COMSOL Multiphysics

Release Notes
COMSOL Multiphysics Release Notes
© 1998–2021 COMSOL
Protected by patents listed on www.comsol.com/patents, or see Help>About COMSOL Multiphysics on the
File menu in the COMSOL Desktop for a less detailed lists of U.S. Patents that may apply. Patents pending.
This Documentation and the Programs described herein are furnished under the COMSOL Software License
Agreement (www.comsol.com/comsol-license-agreement) and may be used or copied only under the terms
of the license agreement.
COMSOL, the COMSOL logo, COMSOL Multiphysics, COMSOL Desktop, COMSOL Compiler,
COMSOL Server, and LiveLink are either registered trademarks or trademarks of COMSOL AB. All other
trademarks are the property of their respective owners, and COMSOL AB and its subsidiaries and products
are not affiliated with, endorsed by, sponsored by, or supported by those trademark owners. For a list of such
trademark owners, see www.comsol.com/trademarks.
Version: COMSOL 6.0

Contact Information
Visit the Contact COMSOL page at www.comsol.com/contact to submit general inquiries
or search for an address and phone number. You can also visit the Worldwide Sales Offices
page at www.comsol.com/contact/offices for address and contact information.

If you need to contact Support, an online request form is located at the COMSOL Access
page at www.comsol.com/support/case. Other useful links include:

• Support Center: www.comsol.com/support

• Product Download: www.comsol.com/product-download
• Product Updates: www.comsol.com/support/updates
• COMSOL Blog: www.comsol.com/blogs
• Discussion Forum: www.comsol.com/forum
• Events: www.comsol.com/events
• COMSOL Video Gallery: www.comsol.com/videos
• Support Knowledge Base: www.comsol.com/support/knowledgebase

Part number: CM010001

C o n t e n t s

Chapter 1: Release Notes

New Products 10
Model Manager . . . . . . . . . . . . . . . . . . . . . . . 10
Uncertainty Quantification Module . . . . . . . . . . . . . . . . 10

COMSOL Multiphysics 11
General New Functionality . . . . . . . . . . . . . . . . . . . 11
New Functionality in the Application Builder . . . . . . . . . . . . 11
New Functionality in COMSOL Compiler™ . . . . . . . . . . . . 15
New and Improved General Functionality in COMSOL Multiphysics . . . 15
New Geometry and Mesh Functionality . . . . . . . . . . . . . . 20
New Functionality in the Physics Interfaces . . . . . . . . . . . . . 23
New and Updated Operators, Functions, and Definitions . . . . . . . 24
New Functionality in Studies and Solvers . . . . . . . . . . . . . . 27
New and Improved Results and Visualization Functionality . . . . . . . 30
®
New Java API Methods . . . . . . . . . . . . . . . . . . . . 37
General Backward Compatibility Considerations . . . . . . . . . . . 38
Backward Compatibility with Version 5.6. . . . . . . . . . . . . . 38
Backward Compatibility with Version 5.5. . . . . . . . . . . . . . 39
Backward Compatibility with Version 5.4. . . . . . . . . . . . . . 39

AC/DC Module 41
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 41
New and Updated Models in Version 6.0. . . . . . . . . . . . . . 45
Backward Compatibility with Version 5.4 and Earlier . . . . . . . . . 46

Acoustics Module 48
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 48
New and Updated Model Examples in 6.0 . . . . . . . . . . . . . 55
Important Fixes in Version 6.0 . . . . . . . . . . . . . . . . . . 55
Backward Compatibility with Version 5.5. . . . . . . . . . . . . . 56
Backward Compatibility with Version 5.4. . . . . . . . . . . . . . 56

CONTENTS |3
 Battery Design Module 57
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 57
New and Updated Applications and Models in Version 6.0 . . . . . . . 60

CAD Import Module, Design Module, and LiveLink™ Products

for CAD 61
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 61
New Functionality in the Design Module in Version 6.0 . . . . . . . . 61
New Functionality in LiveLink™ for AutoCAD® in Version 6.0 . . . . . 62
New Functionality in LiveLink™ for Inventor® in Version 6.0 . . . . . . 62
New Functionality in LiveLink™ for PTC® Creo® Parametric™ in
Version 6.0. . . . . . . . . . . . . . . . . . . . . . . . 63
New Functionality in LiveLink™ for Revit® in Version 6.0 . . . . . . . 63
®
New Functionality in LiveLink™ for Solid Edge in Version 6.0 . . . . . 63
®
New Functionality in LiveLink™ for SOLIDWORKS in Version 6.0 . . . 64

CFD Module 65
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 65
New Model in Version 6.0 . . . . . . . . . . . . . . . . . . . 66
Backward Compatibility with Version 5.6 and Earlier . . . . . . . . . 66
Backward Compatibility with Version 5.5 and Earlier . . . . . . . . . 66
Backward Compatibility with Version 5.4 and Earlier . . . . . . . . . 67

Chemical Reaction Engineering Module 68

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 68
New Models in Version 6.0 . . . . . . . . . . . . . . . . . . . 71
Updated Applications in Version 6.0 . . . . . . . . . . . . . . . 72
Backward Compatibility with Version 5.4 and Earlier . . . . . . . . . 73

Composite Materials Module 74

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 74
New Models in Version 6.0 . . . . . . . . . . . . . . . . . . . 75

Corrosion Module 76
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 76
New and Updated Models in Version 6.0. . . . . . . . . . . . . . 79

4 | CONTENTS
ECAD Import Module 80
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 80

Electrochemistry Module 81
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 81
New and Updated Models in Version 6.0. . . . . . . . . . . . . . 84

Electrodeposition Module 85
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 85
New and Updated Models in Version 6.0. . . . . . . . . . . . . . 88

Fatigue Module 89
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 89
New Models in Version 6.0 . . . . . . . . . . . . . . . . . . . 89

Fuel Cell & Electrolyzer Module 90

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 90
New and Updated Models in Version 6.0. . . . . . . . . . . . . . 96

Geomechanics Module 97
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . . 97
New Model in Version 6.0 . . . . . . . . . . . . . . . . . . . 99

Heat Transfer Module 100

New and Improved Functionality in Version 6.0 . . . . . . . . . . 100
New Applications in Version 6.0 . . . . . . . . . . . . . . . . 108
Updated Applications in Version 6.0 . . . . . . . . . . . . . . 108
Backward Compatibility with Version 5.6. . . . . . . . . . . . . 109
Backward Compatibility with Version 5.5. . . . . . . . . . . . . 110
Backward Compatibility with Version 5.4. . . . . . . . . . . . . 111

Liquid & Gas Properties Module 113

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 113

MEMS Module 114

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 114
New Models in Version 6.0 . . . . . . . . . . . . . . . . . . 122

CONTENTS |5
 Metal Processing Module 124
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 124

Microfluidics Module 125

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 125
Backward Compatibility with Version 5.5. . . . . . . . . . . . . 125

Mixer Module 126

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 126

Multibody Dynamics Module 127

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 127
New Models in Version 6.0 . . . . . . . . . . . . . . . . . . 130

Nonlinear Structural Materials Module 131

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 131
New Models in Version 6.0 . . . . . . . . . . . . . . . . . . 134

Optimization Module 136

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 136
New and Updated Models in Version 6.0. . . . . . . . . . . . . 137
Backward Compatibility with Version 5.4. . . . . . . . . . . . . 137

Particle Tracing Module 138

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 138
New and Updated Applications in Version 6.0 . . . . . . . . . . . 139
Important Fixes in Version 6.0 . . . . . . . . . . . . . . . . . 139
Backward Compatibility with Version 5.5. . . . . . . . . . . . . 141
Backward Compatibility with Version 5.4. . . . . . . . . . . . . 142

Pipe Flow Module 143

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 143
Updated Models in Version 6.0 . . . . . . . . . . . . . . . . 143

Plasma Module 144

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 144
New and Updated Applications in Version 6.0 . . . . . . . . . . . 145
Updated Applications in Version 6.0 . . . . . . . . . . . . . . 146

6 | CONTENTS
Backward Compatibility with Version 5.5 and Earlier . . . . . . . . 146

Porous Media Flow Module 147

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 147
Backward Compatibility with Version 5.6. . . . . . . . . . . . . 150

Polymer Flow Module 152

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 152
New Models and Apps in Version 6.0 . . . . . . . . . . . . . . 153

Ray Optics Module 154

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 154
New Applications in Version 6.0 . . . . . . . . . . . . . . . . 156
Important Fixes in Version 6.0 . . . . . . . . . . . . . . . . . 157
Backward Compatibility with Version 5.5. . . . . . . . . . . . . 158
Backward Compatibility with Version 5.4. . . . . . . . . . . . . 158

RF Module 160
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 160
New and Updated Models in Version 6.0. . . . . . . . . . . . . 162
Backward Compatibility with Version 5.6 and Earlier . . . . . . . . 163

Rotordynamics Module 164

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 164
New Models in Version 6.0 . . . . . . . . . . . . . . . . . . 165

Semiconductor Module 166

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 166
New Models in Version 6.0 . . . . . . . . . . . . . . . . . . 168

Structural Mechanics Module 171

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 171
New and Updated Models in Version 6.0. . . . . . . . . . . . . 186

Subsurface Flow Module 188

New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 188
New Models in Version 6.0 . . . . . . . . . . . . . . . . . . 190

CONTENTS |7
 Wave Optics Module 191
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 191
New and Updated Models in Version 6.0. . . . . . . . . . . . . 192
Backward Compatibility with Version 5.6 and Earlier . . . . . . . . 193

Material Library 194

New and Updated Material Data in Version 6.0 . . . . . . . . . . 194
®
LiveLink™ for Excel 197
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 197
®
LiveLink™ for MATLAB 198
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 198
Backward Compatibility with Version 5.6. . . . . . . . . . . . . 199
Backward Compatibility with Version 5.4. . . . . . . . . . . . . 199
®
LiveLink™ for Simulink 200
New Functionality in Version 6.0 . . . . . . . . . . . . . . . . 200
®
The COMSOL API for Use with Java 201
COMSOL 6.0 API Changes . . . . . . . . . . . . . . . . . . 201
COMSOL 5.6 API Changes . . . . . . . . . . . . . . . . . . 202
COMSOL 5.4 API Changes . . . . . . . . . . . . . . . . . . 202

8 | CONTENTS
 1

Release Notes

COMSOL Multiphysics® version 6.0 includes the all-new Model Manager, which
provides simulation process and data management (SPDM) tools for managing
your COMSOL modeling projects. There is also a new Uncertainty Quantification
Module for performing uncertainty quantification (UQ) of parameters in
COMSOL Multiphysics simulations. These Release Notes provide information
regarding new functionality in version 6.0 for all COMSOL® software products.

9
 New Products
Model Manager
The Model Manager provides all-new functionality for working with COMSOL
projects. You can save drafts of model files, save versions of models with version control
and comments, compare versions, add auxiliary files in project databases, search and
tag models, and much more. This functionality makes the COMSOL software work as
a part of SPDM or product lifecycle management (PLM).

The Model Manager is included in COMSOL Multiphysics version 6.0. The Model
Manager server requires a floating network license and makes it possible to share
database content. The Model Manager is not available with a class kit license.

See the new Model Manager Reference Manual and Model Manager Server Manual
in the COMSOL documentation set for more information.

Uncertainty Quantification Module

The new Uncertainty Quantification Module adds functionality for studying UQ using
the following analysis types:

• Screening, MOAT for Morris one-at-a-time screening

• Sensitivity analysis
• Uncertainty propagation
• Reliability analysis

Furthermore, you specify quantities of interest and input parameters for the UQ
analysis. For the input parameters, you can choose from a number of distributions and
specify parameters for those distributions, such as mean and standard deviation, and
also add optional lower and upper bounds. For the results analysis, a number of UQ
results quantities are available in tables, and a new Sobol Index plot is available for
visualizing Sobol indices.

10 | CHAPTER 1: RELEASE NOTES

C O M S O L M ult i ph ysi cs
General New Functionality

UPDATED LICENSE SERVER

The FlexNet license server software has been upgraded to version 11.18. For that
reason, update the COMSOL license server before you install and run version 6.0.

UPDATED SYSTEM REQUIREMENTS

For up-to-date information about the system requirements for version 6.0, including
supported versions of Windows®, Linux®, and macOS operating systems, see
https://www.comsol.com/system-requirements.

New Functionality in the Application Builder

Also see the Introduction to the Application Builder, Application Builder Reference
Manual, and Application Programming Guide for more information.

GENERAL IMPROVEMENTS
• The Main Window feature includes a new Prefer inner scrollbars check box, selected
by default. This setting controls where scrollbars appear when shrinking the height
of the app window. If you run an app with the default setting and shrink the window,
you will get a vertical scrollbar to the right of the input sections. If you clear the
Prefer inner scrollbars check box and run the app again, then, when you shrink the
window, you will instead get a vertical scrollbar to the very right.
• The icons included by default in the Images library have changed, and there are new
icons (on disk) that can be used when making apps.

IMPROVEMENTS TO THE FORM EDITOR

Editor for the Main Window

You can now right-click the Main Window node and choose Edit to open the Main
Window contextual toolbar. It contains menus and menu items for creating main
window menus:

• The Classic Menu and Ribbon buttons in the Menu Type section can be used to switch
the menu type of the application. The corresponding options in the Main Window
settings window are updated accordingly. Depending on the chosen menu type, you

COMSOL MULTIPHYSICS | 11
 can add Menu Bar and Toolbar items in the Classic Menu section or File Menu, Ribbon
Tab, and Ribbon Section items in the Ribbon section.
• Open Menu opens the selected menu, and Close Menu closes the selected menu if open
or the selected item’s parent menu otherwise. After closing, the menu that was
closed is selected, if not selected already. This makes it possible to close a “tree of
menus” by clicking the button multiple times.
• Click the Item and similar buttons in the Items section to create items of the correct
type depending on the current selection (for example, toolbar items if a toolbar is
selected and menu items if a menu is selected).

Gauge Form Object

A new Gauge form object is available as a way to display a scalar value in an app using
a circular gauge with an arrow or needle pointing to the current value. Optional tick
marks and labels indicate the range of values that the gauge can display. There are also
color ranges along the scale. You can configure the colors.

More Customization to Buttons

• Outlined and flat button styles are now also available for small buttons.
• Toggle buttons can now have “radio button behavior”.
• A new Selected style list with the options Color highlight and None. This setting makes
it possible to skip the highlight in cases where the icon itself indicates the state. It is
only available for large buttons using the flat or outlined style and is not available on
Linux and macOS because apps use built-in buttons there.

Visibility for Forms in Form Collections

It is now possible to specify which form is visible in form collections. You can also
change the visible and expanded states at runtime. The Settings window for Form
Collection objects has also been updated and improved.

Error and Warning Column in Table of Toolbar Items

The table of toolbar items in the Settings windows of Toolbar, Graphics, and Table form
objects now has a column that shows error or warning icons for items that have any
error. For example, a toggle item that is missing a source.

Player Buttons in the Graphics Toolbar

When an animation is the source of a Graphics form object in an app, the toolbar will
include player buttons for playing, stopping, and stepping in the animation.

12 | CHAPTER 1: RELEASE NOTES

IMPROVEMENTS TO THE METHOD EDITOR

Improved Debugger
The debugging tools have been significantly expanded and improved in version 6.0:

• A new Variables window that opens by default when a method is paused at a

breakpoint. In the Variables window you can inspect and modify local variables,
declarations, and parameters.
• A new Call Stack window that opens by default when a method is paused at a
breakpoint. In that window, it is possible to select a stack frame to see the
corresponding local variables in the Variables window.
• A new Breakpoints window that contains a list of all breakpoints in all methods in
the application. You can open this window from the Method ribbon toolbar or from
a new Windows menu in the Home ribbon toolbar.
• It is possible to enable and disable individual breakpoints, either from the
Breakpoints window or by right-clicking on a breakpoint in the left margin in any
method editor window.
• Two new debugger actions have been added:
- Step Out: Step until the current method returns.
- Break: Suspend method execution at the next reached line in the method code,
as if there had been a breakpoint there.

Improved Method Support for Events

It is now possible to pass arguments to methods that are used in the Events section for
form objects. You can now choose Create Global Method, Create Form Method, Create
Local Method, and Create Command Sequence next to an On data change list, for example.

• Clicking directly on the plus sign corresponds to Create Global Method.

• Clicking Create Global Method or Create Form Method opens a dialog box where you
can enter the name of the method to create. A method of the corresponding type is
then created. It has a first argument called newValue, which is based on the form
object type in the same manner that the newValue argument type is determined for
local methods. The argument added to the method contains the new value in the
form object when the event is triggered. The newly created method is also selected
in the Application Builder tree and opened in the editor area.
• Clicking Create Command Sequence opens a Create Command Sequence dialog box
where a command sequence to run when the event is triggered can be specified. The
Go to Method button in the dialog box first performs an “OK operation” and then

COMSOL MULTIPHYSICS | 13
 selects the method in the Application Builder tree and opens the method in the
editor area. When OK is clicked in the Create Command Sequence dialog box, a
Command sequence item is added to the On data change list after None, and that item
is selected.

IMPROVEMENTS AND CHANGES FOR THE APPLICATION LANGUAGE

See also the Application Programming Guide for a complete overview of the
application language for programming methods in the Application Builder for use in
applications, models, and add-ins.

Methods for Calling Go to View Actions

The first method parameter in the following methods is a view orientation string from
the set {"xy", "xz", "yx", "yz", "zx", "zy"}.

// Go to view in main graphics

goToView("xy");
// Go to view in the given graphics form object.
goToView("xz", "form1/graphics1");
// Go to view in the given graphics form object.
goToView("yz", app.form("form1").formObject("graphics1"));

Methods for Clearing the Logs

In apps with multiple Log or Message Log form objects, all Log and Message Log form
objects are cleared of text. These methods are also supported in methods run from the
Model Builder.

// Clears all logs

clearLog();
// Clears all message logs.
clearMessageLog();

New FormObject Methods

The following methods are now available for a FormObject:

void visible(String form, boolean visible)

sets the form’s visible state, and

void expanded(String form, boolean expanded)

sets the form’s expanded state when using the Sections type. There are also new
Boolean formvisible and sectionexpanded properties for the visible state of forms
in form collections and the expanded state of forms in form collection when using
sections type, respectively.

14 | CHAPTER 1: RELEASE NOTES

New Functionality in COMSOL Compiler™
It is now possible to sign compiled apps on Windows.

New and Improved General Functionality in COMSOL Multiphysics

BETTER FILE SELECTION TOOLS

For opening ans saving files, you can now choose Open From and Save To on the File
menu; Insert Components From, Insert Sequence From, Insert Physics From, Import
Materials From from context menus for component and material nodes; and choose
Browse From when choosing files to insert or import. All these actions open one of
these new fullscreen windows for managing files:

• The Open window for opening files. You can choose among recently used files, add
databases, and browse for files on the file system.
• The Save window for saving files. You can add databases and browse for a location
to save the file on the file system.
• The Select Model window for selecting files for inserting components, physics, and
materials. You can add databases and browse for model files on the file system.
• The Select File window for selecting a file to insert of import. You can choose among
available files of different types (such as text files), add databases, and browse for files
on the file system.

When insert or importing files, you can also choose (when applicable) Show in Auxiliary
Data, to add it to the Auxiliary Data window, and Copy File Location and Paste File
Location for copying the file location to the clipboard and pasting an already copied file
location in any text field or document.

IMPROVED GRAPHICS AND RENDERING

• It is now easier to determine the number or name of a domain, boundary, edge,
point, or geometry object. By default, the number or name now appears in the
upper-left corner of the Graphics window when hovering over the geometry. There
is also information about what is the up and down sides of interior boundaries in
assembly geometries (requires OpenGL rendering). You can turn off this display by
clearing the Show information on hover check box on the Graphics Interaction page in
the Preferences dialog box.

COMSOL MULTIPHYSICS | 15
 • You can also control the amount of glow when hovering over the geometry. On the
Graphics Interaction page in the Preferences dialog box, choose Full glow effect (the
default), Only glow lines, No glow, or Only lines from the Primary hover effect type list.
• You can now add screen-space ambient occlusion (SSAO), which can improve the
quality of images. Select the Ambient occlusion check box in the 3D View node’s
settings to include SSAO in the view. Ambient occlusion adds soft shadows from
ambient and indirect lighting to the scene.
• For transparency plots, Fresnel transmittance is now available. In the Fresnel
transmittance field in the Transparency section of the 3D View node’s settings, or
using the slider underneath, you can specify the Fresnel transmittance as a value
between 0 (no transmittance) and 1 (full transmittance). Fresnel transmittance
results in that the transparency depends on the angle of incidence, which can
improve realism.
• When clipping solid domains, you can now show a cross section as well as colorize
it. It is also possible to highlight overlapping domains and to make the cross section
transparent.
• By default, antialiasing is now enabled for transparency rendering. You can turn the
antialiasing on or off using the Enable antialiasing for on-screen transparency effects
check box on the Graphics and Plot Windows page in the Preferences dialog box.
• You can now zoom by scrolling. Either rotate the scroll wheel while holding down
the CTRL key or rotating the scroll wheel while keeping it pressed.

IMPROVED RENDERING OF MATERIALS

Four new material types are now available for rendering of materials in the Graphics
window: Brass, Aluminum (anodized), Steel (anodized), and Silicon. A Color list for
quickly changing the color of a material is now also available for Aluminum (anodized),
Steel (anodized), Car paint, and Leather. There are also settings for specifying a custom
basis for the material coordinates and new options for brush lines.

NEW XML FORMAT FOR IMPORTING AND EXPORTING MATERIALS

There is a new XML format for importing materials into COMSOL Multiphysics,
either using a dedicated import tool or as a material library in the Material Browser. The
main purpose of this format is to provide a well-defined and documented XML syntax,
simplifying collaboration. A material supplier, for example, can generate material
libraries only with the aid of the documentation and examples that you generate by the
export tool.

16 | CHAPTER 1: RELEASE NOTES

Use the new Import Materials and Import Materials From options to import materials
from an MPH-file or a model XML file and insert it into the open model. Use the
Export Materials option to export the materials under the selected node to a Model
XML-file. You can use the export to see the correct XML syntax for the materials.

MERGING CHANGES IN COMPARISON RESULTS

For nodes that correspond to a node in the Model Builder tree, right-click the node
and choose Override Difference in Opened to overwrite that difference in the open
model. After this operation, the open model is altered to match the other model as
closely as possible. When comparing two models that both can be altered, for example,
two models saved on your file system, you can also right-click and choose Override
Difference in Remote File to overwrite the difference in the other model instead. See the
Model Manager documentation for information on merging model files with version
information.

SAVING WITHOUT BUILT, COMPUTED, OR PLOTTED DATA

In the Settings window for the root node, there is a new Save section where you can
specify if the model should be saved without built, computed, or plotted data. There
are separate settings for saving to an MPH-file and to a database. If you choose to
exclude such data, the model will be saved without the binaries for the geometry,
mesh, solution, and postprocessing data.

There are also settings on the new Save page in the Preferences dialog box to handle
the default for new models or old models that are opened.

IMPROVED DEFAULTS FOR PLOT GENERATION AND DATASETS

You can now right-click a Study node and choose Show Default Plots to create default
plots for a study and Reset Default Plots to reset the plots related to the study to the
defaults. These commands are also available in the Solver section on the Study toolbar.
In both cases, nothing is computed. There is also a new Generate default plots for
intermediately stored solutions check box in the Settings window for Study nodes.
Select it to generate a plot group for each Solution Store node in the solver sequence of
the study if such nodes appear after a Compile Equations node.

There is also a new setting on the Results page in the Preferences dialog box. Under
Dataset generation policy, select the Add datasets only for components solved for check
box to change the behavior of the dataset creation functionality and avoid creation of
unnecessary datasets.

COMSOL MULTIPHYSICS | 17
 You can also create plots automatically when running COMSOL batch jobs from the
command line using the new -createplots argument.

AUXILIARY DATA WINDOW

The Auxiliary Data window is a new window that contains a table of all files referenced
by the current model with their properties, source locations, and status. It is available
from the Windows menu in the Main toolbar. This window also includes auxiliary data
files stored in a database when using the Model Manager.

VERSIONS WINDOW
The new Versions window is intended for managing versions of files in your modeling
projects when using the Model Manager. You can open, run, compare, and restore
versions of files in the Versions window. It is available from the Windows menu in the
Main toolbar.

OTHER IMPROVEMENTS AND CHANGES

• There is a new preference for what unit system COMSOL will use for new models.
In the General section of the Preferences dialog box, under Unit system, you will find
the new In new models list. The unit system selected here will be used when a new
model is created.
• In the Windows version, you can now insert ASCII characters in text fields by
holding down ALT and then typing the character code using the number pad.
• The ribbon toolbar section previously called Physics Variables is now called Physics
Utilities. Physics Utilities also now appears as a separate submenu in the context menu
for Definitions nodes. They are no longer in the Variables Utilities submenu. The
physics utilities are always visible (provided your license includes them) and no
longer require that you activate Variable Utilities in the Show More Options dialog
box.
• In the Model Builder toolbar, the Expand All and Collapse All buttons now include a
menu that includes those actions and the new actions Expand Selected and Collapse
Selected.
• The Save recovery file preference setting has been changed to Keep recovery files
when the application terminates unexpectedly for clarity.
• Some new actions are available for the Quick Access Toolbar. They appear if you
choose More Commands and open the Quick Access Toolbar dialog box. The new
actions are Model Manager, Model Builder, Application Builder (to move to those
workspaces), and Refresh Add-ins.

18 | CHAPTER 1: RELEASE NOTES

• The Messages window has got a new Show on New Message toolbar button. When
selected, the Messages window will be brought on top when there is a new message.
Previously, the Messages window was sometimes brought on top when there was a
new message, but not in a way that you could control.

COMSOL API ADDITIONS

Setting and Reading the Model’s Title and Description

New API commands for setting and reading the model’s title and description:

model.title(title)
title = model.title()
model.description(description)
description = model.description()

Previously, the title and description were part of model.comments().

Saving Without Computed Data

There are two new methods

model.excludeComputedDataInMph(<exclude>)
model.excludeComputedDataInMph()

With the first method, you can specify whether to exclude the computed data by
providing a Boolean input variable. The second method returns the current value.

Blocking Other Clients

There is a new API command

ModelUtil.blockOtherClients(boolean blockOtherClients)

that blocks other clients from a server. This command is only meaningful when there
are multiple clients connected to the server. Calling
ModelUtil.blockOtherClients(true) will block other clients. The blocking is
lifted when clients request the block call ModelUtil.blockOtherClients(false) or
when it disconnects from the server. While the server is blocked, the COMSOL
Multiphysics desktop will show the Server Busy window.

Creating Default Plots When Running Study Steps

The runAll, runFrom, and runFromTo methods can all now take an extra Boolean
input argument (boolean createPlots) to create default plots as part of running the
study steps when it is set to true.

Trigger ModelChangedHandler When the Model Changes

The ModelChangedHandler interface in the COMSOL API now includes a new
method. It was originally intended to be used when there are multiple clients

COMSOL MULTIPHYSICS | 19
 connected to a COMSOL Multiphysics server. Now it can also be triggered when the
client registering it makes changes to the model using the new method

boolean triggeredByChangesFromThisClient()

It has a default implementation returning false, which gives the behavior in earlier
versions. In this case, the method

handleModelChangeOnServer(<modelChangeInfo>)

is only called when another client changes the model.

If triggeredByChangesFromThisClient() returns true then

handleModelChangeOnServer(<modelChangeInfo>) is also called when the client
registering the ModelChangedHandler changes the model.

New Geometry and Mesh Functionality

GEOMETRY FUNCTIONALITY
• Faster geometry build: After building the geometry sequence to a certain node,
building to a node above it is now very fast. Also, building a node below is very fast,
if it has been built before.
• You can now group nodes in the Geometry sequence, which can be useful for long
sequences. Note that the order in which the nodes appear in the sequence is still
important.
• The Selection List now includes a section for named selections, which gives an
overview of all available named selections.
• You can mark a 2D or 3D geometry object as Construction geometry. The
construction objects are used only to help with the geometry construction and are
not included in the geometry once Form Union or Form Assembly has been built. This
is mainly useful to draw construction lines (visualized as dashed lines) and points in
a 2D sketch, for use in constraints and dimensions.
• For the Difference operation, there is finer control of which objects to keep with the
Keep objects to add and Keep objects to subtract options.
• Mirror (in 2D) now supports specifying a straight edge as the line of reflection.
• The Rotate operation now supports specifying a straight edge (3D) or a point (2D)
as the center of rotation.

20 | CHAPTER 1: RELEASE NOTES

• In the settings for Intersection, there is a new option, Keep lower-dimensional
intersections, which can make the intersection of two surfaces become a curve
(instead of an empty result).
• Use the new Offset operation to offset the edges of a 2D curve or solid object.
• You can use the new Extract operation to extract entities of a geometry object. The
default is to create a new object containing the extracted entities while keeping the
original object. It is also possible to create one object for the extracted entities and
one for the remainder entities.
• Use the new Thicken operation to create a thickened solid object of a curve object
in 2D.
• Many of the primitives for creating curves in 2D are now located under the More
Primitives menu on the Geometry toolbar and in the Model Builder menu.

MESHING FUNCTIONALITY
• The Mesh toolbar and the Model Builder menu have been reorganized for better ease
of use.
• Some mesh Warnings have been converted into Information nodes to better indicate
the severity of the reported problems. For all Information, Warning, and Error nodes
that contain problem coordinates, it is possible to zoom in on the coordinate and
also add a Clip Sphere to inspect the region more closely.
• The Settings window of all mesh operations now contains an Information section
with information about building time, build version, and more.
• The Adjust edge mesh option for Mapped has been improved to better reduce
element skewness.
• The new functionality for Mesh Plots include:
- A new default color table, Traffic Flow, designed to highlight elements of poor
quality more clearly. The color table also works better for people with color
blindness.
- The possibility to color the elements by their type. Set Element color: Type in the
Settings window of the Mesh plot node.
• You can now do componentwise copies from the entire source to each destination
component. Select Copy method: Array copy in the Settings window of the operation
to specify this. The option Copy method: Automatic allows for a mixture of single and
array copy.

COMSOL MULTIPHYSICS | 21
 • For the Distribution attribute, some names have been updated when Distribution type
is set to Predefined. The change also affects similar options for the Create Edges
operation:
- Growth formula is now Growth rate.
- Arithmetic sequence is now Linear.
- Growth formula is now Exponential.
• The Boundary Layers operation is updated with new functionality:
- Specification of the Total thickness for the layers.
- No trimming is done where isolated boundary-layer boundaries connect to
exterior boundaries.
- Support to add boundary layers for imported 2D and 3D meshes.
• The Import domain elements option is now available for the Import operation. Clear
this option to import the surface mesh only. This allows for remeshing of the faces
or other types of editing of the imported mesh.
• Curvature of imported linear 2D and 3D meshes is now taken into account. The
software estimates the shape of the faces and edges from its linear elements. This
curved representation is used to:
- Place higher-order nodes.
- Export of second-order elements for an imported linear mesh.
- Place new or moved mesh vertices when modifying imported meshes with the
Refine or Adapt operations.
• From a Filter dataset, it is now possible to select Create Mesh in New Component in
addition to the Create Mesh Part option.
• Intersecting and uniting surface meshes and repairing self-intersecting surface
meshes is now possible using the new Union operation.
• The new Transform attribute is now available to position, rotate, or scale an
imported mesh. The attribute is only available for the Import feature.
• You can now use the Fill Holes operation to fill slit-like holes in an imported surface
mesh. This is done by merging the two sides of the hole, introducing new mesh
edges when needed.
• It is now possible to combine imported mesh (STL) with CAD/geometry using one
component only. The updated tutorial STL Import 2 — Remeshing an Imported
Mesh shows this workflow.

22 | CHAPTER 1: RELEASE NOTES

• The Build automatically when leaving mesh option controls if the mesh operations are
automatically built when leaving the meshing sequence. This option is only available
for imported meshes on which the physics is defined.

New Functionality in the Physics Interfaces

FLUID-SOLID INTERACTION, FIXED GEOMETRY

The Fluid-Solid Interaction, Fixed Geometry interface is now available without any
additional module license.

FULLY DEVELOPED FLOW

The Fully developed flow option for the Inflow and Outflow boundary conditions in
the Laminar Flow interface is now available without any additional module license.

INCLUDE GRAVITY AND OPTION TO USE THE REDUCED PRESSURE

The Include gravity option and the associated Use reduced pressure option in the
Settings window for the main Laminar Flow physics interface node are now available
without any additional module license.

MASS FLOW OPTION FOR THE PERIODIC FLOW CONDITION

In the Settings window for the Periodic Flow Condition in the Laminar Flow interface,
you can now choose Mass flow from the Flow condition list without any additional
module license.

IMPROVED PRESSURE CONDITION FOR INCOMPRESSIBLE FLUID FLOW

The total (or stagnation) pressure can now be specified on inlet and outlet boundaries
for incompressible flow. Both local and averaged values of the total pressure can be
imposed. This new feature functionality can, for example, be used in pump
simulations.

WEAK INEQUALITY CONSTRAINT

You can now add a Weak Inequality Constraint node on a domain, boundary, edge, or
point in any physics interface. You must first select Equation-Based Contributions in the
Show More Options dialog box. This constraint feature makes it possible to specify a
complementary nonlinear inequality constraint.

COMSOL MULTIPHYSICS | 23
 IMPROVEMENTS FOR MOVING MESHES AND DEFORMED GEOMETRIES
You can now press Ctrl+Space to add expressions to moving mesh and deformed
geometry features.

New and Updated Operators, Functions, and Definitions

NEW IMPROVED PAIR FUNCTIONALITY

The handling of pairs — identity pairs and contact pairs — has been improved in the
following ways:

• The fallback features under each pair feature have been removed. For pairs, there
can be parts of the boundaries that do not overlap, which typically represent exterior
boundaries outside of the overlapping area. For those parts, the other boundary
conditions under the physics interface will contribute with proper conditions, acting
as fallback boundary conditions on the nonoverlapping parts. These boundary
conditions have an indicator in the lower-left corner displaying their status as
fallback boundary conditions.
• Most physics interfaces now have a default pair feature, and the default boundary
condition will act as the default fallback condition for all pairs. If you do not want
to use one of the pairs under Definitions in your physics interface, you can select the
Disconnect pair check box to disconnect all the pairs. Then, the default pair feature
will not contribute to any conditions, and the fallback conditions from the other
boundary condition will apply to the selection of the default pair feature.
• Pair features can override each other.

See also Pair Features under COMSOL 6.0 API Changes.

NEW NORMAL DISTRIBUTION FUNCTION

The new Normal Distribution function represents a normal distribution with a bell-
shaped (Gaussian) curve. It supports the following statistical functions related to a
normal distribution: density function, cumulative distribution function, quantile
function, and randomly generated function.

IMPROVED POROUS MATERIAL FUNCTIONALITY

The Porous Material node, used for defining and sharing properties for modeling
porous media processes, has been significantly improved.

• The Phase-Specific Properties section lists all properties requested by physics features
with support for a porous material.

24 | CHAPTER 1: RELEASE NOTES

• For each phase-specific property, the phase is indicated in the right-most column.
• When requested, multiple entries of the same quantity, one for each phase, can be
listed.
• Tables listing properties can be sorted using any of the columns.
• By right-clicking any property, it is now possible to select Go to Requesting Mode to
navigate to where it is used. In the same manner, it is also possible to select Go to
Source to inspect how the property is defined.
• The Homogenized Properties section lists all properties that are not phase specific, like
the permeability. It also lists properties requested by interfaces without support for
a porous material. The homogenized properties can be entered directly by selecting
Locally defined in the Material list.
• When defining properties in a Fluid, Solid, or Pellet subnode to Porous Material,
Locally defined can be selected, in the Material list, to enter the properties directly.

A new subnode, Pellet, is available for Porous Material. The new subnode is used to
provide properties for solving mass and energy transport inside porous pellets in a
packed bed.

IMPROVED UNIT SETTINGS IN FUNCTIONS

You now define units of input arguments in Analytic functions using a table with a row
for each argument. A new Unit column is also available in the Plot Parameters section
for Analytic function nodes.

The Step, Ramp, Random, and Rectangle functions now support unit syntax in their
settings.

SELECTION INFORMATION
The Override and Contribute option in the Show More Options dialog box has been
renamed Selection Information. You would typically use it to turn this functionality off
to improve performance in complex models with a lot of physics features.

OTHER NEW FUNCTIONALITY

• For Interpolation functions, None is now available in the Extrapolation list. It can be
useful to prevent extrapolation of interpolation functions outside the range of the
given data.
• The Base Vector System and Rotated System coordinate systems now include settings
for specifying an origin.

COMSOL MULTIPHYSICS | 25
 • A new Events Sequence node is now available in the Events interface. This node
defines an ordered sequence of events, as defined by the added Sequence Member
subnodes.
• The sqrteps and poweps built-in functions are now available. They are similar to
sqrt and pow, but with an added, very small eps value in their derivatives, which
can be needed when the derivatives of the expressions need to be evaluated near
zero.
• You can now control the identification method for identifying meshes in the source
and destination in the Settings windows for Period Boundary Conditions and Identity
Pair nodes, if desired. Choose one of the following alternatives from the Elementwise
mapping for compatible meshes list: On, Off, or Require compatible meshes. Choose On
if you expect compatible meshes and want to use an elementwise mapping, but you
want to get a mapping even if source and destination cannot be identified. Choose
Off if you do not expect compatible meshes or if you want to use elemental
constraints for continuous fields. Choose Require compatible meshes if you expect
compatible meshes and you would rather fail than use noncompatible meshes.
• Import of images on the TIFF format (*.tif and *.tiff files) is now available in
Image functions.
• A fast alternative for coupling operators is now available by selecting Walk in
connected component from the Mesh search method list.
• In the settings for Expression Operator nodes, there is a new Argument type column,
where you can choose Expression (the default) or Value. Use Value to pass expression
operator arguments by value instead of as an expression.
• In the Settings window for State Variables and Expression Operator nodes, you can
now use Ctrl+Space to add parameters, built-in mathematical constants and
functions, operators, and physical constants to define expressions.
• In the settings for the State Variables node, Only initialization is now available in the
Update list. It can sometimes be useful to ensure that a complicated or expensive
expression is evaluated only once.
• Material nodes now have a Name field to change their identifier. This setting also
changes the tag at the same time and makes it possible to control the scope of
variables and functions that the material defines. There is also a new Name field
available for property groups under materials that change the identifier of the
property group.

26 | CHAPTER 1: RELEASE NOTES

New Functionality in Studies and Solvers

STUDIES AND STUDY STEP FUNCTIONALITY

• For each individual study step you can alter the equation form of a physics interface
if the study step shows the tree view in the Physics And Variables Selection section.
When the study step shows the table view, you can see what equation form applies
to a physics interface in a new Equation form column, and an icon representing the
type of changes made for the study step is available. The Discretization column
available in earlier versions has been removed. Also, when you have selected the
Modify model configuration for study step check box, you can now modify state
variables and matrix and vector utilities in the model tree.
• The Model Reduction study step, Modal Solver solver node, and generated reduced-
order model nodes under Global Definitions>Reduced-Order Modeling have all been
updated and extended to support component mode synthesis (CMS).
- For the reduced-order modeling, you can choose a stateful or stateless interface
in the Time-Dependent Modal Reduced-Order Model and Frequency-Domain Modal
Reduced-Order Model nodes, which appear under Global Definitions>Reduced-Order
Modeling. A reduced-order model with a stateless interface is a black box with a
number of inputs and outputs. A reduced-order model with a stateful interface,
in contrast, relies on the main COMSOL solvers to solve a set of reduced-order
equations in terms of states (degrees of freedom), which are exposed to the main
solver and subsequently stored in the main solution vector.
- In the settings for the Model Reduction study step, you can now choose a
training study and a study step for constraint modes. There is a new Use extra
Compile Equations for Results check box. This setting controls whether an extra,
active Compile Equations node will be added last in the default solver sequence. It
then includes the newly created reduced-order model feature in the final solution
output by the study so that the reduced-order model’s outputs can be evaluated
directly using this solution.
- The Modal Reduction solver node’s settings includes a new Constraint Modes
section with settings for specifying which solution to be used when constructing
the reduced model and which of the constraint modes present in the solution to
include when constructing the reduced model.
- In the Compile Equations node’s settings, there is a new Geometric Entity
Selection section. This section contains a Use entities list, where you can choose
All (the default) and Selected. When Selected is chosen, you can choose a named
selection in the Selections list. Solving on a subset of the entire geometry can be

COMSOL MULTIPHYSICS | 27
 useful for component mode synthesis (CMS) and general submodeling
simulations.
- In the settings for the Parametric Sweep node, you can now select the Keep
reduced-order models check box. If selected, such models are not discarded, and
you can specify what to add to the tag for those models (the parameter name and
value, for example). You can also choose to clear previous models so that you can
update all reduced-order models by running the study again and to keep the
generating reduced-order models, so that they are saved in addition to the
reduced-order model copies.
• In the settings under Adaptation and Error Estimates in study step nodes, when you
choose Functional from the Error estimate list, a new Interpolation option is available
in the Adjoint solution error estimate list. It provides an estimation of the
interpolation error in an expression to get a better weight related to the dual
solution in the error estimate based on a functional.
• The new Auxiliary-Space AMG iterative solver is suitable for solving complex-valued
RF problems and other electromagnetics simulations that use vector elements
(Nédélec elements). Compared to the Auxiliary-Space Maxwell (AMS) solver, the
Auxiliary-Space AMG solver supports complex arithmetic. More intermediate
smoothers are available, which makes it possible to, for example, use Auxiliary-Space
AMG when a gauge feature is used or, in general, to use it combined with Vanka
smoothers.
• You can now use automatic remeshing for stationary parametric studies that use one
continuation parameter. If Auxiliary sweep is enabled in the Study Extension section
for a stationary study step, you can now select an Automatic remeshing check box and
select in which geometry to remesh.
• The linearization point settings have been removed from the study level for the
following study types: Frequency Domain, Frequency Domain Modal, Eigenvalue,
Eigenfrequency, Linear Buckling, Boundary Mode Analysis, and Mode Analysis.
• Equation-related settings now have a separate Equations section in domain
decomposition and multigrid solvers. You no longer have to activate the Prefer the
matrix free format option when using absorbing boundary conditions and shifted
Laplace contributions. You can now use the Schwarz method for domain
decomposition to solve Helmholtz equation problems.
• In the Domain Decomposition (Schwarz) solver node’s settings, you can now choose
Nicolaides from the Use coarse level list. The Nicolaides coarse space is a simple way
of constructing a coarse space for Schwarz domain decomposition. It does not need

28 | CHAPTER 1: RELEASE NOTES

 any additional information, such as a coarser mesh, except the local system matrices
from the local domains.
• There are new options for filtering lines in the SCGS smoother to provide slimmer
line construction for use with the AMG solver.
• A local Schur complement approximation for the Vanka and SCGS smoothers is
available as a more “lightweight” version of the Vanka solver that can store the data
permanently without using too much memory. For Vanka, choose Direct, stored
factorization from the Block solver list. For SCGS, select the Vanka check box to access
these settings.
• In the settings for parameters, for example, in the Values of Dependent Variables
section in the settings for study step nodes, the following improvements and
changes have been made:
- The previous Automatic and All options have been renamed to Automatic (single
solution) and Automatic (all solutions), respectively.
- The Automatic (all solutions) option is now also available in the Initial values of
variables solved for list. When selected, all solutions are passed down to the solver
and the solver selects the initial values based on if there is a match to the current
parameter tuple; otherwise, the logic from previous versions is used.
• To improve time-dependent adjoint sensitivity analysis for DAEs, new backward
time stepping is available. In the Sensitivity node’s settings, a new Backward Time
Stepping section is available when you have selected Adjoint as the sensitivity method.
Use the new From forward or Manual options if needed.
• A new Drop tolerance for local matrix contributions field is available in the Advanced
solver node’s settings. It is primarily useful for semiconductor simulations.
• In the settings for the Model Reduction study step node, the Compute list in the
Model Reduction Settings section has been removed because it is no longer
transparent, useful, or safe to use. The corresponding API property,
trainingRecompute, has also been removed. The behavior in 6.0 corresponds to
the previous Initially option, and you will have to compute or recompute any
dependent studies manually.

CLUSTER AND BATCH FUNCTIONALITY

• A new Parallel modified independent set parallel coarsening method for the algebraic
multigrid solver is now available. It makes the algebraic multigrid solver more
efficient in distributed computing. The old coarsening method is still available
under the name Classical. You can select these methods in the Coarsening method list
for the Algebraic multigrid solver: Parallel modified independent set (the new default)

COMSOL MULTIPHYSICS | 29
 and Classical (the previous implementation). New options are also available to
further reduce memory consumption and computation time per iteration. The
Sparser prolongator option (enabled by default) gives better performance per
iteration (less memory and computation time) but the solver may require more
iterations to converge. You can use the Prolongator truncation factor to control the
fill-in of the prolongator.
• Improved file handling for batch jobs in the settings for Batch Sweep and Cluster
Sweep nodes:
- Under Before sweep: The Clear meshes and Clear solutions check boxes are now
available and selected by default.
- Under During sweep: The Clear synchronization data and Save model to file after
synchronization is completed check boxes are now available when you select the
Synchronize solutions check box.
- There is also a new Clear All Data option in the settings for the Batch Data and
External Process nodes.
• There are new settings for controlling the use of METIS for DOF numbering on
cluster. In the settings for the Solution and Compile Equations nodes under Solver
Configurations, a new Partitioning method for distributed computing list is now
available. This setting affects the partitioning of the mesh data on cluster for the
purpose of creating a DOF enumeration optimized for the particular cluster
configuration (number of nodes).
• There are new options for running Hydra when initializing the MPI environment
on Linux. See COMSOL Commands on Linux in the COMSOL Multiphysics
Reference Manual for more options.

New and Improved Results and Visualization Functionality

GENERAL NEW RESULTS FUNCTIONALITY

In version 6.0, only the relevant postprocessing features, dependent on the physics in
the current COMSOL Multiphysics simulation model, are visible. To show all features
under Results, select the All Dataset Types, All Plot Types, All Numerical Evaluation Types,
and All Export Types check boxes in the Show More Options dialog box.

30 | CHAPTER 1: RELEASE NOTES

ROTATION CENTER IMPROVEMENTS
Setting the rotation center now works on all platforms, including macOS. The
following functionality has also been improved:

• The automatic placement of the rotation center has been improved so that it snaps
to the center of the axis-aligned bounding box of whatever geometrical entity
(geometry objects, domains, boundaries, or edges) that is found at the center of the
screen (if the bounding box is found suitable; that is, it is mostly seen on screen and
its center is seen on screen).
• Moreover, the algorithm for the automatic placement of the rotation center now
searches around the position where you zoom into or out from after an interactive
zoom with the mouse.
• You can Ctrl+middle-click or Ctrl+click the scroll wheel to set the rotation center at
the center of the axis-aligned bounding box of the geometrical entity that you
currently hover over. As before, middle-clicking only will explicitly set the rotation
center on the surface of the geometry over which you have positioned your cursor.

NEW AND IMPROVED DATASET AND EVALUATION FUNCTIONALITY

• New Transformation 3D and Transformation 2D datasets are available for rotations and
other affine transformation of data.
• Under Derived Values>Measure, you can now add the following nodes for computing
various geometric measurements: Volume Measurement, Surface Measurement, Line
Measurement, and Distance Measurement.
• The Global Matrix Evaluation feature has been extended to handle the following
cases:
- A parameter sweep where the innermost two parameter levels both are inner.
- A parameter sweep where the innermost two parameter levels both are outer.
The Inner solutions and Outer solutions lists in previous versions have been renamed
to Inner parameters and Outer parameters, respectively, or show a dynamic level for
structured sweeps where the two innermost columns are both inner and both outer.
• The new Kernel Density Estimation dataset takes a table and computes the kernel
density estimation (KDE), which is a nonparametric way to estimate the probability
density function of a random variable.
• You can now add a Selection subnode to Shell dataset nodes, making it easier to
postprocess large shell models. Selection subnodes are now also available for Cut
Plane datasets.
• It is now possible to specify rotations in 2D and 3D Array datasets.

COMSOL MULTIPHYSICS | 31
 • There is a new Spatial FFT dataset for use in applications where the spatial resonance
and harmonics of a scalar variable or expression are of interest.
• In Grid datasets, the point distribution for the grids can now be uniform,
exponential, or mixed uniform and exponential.
• For the following numerical evaluation features under Derived Values and evaluation
groups, it is now possible to normalize the values within each column:
- Average
- Global
- Integration
- Maximum
- Measure
- Minimum
From the Normalization list, you can choose to normalize values with respect to the
first, last, and maximum absolute value within each column. Also, the Operation list
in earlier versions has been renamed as Transformation in version 6.0.
• For Filter and Partition datasets, you can now right-click and choose Create Mesh in
New Component to add a new Component node and import the modified geometry
from the dataset as a mesh in that component. You can also right-click and choose
Mesh Import Parameters to control which solution parameters to use for the data
transferred from the dataset to a mesh part.
• You can now use a transparent background for exported animations.
• Through-thickness evaluation options have been added to Cut Point datasets that
refer to Layered Material datasets.
• Area and line scale factors are now available in Shell datasets. Also, a Through-
Thickness section has been added to Shell datasets that are chosen in a numerical
evaluation feature that evaluates on a geometry.
• Shell datasets now include a location setting to specify where in the Shell dataset that
evaluation takes place.
• It is now possible to create 2D contour plots from the data in a Table or Evaluation
Group. This could be useful in, for example, AC/DC models for generating
efficiency map plots.
• For Extrusion datasets, you can now specify the extrusion direction.
• You can now snap cut points in 3D to the closest edges in addition to the closest
boundaries (surfaces).

32 | CHAPTER 1: RELEASE NOTES

• Edge and Surface datasets are now supported as inputs to Join datasets.
• For Cut Line and Parametric Curve datasets, there are now variables with unit
length that evaluate to the tangent direction.
• In the Integration data series transformation, there is a new Cumulative option that
makes it possible to compute the cumulative sum or integral.
• When exporting plots on the PLY and 3MF file formats, there is now an option to
include vertex colors, which can make duplicated vertices unique.
• The Touchstone export capability is now available with the MEMS Module and
AC/DC Module.
• For WAV file export, there is now an option to add a reference value for
normalization of the WAV data.
• Tables can now be exported to Microsoft® PowerPoint®, maintaining a link to
COMSOL Multiphysics so that you can modify the layout of the tables and update
them using a new COMSOL Table ribbon in PowerPoint.
• When you are evaluating a minimum or maximum under Derived Values, there is
now an option in the Configuration section of the Settings window to include the
position of that maximum or minimum by adding the coordinate values for that
position to the table for the evaluated quantities.

NEW AND IMPROVED GRAPHICS AND PLOT FUNCTIONALITY

• With the new Plot Array functionality, you can easily arrange the plots in a 2D or 3D
plot group in an array (linear or square), with options to exclude some of the plots
and to control the amount of padding (empty space) between each plot in the plot
array. Such a plot array can be useful for comparing and presenting multiple
solutions with different physical quantities, time steps, parameter values, or
eigenmodes.
• A new Translation subnode can also be used to shift a plot a given distance to put
two plots side by side, for example. You no longer need to use an “artificial
deformation” in a Deformation subnode for this purpose.
• A new Streamline Multislice plot simplifies the creation of plots that combine
streamlines with slices and for which you previously had to combine multiple Cut
Plane datasets and as many Streamline Surface plots.
• A new 1D Sobol Index plot is available with the Uncertainty Quantification Module
to visualize Sobol indices for an uncertainty quantification study.

COMSOL MULTIPHYSICS | 33
 • A number of new and updated color tables have been added:
- New Dipole and DipoleDark color tables, which are useful for electromagnetic
fields and other field quantities with large variations.
- New Prism and PrismDark color tables, which are alternatives to the Rainbow
color scale. Prism is now used as the default color table for solid mechanics
models.
- New Cividis and Viridis color tables, which are suitable for users with red-green
color vision deficiency.
- New Inferno, Magma, and Plasma color tables, which all use a blue-red-yellow
sequence. They are also relatively friendly to common forms of color vision
deficiency.
- New Gaia and GaiaLight color tables span colors that make them suitable for
plots related to topography and bathymetry.
- The new TrafficFlow color table is specifically designed for visualizing mesh
quality. The green half has a very soft gradient, so all good mesh elements have
about the same color when using a reversed TrafficFlow color table. This makes
it much easier to focus on the mesh elements with a low quality.
- DiscoDark and RainbowDark are new variations of the Disco and Rainbow color
tables, respectively.
- The Cyclic, Disco, DiscoLight, Rainbow, RainbowLight, Spectrum, Traffic,
TrafficLight, Wave, and WaveLight color tables have been updated.
- New GrayBody and GrayBodyLight color tables are based on the Planckian locus
for blackbody radiation. They are useful within the context of metal processing,
for example.
- New SpectrumLight color table that complements the Spectrum color table.
- The Thermal color tables have been revised. There is a new ThermalDark color
table, and the Thermal and ThermalLight color tables have been revised. The
ThermalNarrow and ThermalEquidistant color tables from earlier versions have
been removed and are now replaced with the new Thermal and ThermalDark
color tables, respectively. There are also new ThermalWave and
ThermalWaveDark color tables that are designed for wave phenomena of a
thermal nature, such as thermoviscous acoustics and linearized Navier–Stokes
physics.
The color tables have been revised to, for example, have less saturated colors, have
smoother luminance curves, or be more symmetric. For the color tables that have

34 | CHAPTER 1: RELEASE NOTES

 been revised, the corresponding color tables from earlier versions are available
with a Classic suffix (for example, RainbowClassic).
A small plot of the color table is now visible when choosing a color table from a
Color table list to make it easier to choose a suitable color table.
• There is now a Color table transformation list in the plot nodes’ Settings windows.
You can choose None, Reverse, Nonlinear, and Nonlinear symmetric. For the nonlinear
options, you can reverse the color table and control the nonlinearity using a color
calibration parameter. For the None and Reverse options, you can now use a
logarithmic color scale in the plots. Also, the previous Symmetrize color range check
box has been moved to a new Scale list as Linear symmetric together with the new
Logarithmic option and the default Linear option. In connection with this change,
the new API property colorscalemode replaces the colortablesym property in
earlier versions of COMSOL Multiphysics.
• Color tables can now span from the minimum mesh quality to the maximum mesh
quality for mesh plots.
• There is a new Information section in the Settings window for all plots and graphs to
display different quantities useful for benchmarking against different settings and
setups, such as the time that it took to prepare the data for the plot. The information
appears in a table that is saved with the model and updated each time the plot had
to be redrawn. This section is hidden by default, and you can enable it in the Show
More Options dialog box by selecting the Plot Information Section check box.
• For Isosurface plots, you can now choose to plot the isosurfaces as closed volumes
using a new Filled isosurface type.
• You can now choose Discrete Fourier transform from the Transform list for the x-axis
data to use a discrete Fourier transform (DFT) in Table Graph, Impulse Response, and
Global 1D plots.
• You can now add transparency to 2D plots with a height attribute.
• A From theme option for colors is now available in streamline plots.
• There are new filtering options in 1D Global and Point graph plots that make it
possible to draw noncontinuous lines when the x-axis data expression is
nonmonotonic. They are available in the Line Segment Selection section in the
Settings window for the Filter subnode.
• For streamline plots, you can now use mesh-based streamline starting points on
selected boundaries.
• It is now possible to use the element type as element color in 2D and 3D mesh plots.

COMSOL MULTIPHYSICS | 35
 • In plot groups, there is a new option for dataset edges to use the same frame as the
underlying dataset.
• Point Graph and Line Graph plots now support Layered Material datasets and
include a setting for specifying the through-thickness location.
• The layer/interface settings in Layered Material datasets have been extended to
handle models containing multiple shells.
• It is now possible to filter line segments with respect to direction for Global, Line
Graph, and Point Graph 1D plots to remove unwanted edges in such plots.
• There is now an option to add a grid in the Matrix Histogram plot for use in fatigue
simulations.
• It is now possible to use evaluations of expressions in legends in graph plots.
• When evaluating variable expressions in annotations in plots, the eval() operator
supports a new third input argument for the numerical precision.
• It is now possible to reverse the direction of Edge datasets.
• You can now choose Dataset space variable from the Parameter list for the x-axis data
in Line Graph plots, which can be useful for line graphs of Spatial FFT datasets, for
example.
• The Table Graph plots now also support row-based tables (that is, tables with data
on rows and sample points on columns).
• You can now add padding to Table Annotation plots, which makes it possible to get
better-looking results when zooming.

NEW AND IMPROVED REPORT AND PRESENTATION FUNCTIONALITY

• For PowerPoint presentations, you can now use two-content PowerPoint slide
layouts.
• Presentations now include link information for the COMSOL add-in for
PowerPoint for images and tables in presentations.
• You can now use a custom view for general report images.
• Reports can now include styled text area toolbar buttons and highlighting for
bulleted and numbered lists.
• Reports now have full support for Microsoft Word templates, including background
(for example, colors, gradients, and graphics); header (for example, watermarks);
and footer settings.

36 | CHAPTER 1: RELEASE NOTES

New Java® API Methods
The following methods are now available:

GENERAL GEOMETRY COMMANDS

model.component(<tag>).geometricModel();

Returns the tag of the component’s geometric model on which the physics is defined.

model.component(<tag>).geometricModel(<sequencetag>);

This method specifies the geometric model used in the component.

GEOMETRY OBJECT INFORMATION

Boolean assem = model.component(<ctag>).geom(<tag>).isAssembly();

Use this method to check if this is a geometry of assembly type (originating from a
Form Assembly feature).

Boolean virt = model.component(<ctag>).geom(<tag>).isVirtual();

Use this method to check if the geometry is a virtual geometry.

Boolean gM = model.component(<ctag>).geom(<tag>).isMesh();

Returns true if the component’s geometric model is represented by a mesh (rather than
a geometry).

MESH SETTINGS
String gM = model.component(<ctag>).mesh(<tag>).geometricModel();

This method returns a string with the tag of the geometry or meshing sequence, which
defines the geometric model of the mesh.

model.component(<ctag>).mesh(<tag>).geometricModel(boolean);

Use this method to specify if the meshing sequence defines its own geometric model.

model.component(<ctag>).mesh(<tag>).autoRebuild(boolean);

This is a user interface setting that controls if the mesh sequence is automatically
rebuilt when clicking a node in the model tree outside the mesh sequence.

model.component(<ctag>).definedByMesh(<tag>);

Use this method to control on which geometric model to define the physics.

COMSOL MULTIPHYSICS | 37
 General Backward Compatibility Considerations
COMSOL Multiphysics version 6.0 can open MPH-files saved from COMSOL
Multiphysics versions 4.0–5.6.

COMSOL Multiphysics version 6.0 can run Java® files saved from COMSOL
Multiphysics versions 4.0–5.6. However, the Java® files may need to be modified in
accordance with information in this document and may need to be recompiled with
the comsol compile command in version 6.0.

Backward Compatibility with Version 5.6

PAIR FEATURES
The new framework for pair features can affect old models, although the goal is that
the effect is as small as possible. When opening an old model in the Model Builder, pair
features are migrated using the following main steps:

1 Add a default pair feature with the Disconnect pair check box selected to ensure that
the default pair feature does not contribute any new constraints or equations to the
model.
2 Move all fallback features that were located under the pair features in 5.6 to the level
below the physics interface. The selection of these moved features will typically get
the same active selection as the pair feature they belonged to in 5.6. Because those
moved features now have an overlapping selection with that pair feature, they will
act as an equivalent fallback feature to the pair feature.
3 Optimize all moved fallback features by merging exact duplicates into one feature,
including existing features such as the default features. In many situations, the result
will be that only the default feature of the physics will remain because most fallback
features under pair features in 5.6 were identical to the default feature of the physics
interface.

Potential Issues
If the optimization in Step 3 above cannot merge all features, some moved fallback
features will remain and be placed before their pair feature. Normally, a pair feature in
5.6 overrode all features before it, so if the moved fallback also overrides these features,
the selections should not change for them. However, there can be configurations
where the selection may change because of special schemes for overriding and
contributing between features. One example is shell interfaces that have pair features
on the same level as the physics interface. The pair feature can then contribute with a

38 | CHAPTER 1: RELEASE NOTES

feature that the moved feature will override. Such cases are difficult to detect during
migration and can create a discrepancy in some selections between the versions.

PHYSICS AND VARIABLES SELECTIONS

The Discretization column is removed from the table to the more advanced tree option,
which is shown when the Modify model configuration for study step check box is
selected.

USER DOCUMENTATION NODES IN THE PHYSICS BUILDER

The User Documentation nodes in the Physics Builder are no longer available in
version 6.0.

Backward Compatibility with Version 5.5

The selections imported from Filter datasets in version 5.5 are now given tags by a
corrected scheme that makes them independent of the language setting in COMSOL.
If you have written API code using such selections, the code might need to be updated
to handle the new naming principle.

In version 5.6, the getAdjSparse method has been deprecated and may be removed
in a future version.

In the Octave Plot, the Solid setting for Type is now called Bar, and the Curve setting
is now called Outline. The corresponding API property values are now bar and
outline.

The default value for the compact aggregation setting for the smooth aggregation
AMG multigrid solver, from version 5.6, is set to off.

Backward Compatibility with Version 5.4

GEOMETRY REPRESENTATION ARRAYS

To reflect the new serialization format for 3D objects represented using the COMSOL
kernel, the meaning of the arrays returned from the GeomInfo methods getFace,
getPVertex, and getPEdge applied to a 3D geometry object has changed slightly in
version 5.5. In version 5.4, row 3 of getFace contained the index of the surface in the
list of all manifolds (curves and surfaces) in the geometry object. In version 5.5, it is
instead an index in the list of all surfaces in the geometry object. Similarly, row 5 of
getPVertex and row 9 of getPEdge now contain an index in the list of surfaces
instead of an index in the list of manifolds.

COMSOL MULTIPHYSICS | 39
 INTERPOLATION CURVES
For the Interpolation Curve geometry feature, the default end conditions have been
changed from Zero curvature to None. In the API, the default has been changed from
zerocurv to none. In addition, the algorithm has been changed for the case when Type
is Open curve and Relative tolerance is 0, which can give a slightly different shape of the
curve.

SCALING 2D GEOMETRY OBJECTS

In version 5.4, you could scale and move a set of selected 2D objects in the Graphics
window by dragging its bounding box. In version 5.5, you can still move objects by
dragging, but scaling requires adding a Scale feature in some cases.

FREE QUAD MESHING

Models saved in versions 5.3 to 5.4 with the free quad tessellation method set to
Automatic (in a Free Quad feature) or face meshing method set to Quadrilateral (in a
Swept feature), will, when opened in version 5.5, have the tessellation method or face
meshing method set to Legacy version 5.4 or Quadrilateral (legacy version 5.4),
respectively.

PARTICLE TRAJECTORIES STUDY STEP

The Particle Trajectories study step has been removed in version 5.5. Instead, use a
Time Dependent study step, which has exactly the same settings.

IMAGE EXPORT PROPERTIES

For image export features, the properties options, title, legend, logo, and axes
have been split into three properties in version 5.5, one for each spatial dimension. For
example, options1d, options2d, and options3d. axes does not exist in 3D. The
same change has been done for all image() features.

40 | CHAPTER 1: RELEASE NOTES

AC / D C M o dule
New Functionality in Version 6.0

MAGNETOMECHANICS MULTIPHYSICS INTERFACES

Two new physics interfaces for analysis of coupled magnetic and mechanical effects
have been added: Magnetomechanics and Magnetomechanics, No Currents. When
you add such an interface, two physics interfaces are added to the model: Solid
Mechanics and either Magnetic Fields or Magnetic Fields, No Currents. The new
Magnetomechanical Forces multiphysics coupling is also added. You can find these
interfaces under the Electromagnetics and Mechanics branch in the AC/DC folder in the
Add Physics tree.

Magnetomechanics is demonstrated in the new plate_deflected_by_magnet tutorial

model. In addition to the AC/DC Module, these interfaces require either the MEMS
Module, Structural Mechanics Module, or Acoustics Module.

PERIODIC PAIR FOR THE MAGNETIC FIELDS INTERFACE

The new Periodic Pair feature in the Magnetic Fields interface simplifies the modeling
of linear machines in 2D (machines such as motors, generators, actuators, and sensors).
Instead of modeling the entire device, periodicity is assumed: Typically one or a couple
of unit cells are included, ignoring the end effects.

The Periodic Pair feature is demonstrated in the new linear_motor_2D tutorial model.

ARKKIO TORQUE CALCULATION FOR THE ROTATING MACHINERY,

MAGNETIC INTERFACE
Rotating machines typically have a cylindrical air gap between rotor and stator, where
electromagnetic fields are exchanged and forces and torque are exerted. The new
Arkkio Torque Calculation feature in the Rotating Machinery, Magnetic interface
analyzes the fields in this gap and makes it possible to easily evaluate the total torque
exerted on the rotor and stator in both 2D and 3D. Also, several postprocessing
quantities important to motor design are now available. These include the radial force
density, azimuthal force density, radial magnetic flux density, and azimuthal magnetic
flux density in the air gap.

AC/DC MODULE | 41
 The Arkkio Torque Calculation feature is demonstrated in the following tutorial models:
pm_motor_2d_introduction, pm_motor_3d, and
electric_motor_parameter_optimization.

MAGNETOSTRICTION FOR THE ROTATING MACHINERY, MAGNETIC

INTERFACE
Magnetostrictive materials exhibit free strain under the influence of a magnetic field.
This is the effect causing the low-pitched humming sound coming from a transformer,
for instance. While previously only available for the Solid Mechanics and Magnetic
Fields interfaces, the Magnetostriction multiphysics coupling can now also be used to
connect the Solid Mechanics and Rotating Machinery, Magnetic physics interfaces. It
is available in 3D, 2D, and 2D axisymmetry.

Magnetostriction is demonstrated in the nonlinear_magnetostriction tutorial model.

In addition to the AC/DC Module, this multiphysics coupling requires either the
MEMS Module, Structural Mechanics Module, or Acoustics Module.

FREQUENCY DOMAIN SUPPORT FOR THE MAGNETIC FIELDS, CURRENTS

ONLY INTERFACE
The Magnetic Fields, Currents Only interface, designed to efficiently compute the
lumped inductance matrix of complex circuits in 3D, has now been extended to
include frequency-domain support. This is done using the new Frequency Domain
Source Sweep with Initialization study step. It can be used to calculate the partial
contributions from magnetic fields generated by open conductors, such as those
commonly found on printed circuit boards (PCBs).

The main difference compared to the Stationary Source Sweep, introduced in version
5.6, is that the frequency-domain solution considers skin and proximity effects. This
makes it possible to extract inductance matrices at low to medium frequencies (in the
range where capacitive effects are not yet dominant). Also, the automated
postprocessing for resistance and inductance matrices has been significantly improved.

The Frequency Domain Source Sweep with Initialization study step is demonstrated in the
inductance_matrix_pcb_coils tutorial model.

SPLIT BY CONNECTIVITY OPTION FOR TERMINALS, CONDUCTORS, AND

THE LUMPED PORT
The new Split by Connectivity option is available on the context menu for Terminal
nodes in (for example) the Electrostatics and the Magnetic and Electric Fields

42 | CHAPTER 1: RELEASE NOTES

interface, for Conductor nodes in the Magnetic Fields, Currents Only interface, and for
the Lumped Port in the Magnetic Fields interface.

This option is useful when a model contains many conductors — as is the case for a
PCB or a touchscreen. In those cases, having a single selection containing all
conductors is enough: You can add a single Terminal feature to your model, choose all
conductors as its selection, and then choose Split by Connectivity to automatically
generate a separate Terminal feature for each disconnected part of the selection.
Especially for terminal sweep studies and inductance matrix extraction, this will greatly
speed up your workflow.

This functionality is demonstrated in the inductance_matrix_pcb_coils and

capacitive_position_sensor tutorial models.

LAYERED TRANSITION BOUNDARY CONDITION FOR THE MAGNETIC

FIELDS INTERFACE
At low to medium frequencies, the new Layered Transition Boundary Condition in the
Magnetic Fields interface can be used to model composite electromagnetic shielding
materials such as those containing mu-metal and copper layers or, for instance, gold-
plated copper on a circuit board. At higher frequencies, this feature can even be used
for modeling RF metamaterials and antireflection coatings, although for those
applications, you would typically use the same feature in the RF Module or the Wave
Optics Module. The feature is used in combination with a Layered Material in the
global Materials node, and a Layered Material Link in the Materials node. Together with
the extended Electromagnetic Heating, Layered Shell multiphysics coupling, the Layered
Transition Boundary Condition supports fully coupled induction heating in the layer
stack.

FULLY COUPLED ELECTROMAGNETIC HEATING FOR THE LAYERED

TRANSITION BOUNDARY CONDITION
The Electromagnetic Heating, Layered Shell multiphysics coupling has been extended to
support the Layered Transition Boundary Condition in the Magnetic Fields interface for
frequency-stationary and frequency-transient study types in 2D and 3D. The Layered
Transition Boundary Condition takes the layer-averaged (temperature dependent)
material properties for each individual layer, and uses them to compute the
electromagnetic field propagation and power dissipation in the layer stack. The layer-
averaged heat sources are then fed back into the thermal simulation. Together with the
Thin Layer feature in the Heat Transfer in Solids interface, this allows you to investigate
the temperature distribution in the layer stack. Typical use cases include the induction
welding of layered composite materials, for example.

AC/DC MODULE | 43
 SYMMETRY PLANE FEATURES FOR VARIOUS INTERFACES
The new Symmetry Plane feature is available for the following physics interfaces:
Electrostatics; Electric Currents; Magnetic Fields; and Magnetic Fields, No Currents.
The symmetry planes for Electrostatics and Electric Currents provide symmetry and
antisymmetry conditions for the electric field. For the Magnetic Fields and Magnetic
Fields, No Currents interfaces, the symmetry type can be set for the magnetic flux
density B and the magnetic field H, respectively.

The Symmetry Plane features are demonstrated in several tutorial models, including:
vector_hysteresis_modeling, permanent_magnet, and
dielectric_shielding_comparison.

DEBYE DISPERSION MODELS FOR DIELECTRICS

New damping models have been added for dielectric materials. Under Charge
Conservation, when the material type is set to Solid, you can now use the Dispersion
dielectric material model. In the Dispersion subnode, you can choose between the
Debye and Multipole Debye dispersion models.

NEW ALGEBRAIC MULTIGRID PRECONDITIONER FOR LINEAR CURL

ELEMENTS
The Auxiliary-Space Algebraic Multigrid (ASAMG) preconditioner is an implementation
of the AMS (Auxiliary-Space Maxwell) solver. For the AC/DC Module, it is now
available as an alternative preconfigured solver in the default solver configuration for
the Magnetic Fields interface (when the discretization of that interface is set to linear
in 3D).

GEOMETRY PART LIBRARY FOR ELECTRIC MOTORS

A new Geometry Part Library has been added specifically for modeling rotating
machines in 2D. It contains three internal permanent magnet rotors and one external
stator (equipped with coils). The rotor designs include the following types: embedded
magnet, embedded magnet (V-shape configuration), and surface-mounted magnet.

The geometry parts are demonstrated in the following tutorial models:

pm_motor_2d_introduction, generator_2d, sector_generator_3d, and
electric_motor_parameter_optimization.

IMPROVED PLOT DEFAULTS

As an extension to the default plot improvements included in version 5.6, the AC/DC
Module’s plot defaults now use new color tables and new plot features such as the
Streamline Multislice plot. A new Prism color table has been designed specifically for

44 | CHAPTER 1: RELEASE NOTES

vector field norms with singular tendencies such as the electric field norm or the
magnetic flux density norm. Multislice and streamline plots are combined to indicate
both field strength and direction. For physics interfaces that use a scalar potential
formulation — Electrostatics; Electric Currents; Magnetic and Electric Fields; and
Magnetic Fields, No Currents — extra plot groups are added that use the symmetric
Dipole color table.

TOUCHSTONE EXPORT SUPPORT IN POSTPROCESSING

The Electrostatics, Electric Currents and Magnetic Fields interfaces can generate S-
parameter matrices through port or terminal sweeps. In addition to the old
Touchstone export functionality, these physics now support the Touchstone Export
functionality available during postprocessing.

ELECTRIC CIRCUIT IMPROVEMENTS

The Electric Circuit interface has received a number of fixes and extensions. These
improvements include the support for importing and exporting the mutual inductance
element K and more robust and general SPICE import.

New and Updated Models in Version 6.0

The following models are new in version 6.0:

ELECTRICAL HEATING IN A BUSBAR WITH TERMINALS

This example demonstrates how to use a terminal fed by a current to compute the
resistive heating in a busbar.

PERMANENT MAGNET MOTOR 2D

In this tutorial, a PM motor with 10 rotor poles and 12 stator slots is modeled in 2D
to capture the torque ripple over an electrical period and map the volumetric loss
density in the rotor and stator iron.

LINEAR MOTOR IN 2D
In this model, a synchronous motor with a three-phase winding on the stationary part
and permanent magnets on the moving part is investigated. A moving mesh and
periodic conditions are utilized in order to model a periodic part of a complete motor.
It is also explained how to use this model for a 2D representation of an axial flux motor
(rotational motion).

AC/DC MODULE | 45
 PARAMETER OPTIMIZATION OF AN ELECTRIC MOTOR
This example demonstrates how to use parameter optimization to identify the optimal
position and shape of the permanent magnets in an electric motor. The torque is used
as the objective function, but different designs are found depending on whether the
minimum or average torque is considered.

SHAPE OPTIMIZATION OF COILS

This example shows how to use the Optimization Module to find a coil geometry
giving a uniform magnetic field on axis and minimal field near the axis ends.

DEFORMATION OF AN IRON PLATE BY MAGNETIC FORCE

This example studies the elastic deformation and stress in a ferromagnetic plate when
deflected by a permanent magnet. The deformation of the plate has an influence on
the distribution of the magnetic field. This effect is accounted for using a moving mesh
in the air surrounding the plate.

UPDATED MODELS
In addition, the following models have been updated:

• Magnetic Damping of Vibrating Conducting Solids

• Inductance Matrix Calculation of PCB Coil
• Topology Optimization of a Magnetic Circuit
• Inductance of a Power Inductor

Backward Compatibility with Version 5.4 and Earlier

REMANENT FLUX DENSITY MATERIAL PROPERTY GROUP

With the introduction of the remanent flux density material property group, support
for the remanent flux density constitutive relation with the relative permeability given
by an effective medium has been dropped. In other words, a linear (hard) magnetic
material where Br is set and μr given by an Effective medium is no longer supported.
User-defined expressions for μr are still possible and may serve as a substitute.

REFACTORING OF MODEL INPUTS

For all AC/DC Module interfaces, some model input functionality has been moved to
the Default Model Inputs node.

For a temperature-dependent boundary material used in an Impedance boundary

condition, for example, this means the default model inputs should be used, rather

46 | CHAPTER 1: RELEASE NOTES

than the feature’s model input section (the model input section in the Impedance
boundary condition’s Settings window).

For more frequently occurring combinations, like a temperature-dependent domain

material and a current conservation law, both the default model inputs as well as the
feature’s model input section are available as a means of specifying the temperature.

AC/DC MODULE | 47
 Acoustics Module
New Functionality in Version 6.0

PIEZOELECTRIC WAVES, TIME EXPLICIT PHYSICS INTERFACE

With the Piezoelectric Waves, Time Explicit multiphysics interface, there are new
capabilities for modeling piezoelectric phenomena in the time domain for wave
propagation. Both the direct and inverse piezoelectric effects can be modeled, and the
piezoelectric coupling can be formulated using the strain-charge or stress-charge
forms. The new interface couples the Elastic Waves, Time Explicit interface with the
Electrostatics interface using the new Piezoelectric Effect, Time Explicit multiphysics
coupling. The interface is based on the discontinuous Galerkin (dG or dG-FEM)
method and uses a time-explicit solver. The electrostatics problem is solved at every
time step through an algebraic system of equations solved with classical FEM. This
ensures a very computationally efficient hybrid method that can solve very large
models with many millions of degrees of freedom (DOFs). The method is well suited
for distributed computing on cluster architectures.

PHYSICS-CONTROLLED MESH FOR PRESSURE ACOUSTICS

Physics-Controlled mesh is introduced for the Pressure Acoustics, Frequency Domain and
Pressure Acoustics, Transient interfaces. The generated mesh follows best practices to
resolve the wave problem with an adequate number of mesh elements. Perfectly
matched layers are meshed with a structured mesh, periodic conditions use copy mesh
operations, and a single boundary-layer mesh is used for exterior field calculations. For
the frequency domain, the maximum frequency is automatically picked up from the
study. For time-domain models, the maximum frequency is taken from physics,
ensuring a consistent spatial resolution with the mesh and temporal resolution of the
transient solver. For all other acoustic interfaces and for Solid Mechanics, only PMLs and
Periodic Conditions are handled in this version.

INTRODUCING FLOW-INDUCED NOISE

With this release, a hybrid computational aeroacoustic (CAA) method is introduced
for modeling flow-induced noise. It is based on a one-way coupling between the
turbulent flow sources and the acoustic problem. It is assumed that no back-coupling
exists from the acoustics field to the flow. The computational method is based on the
finite element method (FEM) discretization of Lighthill’s acoustic analogy (wave
equation). This formulation of the equations ensures that any solid (fixed or vibrating)

48 | CHAPTER 1: RELEASE NOTES

boundaries are implicitly taken into account. Both the Lighthill and the simpler
aeroacoustic wave equation (AWE) analogies can also be selected.

The new functionality relies on coupling a large eddy simulation (LES) fluid flow
model to the Aeroacoustic Flow Source domain feature in Pressure Acoustics, Frequency
Domain. The coupling is achieved by using the Aeroacoustic Flow Source Coupling
multiphysics coupling and the dedicated Transient Mapping study.

NEW HIGH-FREQUENCY PRESSURE ACOUSTICS PHYSICS INTERFACES

Two new dedicated physics interfaces are based on the Kirchhoff–Helmholtz integral,
so they both rely on a high-frequency assumption. The Pressure Acoustics, Asymptotic
Scattering interface is useful for modeling scattering problems, while the Pressure
Acoustics, Kirchhoff-Helmholtz interface is primarily for modeling radiation problems.

Pressure Acoustics, Asymptotic Scattering

The Pressure Acoustics, Asymptotic Scattering interface is used to model scattering
problems at high frequencies using the Kirchhoff–Helmholtz integral formulation.
The acoustic field is assumed to be a locally plane field such that the scattered field can
be expressed analytically. The scattering object surface can be treated as perfectly
reflecting or having absorbing properties by defining a surface normal impedance, a
reflection coefficient, or an absorption coefficient. The latter two can depend on the
angle of incidence. The interface can model scattering from spherical and plane waves.
The interface has built-in functionality to compute the visibility factor using both a
simple angle consideration and a more advanced hemicube method.

Pressure Acoustics, Kirchhoff–Helmholtz

The Pressure Acoustics, Kirchhoff-Helmholtz interface is used to model radiation
problems at high frequencies using the Kirchhoff–Helmholtz integral formulation.
The acoustic field is assumed to be a locally plane field. This technique is also often
referred to as high-frequency BEM or simply HFB. The method is often used for
computing the radiated acoustic field from vibrating structures at high frequencies,
without the need to model the surrounding fluid. The method is valid as long as the
acoustic wavelength in the fluid is smaller than the structure and the structural modes.
For a flat vibrating surface, the method reduces to solving the Rayleigh integral. A
user-defined option gives access to the full Kirchhoff–Helmholtz integral formulation
for defining both pressure and its normal gradient.

ACOUSTICS MODULE | 49
 LUMPED SPEAKER BOUNDARY AND INTERIOR LUMPED SPEAKER
BOUNDARY
For many electroacoustics applications, speakers can be efficiently modeled by
combining a lumped Thiele–Small representation and the finite element method. The
electromagnetic components of the motor are modeled with an Electric Circuit, while
the acoustics are solved with the Pressure Acoustics, Frequency Domain interface. This
approach assumes piston motion of the speaker diaphragm, so it is of particular interest
for microspeakers. The Interior Lumped Speaker Boundary includes the effects of the air
on both sides of the diaphragm. The Lumped Speaker Boundary has options to include
the compliance of a back volume through an impedance.

SECTOR SYMMETRY OPTIONS FOR EXTERIOR FIELD CALCULATION

The Exterior Field Calculation feature has been extended with new options to handle
models with sector symmetry. Two options extend the existing symmetry planes
functionality: the Sector symmetry and Sector symmetry with one symmetry plane
options. The latter is of particular interests when modeling loudspeaker drivers placed
in an infinite baffle. The analysis of the exterior field can also be extended with an
Azimuthal mode number for advanced sector-symmetric problems.

OPTIMIZATION WITH NEW EXTERIOR FIELD VARIABLES

In 3D, new dedicated exterior field variables exist for use in gradient-based
optimization problems, such as shape optimization or topology optimization. The
optimization objective can now be defined as a variable evaluated in the exterior field,
defining, for example, values on a radiation pattern or an off-axis response. The
variable exists only for the Symmetry planes option in the Exterior Field Calculation
feature. The new variables are defined with _opt appended to the existing exterior field
variables. The operator for the pressure is pext_opt(x,y,z), and for the sound
pressure level it is Lp_pext_opt(x,y,z).

PERFECTLY MATCHED BOUNDARY (PMB) RADIATION CONDITION

The new Perfectly Matched Boundary is effectively a perfectly matched layer (PML) that
is applied to the open boundary as a radiation condition, without the need to define a
domain (a layer in the geometry). The condition automatically applies the PML
formulation using the extra dimension functionality in COMSOL Multiphysics. This
approach also simplifies the requirements to the radiating boundary as it can, in
principle, have any convex shape. Different options exist for controlling the
attenuation direction. The condition exists for the Pressure Acoustics, Frequency Domain
physics interface in all relevant space dimensions.

50 | CHAPTER 1: RELEASE NOTES

LINEARIZED NAVIER-STOKES, BOUNDARY MODE PHYSICS INTERFACE
The Linearized Navier-Stokes, Boundary Mode interface is used to compute and identify
propagating and nonpropagating modes in waveguides and ducts in the presence of
any stationary isothermal or nonisothermal background mean flow. It performs a
boundary mode analysis on a boundary, inlet, or cross section of the waveguide. The
interface considers all thermal and viscous loss effects and the interaction with the
background flow. This includes the acoustic boundary layer losses, if necessary.

LINEARIZED EULER, BOUNDARY MODE PHYSICS INTERFACE

The Linearized Euler, Boundary Mode interface is used to compute and identify
propagating and nonpropagating modes in waveguides and ducts in the presence of a
stationary background mean-flow that is well approximated by an ideal gas flow. The
interface performs a boundary mode analysis on a boundary, inlet, or cross section of
the waveguide.

OUT-OF-PLANE AND BOUNDARY MODE ANALYSIS FOR LINEARIZED

EULER AND LINEARIZED NAVIER-STOKES
In both the Linearized Euler and the Linearized Navier-Stokes interfaces, it is now
possible to add optional out-of-plane and circumferential wave numbers in 2D and 2D
axisymmetry, respectively. In the same space dimensions, there is now also support for
the mode analysis study. This functionality makes it possible to set up so-called 2.5D
simulations in these physics.

COMPUTE DISPLACEMENT POSTPROCESSING FEATURE IN ELASTIC

WAVES, TIME EXPLICIT
A new postprocessing feature, Compute Displacement, has been added to the Elastic
Waves, Time Explicit physics interface. The feature allows the computation of the
displacement in points, along edges, on boundaries, or in domains by solving auxiliary
ODEs. The new features are added as subfeatures to a material model such as the
Elastic Waves, Time Explicit Model or Piezoelectric Material nodes. The feature is a pure
postprocessing type, and the variables can be used to visualize displacements. This
feature adds and solves an ODE and thus requires additional computational resources.

MINIMUM AND MAXIMUM PRESSURE POSTPROCESSING FEATURE IN

NONLINEAR PRESSURE ACOUSTICS, TIME EXPLICIT
A new postprocessing feature, Compute Minimum and Maximum Pressure, has been
added to the Nonlinear Pressure Acoustics, Time Explicit interface. This feature computes
the maximum and minimum pressure over time and space in a domain or on a

ACOUSTICS MODULE | 51
 boundary. Two variables are automatically created, nate.p_min and nate.p_max,
which can be used in postprocessing to, for example, evaluate the size of a focal zone.

PAIR MULTIPHYSICS COUPLINGS FOR VIBROACOUSTIC SIMULATIONS

Two new pair multiphysics couplings have been added to the Acoustics Module for
coupling acoustic and solid domains in an assembly. These couplings make it possible
to use nonconforming meshes at the interface between the two domains.

• Pair Acoustic-Structure Boundary multiphysics coupling for coupling the Pressure

Acoustics, Frequency Domain or the Pressure Acoustics, Transient interface to the Solid
Mechanics interface, in an assembly geometry where identity pairs have been created.
• Pair Thermoviscous Acoustic-Structure Boundary multiphysics for coupling
Thermoviscous Acoustics, Frequency Domain or the Thermoviscous Acoustics, Transient
interface to the Solid Mechanics interface, in an assembly geometry where identity
pairs have been created. The condition uses a penalty formulation for more efficient
computations in the time domain.

UPDATES TO THERMOVISCOUS ACOUSTICS

• For ports, a new Plane Wave option for handling slip and adiabatic cases or models
that are not set in a waveguide context. The names of the existing waveguide modes
have been updated to Numeric (0,0)-mode and Circular (0,0)-mode.
• The default discretization in all thermoviscous interfaces has been changed from
Lagrange to serendipity elements. This results in important performance gains when
solving models that use a structured mesh (for example, a mapped or extruded
mesh).
• Improved solver settings when solving models with Nonlinear Thermoviscous
Acoustics Contributions. The settings are especially important for highly nonlinear
problems. If you open a version 5.6 model, remember to reset the solvers to default.
• Improved coupling between Thermoviscous Acoustics and structures that have
curved surfaces by using a Nodal constraint type.
• In the Boundary Layers mesh, a new Thickness specification option, All layers, has been
added. This allows to more easily set up a mesh that spans the entire thermoviscous
acoustic boundary layer.
• The material properties for the second-order density approximation option in
Nonlinear Thermoviscous Acoustics Contributions can now be evaluated if the air
material (moist or dry) is specified using a Thermodynamics system.

52 | CHAPTER 1: RELEASE NOTES

• In the Thermoviscous Acoustics, Frequency Domain interface, it is now possible to add
an optional circumferential wave number for 2D axisymmetric models. This allows
for advanced 2.5D analysis of complex propagation patterns.

SOURCE WITH DIRECTIVITY

In the Ray Acoustics interface, the new Source with Directivity node can be used to
release a distribution of rays with initial intensity or power based on a user-defined
spatial directivity function. This is of particular interest when defining loudspeaker
sources in ray acoustic simulations.

NEW RAY ATTENUATION MODELS

In the Ray Acoustics interface, the attenuation of rays can now be specified by entering
any of the following:

• Pressure amplitude attenuation coefficient in nepers per meter

• Intensity attenuation coefficient in nepers per meter
• Pressure amplitude attenuation coefficient in decibels per meter
• Pressure amplitude attenuation coefficient in decibels per wavelength

All of these options are supported in domains as well as the exterior void domain
outside the geometry.

SIMPLIFIED NONLOCAL COUPLING NAMES IN RAY ACOUSTICS

The built-in couplings that evaluate the sum, average, maximum, and minimum of
expressions over all rays have been renamed. In an instance of the Ray Acoustics
interface with the default name rac, the old name for the sum was rac.racop1(); the
new name is rac.sum(). Similarly, the average, maximum, and minimum are
rac.ave, rac.max, and rac.min, respectively.

Backward Compatibility
The old coupling names, such as rac.racop1() in the example above, will be labeled
as deprecated but they will still be supported in version 6.0.

TRANSFORMATIONS OF LOADED RAY COORDINATES

For the Ray Acoustics interface, when loading the ray position and direction from a file
using the Release from Data File node, the distribution of initial ray positions can now
be dilated, rotated, and translated. If a rotation is applied, you can also choose to apply
the same rotation to the initial ray directions if they are also loaded from the same file.

ACOUSTICS MODULE | 53
 RELEASE FROM EXTERIOR FIELD CALCULATION
In the Ray Acoustics interface, the Release from Exterior Field Calculation node can be
used to launch a distribution of rays with initial intensity or power proportional to the
solution of an Exterior Field Calculation from a previous study. Several copies of the
exterior field calculation can be released, each with its own translation and rotation.

POSTPROCESSING NEWS FOR ACOUSTICS

• New Discrete Fourier transform (DFT) and Moving average options have been added
when analyzing an Impulse Response plot in Ray Acoustics.
• A new Information section is available on plots when Plot Information Section is
enabled in the Show More Options dialog box. The new section displays the plot time,
which can be considerable in Ray Acoustic models or models using the new
Kirchhoff–Helmholtz-based high-frequency interfaces.
• In the Octave Band plot, the plotted Quantity can now be selected as Continuous
power spectral density, Band power, or Band average power spectral density.
• A warning message has been added in the Energy Decay subfeature when the
behavior of the level decay deviates from normal.
• New Thermal Wave and Thermal Wave Dark color schemes have been added. The
Thermal Wave color table is the new default for the temperature variation default
plot in the thermoviscous acoustics and linearized Navier–Stokes physics interfaces.
• New updated Wave and Wave Light color tables have been added and are used as
default colors for all plots that represent the acoustic pressure.
• There is an important speedup of postprocessing of large surface plots.
• The preview of Receiver datasets is now much faster.

IMPORTANT ENHANCEMENTS AND UPDATES IN THE ACOUSTICS MODULE

• Improved formulation of damping for the dG-FEM-based time-explicit interfaces.
The new formulation is more stable, more accurate, and results in faster simulations.
• Better handling of lumped circular approximations in thermoviscous acoustics
models (Bessel functions) for narrow region acoustics and for interior perforated
plates.
• The robustness of the mapping performed by the Mapping study, in combination
with the Background Fluid Flow Coupling multiphysics feature, has been improved.
This is particularly useful for models with an imported geometry and for models that
have a CFD boundary layer mesh on curved boundaries.

54 | CHAPTER 1: RELEASE NOTES

• The Compressible Potential Flow physics interface is now applicable in PML domains,
which simplifies setting up the background flow for the subsequent analysis in the
Linearized Potential Flow interface.

API RELEASE NOTES

Note 1
The default discretization has changed for the Thermoviscous Acoustics interfaces.
Add the following lines to obtain the old behavior of a physics interface ta:

model.component("comp1").physics("ta").prop("ShapeProperty").set(
"shapeorder_u", 2);
model.component("comp1").physics("ta").prop("ShapeProperty").set(
"shapeorder_T", 2);

Note 2
In the Linearized Navier–Stokes physics interfaces, the default value for the
Stabilization constant has changed. Add the following line to obtain the old behavior:

model.component("comp1").physics("lnsf").prop("Stabilization").se
t("alpha", "1e-2");

New and Updated Model Examples in 6.0

• Chamber Music Hall
• Schroeder Diffuser 2D
• Loudspeaker Spider Optimization
• Tweeter Dome and Waveguide Shape Optimization
• Angle Beam Nondestructive Testing
• Submarine High-Frequency Asymptotic Scattering
• Several models have been updated to use the new Physics-Controlled mesh available
for pressure acoustics.
• Updated Ultrasonic Flowmeter with Piezoelectric Transducers to use the new
Piezoelectric Waves, Time Explicit multiphysics interface.

Important Fixes in Version 6.0

ACCURACY IMPROVEMENTS FOR BOUNDARY ACCUMULATORS

For the Ray Acoustics interface, the accuracy of accumulated variables defined on
curved boundaries has been significantly improved.

ACOUSTICS MODULE | 55
 COUNT REFLECTIONS CHECK BOX NO LONGER CAUSES AN ERROR
In the Ray Acoustics interface, an error message was given if the intensity formulation
for graded media was selected in a 3D model and the Count reflections check box was
selected. In version 6.0, these features can be used together without causing an error
message to appear.

ERROR MESSAGES FROM INTERFACES NOT SOLVED FOR

In version 5.6, an error message was given when trying to run any study if an instance
of the Ray Acoustics interface had no ray release features, or if the ray release features
had no selection. In version 6.0, such error messages will be ignored when running a
study in which the Ray Acoustics interface is not solved for.

Backward Compatibility with Version 5.5

The handling of tables with data entered in frequency bands in the Acoustic Diffusion
Equation interface has been improved. Data can now be stored for all three input types.
To retrieve the old behavior for user-defined bands, change the parameter names
N_band, alphaN, JiN, QN, alphafN, TLN, and qpn to N_band_ud, alphaN_ud,
JinN_ud, QN_ud, alphafN_ud, TLN_ud, and qpn_ud where needed.

For example:

model.component("comp1").physics("ade").feature("ps1").set("N_ban
d_ud", new int[][]{{1},{2},{3}});

model.component("comp1").physics("ade").feature("ps1").set("qpN_u
d", new int[][]{{1},{0},{0}})

Backward Compatibility with Version 5.4

For the Pressure Acoustics interfaces, the following features have been removed from
the Fallback feature submenu to the Pair features: Matched Boundary, Exterior Field
Calculation, Port, Plane Wave Radiation, Circular Wave Radiation, and Spherical Wave
Radiation.

56 | CHAPTER 1: RELEASE NOTES

B a t t e r y De s i g n Mod u l e
New Functionality in Version 6.0

GENERAL NEW FUNCTIONALITY IN THE ELECTROCHEMISTRY

INTERFACES
• The Highly Conductive Porous Electrode domain node is no longer available when
using a Primary Current Distribution type in the Primary Current Distribution and
Secondary Current Distribution interfaces.
• The Thermodynamic Equilibrium electrode kinetics option is no longer available in
the Highly Conductive Porous Electrode domain node in the Primary Current
Distribution, Secondary Current Distribution, and Tertiary Current Distribution
interfaces as well as for the Lithium-Ion Battery and Battery with Binary Electrolyte
interfaces.
• A new Adsorbing-Desorbing Species section has been added to the Electrode Surface,
Porous Electrode, and related nodes in the electrochemistry interfaces. The section
can be used to model adsorption-desorption and the resulting surface coverage.
• The implementation of the Concentration-Dependent Kinetics kinetics type in the
Electrode Reaction and Porous Electrode Reaction (and related) nodes has changed in
order to improve accuracy and numerical convergence for well-posed problems.
Prior to version 6.0, the max() operator was used on the CO and CR parameter
expressions before applying them internally in the definition expression for the iloc
variable. The usage of the max operator has been removed in version 6.0. To get old
models to behave exactly as before, rewrite the CO and CR parameter expressions as
max(CO,eps^2) and max(CR,eps^2), respectively.
• New Adsorption-Desorption Voltammetry tutorial added to the Electrochemistry
Module Application Library.

NEW BATTERY-SPECIFIC FUNCTIONALITY

• Battery Material Library: NMC 622 was added as an electrode material. The solid
diffusion coefficient value and reference information was updated for the electrode
material NMC 111. The volumetric intercalation strain, Young’s modulus, and
Poisson’s ratio data have been added for a number of electrode materials. Lithium
metal was added.

BATTERY DESIGN MODULE | 57

 • A Porous Conductive Binder domain node has been added to the Lithium-Ion Battery
and Battery with Binary Electrolyte interfaces.
• A new Intercalation Strain subnode to Linear Elastic Material in Solid Mechanics
interface is now available with a Battery Design Module license.
• A new rate constant option for specifying the exchange current density in the
Lithium Insertion kinetics option has been added to the Lithium-Ion Battery and
Single Particle Battery interfaces.
• A new postprocessing variable, Cell State-of-Health, has been added to the Capacity
Loss feature of the Lumped Battery interface.
• A new Lithium-Ion Battery, Deformed Geometry model wizard entry has been added.
• The Nondeforming Boundary and Deforming Electrode Surface multiphysics coupling
features are now available with a Battery Design Module license.

GENERAL NEW FUNCTIONALITY IN THE CHEMICAL TRANSPORT

INTERFACES
• The Porous Material node, used for defining and sharing properties for modeling
porous media processes, has been significantly improved.
- The Phase-Specific Properties section lists all properties requested by physics
features with support for a porous material.
- For each phase-specific property, the phase is indicated in the right-most column.
- When requested, multiple entries of the same quantity, one for each phase, can
be listed.
- Tables listing properties can be sorted using any of the columns.
- By right-clicking any property, it is now possible to select Go to Requesting Node
to navigate to where it is used. In the same manner, it is also possible to select Go
to Source to inspect how the property is defined.
- The Homogenized Properties section lists all properties that are not phase specific,
like the permeability. It also lists properties requested by interfaces without
support for a porous material. The homogenized properties can be entered
directly by selecting Locally defined in the Material list.
- When defining properties in a Fluid or Solid subnode to Porous Material, Locally
defined can be selected, in the Material list, to enter the properties directly.
• A new Reacting Flow multiphysics coupling feature has been introduced. In addition
to isothermal flow, it also supports nonisothermal flow, provided that both a
Chemistry interface and a Heat Transfer interface are included in the coupling.

58 | CHAPTER 1: RELEASE NOTES

 Opening a model created in a previous version, the old coupling feature is migrated
into a new one.
• Postprocessing variables for the resulting flow rates (mass or molar) on boundaries
are now available in all mass transport interfaces and are listed under the respective
boundary conditions used. They can, for example, be used to evaluate the mass
conservation.
• The out-of-plane geometry extension can now be specified in all mass transport
interfaces. In 2D, the out-of-plane thickness is available, while in 1D, it is the cross-
sectional area. The out-of-plane dimension is accounted for in the governing
equations and in postprocessing variables for the flow rate.
• A new Out-of-Plane Flux feature is available in the Transport of Diluted Species and
Transport of Concentrated Species interfaces. It can be used to model flux out of
slender geometries in 2D and 1D.
• The Partition Condition feature now includes the possibility to specify a user-defined
relation, possibly nonlinear, between the concentration on the two sides.

• A new Porous Medium feature is available in the Transport of Concentrated Species

interface. The new feature replaces the Porous Media Transport Properties feature.
When opening a model created in an older version, the previous features are kept in
the model.
• A new Transport of Concentrated Species in Porous Media interface is available in the
Chemical Species Transport branch when adding a physics interface.

REACTING FLOW, DILUTED SPECIES NOW SUPPORTS TURBULENT FLOW

• The Reacting Flow, Diluted Species coupling feature can now be used to model
transport of dilute solutes in a fluid subjected to turbulent flow. The coupling
feature applies turbulence modeling for the species transport. Turbulent mixing is
accounted for, and mass transfer close to walls is modeled using wall functions.
• When licensed to use models for turbulent flow, multiphysics interfaces are available
in the Turbulent Flow, Diluted Species group under Reacting Flow in the Chemical
Species Transport branch.
• The Turbulent Mixing node is now obsolete. It was previously available under the
Transport Properties feature in the Transport of Diluted Species and Transport of
Concentrated Species interfaces.

BATTERY DESIGN MODULE | 59

 CHEMISTRY INTERFACE IMPROVEMENTS
• It is no longer needed to explicitly type in variable names to specify the mixture
composition in the Species Matching section. Instead, use the new Species solved for
list to first select a mass transport interface solving for concentrations. Then, for
each of the bulk species in the Chemistry interface, select which of the
concentrations to use.
• Using the Chemistry interface, it is no longer possible to define a species as solvent
when modeling a concentrated mixture (when Type is set to Concentrated Species in
the Mixture Properties section).

THERMODYNAMICS
• The Thermodynamics wizards can now be started from the Physics toolbar.
• The filter search used when including species in a thermodynamic system has been
improved. Species whose molecular formula includes any combination of the given
atoms are now shown. When searching for ammonia, NH3 or H3N can hence be
used. It is also possible to filter species using aliases defined in the database. For
example, “dihydrogen oxide” can be used to find water.
• The Davidson model (including pressure correction) can now be used for the
dynamic viscosity.
• Second derivatives, with respect to temperature and pressure, are now available for
the density and the specific volume.

New and Updated Applications and Models in Version 6.0

• Electrochemical Capacitor with Porous Electrodes
• Electrochemical Capacitor with Side Reactions
• Heterogeneous Electrode Geometry Generation
• Heterogeneous NMC Electrode
• Homogenizing a Heterogeneous Electrode Model
• Lithium Plating with Deformation
• Updated li_battery_1d and li_battery_single_particle tutorials to use the new rate
constant option for specifying the exchange current density in lithium insertion
kinetics. Also, a few parameters for these two models were updated for consistency
with the reference paper.

60 | CHAPTER 1: RELEASE NOTES

C AD I m po r t M od u l e, D esi gn Mod u l e,
and LiveLink™ Products for CAD
New Functionality in Version 6.0

CAD IMPORT MODULE GEOMETRY KERNEL UPGRADE

The CAD Import Module, the Design Module, and the LiveLink™ products for CAD
utilize the Parasolid® geometry kernel from Siemens PLM for solid modeling
operations, geometry repair, and defeaturing. (Without these products, a COMSOL-
native geometry modeling kernel is used.) The CAD Import Module released with
COMSOL version 6.0 includes an upgraded version of the Parasolid kernel. As a result,
a number of stability issues have been fixed, which makes geometry modeling
operations and the import of CAD models more robust.

• The CAD file import and export functionality included with these products has
been extended to support new versions for some of the supported file formats (for
details, see under Read from File, CAD on
www.comsol.com/products/specifications/cad).
• The Detect Interferences tool has been improved to better present and visualize the
detected touches, intersections, gaps, and containments.
• The new Projection feature can be used to project 3D objects and entities to work
planes and 2D geometries.

New Functionality in the Design Module in Version 6.0

• All buttons for constraints and dimensions on the Sketch toolbar are now sticky, so
that after adding a constraint or dimension, it is possible to continue with adding
constraints or dimensions of the same kind without the need to select the button
again. Previously only the smart Constraint and Dimension buttons were sticky.
• When a constraint or dimension button is selected from the Sketch toolbar, the
mouse cursor is decorated with the icon for the constraint or dimension to indicate
which button is selected.

CAD IMPORT MODULE, DESIGN MODULE, AND LIVELINK™ PRODUCTS FOR CAD | 61
 New Functionality in LiveLink™ for AutoCAD® in Version 6.0
• The LiveLink™ interface now supports AutoCAD® 2022.
• The new offline synchronization functionality enables synchronization when
COMSOL Multiphysics and the CAD software cannot be run side by side on the
same computer. A synchronization file, which can contain one or several versions of
the geometry, is prepared in advance using the LiveLink add-in inside the CAD
software. The synchronization file is then loaded by the LiveLink feature in a
COMSOL model. Synchronization files can also be saved in response to a request
saved to file from the LiveLink feature. Offline synchronization supports parametric
sweeps and can be used on all supported operating systems. In addition to being
installed by the COMSOL installer, the LiveLink add-in is also installed by the
COMSOL Client installer, available for download from https://www.comsol.com.
Such an installation enables the functionality for preparing designs for
synchronization, saving files for offline synchronization, and running simulation
apps that synchronize geometry.

New Functionality in LiveLink™ for Inventor® in Version 6.0

• The LiveLink™ interface now supports Inventor® 2022.
• The new offline synchronization functionality enables synchronization when
COMSOL Multiphysics and the CAD software cannot be run side by side on the
same computer. A synchronization file, which can contain one or several versions of
the geometry, is prepared in advance using the LiveLink add-in inside the CAD
software. The synchronization file is then loaded by the LiveLink feature in a
COMSOL model. Synchronization files can also be saved in response to a request
saved to file from the LiveLink feature. Offline synchronization supports parametric
sweeps and can be used on all supported operating systems. In addition to being
installed by the COMSOL installer, the LiveLink add-in is also installed by the
COMSOL Client installer, available for download from https://www.comsol.com.
Such an installation enables the functionality for preparing designs for
synchronization, saving files for offline synchronization, and running simulation
apps that synchronize geometry.
• The new tutorial Stress Analysis of a Pipe Fitting has been added to demonstrate
how to set up a simulation on a 2D cross section of a 3D geometry that has been
synchronized using the LiveLink™ interface.

62 | CHAPTER 1: RELEASE NOTES

New Functionality in LiveLink™ for PTC® Creo® Parametric™ in
Version 6.0
• Support for PTC® Creo® Parametric™ 4.0, 5.0, and 6.0 by the LiveLink™ interface
will be removed in the next major release of COMSOL Multiphysics®. File import
of PTC® Creo® Parametric™ 4.0, 5.0, and 6.0 files will continue to be supported.
• The LiveLink™ interface now supports PTC® Creo® Parametric™ 8.0.
• The new offline synchronization functionality enables synchronization when
COMSOL Multiphysics and the CAD software cannot be run side by side on the
same computer. A synchronization file, which can contain one or several versions of
the geometry, is prepared in advance using the LiveLink add-in inside the CAD
software. The synchronization file is then loaded by the LiveLink feature in a
COMSOL model. Synchronization files can also be saved in response to a request
saved to file from the LiveLink feature. Offline synchronization supports parametric
sweeps and can be used on all supported operating systems. In addition to being
installed by the COMSOL installer, the LiveLink add-in is also installed by the
COMSOL Client installer, available for download from https://www.comsol.com.
Such an installation enables the functionality for preparing designs for
synchronization, saving files for offline synchronization, and running simulation
apps that synchronize geometry.
• The new tutorial Stress Analysis of a Pipe Fitting has been added to demonstrate
how to set up a simulation on a 2D cross section of a 3D geometry that has been
synchronized using the LiveLink™ interface.

New Functionality in LiveLink™ for Revit® in Version 6.0

The LiveLink™ interface now supports Revit® 2022.

New Functionality in LiveLink™ for Solid Edge® in Version 6.0

• The LiveLink™ interface now supports Solid Edge® 2022.
• The new offline synchronization functionality enables synchronization when
COMSOL Multiphysics and the CAD software cannot be run side by side on the
same computer. A synchronization file, which can contain one or several versions of
the geometry, is prepared in advance using the LiveLink add-in inside the CAD
software. The synchronization file is then loaded by the LiveLink feature in a
COMSOL model. Synchronization files can also be saved in response to a request
saved to file from the LiveLink feature. Offline synchronization supports parametric

CAD IMPORT MODULE, DESIGN MODULE, AND LIVELINK™ PRODUCTS FOR CAD | 63
 sweeps and can be used on all supported operating systems. In addition to being
installed by the COMSOL installer, the LiveLink add-in is also installed by the
COMSOL Client installer, available for download from https://www.comsol.com.
Such an installation enables the functionality for preparing designs for
synchronization, saving files for offline synchronization, and running simulation
apps that synchronize geometry.
• The new tutorial Stress Analysis of a Pipe Fitting has been added to demonstrate
how to set up a simulation on a 2D cross section of a 3D geometry that has been
synchronized using the LiveLink™ interface.

New Functionality in LiveLink™ for SOLIDWORKS® in Version 6.0

• The LiveLink™ interface now supports SOLIDWORKS® 2022.
• The new offline synchronization functionality enables synchronization when
COMSOL Multiphysics and the CAD software cannot be run side by side on the
same computer. A synchronization file, which can contain one or several versions of
the geometry, is prepared in advance using the LiveLink add-in inside the CAD
software. The synchronization file is then loaded by the LiveLink feature in a
COMSOL model. Synchronization files can also be saved in response to a request
saved to file from the LiveLink feature. Offline synchronization supports parametric
sweeps and can be used on all supported operating systems. In addition to being
installed by the COMSOL installer, the LiveLink add-in is also installed by the
COMSOL Client installer, available for download from https://www.comsol.com.
Such an installation enables the functionality for preparing designs for
synchronization, saving files for offline synchronization, and running simulation
apps that synchronize geometry.

64 | CHAPTER 1: RELEASE NOTES

CFD Module
New Functionality in Version 6.0
• The Rotating Machinery, High Mach Number Flow interfaces are now available for
laminar flow and for turbulent flow using either the k-epsilon or the Spalart–
Allmaras turbulence models.
• Boundary-level interfaces for Phase Field Thin-Film Flow, Shell and Phase-Field
Thin-Film Flow, Edge have been added under the Moving Interface section in the
Mathematics branch, and predefined Two-Phase Thin-Film Flow, Phase Field
multiphysics couplings to the corresponding Thin-Film Flow interfaces have been
added under the Multiphase Flow section in the Fluid Flow branch.
• A Fluid-Solid Interaction, Fixed Geometry interface for Two-Phase Flow, Phase
Field has been added under the Fluid-Structure Interaction section in the Fluid
Flow branch.
• The Brinkman Equations, Free and Porous Media Flow and Darcy’s Law interfaces
have a new Porous Medium feature that replaces the old Fluid and Matrix Properties
feature. The new feature contains two subfeatures, Porous Matrix and Fluid, in which
the characteristic properties can be specified directly or by linking to a Material node.
• A predefined multiphysics coupling between the Brinkman Equations and Heat
Transfer in Porous Media interfaces has been added under the Nonisothermal Flow
section in the Fluid Flow branch.
• The Brinkman Equations and Free and Porous Media Flow interfaces have a new
option, Porous treatment of no slip condition, with two options, Standard no slip
formulation and Porous slip, where the latter one is suitable for models with under-
resolved boundary layers.
• Automatic wall treatment has been added for the LES interfaces. For nonisothermal
LES, thermal wall functions are automatically added when Wall treatment is set to
Automatic in the LES interface.
• Source terms for mass and momentum have been added in the Shallow Water
Equations interfaces.
• The Phase and Porous Media Transport Properties node in the Phase Transport in
Porous Media interface has a new Capillary diffusion option in the Capillary pressure
setting.

CFD MODULE | 65
 • The Phase Transport Mixture Model interfaces have a new Low dispersed phase
concentration option in the Mixture Model multiphysics coupling node.
• The Flow condition setting in the Grille feature has a new Loss coefficient option for
quadratic losses.
• The Interior Fan feature has a new Prescribed flux option in the Flow condition setting.
• A warning for missing Wall Distance Initialization study steps will be displayed when
this step is missing for RANS turbulence models requiring the wall distance.

New Model in Version 6.0

DAM BREAKING ON A COLUMN, LEVEL SET

This is a transient model for the impact of a water wave on a column. A body of water
is initially contained behind a gate. At the start of the simulation, the gate is suddenly
released and the body of water forms a wave moving toward the structure. After
impacting on the structure, the water continues until it is reflected at the wall of the
tank and impinges a second time on the column.

Backward Compatibility with Version 5.6 and Earlier

BRINKMAN EQUATIONS, FREE AND POROUS MEDIA FLOW AND DARCY’S

LAW
When opening a model created in version 5.6 or earlier, the old Fluid and Matrix
Properties feature is not migrated, but a warning message is prompted. This feature
should be manually replaced by the new Porous Medium feature (which adds Porous
Matrix and Fluid subfeatures) to allow an updated definition of the material properties.
Support for obsolete features is not ensured in future versions; thus it is strongly
recommended to transfer the old feature to the new feature.

Backward Compatibility with Version 5.5 and Earlier

TURBULENT INTENSITY AND TURBULENCE LENGTH SCALE

The new default value for Turbulent intensity is the same as the old default value,
whereas the new Geometry based default value in general will differ from the old one.
Old Java files with default values for Turbulence length scale should add the line,

model.component("comp1").physics("spf").feature("inl1").
set("LT_list", "user_defined");

66 | CHAPTER 1: RELEASE NOTES

in the Model Java-file to retrieve the old length scale.

EULER-EULER MODEL
Results from models using the Mass Transfer feature in the Euler-Euler Model
interfaces may differ in version 5.6 due to the improved momentum conservation.

PHASE TRANSPORT, MIXTURE MODEL

Because the Phase Transport, Mixture Model interfaces have been revised using a
mass-averaged formulation to support compressible mixture flows, models built in
version 5.5 can return slightly different results when run in version 5.6.

TWO-PHASE FLOW
In the Level Set and Phase Field interfaces, the Initial Interface feature is now obsolete.
It does not need to be replaced because the phase interfaces are automatically detected
from the selection of the Initial Values features, and the initial distribution of the level
set or phase field function is solved for in the Phase Initialization study step.

Backward Compatibility with Version 5.4 and Earlier

Models created using the Brinkman Equations and Free and Porous Media Flow
interfaces need to be modified if they apply Forchheimer drag. The previous subfeature
has been replaced by a new setting in the Porous Matrix Properties section.

Old Java files need to replace the line

model.component("comp1").physics("fp").feature("fmp1").
create("fd1", "Forchheimer", 2);

with

model.component("comp1").physics("fp").feature("fmp1").
set("ktype", "nonDarcianModel");

Forchheimer is the default. Also replace

model.component("comp1").physics("fp").feature("fmp1").
feature("fd1").set("betaF", "value");

with

model.component("comp1").physics("fp").feature("fmp1").
set("cf", "value");

CFD MODULE | 67
 Chemical Reaction Engineering
Module
New Functionality in Version 6.0

GENERAL NEW FUNCTIONALITY

• The Porous Material node, used for defining and sharing properties for modeling
porous media processes, has been significantly improved.
- The Phase-Specific Properties section lists all properties requested by physics
features with support for a porous material.
- For each phase-specific property, the phase is indicated in the right-most column.
- When requested, multiple entries of the same quantity, one for each phase, can
be listed.
- Tables listing properties can be sorted using any of the columns.
- By right-clicking any property, it is now possible to select Go to Requesting Mode
to navigate to where it is used. In the same manner, it is also possible to select Go
to Source to inspect how the property is defined.
- The Homogenized Properties section lists all properties that are not phase specific,
like the permeability. It also lists properties requested by interfaces without
support for a porous material. The homogenized properties can be entered
directly by selecting Locally defined in the Material list.
- When defining properties in a Fluid, Solid, or Pellet subnode to Porous Material,
Locally defined can be selected, in the Material list, to enter the properties directly.
• A new subnode Pellet is available for Porous Material. The new subnode is used to
provide properties for solving mass and energy transport inside porous pellets in a
packed bed.
• A new Reacting Flow multiphysics coupling feature has been introduced. In addition
to isothermal flow, it now also supports nonisothermal flow, provided that both a
Chemistry interface and a Heat Transfer interface are included in the coupling.
Opening a model created in a previous version, the old coupling feature is migrated
into a new one.
• Postprocessing variables for the resulting flow rates (mass or molar) on boundaries
are now available. These are available in all mass transport interfaces, and are listed

68 | CHAPTER 1: RELEASE NOTES

 under the respective boundary conditions used. They can, for example, be used to
evaluate the mass conservation.
• The out-of-plane geometry extension can now be specified in all mass transport
interfaces. In 2D, the out-of-plane thickness is available, while in 1D, it is the cross-
sectional area. The out-of-plane dimension is accounted for in the governing
equations and in postprocessing variables for the flow rate.
• A new Out-of-Plane Flux feature is available in the Transport of Diluted Species and
Transport of Concentrated Species interfaces. It can be used to model flux out of
slender geometries in 2D and 1D.
• The Partition Condition feature now includes the possibility to specify a user-defined
relation, possibly nonlinear, between the concentration on the two sides.

REACTING FLOW, DILUTED SPECIES NOW SUPPORTS TURBULENT FLOW

• The Reacting Flow, Diluted Species coupling feature can now be used to model
transport of dilute solutes in a fluid subjected to turbulent flow. The coupling
feature applies turbulence modeling for the species transport. Turbulent mixing is
accounted for, and mass transfer close to walls is modeled using wall functions.
• When licensed to use models for turbulent flow, multiphysics interfaces are available
in the Turbulent Flow, Diluted Species group under Reacting Flow in the Chemical
Species Transport branch.
• The Turbulent Mixing node is now obsolete. It was previously available under the
Transport Properties feature in the Transport of Diluted Species and Transport of
Concentrated Species interfaces.

CHEMISTRY INTERFACE IMPROVEMENTS

• It is no longer needed to explicitly type in variable names to specify the mixture
composition in the Species Matching section. Instead, use the new Species solved for
list to first select a mass transport interface solving for concentrations. Then, for
each of the bulk species in the Chemistry interface, select which of the
concentrations to use.
• Using the Chemistry interface, it is no longer possible to define a species as solvent
when modeling a concentrated mixture (when Type is set to Concentrated Species in
the Mixture Properties section).
• To create variables available in pellets of a packed bed, select the Define variables for
porous pellets check box in the Pellet Chemistry section (seen in the Settings window
when selecting the interface). The check box was formerly called Define variables in
extra dimension and was available in a section named Extra Dimension.

CHEMICAL REACTION ENGINEERING MODULE | 69

 NEW AND IMPROVED POROUS MEDIA INTERFACES AND FEATURES
• New Packed Bed features are available in the Transport of Diluted Species and
Transport of Concentrated Species interfaces. The features can be used to study
mass transport in beds filled with reactive pellets and solves for interparticle
transport (surrounding the pellets) as well as intraparticle transport (inside the
pellets). The properties of the bed-filling porous pellets are defined using a Porous
Material, including one or several Pellet subnodes (one per pellet type). The Packed
Bed features replace the previous Reacting Pellet Bed features. Opening a model
created in an older version, the previous features are kept in the model.
• The Heat Transfer in Packed Beds interface can be used to solve for the temperature
in a bed filled with porous pellets. Similar to the mass transport feature, both
interparticle energy transport (surrounding the pellets) as well as intraparticle
energy transport (inside the pellets) is solved for.
• New Porous Catalyst features are available in the Transport of Diluted Species and
Transport of Concentrated Species interfaces. The features can be used to study
adsorption and desorption as well as heterogeneous reactions in porous media.
• A new Porous Medium feature is available in the Transport of Concentrated Species
interface. The new feature replaces the Porous Media Transport Properties feature.
Opening a model created in an older version, the previous features are kept in the
model.
• A new Transport of Concentrated Species in Porous Media interface is available in the
Chemical Species Transport branch when adding a physics interface.
• Two new multiphysics interfaces, Transport of Diluted Species in Porous Catalysts and
Transport of Concentrated Species in Porous Catalysts, are available in the Reacting
Flow in Porous Media group of the Chemical Species Transport branch when adding
physics interfaces.
• Two new multiphysics interfaces, Transport of Diluted Species in Packed Beds and
Transport of Concentrated Species in Packed Beds, are available in the Reacting Flow in
Porous Media group of the Chemical Species Transport branch when adding physics
interfaces.
• Physics interfaces using any of the Porous Medium, Porous Catalyst, or Packed Bed
features can be created and automatically set up using the Generate Space-Dependent
Model node in the Reaction Engineering interface.

70 | CHAPTER 1: RELEASE NOTES

• The Porous Catalyst and Packed Bed features include the possibility to solve for
surface species adsorbed to the solid substrate in a porous medium domain.
- Selecting an interface containing any of these features in a Generate Space-
Dependent Model node, the surface reaction kinetics defined in the Reaction
Engineering interface will be transferred to the space-dependent model.
- When none of these domain features are created by the Generate Space-Dependent
Model feature, the surface reaction kinetics are instead implemented in a Surface
Reactions interface and solved for on boundaries adjacent to the model geometry.
- The Surface reactions on setting is no longer present in the Generate Space-
Dependent Model node.

THERMODYNAMICS
• The Thermodynamics wizards can now be started from the Physics toolbar.
• The filter search used when including species in a thermodynamic system has been
improved. Species whose molecular formula includes any combination of the given
atoms are now shown. When searching for ammonia, NH3 or H3N can hence be
used. It is also possible to filter species using aliases defined in the database. For
example, “dihydrogen oxide” can be used to find water.
• The Davidson model (including pressure correction) can now be used for the
dynamic viscosity.
• Second derivatives, with respect to temperature and pressure, are now available for
the density and the specific volume.

New Models in Version 6.0

CUBIC AUTOCATALYSIS: EXPLORING THE GRAY–SCOTT MODEL

A deceptively simple system of two species and two reactions, describing the
autocatalytic conversion of a substance, is shown to display a surprisingly exotic
behavior. Starting from a 0D CSTR description, a 2D reaction-diffusion problem is
formulated showing rich patterns stemming from the multistability of the underlying
reaction network.

GOLD RECYCLING THROUGH OXIDATIVE DISSOLUTION

This model studies the oxidative dissolution of this noble metal in an air-saturated
cyanide solution. The system encompasses three phases: a gaseous phase (air), an
aqueous phase, and a solid gold phase. The system is assumed to be homogeneous on

CHEMICAL REACTION ENGINEERING MODULE | 71

 a macroscopic scale — for example, fine particulate matter dispersed in water, which is
continuously agitated by a stream of air bubbles.

MASS TRANSFER FROM A THIN DOMAIN

This model demonstrates how mass transfer out from a thin domain can be
approximated with a 2D model using the Out-of-Plane Flux feature. To check the
validity of the 2D approximation, the equivalent 3D model is also solved. Additionally,
this model illustrates how the out-of-plane thickness affects the relative error of the 2D
approximation and by how much the computational time can be decreased.

OSCILLATIONS IN METABOLIC REACTION NETWORKS

Under certain conditions, reaction rates involved in the glycolysis may exhibit a limit
cycle, where the concentrations vary periodically in a manner not usually seen in
chemical kinetics. Here, a lumped kinetic model from the literature is used to study
how the rate of consumption of glucose varies along with the coenzyme NADH and
the energy carrier ATP. The time evolution of the swings in concentration is studied in
a homogeneous system, and transport processes across the cell membrane is treated in
an implicit manner.

TURBULENT MIXING IN A STIRRED TANK

This example shows how the mixing capability of a stirred vessel can be evaluated. To
efficiently reach the steady operational conditions, the turbulent flow is solved for
using a frozen rotor analysis followed by a time-dependent simulation. When a
quasisteady flow field has developed, the turbulent mixing of a trace species is
simulated, and the time until it is fully mixed is evaluated.

Updated Applications in Version 6.0

STEAM REFORMER
The model has been extended to include a water-gas shift reaction in order to study
carbon monoxide formation. Three different setups involving cocurrent and
countercurrent flow of the heating media are modeled, and the impact on the carbon
monoxide production is evaluated.

72 | CHAPTER 1: RELEASE NOTES

Backward Compatibility with Version 5.4 and Earlier

SPECIFYING VOLUMETRIC FLOW RATE OF A PLUG FLOW REACTOR IN

REACTION ENGINEERING
When using the Reaction Engineering interface to model gas flow in a plug flow
reactor, you must now specify either the pressure or the volumetric flow rate, not both.
Running Java® files from previous versions may fail if the volumetric flow rate is set
while also prescribing the pressure. In this case, specify to instead compute the pressure
from the ideal gas law in the following manner:
model.component("comp1").physics("re").prop("mixture").set("psource",
"idealGas");

This selects Ideal gas law in the Reactor pressure list in the Mixture Properties section.
When this is done, the User defined option is available in the Volumetric rate list in the
Mass Balance section.

WALL FUNCTION VARIABLES OMITTED IN LAMINAR REACTING FLOW

Dependent variables for the mass fraction wall functions are no longer added when
using the Reacting Flow multiphysics interface for laminar flow. Due to the now-
omitted variables, running Java® files from previous versions may fail during the solver
sequence generation.

When specifying the content of a segregated group in the manner of

model.sol("sol1").feature("s1").feature("se1").feature("ss1")
.set("segvar", new String[]
{"comp1_w1", "comp1_w2", "comp1_tcs_wWall_w1","comp1_tcs_wWall_w2"});

remove entries for the wall functions variables of the kind comp1_tcs_wWall_name to
result in
model.sol("sol1").feature("s1").feature("se1").feature("ss1")
.set("segvar", new String[]{"comp1_w1", "comp1_w2"});

CHEMICAL REACTION ENGINEERING MODULE | 73

 Composite Materials Module
New Functionality in Version 6.0

IMPROVEMENTS TO SAFETY FACTOR COMPUTATIONS

Some of the models for safety factor computation in the Safety node have been
updated.

• In the Tsai–Hill criterion, there is now a separate input for the compressive
strengths.
• The Modified Tsai-Hill criterion, intended for 2D analysis, has been removed from the
list of failure models. To obtain the same effect, use the Tsai-Hill criterion, and select
the new Use plane stress formulation check box. When you open an old model that
contained the Modified Tsai-Hill criterion, it will automatically be updated to the new
settings.
• 2D formulations of the Tsai–Wu and Hashin criteria have been added. To use them,
choose Tsai-Wu orthotropic or Hashin as the failure criterion and then select the new
Use plane stress formulation check box.
• In the Layered Shell interface, several new criteria were added or extended: The
Norris and Azzi–Tsai–Hill criteria are new. The Tsai–Wu and Tsai–Hill criteria have
been extended with a 2D formulation.

NEW TYPE OF FOLD-LINE CONSTRAINTS IN LAYERED SHELL

A new type of constraint, called a simplified fold-line constraint, has been
implemented in the Layered Shell interface. It is computationally more efficient than
the standard full fold-line constraint but is less accurate for thick shells. The new
constraint type is particularly advantageous for large geometrically linear problems
because it has a special linear formulation in that case.

You select the type of fold-line constraint in the Fold-Line Settings section of the
interface settings.

NEW COUPLING BETWEEN SHELL AND LAYERED SHELL

The new Layered Shell-Shell Connection multiphysics coupling makes it possible to mix
the Shell and Layered Shell interfaces by defining different stack members within a
single layup using either of the interfaces. This modeling approach is sometimes called
the multiple model method.

74 | CHAPTER 1: RELEASE NOTES

STACK MEMBER SELECTION
In the Layered Linear Elastic Material in the Shell interface, it is now possible to select
one single stack member from a Layered Material Stack by clearing the Use all layers
check box. The main use for this facility is when mixing different shell formulations on
the same boundary in the multiple model method.

NEW OPERATORS FOR LAYERED MATERIALS

For evaluation at a certain location in the extra dimension (the thickness direction),
two new physics scoped operators have been added: <phys>.atxd0 (for evaluation in
points) and <phys>.atxd1 (for evaluation on edges). These operators take two
arguments, the through-thickness location and the expression to be evaluated — for
example, lshell.atxd1(2*th,mean(lshell.mises)).

The new operators are available in the Layered Shell interface as well as for the layered
materials in the Shell and Membrane interfaces.

New Models in Version 6.0

ANALYSIS OF A COMPOSITE BLADE USING MULTIPLE MODEL METHOD

This example demonstrates the modeling of a sandwiched composite blade made up
of carbon-epoxy, glass-vinylester, and PVC foam materials. To model the sandwich
composite structure, three different methods are used: layerwise theory, equivalent
single layer (ESL) theory, and a multiple model method. Modal as well as frequency
response analyses are performed for the specified loading, and the multiple model
method is compared with the traditional layerwise and ESL theory in terms of
performance and accuracy in the results, including the through-thickness stress
distribution.

COMPOSITE MATERIALS MODULE | 75

 Corrosion Module
New Functionality in Version 6.0

GENERAL NEW FUNCTIONALITY IN THE ELECTROCHEMISTRY

INTERFACES
• The Highly Conductive Porous Electrode domain node is no longer available when
using a Primary Current Distribution type in the Primary Current Distribution and
Secondary Current Distribution interfaces.
• The Thermodynamic Equilibrium electrode kinetics option is no longer available in
the Highly Conductive Porous Electrode domain node in the Primary Current
Distribution, Secondary Current Distribution, and Tertiary Current Distribution
interfaces.
• A new Adsorbing-Desorbing Species section has been added to the Electrode Surface,
Porous Electrode, and related nodes in the electrochemistry interfaces. The section
can be used to model adsorption-desorption and the resulting surface coverage.
• The implementation of the Concentration Dependent Kinetics kinetics type in the
Electrode Reaction, Porous Electrode Reaction, and related nodes has changed in order
to improve accuracy and numerical convergence for well-posed problems. Prior to
version 6.0, the max() operator was used on the CO and CR parameter expressions
before applying them internally in the definition expression for the iloc variable.
The usage of the max operator has been removed in version 6.0. To get old models
to behave exactly as before, rewrite the CO and CR parameter expressions as
max(CO,eps^2) and max(CR,eps^2), respectively.

NEW CORROSION MODULE-SPECIFIC FUNCTIONALITY

• The Corrosion, Deformed Geometry model wizard entries now make use of the new
Deformed Geometry formulation. The new multiphysics interfaces also use new
Nondeforming Boundary and Deforming Electrode Surface multiphysics features.
• A Seawater material entry was added to the Corrosion material library. The new
material contains temperature- and salinity-dependent electrolyte conductivity data.
• The Phase Initialization study node for Level Set and Phase Field interfaces is now
available.
• The Corrosion material library polarization data has been cleansed of
nonmonotonic polarization data, when deemed caused by noise.

76 | CHAPTER 1: RELEASE NOTES

• A new Cathodic Protection model wizard entry has been added.
• A number of new Cathodic Protection feature nodes have been added to the
Primary and Secondary Current Distribution interfaces and are active in the new
Cathodic Protection model wizard entry: Connection Surface, Impressed Current
Surface, Passive Metal Surface, Protected Metal Surface, Thin Passive Metal Surface, Thin
Protected Metal Surface, Connection Point (under Edge Electrode), and Impressed
Current Point.
• Stabilization sections are added to the Current Distribution, Boundary Elements
interface for dissolving-depositing species, which are enabled when Deformed
Geometry is active.

GENERAL NEW FUNCTIONALITY IN THE CHEMICAL TRANSPORT

INTERFACES
• The Porous Material node, used for defining and sharing properties for modeling
porous media processes, has been significantly improved.
- The Phase-Specific Properties section lists all properties requested by physics
features with support for a porous material.
- For each phase-specific property, the phase is indicated in the right-most column.
- When requested, multiple entries of the same quantity, one for each phase, can
be listed.
- Tables listing properties can be sorted using any of the columns.
- By right-clicking any property, it is now possible to select Go to Requesting Mode
to navigate to where it is used. In the same manner, it is also possible to select Go
to Source to inspect how the property is defined.
- The Homogenized Properties section lists all properties that are not phase specific,
like the permeability. It also lists properties requested by interfaces without
support for a porous material. The homogenized properties can be entered
directly by selecting Locally defined in the Material list.
- When defining properties in a Fluid or Solid subnode to a Porous Material node,
Locally defined can be selected, in the Material list, to enter the properties directly.
• A new Reacting Flow multiphysics coupling feature has been introduced. In addition
to isothermal flow, it now also supports nonisothermal flow, provided that both a
Chemistry interface and a Heat Transfer interface are included in the coupling.
Opening a model created in a previous version, the old coupling feature is migrated
into a new one.

CORROSION MODULE | 77
 • Postprocessing variables for the resulting flow rates (mass or molar) on boundaries
are now available. These are available in all mass transport interfaces and are listed
under the respective boundary conditions used. These can, for example, be used to
evaluate the mass conservation.
• The out-of-plane geometry extension can now be specified in all mass transport
interfaces. In 2D the out-of-plane thickness is available, while in 1D it is the cross-
sectional area. The out-of-plane dimension is accounted for in the governing
equations and in postprocessing variables for the flow rate.
• A new Out-of-Plane Flux feature is available in the Transport of Diluted Species and
Transport of Concentrated Species interfaces. It can be used to model flux out of
slender geometries in 2D and 1D.
• The Partition Condition feature now includes the possibility to specify a user-defined
relation, possibly nonlinear, between the concentration on the two sides.

• A new Porous Medium feature is available in the Transport of Concentrated Species

interface. This feature replaces the Porous Media Transport Properties feature.
Opening a model created in an older version, the previous features are kept in the
model.
• A new Transport of Concentrated Species in Porous Media interface is available in the
Chemical Species Transport branch when adding a physics interface.

REACTING FLOW, DILUTED SPECIES NOW SUPPORTS TURBULENT FLOW

• The Reacting Flow, Diluted Species coupling feature can now be used to model
transport of dilute solutes in a fluid subjected to turbulent flow. The coupling
feature applies turbulence modeling for the species transport. Turbulent mixing is
accounted for, and mass transfer close to walls is modeled using wall functions.
• When licensed to use models for turbulent flow, multiphysics interfaces are available
in the Turbulent Flow, Diluted Species group under Reacting Flow in the Chemical
Species Transport branch.
• The Turbulent Mixing node is now obsolete. It was previously available under the
Transport Properties feature in the Transport of Diluted Species and Transport of
Concentrated Species interfaces.

CHEMISTRY INTERFACE IMPROVEMENTS

• It is no longer needed to explicitly type in variable names to specify the mixture
composition in the Species Matching section. Instead, use the new Species solved for
list to first select a mass transport interface solving for concentrations. Then, for

78 | CHAPTER 1: RELEASE NOTES

 each of the bulk species in the Chemistry interface, select which of the
concentrations to use.
• Using the Chemistry interface, it is no longer possible to define a species as solvent
when modeling a concentrated mixture (when Type is set to Concentrated Species in
the Mixture Properties section).

New and Updated Models in Version 6.0

• New Oxide Jacking of Reinforced Concrete tutorial
• New Pitting Corrosion tutorial
• The Stress Corrosion model was extended to account for deformations at the
corrosion defect also in the stress calculation
• The oil_platform, ship_hull, and monopile tutorials were updated to use the new
Cathodic Protection interface and related features.
• Updated localized_corrosion_ls and localized_corrosion_pf models using the Time
Dependent with Phase Initialization study node.
• Resistor-Modulated Pipeline Cathodic Protection (changed name from prior
version).
• The following models are updated to use the new Deformed Geometry node:
crevice_corrosion_with_deformation, galvanic_corrosion_with_deformation,
localized_corrosion, and under_deposit_corrosion.
• The oil_platform, anode_film_resistance, and monopile tutorials were updated to
use Seawater material for the electrolyte conductivity.

CORROSION MODULE | 79
 ECAD Import Module
New Functionality in Version 6.0
The ECAD Import Module includes significant improvements for generating the 2D
geometries from the layouts imported from IPC-2581, ODB++, and GDS files. The
geometry for complex layouts is now generated in a more robust way, which eliminates
the need for slower fallback methods and thereby improves the performance for the
import.

80 | CHAPTER 1: RELEASE NOTES

Electrochemistry Module
New Functionality in Version 6.0

GENERAL NEW FUNCTIONALITY IN THE ELECTROCHEMISTRY

INTERFACES
• The Highly Conductive Porous Electrode domain node is no longer available when
using a Primary Current Distribution type in the Primary Current Distribution and
Secondary Current Distribution interfaces.
• The Thermodynamic Equilibrium electrode kinetics option is no longer available in
the Highly Conductive Porous Electrode domain node in the Primary Current
Distribution, Secondary Current Distribution, and Tertiary Current Distribution
interfaces.
• A new Adsorbing-Desorbing Species section has been added to the Electrode Surface,
Porous Electrode, and related nodes in the electrochemistry interfaces. The section
can be used to model adsorption-desorption and the resulting surface coverage.
• The implementation of the Concentration Dependent Kinetics kinetics type in the
Electrode Reaction, Porous Electrode Reaction, and related nodes has changed in order
to improve accuracy and numerical convergence for well-posed problems. Prior to
version 6.0, the max() operator was used on the CO and CR parameter expressions
before applying them internally in the definition expression for the iloc variable.
The usage of the max operator has been removed in version 6.0. To get old models
to behave exactly as before, rewrite the CO and CR parameter expressions as
max(CO, eps^2) and max(CR, eps^2), respectively.

• New Adsorption-Desorption Voltammetry tutorial added to the Electrochemistry

Module Application Library.

ELECTROCHEMISTRY MODULE | 81
 GENERAL NEW FUNCTIONALITY IN THE CHEMICAL TRANSPORT
INTERFACES
• The Porous Material node, used for defining and sharing properties for modeling
porous media processes, has been significantly improved.
- The Phase-Specific Properties section lists all properties requested by physics
features with support for a porous material.
- For each phase-specific property, the phase is indicated in the right-most column.
- When requested, multiple entries of the same quantity, one for each phase, can
be listed.
- Tables listing properties can be sorted using any of the columns.
- By right-clicking any property, it is now possible to select Go to Requesting Mode
to navigate to where it is used. In the same manner, it is also possible to select Go
to Source to inspect how the property is defined.
- The Homogenized Properties section lists all properties that are not phase specific,
like the permeability. It also lists properties requested by interfaces without
support for a porous material. The homogenized properties can be entered
directly by selecting Locally defined in the Material list.
- When defining properties in a Fluid or Solid subnode to a Porous Material node,
Locally defined can be selected, in the Material list, to enter the properties directly.
• A new Reacting Flow multiphysics coupling feature has been introduced. In addition
to isothermal flow, it now also supports nonisothermal flow, provided that both a
Chemistry interface and a Heat Transfer interface are included in the coupling.
Opening a model created in a previous version, the old coupling feature is migrated
into a new one.
• Postprocessing variables for the resulting flow rates (mass or molar) on boundaries
are now available in all mass transport interfaces and are listed under the respective
boundary conditions used. They can, for example, be used to evaluate the mass
conservation.
• The out-of-plane geometry extension can now be specified in all mass transport
interfaces. In 2D the out-of-plane thickness is available, while in 1D it is the cross-
sectional area. The out-of-plane dimension is accounted for in the governing
equations and in postprocessing variables for the flow rate.

82 | CHAPTER 1: RELEASE NOTES

• A new Out-of-Plane Flux feature is available in the Transport of Diluted Species and
the Transport of Concentrated Species interfaces. It can be used to model flux out
of slender geometries in 2D and 1D.
• The Partition Condition feature now includes the possibility to specify a user-defined
relation, possibly nonlinear, between the concentration on the two sides.

• A new Porous Medium feature is available in the Transport of Concentrated Species

interface. The new feature replaces the Porous Media Transport Properties feature.
If you open a model created in an older version, the previous features are kept in the
model.
• A new Transport of Concentrated Species in Porous Media interface is available in the
Chemical Species Transport branch when adding a physics interface.

REACTING FLOW, DILUTED SPECIES NOW SUPPORTS TURBULENT FLOW

• The Reacting Flow, Diluted Species coupling feature can now be used to model
transport of dilute solutes in a fluid subjected to turbulent flow. The coupling
feature applies turbulence modeling for the species transport. Turbulent mixing is
accounted for, and mass transfer close to walls is modeled using wall functions.
• When licensed to use models for turbulent flow, multiphysics interfaces are available
in the Turbulent Flow, Diluted Species group under Reacting Flow in the Chemical
Species Transport branch.
• The Turbulent Mixing node is now obsolete. It was previously available under the
Transport Properties feature in the Transport of Diluted Species and Transport of
Concentrated Species interfaces.

CHEMISTRY INTERFACE IMPROVEMENTS

• It is no longer needed to explicitly type in variable names to specify the mixture
composition in the Species Matching section. Instead, use the new Species solved for
list to first select a mass transport interface solving for concentrations. Then, for
each of the bulk species in the Chemistry interface, select which of the
concentrations to use.
• Using the Chemistry interface, it is no longer possible to define a species as solvent
when modeling a concentrated mixture (when Type is set to Concentrated Species in
the Mixture Properties section).

ELECTROCHEMISTRY MODULE | 83
 New and Updated Models in Version 6.0
A New Adsorption-Desorption Voltammetry tutorial has been added to the
Electrochemistry Module Application Library.

84 | CHAPTER 1: RELEASE NOTES

E le c t r o de po s i t i on Mod u l e
New Functionality in Version 6.0

GENERAL NEW FUNCTIONALITY IN THE ELECTROCHEMISTRY

INTERFACES
• The Highly Conductive Porous Electrode domain node is no longer available when
using a Primary Current Distribution type in the Primary Current Distribution and
Secondary Current Distribution interfaces.
• The Thermodynamic Equilibrium electrode kinetics option is no longer available in
the Highly Conductive Porous Electrode domain node in the Primary Current
Distribution, Secondary Current Distribution, and Tertiary Current Distribution
interfaces.
• A new Adsorbing-Desorbing Species section has been added to the Electrode Surface,
Porous Electrode, and related nodes in the electrochemistry interfaces. The section
can be used to model adsorption-desorption, and the resulting surface coverage.
• The implementation of the Concentration Dependent Kinetics kinetics type in the
Electrode Reaction, Porous Electrode Reaction, and related nodes has changed in order
to improve accuracy and numerical convergence for well-posed problems. Prior to
version 6.0, the max() operator was used on the CO and CR parameter expressions
before applying them internally in the definition expression for the iloc variable.
The usage of the max operator has been removed in version 6.0. To get old models
to behave exactly as before, rewrite the CO and CR parameter expressions as
max(CO,eps^2) and max(CR,eps^2), respectively.

• A new Adsorption-Desorption Voltammetry tutorial has been added to the

Electrochemistry Module Application Library.

NEW ELECTRODEPOSITION MODULE-SPECIFIC FUNCTIONALITY

• The Electrodeposition, Deformed Geometry model wizard entries now make use of the
new Deformed Geometry formulation. The new multiphysics interfaces also use
new Nondeforming Boundary and Deforming Electrode Surface multiphysics features.
• Stabilization sections are added to the Current Distribution, Boundary Elements
interface for dissolving-depositing species, which are enabled when Deformed
Geometry is active.

ELECTRODEPOSITION MODULE | 85
 GENERAL NEW FUNCTIONALITY IN THE CHEMICAL TRANSPORT
INTERFACES
• The Porous Material node, used for defining and sharing properties for modeling
porous media processes, has been significantly improved.
- The Phase-Specific Properties section lists all properties requested by physics
features with support for a porous material.
- For each phase-specific property, the phase is indicated in the right-most column.
- When requested, multiple entries of the same quantity, one for each phase, can
be listed.
- Tables listing properties can be sorted using any of the columns.
- By right-clicking any property, it is now possible to select Go to Requesting Node
to navigate to where it is used. In the same manner, it is also possible to select Go
to Source to inspect how the property is defined.
- The Homogenized Properties section lists all properties that are not phase specific,
like the permeability. It also lists properties requested by interfaces without
support for a porous material. The homogenized properties can be entered
directly by selecting Locally defined in the Material list.
- When defining properties in a Fluid or Solid subnode to a Porous Material node,
Locally defined can be selected, in the Material list, to enter the properties directly.
• A new Reacting Flow multiphysics coupling feature has been introduced. In addition
to isothermal flow, it now also supports nonisothermal flow, provided that both a
Chemistry interface and a Heat Transfer interface are included in the coupling.
Opening a model created in a previous version, the old coupling feature is migrated
into a new one.
• Postprocessing variables for the resulting flow rates (mass or molar) on boundaries
are now available in all mass transport interfaces and are listed under the respective
boundary conditions used. They can, for example, be used to evaluate the mass
conservation.
• The out-of-plane geometry extension can now be specified in all mass transport
interfaces. In 2D the out-of-plane thickness is available, while in 1D it is the cross-
sectional area. The out-of-plane dimension is accounted for in the governing
equations and in postprocessing variables for the flow rate.

86 | CHAPTER 1: RELEASE NOTES

• A new Out-of-Plane Flux feature is available in the Transport of Diluted Species and
Transport of Concentrated Species interfaces. It can be used to model flux out of
slender geometries in 2D and 1D.
• The Partition Condition feature now includes the possibility to specify a user-defined
relation, possibly nonlinear, between the concentration on the two sides.

• A new Porous Medium feature is available in the Transport of Concentrated Species

interface. The new feature replaces the Porous Media Transport Properties feature.
If you open a model created in an older version, the previous features are kept in the
model.
• A new Transport of Concentrated Species in Porous Media interface is available in the
Chemical Species Transport branch when adding a physics interface.

REACTING FLOW, DILUTED SPECIES NOW SUPPORTS TURBULENT FLOW

• The Reacting Flow, Diluted Species coupling feature can now be used to model
transport of dilute solutes in a fluid subjected to turbulent flow. The coupling
feature applies turbulence modeling for the species transport. Turbulent mixing is
accounted for, and mass transfer close to walls is modeled using wall functions.
• When licensed to use models for turbulent flow, multiphysics interfaces are available
in the Turbulent Flow, Diluted Species group under Reacting Flow in the Chemical
Species Transport branch.
• The Turbulent Mixing node is now obsolete. It was previously available under the
Transport Properties feature in the Transport of Diluted Species and Transport of
Concentrated Species interfaces.

CHEMISTRY INTERFACE IMPROVEMENTS

• It is no longer needed to explicitly type in variable names to specify the mixture
composition in the Species Matching section. Instead, use the new Species solved for
list to first select a mass transport interface solving for concentrations. Then, for
each of the bulk species in the Chemistry interface, select which of the
concentrations to use.
• Using the Chemistry interface, it is no longer possible to define a species as solvent
when modeling a concentrated mixture (when Type is set to Concentrated Species in
the Mixture Properties section).

ELECTRODEPOSITION MODULE | 87
 New and Updated Models in Version 6.0
• A new model in the Electrodeposition Application Library: Copper Deposition in a
Through-Hole Via, which demonstrates the effect of additives during
electrodeposition. The new model uses the Adsorbing-Desorbing Species feature.
• The following models were updated to use the new Deformed Geometry node:
cu_trench_deposition, ecm_microbore, inductor_coil, microconnector_bump_3d,
pulse_reverse_plating, and electrode_insulator_growth.

88 | CHAPTER 1: RELEASE NOTES

Fatigue Module
New Functionality in Version 6.0

RANDOM VIBRATION FATIGUE

A new Random Vibration feature is provided. It is used to compute the fatigue life of a
structure subjected to random vibrations. The response of the structure is defined by
spectral moments of stress, computed from the results of a preceding random vibration
analysis. Two cycle counting models are available — the Dirlik and Bendat models.

As an effect of this addition, the old Vibration Fatigue feature has been renamed to
Harmonic Vibration in order to improve clarity.

IMPROVEMENTS TO CUMULATIVE DAMAGE

Additional S-N curve types have been added to the Cumulative Damage feature. You can
now use an S-N curve with R-value dependence, an S-N curve with mean stress
dependence, or an S-N curve for amplitude stress. For the S-N curve for amplitude stress
option, you have the option of using a Gerber, Goodman, or Soderberg mean stress
correction.

IMPROVEMENTS TO THE EVALUATION OF FATIGUE

The various features in the Fatigue interface now base the fatigue computations on
stresses, strains, and energies evaluated at the integration points of the elements.

New Models in Version 6.0

BRACKET — RANDOM VIBRATION FATIGUE

This model uses a solution from a random vibration analysis of a bracket, and computes
the fatigue life using different cycle counting models. The fatigue properties of the
material are represented by the Basquin fatigue model.

FATIGUE MODULE | 89
 Fuel Cell & Electrolyzer Module
New Functionality in Version 6.0

GENERAL NEW FUNCTIONALITY IN THE ELECTROCHEMISTRY

INTERFACES
• The Highly Conductive Porous Electrode domain node is no longer available when
using a Primary Current Distribution type in the Primary Current Distribution and
Secondary Current Distribution interfaces.
• The Thermodynamic Equilibrium electrode kinetics option is no longer available in
the Highly Conductive Porous Electrode domain node in the Primary Current
Distribution, Secondary Current Distribution, and Tertiary Current Distribution
interfaces as well as for the Lithium-Ion Battery and Battery with Binary Electrolyte
interfaces.
• A new Adsorbing-Desorbing Species section has been added to the Electrode Surface,
Porous Electrode, and related nodes in the electrochemistry interfaces. The section
can be used to model adsorption-desorption and the resulting surface coverage.
• The implementation of the Concentration Dependent Kinetics kinetics type in the
Electrode Reaction and Porous Electrode Reaction (and related) nodes has changed in
order to improve accuracy and numerical convergence for well-posed problems.
Prior to version 6.0, the max() operator was used on the CO and CR parameter
expressions before applying them internally in the definition expression for the iloc
variable. The usage of the max operator has been removed in version 6.0. To get old
models to behave exactly as before, rewrite the CO and CR parameter expressions as
max(CO,eps^2) and max(CR,eps^2), respectively.
• New Adsorption-Desorption Voltammetry tutorial added to the Electrochemistry
Module Application Library.

NEW FUEL CELL AND ELECTROLYZER-SPECIFIC FUNCTIONALITY

• A new Fuel Cell and Electrolyzer Material Library was added.

90 | CHAPTER 1: RELEASE NOTES

New features in the Hydrogen Fuel Cell and Water Electrolyzer interfaces:

• The “No potential constraint”, “There are domains selected on interface level that
are not selected by any phase node”, and “Zero convective velocity” model warning
messages have been added.
• A new Water Condensation-Evaporation domain subnode has been added to the H2
Gas Phase and 02 Gas Phase domain nodes.
• A new Water Gas Shift Reaction domain subnode has been added to the H2 Gas Phase
domain node
• A new Flux boundary subnode has been added to the H2 Gas Phase and 02 Gas Phase
domain nodes.
• A new Separator domain node has been added.
• Updated description of H2 Flow Channel and O2 Flow Channel domain nodes to
H2 Gas Flow Channel and O2 Gas Flow Channel, respectively.
• New functionality added for gas crossover transport of H2, O2, and N2 species
across the membrane domain for proton exchange and hydroxide exchange
electrolyte types.
• New functionality has been added for electroosmotic water drag in the membrane
domain for proton exchange and hydroxide exchange electrolyte types.
• New liquid water properties (density and dynamic viscosity) have been added to the
H2 Gas Phase and O2 Gas Phase domain nodes for announcement to other interfaces,
if H2O selected in the corresponding H2 or O2 gas mixture. Additionally, the

FUEL CELL & ELECTROLYZER MODULE | 91

 descriptions of the announced gas phase properties have been changed to Density of
gas phase and Dynamic viscosity of gas phase, for clarity.
• New H2 Gas-Electrolyte Compartment and O2 Gas-Electrolyte Compartment domain
nodes have been added.
• New H2 Electrode Surface, O2 Electrode Surface, Internal H2 Electrode Surface, and
Internal O2 Electrode Surface boundary nodes have been added.
• A new composition option of Humidified mixture has been added to H2 Inlet, O2
Inlet, and Initial Values nodes, and to the Composition section of H2 Gas Phase and O2
Gas Phase nodes, if an H2O species is present in the corresponding gas mixture.
• A new composition option of Humidified air has been added to O2 Inlet and Initial
Values nodes, and to the Composition section of O2 Gas Phase nodes, if H2O and N2
species are present in the O2 gas mixture.
• A new option of From initial values has been added to the H2 Inlet and O2 Inlet nodes.
The default has been changed from Mole fractions to From initial values. Note that
COMSOL API backward compatibility for this change is not supported.
• Corrected the built-in reference equilibrium potential calculation to account for
reference pressures in the Hydrogen Fuel Cell and Water Electrolyzer interfaces.

92 | CHAPTER 1: RELEASE NOTES

 This bug fix will affect models that have reference pressures set to values other than
1 atm.
• Built-in calculation of gas phase thermal conductivities has been added in the H2 Gas
Phase and O2 Gas Phase domain nodes. The variables (named fc.kgas_mix or
we.kgas_mix) can be used in, for instance, heat transfer simulations.

• The description of sigmas has changed to “Effective electrical conductivity” in

nodes for gas diffusion electrodes and gas diffusion layers. Description remains as
before (“Electrical Conductivity”) in Current Collector nodes.
• The species CH4 (methane) is now supported in the hydrogen gas mixture.
• The default value of the Electrolyte conductivity material input in the Electrolyte
Phase node has been changed to From material (from User defined). Note that
COMSOL API backward compatibility for this change is not supported.
• A Domain material option has been added to the Electrolyte material list in the
Electrolyte Phase node and is set as the default value. The previously available None
option has been removed.
• The new permeability option for Straight Channels in H2/O2 Gas Flow Channel is
now available when Darcy’s law is enabled.
• The Continuity node has been updated to support mass fraction and pressure
continuities on internal pair gas phase boundaries. The updated continuity node
now supports continuity of all domain-dependent variables solved for.

FUEL CELL & ELECTROLYZER MODULE | 93

 GENERAL NEW FUNCTIONALITY IN THE CHEMICAL TRANSPORT
INTERFACES
• The Porous Material node, used for defining and sharing properties for modeling
porous media processes, has been significantly improved.
- The Phase-Specific Properties section lists all properties requested by physics
features with support for a porous material.
- For each phase-specific property, the phase is indicated in the right-most column.
- When requested, multiple entries of the same quantity, one for each phase, can
be listed.
- Tables listing properties can be sorted using any of the columns.
- By right-clicking any property, it is now possible to select Go to Requesting Node
to navigate to where it is used. In the same manner, it is also possible to select Go
to Source to inspect how the property is defined.
- The Homogenized Properties section lists all properties that are not phase specific,
like the permeability. It also lists properties requested by interfaces without
support for a porous material. The homogenized properties can be entered
directly by selecting Locally defined in the Material list.
- When defining properties in a Fluid or Solid, subnode to Porous Material, Locally
defined can be selected, in the Material list, to enter the properties directly.
• A new Reacting Flow multiphysics coupling feature has been introduced. In addition
to isothermal flow, it also supports nonisothermal flow, provided that both a
Chemistry interface and a Heat Transfer interface are included in the coupling.
Opening a model created in a previous version, the old coupling feature is migrated
into a new one.
• Postprocessing variables for the resulting flow rates (mass or molar) on boundaries
are now available in all mass transport interfaces and are listed under the respective
boundary conditions used. They can, for example, be used to evaluate the mass
conservation.
• The out-of-plane geometry extension can now be specified in all mass transport
interfaces. In 2D the out-of-plane thickness is available, while in 1D it is the cross-
sectional area. The out-of-plane dimension is accounted for in the governing
equations and in postprocessing variables for the flow rate.

94 | CHAPTER 1: RELEASE NOTES

• A new Out-of-Plane Flux feature is available in the Transport of Diluted Species and
Transport of Concentrated Species interfaces. It can be used to model flux out of
slender geometries in 2D and 1D.
• The Partition Condition feature now includes the possibility to specify a user-defined
relation, possibly nonlinear, between the concentration on the two sides.

• A new Porous Medium feature is available in the Transport of Concentrated Species

interface. The new feature replaces the Porous Media Transport Properties feature.
When opening a model created in an older version, the previous features are kept in
the model.
• A new Transport of Concentrated Species in Porous Media interface is available in the
Chemical Species Transport branch when adding a physics interface.

REACTING FLOW, DILUTED SPECIES NOW SUPPORTS TURBULENT FLOW

• The Reacting Flow, Diluted Species coupling feature can now be used to model
transport of dilute solutes in a fluid subjected to turbulent flow. The coupling
feature applies turbulence modeling for the species transport. Turbulent mixing is
accounted for, and mass transfer close to walls is modeled using wall functions.
• When licensed to use models for turbulent flow, multiphysics interfaces are available
in the Turbulent Flow, Diluted Species group under Reacting Flow in the Chemical
Species Transport branch.
• The Turbulent Mixing node is now obsolete. It was previously available under the
Transport Properties feature in the Transport of Diluted Species and Transport of
Concentrated Species interfaces.

CHEMISTRY INTERFACE IMPROVEMENTS

• It is no longer needed to explicitly type in variable names to specify the mixture
composition in the Species Matching section. Instead, use the new Species solved for
list to first select a mass transport interface solving for concentrations. Then, for
each of the bulk species in the Chemistry interface, select which of the
concentrations to use.
• Using the Chemistry interface, it is no longer possible to define a species as solvent
when modeling a concentrated mixture (when Type is set to Concentrated Species in
the Mixture Properties section).

THERMODYNAMICS
• The Thermodynamics wizards can now be started from the Physics toolbar.

FUEL CELL & ELECTROLYZER MODULE | 95

 • The filter search used when including species in a thermodynamic system has been
improved. Species whose molecular formula includes any combination of the given
atoms are now shown. When searching for ammonia, NH3 or H3N can hence be
used. It is also possible to filter species using aliases defined in the database. For
example, “dihydrogen oxide” can be used to find water.
• The Davidson model (including pressure correction) can now be used for the
dynamic viscosity.
• Second derivatives, with respect to temperature and pressure, are now available for
the density and the specific volume.

New and Updated Models in Version 6.0

• A new alkaline_electrolyzer model, coupled to Euler–Euler two phase laminar flow.
The model also makes use of the new Separator, H2 Electrolyte-Gas Compartment,
O2 Electrolyte-Gas Compartment, H2 Electrode Surface, and O2 Electrode
Surface nodes.
• A new pemfc_serpentine_flow_field model. The model makes use of the new
electroosmotic drag feature in Membrane Transport.
• A new nonisothermal_pem_fuel_cell model. This tutorial models the intercoupled
electrochemical reactions, charge and species transport, and heat transfer in a
polymer electrolyte membrane (PEM) fuel cell.
• The pem_mea_1d model was updated to use the Hydrogen Fuel Cell interface,
including the new Membrane Transport features.

96 | CHAPTER 1: RELEASE NOTES

Geomechanics Module
New Functionality in Version 6.0

NONLOCAL PLASTICITY
Nonlocal plasticity has been added to Plasticity, Soil Plasticity, and Elastoplastic Soil. You
can enable and control it in the new Nonlocal Plasticity Model section in the settings for
these material models.

With nonlocal plasticity, the equivalent plastic strain is a distributed field, rather than
defined at a point. This makes it possible to solve problems that would otherwise fail
due to high strain localization.

NEW INPUT PROPERTY FOR HARDENING SOIL MODEL

In the Hardening Soil material model, the Ellipse aspect ratio has been added as a new
material parameter. It has the options From material, From coefficient of earth pressure
at rest, and User defined.

NEW INPUT PROPERTY FOR EXTENDED BARCELONA BASIC MODEL

In the Extended Barcelona Basic material model, the Slope of critical state line has been
added as a new material parameter. It has the options From material, Match to Mohr-
Coulomb criterion, Match to Matsuoka-Nakai criterion, and User defined.

Also, the Match to Matsuoka-Nakai criterion option has been added to the
corresponding material parameter in the Modified Cam-Clay and Modified Structured
Cam-Clay material models.

CREEP IMPROVEMENTS
The functionality in the Creep node has been subject to major changes:

• A new general framework gives large improvements in computational speed and

memory usage. For larger models, a speedup by a factor of 10 or more can be
expected.
• You can select the type of scheme that is used for the time integration of the creep
equations. In the Time Stepping section, you can set Method to Automatic, Backward
Euler, Forward Euler, or Domain ODEs. The last method provides backward
compatibility with versions prior to 6.0 and is usually significantly slower for large
models.

GEOMECHANICS MODULE | 97
 • It is no longer possible to add several Creep nodes acting in parallel on the same
selection. When more than one creep mechanism is acting, you can instead add one
or more Additional Creep nodes under a Creep node. As an effect, only one set of
creep strain degrees of freedom is needed, even if several creep models are used. This
gives additional performance gains for models of this type.
• The Norton–Bailey creep model has been removed. It is replaced by a general
isotropic “Bailey-type” hardening formulation, which can be combined with all
creep models, not only the Norton model.
• The type of equivalent stress used to determine the creep rate is now a user input.
You can select von Mises, Hill orthotropic, Pressure, or User defined. Thus,
nonisotropic creep can now be modeled.
• The old creep models Potential, Volumetric, and Deviatoric have been removed. The
same functionality is now achieved either through the User defined creep model, or
by using the new Inelastic Strain Rate node.

PRESCRIBED INELASTIC STRAIN RATE

A new Inelastic Strain Rate subnode has been added to Linear Elastic Material, Layered
Linear Elastic Material, and Nonlinear Elastic Material. The effect of this node is similar
to the existing External Strain node, in that it injects an inelastic strain in the material
model. The difference is that here you instead specify a rate, which is then integrated
into an inelastic strain contribution. By using this feature, you do not need to add an
extra ODE interface when you have a rate equation describing an extra strain
contribution.

The strain rate can be specified in terms of a strain tensor, a deformation gradient, an
inverse deformation gradient, or three orthogonal stretches.

REDUCED INTEGRATION
The new framework for reduced integration will in many cases make the solution of
models with nonlinear materials, such as plasticity and creep, significantly faster.
Reduced integration is controlled from the Quadrature Settings section in material
models such as Linear Elastic Material, Nonlinear Elastic Material, and Elastoplastic Soil
Material.

98 | CHAPTER 1: RELEASE NOTES

New Model in Version 6.0

TRANSIENT RESPONSE OF A SHALLOW FOUNDATION ON UNSATURATED

SOIL
In this example, the transient response of a shallow foundation resting on an
unsaturated soil stratum is investigated. The changes in pore suction due to water
infiltration and evaporation is investigated using the Extended Barcelona Basic soil
model. The results show the deformation of the clay stratum due to footing pressure
and changes in pore suction.

GEOMECHANICS MODULE | 99
 Heat Transfer Module
New and Improved Functionality in Version 6.0

IMPROVED COMPUTATIONAL PERFORMANCE FOR SURFACE-TO-SURFACE

RADIATION
The Surface-to-Surface Radiation interface has a new Jacobian Contribution section
where you can control the contribution of the radiosity terms to the Jacobian matrix.
The default option, Only local contributions to radiosity, computes an approximate of
the Jacobian matrix. Combined with improved solver settings, it provides a dramatic
decrease of the CPU time and memory requirements for surface-to-surface radiation
computations, in particular with the hemicube method. The previously used approach,
computing the full Jacobian matrix, remains available and can be activated by selecting
the Include contributions from total irradiation option.

In addition, the Surface-to-Surface Radiation interface has a new Store view factors on
disk option in the View Factor section. The view factors can be stored in the model
when they have been computed once, provided that the mesh, the selection of the
radiating boundaries, the radiation direction, and the external radiation sources or
symmetry conditions are unchanged. This may significantly reduce the computational
time for cases where the view factor calculation is demanding, typically with specular
or semitransparent surfaces.

EXTENDED FEATURES FOR SURFACE-TO-SURFACE RADIATION

The Ray shooting option is now available as a Surface-to-surface radiation method, as well
as in 2D axisymmetric models, for the computation of the view factors.

The user interface of the surface-to-surface radiation features has been simplified as
follows:

• The new View factor update threshold list in the View Factor section makes it possible
to easily control the frequency of view factor computations in time-dependent
studies, when using moving symmetry planes or time-dependent surface properties.
• In the Opaque Surface and Semitransparent Surface features, a single Define properties
on each side check box controls the display of side-specific user inputs for the

100 | CHAPTER 1: RELEASE NOTES

 definition of the emissivity, reflectivity, transmissivity, and critical angle on two sides
of a shell.
• The Wavelength unit for the spectral band endpoints can now be defined once for all
at the interface level in the Radiation Settings section.

In the View Factor section, the new High order mesh elements check box enables the use
of mesh elements defined by the geometry shape functions to compute the view factors
with high accuracy, or the use of linearized mesh elements to reduce computational
cost. In addition, the Mesh element self-irradiation check box has been added to
account for self-irradiation in concave elements.

The Prescribed Radiosity feature now has a Polar option to specify the Directional
dependence of emissivity. In addition, this feature has a new user input for the external
temperature, Text, when the external radiation is defined from a blackbody or
graybody model. Either a user-defined temperature can be entered or an ambient
temperature can be selected.

In the Prescribed Radiosity and Semitransparent Surface features, there is a new

Coordinate System Selection section for the specification of the reference vector of the
azimuthal angle. The choice must be made among boundary coordinate systems.

The radiosity is now assumed to be discontinuous by default on edges, which

corresponds to the definition of a slit on the radiosity variables. This benefits cases
where only one of two faces connected by an edge is exposed to solar radiation, and
this improves the robustness of the computation.

Finally the adaption algorithm available for the Ray shooting method has been
improved to more efficiently detect radiosity or topology changes.

EXTENDED FEATURES FOR HEAT TRANSFER IN POROUS MEDIA

The Porous Medium domain feature now includes three different models, which
correspond to the thermal equilibrium and thermal nonequilibrium assumptions and
to the particular configuration of packed beds of pellets. These models can be selected
from the new Porous medium type list in the Porous Medium section:

• When the Porous Medium feature is added by the Heat Transfer in Porous Media
predefined interface, the Porous medium type is set by default to Local thermal
equilibrium. This corresponds to the model previously available within the Porous
Medium feature, solving for a single equation for the temperature with effective
properties, with some extra functionality. New averaging models are available for the
Effective thermal conductivity: Solid spherical inclusions, Fluid spherical inclusions,

HEAT TRANSFER MODULE | 101

 Wrapped screen, and Sintered metal fibers. In addition, the Volume average and
Reciprocal average models have been renamed Plane layers parallel to heat flow and
Plane layers perpendicular to heat flow, respectively, for clarity. Finally, using the new
Equivalent thermal conductivity option, you can specify a user-defined expression.
• When the Porous Medium feature is added by the new Local Thermal
Nonequilibrium predefined interface, the Porous medium type is set by default to
Local thermal nonequilibrium. This replaces the model previously available within the
Local Thermal Nonequilibrium multiphysics interface (including the Heat Transfer
in Solids interface, the Heat Transfer in Fluids interface, and the Local Thermal
Nonequilibrium multiphysics coupling). The settings for the fluid and solid
temperature equations are available in the Fluid and Solid subnodes of the Porous
Medium feature, which provides the setting of the Interstitial convective heat transfer
coefficient (previously available in the multiphysics coupling node). In this
configuration, the Local Thermal Nonequilibrium Boundary feature is automatically
added to delimit the domain in which a two-temperature model is used.
• When the Porous Medium feature is added by the new Heat Transfer in Packed Beds
predefined interface, the Porous medium type is set by default to Packed bed. This
option corresponds to a new multiscale modeling approach specifically tailored for
pellet beds, as described in the section below.

For the three Porous medium type options mentioned above, the postprocessing
variables (thermal properties, heat fluxes, heat sources, heat and energy balance) are
available in a unified way for homogenized quantities.

MULTISCALE HEAT TRANSFER IN PELLETS BEDS

The new Heat Transfer in Packed Beds predefined interface is available under the
Porous Media group in the Heat Transfer branch in the model wizard. It adds the Porous
Medium feature, with Porous medium type set to Packed bed, to model the pellets of a
packed bed as spherical homogenized porous particles in which the temperature varies
radially. The temperature distribution in the pellets is computed for every position in
the packed bed and is coupled to the temperature in the surrounding fluid through an
interstitial heat flux or a temperature continuity condition between the pellets’ surface
and the fluid.

In this configuration, the Porous Medium feature has three subnodes (Fluid, Pellets, and
Pellet-Fluid Interface), and the Local Thermal Nonequilibrium Boundary feature is
automatically added to delimit the domain in which a model solving for multiple
temperatures is used. Note that the new Pellet subnode, available under the Porous
Material node, can be used to define the properties of the pellet phase.

102 | CHAPTER 1: RELEASE NOTES

EXTENDED FEATURES FOR HEAT TRANSFER IN THIN STRUCTURES
In the Heat Transfer in Shells interface, the formulation of heat sources and fluxes has
been improved to account for the effect of the surface curvature on the layers’ area and
length. Because of the use of an extra dimension through the shell’s thickness instead
of representing it explicitly in the geometry, the dimensions of the selected boundary
or edge in the geometry have to be scaled when computing fluxes and sources through
the thickness. The application of this scaling factor to variables and equations makes it
possible to accurately handle thick and curved layered shells.

The new Lumped System Connector and Lumped System Connector, Interface features are
now available in the Heat Transfer in Shells interface to connect a lumped thermal
system to a shell through a layer’s edge or interface, respectively. When a Lumped
Thermal System interface with an External Terminal node is present in the model, this
terminal can be selected as a Source in the Terminal Inputs section of any of the new
features. The use of the Lumped System Connector, Interface feature is illustrated in the
Lumped Composite Thermal Barrier with Shells tutorial.

The new Symmetry subnode is available under the Thin Layer, Thin Film, and Fracture
boundary features of the Heat Transfer interface.

The new Thermal Insulation, Interface feature is added by default in the Heat Transfer
in Shells interface and specifies thermal insulation on the exterior interfaces of the shell.
When added manually, the thermal insulation condition can be applied on any
interface of the layers.

A new Layerwise constant properties check box is available and active by default when
the Layer type is Thermally thin approximation in the Solid and Fluid features of the Heat
Transfer in Shells interface and in the Thin Layer and Thin Film features of the Heat
Transfer interface. A significant speedup can be obtained with this option, which
assumes that the material properties are constant per layer but still changing with layers
or the layers’ temperature.

NONISOTHERMAL REACTING FLOW

The new Nonisothermal Reacting Flow multiphysics interface is available in the
Laminar Flow and Turbulent Flow versions, under the Nonisothermal Reacting Flow group
in the Chemical Species Transport branch. It adds the Reacting Flow multiphysics
coupling and the Chemistry, Transport of Concentrated Species, and Heat Transfer in
Fluids interfaces, together with a laminar or turbulent version of the Single Phase Flow
interface. It provides the heat transfer functionality to reacting flow models, by
accounting for the cross contributions between the heat and species equations, in
particular for the enthalpy of phase change and the enthalpy diffusion term. It handles

HEAT TRANSFER MODULE | 103

 the temperature, pressure, and concentration dependence of the material properties;
and accurate heat and energy balance is therefore expected using the predefined
variables.

EXTENDED FEATURES FOR MOISTURE TRANSPORT

The Moisture Flow multiphysics coupling has a new option to account for the leakage
velocity at walls where evaporation and condensation take place. The Account for Stefan
velocity at walls check box is available in the Moisture Transport at Walls section, when
the Mixture type for moist air is set to Concentrated species in the Moisture Transport
interface. This functionality is available for free and porous media and automatically
adds the corresponding contribution to the velocity of the fluid flow interface.

In the Moisture Transport interface, the Hygroscopic Porous Medium feature has been
updated to match the classic design for features using porous materials. It now contains
the Liquid Water, Moist Air, and Porous Matrix subnodes for the settings of each phase.

In addition, the Moisture Transport interface has extended functionality for several
boundary conditions:

• The Periodic Condition feature is now available. Discontinuous Galerkin constraints

or classical constraints are applied on the relative humidity, vapor concentration, or
mass fraction, depending on the adjacent domain features.
• In the Inflow, Open Boundary, and Moisture Content features, it is now possible to
specify the vapor mass fraction.
• In the Open Boundary feature, the formulation of the Discontinuous Galerkin
constraints has been improved.
• In the Moisture Content and Continuity features, the Discontinuous Galerkin
constraints have been added.

Finally, several postprocessing variables are now available to check the mass balance. In
particular, the mt.massBalance, mt.dwcInt, mt.ntfluxInt, and mt.GInt
predefined variables make it possible to control the mass balance, the total
accumulated moisture rate, the total net moisture rate, and the total mass source,
respectively.

EXTENDED FEATURES FOR PHASE CHANGE

The Phase Change Interface feature now accounts for the solid translation velocity in the
velocity of the fluid phase at the interface. This can be applied to continuous casting
processes, for example.

104 | CHAPTER 1: RELEASE NOTES

The new Phase Change Interface, Exterior feature is available to define the temperature
and position of a phase change interface with the Stefan condition, in the same way as
the Phase Change Interface feature, but applicable on exterior boundaries adjacent to
Fluid, Solid, or Porous Medium domains. The properties of the phase not represented in
the computational domain have to be set in the user interface instead of being gathered
from an adjacent domain.

THERMAL CONDUCTIVITY MODEL FOR LARGE STRAINS

When the material and spatial frames differ, a new Deformation model for thermal
conductivity list is available in the features of the heat transfer interfaces representing
solid domains (Solid, Biological Tissue, and Shape Memory Alloy).

The Standard (previously existing) option assumes that thermal conduction is not
affected by the deformation of the solid. In other words, a temperature difference
between two faces of a solid induces the same heat flux as in the undeformed
configuration, whatever the deformation is. The new option, Large strain, defines the
effective thermal conductivity in deformed configurations accounting for the rotation
between the undeformed and the deformed configuration but not for the strain. In the
case of isotropic thermal conductivity, this option ensures that the thermal
conductivity in the deformed solid is the same as in the undeformed solid.

NUCLEATE BOILING HEAT FLUX

The Rohsenow correlation, describing the heat flux due to nucleate boiling at a solid
surface immersed in a fluid, is now available as a predefined option in the following
heat flux features: Heat Flux, Out-of-Plane Heat Flux, Point Heat Flux, and Isothermal
Domain Interface in the Heat Transfer interface; Heat Flux and Heat Flux, Interface in the
Heat Transfer in Shells interface; and Nucleate Boiling Thermal Resistor and Heat Rate in
the Lumped Thermal System interface. The Nucleate boiling heat flux option is available
from the Flux type list, and the coefficients needed to define the correlation are
predefined for some liquids and solid surfaces. Alternatively, user-defined coefficients
can be specified. The critical heat flux that can be expected with nucleate boiling is
available as a predefined variable for postprocessing.

MARANGONI EFFECT
The Marangoni Effect multiphysics coupling has been updated to account for both the
tangential and normal effects of surface tension. In addition, the contact angle can now
be set in the user interface, either directly or through Young’s equation. The new
formulation results in improved computational performance.

HEAT TRANSFER MODULE | 105

 MISCELLANEOUS IMPROVEMENTS
• A new set of predefined variables is available for the efficient postprocessing of heat
and energy balances. These variables are exact in cases of constant material
properties and can be a good approximate of the default variables in cases of
moderate material property variation in the temperature and pressure ranges present
in the model. These variables are much faster to evaluate than the default variables
for heat and energy balances. The ht.dEiInt_cst, ht.dEi_cst, ht.Ei_cst,
ht.H_cst, and ht.DeltaH_cst variables are available for heat balance. The
ht.dEi0Int_cst, ht.dEi0_cst, ht.Ei0_cst, ht.H_cst, and ht.DeltaH_cst
variables are available for energy balance.
• The default color table for the temperature plots has been changed from
ThermalLight to HeatCameraLight. The default color table for the relative humidity
has been changed from Rainbow to JupiterAuroraBorealis with a Reverse color map
transformation.
• An efficient formulation of the sensible enthalpy predefined variable, ht.DeltaH, is
now available for free and porous media containing moist air. This analytical
formulation, taking advantage of the polynomial definition of the heat capacity for
moist air, drastically reduces the evaluation time of the accumulated heat and energy
rates in postprocessing.
• In fluid domains, when Fluid type is From material or Gas/Liquid, the ratio of specific
heats is now computed by default from the density and the heat capacity, using
Mayer’s relation.
• The Semitransparent Surface feature is now available in the Radiation in Absorbing
and Scattering Media interface. The settings are similar to those of the
Semitransparent Surface feature available in the Radiation in Participating Media
interface.
• The Semitransparent Surface and Incident Intensity boundary features in the
Radiation in Participating Media and Radiation in Absorbing-Scattering Media
physics interfaces have an additional user input for the external temperature, Text,
when the external radiation is defined from a blackbody or graybody model. Either
a user-defined temperature can be entered or an ambient temperature can be
selected.
• In the Incident Intensity feature of the Radiative Beam in Absorbing Media interface,
the definition of the radiative intensity has been fixed when using built-in beam
profiles in configurations where the beam is not perpendicular to the surface.

106 | CHAPTER 1: RELEASE NOTES

• In the Deposited Beam Power feature of the Heat Transfer interface and in the
Incident Intensity feature of the Radiative Beam in Absorbing Media interface, the
definition of the distribution function has been fixed for 2D models.
• The contributions of Heat Transfer with Surface-to-Surface Radiation and Heat
Transfer with Radiative Beam in Absorbing Media interfaces to the radiative heat
flux of the Heat Transfer interface, ht.rflux, have been added. Contributions of
the Heat Transfer with Radiation in Participating Media and Heat Transfer with
Radiation in Absorbing-Scattering Media interfaces have been fixed.
• In the Surface-to-Surface Radiation interface, the operator rad.radGpEval(expr),
used to evaluate expressions at Gauss points, has been replaced by the built-in
operator gpeval(order, expr).
• In the Moist Porous Medium feature of the Heat Transfer interface, the definition of
the effective properties has been fixed to include the liquid water contribution.
• The Stationary, One-way with Initialization, Moisture Flow and Time Dependent, One-
way with Initialization, Moisture Flow preset studies are now available for turbulent
moisture flow models with wall functions.
• The Heat Transfer in Shells interface now supports the Eigenfrequency study.
• In the Thin Layer, Thin Film, and Fracture features of the Heat Transfer interface, the
handling of the initial condition has been fixed when Layer type is General. The Initial
Values subnode is now also available when the Layer type is set to Thermally thick
approximation. For the Thermally thick approximation layer type, it is applicable on
exterior boundaries only to define the temperature on the thin structure’s outer
side, the temperature on inner sides being defined from the adjacent domain’s initial
temperature. In addition, for the General and Thermally thick layer types, there is a
new From adjacent domains option in the Initial Values subnode, to get the average
value of the initial temperature in the adjacent domains.
• The default solver suggestions have been improved when solving for the thermal
damage indicator with the Irreversible Transformation feature.
• A new Coordinate System Selection section is available in the Thermoelectric Effect
multiphysics coupling.
• The Longitudinal dispersivity and Transverse dispersivity settings in the Thermal
Dispersion feature are now available with all COMSOL products that include the
feature.

HEAT TRANSFER MODULE | 107

 • The definition of the isothermal Joule–Thomson coefficient variable, ht.mujtT, has
been improved.
• Pair features have been updated following the new framework and are now also
applicable on interior boundaries.

New Applications in Version 6.0

LUMPED COMPOSITE THERMAL BARRIER WITH SHELLS

This example is a variant of the Composite Thermal Barrier tutorial and shows how
to set up multiple sandwiched thin layers with different thermal conductivities in two
different ways. First, the composite is modeled as a 3D object. In the second approach,
the Lumped Thermal System physics interface is used to avoid resolving the thin
domains by use of thermal circuit modeling. In addition, the top and bottom parts of
the steel column are represented with two shells.

THE CONTINUOUS CASTING — ARBITRARY LAGRANGIAN-EULERIAN

METHOD
This example describes the cooling and solidification, from melt to solid metal, in a
continuous casting process. The model includes nonisothermal properties,
temperature distribution, a flow field, and phase change. This example uses the
arbitrary Lagrangian–Eulerian (ALE) method for modeling the phase change, and the
casting velocity is handled automatically.

Updated Applications in Version 6.0

The tutorials have been updated to take advantage of the latest geometry, mesh, and
solver features in order to have simplified step-by-step instructions for the model
construction. In addition to these updates, some tutorials have been updated
regarding physical modeling:

• The Evaporation in Porous Media with Large Evaporation Rates tutorial has
been replaced with the Drying of a Potato Sample tutorial, with simplified mesh
and solver settings.
• The use of the new predefined variables for mass balance is demonstrated in the
Evaporative Cooling of Water, Condensation Detection in an Electronic Device
with Transport and Diffusion, Condensation Risk in a Wood Frame Wall, and
Drying of a Potato Sample tutorials.

108 | CHAPTER 1: RELEASE NOTES

• The Heat Sink tutorial has been updated with the control of energy balance using
the new predefined variable ht.energyBalance.
• In the Electronic Chip Cooling tutorial, Study 1 is not used anymore to get the
initial conditions for Study 2.
• The Vacuum Flask tutorial has been updated with the use of a Fluid feature with 0
velocity in the air domain for clarity.
• In the Thermal Performances of Windows and Glazing Influence on Thermal
Performances of a Window tutorials, the Open Boundary features have been replaced
by Heat Flux features.

Backward Compatibility with Version 5.6

• In the Prescribed Radiosity feature in the Surface-to-Surface Radiation interface, if the
graybody model is selected to define the radiative intensity, the emissivity is
temperature dependent, and the temperature model input is automatically defined
from a multiphysics coupling in an API script, then you must update the script to
define Text. In other cases, Text is set to the value that was set to the model input
in version 5.6 or earlier so that the model behavior is unchanged.
• In Heat Transfer interfaces, in the Deposited Beam Power boundary condition, for
User defined beam profiles, the beam intensity is now defined from its value in the
beam cross section, which corresponds to the deposited beam power divided by the
cosine of the incidence angle. Nevertheless, for User defined beam profiles, the
models and API scripts created in version 5.6 or earlier behave identically in version
6.0.
• In the Opaque Surface feature of the Surface-to-Surface Radiation interface, the
coordinate system choice is restricted to the boundary system. When opening a
model created in version 5.6 or earlier with another coordinate system selected, this
input should be checked for consistency.
• In the Surface-to-Surface Radiation interface, when the Expression or Tolerance
settings had been modified in the View Factor Update section of a model created in
version 5.6 or earlier, these settings are reported in the new View Factor section.
• A modification of the numerical results within the solver tolerance range may be
observed when computing surface-to-surface radiation models created in version
5.6 or earlier, due to the introduction in the Surface-to-Surface interface of the
Jacobian contribution option set by default to Only local contributions to radiosity.
When opening a model created in version 5.6 or earlier, a warning is displayed under

HEAT TRANSFER MODULE | 109

 the Surface-to-Surface Radiation interface. Note that the accuracy level remains the
same because the convergence criteria and the solver tolerance are unchanged.
• The Local Thermal Nonequilibrium predefined multiphysics interface and the Local
Thermal Nonequilibrium multiphysics coupling are now obsolete. When opening a
model created in version 5.6 or earlier, this feature is not migrated. It should be
manually replaced by the Porous Medium feature with the Porous medium type set to
Local thermal nonequilibrium. Support for features marked as obsolete is not ensured
in future versions; thus it is strongly recommended to transfer this feature to the new
feature.
• In the Moisture Transport interface, the Hygroscopic Porous Medium feature not
supporting the Porous Material functionality is now obsolete. When opening a model
created in version 5.6 or earlier, this feature is not migrated. It should be manually
replaced by the new Hygroscopic Porous Medium feature. Support for features marked
as obsolete is not ensured in future versions; thus it is strongly recommended to
transfer this feature to the new feature.

Backward Compatibility with Version 5.5

• The Heat and Moisture multiphysics coupling feature is no longer applicable to
domains where the Fluid (Heat Transfer interface) feature is active. Models built in
version 5.5 and before, where the Heat and Moisture multiphysics feature and the
Fluid feature with the Moist air option have been used, are updated when opened in
version 5.6. The Fluid feature is then replaced by the Moist air feature so the
multiphysics coupling is still active. In the unexpected case where the Heat and
Moisture multiphysics feature and the Fluid feature with an option other than Moist
air were combined, the multiphysics coupling is no longer active in version 5.6,
which is likely to modify the model solution.
• In the Surface-to-Surface Radiation interface, the Opaque Surface and
Semitransparent Surface features not supporting directional dependence of surface
properties are now obsolete. When opening a model created in version 5.5 or earlier
in the user interface, these features are not migrated. They should be manually
replaced by the new Opaque Surface and Semitransparent Surface features to account
for directional dependence of surface properties. Support for features marked as
obsolete is not ensured in future versions; thus it is strongly recommended to
transfer all these features to the new features.
• In the Heat Transfer interfaces, the Porous Medium feature not supporting the new
Porous Material functionality is now obsolete. When opening a model created in
version 5.5 or earlier in the user interface, this feature is not migrated. It should be

110 | CHAPTER 1: RELEASE NOTES

 manually replaced by the new Porous Medium feature (which comes with Porous
Matrix and Fluid subfeatures) to allow an updated definition of the material
properties of the porous medium. Support for features marked as obsolete is not
ensured in future versions; thus it is strongly recommended to transfer this feature
to the new feature.
• The Temperature feature is no longer applicable on boundaries adjacent to a domain
where the Isothermal Domain feature is active. When opening a model built in
version 5.5 and before, where the Temperature feature is applied in this way, the
node is present but the feature is not active, which is likely to modify the model
solution.
• The Open Boundary feature is no longer applicable on boundaries adjacent to a
nonfluid domain. When opening a model built in version 5.5 and before, where the
Open Boundary feature is applied in this way, the node is present but the feature is
not active, which is likely to modify the model solution.
• The Surface-to-surface radiation interface is no longer applicable on boundaries that
belong to infinite element domains, perfectly matched layer domains or absorbing
domains since the view factor computation does not account for the corresponding
coordinate system change.

Backward Compatibility with Version 5.4

• In the Inflow feature (Heat Transfer interface), the Upstream absolute pressure user
input is now inactive, unless the new Specify upstream absolute pressure check box
is selected. Scripts using the COMSOL API that used to fill this input field need to
add an instruction for selecting the Specify upstream absolute pressure check box
beforehand.
• The Open Boundary feature now comes with a new default formulation based on the
Danckwerts inflow condition to specify the incoming temperature. The former
default formulation is still available by choosing the Discontinuous Galerkin
constraints option in the Inflow Condition section. To keep the former default
formulation in scripts using the COMSOL API, you need to add the corresponding
instruction to set the Discontinuous Galerkin constraints option in the Inflow Condition
section.
• When the Relative humidity option is used in the Moisture Content boundary
condition, the associated temperature now has to be defined. In version 5.4 and
before, the moisture transport local temperature was used instead. In order to
reproduce previous results, models created in version 5.4 or earlier will use the

HEAT TRANSFER MODULE | 111

 moisture transport interface temperature when opened in the GUI. The scripts
using the COMSOL API will use the default temperature 293.15 K instead and may
need to be updated to reproduce existing results.
• In the Heat Flux, Interface and Heat Source, Interface features, when the Heat Rate
option is selected in the Heat Flux and Boundary Heat Source sections, respectively,
the surface definition has been corrected in order to take into account several
interfaces in the Interface Selection. For example, if two interfaces are selected, the
surface definition is now multiplied by two.
• The Deposited Beam Power, Interface and Surface-to-Ambient Radiation, Interface
features are now applicable on the interior interfaces. In the scripts using the
COMSOL API in which the All interfaces option was selected, the interior interfaces
were selected but no contribution was added on these interfaces. Changes in the
numerical results are expected now that the contribution is added also on the
interior interfaces.

112 | CHAPTER 1: RELEASE NOTES

Liquid & Gas Properties Module
New Functionality in Version 6.0
• The Thermodynamics wizards can now be started from the Physics toolbar.
• The filter search used when including species in a thermodynamic system has been
improved. Species whose molecular formula includes any combination of the given
atoms are now shown. When searching for ammonia, NH3 or H3N can hence be
used. It is also possible to filter species using aliases defined in the database. For
example, “dihydrogen oxide” can be used to find water.
• The Davidson model (including pressure correction) can now be used for the
dynamic viscosity.
• Second derivatives, with respect to temperature and pressure, are now available for
density and specific volume.

LIQUID & GAS PROPERTIES MODULE | 113

 MEMS Module
New Functionality in Version 6.0

MAGNETOMECHANICS MULTIPHYSICS INTERFACE

Two new physics interfaces for the analysis of coupled magnetic and mechanical effects
have been added: Magnetomechanics and Magnetomechanics, No Currents. When you add
such an interface, two physics interfaces are added to the model: Solid Mechanics and
either Magnetic Fields or Magnetic Fields, No Currents. The new Magnetomechanical
Forces multiphysics coupling is also added.

You can find these interfaces under the Electromagnetics-Structure Interaction branch in
the Structural Mechanics folder in the Add Physics tree.

This interface also requires the AC/DC Module.

MAGNETOSTRICTION COUPLED TO ROTATING MACHINERY, MAGNETIC

INTERFACE
The Magnetostriction multiphysics coupling can now also be used to connect the Solid
Mechanics and Rotating Machinery, Magnetic physics interfaces.

COMPONENT MODE SYNTHESIS

Linear components built using the Solid Mechanics and Multibody Dynamics
interfaces can be reduced to computationally efficient reduced-order models using the
Craig–Bampton method. A reduced component can then be used in dynamic or
stationary analyses, either in a model consisting of only reduced components, or
together with ordinary full FE models. The latter can then be nonlinear.

To create such components, you add a Reduced Flexible Components node in the physics
interface.

RIGID JOINT IN SOLID MECHANICS INTERFACE

A Rigid Joint node has been added to the Solid Mechanics interface. Such a joint can
be used to connect two attachments rigidly. The feature is similar to the same joint in
the Multibody Dynamics interface. The main purpose is for joining CMS components.

114 | CHAPTER 1: RELEASE NOTES

ATTACHMENTS NO LONGER REQUIRE MULTIBODY DYNAMICS LICENSE
You can now add the Attachment node with a Structural Mechanics Module or MEMS
Module license. The main usage is in the context of creating and joining CMS
components.

POINT LOADS AT ARBITRARY LOCATIONS IN SOLID MECHANICS

With the new Point Load, Free and Ring Load, Free features, you can apply point loads
at arbitrary locations that do not coincide with a geometrical point or a mesh node.
This is particularly useful when working with moving loads.

NEW DEFAULT BEHAVIOR FOR CONTACT AND CONTINUITY

If any contact pairs are defined under Definitions, a default Contact node will be added
in the Solid Mechanics interface. Similarly, if there are any identity pairs, a default
Continuity node will be added. In many cases, the pair handling can be considered
automatic because pairs can be generated during the finalization of the geometry
sequence.

As an effect of the general reformulation of the pair functionality, the Source external
to current physics check box is no longer needed and has been removed.

TWISTING DEGREE OF FREEDOM IN 2D AXISYMMETRY

In the Solid Mechanics interface in 2D axisymmetry, it is now possible to include
circumferential deformation. To do so, you select the Include circumferential
displacement check box in the Axial Symmetry Approximation section in the settings for
the physics interface.

INITIAL IMPERFECTION IN BUCKLING ANALYSIS

You can now use the buckling modes from a linear buckling analysis as initial
imperfections to the geometry when performing a full nonlinear buckling analysis.
This functionality is controlled from the new Buckling Imperfection node, which can be
added from Physics Utilities under Definitions.

VISCOELASTICITY IMPROVEMENTS
There are several important additions to the viscoelastic material models:

• For frequency-domain and time-dependent analyses, all of the viscoelasticity models

have been augmented with the possibility to include viscoelasticity in the volumetric
deformation.
• The Generalized Maxwell model now includes the possibility to prune branches
representing frequency ranges outside the bandwidth of prescribed loads, improving

MEMS MODULE | 115

 the performance in time-dependent analysis for models with many viscoelastic
branches.
• For frequency-domain analyses, a new user-defined viscoelasticity model makes it
possible to enter frequency-dependent expressions for the loss and storage moduli
or compliances.
• Through a new formulation of the viscoelastic equations, it is now possible to solve
for eigenfrequencies in a structure containing viscoelastic materials using a standard
procedure for damped eigenfrequency problems. Previously, the eigenvalue
problem was nonlinear in the frequency, and only one eigenfrequency at a time
could be computed. The new approach is available for all viscoelastic models except
user defined, under the assumption that fractional derivatives are not used.

IMPROVED MIXED FORMULATION

In material models that have an option to select a mixed formulation, you can now
modify the discretization for the auxiliary dependent variable (pressure or volumetric
strain). To do so, Advanced Physics Options must be selected. Then, a new section
named Discretization will appear in the settings for the material model as soon as a
mixed formulation is selected. In this section, you can choose between different types
of shape functions for the auxiliary dependent variable.

Another change is that the auxiliary dependent variable now has feature scope, rather
than physics scope.

NEW DAMPING MODELS

Two new damping models have been added for the mechanical material models:

• The Wave attenuation model is essentially a viscous model but with parameters given
by measured data for the attenuation of elastic waves in the material. It is available
in the Linear Elastic Material in Solid Mechanics.
• The Maximum loss factor model is mainly intended for time-domain analysis of
materials for which a loss factor representation provides a good description in the
frequency domain. This damping model is available for all material models that
support viscous damping.

The new damping models can be used in both frequency-domain and time-domain
studies.

116 | CHAPTER 1: RELEASE NOTES

DAMPING FOR PIEZOELECTRIC MATERIALS
Two new options have been added to the Dielectric Loss subnode under Piezoelectric
Material in the Solid Mechanics interface: Complex Permittivity, intended for frequency-
domain analysis, and Maximum loss factor, intended for time-domain analysis.

MULTIPOLE DEBYE DISPERSION FOR PIEZOELECTRIC MATERIALS

New damping models have been added for piezoelectric materials. Under Charge
Conservation, Piezoelectric, you can now add a Dispersion subnode. There, you can
choose between the Debye and Multipole Debye dispersion models.

REDUCED INTEGRATION
A new framework has been added for reduced integration. It is now available in the
Solid Mechanics interface. Reduced integration is particularly useful when the
computational cost per integration point is high, which is true for many advanced
material models.

Reduced integration is controlled from the Quadrature Settings section in various

material models. It is available in top-level material models like Linear Elastic Material.
The selected integration rule will then be inherited by any subnodes that may be
added.

EASIER GAUSS POINT EVALUATION IN DOMAINS

A set of physics-scoped operators have been added as tools for easier evaluation of
Gauss point results in domains. The operators automatically adapt to the Gauss point
pattern in each element. These operators are:

• <phys>.gpeval(expr): Map a field of suitable order to Gauss point data.

• <phys>.elemgpmax(expr): Maximum value in any Gauss point in an element.
• <phys>.elemgpmin(expr): Minimum value in any Gauss point in an element.
• <phys>.elemavg(expr): Average value of the Gauss points in an element.
• <phys>.elemavg(expr): Integration over an element using Gauss point data.

INTERCALATION STRAIN
A new subnode, Intercalation Strain, has been added to Linear Elastic Material and
Nonlinear Elastic Material in the Solid Mechanics interface. The purpose is to model
inelastic strains caused by ion transport.

This feature requires the Battery Design Module.

MEMS MODULE | 117

 INPUT OF RESIDUAL STRESS
In the External Stress node, there is now an option called Residual Stress. A residual
stress will not directly affect the displacements. That is, if you just enter a residual stress
and no other loads, there will be no displacements. The stress is, however, added to
the stress tensor in the sense that it will be part of the stress state that is used in various
material models.

EXTERNAL STRESS CAN BE ENTERED AS FIRST PIOLA-KIRCHHOFF

STRESS TENSOR
In the External Stress node in the Solid Mechanics interfaces, it is now possible to enter
the stress state in the form of a 1st Piola–Kirchhoff stress tensor.

SYMMETRIC FORMULATION FOR RIGID CONNECTOR AND RIGID DOMAIN

Under geometric nonlinearity, the consistent stiffness and mass matrices generated by
the Rigid Connector and Rigid Domain features may become nonsymmetric. The solution
of a nonsymmetric problem requires significantly more memory and computation time
than solving a symmetric problem of the same size. In version 6.0, you can choose to
use a less accurate symmetric matrix instead. Doing so may require some more
iterations, but each iteration will be significantly faster.

In order to use the symmetric formulation, select Use symmetric formulation for
geometric nonlinearity in the Formulation section in the settings for the feature.

COORDINATE SYSTEM TYPES RESTRICTED FOR RIGID CONNECTOR

For a rigid connector, it is not meaningful to select a coordinate system with position-
dependent axis orientations. This is no longer allowed, neither in the Rigid Connector
node itself nor in its subnodes.

EXPRESSIONS FOR THE ROTATION VECTOR CHANGED IN RIGID

CONNECTOR AND RIGID DOMAIN
The expression defining the variables giving the rotation vector in a rigid connector
and rigid domain has been changed in such a way that it is more robust with respect
to computing Jacobian contributions. This can be important if you want to, for
example, control such a variable using other equations.

The affected variables for a rigid connector are named <phys>.th<dir>_<tag> (for
example, solid.thz_rig1). For a rigid domain, the names are of the type
<phys>.<tag>.th<dir> (for example, solid.rd1.thy).

118 | CHAPTER 1: RELEASE NOTES

CONSTRAINT GROUPS FULLY SUPPORTED FOR RIGID CONNECTORS AND
RIGID DOMAINS
In earlier versions, constraint groups could not be used for rigid connectors or rigid
domains if the option Evaluate reaction forces was selected. This limitation has been
removed.

STRICTER DEFAULT TOLERANCE FOR TIME-DEPENDENT ANALYSIS

The default tolerance for time-dependent analysis has been decreased from 0.01 to
0.001 for the Solid Mechanics interface. This will affect both new and old models in
which the solver sequence has not been manually edited. The effect is that solutions
will be more accurate but typically at the cost of longer solution times.

THE TIME-DEPENDENT, PRESTRESSED, MODAL STUDY TYPE REMOVED

Due to problems with obtaining a correct linearization in all situations, the Time-
Dependent, Prestressed, Modal study type is no longer available.

AZIMUTHAL MODE NUMBER AUTOMATICALLY TRANSFERRED TO

REVOLUTION DATASET
When using circumferential mode extension for 2D axisymmetry, the azimuthal mode
number (for example, solid.mk) is now automatically transfered to the revolution
datasets that are created as default plots. This means that you do not need to modify
the settings in the dataset when switching between results for different mode numbers.

IMPROVEMENTS IN RANDOM VIBRATION FUNCTIONALITY

There are several improvements within the field of random vibration analysis:

• The number of study steps generated by a Random Vibration (PSD) study sequence
is now two instead of three. The “dummy” Frequency Domain study now resides
inside the Model Reduction study.
• A new operator, q2sq, has been added. It is similar to the existing q2 operator but
has more convenient arguments and output. In particular, it is targeted at
computing the RMS of the von Mises stress and vector norms.
• In the Random Vibration node, it is possible to enter default parameters for operators
like rms, which perform an integration. It is then possible to invoke these operators
without the last three arguments.

MEMS MODULE | 119

 • It is now easy to add the base acceleration to computed acceleration PSDs, thus
facilitating direct evaluation of absolute as well as relative spectra.
• There are significant performance improvements in the random vibration
computations.

CHANGED NOMENCLATURE FOR FORCE LINEAR STRAIN

The label of the Force linear strain check box, which is present in many material models,
has been changed to Geometrically linear formulation in order to improve clarity.

CHANGES IN DEFAULT PLOTS

There are many improvement in the default plots. Some of the more prominent are:

• The default stress plot in 3D now uses a Volume Plot, rather than a Surface Plot.
There are several advantages, including easier selection of individual domains and
more consistent evaluation of some expressions defined at the domain level.
However, in larger models, you may find that the plotting is slower. Also, the
visualization of curved boundaries with a coarse mesh is smoother in a surface plot.
• A new Prism color table is used for the default stress plots. This color table will
emphasize regions with high stresses.
• In the default stress plots, the smoothing is conditional, so that stress jumps larger
than 20% between adjacent elements will be plotted without smoothing. This makes
it easier to identify a solution computed on a too coarse mesh.

EASIER USE OF MULTIPLE CELL PERIODICITY NODES

The way studies are generated has been changed so that unique studies are created for
each Cell Periodicity node. This makes it possible to, for example, run nested
homogenizations.

A new section, Advanced, has been added to the settings for Cell Periodicity. Here you
can, if needed, select to use the old scheme to keep the old behavior.

SYMMETRY PLANE FOR ELECTROSTATICS AND ELECTRIC CURRENTS

A new Symmetry Plane feature is available for the Electrostatics and Electric Currents
interfaces. It provides symmetry and antisymmetry conditions for the electric field. For
the antisymmetric case, a reference potential can be set to specify around what electric
potential the field is antisymmetric (the default is ground).

120 | CHAPTER 1: RELEASE NOTES

IMPROVED HANDLING OF LOAD GROUPS FOR THERMAL EXPANSION AND
HYGROSCOPIC SWELLING
The handling of load groups in the Thermal Expansion and Hygroscopic Swelling features
has been improved. In particular, using several such features with the same selection
now works in a more intuitive way.

REMOVAL OF SEVERAL PAIR FEATURES

As part of a major overhaul of the pair handling, several pair features have been
removed. These nodes were seldom or never used, and the interpretation was difficult
and error prone. The removed pair features are Fixed Constraint, Prescribed
Displacement, Roller, Symmetry, Antisymmetry, Rigid Connector, and Boundary Load. In
case any such feature was used in a model, it can easily be replaced by standard
boundary conditions.

CHANGE IN PRESCRIBED DISPLACEMENT WHEN ELASTIC

PREDEFORMATION IS USED
The interpretation of the input in the Prescribed Displacement node in connection with
Elastic Predeformation has been changed. It is now the total displacement, which is
given in the same way as if Elastic Predeformation is not present.

CHANGED API FOR INITIAL VALUES

Due to changes in the implementation of the Initial Values node in Solid Mechanics,
there are changes in the API syntax. In version 5.6 and earlier, if the displacement field
had a name that was not u, the field name was reflected in the API syntax when setting
an initial value. Also, the syntax for the initial velocity has changed. Assume that the
field name is u_solid:

Earlier versions:
...feature("init1").set("u_solid", new int[]{1, 0, 0});
...feature("init1").set("du_solid/dt", new int[]{2, 0, 0});

Version 6.0:
...feature("init1").set("u", new int[]{1, 0, 0});
...feature("init1").set("ut", new int[]{2, 0, 0});

If you are reading a file containing the old syntax, there is an automatic conversion for
some common cases. The cases that are handled include when the field name is u* or
u_solid*, where * is a number less than or equal to 10.

For all other cases when a nondefault field name is used, you must manually replace
the argument in the call to set() by the new names, u or ut, to set the initial
displacements or velocities.

MEMS MODULE | 121

 BUG FIXES
The Compliance matrix and Elasticity matrix of the piezoelectric material
Polyvinylidene fluoride (PVDF) has been corrected to be consistent with the Voigt
notation (as of 5.6 update 1).

The formulation for rarefied (general accommodation) thin-film flow has been
corrected (as of 5.6 update 2).

New Models in Version 6.0

SOLIDLY MOUNTED RESONATOR 2D

A solidly mounted resonator (SMR) is a piezoelectric MEMS resonator formed on top
of an acoustic mirror stack deposited on a thick substrate. This tutorial shows how to
simulate an SMR in 2D. In this example, the eigenmodes were computed and the
frequency response from 500 to 1200 MHz was analyzed. The computed impedance
curve and displacement profile agree reasonably well with the simulation results found
in literature.

SOLIDLY MOUNTED RESONATOR 3D

A solidly mounted resonator (SMR) is a piezoelectric MEMS resonator formed on top
of an acoustic mirror stack deposited on a thick substrate. This tutorial shows how to
simulate an SMR in 3D. In this example, the eigenmodes were computed with a
varying number of particles attached to the sensor surface for computing the
sensitivity, and the corresponding change in frequency response was analyzed. The
resonant frequency decreases with more attached particles, and the sensitivity depends
on the attachment location relative to the mode shape — both observations are as
expected.

COMPONENT MODE SYNTHESIS TUTORIAL

In this tutorial example, the concept of component mode synthesis (CMS) is
introduced through a simple solid model of a beam. Parts of the beam are reduced into
CMS components. Dynamic as well as stationary analyses are performed. Moreover,
you will learn how to apply loads directly on a CMS component and work with such
components during result evaluation. The results from the full and reduced versions of
the model are compared.

MODEL UPDATES
The biased_resonator_3d_freq model has been updated to use the same AC drive
voltage as in the reference paper.

122 | CHAPTER 1: RELEASE NOTES

The comb_drive_tuning_fork_gyroscope_manufacturing_variation model has been
updated to use Component>Deformed Geometry branch features.

MEMS MODULE | 123

 Metal Processing Module
New Functionality in Version 6.0

IMPORT OF PHASE TRANSFORMATION DATA

It is now possible to import phase transformation data from JMatPro® for steel
quenching purposes. Phase transformation data is imported for both diffusive and
diffusionless phase transformations. The appropriate phase transformation models are
configured automatically, and parameters for start and finish temperatures, latent heat,
and phase transformation strains are also imported.

NEW PHASE TRANSFORMATION MODEL

The Kirkaldy-Venugopalan phase transformation model has been added to the Phase
Transformation node in the Metal Phase Transformation and Austenite Decomposition
interfaces. This phase transformation model is mainly suitable for diffusive phase
transformations during steel quenching.

MISCELLANEOUS
• The Leblond-Devaux, Johnson-Mehl-Avrami-Kolmogorov (JMAK), and Kirkaldy-
Venugopalan phase transformation models can use TTT diagram data directly, as input.
• The Koistinen-Marburger phase transformation model can now be defined in terms
of the Martensite start temperature and Martensite finish temperature.
• In 0D, it is now possible to compute the time at which a phase transformation has
stopped. This effectively corresponds to the time at which the rate of formation of
the destination phase falls below a lower threshold. You can model this by using the
Compute time to completion option in a Metallurgical Phase node, where the
numerical lower threshold, the Completion rate, is additionally supplied.

124 | CHAPTER 1: RELEASE NOTES

Microfluidics Module
New Functionality in Version 6.0
• A Fluid-Solid Interaction, Fixed Geometry interface for Two-Phase Flow, Phase
Field has been added under the Fluid-Structure Interaction section in the Fluid
Flow branch.
• The Brinkman Equations, Free and Porous Media Flow and Darcy’s Law interfaces
have a new Porous Medium feature, which replaces the old Fluid and Matrix Properties
feature. The new feature contains two subfeatures, Porous Matrix and Fluid, in which
the characteristic properties can be specified directly or by linking to a Material
node.
• The Brinkman Equations and Free and Porous Media Flow interfaces have a new
option, Porous treatment of no slip condition, with two options, Standard no slip
formulation and Porous slip. The latter is suitable for models with underresolved
boundary layers.

Backward Compatibility with Version 5.5

In the Level Set and Phase Field interfaces, the Initial Interface feature is now obsolete.
It does not need to be replaced, since the phase interfaces are automatically detected
from the selection of the Initial Values features and the initial distribution of the level
set or phase field function is solved for in the Phase Initialization study step.

MICROFLUIDICS MODULE | 125

 Mixer Module
New Functionality in Version 6.0

ROTATING MACHINERY, PHASE TRANSPORT MIXTURE MODEL

The predefined Rotating Machinery, Phase Transport Mixture Model interfaces for
laminar and turbulent flow of multiple dispersed phases have been added under the
Multiphase Flow section in the Fluid Flow branch. These interfaces can be used to
simulate phase separation; fractionation by density, size, or shape; and the mixing of
several phases that would otherwise separate due to sedimentation or flotation
processes.

126 | CHAPTER 1: RELEASE NOTES

Multibody Dynamics Module
New Functionality in Version 6.0

IMPROVED RIGID BODY CONTACT

The Rigid Body Contact feature has been extended so that more shapes are available in
3D. The source geometry can be a sphere or cylinder. The destination geometry can
be a sphere, cylinder, planar shape, or user-defined arbitrary shape.

In both 2D and 3D, the geometrical parameters of the standard shapes can now be
automatically determined from the geometry of a selected rigid domain.

COMPONENT MODE SYNTHESIS

Linear components built using the Solid Mechanics and Multibody Dynamics
interfaces can be reduced to computationally efficient reduced-order models using the
Craig–Bampton method. A reduced component can then be used in dynamic or
stationary analyses, either in a model consisting of only reduced components or
together with ordinary full finite-element models. The latter can then be nonlinear.

To create such components, you add a Reduced Flexible Components node in the physics
interface.

POINT LOADS AT ARBITRARY LOCATIONS

With the new Point Load, Free feature, you can apply point loads at arbitrary locations
that do not coincide with a geometrical point or a mesh node. This is particularly useful
when working with moving loads.

IMPROVED MIXED FORMULATION

When using a mixed formulation in the Linear Elastic Material, you can now modify the
discretization for the auxiliary dependent variable (pressure or volumetric strain). To
do so, Advanced Physics Options must be selected. Then, a new section named
Discretization will appear in the settings for Linear Elastic Material as soon as a mixed
formulation is selected. In this section, you can choose between different types of
shape functions for the auxiliary dependent variable.

Another change is that the auxiliary dependent variable now has feature scope rather
than physics scope.

MULTIBODY DYNAMICS MODULE | 127

 NEW DAMPING MODELS
Two new damping models have been added for the Linear Elastic Material:

• The Wave attenuation model is essentially a viscous model, but with parameters
given by measured data for the attenuation of elastic waves in the material.
• The Maximum loss factor model is mainly intended for time-domain analysis of
materials for which a loss factor representation provides a good description in the
frequency domain.

The new damping models can be used in both frequency-domain and time-domain
studies.

REDUCED INTEGRATION
A new framework has been added for reduced integration. It is controlled from the
Quadrature Settings section in the Linear Elastic Material node’s settings.

EASIER GAUSS POINT EVALUATION IN DOMAINS

A set of physics-scoped operators have been added as tools for easier evaluation of
Gauss point results in domains. The operators automatically adapt to the Gauss point
pattern in each element. These operators are:

• <phys>.gpeval(expr): Map a field of suitable order to Gauss point data.

• <phys>.elemgpmax(expr): Maximum value in any Gauss point in an element.
• <phys>.elemgpmin(expr): Minimum value in any Gauss point in an element.
• <phys>.elemavg(expr): Average value of the Gauss points in an element.
• <phys>.elemavg(expr): Integration over an element using Gauss point data.

SYMMETRIC FORMULATION FOR RIGID CONNECTOR AND RIGID DOMAIN

Under geometric nonlinearity, the consistent stiffness and mass matrices generated by
the Rigid Connector and Rigid Domain features may become nonsymmetric. The solution
of a nonsymmetric problem requires significantly more memory and computation time
than solving a symmetric problem of the same size. In version 6.0, you can choose to
use a less accurate symmetric matrix instead. Doing so may require some more
iterations, but each iteration will be significantly faster.

In order to use the symmetric formulation, select the Use symmetric formulation for
geometric nonlinearity check box in the Formulation section in the settings for the
feature.

128 | CHAPTER 1: RELEASE NOTES

COORDINATE SYSTEM TYPES RESTRICTED FOR RIGID CONNECTOR
For a rigid connector, it is not meaningful to select a coordinate system with position-
dependent axis orientations. This is no longer allowed, neither in the Rigid Connector
node itself nor in its subnodes.

EXPRESSIONS FOR THE ROTATION VECTOR CHANGED IN RIGID

CONNECTOR AND RIGID DOMAIN
The expression defining the variables giving the rotation vector in a rigid connector
and rigid domain has been changed in such a way that it is more robust with respect
to computing Jacobian contributions. This can be important if you want to, for
example, control such a variable using other equations.

The affected variables for a rigid connector are named <phys>.th<dir>_<tag> (for
example, solid.thz_rig1). For a rigid domain, the names are of the type
<phys>.<tag>.th<dir> (for example, solid.rd1.thy).

CONSTRAINT GROUPS FULLY SUPPORTED FOR RIGID CONNECTOR AND

RIGID DOMAIN
In earlier versions, constraint groups could not be used for rigid connectors or rigid
domains if the Evaluate reaction forces option was selected. This limitation has been
removed.

STRICTER DEFAULT TOLERANCE FOR TIME-DEPENDENT ANALYSIS

The default relative tolerance for time-dependent analysis has been decreased from
0.01 to 0.001 for the Multibody Dynamics interface. This will affect both new and old
models in which the solver sequence has not been manually edited. The effect is that
solutions will be more accurate but typically at the price of longer solution times.

THE TIME-DEPENDENT, PRESTRESSED, MODAL STUDY TYPE REMOVED

Due to problems with obtaining the correct linearization in all situations, the Time-
Dependent, Prestressed, Modal study type is no longer available.

CHANGED NOMENCLATURE FOR FORCE LINEAR STRAIN

The Force linear strain check box, which is present in Linear Elastic Material, has been
changed to Geometrically linear formulation in order to improve clarity.

IMPROVED HANDLING OF LOAD GROUPS FOR THERMAL EXPANSION

The handling of load groups in Thermal Expansion has been improved. In particular,
using several such features with the same selection now works in a more intuitive way.

MULTIBODY DYNAMICS MODULE | 129

 REMOVAL OF SEVERAL PAIR FEATURES
As part of a major overhaul of the pair handling, several pair features have been
removed. These nodes were seldom or never used, and the interpretation was difficult
and error prone. The removed pair features are Fixed Constraint, Prescribed
Displacement, Roller, Rigid Connector, and Boundary Load. In case any such feature was
used in a model, it can easily be replaced by standard boundary conditions.

New Models in Version 6.0

DYNAMICS OF A ROLLER CONVEYOR

This model simulates the dynamics of a roller conveyor belt transporting a spherical
object or a ball. The ball is placed at one end of a roller conveyor and then transported
to the other end, where it is collected in a rectangular tray. The conveyor system
consists of cylindrical rollers inserted between two frames. Guideways are provided on
both sides of the conveyor to prevent the ball from falling off. All the components of
the system are assumed to be rigid. The rollers are assumed to be connected to a power
source to rotate at a prescribed speed, whereas the frame, guideways, and tray are fixed.

The contact between the ball and rollers is modeled using rigid body contact with
friction. Similar contact conditions are also used between the ball and other
components. The connection between the rollers and frame is simplified using hinge
joints. A transient study is performed to analyze the system dynamics and also find the
contact forces, friction forces, and energy dissipation due to friction.

COMPONENT MODE SYNTHESIS TUTORIAL

In this tutorial example, the concept of component mode synthesis (CMS) is
introduced through a simple solid model of a beam. Parts of the beam are reduced into
CMS components. Dynamic as well as stationary analyses are performed. Moreover,
you will learn how to apply loads directly on a CMS component and work with such
components during result evaluation. The results from the full and reduced versions of
the model are compared.

MODELING VIBRATION AND NOISE IN A GEARBOX: CMS VERSION

This example illustrates the modeling of vibration and noise in a 5-speed synchromesh
gearbox of a manual transmission vehicle. The example is essentially the same as
Modeling Vibration and Noise in a Gearbox but with some elastic components
reduced to computationally efficient lower-order models using the component mode
synthesis (CMS) technique.

130 | CHAPTER 1: RELEASE NOTES

Nonlinear Structural Materials
Module
New Functionality in Version 6.0

NEW HYPERELASTIC MATERIAL MODEL: VAN DER WAALS

A new hyperelastic material model has been added, called van der Waals. It is based
on thermodynamical principles for polymer chains. This material model is also known
as the Kilian model.

FIBER-REINFORCED HYPERELASTIC AND NONLINEAR ELASTIC

MATERIALS
By adding one or more Fiber subnodes under a Hyperelastic Material or Nonlinear Elastic
Material node, you can augment the stiffness by the effect of distributed fibers. For
hyperelastic materials, there are three material models for the fibers: Holzapfel-Gasser-
Ogden, Linear elastic, and User defined. The fiber content is assumed to be a small
fraction of the total material volume. The new Fiber node makes it easy to model
anisotropic hyperelastic materials in general.

For nonlinear elastic materials, the only material model for the fibers is linear elastic.

You can model thermal expansion of the fibers by adding a Thermal Expansion subnode
under the Fiber node.

ALTERNATIVE INPUT PROPERTIES FOR SOME HYPERELASTIC MODELS

In the Neo-Hookean and St. Venant–Kirchhoff hyperelastic material models, you can
now enter the constitutive parameters on several different forms, for example, by
Young’s modulus and Poisson’s ratio.

NEW VARIANT OF MULLINS EFFECT

When using Ogden-Roxburgh as a Damage function in the Mullins Effect node, it is now
possible to choose between two different interpolation functions for the damage. In
the original version, a tanh function is used. By selecting the Use error function check
box, you can replace it with the erf function.

NONLINEAR STRUCTURAL MATERIALS MODULE | 131

 NEW DEFAULT VALUE FOR BULK MODULUS IN HYPERELASTIC
MATERIALS
In hyperelastic materials that are of a nearly incompressible type, the bulk modulus κ
now has the default value 100*phys.Gequ. The built-in variable phys.Gequ is the
shear modulus at zero strain. In earlier versions, the default value was zero; a value that
always had to be replaced by something reasonable.

NONLOCAL PLASTICITY
Nonlocal plasticity has been added to the Plasticity and Porous Plasticity features. You
can enable and control it in the new Nonlocal Plasticity Model section in the settings for
these material models.

With nonlocal plasticity, the equivalent plastic strain is a distributed field, rather than
defined at a point. This makes it possible to solve problems that would otherwise fail
due to high strain localization.

POROUS PLASTICITY ENHANCEMENTS

There are several enhancements to the porous plasticity material model:

• The hardening variable in the matrix (<item>.epm) is constrained to be positive.

• When the effective void volume function reaches its maximum value, a small
stiffness now remains, thus avoiding a breakdown of the analysis. You can control
the Maximum damage from the Advanced section in the Porous Plasticity node.
• In the Gurson–Tvergaard–Needleman model, the effective void volume fraction
function fe can now be selected and has the options Bilinear, Asymptotic, and User
defined.
• It is possible to account for void growth in shear in 2D axisymmetry.
• Gauss point interpolated variables have been added for more accurate result
evaluation.

CREEP IMPROVEMENTS
The functionality in the Creep node has been subject to major changes:

• A new general framework gives large improvements in computational speed and

memory usage. For larger models, a speedup by a factor of 10 or more can be
expected.
• You can select the type of scheme that is used for the time integration of the creep
equations. In the Time Stepping section, you can set Method to Automatic, Backward
Euler, Forward Euler, or Domain ODEs. The last method provides backward

132 | CHAPTER 1: RELEASE NOTES

 compatibility with versions prior to 6.0 and is usually significantly slower for large
models. The Backward Euler method is not available in the Layered Shell interface
or in the layered material models in the Membrane and Shell interfaces.
• It is no longer possible to add several Creep nodes acting in parallel on the same
selection. When more than one creep mechanism is acting, you can instead add one
or more Additional Creep nodes under a Creep node. As an effect, only one set of
creep strain degrees of freedom are needed, even if several creep models are used.
This gives additional performance gains for this type of model.
• The Norton-Bailey creep model has been removed. It is replaced by a general
isotropic “Bailey-type” hardening formulation, which can be combined with all
creep models, not only the Norton model.
• The type of equivalent stress used to determine the creep rate is now a user input.
You can select von Mises, Hill orthotropic, Pressure, or User defined. Thus,
nonisotropic creep can now be modeled.
• The old creep models Potential, Volumetric, and Deviatoric have been removed. The
same functionality is now achieved either through the User defined creep model or
by using the new Inelastic Strain Rate node.

PRESCRIBED INELASTIC STRAIN RATE

A new subnode, Inelastic Strain Rate, has been added to Linear Elastic Material, Layered
Linear Elastic Material, Nonlinear Elastic Material, Hyperelastic Material, and Layered
Hyperelastic Material. The effect of this node is similar to the existing External Strain
node, in that it injects an inelastic strain in the material model. The difference is that
here you instead specify a rate, which is then integrated into an inelastic strain
contribution. By using this feature, you do not need to add an extra ODE interface
when you have a rate equation describing an extra strain contribution.

The strain rate can be specified in terms of a strain tensor, a deformation gradient, an
inverse deformation gradient, or three orthogonal stretches.

UPDATE OF ANAND VISCOPLASTIC MODEL

The Anand model under Viscoplasticity has been updated. The main change is with
respect to the time integration. You can now select the type of scheme that is used. In
the Time Stepping section, you can set Method to Automatic, Backward Euler, or Domain
ODEs. The last method provides backward compatibility with versions prior to 6.0 and
is usually significantly slower for large models. This section is not present in the
Layered Shell interface or in the layered material models in the Membrane and Shell
interfaces.

NONLINEAR STRUCTURAL MATERIALS MODULE | 133

 As an effect of these changes, there are some potential backward compatibility issues:

• API code that references the old field names in order to modify solver settings will
fail.
• Manual variable scaling of the viscoplastic degrees of freedom will be lost, even
when Domain ODEs is selected.

DAMPING FOR SHAPE MEMORY ALLOY

In a dynamic analysis, you can now use a Damping subnode under Shape Memory Alloy
in both the Solid Mechanics and Truss interfaces.

ENHANCEMENTS IN PHASE-FIELD-BASED DAMAGE

When using the Damage node, there are two new options for the Crack driving force:
Total strain energy density and User defined. The old option, Strain energy density, has
been renamed to Elastic strain energy density for the sake of clarity.

IMPROVED PERFORMANCE FOR LARGE-STRAIN PLASTICITY

Large-strain plasticity is up to 20% faster than in earlier versions because a more
efficient formulation is used.

REDUCED INTEGRATION
The new framework for reduced integration will, in many cases, make the solution of
models with nonlinear materials, such as plasticity and creep, significantly faster.
Reduced integration is controlled from the Quadrature Settings section in material
models such as Linear Elastic Material, Nonlinear Elastic Material, Hyperelastic Material,
and Shape Memory Alloy.

New Models in Version 6.0

CREEP ANALYSIS OF A TURBINE STATOR BLADE

In this example, the deformations caused by secondary creep in a turbine stator blade
are computed. The creep rate is highly influenced by temperature, and the deformation
and stress relaxation is thus to a large extent controlled by the temperature field, which
is computed.

PHARMACEUTICAL TABLETING PROCESS

Powder compaction is a popular manufacturing process not only in powder metallurgy,
but also in the pharmaceutical industry. The capped Drucker–Prager model is
commonly used for simulating the compaction processes of pharmaceutical powders,

134 | CHAPTER 1: RELEASE NOTES

where the material properties depend on the powder density. In this example,
microcrystalline cellulose (MCC) is compacted, and the constitutive material
properties are calibrated based on experimental data.

TWO-STAGE POWDER COMPACTION PROCESS

Powder compaction is a key process in powder metallurgy, where it gives the flexibility
to produce quality products of complex shapes for sintering. The density of the
compact is a key factor to determine the overall quality of the sintered product, as
regions with lower density could reduce its mechanical strength.

A multistage compaction process is one way to achieve homogeneous density in the

workpiece. In this example, a two-stage compaction process is demonstrated in a
simple setup, and the outcome is compared with the results of a single-stage process.

BIVENTRICULAR CARDIAC MODEL

The standard biventricular cardiac model of the heart is used to show how to set up
fiber directions in a complex geometry. The fibers are then used to model the large
deformation of the myocardium with the Holzapfel–Gasser–Ogden anisotropic
material model. The Aliev–Panfilov equations are included to model the electric
cardiac excitation.

NONLINEAR STRUCTURAL MATERIALS MODULE | 135

 Optimization Module
New Functionality in Version 6.0

EASIER ACCESS TO SHAPE AND TOPOLOGY OPTIMIZATION

The shape and topology optimization features have been moved up the model tree. In
version 5.6, they were available from the Component node’s Definitions branch, but in
6.0, they reside directly under the Component (similar to physics interfaces).

It is now also possible to add a Control Variable Field directly under the Component
node, so in many cases, it is no longer necessary to use the Optimization interface.

PARAMETER ESTIMATION IMPROVEMENTS

The user interface for parameter estimation has been improved with the addition of a
new Global Least-Squares Objective feature (available under the Component node) and a
new Parameter Estimation study step. The old Parameter Estimation study step has
been renamed to Curve Fitting. For most parameter estimation models, this means that
it is no longer necessary to use the Optimization interface.

SENSITIVITY ANALYSIS IMPROVEMENTS

There have been several improvements to the robustness of transient sensitivity
analysis, and it is now possible to force solving of the adjoint problem using the same
time steps as the forward problem. This setting is available in the Optimization Solver
node. Oftentimes, robustness can be further improved by the use of fixed time steps.

Additional Improvements
A Control Function feature has been added, which can be used with arbitrary arguments.
One possibility is to use it for introducing a scalar transient control variable.
Regularization using a Bernstein polynomial or a Helmholtz filter is supported by this
feature.

The Transformation shape optimization feature has been introduced to simplify simple
operations such as translation, rotation, and scaling. This feature interacts with the
other shape optimization features, and it also supports scaling and translation in
custom coordinate systems.

136 | CHAPTER 1: RELEASE NOTES

New and Updated Models in Version 6.0
The following new models are now available:

• The coil_shape_optimization model demonstrates how to find a coil geometry

giving a uniform magnetic field on an axis and minimal field near where the axis
ends. This model requires the AC/DC Module.
• The optimal_heating_control model provides an example of optimal control,
where the input heating power to a rod is optimized to give a certain temperature
on the outside of the rod. This problem corresponds to the situation where you
want to reach a certain minimum temperature as fast as possible all over, without
exceeding some maximum temperature.
• The rectangular_horn_shape_optimization model demonstrates how to use the
Control Function feature to perform shape optimization on a rectangular horn for
an improved on-axis response. The off-axis spatial response is improved using the
Free Shape Boundary and Transformation shape optimization features. This model
requires the Acoustics Module and either the AC/DC Module, Battery Design
Module, MEMS Module, Plasma Module, RF Module, or Semiconductor Module.
• The tesla_microvalve_transient_optimization model performs a topological
optimization for a Tesla microvalve for an oscillating pressure drop. A Tesla
microvalve inhibits backward flow using friction forces rather than moving parts,
and therefore the objective is to maximize the average flow rate. The design can be
optimized by distributing material within the modeling domain.
• In the waveguide_filter_optimization_transformation model, a waveguide filter
is designed using shape optimization by moving and scaling rectangles in the
geometry. The irises of the initial geometry are optimized to ensure good bandpass
response and out-of-band rejection, while maintaining the double mirror symmetry.
This model requires the RF Module.

Backward Compatibility with Version 5.4

The computed solution in models using the Nelder-Mead, Coordinate search, or
Monte-Carlo method should not change with Enforce design constraints strictly
enabled, compared to the solution obtained in the previous version. The number of
optimization iterations should either decrease or remain unchanged depending on the
number of infeasible iterations. The computed solution in models using COBYLA can
differ with the Enforce design constraints strictly enabled.

OPTIMIZATION MODULE | 137

 Particle Tracing Module
New Functionality in Version 6.0

SIMPLIFIED NONLOCAL COUPLING NAMES

The built-in couplings that evaluate the sum, average, maximum, and minimum of
expressions over all particles have been renamed. In an instance of the Mathematical
Particle Tracing interface with the default name pt, the old name for the sum was
pt.ptop1(); the new name is pt.sum(). Similarly, the average, maximum, and
minimum are now pt.ave, pt.max, and pt.min, respectively.

Backward Compatibility
The old coupling names, such as pt.ptop1() in the example above, will be labeled as
deprecated but they are still supported in version 6.0.

TRANSFORMATIONS OF LOADED PARTICLE COORDINATES

When loading initial particle position and velocity from a file using the Release from
Data File node, the distribution of initial particle positions can now be dilated, rotated,
and translated. If a rotation is applied, you can also choose to apply the same rotation
to the initial particle velocity vectors if they are also loaded from the same file.

FAST BOUNDARY SEARCH FOR DRAG IN WALL-BOUNDED FLOWS

In the Drag Force for Particle Tracing for Fluid Flow interface, when you select the
Include wall corrections check box to define a more accurate drag force in narrow pipes
or channels, you can now specify the Mesh search method used to compute the
direction and distance from each particle to the closest wall. Similar functionality is
used to detect the closest wall when activating anisotropic turbulent dispersion in the
Drag Force or to find the distance to the closest wall for the wall-induced Lift Force.

The default option in the Mesh search method list, Closest point, is the behavior from
versions 5.6 and earlier. It is the slowest but most robust option. The Use tolerance and
Walk in connected component options may be significantly faster in finely meshed 3D
geometries.

ACCUMULATED DOSE FOR PARTICLE-MATTER INTERACTIONS

In the Particle-Matter Interactions node, you can now select any combination of the
Absorbed dose, Absorbed dose from ionization losses, or Absorbed dose from nuclear
stopping check boxes. Selecting a check box causes an accumulated variable to be

138 | CHAPTER 1: RELEASE NOTES

defined in the selected domains to compute the accumulated dose and dose equivalent
from the passage of energetic particles (usually ions) through the medium.

DISSIPATED PARTICLE HEAT

You can use the new Dissipated Particle Heat node to accumulate a heat source term
based on the heating or cooling of particles by their surroundings. To use this node,
the particle temperature must be solved for, and at least one Convective Heat Losses or
Radiative Heat Losses node must be present. A particle that is heated by its surroundings
will define a volumetric heat sink, while a particle that is cooled by its surroundings will
define a volumetric heat source.

New and Updated Applications in Version 6.0

PINCHED FLOW FRACTIONATION

This tutorial model demonstrates the continuous separation of particles by size via
purely inertial means. Particles enter a microchannel through one inlet while pure
liquid enters the same microchannel through a second inlet at a much higher velocity.
When the two streams are joined, the distribution of particles is pushed against the wall
of the microchannel opposite the high-velocity stream. In close proximity to the
microchannel wall, the wall-induced lift force separates out the different particle sizes.
Then, at the other end of the channel, particles exit through different outlet branches
based on their relative size.

THREE-BODY PROBLEM
The classical three-body problem involves the prediction of the trajectories of three
point masses under mutual gravitational attraction. This example compares two known
periodic solutions to the three-body problem: the stable figure-eight configuration and
the unstable Lagrange configuration.

Important Fixes in Version 6.0

ACCURACY IMPROVEMENTS FOR BOUNDARY ACCUMULATORS

Significantly improved the accuracy of accumulated variables defined on curved
boundaries.

PARTICLE TRACING MODULE | 139

 CHARGE ACCUMULATION WITH MULTIPLE SPECIES
In the Particle Tracing for Fluid Flow interface, using the Charge Accumulation node
with two or more different particle species no longer fails with an “Undefined value
found” error message.

NUMBER OF PARTICLE DIAMETER VALUES PARSING ISSUE

When computing particle diameters in the Particle Tracing for Fluid Flow interface,
most particle release features allow you to sample from a normal, lognormal, or
uniform distribution of initial diameter values. In version 5.6, entering an expression
or a parameter for the Number of values field resulted in the error message “Field is not
an integer” even if an integer value was given. This has been fixed in version 6.0.

VOLUME FORCE IN THE INERTIALESS FORMULATION

In the Particle Tracing for Fluid Flow interface, using the Newtonian, ignore inertial
terms formulation together with the Volume Force Calculation node caused an
“Undefined variable” error in version 5.6. This has been fixed in version 6.0.

WALL CORRECTIONS AFFECTING THE SOLUTION WHILE DISABLED

In the Particle Tracing for Fluid Flow interface, the Drag Force node can be used to
model drag in wall-bounded flows by selecting the Include wall corrections check box.
This check box is only supported for the Stokes drag law. However, in version 5.6, it
was possible to define drag corrections in other drag laws by selecting this check box
and then switching to a different drag law. This has been fixed in version 6.0.

WALL CORRECTIONS FROM PAIR CONTINUITY BOUNDARIES

In the Particle Tracing for Fluid Flow interface, the Drag Force node can be used to
model drag in wall-bounded flows by selecting the Include wall corrections check box.
In version 5.6, interior boundaries of an assembly where the Pair Continuity boundary
condition was used would count as walls for the purpose of applying such wall
corrections. In version 6.0, Pair Continuity boundaries will be ignored when computing
wall corrections for the drag force.

THERMAL RE-EMISSION ON INTERIOR MOVING BOUNDARIES

In version 5.6, it was possible for the Thermal Re-Emission boundary condition to reflect
particles on the wrong side of interior boundaries if those boundaries were moving at
speeds comparable to or faster than the thermal velocity of the particles. This has been
corrected in version 6.0.

140 | CHAPTER 1: RELEASE NOTES

CONFLICT WITH MULTIPLE PARTICLE-MATTER INTERACTIONS NODES
In the Charged Particle Tracing interface, variables defined by multiple copies of the
Particle-Matter Interactions node will no longer overwrite each other.

ERROR MESSAGES FROM INTERFACES NOT SOLVED FOR

In version 5.6, an error message was given when trying to run any study if a particle
tracing interface in the model had no particle release features, or if the particle release
features had no selection. In version 6.0, such error messages will be ignored when
running a study in which the particle tracing interface is not solved for.

Backward Compatibility with Version 5.5

PARTICLE PROPERTIES FROM MATERIAL

In the Particle Tracing for Fluid Flow interface, the option to take particle properties
such as density From material was added in version 5.6. When opening a model built
in version 5.5 or earlier, User defined will be selected from the list instead of From
material in the settings for the Particle Properties node.

SAMPLING OF AUXILIARY DOFS FROM UNIFORM DISTRIBUTIONS

When a particle tracing model includes an Auxiliary Dependent Variable, or in the
Particle Tracing for Fluid Flow interface when solving for particle mass or diameter,
you have the option of sampling the initial values of these extra variables from a Uniform
distribution. In versions 5.5 and earlier, the mean and standard deviation were
specified. In versions 5.6 and later, the maximum and minimum are specified.

When opening a model built in version 5.5 or earlier, the text fields for the maximum
and minimum values in uniform distributions of auxiliary variables are automatically
defined to give the same result as in the older model:

New minimum = (Old mean) - (Old standard deviation)*sqrt(3)

New maximum = (Old mean) + (Old standard deviation)*sqrt(3)

TEMPERATURES MOVED TO MODEL INPUT SECTION

In the Convective Heat Losses and Radiative Heat Losses nodes (available for the Particle
Tracing for Fluid Flow interface when the Compute particle temperature check box is
selected), the input for the ambient temperature has been moved to the Model Input
section.

PARTICLE TRACING MODULE | 141

 Backward Compatibility with Version 5.4

REMOVED THE PARTICLE TRAJECTORIES STUDY STEP

The Particle Trajectories study step is obsolete. All instances of this study step have been
replaced with Time Dependent study steps with the same settings.

142 | CHAPTER 1: RELEASE NOTES

P i pe Flo w M o d u l e
New Functionality in Version 6.0

NON-NEWTONIAN FLOW IN PIPES

A new Herschel–Bulkley fluid model is now available for the Pipe Flow interface. Using
the Herschel–Bulkley model, you can simulate flow of fluids exhibiting viscoplastic
behavior.

ADDITIONAL OPTION FOR T-JUNCTIONS

The Pipe Flow interface now includes an additional option for the T-junction
specification called Loss coefficient with respect to common branch. This option
automatically calculates the dimensionless loss coefficients accounting for a particular
flow situation, for example, if the flow in the junction is joining or separating.

Updated Models in Version 6.0

STRESS IN COOLING PIPELINE NETWORK

This example demonstrates how to model coupled flow, heat transfer, and structural
deformation and stress in a pipeline network. Gravity loads from the pipe and fluid are
also taken into account.

TOPOLOGY OPTIMIZATION OF A DISTRICT HEATING NETWORK

This example shows how to design the layout of a district heating network using
topology optimization. The model requires the Pipe Flow Module and the
Optimization Module.

PIPE FLOW MODULE | 143

 Plasma Module
New Functionality in Version 6.0

CONSISTENT STABILIZATION
Streamline diffusion for electrons and ions is available when the Formulation is set to
Finite element (linear shape function) and Finite element (quadratic shape function). This
stabilization technique makes it easier to solve plasma problems with a large drift
component like corona discharges and streamers.

INCONSISTENT STABILIZATION
Isotropic diffusion for electrons and ions is available when the Formulation is set to
Finite element (linear shape function) and Finite element (quadratic shape function). This
stabilization technique adds a diffusion component to the transport equations in order
to dampen the effect of oscillations.

STATIONARY AND FREQUENCY-STATIONARY STUDIES ARE AVAILABLE

FOR THE PLASMA INTERFACE
Stationary and frequency-stationary studies are available in the Plasma interface. These
types of studies increase computational speed for DC and ICP discharges, allows for
fast parameterization, and makes the use of optimization technology more practical.

TIME DEPENDENCY ADDED TO THE BOLTZMANN EQUATION IN THE

TWO-TERM APPROXIMATION
Time dependency was added to the isotropic part of the two-term expansion of the
Boltzmann equation in the Boltzmann Equation, Two-Term Approximation interface and
to the Plasma interface when using the Local Field Approximation. This new
functionality makes it possible to study cases in which the isotropic part of the electron
distribution function cannot be assumed to be stationary.

IMPROVEMENTS IN THE STABILITY OF THE EQUILIBRIUM DISCHARGE

INTERFACES
Several improvements were made in the equilibrium discharge interfaces to increase
their stability. In general, models solve smoother and faster. The improvements include
changes in the solver settings and how the fluid properties are interpolated and
extrapolated.

144 | CHAPTER 1: RELEASE NOTES

New and Updated Applications in Version 6.0

DC CORONA DISCHARGE IN AIR IN A POINT-TO-PLANE CONFIGURATION

This tutorial model presents a study of a negative corona discharge in dry air at
atmospheric pressure in a point-to-plane configuration. A millimeter-sized elliptical
electrode creates a high-intensity electric field where the corona discharge occurs. A
ground plane is placed 10 cm away from the corona electrode. The voltage is varied
between −5 and −50 kV. The creation and transport of charged species is solved self-
consistently coupled with Poisson’s equation. The local field approximation is used,
which means that transport and source coefficients are assumed to be well
parameterized through the reduced electric field.

DOUBLE-HEADED STREAMER IN NITROGEN

This tutorial model presents a study of a double-headed streamer in nitrogen at
atmospheric pressure. An initial seed of electrons is placed between two electrodes that
apply an initial electric field of 52 kV/cm to the gas. A negative and positive streamer
propagate toward the electrodes. The electron density, electric field, and the
propagation velocity of the streamer agree well with simulation results published by
Bessières and others.

PLASMA DC ARC
This example investigates the electrical and thermal characteristics of a plasma DC arc
created in a point-to-plane configuration. The discharge is assumed to be in local
thermodynamic equilibrium. The plasma is considered a conductive fluid medium and
is modeled using a magnetohydrodynamics approach. This model shows how to use
the Combined Inductive/DC Discharge interface to simulate the plasma generated in
a DC arc.

PLASMA-PULSED ARC
This example investigates the electrical and thermal characteristics of a plasma-pulsed
arc created in a point-to-plane configuration. The discharge is assumed to be in local
thermodynamic equilibrium. The plasma is considered a conductive fluid medium and
is modeled using a magnetohydrodynamics approach. This model shows how to use
the Combined Inductive/DC Discharge interface to simulate the plasma generated in
a pulsed arc.

PLASMA MODULE | 145

 COIL OPTIMIZATION OF AN ICP REACTOR
This model shows how shape optimization can be used to design the coils of an ICP
reactor to obtain plasma uniformity. The reactor in study is a planar ICP with the coils
distributed along the radial direction. The Optimization study step is used to find the
best coil placement so that the gradient of the electron density in the reactor center is
minimized. A target electron density in the reactor center is also imposed by varying
the coil power.

TRANSIENT NEGATIVE MOBILITY AND NEGATIVE DIFFERENTIAL

CONDUCTIVITY EFFECTS IN XENON
This tutorial model presents a study showing the transient negative mobility and the
negative differential conductivity effects in xenon. The stationary and time-dependent
Boltzmann equation in the two-term approximation is used to compute the electron
energy distribution function.

Updated Applications in Version 6.0

INDUCTIVELY COUPLED PLASMA (ICP) TORCH

This model before version 6.0 was solved using a time-dependent study. Now, it is
solved with a stationary solver that allows for much faster computation times.

Backward Compatibility with Version 5.5 and Earlier

There are new versions of the following interfaces: Plasma; Drift Diffusion; Heavy
Species Transport; Inductively Coupled Plasma; Microwave Plasma; and Boltzmann
Equation, Two-Term Approximation. When opening models created with version 5.5
and earlier, a warning sign appears saying “This physics interface is obsolete and will
be removed in a future version. A new version of the physics interface is available from
the Add Physics window”. These models can still be changed and will run as before,
but it is encouraged to recreate the models using version 5.6.

146 | CHAPTER 1: RELEASE NOTES

Porous Media Flow Module
New Functionality in Version 6.0

POROELASTICITY, LARGE DEFORMATION, SOLID MULTIPHYSICS

INTERFACE
The new Poroelasticity, Large Deformation, Solid multiphysics interface is used for
modeling poroelasticity under finite structural deformations. It is similar to the
existing Poroelasticity, Solid multiphysics interface but with an added Elastic
Predeformation node in the Solid Mechanics interface.

You can find this interface under the Poroelasticity branch in the Structural Mechanics
folder in the Add Physics tree. This new interface also requires the Structural Mechanics
Module.

IMPROVED HANDLING OF POROUS MATERIALS

Porous materials are now defined using phase-specific properties tabulated in the
Porous Material node. In addition, subnodes can be added for the solid and fluid
features. Several subnodes can be defined for each phase. This functionality makes it
possible to use one and the same porous material for fluid flow, chemical species
transport, and heat transfer without having to duplicate material properties and
settings.

IMPROVEMENTS TO BRINKMAN EQUATIONS INTERFACE

In the Brinkman Equations interface, the new Porous Medium feature now supports the
Porous Material functionality. When opening a model created in version 5.6 or earlier,
the old Fluid and Matrix Properties feature is not migrated, but a warning message is
prompted. It should be manually replaced by the new Porous Medium feature (which
adds a Porous Matrix and Fluid subfeatures) to allow an updated definition of the
material properties. Support for obsolete features is not ensured in future versions;
thus it is strongly recommended to transfer the old feature to the new feature.

POROUS SLIP IN THE BRINKMAN EQUATIONS INTERFACE

The boundary layer between free and porous media flow may be very thin and
impractical to resolve. The new porous-slip wall treatment makes it possible to account
for porous walls without resolving the full flow profile in the boundary layer. Instead,
a stress condition is applied at the porous surfaces by utilizing an asymptotic solution,

POROUS MEDIA FLOW MODULE | 147

 yielding a good accuracy in the bulk flow. The functionality is activated in the new
Porous treatment of no slip condition settings, and it is then used as the default wall
condition. This new feature can be used in most problems where the model domain is
large.

TWO-PHASE FLOW IN POROUS MEDIA

The new Brinkman Equations multiphysics interface combines the Brinkman Equations
and Level Set interfaces and automatically adds a Two-Phase Flow, Level Set coupling
node. It solves the conservation of momentum and the continuity of mass with the
Brinkman equations. The front between two immiscible fluids in porous media is
tracked with the level-set function.

You can find this interface under the Multiphase Flow>Two-Phase Flow, Level Set branch
in the Fluid Flow folder in the Add Physics tree.

NONISOTHERMAL FLOW IN POROUS MEDIA

The new Nonisothermal Flow, Brinkman Equations multiphysics interface automatically
adds the coupling between heat transfer and fluid flow in porous media. It combines
the Heat Transfer in Porous Media and Brinkman Equations interfaces and automatically
adds a Nonisothermal Flow coupling node.

You can find this interface under the Nonisothermal Flow branch in the Fluid Flow folder
in the Add Physics tree.

EXTENDED FEATURES FOR HEAT TRANSFER IN POROUS MEDIA

The Porous Medium domain feature now includes three different models, which
correspond to the thermal equilibrium, thermal nonequilibrium assumptions, and the
configuration of packed beds. These models can be selected from the new Porous
medium type list in the Porous Medium section:

• The default Porous Medium feature in the Heat Transfer in Porous Media interface
by default selects the Local thermal equilibrium option under the Porous medium type
list. This corresponds to the previously available averaging techniques within the
Porous Medium feature, which solved a single equation for the average temperature
field with effective properties. New averaging techniques are available for
computing the Effective thermal conductivity in porous media: Solid spherical
inclusions, Fluid spherical inclusions, Wrapped screen, and Sintered metal fibers. In
addition, the Volume average and Reciprocal average techniques have been renamed
to Plane layers parallel to heat flow and Plane layers perpendicular to heat flow,

148 | CHAPTER 1: RELEASE NOTES

 respectively, for clarity. Finally, using the new Equivalent thermal conductivity option,
you can enter user-defined expressions.
• When the new Local Thermal Nonequilibrium interface is added, the Porous
medium type list is set to Local thermal nonequilibrium. This setting replaces the
previously available Local Thermal Nonequilibrium multiphysics interface (which
included the Heat Transfer in Solids interface, Heat Transfer in Fluids interface, and
Local Thermal Nonequilibrium multiphysics coupling node). The settings for the
fluid and solid materials are available in the Fluid and Solid subnodes of the Porous
Medium feature, which provides the setting of the Interstitial convective heat transfer
coefficient (previously available in the multiphysics coupling node). The Local
Thermal Nonequilibrium Boundary feature is automatically added to surrounding
domains where the fluid and solid temperatures are not in equilibrium.
• When the new Heat Transfer in Packed Beds interface is added, the Porous medium
type list is set to Packed bed. This option corresponds to a new multiscale modeling
approach tailored for packed beds.

For these three Porous medium type options, the postprocessing variables (thermal
properties, heat fluxes, heat sources, and heat and energy balance) are available in a
unified way for homogenized quantities.

MULTISCALE HEAT TRANSFER IN PACKED BEDS

The new Heat Transfer in Packed Beds interface is useful to model packed beds made
of spherical homogenized porous particles in which the temperature varies radially.
The temperature distribution in the pellets is computed for every position in the bed,
and it is coupled to the temperature in the surrounding fluid through an interstitial
heat flux or a temperature continuity condition between the pellets’ surface and the
fluid.

In this configuration, the Porous Medium feature has three subnodes (Fluid, Pellets, and
Pellet-Fluid Interface), and a Local Thermal Nonequilibrium Boundary feature is
automatically added to surrounding domains where the fluid and solid temperatures
are not in equilibrium. The new Pellet subnode, available under the Porous Material
node, can be used to define the properties of the pellet and fluid.

You can find this interface under the Porous Media branch in the Heat Transfer folder in
the Add Physics tree.

EXTENDED FEATURES FOR MOISTURE TRANSPORT

The Moisture Flow multiphysics coupling has a new option to account for the leakage
velocity at walls where evaporation and condensation take place. When the Mixture

POROUS MEDIA FLOW MODULE | 149

 type for moist air is set to Concentrated species in the Moisture Transport interface, the
Account for Stefan velocity at walls check box is available in the Moisture Transport at
Walls section. This functionality is available for free and porous media, and it
automatically adds the corresponding contribution to the fluid flow velocity.

In the Moisture Transport interface, the Hygroscopic Porous Medium feature has been
updated to match the new design for features using porous materials. It now contains
the Liquid Water, Moist Air, and Porous Matrix subnodes for the settings of each phase.

In addition, the Moisture Transport interface has extended functionality for several
boundary conditions.

Several postprocessing variables are now available to check mass balance. In particular,
the mt.massBalance, mt.dwcInt, mt.ntfluxInt, and mt.GInt predefined variables
make it possible to check the mass balance, the total accumulated moisture rate, the
total net moisture rate, and the total mass source, respectively.

MISCELLANEOUS
• An efficient formulation of the sensible enthalpy predefined variable, ht.DeltaH, is
now available for free and porous media containing moist air. This analytical
formulation, taking advantage of the polynomial definition of the heat capacity for
moist air, drastically reduces the evaluation time of the accumulated heat and energy
rates in postprocessing.
• In the Moist Porous Medium feature of the Heat Transfer interface, the definition of
the effective properties has been fixed to include the liquid water contribution.
• The Longitudinal dispersivity and Transverse dispersivity settings in the Thermal
Dispersion feature are now available with all the products that include the feature.

Backward Compatibility with Version 5.6

• The Local Thermal Nonequilibrium predefined multiphysics interface and the Local
Thermal Nonequilibrium multiphysics coupling are obsolete. These features are not
migrated when opening a model created in version 5.6 or earlier. They should be
manually replaced by the new Porous Medium feature, with the Porous medium type
set to Local thermal nonequilibrium. Support for features marked as obsolete is not
ensured in future versions, so it is strongly recommended to transfer the old features
to the new feature.
• In the Moisture Transport interface, the old Hygroscopic Porous Medium feature not
supporting the Porous Material functionality is now obsolete. This feature is not

150 | CHAPTER 1: RELEASE NOTES

migrated when opening a model created in version 5.6 or earlier. It should be
manually replaced with the new Hygroscopic Porous Medium feature. Support for
features marked as obsolete is not ensured in future versions, so it is strongly
recommended to transfer the old feature to the new feature.

POROUS MEDIA FLOW MODULE | 151

 Polymer Flow Module
New Functionality in Version 6.0

NEW VISCOELASTIC MODEL

The exponential Phan–Thien Tanner viscoelastic model (EPTT) has been added in the
Viscoelastic Flow interface.

SISKO MODEL FOR INELASTIC NON-NEWTONIAN MODEL

The new Sisko inelastic model has been added to model the flow of non-Newtonian
fluids at moderate and high shear rates.

CURING REACTION INTERFACE

The new Curing Reaction interface has been added to model the crosslinking of
thermosetting polymers. The interface includes predefined models for the curing
reaction rate and models for correlating the viscosity and degree of cure.

FLUID-STRUCTURE INTERACTION FOR VISCOELASTIC FLUIDS

New multiphysics interfaces — Fluid-Solid Interaction, Viscoelastic Flow; Fluid-Solid
Interaction, Viscoelastic Flow; and Fixed Geometry — have been added to facilitate a
convenient setup of models combining flow of viscoelastic fluids and deformations of
solids. The Fluid-Solid Interaction, Viscoelastic Flow interface also requires a license
for either the Structural Mechanics Module, MEMS Module, or Multibody Dynamics
Module.

LICENSE UPDATE FOR BRINKMAN EQUATIONS AND FREE AND POROUS

MEDIA FLOW INTERFACES
The Brinkman Equations and Free and Porous Media Flow interfaces are now available
in the Polymer Flow Module.

TWO-PHASE FLOW IN POROUS MEDIA

A new multiphysics interface combines the Brinkman Equations and the Level Set
interfaces and automatically adds a Two-Phase Flow, Level Set coupling node. It solves
for conservation of momentum and mass with the Brinkman equations. The front
between two immiscible fluids in porous media is tracked with the level set function.

152 | CHAPTER 1: RELEASE NOTES

New Models and Apps in Version 6.0

PASTA EXTRUSION
This example shows how to simulate the nonisothermal flow of the dough in the
metering zone of the pasta extruder. To account for shear thinning behavior of the
dough, the power law non-Newtonian model with temperature-dependent parameters
is used.

VISCOELASTIC FLOW THROUGH A CHANNEL WITH A FLEXIBLE WALL

This model demonstrates how to use the new Fluid-Solid Interaction, Viscoelastic
Flow interface. This example models the flow of a viscoelastic fluid in a channel in
which part of the upper wall is replaced by an elastic plate subjected to an external
pressure.

POLYMER FLOW MODULE | 153

 Ray Optics Module
New Functionality in Version 6.0

OPTICAL LIBRARY IMPROVEMENTS

The Optical Material Library has been significantly improved in version 6.0. Glasses
from CDGM Glass Co., Ltd.; Corning Incorporated; Ohara Inc.; and Schott AG now
include a more complete set of material properties. In addition to optical dispersion
coefficients or tabulated values of the refractive index, many of these materials now
include the internal transmittance, thermal conductivity, specific heat capacity, Young’s
modulus, Poisson’s ratio, and coefficient of thermal expansion. This extension of the
Optical material library greatly facilitates coupled structural-thermal-optical
performance (STOP) analysis.

SIMPLIFIED NONLOCAL COUPLING NAMES

The built-in couplings that evaluate the sum, average, maximum, and minimum of
expressions over all rays have been renamed. In an instance of the Geometrical Optics
interface with the default name gop, the old name for the sum was gop.gopop1(); the
new name is gop.sum(). Similarly, the average, maximum, and minimum are now
gop.ave, gop.max, and gop.min, respectively.

Backward Compatibility
The old coupling names, such as gop.gopop1() in the example above, will be labeled
as deprecated, but they are still supported in version 6.0.

GAUSSIAN BEAM RAY RELEASE FEATURE

In a Geometrical Optics model that solves for ray intensity or power, you can add the
Gaussian Beam node to release rays with a Gaussian intensity or power distribution.

Because geometric ray tracing does not consider diffraction in the vicinity of the beam
waist, the Gaussian Beam node is appropriate only when the wavefront normals can be
approximated as following straight lines. This is a reasonable assumption when the
model geometry is much larger than the Rayleigh range of the beam, in which case the
rays can be seen to emanate out from a point with angle-dependent intensity or power;
or occasionally when the Rayleigh range is much larger than the geometry, in which
case the released beam is collimated and the intensity or power is a Gaussian function
of radial distance from the beam axis.

154 | CHAPTER 1: RELEASE NOTES

BLACKBODY RADIATION RAY RELEASE FEATURE
You can use the Blackbody Radiation node to release rays diffusely from a surface, with
initial ray intensity or power based on the surface temperature. If tracing of
polychromatic light has been enabled in the Geometrical Optics interface settings,
then the wavelength or frequency of the released rays is sampled from the Planck
distribution.

The more general-purpose ray release features, such as Release from Grid and Release
from Boundary, now also support sampling the initial wavelength or frequency from the
Planck distribution, but for these release features, the power or intensity must be
explicitly given; it is not automatically derived from a surface temperature.

NEW WAYS TO ASSIGN A WEIGHTED INTENSITY DISTRIBUTION

When releasing a distribution of rays from a point, using a node such as Release from
Grid, you can now choose to assign a Weighted distribution of initial intensity or power.
The sum of the initial power over all released rays will still add up correctly to the
specified total source power, but the intensity and power of individual rays can be a
function of initial position or direction.

TRANSFORMATIONS OF LOADED RAY COORDINATES

When loading initial ray position and direction from a file using the Release from Data
File node, the distribution of initial ray positions can now be dilated, rotated, and
translated. If a rotation is applied, you can also choose to apply the same rotation to
the initial ray directions if they are also loaded from the same file.

EASIER POSTPROCESSING OF REFRACTIVE INDEX

Built-in postprocessing variables allow you to easily visualize the refractive index of
materials at the d-line, F-line, and C-line; as well as the Abbe number. Usually, the
refractive indices of dispersive materials can only be plotted on discrete rays because
they are expressed as functions of the ray frequency or vacuum wavelength.

NEW WAYS TO DEFINE ABSORBING MEDIA

There are several new ways to define the material properties of absorbing media in the
Geometrical Optics interface. The default behavior is to specify the imaginary part of
the refractive index directly. Alternatively, you can specify the intensity attenuation
coefficient or the internal transmittance for a material sample of a certain thickness.

RAY OPTICS MODULE | 155

 RELEASE FROM ELECTRIC FIELD IMPROVEMENT
When releasing rays from a boundary with the Release from Electric Field node, which
initializes the ray Stokes parameters based on the electric field in an adjacent domain,
you can now choose to initialize the ray direction using the Poynting vector of the
adjacent field. In previous versions, rays could only be released normal to the
boundary.

IMPROVED HANDLING OF CURVED GRATINGS

The interpretation of diffraction gratings on curved surfaces has been fine-tuned in
version 6.0. You can now control whether the grating constant should be interpreted
as the distance between unit cells of the grating along the curved surface, or as the
projected distance between unit cells on a tangent plane.

New Applications in Version 6.0

PETZVAL LENS OPTIMIZATION

This example extends the basic Petzval Lens tutorial by demonstrating how an optical
prescription can be modified when introducing a replacement glass. To maintain the
same level of image quality using the replacement glass, small adjustments to the radii
of curvature of the lenses are required. The changes in the lens radii of curvature were
computed using the BOBYQA gradient-free optimization method. The perturbations
in the lens radii of curvature were used as control parameters, while the global
objective function was the sum of the root mean square spot size at three different field
angles.

MICROLITHOGRAPHY LENS
In this example, geometrical optics simulation is performed on a system of 21 fused
silica spherical lens elements for use in deep ultraviolet (DUV) photolithography. The
spot diagram shows excellent image quality over a 23.4 mm image circle.

ROWLAND CIRCLE SPECTROMETER

The Rowland circle is a circle of radius R that lies tangent to a concave curved
diffraction grating of radius 2R. When light enters through an arrow slit located
anywhere along the Rowland circle and then hits the grating, the reflected rays of each
diffraction order will be focused at different points along the same circle. For nonzero
diffraction orders, the rays are also sorted by wavelength.

156 | CHAPTER 1: RELEASE NOTES

Important Fixes in Version 6.0

WAVEFRONT CURVATURE REINITIALIZATION BY GRATINGS

When solving for ray intensity, the reinitialization of the wavefront radii of curvature
has been improved when releasing rays of nonzero diffraction order at the grating
surface.

SCATTERING BOUNDARY CONDITION IMPROVEMENTS

When solving for ray intensity or power, the Scattering Boundary node shows
significantly improved energy conservation in version 6.0 compared to version 5.6.
This is because the ray splitting algorithm that produces reflected and refracted rays at
the rough surface now uses the same value of the random perturbation to the surface
normal to compute the Fresnel coefficients of the reflected ray as it does for the
refracted ray.

ACCURACY IMPROVEMENTS FOR BOUNDARY ACCUMULATORS

The accuracy of accumulated variables defined on curved boundaries has been
significantly improved.

COUNT REFLECTIONS CHECK BOX NO LONGER CAUSES AN ERROR

Previously, an error message was given if the intensity formulation for graded media
was selected in a 3D model and the Count reflections check box was selected. In version
6.0, these features can be used together without causing an error message to appear.

UNIT SYNTAX IN PART LIBRARIES

Some of the parts in the Part Libraries for the Ray Optics Module have been updated
to use proper unit syntax in their definitions. This makes the parts more robust when
toggling between different unit systems.

ERROR MESSAGES FROM INTERFACES NOT SOLVED FOR

In version 5.6, an error message was given when trying to run any study if an instance
of the Geometrical Optics interface had no ray release features, or if the ray release
features had no selection. In version 6.0, such error messages will be ignored when
running a study in which the Geometrical Optics interface is not solved for.

IMPROVED PHASE REINITIALIZATION FOR POLARIZED RAYS

A stability improvement has been made to the reinitialization of phase when rays are
reflected or refracted at boundaries. The phase variable is solved for when ray intensity

RAY OPTICS MODULE | 157

 is solved for and the Compute phase check box is selected in the Settings window for
the Geometrical Optics interface.

Backward Compatibility with Version 5.5

NEW OPTIONS FOR SPECIFYING REFRACTIVE INDEX

Beginning in version 5.6, in the settings for the Medium Properties node you can either
specify the absolute refractive index or the refractive index relative to air at a given
temperature and pressure. If you selected Specify refractive index from the Optical
Dispersion Model list in version 5.5 or earlier, then in a later version Specify absolute
refractive index will be selected from the Refractive index of domains list.

SAMPLING OF AUXILIARY DOFS FROM UNIFORM DISTRIBUTIONS

When a ray optics model includes an Auxiliary Dependent Variable, or when
polychromatic light is released with a distribution of frequency or vacuum wavelength,
you have the option of sampling the initial values of these extra variables from a Uniform
distribution. In versions 5.5 and earlier, the mean and standard deviation were
specified. In versions 5.6 and later, the maximum and minimum are specified.

When opening a model built in version 5.5 or earlier, the text fields for the maximum
and minimum values in uniform distributions of auxiliary variables are automatically
defined to give the same result as in the older model:

New minimum = (Old mean) - (Old standard deviation)*sqrt(3)

New maximum = (Old mean) + (Old standard deviation)*sqrt(3)

ILLUMINATED SURFACE SETTINGS

Many of the settings for the Illuminated Surface ray release feature have been renamed.
However, in older models, the values of the variables with the old names will be
mapped to the variables with the new names. In addition, the old names can still be
used in model methods in the Application Builder or in Model Java® files.

Backward Compatibility with Version 5.4

HANDLING OF GRATING ORIENTATION DIRECTION

The handling of grating orientation was made more robust and consistent in version
5.5 compared to previous versions. Occasionally, this means that the directions
associated with positive and negative diffraction orders will appear reversed when
opening an older model in version 5.5 or newer.

158 | CHAPTER 1: RELEASE NOTES

To assist in the reassignment of any grating orientation directions, the direction
associated with positive diffraction orders will be indicated by an arrow in the Graphics
window.

RENAMED RAY RELEASE FEATURES

In version 5.5 the Inlet feature was renamed to Release from Boundary and the Inlet on
Axis feature was renamed to Release from Symmetry Axis.

RENAMED RAY STATISTICS

Since version 5.5 some variables for the ray statistics have been renamed. Here, <phys>
is the physics identifier and <feat> is the tag of a ray release feature.

• <phys>.rmaxall has been renamed to <phys>.rmax.

• <phys>.rmidrms has been renamed to <phys>.rmrms.
• <phys>.<feat>.rmaxrel has been renamed to <phys>.<feat>.rmax.
• <phys>.<feat>.rmidrms has been renamed to <phys>.<feat>.rmrms.

RAY OPTICS MODULE | 159

 RF Module
New Functionality in Version 6.0

ELECTROMAGNETIC WAVES, BOUNDARY ELEMENTS

This new physics interface is for modeling the scattering properties of objects,
evaluating electric fields far from the scatterer or the far fields of an antenna placed on
an electrically large platform.

LAYERED TRANSITION BOUNDARY CONDITION

Multiple thin layers, such as the gold-plated copper of a circuit board trace or close-to-
normal incidence on an antireflection coating to an optical lens, can be described by
the Layered Transition Boundary Condition feature. It requires combining this physics
feature, Layered Material in the global Materials, and Layered Material Link in the
Materials node.

THREE-PORT NETWORK FEATURE

This feature characterizes the response of a three-port network component using S-
parameters. Import a Touchstone file to describe the physical behavior and response
of a three-port device or system through three-port boundaries without addressing a
complicated geometry.

B-H CURVE MAGNETIC FIELD CONSTITUTIVE RELATION

The B-H curve option is available in the magnetic field constitutive relation for
modeling nonlinear magnetic phenomena. The material properties from the Nonlinear
Magnetic material library, available with the RF Module, can be used to relate the
magnetic field and magnetic flux density.

REFINE CONDUCTIVE EDGES

This new physics-controlled mesh option quickly identifies the exterior edges of signal
path boundaries, configured by perfect electric conductors or transition boundary
conditions, and applies a user-specified mesh size.

SHIFTED LAPLACE CONTRIBUTION ON MULTIGRID LEVELS

If no geometrical feature sizes are smaller than half a wavelength, when the operating
frequency is high, modeling with a higher-order element such as cubic element

160 | CHAPTER 1: RELEASE NOTES

discretization is beneficial for faster computation. The computation efficiency can be
further boosted by imposing the Shifted Laplace contribution on multigrid levels.

SUGGESTED ITERATIVE SOLVER FOR PERIODIC STRUCTURES

Typical periodic problems are solved with a direct solver. However, the direct solver
consumes a lot of memory, when the periodic unit cell size is not subwavelength. In
this case, switch to the Suggested Iterative Solver to finish the computation faster with
less memory usage.

CONSTRAINT-FREE PORT OPTION

The constraint-free port option is available to calculate the expansion coefficients as an
overlap integral, while in the default port formulation, the expansion coefficients (or
S-parameters) are calculated by adding a scalar dependent variable for each coefficient
and then adding a constraint to enforce the series expansion.

FREQUENCY DOMAIN, RF ADAPTIVE MESH STUDY

The dedicated study step for mesh adaptation makes the workflow much easier when
setting up mesh adaptation for modeling microwave and millimeter-wave antennas and
circuits. This dedicated study step automatically provides a physics-controlled
functional, goal-oriented termination expressions, and other settings. It requires the
use of linear element discretization in the physics interface. The adapted mesh can be
used for further frequency-domain simulations to characterize the device under test.

REFLECTION COEFFICIENT WITH MULTIPLE EXCITATION

When exciting all ports as in a phased antenna array, it is possible to compute the
reflection coefficient at each excited port that includes the impedance mismatching as
well as coupling by adjacent active ports.

SYMMETRY AXIS REFERENCE POINT

The Symmetry Axis Reference Point subnode is available in 2D axisymmetry, when there
is an incident field defined in the parent node. Then this subnode is available as a
default subnode to the Scattering Boundary Condition parent node. The Symmetry Axis
Reference Point subnode defines a reference position at the intersection point between
the parent node’s boundary selection and the symmetry axis.

DEFAULT PLOTS FOR NUMERIC PORT MODE FIELDS

Plots are automatically created of the port mode fields for numeric ports.

RF MODULE | 161
 New and Updated Models in Version 6.0

MODELING OF A CPW USING NUMERIC TEM PORTS

This model shows how to set up a coplanar waveguide (CPW) using numeric ports
analyzed as a TEM option, under the assumption that the ports mode fields are close
to transverse electromagnetic (TEM).

MODELING OF A GROUNDED CPW USING NUMERIC TEM PORTS

This model shows how to set up a grounded coplanar waveguide (GCPW), using
TEM-like numeric ports. The edges of metalized vias are included on the numeric port
boundaries to capture a proper mode propagating through the GCPW circuit.

FEM-BEM COUPLING OF A MICROSTRIP PATCH ANTENNA

This example shows how to couple the finite element method (FEM), analyzing a
microstrip patch antenna, to the boundary element method (BEM), for evaluating the
field outside the FEM computational domain. The model computes the S-parameter,
near-field distribution, and far-field radiation pattern through the FEM and the electric
fields outside the near-field air domain sphere with the BEM.

CPW RESONATOR FOR CIRCUIT QUANTUM ELECTRODYNAMICS

This model demonstrates a transmission line resonator, one of the main components
of CPWs. The resonator can be built from CPW transmission lines, terminated with a
combination of open and short ends. These ends create a resonator out of a CPW, with
the open and short ends functioning as zero current and zero voltage boundary
conditions, respectively.

WI-FI BOOSTER YAGI-UDA ANTENNA

This models demonstrates how to design a conventional Yagi–Uda antenna for the
2.4−2.5 GHz band. The antenna can direct the signal in a particular direction and
prevent intruders from eavesdropping on personal Wi-Fi signals.

RAT-RACE COUPLER
The rat-race coupler model has been updated to show how to use the layered transition
boundary condition — a new physics feature — to model two thin metal layers as a
boundary condition.

162 | CHAPTER 1: RELEASE NOTES

Backward Compatibility with Version 5.6 and Earlier
The port mode fields are now defined as the outgoing waves, whereas in version 5.6
and earlier the port mode fields were defined as the incoming (exciting) waves. This
change only affects User defined and Periodic port types because input fields can be
provided for these types. When loading MPH-files from version 5.6 or earlier, the
input fields for User defined and Periodic ports are migrated to function with the new
mode field definition. However, for User defined ports, the migration assumes that the
material adjacent to the port is homogeneous.

RF MODULE | 163
 Rotordynamics Module
New Functionality in Version 6.0

NEW INTERFACE: SOLID ROTOR, FIXED FRAME

A new physics interface, Solid Rotor, Fixed Frame, has been added. Its functionality is
essentially the same as in the existing Solid Rotor interface, but the rotor must be
axially symmetric. The difference is that it considers the rotor from a fixed frame rather
than from a corotating frame, just like the Beam Rotor interface.

In long rotors, there can be mixed whirling modes that are difficult to interpret in a
corotational frame. This is one example where you would prefer the new interface over
the existing Solid Rotor interface.

LIQUID ANNULAR SEAL

A new feature, Liquid Annular Seal, has been added to all rotordynamics interfaces. It
can be used to model the effects of the pressure drop across the seal, as well as swirling
of the fluid at the entrance of the seals. Both these effects affect the stability
characteristics of the rotor.

TURBULENCE AND SURFACE ROUGHNESS IN HYDRODYNAMIC THRUST

BEARING
The Hydrodynamic Thrust Bearing feature in the Hydrodynamic Bearing interface has
been enhanced by the possibility to take turbulence and surface roughness into
account. You can control this from the new Contact Surface Property section in the
settings for this node.

REMOVAL OF NONLINEAR MATERIAL MODELS FROM SOLID ROTOR

The possibility to model plasticity, creep, and viscoelasticity in the Solid Rotor interface
has been discontinued. This capability seems never to have been used and would block
other important developments.

REDUCED INTEGRATION
A new framework has been added for reduced integration. It is now available in the
Solid Rotor and Solid Rotor, Fixed Frame interfaces. Reduced integration is controlled
from the Quadrature Settings section in Linear Elastic Material.

164 | CHAPTER 1: RELEASE NOTES

COMPONENT MODE SYNTHESIS
Linear components built using the Solid Mechanics and Multibody Dynamics
interfaces can be reduced to computationally efficient reduced-order models using the
Craig–Bampton method. A reduced component can then be used in dynamic or
stationary analyses (for example, as foundations for a rotor).

New Models in Version 6.0

COMPARISON OF CAMPBELL PLOTS USING DIFFERENT ROTOR

INTERFACES
Different types of elements can be used for modeling a rotor, depending on the level
of complexity and the type of the system being modeled. The modeling steps and
representation of the results will vary with the type of idealization. In this tutorial
model, an eigenfrequency analysis is performed on a stepped rotor, using three
different physics interfaces for rotordynamics: Solid Rotor; Solid Rotor, Fixed Frame;
and Beam Rotor. The resulting Campbell plots from these interfaces are compared
with each other. The model also helps in understanding the different steps involved
when using each interface.

RESPONSE OF A ROTOR UNDER INFLUENCE OF SEAL FORCES

The fluid flow in seals causes a strong restoring radial force, opposing the shaft motion.
Also, there is a cross-coupled force due to the axial variation of the circumferential
flow. The first type of force always has a stabilizing effect on the rotor, whereas the
second one in certain situations can cause an instability.

In this example, an axial compressor is modeled using the Beam Rotor interface. The
compressor has 10 stages of impellers, with a seal near each impeller to avoid flow
leakage. In addition, a balance piston seal is present at the end of the impeller stages.
The time-dependent response of the system is studied for a gradually increasing rotor
speed. Two cases are considered for the simulation: excluding all seals and including
the seals. The responses of the system for the two cases are compared. The result shows
a better stability in the presence of seals.

ROTORDYNAMICS MODULE | 165

 Semiconductor Module
New Functionality in Version 6.0

TRANSITION BETWEEN DISCRETE LEVELS

A new feature of transition between discrete trap levels has been added to the three
trap-assisted recombination features (domain, boundary, and heterointerface) when
the explicit trap option is selected and more than one discrete trap level is created. This
makes it possible to simulate the transition between the trap levels and to simulate the
transition between the quantized levels of a quantum well and a quantum dot.

You can view this new feature in the A Solar Cell with InAs Quantum Dots Embedded
in AlGaAs/GaAs Quantum Wells tutorial model.

CONTACT RESISTANCE
A new Contact resistance option has been added to the Metal Contact boundary
condition for both ohmic and Schottky contact types and all five driving modes:
voltage, current, power, circuit current, and circuit voltage.

This feature can be enabled by selecting the Contact resistance check box (disabled by
default) in the Settings window.

The A Cross-Bridge Kelvin Resistor Model for the Extraction of Specific Contact
Resistivity tutorial model shows this new option.

KLAASSEN UNIFIED MOBILITY MODEL (LIC)

The Semiconductor interface now includes a Klaassen Unified Mobility Model (LIC)
feature (sometimes referred to as Philips Unified Mobility Model) available under the
Semiconductor Material Model node. In this model, the total carrier mobility is given
by combining the lattice (L), donor (I), acceptor (I), and carrier-carrier (C) scattering
effects. Screening of the impurities by charge carriers and clustering of impurities at
high doping levels are also included. Both the Trench-Gate IGBT 3D and Trench-
Gate IGBT 2D tutorial models use this new mobility model.

HETEROJUNCTION HEAT SOURCE

The Joule heating contribution to the boundary heat sources at heterojunctions has
been included in the built-in variables of the Semiconductor interface. This makes it
easier to perform coupled thermal analysis of heterostructures.

166 | CHAPTER 1: RELEASE NOTES

DEFAULT SOLVER
A new physics suggested solver sequence has been added to the Semiconductor
interface to streamline the study setup for models with field-dependent mobility or
impact ionization generation. These existing Application Library models are now more
efficient with the new solver sequence: mosfet_mobility, lombardi_surface_mobility,
caughey_thomas_mobility, and insb_pfet_density_gradient.

TRAPPING FEATURES
A new option of user-defined additional electron and hole capture rates has been added
to the three discrete trap level features (domain, boundary, and heterointerface) when
the explicit trap option is selected.

The continuous trap distribution feature has been enhanced by expanding the range
of the trap energy levels outside of the band gap and allowing the capture probability
to depend on the trap energy level.

The trapping feature for the density-gradient formulation has been enhanced with the
new option of solving for the trap occupancy (as opposed to the trap Fermi level).

MOBILITY MODELS
The Caughey–Thomas Mobility Model has been enhanced with several new options
for the formulation of the driving forces. This adds versatility to the feature for users
interested in different types of driving forces.

METAL CONTACTS
A new built-in global variable for the average terminal current density has been added
for the Metal Contact boundary condition. This makes it convenient to obtain scaling
independent output quantities for comparison purposes.

CORRECTIONS AND BACKWARD COMPATIBILITY

• The default expression, description, and equation display for the user-defined
energy distribution of continuous trap levels have been updated.
• The selection of the Schrödinger–Poisson multiphysics coupling has been extended
to include domains of the Second Order Hamiltonian domain condition (in 5.6
update 1).
• The Caughey–Thomas Mobility Model and Lombardi Surface Mobility Model have
been improved for the finite volume formulation by applying the circumcenter
operator on the driving forces.
• A formulation issue of WKB tunneling feature for Schottky contacts has been fixed.

SEMICONDUCTOR MODULE | 167

 • The trapping Formulation list under the discrete and continuous trap energy level
subnodes has been moved to the parent trap-assisted recombination node (three
kinds: domain, boundary, and heterointerface) to simplify model setup.
• The unit of the default value for the Hole thermal velocity under the Carrier Capture
section for explicit traps has been corrected from m/s to cm/s.

New Models in Version 6.0

SURFACE-TRAP-INDUCED HYSTERESIS IN AN INAS NANOWIRE FET, A

DENSITY-GRADIENT ANALYSIS
This tutorial analyzes the hysteresis of the conductance-gate-voltage (G-Vg) curves of
an InAs nanowire FET, using the density-gradient theory to add the effect of quantum
confinement to the conventional drift-diffusion formulation, without a large increase
of computational costs. The hysteresis is caused by the dynamic charging effects of fast
and slow semiconductor-oxide interface traps of continuous energy distributions and
of both donor and acceptor types. The capture probability is modeled as thermally
activated with a barrier height that varies with the trap energy level. The qualitative
behavior and the order of magnitude of the computed G-Vg curves under various
voltage ramping conditions agree well with simulation and experiment results found in
the literature.

A CROSS-BRIDGE KELVIN RESISTOR MODEL FOR THE EXTRACTION OF

SPECIFIC CONTACT RESISTIVITY
This benchmark example builds two models of a cross-bridge Kelvin resistor used for
extracting the specific contact resistivity. The first model simulates the system in 3D,
using the contact resistance feature built in the Semiconductor interface. The other
model is a 2D approximation of the system developed in the reference paper,
implemented using a boundary PDE math interface. The computed Kelvin contact
resistance as a function of the contact window size and the diffusion tap width
correlates well between the two models and agrees well with the result shown in the
reference paper.

A SOLAR CELL WITH INAS QUANTUM DOTS EMBEDDED IN ALGAAS/GAAS

QUANTUM WELLS
This example shows an approximate approach to model a dot-in-well solar cell as
described by Asahi and others in the reference paper. The quantum wells and layers of
quantum dots are each treated as lumped energy levels in the band gap. The authors
specify transitions between the dot/well levels and the energy bands. The continuum

168 | CHAPTER 1: RELEASE NOTES

part of the current density is otherwise unimpeded by the wells and dots. This
description is equivalent to the trapping feature in the Semiconductor interface, so it
is used to model the wells and dots in this example. The computed trend of the
photocurrent and occupancy of the quantum dot states agrees well with the result
shown in the paper.

TRENCH-GATE IGBT 2D
In this first half of a two-part example, a 2D model of a trench-gate IGBT is built,
which will be extended to 3D in the second half. In general, it is the most efficient to
start with a 2D model to make sure everything works as expected before extending it
to 3D. The Caughey–Thomas mobility model is combined with the Klaassen unified
mobility model to account for velocity saturation and phonon, impurity, and carrier-
carrier scattering. The contact resistance option of metal contact boundary conditions
is used to implement the mixed-mode simulation with parasitic resistance at the
collector and emitter as mentioned in the reference paper. The computed collector
current density as a function of the collector voltage agrees reasonably well with the
published result.

TRENCH-GATE IGBT 3D
In this second half of a two-part example, a 3D model of a trench-gate IGBT is built
by extruding the 2D model from the first half. Unlike the 2D model, now it is possible
to arrange the alternating n+ and p+ emitters along the direction of extrusion as in the
real device. This more realistic arrangement leads to better quantitative agreement
with experimental data. The computed collector current density as a function of the
collector voltage agrees reasonably well with the published result.

IMPROVED MODELS
The two tutorial models MOSCAP 1D and MOSCAP 1D Small Signal have been
expanded to demonstrate how to compute the differential capacitance using metal
contacts where the terminal charge is not readily available.

The meshes for the two tutorial models Interface Trapping Effects of a MOSCAP and
MOSCAP 1D Small Signal have been improved for a better compromise between
discretization error and round-off error.

The mosfet_mobility, lombardi_surface_mobility, caughey_thomas_mobility, and

insb_pfet_density_gradient library models have been updated to take advantage of a
new autogenerated solver sequence.

SEMICONDUCTOR MODULE | 169

 The Thermal Analysis of a Bipolar Transistor model has been updated to use the
available temperature and heat source items in the lists for the temperature and heat
source couplings, as opposed to entering the corresponding variables manually.

170 | CHAPTER 1: RELEASE NOTES

Structural Mechanics Module
New Functionality in Version 6.0

MAGNETOMECHANICS MULTIPHYSICS INTERFACE

Two new physics interfaces for the analysis of coupled magnetic and mechanical effects
have been added: Magnetomechanics and Magnetomechanics, No Currents. When you add
such an interface, two physics interfaces are added to the model: Solid Mechanics and
either Magnetic Fields or Magnetic Fields, No Currents. The new Magnetomechanical
Forces multiphysics coupling is also added.

You can find these interfaces under the Electromagnetics-Structure Interaction branch in
the Structural Mechanics folder in the Add Physics tree.

This interface also requires the AC/DC Module.

POROELASTICITY, LARGE DEFORMATION, SOLID MULTIPHYSICS

INTERFACE
The new Poroelasticity, Large Deformation, Solid multiphysics interface is used for
modeling poroelasticity under finite structural deformations. It is similar to the
existing Poroelasticity, Solid multiphysics interface, but with an added Elastic
Predeformation node in the Solid Mechanics interface.

You can find this interface under the Poroelasticity branch in the Structural Mechanics
folder in the Add Physics tree.

This interface also requires the Porous Media Flow Module.

BEAM CROSS SECTION INTERFACE IN 3D

The Beam Cross Section interface is now available in 3D. In the 3D version, you have
the possibility to extrude the cross section and then show a full 3D representation of
the stresses in a straight beam.

As an effect of this addition, the 2D version of the interface has also been significantly
updated. One major change is that the interface can handle more than one cross
section. The selection of the cross section is thus no longer made at the interface level
but in a node named Homogeneous Cross Section. It is possible to have any number of
such nodes. The Hole node is thus no longer a top-level node but a subnode under
Homogeneous Cross Section. Because there now can be results for several sections in the

STRUCTURAL MECHANICS MODULE | 171

 same interface, the computed cross-section property variables now have feature scope
rather than physics scope. For example, a cross-section area that in previous versions
was named bcs.A will now be named bcs.hcs1.A. When opening an old model, it is
automatically converted to the new structure.

STRESS FROM INTERNAL PRESSURE IN BEAM CROSS SECTION

For use with the Pipe Mechanics interface, it is now possible to enter a pressure in the
Hole node in the Beam Cross Section interface. This will be used to compute the stress
contribution from an internal gauge pressure.

TRANSFER OF DATA BETWEEN BEAM CROSS SECTION AND BEAM TYPE

INTERFACES
Two new multiphysics couplings handling data transfer between the Beam Cross
Section and the Beam or Pipe Mechanics interfaces have been added.

• The Beam Cross Section - Beam Coupling multiphysics coupling transfers cross-section
data computed in the Beam Cross Section interface to the Beam or Pipe Mechanics
interface. It is available in 2D and 3D.
• The Beam - Beam Cross Section Coupling multiphysics coupling transfers section
forces computed in a Beam or Pipe Mechanics interface to a Beam Cross Section
interface for visualization. This coupling is only available in 3D.

MAGNETOSTRICTION COUPLED TO ROTATING MACHINERY, MAGNETIC

INTERFACE
The Magnetostriction multiphysics coupling can now also be used to connect the Solid
Mechanics and Rotating Machinery, Magnetic physics interfaces.

COMPONENT MODE SYNTHESIS

Linear components built using the Solid Mechanics and Multibody Dynamics
interfaces can be reduced to computationally efficient reduced-order models using the
Craig–Bampton method. A reduced component can then be used in dynamic or
stationary analyses, either in a model consisting of only reduced components, or
together with ordinary full FE models. The latter can then be nonlinear.

To create such components, you add a Reduced Flexible Components node in the physics
interface.

172 | CHAPTER 1: RELEASE NOTES

RIGID JOINT IN SOLID MECHANICS INTERFACE
A Rigid Joint node has been added to the Solid Mechanics interface. Such a joint can
be used to connect two attachments rigidly. The feature is similar to the same joint in
the Multibody Dynamics interface. The main purpose is for joining CMS components.

ATTACHMENTS NO LONGER REQUIRE MULTIBODY DYNAMICS LICENSE

You can now add the Attachment node with a Structural Mechanics Module or MEMS
Module license. The main usage in the context of creating and joining CMS
components.

LOADS ON SHELLS AT TOP OR BOTTOM SURFACE

It is now possible to apply loads in the Shell and Plate interfaces not only on the
midsurface but also on the top and bottom surfaces. Using the actual location in the
thickness direction can be important in the case of shells with significant thickness, in
particular when curved.

The thickness direction location is selected in the new Through-Thickness Location

section, available in the Face Load, Edge Load, Point Load, Point Load, Free, Ring Load
and Ring Load, Free nodes’ settings.

COUPLINGS BETWEEN DISJOINT SHELLS

Three new nodes have been added in the Shell interface, facilitating easier coupling
between parts of shells that are located so that there are gaps in the geometry. These
are: Edge to Edge, Edge to Boundary, and Boundary to Boundary. The couplings can be
rigid or elastic. Some applications are:

• Imported geometries where there are slits between parts.

• Midsurface generation having left slits between parts.
• Avoiding the extra flexibility caused by using a common edge in, for example, a T-
joint. The shell thickness can be taken into account in a more accurate way.
• Weld modeling, in which case you can use the flexible version of the coupling and
evaluate the forces in it.

POINT LOADS AT ARBITRARY LOCATIONS

With the new Point Load, Free and Ring Load, Free features, you can apply point loads
at arbitrary locations that do not coincide with a geometrical point or a mesh node.
This is particularly useful when working with moving loads.

This functionality is available in the Solid Mechanics, Shell, Plate, Membrane, Beam,
and Truss interfaces.

STRUCTURAL MECHANICS MODULE | 173

 NEW DEFAULT BEHAVIOR FOR CONTACT AND CONTINUITY
If any contact pairs are defined under Definitions, a default Contact node will be added
in the Solid Mechanics, Shell, and Membrane interfaces. Similarly, if there are any
identity pairs, a default Continuity node will be added. In many cases, the pair handling
can be considered as automatic because pairs can be generated during the finalization
of the geometry sequence.

As an effect of the general reformulation of the pair functionality, the Source external
to current physics check box is no longer needed and has been removed except in the
Layered Shell interface.

TWISTING DEGREE OF FREEDOM IN 2D AXISYMMETRY

In the Solid Mechanics interface in 2D axisymmetry, it is now possible to include
circumferential deformation. To do so, you select the Include circumferential
displacement check box in the Axial Symmetry Approximation section in the settings for
the physics interface.

INITIAL IMPERFECTION IN BUCKLING ANALYSIS

You can now use the buckling modes from a linear buckling analysis as initial
imperfections to the geometry when performing a full nonlinear buckling analysis.
This functionality is controlled from the new Buckling Imperfection node, which can be
added from Physics Utilities under Definitions.

COMPUTATION OF SECTION FORCES IN SOLIDS

By adding the new Section Forces node in Solid Mechanics, you can compute section
forces (axial force, shear forces, bending moments, and twisting moment) on a cross
section in a solid structure.

FIBER-REINFORCED LINEAR ELASTIC MATERIAL

By adding one or more Fiber subnodes under a Linear Elastic Material node, you can
augment the stiffness by the effect of distributed fibers. The fiber content is assumed
to be a small fraction of the total material volume.

You can model thermal expansion of the fibers by adding a Thermal Expansion subnode
under the Fiber node.

WRINKLING IN MEMBRANES
When a principal stress in a membrane is less than or equal to zero, it loses its stability.
Physically, this means that wrinkling will occur. This situation can now be handled by

174 | CHAPTER 1: RELEASE NOTES

adding the new Wrinkling subnode under Linear Elastic Material in the Membrane
interface.

VISCOELASTICITY IMPROVEMENTS
There are several important additions to the viscoelastic material models:

• For frequency-domain and time-dependent analyses, all of the viscoelasticity models

have been augmented with the possibility to include viscoelasticity in the volumetric
deformation.
• The Generalized Maxwell model now includes the possibility to prune branches
representing frequency ranges outside the bandwidth of prescribed loads, improving
the performance in time-dependent analysis for models with many viscoelastic
branches.
• For frequency-domain analyses, a new user-defined viscoelasticity model makes it
possible to enter frequency-dependent expressions for the loss and storage moduli
or compliances.
• Through a new formulation of the viscoelastic equations, it is now possible to solve
for eigenfrequencies in a structure containing viscoelastic materials using a standard
procedure for damped eigenfrequency problems. Previously, the eigenvalue
problem was nonlinear in the frequency, and only one eigenfrequency at a time
could be computed. The new approach is available for the all viscoelastic models
except user defined, under the assumption that fractional derivatives are not used.

IMPROVED MIXED FORMULATION

In material models that have an option to select a mixed formulation, you can now
modify the discretization for the auxiliary dependent variable (pressure or volumetric
strain). To do so, Advanced Physics Options must be selected. Then, a new section
named Discretization will appear in the settings for the material model as soon as a
mixed formulation is selected. In this section, you can choose between different types
of shape functions for the auxiliary dependent variable.

Another change is that the auxiliary dependent variable now has feature scope rather
than physics scope.

STRUCTURAL MECHANICS MODULE | 175

 NEW DAMPING MODELS
Two new damping models have been added for the mechanical material models:

• The Wave attenuation model is essentially a viscous model, but with parameters
given by measured data for the attenuation of elastic waves in the material. It is
available in the Linear Elastic Material in Solid Mechanics.
• The Maximum loss factor model is mainly intended for time-domain analysis of
materials for which a loss factor representation provides a good description in the
frequency domain. This damping model is available for all material models that
support viscous damping.

The new damping models can be used in both frequency-domain and time-domain
studies.

DAMPING FOR PIEZOELECTRIC MATERIALS IN TIME DOMAIN

Two new options have been added to the Dielectric Loss subnode under Piezoelectric
Material in the Solid Mechanics interface: Complex Permittivity intended for frequency-
domain analysis and Maximum loss factor intended for time-domain analysis.

MULTIPOLE DEBYE DISPERSION FOR PIEZOELECTRIC MATERIALS

New damping models have been added for piezoelectric materials. Under Charge
Conservation, Piezoelectric, you can now add a Dispersion subnode. There, you can
choose between the Debye and Multipole Debye dispersion models.

This feature requires the MEMS Module or AC/DC Module.

REDUCED INTEGRATION
A new framework has been added for reduced integration. It is now available in the
Solid Mechanics and Membrane interfaces. Reduced integration is particularly useful
when the computational cost per integration point is high, which is true for many
advanced material models.

Reduced integration is controlled from the Quadrature Settings section in various

material models. It is available in top-level material models like Linear Elastic Material.
The selected integration rule will then be inherited by any subnodes that may be
added.

176 | CHAPTER 1: RELEASE NOTES

EASIER GAUSS POINT EVALUATION IN DOMAINS
A set of physics-scoped operators have been added as tools for easier evaluation of
Gauss point results in domains. The operators automatically adapt to the Gauss point
pattern in each element. These operators are:

• <phys>.gpeval(expr): Map a field of suitable order to Gauss point data.

• <phys>.elemgpmax(expr): Maximum value in any Gauss point in an element.
• <phys>.elemgpmin(expr): Minimum value in any Gauss point in an element.
• <phys>.elemavg(expr): Average value of the Gauss points in an element.
• <phys>.elemavg(expr): Integration over an element using Gauss point data.

INTERCALATION STRAIN
A new subnode, Intercalation Strain, has been added to the Linear Elastic Material and
Nonlinear Elastic Material nodes in the Solid Mechanics interface. The purpose is to
model inelastic strains caused by ion transport.

This feature requires the Battery Design Module.

BOLT PRETENSION FOR BEAMS

The Bolt Pretension feature has been added to the Beam interface, thus facilitating a
simplified modeling of prestressed bolts. The functionality is analogous to that in the
Solid Mechanics interface, with the exception that the bolt cross section is represented
by a point selection rather than by a boundary selection.

SEQUENTIAL TIGHTENING OF BOLTS

In the Bolt Selection node, a new check box, Sequential tightening, has been added.
When selected, it is possible to enter expressions for time or parameter values when the
prestress force in the individual bolt is changed. Using this new approach, several bolts
can be sequentially tightened within a single study step.

BOLT PRESTRESS BY TIGHTENING TORQUE

In the Bolt Pretension node, it is now possible to specify the bolt force also in terms of
the tightening torque. In addition, when using one of the other options (bolt force or
bolt stress), the corresponding torque can be computed.

DEFAULT EVALUATION GROUP FOR BOLT FORCES

When a Bolt Pretension node is active in a study, a default evaluation group containing
a list of forces in all bolts selected in Bolt Selection subnodes is automatically generated.

STRUCTURAL MECHANICS MODULE | 177

 AUTOMATIC GENERATION OF BOLT LABELS
When adding a new Bolt Selection subnode, the bolt label automatically gets a
sequentially numbered default name. The base name that is used for the name
generation is given in the new Bolt Label section in the parent Bolt Pretension node.

CHANGES IN BOLT THREAD CONTACT

As a consequence of the general contact improvements, there are significant changes
in how bolt thread contact is handled. The Bolt Thread Contact node, which used to be
a global feature, is now located under Pairs at the boundary level. The old Thread
Boundary Selection subnode has been removed. The selection of pairs is now done
directly in the Bolt Thread Contact node. This means that you need to add one Bolt
Thread Contact node per bolt because the pair selection should represent one single
bolt.

Also, the old Bolt Label input field has been removed because it is not used.

When opening a file created in a version prior to 6.0, the old representation is
automatically converted to the new one. When using the API, there is no such
conversion. Any file that references the old BoltThreadContact and
ThreadBoundarySelection must be updated so that it uses the new
PairBoltThreadContact.

COUPLING FROM BEAM POINTS TO SOLID EDGES

The Solid-Beam Connection multiphysics coupling has been extended with an option for
connecting a point on a beam to edges in a solid. One use case is when combining bolts
idealized by beams with components modeled with solid elements. Then the edge of
the bolt hole can be connected to the end of the beam to form an abstract bolt head.

CIRCUMFERENTIAL MODE EXTENSION FOR AXISYMMETRIC SHELL

The capability to compute different harmonics in the azimuthal direction for
eigenfrequency and frequency-domain analysis has been added to the Shell interface in
2D axisymmetry. You enable it by selecting Circumferential mode extension in the
settings for the Shell interface.

WEAK CONSTRAINTS FOR AXISYMMETRIC SHELL

The foldline constraints, as well as the shell normal displacement constraints, can now
be of the weak form also in the 2D axisymmetric version of the Shell interface. You can
control the constraint formulations in the Foldline Settings and Advanced Settings
sections in the settings for the Shell interface.

178 | CHAPTER 1: RELEASE NOTES

IMPROVED SHELL DATASET
The Shell dataset has received an important upgrade. Now, shells will be visualized as
solids, with thickness and offset taken into account. Any variable that has a dependency
on the through-thickness location will be displayed accordingly. This is, for example,
used in the default stress plots in the Shell interface. In order to make use of the new
functionality, the definitions of some variables have been changed. For example:

• New variables shell.u, shell.v, and shell.w have been added. They contain the
displacement, including the effect of rotation and distance from the midsurface.
These variables can be used in Deformation nodes, for example.
• The displacement magnitude shell.disp is now based on these new displacement
variables, so that it represents a field also in the thickness direction.

UPDATED INPUT OF THICKNESS AND OFFSET IN SHELL

The settings in the Thickens and Offset node in the Shell interface have been changed,
so that it is easier to enter the more common offset values. Top surface on boundary,
Midsurface on boundary, and Bottom surface on boundary are now predefined choices.

FLUID-STRUCTURE INTERACTION WITH SCREEN

The Screen boundary condition on interior boundaries in fluid-flow interfaces will now
automatically work together with the Shell and Membrane interfaces when using the
Fluid-Structure Interaction multiphysics coupling.

COUPLINGS BETWEEN RIGID CONNECTORS

It is now possible to couple a rigid connector in the Pipe Mechanics interface to a rigid
connector in Solid Mechanics, Shell, or Multibody Dynamics interfaces. To do so,
select a rigid connector defined in the other physics interface from the Connect to list
in the Advanced section in the settings for Rigid Connector.

INPUT OF RESIDUAL STRESS

In the External Stress node, there is now an option called Residual Stress. A residual
stress will not directly affect the displacements. That is, if you just enter a residual
stress, and no other loads, there will be no displacements. The stress is, however, added
to the stress tensor in the sense that it will be part of the stress state that is used in
various material models.

STRUCTURAL MECHANICS MODULE | 179

 EXTERNAL STRESS CAN BE ENTERED AS FIRST PIOLA-KIRCHHOFF
STRESS TENSOR
In the External Stress node in the Solid Mechanics, Layered Shell, and Membrane
interfaces, it is now possible to enter the stress state in form of a 1st Piola–Kirchhoff
stress tensor.

SYMMETRIC FORMULATION FOR RIGID CONNECTOR AND RIGID DOMAIN

Under geometric nonlinearity, the consistent stiffness and mass matrices generated by
the Rigid Connector and Rigid Domain features may become nonsymmetric. The solution
of a nonsymmetric problem requires significantly more memory and computation time
than solving a symmetric problem of the same size. In version 6.0, you can choose to
use a less accurate symmetric matrix instead. Doing so may require some more
iterations, but each iteration will be significantly faster.

In order to use the symmetric formulation, select the Use symmetric formulation for
geometric nonlinearity check box in the Formulation section in the settings for the
feature.

IMPROVED SHEAR CORRECTION FACTORS FOR BEAMS

The shear correction factors generated by the standard beam cross sections have been
updated using more accurate expressions that depend on the aspect ratio of the cross
section as well as the material properties.

COORDINATE SYSTEM TYPES RESTRICTED FOR RIGID CONNECTOR

For a rigid connector, it is not meaningful to select a coordinate system with position-
dependent axis orientations. This is no longer allowed, neither in the Rigid Connector
node itself nor in its subnodes.

EXPRESSIONS FOR THE ROTATION VECTOR CHANGED IN RIGID

CONNECTOR AND RIGID DOMAIN
The expression defining the variables giving the rotation vector in a rigid connector
and rigid domain has been changed in such a way that it is more robust with respect
to computing Jacobian contributions. This can be important if you want to, for
example, control such a variable using other equations.

The affected variables for a rigid connector are named <phys>.th<dir>_<tag> (for
example, solid.thz_rig1). For a rigid domain, the names are of the type
<phys>.<tag>.th<dir> (for example, solid.rd1.thy).

180 | CHAPTER 1: RELEASE NOTES

CONSTRAINT GROUPS FULLY SUPPORTED FOR RIGID CONNECTOR AND
RIGID DOMAIN
In earlier versions, constraint groups could not be used for rigid connectors or rigid
domains if the option Evaluate reaction forces was selected. This limitation has been
removed.

SLIP VELOCITY ADDED TO SHELL AND MEMBRANE INTERFACES

The Slip Velocity subnode under Contact, in which you can describe friction with a
given slip direction, has been added also to the Shell and Membrane interfaces.

IMPROVEMENTS TO SAFETY FACTOR COMPUTATIONS

Some of the models for safety factor computation in the Safety node have been
updated.

• In the Tsai–Hill criterion, there is now a separate input for the compressive
strengths.
• The Modified Tsai-Hill criterion, intended for 2D analysis, has been removed from the
list of failure models. To obtain the same effect, use the Tsai-Hill criterion and select
the new Use plane stress formulation check box. When you open an old model that
contained the Modified Tsai-Hill criterion, it will automatically be updated to the new
settings.
• 2D formulations of the Tsai–Wu and Hashin criteria have been added. To use them,
choose Tsai-Wu orthotropic or Hashin as the failure criterion, and then select the new
Use plane stress formulation check box.

STRICTER DEFAULT TOLERANCE FOR TIME-DEPENDENT ANALYSIS

The default tolerance for time-dependent analysis has been decreased from 0.01 to
0.001 for all structural mechanics interfaces. This will affect new and old models in
which the solver sequence has not been manually edited. The effect is that solutions
will be more accurate but typically at the cost of longer solution times.

THE TIME-DEPENDENT, PRESTRESSED, MODAL STUDY TYPE REMOVED

Due to problems with obtaining a correct linearization in all situations, the Time-
Dependent, Prestressed, Modal study type is no longer available.

TORQUE ARROWS FOR LOADS IN 2D

In the default load plots, out-of-plane moments in 2D are now represented by a new
type of arrow, called Torque in the Point plot.

STRUCTURAL MECHANICS MODULE | 181

 AZIMUTHAL MODE NUMBER AUTOMATICALLY TRANSFERRED TO
REVOLUTION DATASET
When using circumferential mode extension for 2D axisymmetry, the azimuthal mode
number (for example, solid.mk) is now automatically transferred to the revolution
datasets that are created as default plots. This means that you do not need to modify
the settings in the dataset when switching between results for different mode numbers.

MODELING OF CRACK CLOSURE

When modeling cracks using the Crack feature, you can now inhibit closure of the crack
by adding a Crack Closure subnode. This subnode will add a contact condition, in which
friction in the crack can also be taken into account.

CRACK FACE LOAD CAN BE ASSIGNED TO LOAD GROUPS

When using a Face Load under Crack, you can now assign a load group, in order to limit
and scale the load for certain load cases.

IMPROVEMENTS IN STRESS INTENSITY FACTOR COMPUTATION

The stress intensity factors KI, KII, and KIII are now computed with a sign. It is thus
possible to determine the range of, for example, KII for a set of load cases. A negative
value of KI indicates crack closure.

In order to control the definition of the signs of the stress intensity factors in 3D, a
new subnode, Reverse Crack Front, has been added under Crack.

The ratios between KI, KII, and KIII are determined through an analysis of the
displacements on the crack surface. In the new Stress Intensity Factor section in the
setting for the J-integral node, you can fine-tune the part of the crack that is used for
these evaluations.

STRESS LINEARIZATION IMPROVEMENTS

The Stress Linearization feature has two important improvements:

• It is no longer necessary to use a geometrical line to define a stress classification line

through the thickness. You can now use a line between two arbitrary points. The
points can be geometrical points or just locations specified by coordinates.
• Stress linearization values can be presented as a field over a boundary. In this case,
the evaluation is performed using a large number of automatically generated lines,
extending orthogonally to the boundaries. By using this method, you can find the
worst location for a stress classification line.

182 | CHAPTER 1: RELEASE NOTES

CHANGES IN RANDOM VIBRATION FUNCTIONALITY
There are several news within the field of random vibration analysis:

• The number of study steps generated by a Random Vibration (PSD) study sequence
is now two instead of three. The “dummy” Frequency Domain study now resides
inside the Model Reduction study.
• A new operator, q2sq, has been added. It is similar to the existing q2 operator but
has more convenient arguments and output. In particular, it is targeted at
computing the RMS of the von Mises stress and vector norms.
• In the Random Vibration node, it is possible to enter default parameters for operators
like rms, which perform an integration. It is then possible to invoke these operators
without the last three arguments.
• It is now easy to add the base acceleration to computed acceleration PSDs, thus
facilitating direct evaluation of absolute as well as relative spectra.
• There are significant performance improvements in the random vibration
computations.

CHANGED NOMENCLATURE FOR FORCE LINEAR STRAIN

The label of the Force linear strain check box, which is present in many material models,
has been changed to Geometrically linear formulation in order to improve clarity.

OVERRIDING OF GEOMETRIC NONLINEARITY IN THE BEAM INTERFACE

In studies that are geometrically nonlinear, it is now possible to, per material selection,
enforce geometrical linearity also in the Beam and Pipe Mechanics interfaces. You do
this by selecting the Geometrically linear formulation check box. This can be useful for
improving efficiency when there are parts where effects of geometric nonlinearity are
negligible. One such case is when bolts are modeled using beams in a contact analysis.

CHANGES IN DEFAULT PLOTS

There are many improvement in the default plots. Some of the more prominent are:

• The default stress plot in 3D now uses a Volume Plot rather than a Surface Plot. There
are several advantages with that change, including easier selection of individual
domains and more consistent evaluation of some expressions defined at the domain
level. However, you may find that the plotting is slower in larger models. Also, the
visualization of curved boundaries with a coarse mesh is smoother in a surface plot.
• A new Prism color table is used for the default stress plots. This color table will
emphasize regions with high stresses.

STRUCTURAL MECHANICS MODULE | 183

 • In the default stress plots, the smoothing is conditional, so that stress jumps larger
than 20% between adjacent elements will be plotted without smoothing. This makes
it easer to identify a solution computed on a too coarse mesh.

EASIER USE OF MULTIPLE CELL PERIODICITY NODES

The way studies are generated has been changed so that unique studies are created for
each Cell Periodicity node. This makes it possible to, for example, run nested
homogenizations.

A new section, Advanced, has been added to the settings for Cell Periodicity. Here you
can, if needed, select to use the old scheme to keep the old behavior.

IMPROVED GUI IN SPRING-DAMPER NODE

In the Spring-Damper node, options that are related to rotational degrees of freedom
are now only active if such degrees of freedom are actually present at the attachment
points.

THERMAL EXPANSION IMPROVEMENTS

The input options Tangent coefficient of thermal expansion and Thermal strain have been
added to the Thermal Expansion node in the Shell interface as well as to the Layered
Thermal Expansion multiphysics coupling. These options have also been added to the
Thermal Expansion subnodes under all constraints.

IMPROVED HANDLING OF LOAD GROUPS FOR THERMAL EXPANSION AND

HYGROSCOPIC SWELLING
The handling of load groups in Thermal Expansion and Hygroscopic Swelling has been
improved. In particular, using several such features with the same selection now works
in a more intuitive way.

REMOVAL OF SEVERAL PAIR FEATURES

As part of a major overhaul of the pair handling, several pair features have been
removed. These nodes were seldom or never used, and the interpretation was difficult
and error prone. The removed pair features are Fixed Constraint, Prescribed
Displacement, Roller, Symmetry, Antisymmetry, Pinned, No Rotation, Rigid Connector, and
Boundary Load. In case any such feature was used in a model, it can easily be replaced
by standard boundary conditions.

184 | CHAPTER 1: RELEASE NOTES

CHANGE IN PRESCRIBED DISPLACEMENT WHEN ELASTIC
PREDEFORMATION IS USED
The interpretation of the input in the Prescribed Displacement node in connection with
Elastic Predeformation has been changed. It is now the total displacement, which is
given in the same way as if Elastic Predeformation is not present.

CHANGED DEFINITION OF CENTROID OF RIGID CONNECTOR IN SHELL

INTERFACE
When the centroid of a Rigid Connector in the Shell interface is defined by Centroid of
selected entities, and the entity type is points, the computation of the centroid has been
changed. In version 6.0, the average of the coordinates of the points is used without
modification. In earlier versions, a possible offset from the reference surface was
included, so that the actual midsurface of the shell was used. The reason for the change
is that since it is now possible to select points that are not part of the shell interface, an
offset cannot be defined in all points.

CHANGED API FOR INITIAL VALUES

Due to changes in the implementation of the Initial Values node in Solid Mechanics,
there are changes in the API syntax. In version 5.6 and earlier, if the displacement field
had a name that was not u, the field name was reflected in the API syntax when setting
an initial value. Also, the syntax for the initial velocity has changed. Assume that the
field name is u_solid:

Earlier versions:
...feature("init1").set("u_solid", new int[]{1, 0, 0});
...feature("init1").set("du_solid/dt", new int[]{2, 0, 0});

Version 6.0:
...feature("init1").set("u", new int[]{1, 0, 0});
...feature("init1").set("ut", new int[]{2, 0, 0});

If you are reading a file containing the old syntax, there is an automatic conversion for
some common cases. The cases that are handled include when the field name is u* or
u_solid*, where * is a number less than or equal to 10.

For all other cases when a nondefault field name is used, you must manually replace
the argument in the call to set() by the new names, u or ut, to set the initial
displacements or velocities.

STRUCTURAL MECHANICS MODULE | 185

 New and Updated Models in Version 6.0

HANGING CABLE
A cable is a structural member that has stiffness only in its tangential direction but
virtually no bending stiffness. When supported only at its two ends, it deflects under
gravitational load forming a curve known as a catenary. This example shows how to
model components such as cables, wires, or strings using the Truss interface.

INFLATION OF A SQUARE AIRBAG

A square airbag is inflated using a constant air pressure. In some regions of the airbag
the material wrinkles during inflation, as a negative principal stress develops. The
wrinkling membrane model within the tension field theory gives a correct description
of the stress distribution and wrinkling pattern in inflated membranes.

TORSION OF A CIRCULAR MEMBRANE

In this example, a torque is applied to the inner edge of a circular annulus-shaped
membrane while the outer edge is fixed, resulting in membrane wrinkling. The
wrinkling membrane model avoids the equilibrium instability that would be produced
by the compressive stresses.

UNIAXIAL STRETCHING OF A RECTANGULAR MEMBRANE

This example demonstrates the wrinkling phenomenon in a thin sheet stretched
uniaxially. The modified membrane theory, which incorporates the wrinkling model,
ensures noncompressive principal stresses in the wrinkled region. The numerical results
from the simulation are compared to analytical results.

AXISYMMETRIC TWIST AND BENDING

Including circumferential displacements in a 2D axisymmetric Solid Mechanics
interface makes it possible to compute twist and bending deformations. This model
determines stress concentration factors for a hollow shaft for load cases of axial
extension, torsion, and bending, using a computationally lean 2D axisymmetric
formulation. To demonstrate the equivalence, the model setup and results are
compared with a full 3D analysis.

DEFORMATION OF AN IRON PLATE BY MAGNETIC FORCE

A strong permanent magnet is placed close to a clamped thin plate made of iron. The
magnetic force causes the plate to be deflected. This example studies the plate elastic
deformation and stress. The deformation of the plate has an influence on the

186 | CHAPTER 1: RELEASE NOTES

distribution of the magnetic field. This effect is accounted for using a moving mesh in
the air surrounding the plate. The model is set up using the Magnetomechanics, No
Currents multiphysics interface.

STRESS IN COOLING PIPELINE NETWORK

This example demonstrates how to model coupled flow, heat transfer, and structural
deformation and stress in a pipeline network. Gravity loads from the pipe and fluid are
also taken into account.

MODELING OF PRETENSIONED BOLTS

This is a tutorial example, showing how to work with pretensioned bolts. Various
aspects of bolt modeling are explored:

• Modeling bolts by solids or by beams

• Connections of beams to solid components
• Bolts ended by nuts or by internal threads in the attached component
• Import of bolt and nut geometries from the Part Libraries
• Connection of bolts using full contact, or by continuity
• Sequential tightening of a set of bolts

COMPONENT MODE SYNTHESIS TUTORIAL

In this tutorial example, the concept of component mode synthesis (CMS) is
introduced through a simple solid model of a beam. Parts of the beam are reduced into
CMS components. Dynamic as well as stationary analyses are performed. Moreover,
you will learn how to apply loads directly on a CMS component and work with such
components during result evaluation. The results from the full and reduced versions of
the model are compared.

STRUCTURAL MECHANICS MODULE | 187

 Subsurface Flow Module
New Functionality in Version 6.0

IMPROVED HANDLING OF POROUS MATERIALS

Porous materials are now defined using phase-specific properties tabulated in the
Porous Material node. In addition, subnodes can be added for the solid and fluid
features. Several subnodes can be defined for each phase. This functionality makes it
possible to use one and the same porous material for fluid flow, chemical species
transport, and heat transfer without having to duplicate material properties and
settings.

IMPROVEMENTS TO BRINKMAN EQUATIONS INTERFACE

In the Brinkman Equations interface, the new Porous Medium feature now supports the
Porous Material functionality. When opening a model created in version 5.6 or earlier,
the old Fluid and Matrix Properties feature is not migrated, but a warning message is
displayed. It should be manually replaced by the new Porous Medium feature (which
adds Porous Matrix and Fluid subfeatures) to allow an updated definition of the material
properties. Support for obsolete features is not ensured in future versions; thus it is
strongly recommended to transfer the old feature to the new feature.

POROUS SLIP IN THE BRINKMAN EQUATIONS INTERFACE

The boundary layer between free and porous media flow may be very thin and
impractical to resolve. Using the new porous-slip wall treatment, you can account for
porous walls without resolving the full flow profile in the boundary layer. Instead, a
stress condition is applied at the porous surfaces by utilizing an asymptotic solution,
yielding a good accuracy in the bulk flow. The functionality is activated in the new
Porous treatment of no slip condition settings, and it is then used as the default wall
condition. This new feature can be used in most problems where the model domain is
large.

NONISOTHERMAL FLOW IN POROUS MEDIA

The new Nonisothermal Flow, Brinkman Equations multiphysics interface automatically
adds the coupling between heat transfer and fluid flow in porous media. It combines
the Heat Transfer in Porous Media and Brinkman Equations interfaces and automatically
adds a Nonisothermal Flow coupling node.

188 | CHAPTER 1: RELEASE NOTES

You can find this interface under the Nonisothermal Flow branch in the Fluid Flow folder
in the Add Physics tree.

EXTENDED FEATURES FOR HEAT TRANSFER IN POROUS MEDIA

The default Porous Medium feature in the Heat Transfer in Porous Media interface by
default selects the Local thermal equilibrium option under the Porous medium type list.
This corresponds to the previously available averaging techniques within the Porous
Medium feature, which solved a single equation for the average temperature field with
effective properties. New averaging techniques are available for computing the Effective
thermal conductivity in porous media: Solid spherical inclusions, Fluid spherical inclusions,
Wrapped screen, and Sintered metal fibers. In addition, the Volume average and
Reciprocal average techniques have been renamed to Plane layers parallel to heat flow
and Plane layers perpendicular to heat flow for clarity. Finally, the new Equivalent thermal
conductivity option makes it possible to enter user-defined expressions.

For these new options, the postprocessing variables (thermal properties, heat fluxes,
heat sources, and heat and energy balance) are available in a unified way for
homogenized quantities.

SOURCE TERMS FOR THE SHALLOW WATER EQUATIONS INTERFACE

The shallow water equations give a 1D or 2D approximation of shallow flows by
averaging the water column along the depth. Rain, local upwelling, pumping devices,
or boundary stresses enter as source terms in the model equations. Introducing source
terms was previously possible through the equation view, but it is now possible to add
momentum and mass sources as predefined quantities directly in the interface.

MISCELLANEOUS
• The Longitudinal dispersivity and Transverse dispersivity settings in the Thermal
Dispersion feature are now available with all COMSOL products that include the
feature.
• A new Discharge boundary condition in the Darcy’s Law and Richards’ Equation
interfaces makes it possible to specify the discharge in outlets in term of cubic meters
of water per second (in general, volume of liquid per unit of time).

SUBSURFACE FLOW MODULE | 189

 New Models in Version 6.0

GLACIER FLOW: A 2D STUDY OF COLD AND TEMPERATE GLACIERS

This example shows how to set up a glacier flow model containing several important
aspects of glacier modeling: the creation of the 2D geometry, modeling of non-
Newtonian flow, and implementation of basal sliding. The example uses two different
glacier types: a cold glacier and a temperate glacier.

190 | CHAPTER 1: RELEASE NOTES

Wave Optics Module
New Functionality in Version 6.0

ELECTROMAGNETIC WAVES, BOUNDARY ELEMENTS

This new physics interface is useful for modeling the scattering properties of objects,
evaluating electric fields far from the scatterer.

PART LIBRARY FOR SLAB AND RECTANGULAR WAVEGUIDE ELEMENTS

The new Wave Optics Module Part Library contains slab waveguide shapes for 2D
geometries and rectangular waveguide shapes for 3D geometries, including the
following waveguide elements:

• Straight waveguides
• Tapered waveguides
• Bent (ring) waveguides
• S-bend waveguides
• Couplers

The parts are built from fully parameterized sequences of geometry instructions.

LAYERED TRANSITION BOUNDARY CONDITION

Multiple thin layers, such as an antireflection coating of an optical lens, can, for close
to normal incidence, be described by the Layered Transition Boundary Condition
feature. It requires combining this physics feature, Layered Material in the global
Materials, and Layered Material Link in the Materials node.

SHIFTED LAPLACE CONTRIBUTION ON MULTIGRID LEVELS

If no geometrical feature sizes are smaller than half a wavelength, when the operating
frequency is high, modeling with a higher-order element such as cubic element
discretization is beneficial for faster computations. The computation efficiency can be
further boosted by imposing the Shifted Laplace contribution on multigrid levels.

SUGGESTED ITERATIVE SOLVER FOR PERIODIC STRUCTURES

Typical periodic problems are solved with a direct solver. However, the direct solver
consumes a lot of memory, when the periodic unit cell size is not subwavelength. If

WAVE OPTICS MODULE | 191

 this is the case, switch to the Suggested Iterative Solver to finish the computation faster
with less memory usage.

CONSTRAINT-FREE PORT OPTION

The constraint-free port option is available to compute the expansion coefficients as an
overlap integral, while in the default port formulation, the expansion coefficients (or
S-parameters) are calculated by adding a scalar dependent variable for each coefficient
and then adding a constraint to enforce the series expansion.

SMOOTHED HEAT SOURCE CALCULATION

Use an averaged loss calculation option in the bidirectional formulation for the
Electromagnetic Waves, Beam Envelopes interface to remove cross terms between the
two waves that are not resolved by the mesh. If this spatially, fast-varying heat source
distribution anyhow is “washed out” by the heat transfer, it can be advantageous to
not include the cross terms when calculating the electromagnetic loss (and heat
source).

SYMMETRY AXIS REFERENCE POINT

The Symmetry Axis Reference Point subnode is available in 2D axisymmetry, when there
is an incident field defined in the parent node. Then this subnode is available as a
default subnode to the Scattering Boundary Condition or Matched Boundary Condition
parent node. The Symmetry Axis Reference Point subnode defines a reference position
at the intersection point between the parent node’s boundary selection and the
symmetry axis.

DEFAULT PLOTS FOR NUMERIC PORT MODE FIELDS

Plots are automatically created of the port mode fields for numeric ports.

New and Updated Models in Version 6.0

OPTIMIZATION OF A PHOTONIC CRYSTAL FOR DEMULTIPLEXING

This model demonstrates how to apply shape optimization to a photonic crystal. The
objective function is to maximize the output power ratio between two narrow
frequency bands, while constraining the loss from below. This effect is achieved by
letting GaAs pillars change position but not shape. The implementation makes use of
the Free Shape Domain and Transformation features, so that gradient-based
optimization can be applied.

192 | CHAPTER 1: RELEASE NOTES

WHISPERING GALLERY MODE RESONATOR
Dielectric microspheres can support whispering gallery modes with high optical quality
factors. This model illustrates how to compute the different eigenmodes and resonance
frequencies. The resonance frequencies are either filtered by their spatial localization
in the resonator or by comparing the losses of the bound modes and the, for this
application undesired, air modes.

HEXAGONAL PLASMONIC COLOR FILTER

Robust and easy-to-fabricate color filters have many applications. This model
demonstrates how to build a model of an absorbing color filter, based on a hexagonal
array of holes in a thin aluminum layer. The structure is hexagonally periodic, but in
the model, it is demonstrated how to define the problem both as a hexagonally and a
rectangularly periodic problem.

MACH-ZEHNDER MODULATOR
The Mach–Zehnder Modulator model has been updated to use parts from the Wave
Optics Module part library.

OPTICAL RING RESONATOR NOTCH FILTER

This optical ring resonator model has been updated to use parts from the Wave Optics
Module part library.

Backward Compatibility with Version 5.6 and Earlier

The port mode fields are now defined as the outgoing waves, whereas in version 5.6
and earlier the port mode fields were defined as the incoming (exciting) waves. This
change only affects User defined and Periodic port types, as input fields can be
provided for these types. When loading MPH-files from version 5.6 or earlier, the
input fields for User defined and Periodic ports are migrated to function with the new
mode field definition. However, for User defined ports, the migration assumes that the
material adjacent to the port is homogeneous.

WAVE OPTICS MODULE | 193

 Material Library
New and Updated Material Data in Version 6.0
In version 6.0, the Material Library contains 10,328 materials and 84,573 material
property datasets. The following improvements and additions have been made:

• Added several new precious metal alloys: CuPd, Cu-Pt, Au-Pd, Au-Ag, Ag-Cu, Au-
Cu, Au-Pt, Ag-Cd, Au-Ag, Ag-Pd, Ag-Mg, Ag-Mn, Ag-Pt, Pt-Fe, Pd-Ag, Pt-Au,
and Pt-Cu-Co.
• Added several new refractory metal alloys: Cr-based, Ti-Ge, Ti-Cu, Ti-Nb, Ti-Sn,
Ti-Zn, Zr-0.14Hf, Zr-1Hf, and Zr-1Hf-1Ta.
• Added thermal expansion data for TRIP 700, TRIP 1000, ALCHROME alloy DK,
Stellite 23, Stellite 30, W – 20 Cu, W – 40 Cu, W – 7 Ni – 2.5 Cu, Al-7Si, Al-13Si,
Al-17Si, AZ81, EK30, HK31, HZ32, CuAu3, Cu3Au, 44 Pd – 33 Ni – 23 Cr (GE-
76 brazing alloy), and Alloy 19-2.
• Added new materials: TD-NiCr, Fe-Si alloys, Nimonic 100, 30 Co - 23 Fe - 21 Cr
- 20 Ni,Mg – 7 Al – 1 Si, Sn-Mg binary alloys, and Zn-Cu binary alloys.
• Added several foods: honey, orange juice, grape juice, apple puree, pear puree, and
peach puree. Foods were also put into their own group.
• Added data for several oxides, silicates, and electroceramics.
• Added data for Controlled expansion 45, Temp compensator 30 (types 2, 3, and 4),
1065, 1095, Kovar, 7068, several Cu alloys, AZ81, HM31, ZK60, and Au2Mn.
• The heat capacity data for diamond was incorrect and has been updated. The specific
heat data is not affected by this error.
• Refit the thermal expansion data for Alloy 39; there was a significant change in the
values.
• The thermal expansion data for: Hastalloy B, Hastalloy C, several Monels, Inconel
X-750 Inconel HX, Alloy 72 and Alloy 22-3 has been changed, the new values differ
by 5–10%.
• The data for Ta2O3 is actually for Ta2O5. The name of the material has been
updated.
• The expansion data for HfTiO4 and ZrTiO4 was changed to a new reference with
the expansion in different crystallographic directions.

194 | CHAPTER 1: RELEASE NOTES

• Added new materials: TaN, Schott Borofloat 33, spider silk, wheat, Inconel
718SPF, UNS N06333, and several cast irons.
• Added specific heat data for many salts, borides, carbides, nitrides, and other
nonmetallic materials.
• Added data for Inconel X-750, AerMet 100, AM-355, FMDP pipeline steel, API-
X70 pipeline steel, Incoloy A-286, 304, A9, 17-7 PH, 10B46, several Ti-based
alloys, Pb – 55 Bi (eutectic), Zr50Cu40Ag10, rare-earth trifluorides, Dy2WO6,
Gd2WO6, Er6WO12, Ho6WO12, Y2W3O12, and Yb6WO12.
• Data for N08330 was removed and combined with Incoloy 330.
• Added apparent viscosity for several polymers.
• The expansion of aluminum was extended down to 10K with a new reference. The
CTE value near 20K changed substantially, the others (dL/L and MTE) were only
slightly affected.
• Added HgCr2Se4.
• Added data for the following oxides: Dy2TiO5, Er2TiO5, Gd2TiO5, Ho2TiO5,
Lu2TiO5, Y2TiO5, La2TiO5, and Nd2TiO5.
• Added data for Lexan 101 and refit the thermal expansion; dL/L increased by 5%.
• Two new magnetic properties were added: magnetic core loss versus flux density
and magnetic exciting power versus flux density.
• Several electrical steels were added.
• Several iron-base alloys, nickel-base alloys, titanium-base alloys, magnesium-base
alloys, and aluminum-base alloys were added.
• The compressive stress-strain curves for Ta, Ta-2.5W, Ta-5W, and Ta-10W were
incorrect and have been fixed.
• Data was added for Zr-based alloys: Zr-1Nb and Zr-2.5Nb; for Mg-based alloys:
ZK10, ZN10, and ZX10; and for Al-based alloys: 1085, 6016, 6110, and 7075.
• Data was added for: NILOMAG alloy 77, Ni-SPAN C alloy 902, Inconel 601,
Inconel 625LCF, Incoloy 25-6HN, 316L, Inconel X-750, Super 304H (304HCu),
Fe-2Mn-1.3Si-0.7Cr-0.4C, X100 pipeline steel, NiCrMo 2.5-IG filler wire, Ductile
Ni-Resist D-5S, SiMo51 ductile iron (3.2 C – 4.5 Si – 1 Mo), and SiMo1000 ductile
iron.
• Data for Haynes 25 and L-605 were combined as Haynes 25.
• Data has been added for several Cu-based alloys and several Nitronic alloys.

MATERIAL LIBRARY | 195

 • The thermal conductivity for Kapton HN was incorrect; it was actually for the MT
grade. This material has been updated with the correct data.
• Data was added for: UNS S31266, LSHR disk alloy, Rene-41, Haynes 25 (L-605),
S350, S420, and G500.

196 | CHAPTER 1: RELEASE NOTES

LiveLink™ for Excel ®
New Functionality in Version 6.0

IMPROVED VBA SUPPORT

Support for Visual Basic for Applications (VBA) has been improved. For all buttons on
the COMSOL ribbon tab, a VBA command has been introduced. This functionality
makes it possible to use toolbar button commands functionality directly from VBA
code, which makes it easier to automate steps that were easy to perform using the
ribbon but perhaps difficult to perform using the COMSOL API.

ADDITIONAL IMPROVEMENTS
• The LiveLink™ interface now supports Excel® 2021.
• When saving Excel data from apps, it is now possible to include expressions in the
exported data.
• The new Break All Links button on the COMSOL ribbon tab removes all the
comments with links to the COMSOL models from the current worksheet. Doing
so breaks the links between all cell ranges in the worksheet and the model.
• Models can be loaded from a database that is stored using Model Manager. Models
cannot be saved in a database using Excel.
• The stability of connections has been improved as well as improved handling when
Excel is disconnected from a COMSOL Multiphysics Server.

LIVELINK™ FOR EXCEL® | 197

 LiveLink™ for MATLAB ®
New Functionality in Version 6.0

SYSTEM REQUIREMENTS
LiveLink™ for MATLAB® is supported using MATLAB 2021a or 2021b. Older
versions may work, but a Java VM version 1.8 or newer is required. It is available in
MATLAB 2017b or newer.

UPDATED USER INTERFACES

All user interfaces have been remade using the App Designer in MATLAB
(mphnavigator, mphopen, mphsearch, and mphmodellibrary). This update provides
more modern and responsive user interfaces that now permit resizing. The new user
interfaces are available for MATLAB versions from 2020a and newer. The old user
interfaces are still available, but they will not be updated and they will be removed in
a future version of COMSOL Multiphysics. Note that some differences may be
observed depending on the exact version of MATLAB used.

AUTOCOMPLETION
Autocompletion for wrapper functions is supported in the Command Window, the
regular Editor, and the Live Editor for Live Scripts. This functionality makes it much
easier to use the wrapper functions.

NEW GEOMETRY AND MESH FUNCTIONS

COMSOL 6.0 permits more advanced uses of geometries and multiple meshes in a
model. In order to make it easier to follow such models from MATLAB, the new
mphgeominfo and mphcomponentinfo functions have been introduced. The function
mphmeshstats has been improved as well.

MODEL MANAGER SUPPORT

Models can be loaded from a database that is stored with Model Manager using
mphload and mphopen. Models cannot be saved in a database using MATLAB.
Instead, use the command mphlaunch to start the COMSOL Desktop and access the
Model Manager there.

198 | CHAPTER 1: RELEASE NOTES

OTHER IMPROVEMENTS
• The mphmin and mphmax functions are now able to return the position of the
minimum or maximum value when using the position property.
• The mphinterp and mpheval functions have been updates with improved error
messages in case the expression used is outside the geometry or the domain where
the expression is used.
• The mphreport function now supports generating reports as Microsoft®
PowerPoint® presentations in addition to creating reports in HTML or Microsoft®
Word® formats.

Backward Compatibility with Version 5.6

COLOR TABLES
All functions that have to do with color tables (for example, cividis, cyclic, and
disco) except for colortable are deprecated. When using the colortable function,
you can use autocompletion to get a list of available color tables. As an alternative to
using, for example, cividis you can use

table = colortable('cividis');

instead.

REPLACED AND DEPRECATED FUNCTIONS

The functions mphint, mphgetp, and mphgetmodelnode are deprecated.
mphgetmodelnode has been replaced by mphcomponentinfo where new features have
been added. Th deprecated functions still exist but will result in a warning when used
and may be removed in a future release of COMSOL Multiphysics. Such warnings can
be disabled using this command

warning('off','COMSOL:DEPRECATED')

Backward Compatibility with Version 5.4

The mphreduction function has been changed in version 5.5 such that it no longer
can be used to define model reduction studies. You have to manually add the necessary
studies to the model. The mphreduction function can then be used to extract the
defined matrices for use in MATLAB.

LIVELINK™ FOR MATLAB® | 199

 LiveLink™ for Simulink ®
New Functionality in Version 6.0
A significant performance increase has been obtained for cosimulation when
simulating time-dependent COMSOL models together with Simulink.®

Initialization and stability of simulations have also been improved on all platforms.

200 | CHAPTER 1: RELEASE NOTES

The COMSOL API for Use with Java ®
COMSOL 6.0 API Changes

PAIR FEATURES
The removal of the fallback features under pair features can break API code that you
have written that access those features. The presence of a default pair feature may also
create a different configuration compared to what previous API runs did.

Java or MATLAB program that accessed fallback features under a pair feature will not
work anymore. This is a necessary limitation because any attempt to support such API
backward compatibility will be both unintuitive and unsafe. For most physics created
from the API, the default feature from the physics will act as fallback to the pair
features. As long as the default fallback feature in 5.6 was of the same type, the behavior
in 6.0 will be equivalent. If you used a special fallback feature for a pair feature in 5.6,
you now place it either before or after the pair feature and use the same selection as the
pair feature. Placing the fallback feature after makes it possible to use the pair feature’s
selection directly. The example below show such code for 5.6 and how the equivalent
code in 6.0 can be:

// Creating a pair feature in both 5.6 and 6.0

model.component("comp1").physics("es").create("cont1",
"Continuity", 2);
// Assign some pairs to the pair feature
model.component("comp1").physics("es").feature("cont1").
set("pairs", new String[]{"ap1", "ap2"});

// Creating a fallback feature under a pair in 5.6 (this will not

work in 6.0)
model.component("comp1").physics("es").feature("cont1").
create("sfcd1", "SurfaceChargeDensity", 2);
// Accessing a settings in 5.6 (this will not work in 6.0)
model.component("comp1").physics("es").feature("cont1").
feature("sfcd1").set("rhoqs", "1e-9");

// Creating an equivalent feature in 6.0 that act as fallback to

the pair feature
model.component("comp1").physics("es").create("sfcd1",
"SurfaceChargeDensity", 2);
// Use the same selection as the pair feature
int[] ent = model.component("comp1").physics("es").
feature("cont1").selection().entities();

THE COMSOL API FOR USE WITH JAVA® | 201

 model.component("comp1").physics("es").feature("sfcd1").
selection().set(ent);
// Accessing a settings in 6.0
model.component("comp1").physics("es").feature("sfcd1").
set("rhoqs", "1e-9");

When a physics interface is created from the API in 6.0, it will also add default pair
features that did not exist in 5.6. As long as the final model use all created pairs the
behavior will be the same in 6.0; otherwise, it may be necessary to disconnect the
default pair feature with an extra command. Below is an example for the Electric
Currents interface, but the actual set operation is identical for all interfaces:

model.component("comp1").physics("ec").feature("dcont1").
set("pairDisconnect", true);

COMSOL 5.6 API Changes

MESH PART UNITS

The unit of the mesh part created by createMeshPart is now set to the unit of the
source geometry. In earlier versions, it was created without a unit.

THE ADAPT FUNCTION — THE EXPRTYPE PROPERTY

For the adapt API function, the default for the exprtype property has changed: It is
error if added in a component and a solution exists; otherwise, it is size. In previous
versions, the default was always error.

COMSOL 5.4 API Changes

CAD IMPORT MODULE, DESIGN MODULE, AND LIVELINK PRODUCTS FOR

CAD CHANGES
The new fillholes property has been added for the Import function. The following
applies:

• The combination of the fillholes property set to on and the knit property set to
solid is equivalent to the knit property set to solid in previous versions.

• The combination of the fillholes property set to off and the knit property set
to surface is equivalent to the knit property set to surface in version 5.3a.
• The combination of the fillholes property set to on and the knit property set to
surface is equivalent to the knit property set to surface in version 5.3 and
previous versions.

202 | CHAPTER 1: RELEASE NOTES

The new fillholes property has been added for the Knit function. The Knit
function with the fillholes property set to on corresponds to the Knit function in
previous versions.

ECAD IMPORT MODULE CHANGES

• The properties selnetcontributetobnd, selnetcontributetodom,
selnetkeepbnd, selnetkeepdom, selnetnamebnd, selnetnamedom,
selnetshowbnd, selnetshowdom, selnettagbnd, and selnettagdom have been
added to the Import function. The properties apply to the IPC-2581 and ODB++
file types and enable working with the selections generated for the electrical nets
found in the imported files.
• The elevation property of the Import function is now initialized when setting the
manualelevation property to off. Previously, it was only initialized when setting
the filename property and when the manualelevation property is off and the
value of either the height or importtype properties is changed.

This concludes the release notes for COMSOL Multiphysics version 6.0.

THE COMSOL API FOR USE WITH JAVA® | 203

204 | CHAPTER 1: RELEASE NOTES
I n d e x
A AC/DC Module E ECAD Import Module
new and updated models in 45 new functionality in 80
new functionality in 41 Electrochemistry Module
Acoustics Module new functionality in 81
new and updated models in 55 Electrodeposition Module
new functionality in 48 new and updated models in 88
Application Builder, new functionality in new functionality in 85
11
F Fatigue Module
B backward compatibility, general consid- new functionality in 89
erations 38 new models in 89
Battery Design Module Fuel Cell & Electrolyzer Module
new and updated models in 60 new and updated models in 96
new functionality in 57 new functionality in 90

C CAD Import Module G general new functionality 11

new functionality in 61 Geomechanics Module
CFD Module new functionality in 97
new functionality in 65 new models in 99
new models in 66 geometry and mesh, new functionality
Chemical Reaction Engineering Module for 20
backward compatibility 67, 73
H Heat Transfer Module
new functionality in 68
backward compatibility 109–111, 125
new models in 71
new applications in 108
Composite Materials Module
new functionality in 100
new functionality in 74
L Liquid & Gas Properties Module
new models in 75
new functionality in 113
COMSOL API changes 201
LiveLink for Excel
COMSOL Compiler
new functionality in 197
new functionality in 15
LiveLink for MATLAB
COMSOL Multiphysics, new functionali-
backward compatibility 199
ty in 15
new functionality in 197–199
Corrosion Module
LiveLink for Simulink
new and updated models in 79, 84
new functionality in 200
new functionality in 76
LiveLink™ products for CAD
D Design Module
new functionality in 61
new functionality in 61

INDEX| 205
 M Material Library Polymer Flow Module
new and updated material data in 194 new functionality in 152
new material data in 194 new models in 153
MEMS Module Porous Media Flow Module
new functionality in 114 new functionality in 147
new models in 122
R Ray Optics Module
Metal Processing Module
backward compatibility 158
new functionality in 124
new applications in 156
Microfluidics Module
new functionality in 154
new functionality in 125
results and visualization, new functionali-
Mixer Module
ty in 30
new functionality in 126
RF Module
Model Manager 10
new and updated apps and models in
Multibody Dynamics Module
162
new functionality in 127
new functionality in 160
new models in 130
Rotordynamics Module
N new products in version 6.0 10 new functionality in 164
Nonlinear Structural Materials Module new models in 165
new functionality in 131
S Semiconductor Module
new models in 134
new applications in 168
O operators, functions, and definitions, new functionality in 166
new and updated 24 Structural Mechanics Module
Optimization Module new functionality in 171
backward compatibility 137 new models in 186
new functionality in 136 studies and solvers, new functionality in
new models in 137 27

P Particle Tracing Module Subsurface Flow Module

backward compatibility 141 new functionality in 188

new applications in 139 new models in 190

new functionality in 138 U Uncertainty Quantification Module 10

Pipe Flow Module
W Wave Optics Module
new functionality in 143
new functionality in 191
new models in 143
new models in 192
Plasma Module
backward compatibility 146
new applications in 145
new functionality in 144

206 | I N D E X