel-
atively fast simulation times and is able to accurately capture the size of additive manufacturing melt
pools. With this simplified heat source model, the penetration depth and absorptivity are required in-
ternal inputs that are generally unknown and vary based on other process parameters. ANSYS pre-
defined materials have been established using a material tuning process that determines penetration
depth and absorptivity coefficients. With the user defined material capability, you can examine trends
and create your own materials that account for these variations in absorptivity and penetration depth.

This chapter documents the input, output, and file specifications for user defined materials in ANSYS
Additive. An in-depth document is available that describes the theory and procedure, including the use
of the Material Tuner tool (Beta) that automates much of the simulation work for you.

Importing a User Defined Material

To add a user defined material, open the Materials Library and click Import Material.

Navigate to the three appropriate files and click Save.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 155
Creating User Defined Materials

Now when you view the Materials Library listing, you will see your new user defined material. Click on
it to see details.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
156 of ANSYS, Inc. and its subsidiaries and affiliates.
 Importing a User Defined Material

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 157
Creating User Defined Materials

Use the Archive button to remove the user defined material from the Materials Library. Simulations
using this material will remain.

User Defined Material File Specifications

Three files are required to define your user defined material. The formats of these files are very specific
and must be created exactly as described in the following sections.

Material Configuration File

The material configuration file contains information on mechanical properties, constant thermal
properties, and other parameters used in Additive Science. This file is read as part of the input when
running any simulation, but not every setting is used in every simulation.

Settings in the material configuration file are defined in the following table.

Setting Definition Units Value

name Name for the material. Must be -
unique.
description Additional comments (optional) -
coreVersion Version of the current material -
configuration file. This can be

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
158 of ANSYS, Inc. and its subsidiaries and affiliates.
 User Defined Material File Specifications

changed when updating new

versions of the same material.
solidThermalConductivityAtRoomTemperature Thermal conductivity of bulk W/m/K
material at room temperature (298
K)
solidDensityAtRoomTemperature Density of bulk material at room kg/m^3
temperature (298 K)
solidSpecificHeatAtRoomTemperature Specific heat of bulk material at J/kg/K
room temperature (298 K)
purgingGasConvectionCoefficient Convection coefficient between the - 12.5
solid and gas during processing
solidusTemperature Maximum temperature at which the K
material is completely solid
liquidusTemperature Minimum temperature at which the K
material is completely liquid
vaporizationTemperature Temperature at which material has K
completely changed from liquid to
vapor
thermalExpansionCoefficient Coefficient of thermal expansion. 1/K
elasticModulus Elastic modulus of build material. Pa
elasticModulusOfBase Elastic modulus of base material. Pa
poissonRatio Poisson ratio -
materialYieldStrength Yield strength Pa
supportYieldStrengthRatio Factor to reduce the yield strength -
and elastic modulus of support
material
anisotropicStrainCoefficientParallel Multiplier on the predicted strain in - 1.5
the direction that the laser is
scanning for the major fill rasters
anisotropicStrainCoefficientPerpendicular Multiplier on the predicted strain - 0.5
orthogonal to the direction that the
laser is scanning for the major fill
rasters and in the plane of the
surface of the build plate
anisotropicStrainCoefficientZ Multiplier on the predicted strain in - 1
the Z direction
hardeningFactor Factor relating the elastic modulus -
to the tangent modulus for
plasticity simulations (Tangent
Modulus =
elasticModulus*hardeningFactor )
nucleationConstantInterface Controls the heterogeneous 1/m/K^2 1000
nucleation rate (on existing solid
interfaces) during solidification for
the microstructure solver

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 159
Creating User Defined Materials

nucleationConstantBulk Controls the homogenous 1/m^2/K^2 1000

nucleation rate (in bulk of the
microstructure simulation domain)
during solidification for the
microstructure solver
penetrationDepthPowderCoeffA Coefficients used in thermal -
penetrationDepthPowderCoeffB analyses to control absorptivity and
penetration depth. These settings
penetrationDepthSolidCoeffA are determined through tuning.
penetrationDepthSolidCoeffB
absorptivityPowderCoeffA
absorptivityPowderCoeffB
absorptivitySolidCoeffA
absorptivitySolidCoeffB
absorptivityMinimum 0.2*
absorptivityMaximum 0.8*
penetrationDepthMinimum 2.7E-5*
penetrationDepthMaximum 4.1E-4*
powderPackingDensity Density of powder material relative - 0.6*
to the solid
available True
key Unique name for the material. Must -
be 16 characters or fewer, contain
no spaces, and can be the same as
the "name."
thermalSimulationApproved True or false flag signifying whether - True
the material can be used for
thermal simulations.
microstructureSimulationApproved True or false flag signifying whether - False
the material can be used for
microstructure simulations.

These inputs are laid out in a .json file format with material properties held within the configuration
parameter. An example of this file is provided as MaterialConfigurationExample.json. When editing
values from the example or creating your own material configuration file, be sure to follow the json
format to avoid errors.

Material Lookup Table

Overview

The material lookup table contains temperature dependent thermal properties for the material. This
data is required by the Thermal Solver used in Additive Science simulations as well as the Thermal
Strain simulation type in Additive Print.

The temperature dependent properties consist of thermal conductivity, specific heat, and density for
an additive manufacturing material. In addition to properties for solid and liquid material, properties

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
160 of ANSYS, Inc. and its subsidiaries and affiliates.
 User Defined Material File Specifications

for the powder state are specified through ratios for each property. For example, if the density of the
powder material is half that of the solid, a value of 0.5 would be provided for the Density Ratio.

Format

The material lookup file is a csv file with seven columns representing Temperature (K), Thermal Con-
ductivity (W/m/K), Specific Heat (J/kg/K), Density (kg/m^3), Thermal Conductivity Ratio, Density Ratio,
and Specific Heat Ratio. An example of this file is provided as MaterialLookupExample.csv.

• Temperature should be formatted from 2K to 15000K in intervals of 2K.

• Thermal Conductivity Ratio values should be as follows:

– Ratio = 0.01 from 2K to 0.6 x the solidus temperature of the material

– Ratio = 0.6 from 0.6 x the solidus temperature to the liquidus temperature of the material

– Ratio = 1 from the liquidus temperature to 15000K.

• Density Ratio values should be as follows:

– Ratio = 0.6 from 2K to the liquidus temperature of the material. (0.6 is a recommended value.
It must correspond to powderPackingDensity in the material configuration file.)

– Ratio = 1 from the liquidus temperature to 15000K.

• Specific Heat Ratio should be 1.

Material Characteristic Width Lookup Table

Overview

The characteristic width lookup table is used to find characteristic widths for different parameter
combinations. This table is generated as part of the single bead tuning process and will be used to
help determine heat source penetration depth and absorptivity for all thermal simulations. The table
contains data points for characteristic width at different power and speed combinations.

Format

The characteristic width lookup table is in csv format with columns for speed (m/s), power (W), and
characteristic width (m). Speeds and powers in the first two columns should span the range of possible
inputs for those two parameters. The power-speed combinations should be organized such that there
is a characteristic width value for each power at each speed, creating a rectilinear grid. An example
of this file is provided as CharacterisiticWidthExample.csv.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 161
 Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
162 of ANSYS, Inc. and its subsidiaries and affiliates.
Glossary
.aasp File
From "ANSYS Additive simulation package," the .aasp file format is a proprietary file
format that contains simulation inputs that can be imported into ANSYS Additive. When
imported, a new simulation form is populated under Draft Simulations. Note that this
file does not contain the part or simulation results.
Activity Status
The section of results where graphical indicators show the status of each function within
the simulation workflow.
After cutoff displacement .vtk and .avz file formats
An output file that contains the voxelized representation of the part with predicted
displacements after cutoff has occurred (either part and support cutoff or support-only
cutoff, depending on the Cutoff Mode option). Data on file include magnitude and x,
y, and z components of displacement. Included in Output Files when a user has selected
to output the displacement after cutoff.
Anisotropic
When the properties of a material vary with different orientations, the material is said
to be anisotropic. The mechanical and thermal properties of these materials differ in
different directions. (Alternately, when the properties of a material are the same in all
directions, the material is said to be isotropic.)
Anisotropic Strain Coefficients
Coefficients used to represent anisotropic strain behavior on coordinate systems aligned
with the local longitudinal, transverse, and Z (depth) scan directions. Positive values
result in compressive base strain (contraction), whereas negative values result in tensile
strain (expansion).
Anisotropic Strain Coefficient (||) Defaults to 1.5.
Multiplier on the predicted strain in the direction that the laser is scanning for the major fill
rasters.
Anisotropic Strain Coefficient ( ) Defaults to 0.5.
Multiplier on the predicted strain orthogonal to the direction that the laser is scanning
for the major fill rasters and in the plane of the surface of the build plate.
Anisotropic Strain Coefficient (Z) Defaults to 1.
Multiplier on the predicted strain in the Z direction.

ANSYS Pre-defined Materials

Materials in the Materials Library that are available for use and may not be edited.
ANSYS pre-defined materials are designed to capture the effect of a material’s chemical

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 163
Glossary

composition, powder-to-liquid and liquid-to-solid state transitions, and high cooling

rates.
ANSYS Viewer
ANSYS Viewer is an interactive 3D image viewer that is either embedded into your
ANSYS application or is available as an exportable file (on the ANSYS Customer Portal
here) to run in your web browser. Designed specifically for sharing and collaboration,
ANSYS Viewer enables you to visualize 3D models created in ANSYS CAE software even
if you do not have ANSYS software installed. ANSYS Viewer files have a .avz extension.
ANSYS Viewer is embedded in ANSYS Additive for seamless 3D visualization.
Assumed Strain Simulation

A simulation method that assumes a constant, isotropic strain (inherent strain) occurs
at every location within a part as it is being built. This is the fastest simulation method.
.avz File

A file format used by ANSYS Viewer for 3D visualization.

Baseplate

The surface of the 3D printing machine upon which the part is built. Also called the
build plate.
Baseplate Temperature ( °C ) Must be a real number between 20 and 500. Defaults to
80 °C.
The controlled temperature of the baseplate.
Bead Length Must be a real number between 1 and 10. Defaults to 3
mm.
The length of the laser scan in a Single Bead simulation.
Bead Type
An indication of how the bead is deposited in a Single Bead simulation. This setting affects
the calculation of material state, which in turn affects the material properties used in the
solving process, as well as in the laser flux model. "Bead on powder layer" is a single bead
deposited on top of a layer of powder of Layer Thickness. "Bead on baseplate" is a single
bead deposited directly on solid material.
Blade Crash

A scenario in which the recoater blade of the printing machine hits into the part already
printed, most likely due to distortion of the part as a result of residual stress.
Build File

Build files are unique to each 3D printing machine and are required to be .zip files
containing the part .stl file as well as files specifying machine scan vectors.
Build File Library

The repository for Build Files that you have imported into the program. Build Files are
formatted .zip files written for specific 3D printing machines.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
164 of ANSYS, Inc. and its subsidiaries and affiliates.
Build File Type

The machine type, such as Additive Industries, Renishaw, or SLM, corresponding to your
build file. When importing a build file, selecting the build file type assures that the
appropriate translator will be used.
Compensated Geometry .stl file format
An output file containing the distortion-compensated 3D surface representation
(tessellated triangles) of the part while the part is still attached to the baseplate. The
compensated geometry is placed flush with the baseplate surface and does not include
the offset for supports between the baseplate and the part.
Compensated Geometry (after .stl file format
cutoff)
An output file containing the distortion-compensated 3D surface representation
(tessellated triangles) of the part after cutoff has occurred (either part and support
cutoff or support-only cutoff, depending on the Cutoff Mode option).
Completed Simulation

Simulations that have either completed or that have been canceled or failed due to
error. These simulations are no longer running. Select a simulation in the Completed
Simulations area of the dashboard to see simulation results along with input parameters
and log files for that simulation.
.csv File

From "comma separated values," this is a file that allows data to be saved in a
table-structured format. Traditionally, a .csv file is in the form of a text file containing
information separated by commas, hence the name.
Customized Materials

Materials that have been edited from the original ANSYS pre-defined materials are
labeled as customized materials.
Cutoff Method

A setting under the Displacement-after-cutoff output option in strain-based simulations

designating how cutoff occurs; either Instantaneous, Directional (baseplate only), or
Legacy (if Legacy options are turned on).
Cutoff Mode

A setting under the Displacement-after-cutoff output option in strain-based simulations

designating what gets cut off; either Part and Support Cutoff (from the baseplate) or
Support-only Cutoff. In Support-only Cutoff, the support voxels are cut at the
part-support boundaries, separating the part from the support, but not the part from
the baseplate.
Dashboard

The main area, or “home,” of the Additive user interface that shows an overview of
Draft Simulations, Running Simulations, and Completed Simulations.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 165
Glossary

Deposit Layer

The layer of metal powder spread over the baseplate or melted material. Simulations
begin at a deposit layer of 1 and layers are numbered sequentially thereafter as each
successive layer is added.
Detect Potential blade crash due to distortion

An output option that activates the blade crash detection feature of Additive. If this
box is checked, there will be a check to determine if the +Z displacement at every point
in each new layer falls within certain criteria for potential blade crash. Locations of
potential blade crash and associated displacement values are provided in a .csv output
file, as well as in the On-Plate Residual Stress/Distortion and Layerwise .vtk files.
Distortion

The deformation that occurs as a result of residual stress in a part.

Distortion Compensated .stl File

An output option that activates the distortion compensation feature of Additive, which
predicts the location and magnitude of displacement and then reverse distorts the
original .stl file. When you build your part using the compensated geometry, the result
will be closer to the original design.
Distortion Compensation

One of the functions within the simulation workflow responsible for reverse distorting
the original .stl file to compensate for the effects of predicted distortion. Shown with
a status indicator in the Activity Status area of Running and Completed Simulations.
Domain

The entirety of the voxels in the simulation as defined by the bounding dimensions of
the part plus the support voxels (if any). Some voxels are part material, some are support
material and some are powder.
Dynamic Load Stepping

An option with J2-plasticity stress mode dictating how the total load will be applied to
each layer. The full load will be applied initially, and the solver will iterate until
equilibrium is achieved. If not achieved with the initial load, it will be halved and
repeated. If equilibrium is achieved, the next incremental step is applied at the current
load fraction until applying the full load, otherwise, it is halved again. A lower limit of
1/(200) load fraction is enforced, after which the solution will terminate.
Elastic Modulus Must be between 45 and 407 GPa. The default is material
dependent.
A material property that is a measure of the material's stiffness. Elastic Modulus, or
Young’s Modulus (E), describes tensile elasticity, or the tendency of an object to deform
along an axis when opposing forces are applied along that axis; it is defined as the ratio
of tensile stress to tensile strain.
Estimated Memory Usage

For Additive Desktop, an estimate of the memory required to run the simulation based
on the dimensions of the part and Voxel Size. This estimate is provided in the Geometry

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
166 of ANSYS, Inc. and its subsidiaries and affiliates.
Selection section of the simulation form as soon as you add a part. The estimate is
calculated without considering support generation.
Experimental

A label applied to any new feature that has not been fully validated, but that we feel
is stable and useful for users.
Fill Rasters

Laser scans associated with the interior regions of the part.

Fixed Load Stepping

An option with J2-plasticity stress mode dictating how the total load will be applied to
each layer. Fixed load stepping divides the load into a user-defined number of load
steps, Number of Load Steps.
Generic

The Generic machine selection in the Machine Configuration section of the simulation
form applies default input parameters that we have determined to be appropriate
through initial testing. The default machine parameters most closely approximate an
EOS machine, but may be similar to most commonly used metal laser powder bed
fusion machines.
Geometry Height, Z Must be a positive real number between 1.1 and 10.
Defaults to 1.5 mm.
The height of the cuboid, in millimeters, for the thermal solution in a Microstructure
simulation. Geometry Height must be at least Sensor Dimension plus 1 mm (for a buffer
above the sensor).
Geometry Length, Y Must be a positive real number between 1 and 10.
Defaults to 1.5 mm.
The length of the cuboid, in millimeters, for the thermal solution in a Microstructure
simulation. Geometry Length must be at least Sensor Dimension plus 0.5 mm (for a 0.25
mm buffer on each side).
Geometry Selection

This section of the simulation form where you select a part (or a build file) for simulation.
Parts (or build files) must first have been imported to the Parts (or Build File) Library.
Geometry Width, X Must be a positive real number between 1 and 10.
Defaults to 1.5 mm.
The width of the cuboid, in millimeters, for the thermal solution in a Microstructure simulation.
Geometry Width must be at least Sensor Dimension plus 0.5 mm (for a 0.25 mm buffer on
each side).
Geometry with Distortion .vtk file format
An output file containing the 3D surface representation (tessellated triangles) of the
original, undistorted part with predicted displacements at the end of the build while
part is still attached to the baseplate. The geometry does not include the offset for
supports between the baseplate and the part.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 167
Glossary

Geometry with Distortion (after .vtk file format

cutoff)
An output file containing the 3D surface representation (tessellated triangles) of the
original, undistorted part with predicted displacements after cutoff has occurred (either
part and support cutoff or support-only cutoff, depending on the Cutoff Mode option).
The geometry does not include the offset for supports between the baseplate and the
part.
Hardening Factor Must be between 0 and 0.5. The default is material
dependent.
Also known as strain hardening coefficient, a material-specific factor used to calculate
the slope of a material’s stress-strain curve (Ep) above the material’s Yield stress.
Hatch Spacing Must be between 10 and 1000 µm. Defaults to 100 µm.
The distance between adjacent scan vectors when rastering back and forth with the
laser. Hatch spacing should allow for a slight overlap of scan vector tracks such that
some of the material re-melts to ensure full coverage of solid material.
High Strain Areas

An output option that allows you to identify regions of the part that may be prone to
forming cracks during or after the build process by highlighting critical strain values.
High Strain Regions .csv file format
An output file containing a list of high strain warning areas during the build with their
associated strain values.
Inherent Strain

The residual, irrecoverable strain caused by a heat source melting or partially melting
a material in a very localized spot such that the thermal contraction of cooling solidified
material is constrained by the surrounding material. Typically associated with a welding
process.
Isotropic

Isotropic materials have identical properties in all directions. For an isotropic medium,
the stiffness tensor has no preferred direction: an applied force will give the same
displacements (relative to the direction of the force) regardless of the direction in which
the force is applied.
J2-Plasticity

One of the options for stress mode in ANSYS Additive strain-based simulations.
J2-plasticity is a part of plasticity theory that applies best to ductile materials, such as
some metals. Ductility is a measure of a material's ability to undergo significant plastic
deformation before rupture. J2-plasticity uses Von Mises stresses to reduce the stress
levels when strain values exceed elastic strain with strain hardening algorithms included.
Simulations run longer with the J-2 plasticity option but stress and strain results will
be more accurate.
Laser Beam Diameter Must be between 20 and 140 µm. Defaults to 100 µm.
The width of the laser on the powder or substrate surface defined using the D4σ beam
diameter definition. Usually this value is provided by the machine manufacturer.
Sometimes called laser spot diameter.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
168 of ANSYS, Inc. and its subsidiaries and affiliates.
Laser Powder Bed Fusion (LPBF)

In Additive Manufacturing, a method of Powder Bed Fusion (PBF) that involves spreading
a layer of metal powder and then using a laser to melt or fuse metal powder material
together to build a part. This is the method being simulated in ANSYS Additive.
Laser Power Must be between 50 and 700 Watts. Defaults to 195
Watts.
The power setting for the laser in the machine.
Layer Rotation Angle Must be between 0 and 180°. Defaults to 67°.
The angle at which the major scan vector orientation changes from layer to layer. This
is commonly 67 degrees.
Layer Thickness Must be between 10 and 100 µm. Defaults to 50 µm.
The thickness of the powder layer coating that is applied with every pass of the recoater
blade. We recommend that you use the actual thickness used for your machine and
build material.
Layerwise .vtk Files .zip file containing many .vtk files
A series of .vtk files that show voxel representation of the part layer by layer during the
build. Use these files to "animate" the build process and to view locations throughout
the part of potential blade crashes and high strain areas that may indicate cracks. You
will have one .vtk file for every voxel layer in your part, as layerwise files are not written
for support-only layers. The .vtk files are compressed into a .zip file. Since the disk space
used can become very large, especially for models with many layers, you can control
the maximum disk space used for the cumulative layerwise .vtk files with the "Maximum
storage used for layer by layer VTK files" option. Once this limit is reached, subsequent
.vtk layer files will not be written, however the simulation will continue and output files
for other selected outputs will be written, as needed. The default is 20 GB.
Linear Elastic

An assumption that a material will undergo strain linearly proportional to the magnitude
of applied stress and that the material will return to its original shape when the loads
are removed. (A simple straight line on a stress strain curve.) One of the options for
stress mode in ANSYS Additive strain-based simulations. Using this option can result in
a higher maximum stress value for a given strain beyond the yield point for the material.
This over-prediction may not be realistic for parts with larger distortions. Stresses and
strains may be unrealistically high. Distortion values will generally be accurate, however,
so the linear elastic option may be useful for analyzing distortion trends while the part
is still on the baseplate. The simulation runs faster with the linear elastic option and is
a good choice if you are just beginning with Additive and you want to run practice
simulations with default options.
Load Stepping Type

An input option if you choose J2-plasticity stress mode specifying how the total load
for each layer will be applied. Options include Dynamic Load Stepping (default) and
Fixed Load Stepping.
Logs

The section of results where the time-stamped log entries are collected. Reading log
messages is useful for following the progress of a simulation.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 169
Glossary

Machine Configuration

The section of the simulation form where you identify machine and process parameters.
You will see this section for Scan Pattern and Thermal Strain simulations only. Assumed
Strain simulations do not require inputs related to the 3D print machine.
MAPDL After Cutoff .zip file containing .rst and .cdb files
An output file containing the two files required to view the cutoff results in the ANSYS
Mechanical Application.
Material Configuration

The section of the simulation form where you identify the material. When you select a
material, properties associated with that material are automatically populated and any
related background information is tied to the simulation.
Material Library

The repository for saved materials, including ANSYS pre-defined materials and user
customized materials.
Maximum storage used for layer Must be between 1 and 8589934592 GB. Defaults to 20
by layer VTK files GB.
An input parameter required when you choose the Layer by Layer Stress/Distortion
output option. The cumulative maximum storage for layer by layer .vtk files (before
zipping). Once this limit is reached, subsequent .vtk layer files will not be written,
however the simulation will continue and output files for other selected outputs will
be written, as needed.
Maximum Wall Distance Must be between 0 and 10000µm. Defaults to 1600 µm.
Must be greater or equal to Wall Thickness.
A parameter used for the optimized volumeless supports. It is the allowed maximum
distance between two neighboring support walls. Regardless of the predicted stress
level in the support structure, the walls in supported regions will be spaced not more
than this value. Too large of a wall distance might result in failures such as the part
breaking away from the support, the development of cracks at the bottom of the part,
or drooping between support hatches. When a laser scans a relatively large area of
powder where the support wall distance is too wide, cracking might happen since
powder has no strength to hold the solidified part in place. The excessive distortion
might cause blade and part collision. We recommended that Maximum Wall Distance
should not exceed 2 mm when a single bead support wall is used.
Maximum Wall Thickness Must be between and 10000 µm. Defaults to 1000 µm (1
mm). Must be greater or equal to Minimum Wall
Thickness.
A parameter used for the optimized solid supports. Support wall thicknesses will not
exceed Maximum Wall Thickness even in areas of high stress.
Mechanics Solver

One of the functions within the simulation workflow responsible for calculating
displacements and stresses. Shown with a status indicator in the Activity Status area of
Running and Completed Simulations.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
170 of ANSYS, Inc. and its subsidiaries and affiliates.
Mesh Resolution Factor Must be between 1 and 12. Defaults to 5.
A parameter that partially controls the mesh size. Recommendation: Use the default value
of 5.
Minimum Overhang Angle Must be between 1 and 89°. Defaults to 45°.
The overhang angle is measured from the powder bed surface (horizontal = 0) up to
the surface of the part. Any point on the surface of the part having an angle less than
the Minimum Overhang Angle will be supported. Recommendation: Avoid using a value
that is the same as the angle of the geometry of your part, as it can cause asymmetric
support structures due to finite rounding errors.
Minimum Support Height Must be between 0 and 25 mm. Defaults to 5 mm.
The height that the part will be elevated off the baseplate. If you set a value of 3 mm
then the part will be elevated such that the lowest point on the part is at least 3 mm
above the baseplate. This value should be set with consideration of approximating a
realistic support height along with care about how many voxels must be created to
add additional support height. It is recommended that this value be set as low as is
realistic for each simulation.
Minimum Wall Thickness Must be greater or equal to Wall Thickness. Defaults to
100 µm.
A parameter used for the optimized solid supports. Minimum Wall Thickness is the
thinnest possible support wall that a machine will build under certain process parameters.
It is usually the thickness of a single bead scan.
Number of Cores Defaults to 4 cores.
Number of processor cores to be used in the simulation (that is, parallel processing).
Depending on your Additive license, you may have up to 12 cores to use.
Number of Load Steps Must be between 1 and 200. Defaults to 8.
The number of increments that a given load will be divided into for J2-plasticity stress
mode if fixed load stepping is chosen. A larger number of load steps will require more
loading calculations, but total time may or may not increase due to potential
improvements in convergence.
On-plate stress/displacement .vtk and .avz file formats
A result file that contains residual stresses and displacements (and optionally, strains)
of the part prior to its removal from the baseplate. Both end-state stresses and maximum
stress during the build are contained in this file, as well as potential blade crash locations
and high strain areas if those output options are selected.
Optimized Solid Support .stl file format
An output file of the optimized support structure as defined by the solid support input
parameters when automatic supports are used. The solid supports are uniformly spaced,
but wall thickness is varied based on the residual stress levels predicted. (Optimized
supports are generated by default when automatic supports are used but can be disabled
to shorten simulation run time.)
Optimized Volumeless Support .stl file format
An output file of the optimized support structure as defined by the volumeless support
input parameters when automatic supports are used. The thin-walled supports are of
a uniform wall thickness, but wall spacing is varied based on the residual stress levels
predicted in the part. (Optimized supports are generated by default when automatic
supports are used but can be disabled to shorten simulation run time.)

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 171
Glossary

Overhang Angle

Angle measured from the horizontal baseplate (0 degrees) to the surface of the part.
Any surface measuring less than the Minimum Overhang Angle will be supported.
Overview

A section of the simulation results where you can quickly see a summary status of the
simulation.
Part

The geometry for the simulation as defined by an .stl file that must be imported to the
Parts Library. This is the most common method for defining geometry.
Part Strain Threshold Must be between 1 and 100%. Defaults to 20%.
An input parameter required when you choose the High Strain Areas output option.
Defined as the percentage strain in the part above which strain will be considered
critical. (Critical regions are shown in the On-plate stress/displacements, Layerwise .vtk,
and High Strain Regions output files.)
Parts Library

The repository for all parts (as .stl files) that have been imported into the system.
Individual .stl files must be smaller than 100MB.
Poisson's Ratio

A material property that is the ratio of transverse contraction strain to longitudinal

extension strain in the direction of stretching force. Tensile deformation is considered
positive and compressive deformation is considered negative.
Positioned Part .stl file format
An output file of input geometry (non-compensated) positioned into its start location
and orientation, that is, offset to account for supports between the baseplate and the
part.
Potential blade crash locations .csv file format
An output file that contains locations of all potential and likely blade crashes and the
magnitude of the +Z displacement at those points.
Random Seed Integer between 0 and 4294967295
An optional input parameter to be used as a starting point for the nucleation pattern
in a Microstructure simulation. Specifying a number here allows you to generate the
same output for the same inputs to check for repeatability.
Residual Stress

Residual stress is the internal stress distribution locked into a material after all external
loading forces have been removed. Stresses are a result of the material obtaining
equilibrium after it has undergone elastoplastic deformation. In additive manufacturing
processes, a part undergoes repeated expansion and contraction from the heating and
cooling of the build process. This repeated heating and cooling can lead to residual
stress—a result that shows up as cracks, warpage, and other forms of deformation in
an object.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
172 of ANSYS, Inc. and its subsidiaries and affiliates.
Reverse Distort

Predicted distortion of a part is automatically passed to a distortion compensation

function providing you with an .stl file that is pre-distorted, or reverse distorted, to
compensate for process generated distortion.
Running Simulation

A simulation that is either actively running or has been queued to begin as soon as
resources are available. Select a simulation in the Running Simulations list on the
dashboard to see input parameters, activity status, and log files for that simulation.
Scale Factor Must be between 0.1 and 5. Defaults to 1.
An input parameter required when you choose the Distortion Compensated .stl File
output option. The Scale Factor will change the magnitude of the displacement applied
to the original .stl file. A Scale Factor of 1 (default) will create an .stl file with
displacement compensated by the same magnitude as the simulated results. A Scale
Factor < 1 will compensate less than the simulation-predicted magnitude and a value
> 1 will compensate more than the simulation-predicted magnitude.
Scan Pattern Simulation

This simulation type uses the same average strain magnitude as in the Assumed Strain
simulation but it subdivides that strain into anisotropic components based on the local
orientation of scan vectors within the part. This strain mode requires the creation of
scan vectors using user-provided scan settings or by reading scan vectors from a
supported machine's build file. This extra step results in a small, increased amount of
simulation time compared to Assumed Strain simulation. For parts where the scan
pattern is randomized, scan pattern and assumed strain should give a similar answer.
For parts where the scan patterns are aligned, scan pattern strain will result in a more
accurate prediction.
Scan Speed Must be between 350 and 2500 mm/sec. Defaults to 1000
mm/sec.
The speed at which the laser spot moves across the powder bed along a scan vector
to melt material, excluding jump speeds and ramp up and down speeds.
Scan Vector

Direction and velocity of one laser scan across the part. Multiple scan vectors make up
a layer’s scan pattern.
Sensor Dimension Must be between 0.1 and 1 mm. Defaults to 0.5 mm.
The dimension (width and length, in millimeters) of the 2D planes for the microstructure
solution in a Microstructure simulation. The sensor point is the intersection of all three planes
and is always 1 mm deep into the cuboid. Sensor Dimension must not exceed Geometry
Width plus the 0.5 mm buffer (0.25 mm on each side) and must not exceed Geometry Length
plus the 0.5 mm buffer (0.25 mm on each side). Sensor Dimension must be between 0.1 and
1.0.
Simulate with Supports

A check box (on/off ) option in the Supports section of the simulation form that controls
whether supports are automatically generated in the simulation.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 173
Glossary

Simulation Form

All simulations are initiated from a simulation form. It holds the inputs and selections
for your simulation. Saved simulation forms are shown under Draft Simulations in the
dashboard. Once you start a simulation, it is removed from Draft Simulations (that is,
it is no longer a “draft”) but all your input options are shown for Running Simulations
and Completed Simulations.
Simulation ID

A unique identifier for each simulation. When reporting a problem or looking for
clarification on a specific simulation, this is the number that needs to be included with
a support request. You will see the simulation ID in the Overview section of Running
and Completed Simulations.
Slicer

One of the functions within the simulation workflow responsible for “slicing” the domain
into scan vectors according to the scan pattern input parameters. Shown with a status
indicator in the Activity Status area of Running and Completed Simulations.
Slicing Stripe Width Must be between 1 and 100 mm. Defaults to 10 mm.
When using the stripe pattern for scan strategy, the geometry can be broken up into
sections, which are called stripes. The stripes are scanned sequentially to break up what
would otherwise be very long continuous scan vectors. Slicing Stripe Width is commonly
set to 10 mm wide. Memory requirements for the thermal solution will expand
significantly as you increase the Slicing Stripe Width much beyond the default.
Solver Voxel Input .vtk file format
An output file with a voxelized representation of the part and supports showing voxel
densities. This file is output early in the simulation, after voxelization but before solution
begins. If you specified automatic supports, the supports in this file are simply a
generalized density, not specifically one of the optimized supports. If you imported
support .stl files, this file shows your imported supports. You can see how well the part
and support geometries are represented by the voxel mesh and to confirm that your
supports are aligned correctly to the part.
Start (button)

Starts a simulation from the simulation form. All your input options are stored when
you run a simulation so that you may see your options at any time when you click on
a simulation in the Running Simulations and Completed Simulation areas of the
dashboard.
Starting Layer Angle Must be between 0 and 180°. Defaults to 57°.
The orientation of fill rasters on the first layer of the part. This is currently measured
from the X axis, such that 0 degrees results in scan lines parallel to the X axis. The
starting layer angle is commonly set to 57 degrees.
Status

The status of a part indicates the readiness of the part for running a simulation. When
you first import a part it will show as "processing", but there are some basic
pre-processing steps that are completed at this time, so the part is not available for a
simulation until "Available" appears in the status. (As a common practice, you can import
a part and then go to a simulation template and by the time the template is ready to

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
174 of ANSYS, Inc. and its subsidiaries and affiliates.
run the part will usually be available. When importing particularly large parts then there
is a chance that you may need to wait for import to complete.)
.stl File

From “stereolithography,” this is a 3D rendering file that approximates the surfaces of

a solid model with triangles. .stl files describe only the surface geometry of a
three-dimensional object without any representation of color, texture or other common
CAD model attributes. The .stl format specifies both ASCII and binary representations.
Binary files are more common, since they are more compact.
Strain Mode

Strain mode refers to the simulation type (Assumed Strain, Scan Pattern, or Thermal
Strain).
Strain Scaling Factor Must be between -99 and 99. Defaults to 1.
The Strain Scaling Factor (SSF) is a calibration factor used to account for differences in
machines that you may use to improve the accuracy of your simulations. This value is
a direct multiplier to the predicted strain values. Using a value of 1 will result in strain
magnitudes as calculated by the solver. Some material and geometry combinations
result in bulging/expansion rather than shrinkage and so a negative SSF is possible.
Values between -1 and 1 will reduce displacement and stress while values outside of
that range will amplify them. Using a value of 0 will result in no strain and the final
displacement will match the input geometry. The default Strain Scaling Factor is 1.
Strain Warning Factor Must be between 0.1 and 1. Defaults to 0.8.
An input parameter required when you choose the High Strain Areas output option.
This factor is multiplied by both the Support Strain Threshold and the Part Strain
Threshold to define limits where strain is labeled as a warning (that is, approaching
critical range).
Stress Mode
An input option that allows you to choose between linear elastic or elastoplastic (using
the J2-plasticity model) material behavior in calculations of stress.
Support (or Support Structure)

Support structures act as fixtures to anchor a part to a baseplate during part fabrication.
In an ideal scenario, the support density should be as low as possible so that less
material is consumed and supports can be easily removed. However, if the support
density is too low, supports can fail due to the intense strain resulting from thermal
stress accumulation in the part. The Additive application uses predicted residual stress
accumulation as criteria for support generation.
Support Factor of Safety Must be between 0.1 and 10. Defaults to 1.
An input parameter that drives the strength of the automatically generated optimized
support structures. If you would like the supports to withstand 2x the expected load,
then you would enter a 2 in this field and the predicted strength of the auto-generated
support structure would be double the predicted stress. The strength of the support
structure is driven by the number and thickness of support walls that are generated.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 175
Glossary

Support Group (or Support STL Group)

A designated group of support .stl files associated with a part that allows you to use
multiple supports in one simulation. You can mix support .stl types, that is, volumeless
supports and solid supports, in a support group.
Support Optimization

One of the functions within the simulation workflow responsible for automatically
generating supports. Shown with a status indicator in the Activity Status area of Running
and Completed Simulations.
Support Strain Threshold Must be between 1 and 100%. Defaults to 10%.
An input parameter required when you choose the High Strain Areas output option.
Defined as the percentage strain in the supports above which strain will be considered
critical. (Critical regions are shown in the Supports stress/displacement and High Strain
Regions output files.)
Support Yield Strength Ratio Must be between 0.01 and 1. Defaults to 0.4375.
The Support Yield Strength Ratio (SYSR) is a factor that is used in the simulation
assumptions to assign a strength to the support material as compared to the solid
material. It is used as a knockdown factor to modify the strength of the support material.
It affects both yield strength and elastic modulus of the support material. For example,
a value of 1.0 results in a support strength equal to the solid material while 0.5 is half
the strength of the solid material. The default SYSR value depends on which type of
supports you use in your simulation.
Supports Stress/Displacement

A result file, output in both .vtk and .avz file formats, containing the voxelized
representation of the support structure with predicted displacements and stresses at
the end of the build (that is, end state) while the part is still attached to the baseplate.
Tags

Tags are used throughout the Additive application to provide optional input for reference
and searching criteria.
Thermal Simulation

This is the method for calculating the thermal interaction of the laser and the material
at every point in a part throughout the entire build. This method takes much longer
than either of the other simulation methods, but is a much higher fidelity result.
Thermal Solver

One of the functions within the simulation workflow responsible for calculating inherent
strain fields dependent upon scan patterns (Scan Pattern simulation), or scan patterns
and thermal history (Thermal Strain simulation). Shown with a status indicator in the
Activity Status area of Running and Completed Simulations.
Threshold Scaling Factor Must be between 0.5 and 1.5. Defaults to 1.
An input parameter required when you choose the Detect potential blade crash due
to distortion output option. This factor is used to modify blade crash calculations so
that you can allow for flexibility in the recoater blade. (Potential blade crash locations

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
176 of ANSYS, Inc. and its subsidiaries and affiliates.
are shown in the On-plate stress/displacements, Layerwise .vtk, and Potential blade
crash locations output files.)
Title

The name used for a simulation. Required input on a simulation form.

Triangle Count

The number of triangular tessellation elements that define the outer surfaces of your
imported .stl geometry, making up the 3D representation of the part. You will see the
triangle count for a part on the detailed description of each part in the Parts Library.
Uniform Volumeless Support .stl file format
An output file of the geometry-based auto-generated support structure. The uniform
thin-walled supports use a uniform wall thickness and spacing and are created based
solely on geometry using the Minimum Overhang Angle parameter. These supports are
not to be used as stress optimized supports and are not recommended to use in building
parts (use the optimized supports instead).
Version

Unique identifier of the release of the ANSYS Additive application. You can find the
version number under Help > About.
Volume (mm³)

The volume of the part is calculated based upon a rough estimation of the part.
Voxel

A hexahedral (six-sided) element used in the finite element method in the Mechanics
Solver. A regular hexahedron is a cube with all its faces square. Skewed, or elongated,
sides are acceptable up to a certain amount in the solver. When combined, voxels define
the domain of the geometry.
Voxelization

One of the functions within the simulation workflow responsible for creating the
voxelized geometry, that is, dividing the domain into voxels for simulation in the
Mechanics Solver. Shown with a status indicator in the Activity Status area of Running
and Completed Simulations.
Voxel Sample Rate Must be an integer between 1 and 10. Defaults to 5.
The number of divisions on each side of a voxel used in determining Voxel Density.
The input value is cubed, for example, a sample rate of 2 means 2 by 2 by 2 = 8 samples
(that is, a voxel is divided into 8 sampling regions called subvoxels). A sample rate of
5 = 5 x 5 x 5 = 125 samples (125 subvoxels). Sample rate affects the accuracy of voxel
density. A higher number yields a more accurate Voxel Density approximation resulting
in preserved edges of a geometry.
Voxel Size Must be between 0.02 and 10 mm. Defaults to 0.5 mm.
The length of any side of the voxel (hexahedral element).

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 177
Glossary

.vtk File

From "Visualization Toolkit," the .vtk file format is an open source specification for
storing 3D computer graphics, images, and visualization data. A right-handed Cartesian
coordinate system is used.
.vtk to .avz Conversion

One of the functions within the simulation workflow responsible for converting
generic-format results files to a format appropriate for ANSYS Viewer. Shown with a
status indicator in the Activity Status area of Running and Completed Simulations.
Wall Distance Must be greater or equal to Maximum Wall Thickness
µm. Defaults to 1600 µm.
A parameter used for the optimized solid supports. It is the distance between support
walls.
Wall Thickness Must be between 20 and 500 µm.
A parameter used for the optimized volumeless supports. It is the wall thickness of the
generated support walls.
Yield Strength Must be between 10 and 4000 MPa. The default is
material dependent.
The material property defined as the stress, in MPa, at which a material begins to deform
plastically. Prior to the yield point the material will deform elastically and will return to
its original shape when the applied stress is removed. Once the yield point is passed,
some fraction of the deformation will be permanent and non-reversible.
Young's Modulus

Also known as the elastic modulus, Young's modulus is a mechanical property of linear,
elastic solid materials and is a measure of their stiffness. It defines the relationship
between stress (force per unit area) and strain (proportional deformation) in a material.

Release 2021 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
178 of ANSYS, Inc. and its subsidiaries and affiliates.