COMSOL ReleaseNotes

COMSOL Multiphysics
Release Notes
COMSOL Multiphysics Release Notes
© 1998–2015 COMSOL
Protected by U.S. Patents listed on www.comsol.com/patents, and U.S. Patents 7,519,518; 7,596,474;
7,623,991; 8,219,373; 8,457,932; 8,954,302; 9,098,106; and 9,146,652. 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.
Support for implementation of the ODB++® Format was provided by Mentor Graphics Corporation
pursuant to the ODB++ Solutions Development Partnership General Terms and Conditions. ODB++ is a
trademark of Mentor Graphics Corporation.
COMSOL, COMSOL Multiphysics, Capture the Concept, COMSOL Desktop, LiveLink, and COMSOL
Server 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 5.2
Contact Information
Visit the Contact COMSOL page at www.comsol.com/contact to submit general
inquiries, contact Technical Support, 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/community
• Events: www.comsol.com/events
• COMSOL Video Gallery: www.comsol.com/video
• Support Knowledge Base: www.comsol.com/support/knowledgebase
Part number: CM010001

1
Release Notes
COMSOL® version 5.2 includes an extended and improved Application Builder,

a more powerful COMSOL Server™, and updates and improvements for all
COMSOL products. These Release Notes provide information regarding new
functionality in version 5.2 for all COMSOL products.
3
COMSOL Multiphysics
General New Functionality
IMPROVED INSTALLER
The COMSOL Installer now includes better support for installing and configuring the
license manager software.
RELEASE OF LICENSE AT NEW/OPEN

The default behavior in 5.2 is that any used COMSOL product licenses are released
when you use New or Open to start working on a new application. The behavior in
earlier versions—that you need to restart COMSOL Multiphysics—is also available as
a preference setting.
New Functionality in the Application Builder
GENERAL NEW FUNCTIONALITY

• It is now possible to add custom toolbar items to the graphics form object toolbar.
• The Model Code window is now called Editor Tools. It can also generate form
objects in addition to generating code for Methods.
• There is now support for adding Check Box and Combo Box form objects for some
Model Data Access sources in the New Form Wizard and Editor Tools window.
• There is now extended support for Model Data Access so that you can also use
array- and matrix-valued data, reference properties, and special settings that use
indexed access to data of higher dimensions. Examples of the latter are time and
parameter settings in results features and positions in geometry features.
• Improved copy-paste operation between applications. The following nodes now
support copy-paste: Methods, Utility methods, External libraries, File declaration,
Choice list declaration, Menu item, Menu divider, Menu, Ribbon section, and
Ribbon tab.
• Plot-while-solving (including probe plots) now works from runnable applications.
Enable it as usual in the study step settings, and ensure that the plot group you are
plotting is set as the source for a graphics form object.
• Graphics form objects can now be updated while a command sequence or method
is running.
4 | RELEASE NOTES
• Declarations can be used in arguments of command sequence commands.
IMPROVEMENTS TO THE FORM EDITOR

• Multiple form references are now allowed, so that you can use the same form in
different tabs, for instance.
• Form Collections can now change between a tiled and a tabbed look.
• The Form Editor now supports specifying the margins that are added between
borders of the cells in the grid and Form Objects added to those cells.
• Toggle buttons are now available as a new form object type. A Toggle button acts
as two buttons in one, with one action (command sequence) triggered when the
button is selected, and another action triggered when the button is deselected.
• It is now possible to control the order of sections or panes in Form Collection
objects using Move Up and Move Down buttons to rearrange the selected panes.
• The Data Display object can now be used to show arrays or matrices with LaTeX
syntax. The source of a data display can be array declarations, matrix declarations,
or array and matrix settings from Model Data Access such as material properties.
• The Card Stack form object now supports two type of cards: local cards (which
work as cards do in earlier versions) and card references, which are new and act as
Form References by directly referencing an existing form and displaying that form
when the activating value is set).
• Images of 3D scenes in reports are now connected to application Graphics objects
with the same source or the same geometry.
• You can now switch views from the graphics toolbar. There are also new methods
for switching the view, resetting the view, and for retrieving and setting view
properties such as the camera position.
IMPROVEMENTS TO THE METHOD EDITOR AND APPLICATION LANGUAGE

• The Method Editor can now indent and whitespace-format code. This happens
automatically when you press the TAB key. Unless disabled in preferences, it also
happens when the keyboard focus leaves a method editor window.
• A print command and a printGraphics method are available for printing
graphics from command sequences and methods.
• The methods long getLastComputationTime(), String
getExpectedComputationTime(), and setExpectedComputationTime(String
time) have been added to the public API. The first one is available for the Model
COMSOL MULTIPHYSICS | 5
object and Study objects, while the other two are only available for the Model
object.
• A new formattedTime method converts a time in milliseconds to a string
representation.
• New utility method matrixSize() to compute the size of a 2D matrix.
• Vectorized versions of insert, append, and remove for 1D and 2D arrays have
been added to the application language.
• A new sleep method has been added that makes the caller sleep for a specified time.
• A new setWebPageSource(...) method allows changing the source of a Web Page
form object.
• New built-in functions getScreenHeight and getScreenWidth can be used for
retrieving the size of the primary screen (or of the browser window if Web Client is
used).
• New application language methods deleteFile and copyFile make it possible to
delete and copy files on the server.
• The fileOpen method is now available as a new built-in command in the command
sequence tree.
For details and a full list of methods and language elements, see the Introduction
to Application Builder.
New Functionality in COMSOL Server
PRELAUNCHING OF APPLICATIONS
To allow faster launching of applications, application processes are prelaunched on the
server. Additionally, a number of applications can be preloaded in these processes.
Typically the default settings work for most cases.
• Number of recently run applications to preload: Default: 1. Preloads this number of

the most recently run applications.
• Default preloaded applications: Default: none. The selected applications are explicitly
preloaded. If there is overlap with the most recently run applications, the
overlapping applications are still only loaded once.
6 | RELEASE NOTES
ALLOWING RECONNECTING TO RUNNING APPLICATIONS
A new preference setting, Allow reconnecting to running applications, allows users to
reconnect to previously launched applications and come back where they left off.
Disconnecting and reconnecting is currently not possible while solving, for example.
LAUNCHING APPLICATION AT LOGIN

A new preference setting, Launch application on login, is available but is off by default.
If an application is selected here, it is automatically started when a nonadministrative
user logs in. Only administrators can the see the full Application Library.
LAUNCHING COMSOL CLIENT FROM WEB PAGE

You can now launch an application directly in COMSOL Client through the
COMSOL Server interface by clicking on the menu and selecting Launch in COMSOL
Client.
IMPROVED FILTERING AND SOR TING

You can now filter and sort applications by favorite status, application visibility (Public,
All groups, Private), or by individual groups.
CREATING DESKTOP SHORTCUTS

You can now create shortcuts to quickly access your most frequently used applications
from your desktop. First, open up an application’s detailed view by clicking on its name
or image. From there, you have the option to create a shortcut link that opens the
application either in COMSOL Client or in your web browser.
NEW COMSOL COMMANDS

The following COMSOL command is now available for COMSOL Server:
• comsolserver / comsol server
You can also use the options -prefsdir, -forcegcc, -forcecomsolgcc, -np,
-tmpdir, -autosave, -recoverydir. -ipv6, -c, and -comsolinifile.
New and Improved General Functionality in COMSOL Multiphysics

• It is now possible to interface to external material functions written in C.
• The behavior when the license server connection is lost has been improved so that
it is now possible to save the current COMOL project and also to reconnect to the
license server.
• Searching in the Application Libraries window is now faster.
New Geometry and Mesh Functionality
BETTER CONTROL OF OUTPUT SELECTIONS FROM GEOMETRIES

To get better control over the output selections that are generated from a geometry
feature, the following new settings are available:
• The Selections of Resulting Entities sections in the geometry features now contain the
following:
- A Contribute to list to which you select the cumulative selection that the result of
this feature contributes to. The default is None. You can contribute to a
cumulative selection even if you do not create a selection.
- A New button to the right of the Contribute to list.
- A Resulting objects selection check box, cleared by default. When you select it, a
selection will be created for each applicable geometry entity level (object,
domain, boundary, edge, point) for use within this geometry sequence.
- A Show in physics list (enabled when the Resulting objects selection check box is
selected) with the following options: All levels, Domain selection, Boundary
selection, Edge selection, Point selection, and Off (but only applicable options
appear). If you choose something other than Off, corresponding selections are
created for the finalized geometry and will be available in applicable settings in
the physics nodes and also for materials and meshes, for example. The default is
typically Domain selection to use the selection to define a material, material
property, or domain source, for example.
• If the feature is in a part, the Show in physics list is replaced with a Show in instances
list, and the created selections appear in the selection tables of Part Instance nodes.
• For Part Instance nodes, the output selection tables now include the following
columns:
- Keep, containing a check box, cleared by default. When this is selected, the
selection is kept for use within the geometry sequence.
- Physics or Instances, containing a check box that is selected by default. This check
box is enabled when the Keep check box is selected. Clear it if you do not want
the selection to appear when defining physics, materials, and so on.
8 | RELEASE NOTES
NEW GEOMETRY FUNCTIONALITY
• New Partition Domain operation in 2D and 3D for partitioning of domains by
implicitly defined partitioning curves or surfaces.
• Improved snapping in 2D with separate Snap to Grid and Snap to Geometry
buttons in the Geometry toolbar.
NEW MESHING ALGORITHM

A new, more robust free meshing algorithm is available as an alternative to the previous
Delaunay algorithm. In COMSOL Multiphysics 5.2, this new, alternative Delaunay
algorithm is used if the previous Delaunay algorithm fails when trying to respect the
geometry boundaries. This fallback mechanism is the default as the Automatic setting
in the Method list, available in the new Tessellation section in the Settings window for
Free Tetrahedral nodes. You can also choose Delaunay or Delaunay (alternative) to
always use the previous or alternative Delaunay algorithm, respectively.
MESH PARTS AND MESH IMPORT

The import of external meshes, for use as geometry or mesh, now uses Mesh Parts,
which are reusable global meshing sequences where the imported mesh (from an STL,
VRML, NASTRAN, or MPHBIN/MPHTXT file) can be prepared for modeling with
partitioning operations, for example. The mesh parts make it easier to use imported
meshes as a geometry in COMSOL Multiphysics simulations. You add mesh parts
under Global Definitions>Mesh Parts, and you can then refer to them from an Import
node in a geometry sequence.
Additionally, a new option, Detect adjacent fillet faces, has been added to the settings
for the Import (when using the Detect faces as the Boundary partitioning setting) and
Detect Faces nodes in 3D mesh import sequences. Detecting fillets in imported
meshes should help to create a natural partitioning and make it easier to create a
geometry from the imported mesh.
DIAGNOSTICS FOR INVER TED ELEMENTS IN DEFORMED GEOMETRIES

If inverted mesh elements occur when modeling a deformed geometry, you now get
an error message pointing out inverted elements in the deformed mesh. Also, the
problematic deformed mesh is saved (for visualization) in a Problematic Deformed Mesh
node.
New Operators, Functions, and Definitions
• The molar or molarity unit M = mol/l is now available as [M]. SI prefixes are
supported.
• New operators: fsensimag, for evaluating a functional sensitivity with respect to the
imaginary part of a control variable, and residual (see the section below).
New Functionality in Studies and Solvers

• You can now choose to store only a subset of all DOFs in the solution, thereby
saving storage by excluding parts of the solution that may not be of interest.
• Improved support for FFT:
- Predefined window functions such as Hamming, Hanning (Hann), Blackman,
and Tukey.
- Input time for a forward FFT is now defined using a start time and an end time.
The number of interpolated input solutions is derived from a user-defined
maximum output frequency and is reported in the solver log.
- A new option to not store negative frequencies for real input removes redundant
information in complex output data for a forward FFT.
- Now using the FFT library from Intel’s MKL.
• Two new efficient Rung-Kutta time-stepping solvers are now available in addition
to the Dormand-Prince method: RK34 and Cash-Karp 5.
• You now have access to the latest assembled residual using the residual operator,
which you can plot while solving, for example, to illustrate the convergence process
and find problematic areas.
• The number of eigenvalues or eigenfrequencies and the shift around which to search
for them are now controlled by the physics, which provide suitable values. You can
select check boxes for both of these properties to specify them manually.
• The settings for the initial values of variables solved for has been improved. There is
now a Settings list, from which you can choose to use physics-controlled or
user-controlled initial values.
• The Store solution for all intermediate study steps check box has been added to the
main Study node’s settings. Select it to store solutions for all intermediate study
steps, regardless of which study steps the COMSOL software determines to store,
which may be only the solution from the last study step.
10 | RELEASE NOTES
• The Clear Solutions operation to clear all solutions in the study is now available
from the main Study node’s context menu.
• The identifiers for solution nodes in the solver sequence in the Model Builder tree
now appear in the Model Builder tree. The corresponding identifiers also appear for
Solution data sets under Results, to make visible which data sets contain a specific
solution without opening any Settings windows.
• If you enable the Advanced Study Options, a Distribute parametric sweep check box
in now available in the Settings windows for the study steps, so that you can activate
a distributed parametric sweep without first creating a Cluster Computing study.
New and Improved Results and Visualization Functionality

• A new Annotation plot type is available for adding annotations to any 3D or 2D plot
group. You can add text, including evaluated expressions, at any position in the plot.
• The Far Field plots are now available in COMSOL Multiphysics without any
additional license requirements. Also, for 3D far-field plots, it is now possible to
control the threshold level to provide a more dynamic visualization.
• You can now choose the color for plot titles and color scale values in plots.
• Filled histograms in 1D histogram plots are now available. For discrete outputs, you
can choose Solid as the type of histogram to create filled histogram plots.
• Filters can now cut mesh elements to provide smoother edges and better-looking
plots where a Filter subnode is used to only include part of the geometry. The new
Smooth option is the default setting in the Element nodes to fulfill expression list.
• Support is now included for engineering format for numerical values in color
legends, axes, and tables.
• The arrow base can now be centered in arrow plots.
• The 3D grid box has been improved:
- Intermediate labels are removed when they overlap with neighboring labels.
- The Settings window for Component>Definitions>View>Camera now contains a
Grid section at the bottom. This section contains the same settings as the
corresponding section in the Axis node in 2D.
• It is now possible to hide the 2D axis.
• In the AC/DC Module, RF Module, and Plasma Module, Smith plots are now
available in a Smith Plot Group node where you can add reflection, impedance,
admittance, and table graphs. Smith plots are useful for displaying and relating
complex-valued S-parameters (reflection coefficients) to input impedances or
admittances of antennas, transmission lines, and other network components.
• In the Acoustics Module, octave band plots are now available as Octave Band Plot
nodes in 1D plot groups.
• You can now suppress small real or imaginary parts in the output of results data, for
example, on the General page in the Preferences dialog box.
• In data and plot export, you can now concatenate several data sets by appending
data to the same file. Under Advanced, select Append from the If the file exists list to
concatenate a data set to an existing data file.
• In data export, there is also a Recover list, where you can add numerical recovery
within domains or everywhere for the data export.
• It is now possible to add the unit to the legends in 1D Global plots.
• Using settings under Floquet periodicity in the Advanced section, you can now
incorporate phase changes from Floquet periodicity in Array data sets.
• For easier navigation in reports, internal cross references are now hyperlinks.
Hyperlinking can be disabled for printing.
New Applications in Version 5.2

• The Helical Static Mixer application is a new app, with which you can simulate a
helical static mixer with variable number of blades and dimensions. A concentration
field is also included in the analysis in order to compute the extent of the contact
between two streams injected into the static mixer through half circle-shaped inlets.
• The Transmission Line Calculator application, previously available in the AC/DC
Module, is now included in the COMSOL Multiphysics Application Library. This
app calculates the transmission line parameters, R, L, G, and C; the characteristic
impedance; and the propagation constant for some common transmission line types.
General Backward Compatibility Considerations

COMSOL 5.2 is backward compatible with COMSOL versions 4.0–5.1. To open
models created using COMSOL version 3.5a in version 5.2, follow these steps:
1 Open and then save the COMSOL 3.5a mode using any of the COMSOL versions
4.0–4.4.
2 Open the model saved in versions 4.0–4.4 in COMSOL 5.2.
12 | RELEASE NOTES
Backward Compatibility with Version 5.1
CREATING SELECTIONS
In version 5.2, the selresult property replaces the createselection property.
createselection is still supported for backward compatibility.
SELECTIONS IN PART INSTANCES

For backward compatibility for selections in part instances, there is a Keep
noncontributing selections check box, cleared by default, in the Selection Settings section
of the Settings window for Part Instance nodes. If you select the Keep noncontributing
selections check box, the Keep column is disabled, and the selection is kept if the
Contribute to value is None. In the COMSOL API, the default is that the Keep
noncontributing selections setting is active.
MESH PARTS
For backward compatibility regarding STL/VRML import, and for COMSOL API
compatibility, the old user interface for mesh import (with a specified file name and
import properties) and the corresponding properties are available in the COMSOL
API but not in the COMSOL Desktop, unless it is the active option. This means that
the old user interface can only be reached if the filename is set in combination with the
STL import type (the API type stlvrml), which cannot be done from the new user
interface.
DORMAND-PRINCE 5 TIME-STEPPING SOLVER

The Dormand-Prince 5 Runge-Kutta solver in 5.1 did not use field norm scaling in
version 5.1 when estimating errors. This means that the errors of, for instance, an ODE
might become very small when solved together with a field with many DOFs, and the
time steps taken might then be too large. In version 5.2, the Runge-Kutta solver uses
field norm scaling, which means that old models using the Dormand-Prince 5 might
need tighter tolerances to produce results in earlier versions that are similar to those in
5.2.
MESH IMPORT
The method used to automatically partition the boundary of imported meshes in 3D
has been improved. If you have an existing model, you can work with it without being
affected. However, if you click the Import button of the Import feature in the meshing
sequence, the file is read again and the new partitioning method is used. The exception
is if the Import feature had Boundary partitioning set to Manual. In this case, the
modified parameter values are preserved under the Feature detection setting, which
uses the same algorithm as in version 5.0.
COMSOL tries to map the old selection on boundaries to the new boundaries, but it
is not always possible to do accurately when new faces have appeared or old faces have
disappeared. You may have to manually review and update boundary, edge, and point
selections after reimporting the mesh.
If you have a Java® or MATLAB® program that imports meshes, the number of
geometric entities may have changed compared to older versions.
SECURITY SETTINGS
In version 5.2, the Allow external process and libraries check box on the Security page
in the Preferences dialog box is cleared by default to not allow applications to start
external processes on the computer. The default setting in 5.0 was set to allow such
external processes.
DISPLAY OF MAX/MIN MARKERS IN PLOTS

The display of max/min markers in plots is now off by default to make plots fit better
when using a small graphics canvas. In previous versions, the display of max/min
markers in plots was on by default.
MERGED MPH-FILE FORMAT

The MPHAPP-file and MPH-file formats have been merged since version 5.1, and all
application files use the .mph file suffix. You can still open MPHAPP-files created in
version 5.0.
COMSOL SER VER

The COMSOL Server from version 4.4 is now called the COMSOL Multiphysics
Server.
MESH IMPORT
An edge in an imported mesh that has a common start and end vertex, or that lacks
start and end vertices, is now split into two edges with distinct start and end vertices.
14 | RELEASE NOTES
For meshes where this happens, the numbering of all geometric entities may change
when the mesh is rebuilt.
For MPH-files created in earlier versions, selections are automatically updated with the
new entity numbers.
LOCAL COORDINATE VALUES

For model components created in version 5.0 and onward, the variable names xi1,
xi2, and xi3 are reserved for the predefined local coordinate variables. To access the
previous behavior (that local coordinate variables do not exist), you can use the API
method model.modelNode(<tag>).defineLocalCoord(false).
TR A N S P O R T O F D I L U T E D S P E C I E S
®
Running Java -files from previous versions may fail due to the new default name for
the Transport of Diluted Species interface. This can be avoided by adding a command
that specifies the identifier in accordance with the name of the interface. For example,
when creating an interface using the old name (chds):
model.physics().create("chds", "DilutedSpecies", "geom1",

new String[][]{{"c"}});
Add the following line to specify the identifier accordingly:
model.physics("chds").identifier("chds");
The Java® API syntax for creating and accessing vectors and tensors in the Transport
of Diluted Species interface has changed as well as the syntax for setting physics
properties. This does not affect MPH-files. See the backward compatibility notes for
the Chemical Reaction Engineering Module for additional information that also
applies to the Transport of Diluted Species interface.
DOCUMENTATION
The COMSOL Multiphysics Programming Reference Manual replaces the
COMSOL API for use with Java ® Reference Manual.
Backward Compatibility with Version 4.3b
MODEL NODES ARE NOW COMPONENT NODES

The Model nodes (as they were called in previous versions), which contained separate
model components in a model file, are called Component nodes in version 5.2.
CHANGES TO PARAMETRIC SWEEPS
Old models that use stationary parametric sweeps are loaded with the Reuse solution
for previous step list set to Yes. The Run continuation for list is set to the parameter used,
unless the continuation algorithm would not have been used for this model in previous
versions (for example, if multiple parameters are used or if the parameter list is not
monotonous).
CHANGES TO THE PHYSICS SELECTION IN STUDY SETTINGS

The names of the states of a physics in the physics tree, which you can modify under
Physics and Variables Selection in the study steps’ settings windows, have changed:
• Provide Degrees of Freedom is now called Disable in Solvers.

• Disable is now called Disable in Model.
REVISED FORMULATION FOR LAMINAR INFLOW AND LAMINAR OUTFLOW

The formulations of the laminar inflow and laminar outflow conditions have been
corrected. The modified formulation gives a more accurate mass flux. Some models
may now produce a slightly different flow field.
Laminar inflow and laminar outflow are available in the fluid flow physics interfaces in
the following modules:
• Batteries & Fuel Cells Module

• CFD Module
• Corrosion Module
• Electrochemistry Module
• Electrodeposition Module
• Heat Transfer Module
• Microfluidics Module
• Plasma Module
• Subsurface Flow Module
NEW DEFAULT FOR GRAPHICS OPTIMIZATION

Under Graphics and Plot Windows in the Preferences dialog box, the default settings in
the Optimize for list is now Quality instead of Performance.
16 | RELEASE NOTES
VE L O C I T Y / A C C E L E R A T I O N I N T E G R A T I O N V A R I A B L E
For the Solid Mechanics interface (and all related multiphysics interfaces) and the
Truss interface, a help variable u0 (velocity integration variable) is used in Prescribed
Velocity and Prescribed Acceleration features for Time Dependent study types. This
variable computes the displacement for each point where the condition is prescribed.
It is changed to use the full feature scope in order to avoid collisions in cases where
several such features exist within the same model component. As a result, when
opening and running an old model that uses such features together with a segregated
solver, an error message appears, stating that not all dependent variables occur in at
least one of the segregated solver steps. The relevant action is to manually add the
velocity integration variable to the segregated step containing the corresponding
displacement field. Alternatively, you can regenerate any affected solver sequence.
MATH LIBRARIES ON AMD PROCESSORS

MKL is now the default math library on AMD processors. Switching to the ACML
math library for AMD processors might improve performance in some cases.
Backward Compatibility with Version 4.3a
ERROR ESTIMATION CHANGES IN THE SOLVERS

The Automatic method for Check error estimate for direct linear solvers and for Validate
error estimate has changed. For nonlinear problems and for time-dependent problems,
the underlying nonlinear solver does not accept termination for a linear solution step
that does not fulfill the error estimate (unless the step size is very small). The
motivation for this change is that the old method can lead to premature termination
of the nonlinear solution process, which in some cases introduces a large error for the
computed solution.
This change can cause the solvers to take more nonlinear steps for stationary problems
and more time steps for time-dependent problems, and it may also lead to convergence
problems. For such cases, use the No method to obtain the old behavior. However,
doing so can hide numerical problems and potentially lead to large numerical errors.
TE R M I N A T I O N C R I T E R I O N F O R S T A T I O N A R Y S O L VE R S
The default termination criterion for stationary solvers has changed. In 4.3a the
settings corresponded to Solution; now the default is Solution or residual. This change
in default termination criterion might affect models created in earlier versions of
COMSOL if you regenerate the solver sequence in 5.2.
BACKWARD EULER INITIALIZATION TIME STEP
A new setting in the Advanced section of the settings window for the Time-Dependent
Solver, called Fraction of initial step for Backward Euler, provides an option for entering
a dimensionless quantity that determines the size of the time step for the backward
Euler method (in terms of the initial step). This value can improve the accuracy of the
initialization step but can also affect the start-up of some models. The default value is
0.001 (this differs from earlier versions, which used a value of 1). When opening
models created in version 4.3a or earlier, the value for this fraction is set to 1 to
maintain compatibility with those versions.
LOADING EXTERNAL PHYSICS BUILDER JAR FILES

External physics builder JAR archives compiled with earlier versions of COMSOL
include a manifest file that contains a reference to the CDO library, which has been
removed. To load such JAR files in version 5.2, you must first do one of the following:
• Delete the META-INF/MANIFEST.MF file in the archive source on the file

system, and then recompile the JAR archive.
• Manually remove the line with org.eclipse.emf.cdo in the META-INF/
MANIFEST.MF file in the JAR archive. You can do this directly in a file archive
manager such as 7-Zip or similar.
HIGHLIGHTING GEOMETRY OBJECTS WHEN DRAWING IN 2D

When you have drawn one geometry object on top of another object, toggling of the
highlighting of these objects occurs when you click several times. It is important that
you click without moving the cursor in a position where the objects overlap in order
to toggle. For example, if you draw a circle (C1) and then draw another smaller circle
(C2) inside of C1, then a first click on C2 may highlight C1, while a second click,
without moving the cursor, highlights C2. In previous versions, you would move the
cursor after the first click in order to highlight C2.
N E W TE R M I N O L O G Y F O R C O N S T R A I N T TY P E S
The following constraint types have new names in version 5.2:
• Bidirectional, symmetric is now Apply reaction terms on: All physics (symmetric).
• Unidirectional is now Apply reaction terms on: Individual dependent variables.
18 | RELEASE NOTES
WE A K CO NS TRA IN TS UPDATE F OR TH E L AMINA R FL OW INT ERF ACE
The weak constraint formulations for the following boundary conditions have been
updated:
• Symmetry
• The Slip option in the Wall feature
These boundary conditions are now formulated using the same set of Lagrange
multipliers as all the other boundary conditions in the Laminar Flow interface. The
Lagrange multiplier un_lm has been removed.
Models saved in version 4.3 will include un_lm until the model is re-solved. In some
cases, occurrences of un_lm in the solver sequence must be replaced manually. This is
the case if un_lm was the only Lagrange multiplier component in a segregated group
or the only Lagrange multiplier component of a Vanka smoother. Alternatively, you
can generate a new automatic solver sequence. Models saved in versions earlier than
4.3 must either be re-solved in version 5.2 for postprocessing, or opened and re-saved
in version 4.3 before being opened in version 5.2.
AUTOMATICALLY CREATED IDENTITY/CONTACT PAIRS

The pairs created by the Form Assembly feature in the geometry can differ from their
forms in 4.3 for certain 3D and 2D geometries. The pairs should now be the same for
the CAD and COMSOL representations.
AC/DC Module
New Functionality in Version 5.2
EFFECTIVE NONLINEAR MAGNETIC CONSTITUTIVE RELATIONS FOR

FREQUENCY DOMAIN STUDIES
There is now support for effective magnetic constitutive relations, in the form of
effective HB or BH curves. This functionality can be used to model nonlinear
magnetic materials, such as saturable metals, in a frequency domain study by
approximating them with an effective inhomogeneous linear material. You can use the
formulation to compute the approximate (first-order harmonic) response of the
nonlinear material subject to time-harmonic excitation, avoiding the computational
cost of a full transient analysis.
EXTERNAL NONLINEAR MAGNETIC MATERIALS

The Magnetic Fields physics interfaces can now use external material models
implemented as user-defined dynamic libraries, increasing the flexibility and the
modeling capabilities of the physics interfaces. This functionality is built into the
External Material feature and makes it possible to define custom advanced material
models such as hysteresis models, state-dependent models, or even models using a
different discretization scheme.
EFFECTIVE NONLINEAR MAGNETIC CURVES CALCULATOR

Companion to the new Effective Nonlinear Constitutive Relations formulation
described above, this utility application can be used to compute the effective HB and
BH curves for a material given the interpolation data for the BH or HB curve. The
computed effective curves are calculated using one of two different energy methods
defined by the user. The resulting effective material properties can be exported as a
COMSOL Material Library file.
TO U C H S C R E E N S I M U L A T O R
This application computes the capacitance matrix of a touchscreen in the presence of
a human finger phantom, where the position and orientation of the finger is controlled
via input parameters. Intended as a tool for early proof-of-concept in capacitive touch
20 | RELEASE NOTES
device development, the resulting capacitance matrix is evaluated as well as the electric
field norm.
MAGNETIC PROSPECTING
With this application, you can simulate the effect of a deposit of magnetic ore (such as
magnetite or hematite) on the geomagnetic field in the region. Users can import
heightmap images or Digital Elevation Model (DEM) files to define the orography of
the region and can download the geomagnetic field data for the specified location from
the Internet. The computation generates a plot of the magnetic anomaly on the
Earth’s surface and numerical data on the expected anomaly at specified measurement
locations in the region.
INDUCTION HEATING OF A STEEL BILLET

This application simulates an induction heating system for the reheating of steel billets.
Specifically, it simulates a billet traveling at constant speed through AC coils, heating
up due to the induced eddy currents. The geometry of the system, including the
number of coils, their placement, and the billet cross section, is parameterized to make
those parameters available as inputs. The app contains preset billet cross sections, and
additional cross sections can be imported as geometry parts from other COMSOL
Multiphysics files. The temperature range in the billet and the power balance of the
system are presented after the computation as analytical results, together with visual
plots of the electric current density and the temperature field.
Backward Compatibility with Version 5.0 and older
COIL CURRENT CALCULATION

When models saved in previous versions are opened in version 5.1, the Coil Current
Calculation study steps are migrated to Coil Geometry Analysis. Old solutions stored
in the model can still be postprocessed, and the solver sequence will be regenerated
automatically (with a Stationary solver) the first time the study is solved.
The Coil Geometry Analysis study step (formerly Coil Current Calculation) will now
solve for all the coils in the active interfaces. To solve only for specific coils (specified
using the CoilName property), set the SpecifyCoil property to 1.
The default solver sequence generated by the Coil Geometry Analysis is different in
version 5.1. Code that accesses specific solver features in the generated solver sequence
may need to be reviewed.
AC/DC MODULE | 21
OTHER COIL IMPROVEMENTS
Harmonic Perturbation subnodes under coil features (Single-Turn Coil, Multi-Turn
Coil) are now global features, so the call to the create method should use the
appropriate space dimension (-1):
model.physics("mf").feature("stcd1").create("hp1",
"CoilHarmonicPerturbation", -1);
Global features have no selections, so code that accesses the selection of the Harmonic
Perturbation features may need to be reviewed.
Some of the improvements in the 3D Multi-Turn Coil features may require a review
of existing code that uses the COMSOL API.
Subnodes required to set up the coil features are now added automatically. Existing
code that uses the coil features may need to be updated.
The parameters eCoil and length have been moved from the Multi-Turn Coil
features (boundary and domain) to the new subfeature UserDefinedCoilGeometry.
OTHER IMPROVEMENTS
New functionality introduced in version 5.1 is disabled by default when opening
models created in previous versions:
• New boundary conditions for Gauge Fixing features

• Accurate coil voltage calculation
This functionality can be enabled using the appropriate inputs in the Settings window.
Refer to the documentation for the individual features for more details.
22 | RELEASE NOTES
Acoustics Module
OCTAVE BAND PLOTS

A new dedicated acoustics plot, called Octave Band Plot, for representing frequency
domain transfer functions, responses, sensitivity curves, insertion and transmission
loss. The plot has several built-in acoustics-specific features such as predefined
weighting (Z-, A-, C-, and user-defined weighting) and the possibility to plot the
desired response function in octaves, 1/3 octaves, or as a continuous curve. The band
option corresponds to plotting the average or integrated value of, for example, the
squared pressure over a given frequency band defined by the mid-frequency and the
bandwidth. The data input to an octave band plot is a frequency-domain solution; for
example, it can be the acoustic pressure resulting from a Frequency Domain study or
a parametric frequency sweep. The plot will automatically plot a given expression type
on the dB scale; this simplifies postprocessing as it is no longer necessary to define the
expression as a variable. The input to geometric entity of the plot can be global, point,
edge, boundary, or domain. On the latter three, an average is automatically performed.
This means that, for example, an average power at an inlet is easy to define and plot.
DISSIPATED EN ERGY VARIABLES IN PRESSURE ACOUSTICS

The dissipated power density for all fluid models has been introduced for Pressure
Acoustics, Poroacoustics, and Narrow Region Acoustics. The variable is called
acpr.Q_pw and is located under the Heating and losses section in the Replace
Expressions menu when working with results. The expression is valid in the plane wave
limit for traveling waves. This predefined variable is used in the Focused Ultrasound
Induced Heating in Tissue Phantom application where dissipated acoustic energy heats
a tissue phantom.
N O R M A L VE L O C I T Y A N D N O R M A L D I S P L A C E M E N T B O U N D A R Y
CONDITIONS IN PRESSURE ACOUSTICS
In the Pressure Acoustics interfaces, the normal acceleration boundary condition is
now supplemented by two new boundary conditions for prescribing a normal velocity
or a normal displacement. These conditions simplify the modeling procedure when
modeling sources in acoustics. The source in the Generic 711 Coupler—An Occluded
Ear-Canal Simulator application, for example, is defined via a normal displacement.
ACOUSTICS MODULE | 23
UPDATED INTENSITY VARIABLES IN ALL ACOUSTICS INTERFACES
The intensity variables have been updated in all acoustics interfaces. The variables are
now consistent across physics and across study types. Intensity is defined in the
frequency domain (averaged values over one period) and the so-called instantaneous
intensity is defined in the time domain. The intensity variables in the Thermoacoustics
and Linearized Navier-Stokes interfaces now also include the viscous stress
contributions. The variables are available for postprocessing when clicking the Replace
Expression button.
MINOR UPDATES AND BUG FIXES

Several minor updates and bug fixes are also available with this release:
• PARDISO solver used in Thermoacoustics and Linearized Navier-Stokes now uses

the Multithreaded forward and backward solve option by default. This setting
provides minor speedup in general, with the most noticeable effect in
eigenfrequency problems.
• Model inputs are updated in Pipe Acoustics.
• An entropy variable, ta.s_entropy, is now available in Thermoacoustics for use in
postprocessing.
• The user-defined band structure behavior is updated in the Acoustic Diffusion
Equation interface.
• The dissipated thermal and viscous power density variable has been updated in
Thermoacoustics.
• The slip condition is updated in Thermoacoustics, Frequency Domain to avoid
locking on curved boundaries.
• The slip condition is updated in Linearized Navier-Stokes to avoid locking on
curved boundaries.
ABSORPTIVE MUFFLER DESIGNER

This is an application for studying and designing a simple resonant muffler with a
porous lining. The analysis includes both inductive and resistive damping in the
muffler. This application enables the user to study the results of modifying the
dimensions of the muffler, the ambient working conditions, and the material
properties of the porous liner. The application is based on the Absorptive Muffler
model located in the Automotive folder in the Acoustics Module.
24 | RELEASE NOTES
ONE-FAMILY HOUSE ACOUSTICS ANALYZER
The One-Family House Acoustics Analyzer is used to asses noise propagation in
coupled rooms inside a two story family house. The house consists of 10 rooms. The
application will determine the sound pressure level (SPL) distribution in the house
given a number of sources interactively placed in the house. The application illustrates
a classical room acoustics problem where engineers or architects want to determine the
noise environment in a system of coupled rooms like homes, office spaces, or
workshops. This is, for example, important when verifying that an acoustic
environment complies with noise and work regulations. An application, inspired by the
one presented here, will allow users to get on-site and test different noise source
scenarios, different wall insulation characteristics, and compare with live
measurements. All in all, such an application allows for a faster characterization and
solution. The application lets you place, remove, and define multiple acoustic sources
in different rooms of the house in order to determine the resulting sound pressure level
distribution. The acoustics are modeled using the Acoustic Diffusion Equation physics
interface, which is fast and efficient at determining SPL distributions.
ORGAN PIPE DESIGN

Working with acoustics, designing instruments or other sound transducing devices it
can be interesting to not only look at a simulated frequency response but also listen to
the sound generated by the device. This is auralization of the acoustics of the device.
This demonstration application shows how this can be achieved using a simple model
of an organ pipe. Changing the organ pipe design will, for example, result in a different
pitch, which can be auralized. The organ pipe is modeled using the Pipe Acoustics,
Frequency Domain interface. The application allows users to analyze how the first
fundamental resonance frequency varies with pipe properties such as radius and wall
thickness, and with the ambient pressure and temperature. The full frequency response
is found including the fundamental frequency and the harmonics. Using a Java
method, the results analysis is extended beyond the built-in functionality. The
application will detect and find the location and amplitude of all harmonics in the
response. It will also allow users to listen to the sound and pitch of a given organ
design.
ACO U STIC R EF LE CTION A NA LYZE R F OR A WATE R- SE DIME N T IN TER FAC E

This application analyzes the reflection and absorption coefficients of plane acoustic
waves, scattering off a water-sediment interface, at different frequencies and angles of
incidence. Moreover, the random-incidence absorption coefficients are determined for
the studied frequencies. The ability of the Poroelastic Waves interface to model the
coupled acoustic and elastic waves in any porous substance (Biot’s theory) is used to
describe the water-sediment system. The application is based on the Acoustic
Reflections Off a Water-Sediment Interface model located in the Underwater
Acoustics folder.
NEW MODELS IN THE APPLICATION GALLERY

Four new applications have been added to the Application Gallery found on
www.comsol.com:
• Vibrating Plate in a 2D Viscous Parallel Plate Flow: This is a small 2D

demonstration model that couples the linearized Navier-Stokes Frequency Domain,
the Solid Mechanics, and the Creeping Flow physics interfaces to model the
vibrations of a plate located in a 2D viscous parallel plate flow.
• Apartment Acoustics Analyzed using the Acoustic Diffusion Equation: The sound
distribution from a TV in a two-room apartment is computed. The simulation
demonstrates the use of the Acoustic Diffusion Equation interface to get a quick and
simple estimate of the local sound pressure level. For increased accuracy, an
analytical expression for the direct sound is added in the living room.
• Acoustic-Solid Interaction with two Perfectly Matched Layers (PMLs): This small
tutorial model shows how to set up a model with both a solid mechanics and a
pressure acoustics domain including two perfectly matched layers (PMLs), one for
each physics.
• Shape Optimization of a Tweeter Waveguide: This application illustrates how to use
the optimization capabilities in COMSOL Multiphysics to automatically develop
novel designs satisfying critical design constraints. The model optimizes a simple
speaker geometry. Examples of constraints could include the radius of the
loudspeaker or a desired minimum achievable sound-pressure level.

• Plane Wave Radiation has been removed from Linearized Potential Flow, Frequency
Domain and Linearized Potential Flow, Transient in 1D axisymmetry.
• A new Lagrange multiplier variable (ta.lm_slip) has been added to the slip
conditions in Thermoacoustics, Frequency Domain. If you run a model with a
segregated solver you will need to add this variable to the group containing the
velocity degree of freedom. You can also regenerate the default solver. In the Java
API, you need to add comp1_ta_lm_slip to the declaration of the segregated step,
for example, adding:
26 | RELEASE NOTES
model.sol("sol1").feature("s1").feature("se1").feature("ss1").
set("segvar", new String[]{"comp1_p", "comp1_u", "comp1_T",
"comp1_ta_lm_slip"});
• The Waveguide end impedance option has been removed from 2D, 1D
axisymmetric, and 1D as it had no physical meaning in these space dimensions.
Backward Compatibility with Version 4.3a and Older

• The old perfectly matched layer (PML) node that is located under the physics node
will be discontinued in the next version of COMSOL. From the next version on
only the PMLs defined under the Definitions node are supported.
• In order for the old perfectly matched layer (PML) to be backward compatible, the
PML feature must be placed after any domain Monopole Source, domain Dipole
Source, or Background Pressure Field features.

• The symbols for volume sources for the acoustics monopole (Qm; was Q) and dipole
(qd; was q) volume sources in pressure acoustics have changed.
• The default value for the Typical wave speed property in the Acoustic-Piezoelectric
Interaction, Frequency Domain interface has changed to 343 m/s.
Batteries & Fuel Cells Module
COMPENSATE FOR TUBE VOLUME IN ELECTROLYTE IN THE CURRENT

DISTRIBUTION ON EDGES, BEM INTERFACE
By using this new feature, the effect of the volume of the tubes is now considered in
the electrolyte charge transfer equations.
NAME CHANGES OF ELECTROLYTE-ELECTRODE BOUNDARY INTERFACE

AND ELECTROLYTE-ELECTRODE DOMAIN INTERFACE NODES
The new names are Electrode Surface and Internal Electrode Surface, respectively.
WEAK CONSTR AINT S NOW USED BY DEFAULT IN PRIMAR Y CURRENT

DISTRIBUTION BY THE ELECTRODE SURFACE AND INTER NAL
ELECTRODE SURFACE NODES
To improve accuracy the default constraint setting has been change to use Weak
Constraints in Primary Current Distributions.
The change may cause Java API backward compatibility issues with regards to slightly
different results or issues with the convergence of certain iterative solvers.
STOICHIOMETRY SECTION IN THE DOUBLE LAYER CAPACITY AND

POROUS DOUBLE LAYER CAPACITY NODES
A Stoichiometry section has been added to the Double Layer Capacity and Porous
Double Layer Capacity nodes in the Tertiary Current Distribution, Nernst-Planck, and
Electroanalysis interfaces. Use the stoichiometry coefficients to control how ions flow
into and out from the double layer. The unused stoichiometry field input in the
(Porous) Electrode Reaction node has also been removed.
C Y C L I C VO L T A M M E T R Y N O W I N C O R P O R A T E S E X P L I C I T E V E N T S A T T H E
VERTEX POTENTIALS
By the use of explicit events at the vertex potentials the accuracy of cyclic voltammetry
in the Electroanalysis interface has been improved. Old models will open with the
settings turned off. To turn them on, you need to enable Show Advance Physics Options
in the Model Builder window and click the corresponding check box under Advanced
Settings in the Electrode Surface node.
28 | RELEASE NOTES
different results.
BO UND AR Y ELECTRIC PO TENT IA L IN IT IA L VA LUE CAN N OW BE SET IN

THE ELECTRODE CURRENT AND THE ELECTROLYTE CURRENT NODES
You can now set the initial value for the boundary potential in the Electrode Current
and Electrolyte Current nodes. By changing the initial value, convergence properties
of certain models can be improved.
NEW HARMONIC PER TURBATION SUBNODES FOR THE ELECTROLYTE

CURRENT AND ELECTRODE CURRENT BOUNDARY NODES, ALONG WITH
DEFAULT NYQUIST PLOTS
Harmonic perturbation (used in AC Impedance studies) is now supported by the
Electrolyte Current and the Electrode Current nodes.
Default Nyquist plots are also added by default when using these nodes.
MATERIAL LIBRARY UPDATE OF BATTERY ELECTROLYTE

CONDUCTIVITIES
The conductivity of battery electrolytes has been set to 10−6 at zero electrolyte salt
concentrations. The change improves accuracy and convergence for battery models
running at very high current densities during prolonged times.
Note that old models will not be affected by the change. You need to remove and add
the electrolyte material to your model again in order to apply new material data.
EXTENDED PHYSICS INTERFACE SUPPOR T FOR THE FLOW COUPLING

MULTIPHYSICS NODE
The Multiphysics Flow Coupling node now supports Darcy’s Law interface as Source,
and the Transport of Concentrated Species, Electroanalysis and the Tertiary Current
Distribution, Nernst-Planck interfaces as Destination.
The change may have implications on the Java API backward compatibility since the
Source or Destination may now have to be explicitly specified.
E X T E N D E D P HY S I C S I N T E R F A C E S U P P O R T F O R T H E TE M P E R A T U RE
COUPLING MULTIPHYSICS NODE
The Multiphysics Temperature Coupling node now supports Darcy’s Law; Transport
of Concentrated Species; Primary Current Distribution; Secondary Current
BATTERIES & FUEL CELLS MODULE | 29

Distribution; Tertiary Current Distribution, Nernst-Planck; and the Lead-Acid
Battery interfaces as Destination.
The change may have implications on the Java API backward compatibility since the
Destination may now have to be explicitly specified.
New and Updated Applications in Version 5.2
NEW DIFFUSE DOUBLE LAYER APPLICATION

A new diffuse_double_layer application has been added to the application library.
UPDATED AND REVISED APPLICATIONS

• The ac_fuel_cell now uses the new Harmonic Perturbation subnode to
Electrode Current node.
Backward Compatibility with Previous Versions
TR A N S P O R T O F C O N C E N T R A T E D S P E C I E S I N T E R F A C E
See Chemical Reaction Engineering Module.
SURFACE REACTIONS INTERFACE

30 | 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 the CAD Import Module, Design Module, and
LiveLink Products for CAD in Version 5.2
The CAD file import functionality included with these products has been extended to
support new versions for some of the supported file formats (see under Read from File,
CAD, on www.comsol.com/products/specifications/cad).
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 5.2 includes an upgraded version of the Parasolid kernel and
as a result, a number of stability issues have been fixed, which makes the import of
CAD models and solid operations more robust.
New Functionality in LiveLink™ f o r Inventor®

The LiveLink™ interface add-in now loads much faster when starting Autodesk®
Inventor®.
New Functionality in LiveLink™ f o r Solid Edge®

The LiveLink™ interface now supports Solid Edge® ST8.
B A C K W A R D C O M P A T I B I L I T Y W I T H VE R S I O N 5 . 0
The default value for the keepfree property of the LiveLinkSolidEdge function is
now set to on. Previously the default was set to off.
New Functionality in LiveLink™ f o r SOLIDWORKS®

• The LiveLink™ interface now supports SOLIDWORKS® 2016.
• Global variables defined with units in SOLIDWORKS can now be synchronized.
CAD IMPORT MODULE, DESIGN MODULE, AND LIVELINK PRODUCTS FOR CAD | 31
B A C K W A RD C O M P A T I B I L I T Y W I T H VE R S I O N 5 . 0
The default value for the keepfree property of the LiveLinkSOLIDWORKS function
is now set to on. Previously the default was set to off.
32 | RELEASE NOTES
CFD Module
T H R E E - P H A S E F L O W, P H A S E F I E L D I N T E R F A C E
The new Three-Phase Flow, Phase Field interface can be used to model the flow and
interaction of three different, immiscible fluids when the exact positions of the
interfaces separating the fluids are of interest, also known as separated flow with surface
tracking. The fluid-fluid interfaces are tracked using a ternary phase field formulation
that accounts for differences in the fluids’ densities and viscosities, and the formulation
includes the effects of surface tension. The phase-field method can handle moving
contact lines on no-slip boundaries. The movement of the fluid-fluid interfaces is
determined by the minimization of free energy.
TE R N A R Y P H A S E F I E L D
The corresponding Ternary Phase Field interface, used to track moving interfaces
between three immiscible phases in the CFD Module and the Microfluidics Module,
is also found as a standalone Mathematics interface.
R O T A T I N G M A C H I N E R Y F L U I D F L O W I N T E R F A C E S : TU R B U L E N T F L O W,
A L G E B R A I C Y P L U S A N D TU R B U L E N T F L O W, L - V E L
Two new physics interfaces have been added in the Rotating Machinery Fluid Flow
branch: the Turbulent Flow, Algebraic yPlus interface and the Turbulent Flow,
L-VEL interface. For these algebraic turbulence models, the turbulent viscosity is
determined using two different extensions of the logarithmic wall law, and the local
Reynolds number based on the distance to the nearest wall. Algebraic turbulence
models are computationally cheaper and more robust but generally less accurate than
transport equation turbulence models, such as the k-ε and k-ω turbulence models.
STATIONARY FREE SURFACE

This new feature in the Rotating Machinery Fluid Flow interfaces estimates the
deformation of a free surface due to the combined effects of fluid flow and volume
forces (such as gravity) when solving a quasi-steady flow using the Frozen Rotor study
type. An average pressure is applied on the selected boundary in the fluid-flow
computation, and the surface elevation is evaluated from the resulting pressure
variations on the boundary in a postprocessing study step.
CFD MODULE | 33
INKJET
The Inkjet application demonstrates how to model the flow of a droplet produced by
an inkjet nozzle, such as one in a printer. An ink droplet is ejected from the nozzle and
travels through air until it hits the target. The fluid flow is incompressible, and the
movement of the interface between the droplet and the ambient air is modeled using
the level-set method. The user can specify liquid properties in designated input fields,
and the geometry—including nozzle, reservoir, compartment, and outlet—is fully
parameterized. In addition, the user can control the injection pulse profile by varying
the pulse interval and smoothing time.
NACA AIRFOIL OPTIMIZATION

The lift and drag coefficients of a fully parametrized NACA airfoil can be computed in
this application. It incorporates an Optimization interface to find the wing geometry
parameters that maximize the lift-to-drag ratio. The flow regime, laminar or turbulent,
is detected automatically based on the user-defined Reynolds number. Low Reynolds
number simulations are performed with the Laminar Flow interface. Higher Reynolds
number simulations use the Spalart-Allmaras turbulence model, which has been
specifically developed for airfoil design simulations.
WA T E R TRE A T M E N T B A S I N
This application runs simulations of a water treatment basin used for the mixing and
reacting of an added chemical species (for example, chlorine). The supported
geometries are rectangular basins with baffles attached to the sides of the basin in
vertical or horizontal configurations, and the inlet and outlet channels can have a
circular or rectangular cross section. The graphical results show the flow field and the
concentration field of the added chemical species given a first-order chemical reaction,
and the numerical results give the space time, half life, and pressure drop.
EULER-EULER MODEL, LAMINAR FLOW

The results from models using the Euler-Euler Model, Laminar Flow interface may
differ in version 5.1 due to the more general formulation for the viscous stress in the
momentum equation for the dispersed phase. The behavior in the old formulation may
be reproduced by dividing the dispersed phase viscosity, μD, by the variable
ee.phidPos.
34 | RELEASE NOTES
TR A N S P O R T O F C O N C E N T R A T E D S P E C I E S
INLET FEATURE
The old inlet features still exist, but have been excluded from the physics context
menu. Old models retain old Inlet features, but adding a new Inlet feature comes with
the new functionality. Old Java files create Inlet features of the old types. The Normal
stress condition can still be prescribed on inlets using an Open Boundary or a
Boundary Stress feature.
N E W N A M E S F O R M A S S TR A N S P O R T I N T E R F A C E S
Running Model Java-files from previous versions may fail due to the new default name
in the Transport of Diluted Species and Transport of Concentrated Species interfaces.
This can be avoided by adding a command that specifies the identifier in accordance
with the Name of the interface. For example, when creating a Transport of
Concentrated Species interface using the old name (chcs):
model.physics().create("chcs", "ConcentratedSpecies", "geom1",

new String[][]{{"wO2", "wH2O", "wN2"}});
Add the following line to specify the identifier accordingly:
model.physics("chcs").identifier("chcs");
WALL DIST ANCE EQUATION

The old Continuity node is still included in the interface but is excluded from the
feature list. Hence, old models retain their old Continuity node when opened in
version 5.0 and old Model Java-files are fully backward compatible.

ee.phidPos.
CFD MODULE | 35
B U B B L E - I N D U C E D TU R B U L E N C E I N B U B B L Y F L OW
Equation terms accounting for bubble-induced turbulence in the Turbulent Bubbly
Flow interface are now added correctly. As a result, models with Turbulent Bubbly
Flow now show higher levels of turbulence.

ee.phidPos.
MIXTURE MODEL EQUATION FORMULATION

The Mixture Model equation formulation has been revised to increase the accuracy
and facilitate the solution of a number of problem setups. As a consequence of the
revision, old models can return slightly different results in the new version than they
did in previous versions. As part of the revision, a penalty diffusion on the dispersed
phase has been deactivated, because after the other revisions, it made the equation
system too nonlinear and therefore more difficult to make it converge. However, you
can still activate the penalty diffusion in the Advanced section of the Mixture Model
physics interface settings.

ee.phidPos.
36 | RELEASE NOTES
FLUID-STRUCTURE INTERACTION
The Fluid-Structure Interaction (FSI) multiphysics interface has been updated. The
separate vWall field is no longer required and has been removed. FSI models from 4.3
and earlier versions that include Time Dependent study steps will be affected in the
following ways:
• Model files for Java will fail. Any reference to the vWall field must be removed.
• Models older than 4.3 (that is, 4.2a and older) must clear their time-dependent
solutions when opened in 5.2. An alternative procedure is to:
- Open the model in version 4.3.
- Right-click any study with a time-dependent study step and select

U pdate Solution.
- Save the model.

- Open the model in version 5.2.
FSI models with only stationary study steps will not be affected. Note that vWall will
still be available as a variable. Hence, references to fsi.vWall made in, for example,
another physics still work.
BRINKMAN EQUATIONS AND FREE AND POROUS MEDIA FLOW

The momentum equations in the Brinkman Equations interface and the Free and
Porous Media Flow interface have been corrected. The term −Qbr·u/εp2in the right
hand side previously lacked the factor 1/εp2, where εp is the porosity.
REACTING FLOW INTERFACE NAME CHANGE

The Reacting Flow interfaces have been renamed Reacting Flow in Porous Media. If
you open a model using either the Reacting Flow, Diluted Species or the Reacting
Flow, Concentrated Species interface, the interface is renamed Reacting Flow in
Porous Media.
WE A K CO NS TRA IN TS UPDATE F OR FL UID F LOW IN TER FA CES

The weak constraint formulations for the following boundary conditions in the
following interfaces have been updated:
CFD MODULE | 37
Laminar Flow and Creeping Flow
• Symmetry
• The Slip boundary condition in the Wall feature
Turbulent Flow, k- ε and Turbulent Flow, k- ω

• Symmetry
• The Wall functions boundary condition in the Wall feature
• The Moving wall (wall functions) boundary condition in the Wall feature
Turbulent Flow, low-Re k- ε and Turbulent Flow, Spalart-Allmaras

• Symmetry
• The Slip option in the Wall feature
Rotating Machinery
• Symmetry
• The Wall functions boundary condition in the Wall feature (turbulent flow only)
• The Moving wall (wall functions) boundary condition in the Wall feature (turbulent
flow only)
• Rotating Wall (turbulent flow only)
Bubbly Flow
• Symmetry
• The Slip boundary condition for the liquid phase in the Wall feature
• The Wall functions boundary condition for the liquid phase in the Wall feature
(turbulent flow only)
Mixture Model
• Symmetry
• The Slip boundary condition for the mixture in the Wall feature
• The Wall functions boundary condition for the mixture in the Wall feature
(turbulent flow only)
Euler-Euler Model, Laminar Flow

• Symmetry
• The Slip boundary condition for either phase in the Wall feature
• The Slip boundary condition for either phase in the Inlet feature
38 | RELEASE NOTES
Brinkman Equations and Free and Porous Media Flow
• Symmetry
Two-Phase Flow, Level Set

• Symmetry
• The Wetted wall boundary condition in the Wall feature
• The Moving wetted wall boundary condition in the Wall feature
• The Wall function boundary condition in the Wall feature (turbulent flow only)
flow only)
Two-Phase Flow, Phase Field

• Symmetry
• The Wall function boundary condition in the Wall feature (turbulent flow only)
flow only)
Non-Isothermal Flow and Conjugate Heat Transfer

• Symmetry
• The Wall functions boundary condition in the Wall feature (turbulent flow k-ε and
turbulent flow k-ω only)
flow k-ε and turbulent flow k-ω only)
High Mach Number Flow

• Symmetry
• The Wall functions boundary condition in the Wall feature (turbulent flow k-ε only)
flow k-ε only)
CFD MODULE | 39
These boundary conditions are now formulated using the same set of Lagrange
multipliers as all the other boundary conditions for the dependent velocity variables.
The previously used Lagrange multiplier un_lm has been removed.
When you open models saved in version 4.3, they include un_lm until the model is
re-solved. In some cases, occurrences of un_lm in the solver sequence must be replaced
manually. This is the case if un_lm was the only Lagrange multiplier component in a
segregated group or the only Lagrange multiplier component of a Vanka smoother.
Alternatively, you can generate a new automatic solver sequence. Models saved in
versions prior to version 4.3 must either be re-solved in version 5.0 for postprocessing,
or opened and re-saved in version 4.3 before being opened in version 5.0.
Weak constraints for the Interior Wall feature are no longer available.
R E V I S I O N O F T H E TU R B U L E N C E M O D E L S
The formulations of some variables in the turbulence models have been revised in
order to improve accuracy. Models using a turbulence model can display a different
convergence behavior in version 5.2 than in version 4.3 and the results can differ
slightly between the versions.

differ in between versions 4.3 and 5.2 due to the more general formulation for the
viscous stress in the momentum equation for the dispersed phase. The behavior in the
old formulation may be reproduced by dividing the dispersed phase viscosity, μD, by
the variable ee.phidPos.
40 | RELEASE NOTES
Chemical Reaction Engineering
Module
NEW CATALYTIC PAR TICLE SHAPES IN REACTIVE PELLET BEDS

In the Reactive Pellet Bed feature, you can now specify non-spherical catalytic pellet
shapes. Cylinder shapes, flakes, and user-defined pellet shapes are now available as
options.
EDITABLE SPECIES REACTION RATES IN SPECIES NODES, REACTION

ENGINEERING AND CHEMISTRY INTERFACES
It is now easy to change the definition of the reaction rate of a species directly in the
Species node. The entered expression overrides the definition set up by the Reaction
feature, where the rate is defined by the stoichiometry of the reactions in which the
species participates.
I M P R O V E D E Q U I L I B R I U M R E A C T I O N F U N C T I O N A L I T Y, R E A C T I O N
ENGINEERING INTERFACE
For equilibrium reactions, the Suppress negative concentrations check box has been
introduced to aid the computation of equilibrium systems. A selected check box
ensures that no negative values of concentrations are accepted as solutions to the
equilibrium condition.
MORE VERSATILE IMPOR T OF EXPERIMENTAL DATA FOR PARAMETER

ESTIMATION, REACTION ENGINEERING INTERFACE
The Experiment feature can now convert the units of the imported data used for
Parameter Estimation. The importance of each data type during the optimization can
also be controlled with the weight functionality.
MOLAR AND MILLIMOLAR ADDED AS CONCENTRATION UNIT

M (moles/l) and mM (mmoles/l) were added as units to visualize concentration, in
addition to the SI unit moles/m3.
CHEMICAL REACTION ENGINEERING MODULE | 41

MULTIPHYSICS NODE
The Multiphysics Flow Coupling node now supports Darcy’s Law interface as Source,
and the Transport of Concentrated Species and Nernst-Planck interfaces as
Destination.
The change may have implications on the Java API backward compatibility because the
E X T E N D E D P H Y S I C S I N T E R F A C E S U P P O R T F O R T H E TE M P E R A T U RE
The Multiphysics Temperature Coupling node now supports Darcy’s Law, Transport
of Concentrated Species, and the Nernst-Planck interfaces as Destination.
MEMBRANE DIALYSIS APPLICATION

This new application computes the contaminant concentration in a blood stream that
undergoes dialysis. The contaminant removal within a single hollow fiber can be
investigated by varying the fiber geometry and mass transport properties.
LIQUID CHROMATOGRAPHY APPLICATION

A new runnable app for studying the separation of two chemical components in a
high-performance liquid chromatography (HPLC) column.
FERMENTATION IN BEER BREWING

The application models the fermentation process in 0D with the Reaction Engineering
interface. Additionally, it models the process in a sphero-conical tank geometry with
mass transfer and heat transfer, and natural convection accounted for. These two model
setups enable in-depth investigation of several parameters affecting the final alcohol
content and the taste of the beer.
42 | RELEASE NOTES
Backward Compatibility with Version 5.0 and Earlier
REACTION ENGINEERING AND CHEMISTRY INTERFACES

The Java API syntax for creating surface species has changed. Running Java files from
previous versions may fail due to the new default name index ads for surface species
and s for solids in the Reaction Engineering and Chemistry interfaces. The surface
species names and solid species names need to be changed accordingly in any Java
code:
• Change any species index (s) to (ads)

• Change any species index (S) to (s)
The following shows an example of the needed code change:
Version 5.0 and earlier:

model.physics("re").feature("rch1").set("formula", "A+B(S)=>C+D(s)");
model.physics("re").feature("spec1").set("specName", "As(s)");
model.physics("chem").feature("rch1").set("formula", "A+B(S)=>C+D(s)");
model.physics("chem").feature("spec1").set("specName", "As(s)");
Change this to
Version 5.2:
model.physics("re").feature("rch1").set("formula", "A+B(s)=>C+D(ads)");
model.physics("re").feature("spec1").set("specName", "As(ads)");
model.physics("chem").feature("rch1").set("formula", "A+B(s)=>C+D(ads)");
model.physics("chem").feature("spec1").set("specName", "As(ads)");
TR A N S P O R T O F C O N C E N T R A T E D S P E C I E S I N T E R F A C E
Updated Internal Variable Names

The internal variable names of vector and tensor components have changed in the
Transport of Concentrated Species interface. In most cases, users will not be affected
by this change. When opening a model file created in a previous version, the previous
variable definitions remain. However, when updating or re-computing the solution,
new variable names will be generated. This means that if any of these variable names
have been used in a user-defined expression or in a plot, they need to be updated
manually as described below.
In version 5.1, the spatial direction indicator string (x, xy and so on) has been moved
to the end of the variable name. This applies to all vector and tensor variables generated
by the interface.

For example, the diffusive flux vector components have changed from (5.0 and
earlier):
tcs.dfluxx_w1, tcs.dfluxy_w1, tcs.dfluxz_w1
to (5.2)
tcs.dflux_w1x, tcs.dflux_w1y, tcs.dflux_w1z
and the components of the diffusion tensor, defined by the Fick’s Law diffusion model
have been changed from (5.0 and earlier):
tcs.Dfxx_w1, tcs.Dfxy_w1, tcs.Dfxz_w1, tcs.Dfyx_w1, ...
to (5.2)
tcs.Df_w1xx, tcs.Df_w1xy, tcs.Df_w1xz, tcs.Df_w1yx, ...
Java API Syntax Changes

The Java API syntax for setting physics properties has been changed in the Transport
of Concentrated Species interface. Now the property name corresponds to the name
of the GUI section containing the property. In order to run Java files from previous
versions, apply the syntax changes as given in the examples below:

model.physics("tcs").prop("DiffusionModel").setIndex("DiffusionModel",
"MaxwellStefan", 0);
"MixtureAveraged", 0);
"FicksLaw", 0);
model.physics("tcs").prop("Convection").setIndex("Convection", "0", 0);
model.physics("tcs").prop("Migration").setIndex("Migration", "0", 0);
model.physics("tcs").prop("Regularization").setIndex("Regularization", "1",
0);
For version 5.2, change this to:

model.physics("tcs").prop("TransportMechanism").set("DiffusionModel",
"MaxwellStefan");
"MixtureAveraged");
"FicksLaw");
model.physics("tcs").prop("TransportMechanism").set("Convection", false);
model.physics("tcs").prop("TransportMechanism").set("Migration", false);
model.physics("tcs").prop("AdvancedSettings").set("Regularization", "0");
44 | RELEASE NOTES
Updated Internal Variable Names

The same pattern for vector and tensor variable name changes as in Transport of
Concentrated Species Interface, described above, holds for the Surface Reactions
interface in version 5.1.
Java API Syntax Changes
The Java API syntax for setting the diffusion tensor of surface species, the molar mass
and density of the bulk species, and initial values have been changed in the Surface
Reactions interface.
Setting Diffusion Tensor
The surface species name needs to be included in the first argument of the set method.
In order to run Java files from previous versions, apply the syntax changes as in the
example below:

model.physics("sr").feature("sp1").set("D_0", new String[]{"Dcs1", "0", "0",
"0", "Dcs1", "0", "0", "0", "Dcs1"});
model.physics("sr").feature("sp1").set("D_1", new String[]{"Dcs2", "0", "0",
"0", "Dcs2", "0", "0", "0", "Dcs2"});

model.physics("sr").feature("sp1").set("D_cs1", new String[]{"Dcs1", "0",
"0", "0", "Dcs1", "0", "0", "0", "Dcs1"});
model.physics("sr").feature("sp1").set("D_cs2", new String[]{"Dcs2", "0",
"0", "0", "Dcs2", "0", "0", "0", "Dcs2"});
Setting Molar Mass and Density
The bulk species name needs to be included in the first argument of the setIndex
method. Apply the Java API syntax changes as in the example below:

model.physics("sr").feature("sp1").set("M_0", "Mcb1");
model.physics("sr").feature("sp1").set("M_1", "Mcb2");
model.physics("sr").feature("sp1").set("rho_0", "rhocb1");
model.physics("sr").feature("sp1").set("rho_1", "rhocb2");

model.physics("sr").feature("sp1").setIndex("M_cb1", "Mcb1", 0);
model.physics("sr").feature("sp1").setIndex("M_cb2", "Mcb2", 0);

model.physics("sr").feature("sp1").setIndex("rho_cb1", "rhocb1", 0);
model.physics("sr").feature("sp1").setIndex("rho_cb2", "rhocb2", 0);
Setting Initial Values
Apply the Java API syntax changes as in the example below:

model.physics("sr").feature("init1").set("cs1", "cs10");
model.physics("sr").feature("init1").set("cs2", "cs20");
model.physics("sr").feature("init1").set("cb1", "cb10");
model.physics("sr").feature("init1").set("cb2", "cb20");

model.physics("sr").feature("init1").setIndex("initcs", "cs10", 0);
model.physics("sr").feature("init1").setIndex("initcs", "cs20", 1);
model.physics("sr").feature("init1").setIndex("initcb", "cb10", 0);
model.physics("sr").feature("init1").setIndex("initcb", "cb20", 1);
46 | RELEASE NOTES
Corrosion Module
COMPENS ATE FO R TUBE VO LUME IN EL ECTROL YTES IN TH E CURREN T


WE A K C O NS TRA IN TS NOW U S ED BY DEFAULT IN PRIMAR Y CURREN T


PO RO US DOUBLE LAYER CAPACIT Y NODES
C Y C L I C VO L T A M M E T R Y N OW I N C O R P O R A T E S E X P L I C I T E VE N T S A T T H E
VER TEX POTENTIALS
CORROSION MODULE | 47
different results.
INITIAL VALUES FOR BO UNDAR Y EL ECTRIC POTENTIAL CAN NOW BE

SET IN THE ELECTRODE CURRENT AND THE ELECTROLYTE CURRENT
NODES
NEW HA RMON IC PER T URBATION SUBNODES FO R THE ELECTROL YTE

T H I N I N S U L A T I N G L A Y E R I N T H E P R I M A R Y, S E C O N D A R Y, A N D TE R T I A R Y
CURRENT DISTRIBUTION INTERFACES
The Thin Insulating Layer node can be used to model a thin insulating sheet, located
on an interior boundary in an electrolyte domain. The node can be used as an
alternative to drawing the actual insulating domain in the model geometry, which may
significantly reduce meshing time, especially in 3D models. Thin insulating sheets are
commonly inserted in the electrolyte in various types of electrochemical cells. For
example, they can be used for optimizing the current distribution in a corrosion
protection application or for optimizing the local deposition rate in a deposition bath.
DISSOLVING-DEPOSITING SPECIES SECTION IN THE ELECTRODE

SURFACE AND INTERNAL ELECTRODE SURFACE NODES
Use dissolving-depositing species in the Electrode Surface and Internal Electrode
Surface nodes in any Current Distribution interface to calculate the accumulated
surface concentration changes of dissolving or depositing processes. An arbitrary
number of species can be added. The dissolving-depositing species can also be used to
calculate the resulting boundary deformation velocity. The deformation velocity can
be used to couple the current distribution interface to the Deformed Geometry
interface (see below).
You specify the reaction rates in the Stoichiometry section in the Electrode Reaction
subnodes.
48 | RELEASE NOTES
REDESIGNED CORROSION, DEFORMED GEOMETRY INTERFACES
The Corrosion, Deformed Geometry interfaces have been redesigned as predefined
multiphysics interfaces. Upon choosing a Corrosion, Deformed Geometry interface in
the Select Physics menu, the corresponding Current Distribution interface and a
Deformed Geometry interface are added to the model, together with a
Non-Deforming Boundary and a Deforming Electrode Surface Multiphysics node.
The automatically added Deforming Electrode Surface Multiphysics node is used to
couple the dissolution/deposition rate of an Electrode Surface node (see the
Dissolving-Depositing Species Section above) in the Current Distribution interface to
the boundary velocity in the Deformed Geometry interface.
In addition, the moving boundary smoothing in 3D now also acts on the intersecting
edges between deforming and non-deforming boundaries.
Models created prior to 5.2 using a Corrosion interface will not be affected by the
change.

MULTIPHYSICS NODE
The Flow Coupling multiphysics node now supports Darcy’s Law interface as Source,
The Temperature Coupling multiphysics node now supports Darcy’s Law; Transport
Distribution; and the Tertiary Current Distribution, Nernst-Planck interfaces as
Destination.
CORROSION MODULE | 49

A new simpler diffuse_double_layer application has been added to the application
library.
The old diffuse double layer application has been renamed

diffuse_double_layer_with_charge_tran.

• The ship_hull application has been revised with updated boundary conditions for
more accurate ICCP control and updated plots.
• The oil_platform application has been revised with regards to the plotting of
potentials.
Backward Compatibility with Earlier Versions

50 | RELEASE NOTES
ECAD Import Module
• The ECAD Import Module now generates selections for imported layers to enable
a more efficient simulation setup. When importing, for example, ODB++ archives
and GDS-II files, the selections will have easily recognizable names that are
generated from the layer names. In the Import Settings window, you can rename
the layers or choose to keep the names found in the file.
• The length unit of a selected ECAD file is now displayed, and a new check box either
enables or disables the update of the geometry sequence length unit to the detected
unit in the file.
• For GDS import, arc recognition is improved to work when interior boundaries are
deleted even when the arc segments are originally not located on the same polygon.

• The default value for the grouping property of the Import function is now layer.
Previously the default was all.
• For ODB++® and ODB++(X) files (the type property is odb) new rules apply for
the initialization of the importlayer property of the Import function. Now only
layers of type Metal and Dielectric are initialized with the string on in the
importlayer string array. Previously this also included layers of type Drill.
ECAD IMPORT MODULE | 51

Electrochemistry Module
COMPENSATE FOR TUBE VOLUME IN ELECTROLYTES IN THE CURRENT


WEAK CONSTR AINT S NOW USED BY DEFAULT IN PRIMAR Y CURRENT


POROUS DOUBLE LAYER CAPACITY NODES
C Y C L I C VO L T A M M E T R Y N O W I N C O R P O R A T E S E X P L I C I T E V E N T S A T T H E
VERTEX POTENTIALS
52 | RELEASE NOTES
different results.
IN IT IA L VA LUES FO R BO UN DA R Y EL ECT RIC PO TEN TIA L C AN NOW BE

NODES
NEW HARMONIC PER TURBATION SUBNODES FOR THE ELECTROLYTE


MULTIPHYSICS NODE
The Flow Coupling multiphysics node now supports Darcy’s Law interface as Source,
Destination.
ELECTROCHEMISTRY MODULE | 53

A new simpler diffuse_double_layer application has been added to the application
library.
The old diffuse double layer application has been renamed

diffuse_double_layer_with_charge_tran.

54 | RELEASE NOTES
E le c t r o de po s i t i on Mod u l e
COMPENS ATE FO R TUBE VO LUME IN EL ECTROL YTES IN TH E CURREN T


WE A K C O NS TRA IN TS NOW U S ED BY DEFAULT IN PRIMAR Y CURREN T


PO RO US DOUBLE LAYER CAPACIT Y NODES
C Y C L I C VO L T A M M E T R Y N OW I N C O R P O R A T E S E X P L I C I T E VE N T S A T T H E
VER TEX POTENTIALS
ELECTRODEPOSITION MODULE | 55
different results.
INITIAL VALUES FOR BO UNDAR Y EL ECTRIC POTENTIAL CAN NOW BE

NODES
NEW HA RMON IC PER T URBATION SUBNODES FO R THE ELECTROL YTE

T H I N I N S U L A T I N G L A Y E R I N T H E P R I M A R Y, S E C O N D A R Y, A N D TE R T I A R Y
CURRENT DISTRIBUTION INTERFACES
The Thin Insulating Layer node can be used to model a thin insulating sheet, located
on an interior boundary in an electrolyte domain. The node can be used as an
alternative to drawing the actual insulating domain in the model geometry, which may
significantly reduce meshing time, especially in 3D models. Thin insulating sheets are
commonly inserted in the electrolyte in various types of electrochemical cells. For
example, they can be used for optimizing the current distribution in a corrosion
protection application or for optimizing the local deposition rate in a deposition bath.
DISSOLVING-DEPOSITING SPECIES SECTION IN THE ELECTRODE

SURFACE AND INTERNAL ELECTRODE SURFACE NODES
Use dissolving-depositing species in the Electrode Surface and Internal Electrode
Surface nodes in any Current Distribution interface to calculate the accumulated
surface concentration changes of dissolving or depositing processes. An arbitrary
number of species can be added. The dissolving-depositing species can also be used to
calculate the resulting boundary deformation velocity. The deformation velocity can
be used to couple the current distribution interface to the Deformed Geometry
interface (see below).
You specify the reaction rates in the Stoichiometry section in the Electrode Reaction
subnodes.
56 | RELEASE NOTES
REDESIGNED ELECTRODEPOSITION, DEFORMED GEOMETRY INTERFACES
The Electrodeposition, Deformed Geometry interfaces have been redesigned as
predefined multiphysics interfaces. Upon choosing an Electrodeposition, Deformed
Geometry interface in the Select Physics menu, the corresponding Current
Distribution interface and a Deformed Geometry interface are added to the model,
together with a Non-Deforming Boundary and a Deforming Electrode Surface
Multiphysics node. The automatically added Deforming Electrode Surface
Multiphysics node is used to couple the dissolution/deposition rate of an Electrode
Surface node (see the Dissolving-Depositing Species Section above) in the Current
Distribution interface to the boundary velocity in the Deformed Geometry interface.
In addition, the moving boundary smoothing in 3D now also acts on the intersecting
edges between deforming and non-deforming boundaries.
Models created prior to 5.2 using an Electrodeposition interface will not be affected
by the change.

MULTIPHYSICS NODE
The Flow Coupling multiphysics node now supports Darcy's Law interface as Source,
Destination.
DIFFUSE DOUBLE LAYER APPLICATION

A new diffuse_double layer application has been added to the application library.
ELECTRODEPOSITION MODULE | 57
• Updated the number of active species in the cu_electroless_deposition
application.

58 | RELEASE NOTES
Geomechanics Module
IMPROVED FORMULATION FOR SMALL STRAIN PLASTICITY WITH

GEOMETRIC NONLINEARITY
When Small plastic strains is selected as the plasticity model in the Plasticity node, and
the Include geometric nonlinearity check box is selected in the settings for the study, the
Cauchy stress tensor is now used to evaluate the yield function and plastic potential. In
previous versions the second Piola-Kirchhoff stress tensor was used instead. This
change will give significant improvements in accuracy for moderate to large strains. As
a result, you might experience changes in stresses and plastic strains for old models
having significant plastic strains when this combination of settings is used.
GEOMECHANICS MODULE | 59
Heat Transfer Module
New and Improved Functionality in Version 5.2
SYMMETRY PLANE FOR SURFACE-TO-SURFACE RADIATION

The new Symmetry Plane for Surface-to-Surface Radiation global feature is available in
2D, 2D axisymmetric, and 3D models when a Heat Transfer with Surface-to-Surface
Radiation interface or a Surface-to-Surface Radiation interface is active. It is used to take
advantage of plane symmetries to reduce the mesh for a more efficient computation of
the view factor. Depending on the dimension, the user interface provides different
inputs to define the symmetry plane through control points.
TE M P E R A T U RE O N E X T E R I O R S I D E O F T H I N L AYE R S
The new External Temperature subfeature is available on external boundaries under the
feature Thin Layer, when the Layer type is set to Resistive or General. It is used to specify
the temperature on the exterior side of the layer.
TE M P E R A T U RE O N E D G E S O F T H I N F I L M S A N D F R A C T U R E S
New Temperature subfeatures are available under the features Thin Film and Fracture.
They are used to specify the temperature on a set of lines (geometrical edges in 3D
models or geometrical points in 2D and 2D axisymmetric models) that represent thin
boundaries of a thin film or a fracture.
LINE HEAT FLUX ON BOUNDARIES OF THIN FILMS AND FRACTURES
New Line Heat Flux subfeatures are available under the features Thin Film and Fracture.
They are used to add a heat flux across boundaries of a thin film or a fracture. Their
settings provide different options for the definition of the flux: General inward heat flux,
Inward heat flux, and for 3D models, Overall heat transfer rate.
SURFACE-TO-AMBIENT RADIATION ON BOUNDARIES OF THIN FILMS

AND FRACTURES
New Surface-to-Ambient Radiation subfeatures are available under the features Thin Film
and Fracture. They are used to add surface-to-ambient radiation on a set of lines
(geometrical edges in 3D models or geometrical points in 2D and 2D axisymmetric
models) that represent thin boundaries of a thin film or a fracture. Their settings
provide inputs for the Ambient temperature and the Surface emissivity, which are used
for the definition of the net inward heat flux.
60 | RELEASE NOTES
HEAT SOURCE IN THIN FILMS AND FRACTURES
New Heat Source subfeatures are available under the features Thin Film and Fracture.
They are used to add an internal heat source within a thin film or a fracture. Their
settings provide two options for the definition of the heat source: General Source and
Overall heat transfer rate.
A D D I T I O N A L C O R R E L A T I O N F O R H E A T TR A N S F E R C O E F F I C I E N T S
A convective heat transfer coefficient correlation has been added to the heat transfer
coefficients library. This correlation corresponds to external flow induced by natural
convection around a vertical thin cylinder. Such a heat transfer coefficient can be used
to reduce the simulation cost when the model configuration corresponds to one of the
predefined correlations. In that case, the flow computation and the heat convection in
the fluid are replaced by a heat flux boundary condition on the solid boundaries using
the corresponding predefined correlation.
MISCELLANEOUS
• The formulation of the Thermal dispersion subfeature contribution has been updated
to improve the convergence when the heat transfer and the flow equations are
solved together.
• The equations of Thin Layer and Thin Layered Shell features have been updated to
support fully anisotropic thermal conductivity.
• The Thin Layer feature with the General option—which defines the layer as an extra
dimension—is now compatible with moving frame deformations. When a Moving
Mesh interface or structural mechanics physics interface is active for example, the
material’s thermal conductivity and density properties are updated to account for
the deformation effects.
• The Thin Layered Shell feature available in the Heat Transfer in Thin Shells interface
has been updated in the same way, to be compatible with moving frame
deformations.
• The detection of moving frames in the heat transfer interfaces has been enhanced.
Unnecessary updates of the view factor have been removed for efficiency when the
physics interface responsible for a frame deformation is not solved. This
improvement also benefits the solver suggestions, leading to better computational
performances.
• In the Thin Layered Shell feature, the initialization of the shape function Txdim—the
extra dimension temperature variable—has been updated to match the
H E A T TR A N S F E R M O D U L E | 61
temperature’s initial value in the adjacent domain. For interior boundaries, Txdim
is initialized as the average of adjacent initial temperatures.
• The weak equation of the Fracture feature has been updated to use the fluid heat
capacity instead of the effective heat capacity in the convective term.
• The automatic substitution mechanism of the temperature T by the variable
representing the temperature on the appropriate side has been revised. This
mechanism is designed to compute the temperature at the appropriate location
when it is used to define material properties or black body radiative intensity. It is
now only active when the default settings are used. When a model input is defined
using the User defined option, no substitution is made. In this case the user should
call the appropriate variable name to use the temperature on a particular side. For
example, ht.Tu should be used to compute the temperature on the upside. See
Boundary Wall Temperature in the Heat Transfer Module User’s Guide to get
the definition of all temperature variables available on boundaries.
• The solver settings for the heat transfer models have been updated. SOR is used
instead of SOR Line as the presmoother in the multigrid solver for large models,
unless the model contains a Non-Isothermal Flow coupling node. A lower limit of
zero has been added for the temperature when a segregated solver is used and the
temperature is in kelvin.
• The default opacity properties of domains adjacent to boundaries where a
Surface-to-Surface Radiation interface is active have been updated. These domains are
set as transparent by default, and can be turned to opaque using the Opaque feature.
• The definition of the velocity residual in time-dependent study steps with the
Non-Isothermal Flow multiphysics coupling feature has been fixed to include the
time-dependent term.
FLASH METHOD
The flash method is one of the most widely used methods for measuring the thermal
diffusivity of a given sample material. It consists of delivering a heat pulse on one side
of a thin piece of material to be tested. The temperature elevation on the other side is
then monitored and contains all the information to recover the thermal diffusivity.
Knowing the density and heat capacity at constant pressure, the thermal conductivity
can finally be deduced. This experiment protocol is reproduced in a 2D axisymmetric
model with a time-dependent study computing the temperature elevation for various
operating conditions, sample dimensions, and Gaussian laser pulse profiles.
62 | RELEASE NOTES
FINNED PIPE
Finned pipes are used for coolers, heaters, and mainly in heat exchangers to increase
thermal performance. They come in different sizes and designs depending on the
application and requirements. This app calculates the thermal performance of a pipe
filled with water and cooled or heated by the surrounding air. The model makes use of
periodicity conditions to be able to compute only a small section of the pipe. Various
geometric configurations are available for the outer fins (disc-stacked blades, circular
grooves, helical blade, helical grooves, or none) and for the inner fins (none or straight
grooves). Other parameters such as the pipe dimensions and the air flow conditions can
also be set.
THERMOELECTRIC COOLER
Thermoelectric coolers are widely used for electronics cooling in different application
areas, ranging from consumer products to spacecraft. Due to this, countless
configurations exist. This app calculates the thermal performance of single-stage
thermoelectric coolers of different sizes and with different thermocouple materials for
different operating conditions. It outputs common performance plots that are used to
describe the thermoelectric cooler characteristics. The underlying 3D model uses the
equations for the thermoelectric effect for thermocouples and the heat conduction
equation for electrical conductors and ceramics.
Updated Application in Version 5.2
GEOTHERMAL HEAT PUMP

An app with a user interface has been built from the existing application. It analyzes
the performance of ground-coupled heat exchangers for different specifications
(depth, pattern, pipe configuration, and heating conditions), temperature conditions,
and the soil’s thermal conductivity and temperature gradient.
The model couples heat transfer in solids with heat transfer in pipes. With the help of
the Events interface, the heater is turned off if the daily heat demand is achieved and is
turned on again after 24 hours.
The temperature at the pipe’s outlet can be controlled and compared to the minimum
temperature required in the heat exchanger specifications.

CHANGES IN VA RIABLE NAMES
The names of heat capacity variables have changed in several features for consistency:
• In the Bioheat feature, ht.cb has become ht.cp_b.

• In the Thin Layer feature, ht.cs has become ht.cp_s.
• In the Thin Rod feature, ht.cl has become ht.cp_l.
• In the Thin Layered Shell feature, htsh.cs1 has become htsh.cp_s1, and so on.
• In the Heat Transfer in Porous Media feature, ht.c_pp has become ht.cp_p,
ht.c_pp1 has become ht.cp_p1, and so on.
• In the Immobile Fluids feature, ht.c_pg1 has become ht.cp_g1, and so on.
• In the Fracture feature, ht.c_pfr has become ht.cp_fr.
The names of thermal resistance and absolute thermal resistance variables have
changed in the Isothermal Domain Interface feature:
• ht.idi1.Rs has become ht.idi1.R_t

• ht.idi1.Rsth has become ht.idi1.R_tth
The name of the fluid viscosity variable has changed in the Local Thermal
Non-Equilibrium multiphysics coupling: ltne1.mu_f has become ltne1.mu.
OBSOLETE FEATURES
A number of features are now obsolete because they have been merged or replaced by
a more general one. However, when models from COMSOL 5.0 or before contain
these features, some of the obsolete features are visible in version 5.1 with a warning
sign notifying that the feature is now obsolete (see Table 1-1). Some features have
been completely removed (see Table 1-2). A model containing any of them can be
opened in 5.1, but the feature will be missing.
Updating the models is recommended. Use the new feature and remove the obsolete
one to make sure that the model will remain compatible with future versions of
COMSOL Multiphysics.
64 | RELEASE NOTES
The table below summarizes the changes:
TABLE 1-1: OBSOLETE FEATURES THAT ARE OBSOLETE BUT THAN CAN BE LOADED IN 5.1
OBSOLETE FEATURE REPLACE IN 5.1 BY
Convective Heat Flux Heat Flux with the Convective heat flux option
selected
Surface-to-Ambient Radiation Diffuse Surface with Include surface-to-surface
radiation deselected
Surface-to-Surface Radiation Diffuse Surface with Include surface-to-surface
radiation selected
Highly Conductive Layer Thin Layer with Layer type set to Conductive
Thin Thermally Resistive Layer Thin Layer with Layer type set to Resistive
TABLE 1-2: OBSOLETE FEATURES THAT ARE NO LONGER AVAILABLE
OBSOLETE FEATURE REPLACE IN 5.1 BY
Infinite Element Infinite Element is no longer defined in the

physic interface. Infinite elements are defined in
the definitions section instead.
Electrochemical Heat flux Boundary Heat Source
Reaction Heat Flux Boundary Heat Source
Heat Transfer in Participating Heat Transfer in Solids contributing to
Media Radiation in Participating Media
Porous Matrix Heat Transfer in Porous Media
Note that it is not possible to add any of the obsolete features from the COMSOL
Desktop in version 5.1; only the new features are available.
O B S O L E T E PO S T P RO C E S S I N G VA R I A BL E S D E L E T E D I N 5 . 1
List of deleted postprocessing variables: ndflux_acc, ndflux_acc_u, ndflux_acc_d,
ntflux_acc, ntflux_acc_u, ntflux_acc_d, nteflux_acc, nteflux_acc_u,
nteflux_acc_d, qout, qout_u, qout_d, qin, qin_u, qin_d, qw, qw_u, and qw_d.
CHANGES IN VA RIABLE NAMES

The variable ht.Tvar used for non-isothermal flow wall functions becomes a fully
scoped variable in the following features:
• Heat Flux (ht.hf1.Tvar), Line Heat flux, Point Heat flux, and Out of plane heat flux.
• Heat Source (ht.hs1.Tvar), Boundary heat source, Line heat source, and Point heat
source.
• Diffuse Surface (ht.ds1.Tvar).
The fully scoped variables Ptot, Pbtot, Pltot, and Pstot become P0, Pb, Pl, and Ps,
respectively, in the following features:
• Heat Flux
• Inflow Heat Flux
• Line Heat Flux
• Heat Source
• Boundary Heat Source
• Line Heat Source
• Line Heat Source (Thin Rod subfeature)
• Line Heat Source on Axis
• Layer Heat Source (Thin Layer subfeature)
In the Thermal Contact feature:
• Qfric, QfricPow, and Qrate become Qb, Pb, and Qbpart, respectively
• Qfric0, Qrate0, Econtact0, hrad0, hgap0, and hconstr0 are removed and
migrated to existing variables Qb, Qbpart, Econtact, hrad, hgap, and hconstr,
respectively, in old models.
The Change Effective Thickness feature from the Heat Transfer in Thin Shells physics
interface merges htsh.de into the existing variable htsh.ds.
The Heat Transfer in Thin Shells physics interface also merges the property de and
variable htsh.de into existing property ds and variable htsh.ds. An API command
such as model.physics("htsh").prop("de").set("de", [...]) now does not
have any effect.
O U T - O F - P L A N E H E A T TR A N S F E R U P D A T E
The Out-of-plane heat transfer property, which was available in 2D and 1D
geometries, has been removed as of COMSOL Multiphysics 5.0. Now all geometries
are considered 3D geometries and thickness or area parameters are always available.
API Users
For the following features:
• Heat Source
• Heat Flux
66 | RELEASE NOTES
• Boundary Heat Source
• Layer Heat Source
• Inflow Heat Flux
With the Overall heat transfer rate option (previously named Total power or Total heat
flux), the thickness parameters (dz_entr and Ac_input) have been removed.
Compatibility with previous code is maintained for most of the cases. However, three
configurations require manual updates of the code:
• If you specified the thickness parameter twice (there were two instances of specifying
the parameter).
• If you specified first the thickness and in a second command the total power value.
• If you first specified the thickness and then activated the out-of-plane heat transfer
option.
For these 3 cases you need to rewrite this part of the Java code by doing the following.
For a heat source example (2D and 1D axisymmetric geometries):
Ptot_50 = Ptot_44_value*ht.d/dz_entr_44_value,
where Ptot_44_value is the total power value used in previous versions, Ptot_50 is
the total power value used since COMSOL Multiphysics 5.0, and dz_entr_44_value
is the value used in previous versions. In 1D geometries, the following update should
be performed: Ptot_50 = Ptot_44_value*ht.d/Ac_input_44_value ,
where Ac_input_44_value is the value used in previous versions.
Flux Variables
The flux variables are always given in W/m² (also for 2D or 1D geometries). To obtain
the corresponding flux variables in W/m on a boundary in 2D for example, the flux
variables have to be multiplied by the thickness ht.d.
The following variables have been updated following this convention:
• Conductive heat flux, ht.dflux

• Convective heat flux , ht.cflux
• Translational heat flux, ht.trlflux
• Total energy flux, ht.teflux
• Total heat flux, ht.tflux
• Boundary convective heat flux, ht.chflux
• Radiative heat flux, ht.rflux
All flux variables now have the following sign convention: positive heat flux
corresponds to heating, whereas negative heat flux corresponds to cooling. The
variables chflux and rflux may have opposite signs compared to previous versions
depending on the context where they are defined.
INFLOW HEAT FLUX IS ONLY APPLICABLE ON NON-SOLID DOMAINS

The Inflow Heat Flux feature is no longer applicable on boundaries adjacent to the
following domain features: Heat Transfer in Solids, Biological Tissue, and Isothermal
Domain. This restriction avoids ill-posed configurations due to the null velocity field in
these features. This may change model results if this boundary condition was
previously applied to a boundary adjacent to a non-fluid domain.
H E A T TR A N S F E R W I T H P H A S E C H A N G E N O W U S E S A VO L U M E T R I C
FORMULATION
The variables ht.thetai now represent volume fractions (they were mass fractions
before). This implies changes in the definition of density and heat capacity at constant
pressure close to the phase change interface. Far from the phase change interface the
density and heat capacity are unchanged.
INTERIOR FLUXES UPDATED

Because the interior flux variable definitions have been revised, models that use interior
fluxes in definitions, physics interfaces, or postprocessing show different results in
versions 5.0 and later than in earlier versions.
IMPROVED ENERGY BALANCE

Changes introduced in the equation formulation to enable better energy balance may
slightly modify the results.
Among the changes, the reference enthalpy, HRef, the value of which is defined up to
a constant, is now set to 0 J/kg at the reference temperature and pressure. This
changes the absolute values of the enthalpy and several variables based on it. In
particular, the value of the convective heat flux is changed. Even the sign may be
changed in some cases. However, the quantity of interest, the net convective heat flux,
is identical because the shift induced by the change of HRef applies for the enthalpy
definition at the inlet as well as at the outlet.
68 | RELEASE NOTES
H E A T TR A N S F E R I N P O RO U S M E D I A
Old models using the Porous feature will be supported in the COMSOL Multiphysics
interface. However, it is recommended to replace this feature with Heat Transfer in
Porous Media (see Obsolete Features and Variables). For a model using the API,
it will be necessary to update the model to use Heat Transfer in Porous Media.
SURFACE-TO-SURFACE RADIATION
The surface radiosity, previously named J, is now named ht.J by default (assuming
that ht is the physics interface tag).
Models created using versions 4.3 and 4.3a that contain surface-to-surface radiation
support the old syntax in multiphysics interfaces. It is, however, recommended that
you update the models to use the new syntax. Models created in a version before 4.2a
need to be opened in version 4.3 or 4.3a and resaved before being opened in 5.0 or a
later version.
For a model using the API, you must update the model to use the new syntax.
RADIATION IN PARTICIPATING MEDIA

The radiative intensity variables names, previously named I1 and so on, became ht.I1
(when there are fewer than 10 discrete directions) or ht.I01 (when there are 10 or
more discrete directions) by default.
Models created in versions 4.3 and 4.3a that include radiation in participating media
support the old syntax in multiphysics interfaces. It is, however, recommended that
you update the models to use the new syntax. Models created in a version before 4.2a
need to be opened in version 4.3 or 4.3a and resaved before being reopened in 5.0 or
later version. In addition, for all versions, it may be necessary to regenerate the default
solver to get the correct solver settings.
The S2 quadrature in 2D has been replaced by the LSE symmetric quadrature. It has
been updated so that it coincides with the other quadratures. LSE symmetric
quadratures were already used for S4, S6, and S8 in 2D and for all 3D quadratures.
This quadrature update can result in significant changes in 2D models using the S2
quadrature. However, the theoretical error is not different and the 2D and 3D
implementations are now consistent.
THE DEFAULT MOD EL LIST HAS BEEN REMOVED

Heat transfer physics interfaces no longer have the default model list. This list was used
to change the default feature for heat transfer models. You can obtain a similar result
by adding the corresponding domain feature immediately below the default feature
and setting its domain selection to All domains.
Any model files for Java that modify the default model value require a manual update.
I M P ROVE D S T A B I L I Z A T I O N O F H E AT T R A N S F E R I N S O L I D S
The streamline diffusion stabilization for Heat Transfer in Solids and Biological
Tissue features has been improved. It now accounts for contributions from linear
source terms from the Heat Source, Out-of-Plane Convective Cooling, Out-of-Plane
Radiation, and Out-of-Plane Heat Flux features. This improves the robustness of the
convergence when these contributions are large. This change may modify the
convergence behavior of existing models.
FRAME SELECTION IN DISCRETIZATION SECTION

The Frame type list, which was previously available when a moving mesh was detected,
has been removed. The frame type is now controlled by the features.
UPDATE OF FEATURE VARIABLE NAMES

In order to avoid name conflicts between feature variable names, some of them have
been renamed. This change concerns contributive features. Because they are
contributing, it is possible to have similar features active on the same boundary. To
make it possible to distinguish the variables from each feature, the feature scope has
been introduced in the variable name so that variable names now use a prefix. These
changes affect the following feature variables:
• Heat flux
• Out-of-plane heat flux
• Convective cooling
• Out-of-plane convective cooling
• Boundary heat source
• Heat source
• Line heat source
• Point heat source
70 | RELEASE NOTES
• Edge/Point heat flux
• Electrochemical reaction heat flux
• Reaction heat flux
For example, in a Boundary Heat Source feature, ht.surf has been renamed into
ht.bhs1.surf (assuming that ht is the physics interface tag). In the Heat Flux
feature, the variable that was previously named ht.q0_hf1 is now ht.hf1.q0.
Any model files for Java that use the old variable names in expressions (such as
expressions used for plotting or data evaluation that include such old variable names)
need to be updated manually.
NEW DEFAULT FLUID FEATURES AND OPAQUE SUBFEATURE

All MPH-files created in version 4.3 are converted to appear and behave the same way
as in 5.0 or a later version.
Due to these new default features, model files for Java can be simplified. In addition,
model files for Java that add the Fluid feature or the Opaque subfeature with the
default tag require a manual update to avoid duplicate tag conflicts.
WE AK CONS TRAINTS UP DATE F OR FLUI D F LOW US ER INTERFACES

The weak constraint formulations for some boundary conditions in the following
interfaces have been updated:
• Laminar Flow
• Turbulent Flow, k-ε
• Turbulent Flow, Low Re k-ε
• Non-Isothermal Flow
• Conjugate Heat Transfer
See the CFD Module release notes for details.
Weak constraints for the Interior Wall feature are no longer available.
R EV I S I O N T O T H E TU R B U L E N C E M O D E L S
The formulations of some variables in the turbulence models have been revised in
order to improve accuracy. Simulations using a turbulence model can display a
different convergence behavior in versions 5.0 and later compared to version 4.3, and
the numerical result can differ slightly between the versions.
Obsolete Features and Variables
The following features have been removed:
• Porous Matrix (replaced by Heat Transfer in Porous Media)

• Infinite Element (replaced by Infinite Element Domain declared in the Definitions node)
• Electrochemical Heat Flux (replaced by Boundary Heat Source)
• Reaction Heat Flux (replaced by Boundary Heat Source)
• Heat Transfer in Participating Media (replaced by Heat Transfer with Radiation in
Participating Media)
When models created in older versions are loaded in version 5.2, these features are
removed from the model tree and the model can still be opened. The new
corresponding feature can be added to replace the removed one. In the particular case
of a feature defined as a default feature of an interface in the model, the model cannot
be opened.
The table below contains a list of variables that have been removed and replaced by
new corresponding expressions.
TABLE 1-3: REMOVED VARIABLES AND CORRESPONDING EXPRESSIONS
REMOVED VARIABLE CORRESPONDING EXPRESSION
qin -qr_in
qout qr_out
qw -qr_net
ndflux_acc ndflux
ndflux_acc_u ndflux_u
ndflux_acc_d ndflux_d
ntflux_acc ntflux
ntflux_acc_u ntflux_u
ntflux_acc_d ntflux_d
nteflux_acc nteflux
nteflux_acc_u nteflux_u
nteflux_acc_d nteflux_d
72 | RELEASE NOTES
M E M S M o dule
MATERIAL MODELS FROM EXTERNALLY PROGRAMMED LIBRARIES

A new way to specify user-defined material models has been added. You can now access
external material functions, written in C or some other programming language, which
have been compiled into a shared library. This makes it possible to program your own
material models and also to distribute such models as an add-on.
The external library is accessed through an External Material feature under

Global Definitions>Materials. The external library can either completely define the
stress-strain relation or only return an inelastic strain contribution to available material
models. Two new features are available in the Solid Mechanics interface to
accommodate this: the External Stress-Strain Relation material model, and the External
Strain subnode under the Linear Elastic Material.
New Application in Version 5.2
MEMS PRESSURE SENSOR DRIFT DUE TO HYGROSCOPIC SWELLING

For protection, electronic devices are often overmolded with an epoxy mold
compound (EMC). Due to hygroscopic swelling, the epoxy might endure dimensional
changes that can affect the performance of the device. In this application, a
parameterized pressure sensor can be analyzed by changing a number of geometrical
parameters, as well as the material properties. The app results show the stability of the
sensor readings due to moisture diffusion in the epoxy. The Transport of Diluted
Species, Solid Mechanics, and Shell interfaces are combined in the application.

For models using contact, plot expressions are not updated when you open an old
model where the pair names have been edited manually. Also, opening an old model
where a pair name has been manually edited might affect the solver configuration. It
is recommended that you regenerate a new default solver configuration.
MEMS MODULE | 73
Microfluidics Module
T H R E E - P H A S E F L O W, P H A S E F I E L D I N T E R F A C E
The new Three-Phase Flow, Phase Field interface can be used to model the flow and
interaction of three different, immiscible fluids when the exact positions of the
interfaces separating the fluids are of interest, also known as separated flow with surface
tracking. The fluid-fluid interfaces are tracked using a ternary phase field formulation
that accounts for differences in the fluids’ densities and viscosities, and the formulation
includes the effects of surface tension. The phase-field method can handle moving
contact lines on no-slip boundaries. The movement of the fluid-fluid interfaces is
determined by minimization of free energy.
TE R N A R Y P H A S E F I E L D
The corresponding Ternary Phase Field interface, used to track moving interfaces
between three immiscible phases in the CFD Module and the Microfluidics Module,
is also found as a standalone Mathematics interface.
INKJET
The Inkjet application demonstrates how to model the flow of a droplet produced by
an inkjet nozzle, such as one in a printer. An ink droplet is ejected from the nozzle and
travels through air until it hits the target. The fluid flow is incompressible, and the
movement of the interface between the droplet and the ambient air is modeled using
the level-set method. The user can specify liquid properties in designated input fields,
and the geometry—including nozzle, reservoir, compartment, and outlet—is fully
parameterized. In addition, the user can control the injection pulse profile by varying
the pulse interval and smoothing time.

The Free Molecular Flow and Transitional Flow user interfaces now form the
Molecular Flow Module. A Molecular Flow Module license is required to use these
interfaces.
74 | RELEASE NOTES
BRINKMAN EQUATIONS AND FREE AND POROUS MEDIA FLOW

The momentum equations in the Brinkman Equations interface and the Free and
Porous Media Flow interface have been corrected. The term −Qbr·u/εp2on the right
hand side previously lacked the factor 1/εp2, where εp is the porosity.
MOBILITY
In physics interfaces modeling migration (for example, Transport of Diluted Species),
anisotropic mobility is now supported, and the mobility can be set to follow the
Nernst-Einstein relation. The default setting when making new models has been
changed to the Nernst-Einstein relation. Model files for Java generated prior to 4.3b
using Migration will have to be modified manually to account for these changes.
MICROFLUIDICS MODULE | 75
Mixer Module
R O T A T I N G M A C H I N E R Y F L U I D F L O W I N T E R F A C E S : TU R B U L E N T F L O W,
A L G E B R A I C Y P L U S A N D TU R B U L E N T F L O W, L - V E L
Two new physics interfaces have been added in the Rotating Machinery Fluid Flow
branch: the Turbulent Flow, Algebraic yPlus interface and the Turbulent Flow,
L-VEL interface. For these algebraic turbulence models, the turbulent viscosity is
determined using two different extensions of the logarithmic wall law, and the local
Reynolds number based on the distance to the nearest wall. Algebraic turbulence
models are computationally cheaper and more robust but generally less accurate than
transport equation turbulence models, such as the k-ε and k-ω turbulence models.
MIXER
This application is intended for the simulation of mixers equipped with axial or radial
impellers. With a given geometry, the application computes the efficiency of the
mixing process, where the measure of efficiency is the eddy diffusivity, from which the
time scale of mixing of chemical species can be estimated. The app user can select
eleven different types of impellers, each with its own distinctive combination of
features, such as blade cuts and folds. Three different types of vessel are available:
dished bottom, flat bottom, and cone bottom vessels, and the vessels can also be
equipped with baffles.
76 | RELEASE NOTES
M o le c ula r Flo w Mod u l e
CHARGE EXCHANGE CELL SIMULATOR

A charge exchange cell consists of a region of gas at an elevated pressure within a
vacuum chamber. When an ion beam interacts with the higher-density gas, the ions
undergo charge exchange reactions with the gas, creating energetic neutral particles.
It is likely that only a fraction of the beam ions will undergo charge exchange reactions,
so in order to neutralize the beam, a pair of charged deflecting plates are positioned
outside the cell. In this way, an energetic neutral source can be produced.
This application simulates the interaction of a proton beam with a charge exchange cell
containing neutral argon. User inputs include several geometric parameters for the gas
cell and vacuum chamber, the beam properties, and the properties of the charged
plates that are used to deflect the remaining ions. The app computes the efficiency of
the charge exchange cell, measured as the fraction of ions that are neutralized, and
records statistics about the different types of collisions that occur.

Due to the fact that multiple species are now supported, some variable names have
changed. In general, variables will now have _G appended to their names, where G is
the name specified for the Incident molecular fluxes in the Dependent Variables section.
So, if a model is updated and resolved, the expressions used in results should be
changed as follows:
• The number density, fmf.N should be changed to fmf.N_G.

• The pressure should be changed from p to fmf.p_G.
• The reconstructed number density, fmf.N should be changed to fmf.ndr1.Nr_G,
where ndr1 is the tag of the Number Density Reconstruction feature.
• Any other variables can be updated using the Replace Expression button available
during results processing.
In addition, new variables exist for the total pressure, number density, and so on:
• The total number density, fmf.ntot.
MOLECULAR FLOW MODULE | 77

• The total pressure fmf.ptot.
• The total incident molecular flow, fmf.Gtot.
78 | RELEASE NOTES
Multibody Dynamics Module
IMPROVED SOLVER SETTINGS FOR FRICTION

The default solver settings when Friction is used in the Multibody Dynamics interface
have been significantly improved.
PRINCIPAL MOMENTS OF INERTIA FOR RIGID DOMAINS

The principal moments of inertia for a Rigid Domain feature are now available as
variables that can be directly displayed during postprocessing.
RIGID DOMAIN RESULTS IN FREQUENCY DOMAIN

All results in a Rigid Domain, including phase information, are now available in a
Frequency Domain study.
RIGID BODY SYSTEM SIMULATOR

In some cases, it can be more convenient to build a model using input data structured
in tables than by working directly in the Model Builder tree. In this app, a multibody
dynamics simulation of an assembly of rigid bodies can be set up by pure tabular input;
the tables can even be prepared off-line and read directly into the app either
individually or in the form of an input file. In this way, data is given for the geometry
in terms of nodes and their connectivities, and the rigid bodies in terms of mass
properties, joint properties, loads, and constraints. Transient as well as stationary
parametric analyses can be performed to obtain configuration plots, trajectory plots,
joint motions, and joint forces.
CENTRIFUGAL GOVER NOR SIMULATOR

A centrifugal governor is a classical angular speed regulation device. In this app, a
parameterized centrifugal governor can be analyzed with respect to steady state,
transient response, and natural frequencies.
MULTIBODY DYNAMICS MODULE | 79

Nonlinear Structural Materials
Module
IMPROVED FORMULATION FOR SMALL STRAIN PLASTICITY WITH

GEOMETRIC NONLINEARITY
When Small plastic strains is selected as the plasticity model in the Plasticity node, and
the Include geometric nonlinearity check box is selected in the settings for the study, the
Cauchy stress tensor is now used to evaluate the yield function and plastic potential. In
previous versions the second Piola-Kirchhoff stress tensor was used instead. This
change will give significant improvements in accuracy for moderate to large strains. As
a result, you might experience changes in stresses and plastic strains for old models
having significant plastic strains when this combination of settings is used.
S T RE S S A N A L Y S I S O F A P RE S S U RE VE S S E L
In this app, a pressure vessel is analyzed in order check the allowed internal pressure.
Some yielding can be allowed, and the plasticity model can be orthotropic to
accommodate for effects introduced by plate rolling. Inputs include the main cylinder
dimensions of the vessel, the geometry of the head, computed automatically for a
chosen standard head type, design parameters, and material properties. The results give
the analytic yielded volume fraction and the graphical stress, plastic strain, yielded
volume and volume fraction, and local directions.
80 | RELEASE NOTES
Optimization Module
The Optimization check box in the Stationary and Time Dependent study steps have
been removed. Instead, use the Optimization study step. The Optimization check box
does appear in models created in earlier versions of COMSOL where you have selected
that check box.
OPTIMIZATION MODULE | 81
Particle Tracing Module
RELEAS E FRO M EDGES AN D PO INT S

The Release from Edge and Release from Point nodes can be used to release particles
from edges and points in a geometry, respectively. When releasing particles along an
edge, the particle positions can be mesh-based, weighted by a user-defined density
function, or uniformly distributed along the edge length.
IMPROVED DENSITY-BASED R ELEASE

The release features that initialize particle position according to a density function have
new settings that can make them much more accurate. It is now possible to specify a
Release distribution accuracy order and Position refinement factor in the settings for the
Release, Inlet, and Particle Beam nodes as well as the new Release from Edge node.
The improvement in accuracy is most noticeable when the underlying mesh is very
coarse or the particle density varies significantly across different mesh elements.
CHARGE EXCHANGE COLLISIONS

Two new collision types can now be added to the Collisions node: Resonant Charge
Exchange and Nonresonant Charge Exchange.
The Resonant Charge Exchange node is used when energetic ions undergo charge
exchange reactions with ambient neutral atoms of the same element or molecules of
the same substance. The Nonresonant Charge Exchange feature is used when the
ionized and neutral species are of different elements or substances. In both cases, after
the collision, it is possible to continue tracking the ionized species, the neutral species,
or both.
PAR TICLE BEAM IMPROVEMENTS

New options are available in the Particle Beam feature to make the transverse position
and velocity distributions easier to specify. This makes it much easier to release beams
with phase space ellipses of a certain size, shape, and orientation. The Equation Display
has been improved and augmented with images to provide a better indication of what
the various options do.
82 | RELEASE NOTES
PARTICLE COUNTERS
A Particle Counter feature is a domain or boundary feature that provides information
about particles arriving on a set of selected domains or surfaces from a release feature.
Such quantities include the number of particles transmitted, the transmission
probability, transmitted current, mass flow rate, and so on. This feature provides
convenient results expressions that can be used in the Filters node of the Particle
Trajectories plot, which allows only the particles that reach the particle counter
selection to be visualized.
PARTICLE-MATTER INTERACTIONS
You can now model the interaction of energetic ions with solid matter using the
dedicated Particle-Matter Interactions feature. This feature supports two subfeatures
for different types of interactions:
• Ionization loss is used to model the continuous loss of energy as ions interact with
electrons in the target material.
• Nuclear stopping is used to model the deflection of energetic ions by target nuclei.
MAGNETIC FORCE USING EARTH'S MAGNETIC FIELD

The Magnetic Force feature now has a new option to use data for the Earth’s magnetic
field. It is possible to specify locations in the world either based on particle coordinates,
entering latitude and longitude, or by selecting one of several built-in city locations.
BUILT-IN DOMAIN ACCUMULATORS

The Space Charge Density Calculation feature is now available for the Charged Particle
Tracing interface. This feature computes the space charge density of particles in a
domain. It is the single-physics equivalent of the Electric Particle-Field Interaction
Multiphysics node. Similarly, a Volume Force Calculation feature is now available for
the Particle Tracing for Fluid Flow interface.
SUPPORT FOR DIELECTROPHORESIS WITH NESTED SHELLS

It is now possible to add a Dielectrophoretic Force to particles with multiple thin
shells. To do so, add two or more Shell subnodes to the Dielectrophoretic Force node.
P A R T I C L E TR A C I N G M O D U L E | 83
CHARGE EXCHANGE CELL SIMULATOR

A charge exchange cell consists of a region of gas at an elevated pressure within a
vacuum chamber. When an ion beam interacts with the higher-density gas, the ions
undergo charge exchange reactions with the gas, creating energetic neutral particles.
It is likely that only a fraction of the beam ions will undergo charge exchange reactions,
so in order to neutralize the beam a pair of charged deflecting plates are positioned
outside the cell. In this way, an energetic neutral source can be produced.
This application simulates the interaction of a proton beam with a charge exchange cell
containing neutral Argon. User inputs include several geometric parameters for the gas
cell and vacuum chamber, the beam properties, and the properties of the charged
plates that are used to deflect the remaining ions. The app computes the efficiency of
the charge exchange cell, measured as the fraction of ions that are neutralized, and
records statistics about the different types of collisions that occur.
LAMINAR STATIC PARTICLE MIXER DESIGNER

In static mixers, a fluid is pumped through a pipe containing stationary mixing blades.
This mixing technique is well suited for laminar flow mixing because it generates only
small pressure losses in this flow regime. However, the performance of the mixer can
vary greatly depending on its geometry.
This application computes the fluid velocity and pressure field in a static mixer and
then the trajectories of particles that are carried by the fluid. Because the particles have
mass, they do not follow the fluid velocity streamlines exactly, causing some particles
to hit the mixing blades. This app computes the transmission probability of particles in
the mixer. It also evaluates the index of dispersion, which is a measurement of the
uniformity with which different species of particle are mixed together.
ION RANGE BENCHMARK

This simple benchmark model confirms that when modeling a combination of
ionization losses and nuclear scattering, the computed average range of protons in
silicon agrees with experimental data. The average range is computed for initial proton
energies from 1 keV to 100 MeV.
SENSITIVE HIGH-RESOLUTION ION MICROPROBE (SHRIMP)

The SHRIMP can be used to transmit ions of a given initial energy and specified
charge-to-mass ratio by subjecting an incoming beam to appropriately tuned electric
84 | RELEASE NOTES
and magnetic forces. The beam is first sent through a curved sector with a radial
electric force, then through a second curved sector with a uniform magnetic flux
density. The Particle Counter feature is used to visualize the transmitted beam and to
compute the transmission probability of the ions.
RENAMED PHYSICS INTERFACE SETTINGS

The Release type list in the settings windows for the Charged Particle Tracing and
Particle Tracing for Fluid Flow interfaces in earlier versions has been renamed Particle
release specification. For the Charged Particle Tracing interface the options Static and
Transient have been renamed to Specify current and Specify release times (the default),
respectively. For the Particle Tracing for Fluid Flow interface the options Static and
Transient have been renamed to Specify mass flow rate and Specify release times (the
default), respectively.
OLD NODES FOR PARTICLE-FIELD AND FLUID-PARTICLE INTERACTIONS

ARE OBSOLETE
The old Particle-Field Interaction and Fluid-Particle Interaction nodes, which could be
added directly to the Charged Particle Tracing and Particle Tracing for Fluid Flow
interfaces, respectively, are now obsolete. They will be removed in a future version.
Simulations of particle-field and fluid-particle interactions should instead use the
dedicated Multiphysics nodes.
ELASTIC COLLISION FORCE NODE HAS BEEN REPLACED

If a model created in Version 5.0 or earlier containing the Elastic Collision Force node
is opened in Version 5.1, this node will be replaced. If the Collision model in the old
Elastic Collision Force node was set to Monte Carlo, a Collisions node with the Elastic
subnode will be created. If instead the Collision model was set to Friction, the Friction
Force subnode will be added.
P A R T I C L E TR A C I N G M O D U L E | 85
Pipe Flow Module
REVISED FORMULATION OF THE PIPE FLOW INTERFACE

The Pipe Flow interface has been amended with an inertial term. This means that
gradually diverging or converging pipe sections will account for a pressure change due
to the velocity change (Bernoulli effect). Also, sudden expansions and contractions will
account for the Bernoulli effect properly. The stabilization and element shape
functions have also been changed to handle compressible flows for Ma<0.3 properly.
Models created in previous versions will be opened using the old interface with a
notification. For new models, only the new interface is available in the physics interface
list.
GROUND HEAT RECOVER Y APPLICATION

The performance of a ground heat exchanger is analyzed. The app covers different
designs and specifications of the heat exchanger at different ambient conditions. It
calculates the temperature distribution, heat production, and heater state.
ORGAN PIPE DESIGN

Working with acoustics, designing instruments or other sound transducing devices it
can be interesting to not only look at a simulated frequency response but also listen to
the sound generated by the device. This is auralization of the acoustics of the device.
This demonstration application shows how this can be achieved using a simple model
of an organ pipe. Changing the organ pipe design will, for example, result in a different
pitch, which can be auralized. The organ pipe is modeled using the Pipe Acoustics,
Frequency Domain interface. The application allows users to analyze how the first
fundamental resonance frequency varies with pipe properties such as radius and wall
thickness, and with the ambient pressure and temperature. The full frequency response
is found including the fundamental frequency and the harmonics. Using a Java
method, the results analysis is extended beyond the built-in functionality. The
application will detect and find the location and amplitude of all harmonics in the
response. It will also allow users to listen to the sound and pitch of a given organ
design.
86 | RELEASE NOTES
Plasma Module
I M P R OVE D S T A B I L I T Y O F T H E F I N I T E VO L U M E D I S C R E T I Z A T I O N
The finite volume discretization has been improved so that numerical overflow errors
no longer occur. This means that nonconvergence of models using the finite volume
discretization should occur less often.
REDUCED ELECTRIC FIELDS STUDY NOW SUPPOR TS UNITS

The Reduced Electric Fields study now interprets units. The SI unit for the reduced
electric field is V·m2, but it is possible to enter a value in Townsend, for example,
10[Td].
IMPROVED MATERIAL DATA FOR EQUILIBRIUM DISCHARGES

Some outlying data points in the Equilibrium Discharge material data have been
removed. This potentially makes models more stable when solving.

Models using the Reduced Electric Fields study will interpret the unit in the edit field
for the Reduced electric fields as Townsend. For example, when opening a model
created in version 5.1, if the value for the Reduced electric fields was 10, this will now
appear as 10[Td] in the study settings.

Old models will be updated to use the new boundary conditions, but re-solving a
model still gives the same results as before. The following changes to the model are
made upon opening:
• If an Electric Potential feature is applied to boundaries adjacent to a Plasma Model

feature, it is converted to a Metal Contact feature. The settings from the original
model are retained. If an Electric Potential feature is applied to boundaries adjacent
to a Charge Conservation feature, it remains the same.
• If an Electric Potential feature is applied to a combination of boundaries adjacent to
a Plasma Model and Charge Conservation feature, a Metal Contact feature is
PLASMA MODULE | 87
created for the boundaries adjacent to the Plasma Model feature, and the Electric
Potential feature remains on boundaries adjacent to the Charge Conservation
feature.
• Any External Surface Charge Accumulation boundary conditions are replaced by a
new Dielectric Contact feature. The settings from the original model are retained.
The External Surface Charge Accumulation boundary condition is now obsolete.
88 | RELEASE NOTES
Ray Optics Module
NEW PART: ASPHERIC LENS

Aspheric lenses are often used as an alternative to spherical lenses to reduce
monochromatic aberrations. An aspheric lens is now included in the Part Library for
the Ray Optics Module. The depth or sag of the curved surface is defined using a
standard aspheric formula in which the values of all coefficients are given as input
parameters.
RELEASE RAYS FROM EDGES AND POINTS

It is now possible to release rays from edges in 3D using the new Release from Edge
feature, and to release rays from selected points using the new Release from Point
feature. When releasing rays from edges, you can select any set of edges and release rays
uniformly, based on the mesh, or with initial positions weighted according to a
user-defined density function.
NEW OPTIONS FOR DENSITY-BASED RELEASE OF RAYS

The release features that initialize ray position according to a density function have new
settings that make them more accurate. You can now specify a Release distribution
accuracy order and Position refinement factor in the settings for the Release, Inlet, and
Illuminated Surface nodes as well as the new Release from Edge node. The
improvement in accuracy is most noticeable when the underlying mesh is very coarse
or the density expression varies significantly across different mesh elements.
DEPOSITED RAY POWER PE RF OR MA NC E IMPROVEMEN TS

The Deposited Ray Power node on domains now uses the latest domain Accumulator
technology, making it much faster and more accurate compared to older versions.
RENAMED SETTINGS
The Wall condition Bounce has been renamed to Specular reflection.
RAY OPTICS MODULE | 89

CHANGES TO FREQUENCY-DEPENDENT MATERIAL PROPERTIES
The Frequency-dependent refractive indices check box has been renamed to Allow
frequency distributions at release features.
Expressions involving the ray frequency must always be included within the noenv()
operator. Some expressions for user-defined refractive indices may fail to be evaluated
properly in version 5.2 unless this operator is applied.
THIN DIELECTRIC FILM SETTINGS

In version 5.0, if the ray intensity was computed, the option to add Thin Dielectric Film
nodes to a Material Discontinuity node was always available. In version 5.2, the Thin
Dielectric Film subnode can only be added if Add layers to surface or Add layers to
surface, repeating is selected from the Thin dielectric films on boundary list in the Material
Discontinuity settings window. If a model created in version 5.0 is opened in version
5.2, and any Thin Dielectric Film nodes have been added, then Add layers to surface is
selected from the Thin dielectric films on boundary list instead of the default None.
CHANGES TO THE CIRC UL AR WAVE RE TA RD ER

The Retardance of the Circular Wave Retarder now indicates the rotation angle of the
polarization ellipse of a ray that passes through the device. In version 5.0, the
retardance corresponded to half of this rotation angle. The behavior of the Circular
Wave Retarder is now consistent with the documentation. If a model created in version
5.0 is opened in version 5.2, the expression for the Retardance of any Circular Wave
Retarder node is multiplied by 2.
OBSOLETE DEPOSITED RAY POWER NODE

The Deposited Ray Power node for absorbing domains is obsolete and will be removed
in future versions. For simulations in which rays generate heat due to absorption, the
Ray Heat Source Multiphysics node should be used instead.
90 | RELEASE NOTES
RF Module
ENHANCED PHYSICS-CONTROLLED MESH HANDLING LOSSY MEDIA

The physics-controlled mesh takes lossy electric and magnetic media and automatically
scales the size by skin depth on the lossy domain boundaries. When Resolve wave in
lossy media is selected, the outer boundaries of lossy media domains are meshed with
a maximum mesh element size, in free space, given by the minimum value of half a skin
depth plus 1/5 of the vacuum wavelength.
MULTI-ELEMENT UNIFORM LUMPED POR T

A new lumped port type is included in COMSOL Multiphysics 5.2, the Multi-element
uniform port for multi-excitation or -termination of, for example, a coplanar
waveguide port or a differential port. The direction of the field in each subelement of
the port is defined by the Uniform Element subnodes, where the direction between
uniform element terminals, ah, defines the electric potential polarity.
SM IT H P LO T, C O NVE NT IO NA L WAY TO P RE SE NT M AT C H IN G P ROP E R T IE S

When an S-parameter plot is generated by default, a Smith plot will be automatically
generated as well.
BOUNDARY MODE, FREQUENCY-STATIONARY AND BOUNDARY MODE,

F RE Q U E N C Y - TR A N S I E N T S T U D Y S E Q U E N C E S
There are new study sequences in the Model Wizard for the Wave Optics Module and
the RF Module multiphysics interfaces called Laser Heating and Microwave Heating,
respectively. The Boundary Mode, Frequency-Stationary study sequence adds a
Boundary Mode Analysis study step and a Frequency-Stationary study step, whereas
the Boundary Mode, Frequency-Transient study sequence adds a Boundary Mode
Analysis study step and a Frequency-Transient study step. The Boundary Mode
Analysis study step is used for solving the mode field for numeric ports in the
electromagnetic interfaces. The Frequency-Stationary and Frequency-Transient study
steps couple stationary and transient analyses for the Heat Transfer in Solids interface
with a frequency domain analysis for the Wave Optics and RF interfaces.
RF MODULE | 91
I N I T I A L VA L U E S E T T I N G S F O R T H E TR A N S I E N T S C A T T E R I N G B O U N D A R Y
CONDITION
In the Settings window for the Scattering boundary condition for time-dependent
simulations, there is a new Initial Values for Incident Wave section for setting the initial
values of the magnetic vector potential for the incident wave. Note that the section is
initially collapsed by default. When the incident wave is defined by an electric field, you
can specify the initial value for the magnetic vector potential for the incident wave.
When the incident wave is defined by a magnetic field, the user can specify the initial
value for the time derivative of the magnetic potential, in addition to the initial value
for the magnetic vector potential. The new settings enable the user to define the exact
wave form for the solved magnetic vector potential.
FREQUENCY SELECTIVE SURFACE SIMULATOR

Frequency selective surfaces (FSS) are periodic structures that generate a bandpass or
a bandstop frequency response. This application simulates a user-specified periodic
structure chosen from the built-in unit cell types. It provides five unit cell types
popularly used in FSS simulations along with two predefined polarizations in one fixed
direction of propagation that has normal incidence on the FSS. The analysis includes
the reflection and transmission spectra, the electric field norm on the top surface of the
unit cell, and the dB-scaled electric field norm shown on a vertical cut plane in the unit
cell domain.
SLOT-COUPLED MICROSTRIP PATCH ANTENNA ARRAY SYNTHESIZER

This application simulates a single slot-coupled microstrip patch antenna fabricated on
a multilayered low temperature co-fired ceramic (LTCC) substrate. Results include the
far-field radiation pattern of the antenna array and its directivity. The far-field radiation
pattern is approximated by multiplying the array factor and the single antenna
radiation pattern to perform an efficient far-field analysis without simulating a
complicated full array model. Phased antenna array prototypes for 5G mobile networks
can be evaluated with the default input frequency, 30 GHz. This app includes two view
types, a Compact View and a Wide View, depending on how much space the user has
to operate the app.
C O P L A NA R WAVEG U I D E (CP W ) B A NDP A S S F IL T E R

Coplanar waveguide (CPW) bandpass filters can be designed using interdigital
capacitors (IDCs) and short-circuited stub inductors (SSIs). Such a filter can readily be
92 | RELEASE NOTES
implemented on a GaAs wafer. The filter in this model is compact in relation to its
resonant frequency and provides a relatively high Q-factor compared to capacitively
coupled microstrip line model designs.

The following models have been moved from the RF Module Application Library to
the Wave Optics Module Application Library:
• Beam Splitter
• Photonic Crystal
• Scattering on Substrate
• Optical Scattering Off of a Gold Nanosphere
• Dielectric Slab Waveguide
• Fabry-Perot Cavity
RF MODULE | 93
Semiconductor Module
EQUATION DISPLAY REDESIGN

The equation display has been completely redesigned to give a clearer indication of the
equations and boundary conditions that are being applied. Auxiliary relations for the
boundary conditions have been moved to their own relevant section.
94 | RELEASE NOTES
Structural Mechanics Module
MATERIAL MODELS FROM EXTERNALLY PROGRAMMED LIBRARIES

A new way to specify user-defined material models has been added. You can now access
external material functions written in C or some other programming language which
have been compiled into a shared library. This makes it possible to program your own
material models, and also to distribute such models as add-ons.
The external library is accessed through an External Material feature under Global
Definitions>Materials. The external library can either completely define the stress-strain
relation, or only return an inelastic strain contribution to available material models.
Two new features are available in the Solid Mechanics interface to accommodate this:
the External Stress-Strain Relation material model, and the External Strain subnode
under the Linear Elastic Material.
CONTACT WITH SMALL RELATIVE DISPLACEMENTS

A new simplified method for computing the distances within contact pairs has been
introduced. It can be used when there is little sliding between the contacting surfaces,
like in a shrink fit or bolted joint. In this method, the mapping between source and
destination is computed only once, a strategy which leads to faster and more stable
convergence. To use this method, set Mapping method to Initial configuration in the
settings for the contact pair.
ADJUSTMENT OF THE INITIAL CONTACT GAP

Sometimes, the finite element discretization of curved boundaries causes the initial
distance between the two boundaries in a contact pair to have noticeable irregularities.
This can now be compensated for through a built-in computation of the initial gap.
This initial gap can then be subtracted in the subsequent analysis if you select the Force
zero initial gap check box in the settings for Contact in the Solid Mechanics interface.
FULL MASS MATRIX IN ADDED MASS

The Added Mass feature has been improved so that it is possible to enter a full mass
matrix.
STRUCTURAL MECHANICS MODULE | 95

OPTIONAL INCLUSION OF THERMOELASTIC DAMPING
In the Thermal Expansion multiphysics coupling, it is now possible to select whether the
time derivative of the stresses should act as a heat source in the heat transfer problem
or not. In previous versions, this thermoelastic damping effect was always active.
PRINCIPAL MOMENTS OF INERTIA FOR RIGID DOMAINS

The principal moments of inertia for a Rigid Domain are now available as variables
which can be directly displayed during postprocessing.
RIGID DOMAIN RESULTS IN FREQUENCY DOMAIN STUDIES

All results in a Rigid Domain, including phase information, are now available in
Frequency Domain studies.
BEAM SECTION CALCULATOR

Using this app, which is built upon the Beam Cross Section interface, it is possible to
evaluate cross section data for a wide range of US and European standard beams. The
computed data can then be transferred for use as input data in a Beam interface. Given
a set of forces and moments acting on the section, you can also compute a detailed
stress distribution.
An additional app is included for users with LiveLink™ for Excel®, which has somewhat
more functionality, such as the possibility to extend it with other beam standards.
INTERFERENCE FIT
In an interference fit, two parts are joined by employing the contact pressure formed
when a slightly oversized part is fitted into a hole. The stress distribution, contact
pressure, and transferable force and moment in an interference fit between two circular
pipes are computed in this app. You can vary the dimensions of the two pipes and the
coefficient of friction to test different designs.
MEMS PRESSURE SENSOR DRIFT DUE TO HYGROSCOPIC SWELLING

For protection, electronic devices are often overmolded with an epoxy mold
compound (EMC). Due to hygroscopic swelling, the epoxy might endure dimensional
changes that can affect the performance of the device. In this app, a parameterized
pressure sensor can be analyzed by changing a number of geometrical parameters, as
well as the material properties. The app results show the stability of the sensor readings
96 | RELEASE NOTES
due to moisture diffusion in the epoxy. The Transport of Diluted Species, Solid
Mechanics, and Shell interfaces are combined in the application.
TR U S S B R I D G E D E S I G N E R
This is an example of an application for studying a civil engineering structure, a Pratt
truss bridge. The underlying model consists of a Shell interface for the roadway and a
Beam interface for the supporting structure. With this app, you can modify all
dimensions of the bridge, including the number of spans and the cross sections of all
members. Additionally, you can configure the load with an arbitrary number of
vehicles and a wind load. In addition to the static results, the lowest natural frequencies
are computed.
L I N E A R B U C K L I N G O F A TR U S S TO W E R
In this app, users can compute the buckling load for unstayed and guyed truss towers.
First, a user enters the tower dimensions, consisting of individual (longitudinal,
transverse, and diagonal) members. Then, choose between the unstayed or the guyed
option, which calls for data for the wires, including prestress. Finally, for material data,
select structural steel, aluminum, or a user-defined material. After running the app,
there are graphical and analytical results for the critical load.
VISCOELASTIC STRUCTURAL DAMPER

The dynamics properties of a viscoelastic damper are studied in this app. You can
modify the viscoelastic material's properties and the loads at the attachment points,
The primary result is the force-displacement relation at the attachment points.
Alternate results include hysteresis plots and storage and loss moduli. You can even
compare two different materials side by side.
INTERFERENCE FIT CONNECTION IN A MOUNTAIN BIKE FORK

This is a tutorial example showing how to model interference fits. The stresses caused
by a shrink fit between the steering tube and the crown in a bike fork are computed.

• If an old model using the Beam interface is opened, the Euler-Bernoulli formulation
is used. The Beam Formulation selection is still shown but cannot be changed from
Euler-Bernoulli. If Advanced Physics Options is enabled, then the new section
Backward Compatibility is shown. If you clear the Use pre 4.4 formulation check box,
then the new formulation is used. Doing this enables the use of Timoshenko beams,
but you must manually handle solver settings like segregation and scaling. This
STRUCTURAL MECHANICS MODULE | 97

legacy option does not support beams mixed with solids or shells in cases where the
same names for degrees of freedom were used in both physics interfaces.
• For models using mechanical contact, plot expressions are not updated when you
open an old model where the pair names have been edited manually. Also, opening
an old model where a pair name has been manually edited might affect the solver
configuration. It is recommended that you regenerate a new default solver
configuration.

• When using version 5.2, in the Solid Mechanics interface (SolidMechanics) the
property ControlsSpatialFrame from version 4.3a no longer exists, and
COMSOL Multiphysics now always assumes that the displacements control the
spatial frame. Remove all lines setting the value of ControlsSpatialFrame. See
Support Knowledge Base article 1156.
• When opening old models that include contact, the penalty factor control is set to
User defined and the contact normal penalty factor is set to the expression used in
the model.
98 | RELEASE NOTES
Subsurface Flow Module
S O L U T E TR A N S P O R T
The Solute Transport interface has been replaced by the new interface Transport of
Diluted Species in Porous Media.
SUBSURFACE FLOW MODULE | 99

Wave Optics Module
ENHANCED PHYSICS-CONTROLLED MESH HANDLING OF LOSSY MEDIA

The physics-controlled mesh takes lossy electric and magnetic media and provides a
boundary layer mesh automatically scaled by the skin depth. When Resolve wave in
lossy media is selected in the properties of the Electromagnetic Waves, Frequency
Domain interface, the outer boundaries of a lossy media domain are processed with a
boundary layer mesh approximately one skin depth thick. The two-layer mesh has a
1.2 stretching factor. As the first layer is half a skin depth thick, the second layer is 0.6
skin depth thick, making the total thickness of the boundary layer mesh 1.1 skin depths
thick.
BOUNDARY MODE, FREQUENCY-STATIONARY AND BOUNDARY MODE,

F RE Q U E N C Y - TR A N S I E N T S T U D Y S E Q U E N C E S
For the Laser Heating multiphysics interface there are new study sequences in the
Model Wizard. The Boundary Mode, Frequency-Stationary study sequence adds a
Boundary Mode Analysis study step and a Frequency-Stationary study step, whereas
the Boundary Mode, Frequency-Transient study sequence adds a Boundary Mode
Analysis study step and a Frequency-Transient study step. The Boundary Mode
Analysis study step is used for solving for the mode field for numeric ports in the
electromagnetic interfaces. The Frequency-Stationary and Frequency-Transient study
steps couple a frequency-domain analysis for the electromagnetic interface with
stationary and transient analyses, respectively, for the Heat Transfer in Solids interface.
I N I T I A L VA L U E S E T T I N G S F O R T H E TR A N S I E N T S C A T T E R I N G B O U N D A R Y
CONDITION
In the settings for the Scattering Boundary Condition for time-dependent simulations,
there is a new section called Initial Values for Incident Wave for setting the initial values
of the magnetic vector potential for the incident wave. This section is collapsed by
default. When the incident wave is defined by an electric field, you can also specify the
initial value for the magnetic vector potential for the incident wave. When the incident
wave is defined by a magnetic field, you can also specify the initial value for the time
derivative of the magnetic potential, in addition to the initial value for the magnetic
vector potential.
100 | RELEASE NOTES

NO SCATTERED FIELD SETTING FOR THE MATCHED BOUNDARY
CONDITION
To prevent spurious solutions to appear when using the Matched boundary condition
with the Electromagnetic Waves, Beam Envelopes interface, a new setting has been
introduced—the No scattered field check box. When this check box is selected, the
scattered wave is constrained to be zero on the boundary that the feature is selected
for. An example model using this new setting is the brewster_interface model.
FIBER SIMULATOR
This application performs mode analysis on a concentric circular dielectric layer
structure. Both step-index fibers and graded-index optical fibers can be simulated. The
geometry and refractive index distribution, the mode fields, the mode-field diameter,
the attenuation, the group delay, and the dispersion coefficient are visualized.
WAVE O P T I C S M O D U L E | 101
Material Library
New Material Data in Version 5.2
Data has been added for the following materials: Be3N2, Ca3N2, HfN, HfN, TiN,
ZrN, HfC, NbC, Ta2O3, Sc2O3, Lu2O3, Tm2O3, Gd2O3, Sm2Oe, CeO2, SrO,
C5H12, Mo2C, WCIN-939, 26NiCrMoV14-5, 18Mn18Cr (ASTM A289), 2507
(S32750), S32053, ASTM B625 (N08031), UNS N08354, NAS NM15M, NAS
NM17, Pyroceram 9606, Corning 1737, Corning Eagle 2000, Corning Eagle XG,
Corning Lotus, Corning Lotus XT, and SiO2-TiO2.
102 | RELEASE NOTES

LiveLink for MATLAB
LiveLink™ for MATLAB® now supports MATLAB® R2015b.
The following sections list new and improved functionality in LiveLink™ for
®
MATLAB :
MPHPLOT FUNCTION
The wrapper function mphplot has been enhanced in many ways to provide plots in
MATLAB that look more like the plot created by COMSOL Multiphysics. This results
in more accurate colors, line styles, symbols, and so on.
A new options 'normals' has been added to mphplot. It is 'on' by default and tells
MATLAB if normals should be used when rendering surfaces in 3D. This is mostly
noticeable when lighting is used or when displaying arrows.
A new option 'createplot' has been added. It is 'on' by default, but when set to
'off' it is possible to extract data from the model without actually creating a plot on
the screen. The data can later be used to recreate the plot. This operation is mostly
useful when working with servers that do not have a display but is also useful for users
who wish to extract data from a plot.
The new plot types Octave Plots, Smith Plots, and Annotation Plots have been added.
MPHGETEXPRESSIONS FUNCTION
The mphgetexpressions function now returns the value and unit of the evaluated
expression using the model’s default unit system.
MPHEVALGLOBALMATRIX FUNCTION
New options 'outersolnum' and 'outerdataseries' have been added to the
mphevalglobalmatrix function for improved support for models that use parametric
sweeps.
MPHINPUTMATRIX FUNCTION
The mphinputmatrix function now supports complex-valued matrices.
L I V E L I N K F O R M AT L A B | 103
MPHXMESHINFO FUNCTION
The mphxmeshinfo function now supports separate DOF coordinates for models that
use extra dimensions.

• The mphint2 command: For data sets other than Solution, Particle, Cut Line,
Cut Plane, Time Integral, Time Average, Surface, and Line, the integration order
now corresponds to the element refinement. The default value is set to 4 instead of
1. This change can affect the solution in cases where the problem is not resolved well
enough.
• The mphapplicationlibraries function replaces the mphmodellibrary function
for showing the Application Libraries window.
104 | RELEASE NOTES

LiveLink for Excel
The following new functionality is now available in LiveLink™ for Excel®:
S U P P O R T E D VE R S I O N S
Microsoft® Excel® 2016 has been added to the list of supported versions of Excel.
Now Excel 2007, 2010, 2013, and 2016 are supported when using Windows. Loading
and Saving of Excel files is supported on Linux and Mac OS X.
PARAMETER UPDATE DIALOG BOX

A new dialog box for updating parameters and other settings in a model. This dialog
box provides improved control of which parameters are updated when there are
multiple sets of parameters in an Excel workbook.
LIVELINK FOR EXCEL | 105

T he C O M S O L A PI for Use wi t h Jav a ®
COMSOL 5.1 API Changes
ACOUSTICS MODULE CHANGES

The ratio of specific heats is now a material parameter. If the default value was used in
a model, add the following line to the Java file to obtain the old behavior:
model.physics("cpf").feature("cpf1").set("gamma_mat", "userdef");
ELECTROCHEMISTRY MODULES CHANGES

There are changes for the Tertiary Current Distribution, Nernst-Planck interface. The
Java API syntax for creating and accessing vectors and tensors has changed, as well as
the syntax for setting physics properties. See the following examples:
• Specifying the diffusion tensor.

Old syntax:
model.physics("tcdee").feature("ice1").set("D_0", new
String[]{"D1", "0", "0", "0", "D1", "0", "0", "0", "D1"});
String[]{"D2", "0", "0", "0", "D2", "0", "0", "0", "D2"});
String[]{"D3", "0", "0", "0", "D3", "0", "0", "0", "D3"});
New syntax (includes the species name in the first argument of the set method):
model.physics("tcdee").feature("ice1").set("D_c1", new
String[]{"D1", "0", "0", "0", "D1", "0", "0", "0", "D1"});
String[]{"D2", "0", "0", "0", "D2", "0", "0", "0", "D2"});
String[]{"D3", "0", "0", "0", "D3", "0", "0", "0", "D3"});
• Accessing tensor components (in the definitions of other variables, for example).
Old syntax:
Dxx_c1, Dxy_c1, Dxz_c1...
New syntax (component index after the species name):

D_c1xx, D_c1xy, D_c1xz...
• Specifying the reaction rate vector.

Old syntax:
106 | RELEASE NOTES

model.physics("tcdee").feature("reac1").set("R_0", new
String[]{"R1"});
String[]{"R2"});
String[]{"R3"});
New syntax (includes the species name in the first argument of the set method):
model.physics("tcdee").feature("reac1").set("R_c1", new
String[]{"R1"});
String[]{"R2"});
String[]{"R3"});
• Accessing vector components.

Old syntax:
tfluxx_c1, tfluxy_c1, tfluxz_c1
New syntax (component index after the species name):

tflux_c1x, tflux_c1y, tflux_c1z
• Specifying initial values:

Old syntax:
model.physics("tcdee").feature("init1").setIndex("c2", "c20", 0);
model.physics("tcdee").feature("init1").setIndex("c3", "c30", 0);
model.physics("tcdee").feature("init1").setIndex("V", "V0", 0);
New syntax:
model.physics("tcdee").feature("init1").setIndex("initc", "c20",
1);
model.physics("tcdee").feature("init1").setIndex("initc", "c30",
2);
model.physics("tcdee").feature("init1").setIndex("initphil",
"V0");
• Selecting properties.
Old syntax:
model.physics("tcdee").prop("Convection").set("Convection", 1,
"0");
model.physics("tcdee").prop("ConvectiveTerm").setIndex("Convectiv
eTerm", "noncons", 0);
New syntax (the property name corresponds to the section in the user interface):
model.physics("tcdee").prop("TransportMechanism").set("Convection
", 1, "0");
THE COMSOL API FOR USE WITH JAVA® | 107

model.physics("tcdee").prop("AdvancedSettings").setIndex("Convect
iveTerm", "noncons", 0);
For information about API backward compatibility for the Surface Reactions interface,
see the Chemical Reaction Engineering Module release notes.
ECAD IMPORT MODULE CHANGES

• The default value for the grouping property of the Import function is now set to
layer. Previously the default was all.
• For ODB++® and ODB++(X) files (the type property is odb) new rules apply for
the initialization of the importlayer property of the Import function. Now only
layers of type Metal and Dielectric are initialized with the string on in the
importlayer string array. Previously this also included layers of type Drill.
LIVELINK FOR SOLID EDGE CHANGES

The default value for the keepfree property of the LiveLinkSolidEdge function is
LIVELINK FOR SOLIDWORKS CHANGES

The default value for the keepfree property of the LiveLinkSOLIDWORKS function is
DOCUMENTATION
The COMSOL Multiphysics Programming Reference Manual replaces the
COMSOL API for use with Java® Reference Manual.
DEPRECATED METHODS
The following methods are deprecated in COMSOL 5.0:
TABLE 1-4: DEPRECATED METHODS AND THEIR REPLACEMENTS
DEPRECATED METHOD NEW METHOD
com.comsol.model.OptFeature.field() OptFeature.comp()
com.comsol.model.OptFeature.field(String) OptFeature.comp(String)
com.comsol.model.Group.identifier() Group.paramName()
com.comsol.model.Material.identifier() ModelEntity.tag()
com.comsol.model.MaterialModel. ModelEntity.tag()
identifier()
108 | RELEASE NOTES

com.comsol.model.ModelNode.identifier() ModelEntity.tag()
com.comsol.model.physics.Physics. Use tag() instead.
identifier()
com.comsol.model.physics. ModelEntity.tag()
MultiphysicsCoupling.identifier()
com.comsol.model.Group.identifier(String) Group.paramName(String)
com.comsol.model.Material. ModelEntity.tag(String)
identifier(String)
com.comsol.model.MaterialModel. ModelEntity.tag(String)
identifier(String)
com.comsol.model.ModelNode. ModelEntity.tag(String)
identifier(String)
com.comsol.model.physics.Physics. Use tag(string) instead.
identifier(String)
com.comsol.model.physics. ModelEntity.tag(String)
MultiphysicsCoupling.identifier(String)
com.comsol.model.ModelEntity.name() ModelEntity.label()
com.comsol.model.ModelEntity.name(String) ModelEntity.
label(String)
com.comsol.model.ParameterEntity. Use the 0-based version
set(String, int, double) (setIndex) instead.
set(String, int, double) (setIndex) instead.
set(String, int, double[]) (setIndex) instead.
set(String, int, int) (setIndex) instead.
set(String, int, int[]) (setIndex) instead.
set(String, int, int, double) (setIndex) instead.
set(String, int, int, int) (setIndex) instead.
set(String, int, int, String) (setIndex) instead.


set(String, int, String) (setIndex) instead.
set(String, int, String[] (setIndex) instead.
PHYSICS INTERFACES
• The default Initial Value features in the following interfaces have been extended:
- Transport of Concentrated Species
- Reacting Flow in Porous Media (rfcs)
- Reacting Flow
- Rotating Machinery, Reacting Flow
Previously, only the initial mass fractions could be specified. Now the initial values
can be specified in terms of mass fractions, mole fractions, molar concentrations,
number densities, or densities. When opening an old Model MPH-file, Mixture
specification is set to Mass fractions, and the mass fractions specified are entered in
the respective text fields. In old model files for Java, commands specifying the initial
value of a mass fraction return an error in version 5.0 and later. To specify the initial
mass fraction of a species w2, which is the second in the list of species (seen in the
Dependent Variables section when clicking the interface), change the COMSOL
Java command
model.physics("chcs").feature("init1").set("w2", 1, "0.1");
to
model.physics("chcs").feature("init1").set("w0", 2, "0.1");
• The Pressure Acoustics, Frequency Domain interface and the Boundary Mode
Acoustics interface have a new default feature. If the default feature has been edited
in a Java file, the following line should be added to the Java file to obtain the old
behavior for Pressure Acoustics, Frequency Domain:
model.physics("acpr").feature().create("pam1",
"PressureAcousticsModel").selection().all();
or
110 | RELEASE NOTES

model.physics("acbm").feature().create("pam1",
"PressureAcousticsModel").selection().all();
for Boundary Mode Acoustics.

This line should be added directly after the line where the physics interface is
created.
• For models using mechanical contact, the variable field name in the solver settings
for the old contact variables is different than in earlier versions. For example:
- In version 5.2: mod1_solid_contact_old_p1, mod1_solid_cm_old_p1.
- In version earlier than 5.0: mod1_solid_contact_p1_old,
mod1_solid_cm_p1_old.
• The parameter form in the property EquationForm can no longer be set to

ModeAnalysis in 3D and 2D axisymmetric calculations for the Pressure Acoustics,
Frequency Domain, and Pressure Acoustics, Transient interfaces. In these cases, the
parameter modeFreq has been removed.
COMSOL 4.3b API Changes
PHYSICS INTERFACES
• The Crosswind diffusion formulation has been updated for all physics interfaces
supporting Crosswind diffusion. Models solved with the new formulation can give
different results than models solved in versions earlier than 4.3b. Java files can be
modified to retain old crosswind formulations. Please contact COMSOL Support
for details.
• The default settings have changed for the Background Pressure Field in the Pressure
Acoustics interfaces. Add the following line to obtain the old behavior of this
feature:
model.physics("acpr").feature("bpf1").set("c", 1, "acpr.c_c");
• In the Acoustics Module, the Far-Field Calculation feature is no longer available in

1D and 1D axisymmetric models.
• The shape function property border has been deprecated and replaced by order.

COMSOL 4.3a API Changes
GENERAL API CHANGES

• Class files compiled with COMSOL Multiphysics version 4.3 or earlier need to be
recompiled if they contain code that changes the contents of model.selection().
• The solutionintitle property governs whether to show solution information in
plot titles. But since solution information has never been shown for 1D plots,
regardless of the content of the property, the property has been removed from 1D
plot groups.
FLUID FLOW INTERFACES API

• Fluid flow features in the CFD Module and Microfluidics Module that previously
created a scalar Lagrange multiplier field with default name model.un_lm now
create a vectorial Lagrange multiplier field with default name model.u_lm. The
default component names in 3D are model.u_lm, model.v_lm, and model.w_lm,
respectively. Java files must be updated accordingly.
• Weak constraints for the Interior Wall feature are no longer available. Any references
to its weak constraint parameter (weakConstraints) or Lagrange multipliers must
be removed.
• The Automatic setting for the Pseudo time stepping property now sets the variable
<phtag>.locCFL to the built-in variable CFLCMP, which in turn triggers a PID
regulator via the automatic solver suggestions. Here, <phtag> is the physics
interface tag. Java files where pseudo time stepping is active and the local CFL
number is set to Automatic must be modified by adding the command
model.physics(<tag>).prop("PseudoTimeProperty").set("CFLNumbExpr"
, 1, "Manual");
before calling a solver.

• In the Acoustics Module, the far-field variables pfar and Lp_far have new names
with full scope. They are now referred to as acpr.ffc1.pfar and
acpr.ffc1.Lp_far, respectively. Any model files for Java that use the old variable
names in expressions (such as expressions used for plotting or evaluation that
include such old variable names) require a manual update.
• The following methods
model.physics(<tag>).feature(<ftag>).params();
112 | RELEASE NOTES

model.physics(<tag>).prop(propname).params();
are deprecated and replaced by the methods

model.physics(<tag>).feature(<ftag>).param();
model.physics(<tag>).prop(propname).param();
COMSOL 4.2a API Changes

• Class files compiled with COMSOL Multiphysics version 4.1 or earlier need to be
recompiled.
• The far-field feature in the Electromagnetic Waves interface has changed from being
a boundary feature with a boundary selection only, to a domain feature with a
domain selection. It also has a subfeature—a boundary selection.
• The dependent variable associated with gauge fixing in the Magnetic Fields,
Magnetic and Electric Fields, Electromagnetic Waves, and Transient
Electromagnetic Waves interfaces is now unique to each interface. It is no longer
available in the model scope, for example, mod1.psi. Instead the gauge fixing field
is only available in the interface scope, for example, as mod1.mf.psi.
• In the scattered field formulation in the Electromagnetic Waves interface, the
scattered field is no longer available in the model scope (for example, mod1.relEx).
Instead the scattered field is only available in the interface scope as, for example,
mod1.emw.relEx.
• In the Solid Mechanics interfaces (SolidMechanics, PiezoelectricDevices,

AcousticStructure, Poroelasticity, ThermallyInducedStresses,
JouleHeatingThermalExpansion, TransientAcousticStructure,
FluidStructureInteraction), tensors in local coordinate systems (el, eel, Sl,
si, and ei) have new names. The coordinates of the local system (for example x1,
x2, and x3) are no longer used. Instead 1, 2, and 3 are used together with double
indexing for all components. As an example, elX2 is replaced by el22 and elx2x3
is replaced by el23. The tensors si and ei are now called Sil and eil.
• In the Darcy’s Law interface and the Richards’ Equation interface in the Subsurface
Flow Module, fluid compressibility is now a material parameter and no longer has a
default value. If the default value was used, you now have to set the value. The
following example sets the permeability to the old default value:
model.physics("dl").feature("smm1").set(chif_mat,userdef);
model.physics("dl").feature("smm1").set(kappa,4e-10);
• In the Poroelasticity interface in the Subsurface Flow Module, the fluid

compressibility and the Biot-Willis coefficient are now material parameters and no

longer have default values. If the default values were used you now have to set the
value. The following example sets the permeability to the old default value:
model.physics("dl").feature("smm1").set(chif_mat,userdef);
model.physics("dl").feature("smm1").set(kappa,4e-10);
• The Level Set and Phase Field interfaces now include the Initial Interface feature by
default. If you have a model that was created in an earlier version of COMSOL, it
will fail to create a feature with the same tag name.

The following changes were made to the COMSOL API between versions 4.0a and
4.1:
• In version 4.0a the property nonlin in the stationary solver could have the values
auto, on, off, and linearized. The default in most cases was auto, but in some
cases, specifically in a frequency-domain study, the default was linearized. In
version 4.1 the possible values are auto, on, off, and linper, with auto as default,
except in the Frequency Domain, Linearized study step where the default is
linper. The value linearized can still be set, but this is treated as auto. If the
problem is linear and the linearization point is zero, auto should give the same
solution as linearized. For nonlinear problems where the linearization point is
zero, the solution given by auto might be slightly different, and probably more
accurate, while changing to off should give exactly the same solution as
linearized. When there is a nonzero linearization point, it is relevant to use the
linper option. This option works like linearized except that source terms that
contribute to the linearized problem must be enclosed by the linper operator.
• In the Darcy’s Law interface, Brinkman interface, Richards’ Equation interface, and
Poroelasticity interface (in the CFD Module and Subsurface Flow Module), the
permeability and porosity are now material parameters and no longer have default
values. If the default values were used in a model file for Java, you now have to set
them. The following example sets the permeability to the old default value:
model.physics("dl").feature("dlm1").set(kappa_mat,userdef);
model.physics("dl").feature("dlm1").set(kappa,3e-11);
• The pressure variable in solid mechanics, typically solid.pw, now only gets
allocated degrees of freedom for incompressible materials. If you have referred to it,
for example, during segregation in the solver, the code will have to be changed to
remove the field.
114 | RELEASE NOTES

• The solution to eigenvalue and eigenfrequency problems may now appear in a
different order.
• The Batteries & Fuel Cells Module’s boundary feature,
BoundaryReactionCoefficient, is tagged by rc instead of brc by default. This
means that files that explicitly refer to the old default tag name must be modified to
refer to the new tag name.
COMSOL 4.0a Java API Changes

The following changes were made to the COMSOL API between versions 4.0 and
4.0a:
• The units of the load face parameter (Fph) of the Phase feature in
AcousticStructure, TransientAcousticStructure,
ThermallyInducedStresses, SolidMechanics,
JouleHeatingThermalExpansion, Poroelasticity, and
FluidStructureInteraction have changed from degrees to radians.
• The physics.field() operator for Beam and Truss are not backward compatible
with 4.0. Scalar fields are now vector fields.
• The variables for strain and stress in the Shell interface have been renamed.
Variable names that began with a lowercase s now begin with an uppercase S. For
variables names that ended in lowercase l (L), the letter was moved to the third
position. For example, emXXl is now emlXX.
• Force loads on solids and trusses have been moved from the material to the spatial
frame. This means that the index notation of some variables has changed from
uppercase XYZ to lowercase xyz.
• The acoustics interfaces (PressureAcoustics, TransientPressureAcoustics,
AcousticStructure, and TransientAcousticStructure) have been moved
from the material to the spatial frame. This means that the index notation of vector
variables has changed from uppercase XYZ to lowercase xyz.
This concludes the release notes for COMSOL Multiphysics version 5.2.

116 | RELEASE NOTES

COMSOL ReleaseNotes

Uploaded by

Copyright:

Available Formats

COMSOL ReleaseNotes

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

COMSOL ReleaseNotes

Uploaded by

Copyright:

Available Formats

COMSOL Multiphysics

• Support Center: www.comsol.com/support

Part number: CM010001

COMSOL® version 5.2 includes an extended and improved Application Builder,

RELEASE OF LICENSE AT NEW/OPEN

New Functionality in the Application Builder

GENERAL NEW FUNCTIONALITY

IMPROVEMENTS TO THE FORM EDITOR

IMPROVEMENTS TO THE METHOD EDITOR AND APPLICATION LANGUAGE

New Functionality in COMSOL Server

• Number of recently run applications to preload: Default: 1. Preloads this number of

LAUNCHING APPLICATION AT LOGIN

LAUNCHING COMSOL CLIENT FROM WEB PAGE

IMPROVED FILTERING AND SOR TING

CREATING DESKTOP SHORTCUTS

NEW COMSOL COMMANDS

• comsolserver / comsol server

New and Improved General Functionality in COMSOL Multiphysics

BETTER CONTROL OF OUTPUT SELECTIONS FROM GEOMETRIES

NEW MESHING ALGORITHM

MESH PARTS AND MESH IMPORT

DIAGNOSTICS FOR INVER TED ELEMENTS IN DEFORMED GEOMETRIES

New Functionality in Studies and Solvers

New and Improved Results and Visualization Functionality

New Applications in Version 5.2

General Backward Compatibility Considerations

SELECTIONS IN PART INSTANCES

DORMAND-PRINCE 5 TIME-STEPPING SOLVER

Backward Compatibility with Version 5.0

DISPLAY OF MAX/MIN MARKERS IN PLOTS

MERGED MPH-FILE FORMAT

Backward Compatibility with Version 4.4

COMSOL SER VER

LOCAL COORDINATE VALUES

model.physics().create("chds", "DilutedSpecies", "geom1",

Add the following line to specify the identifier accordingly:

Backward Compatibility with Version 4.3b

MODEL NODES ARE NOW COMPONENT NODES

CHANGES TO THE PHYSICS SELECTION IN STUDY SETTINGS

• Provide Degrees of Freedom is now called Disable in Solvers.

REVISED FORMULATION FOR LAMINAR INFLOW AND LAMINAR OUTFLOW

• Batteries & Fuel Cells Module

NEW DEFAULT FOR GRAPHICS OPTIMIZATION

MATH LIBRARIES ON AMD PROCESSORS

Backward Compatibility with Version 4.3a

ERROR ESTIMATION CHANGES IN THE SOLVERS

LOADING EXTERNAL PHYSICS BUILDER JAR FILES

• Delete the META-INF/MANIFEST.MF file in the archive source on the file

HIGHLIGHTING GEOMETRY OBJECTS WHEN DRAWING IN 2D

Backward Compatibility with Version 4.3

AUTOMATICALLY CREATED IDENTITY/CONTACT PAIRS

EFFECTIVE NONLINEAR MAGNETIC CONSTITUTIVE RELATIONS FOR

EXTERNAL NONLINEAR MAGNETIC MATERIALS

New Applications in Version 5.2

EFFECTIVE NONLINEAR MAGNETIC CURVES CALCULATOR

INDUCTION HEATING OF A STEEL BILLET

Backward Compatibility with Version 5.0 and older

COIL CURRENT CALCULATION

• New boundary conditions for Gauge Fixing features