Latest Discussions - COMSOL Forums

Issue with Laminar Flow in 2D Axisymmetric Model Depending on Inlet and Outlet Positions

Sun, 27 Apr 2025 07:42:10 +0000

Hello,

I am working on a 2D-axisymmetric model using the Laminar Flow module in COMSOL, coupled with the Plasma and Heat Transfer modules.

I have noticed an issue: when the inlet and outlet are positioned far apart or diagonally across from each other (see the first image), the simulation fails and throws errors. However, when I place the inlet and outlet closer together (as shown in the second image), the simulation runs successfully without any problems.

Based on my observations, I suspect that the issue might be related to mesh quality. I tried refining and adjusting the mesh multiple times, but the simulation still does not converge when the inlet and outlet are distant.

Is this a known issue when coupling Plasma and Heat Transfer with fluid flow in a 2D-axisymmetric setup? Or is there a recommended strategy for meshing or modeling when the inlet and outlet are positioned far apart?

Any advice or tips would be greatly appreciated. Thank you very much!

Error: Underdefined material property 'epsilonr' required by Charge Conservation 1.

Fri, 25 Apr 2025 20:49:44 +0000

To whom it may concern

I am trying to simulate the static electric filed distribution for my first step. It was working. But after the COMSOL automatically removes the small details in my model geometry, the Error shows up. The Error is "Undefined material property 'epsilonr' required by Charge Conservation 1." I checked all my Material Property setting. All materials have epsilonr values and all geometry have been assigned materials. I attached my mph file. Could COMSOL experts help me out? Thank you very much for helping.

Problem with saving table in a method

Fri, 25 Apr 2025 14:46:44 +0000

Hi everyone, I'm currently working on creating small scripts using methods in the Application Builder to simplify the export of data in the Model Builder. I call the equivalent method in the Model Builder to get it running.

Right now, I'm trying to save a table using the following script and ran into two issues:

model.result().evaluationGroup().create("eg10", "EvaluationGroup"); model.result().evaluationGroup("eg10").set("data", "dset1"); model.result().evaluationGroup("eg10").label("Iron Losses"); model.result().evaluationGroup("eg10").create("int1", "IntSurface"); model.result().evaluationGroup("eg10").feature("int1").label("Surface Iron Losses"); model.result().evaluationGroup("eg10").feature("int1").set("intvolume", true); model.result().evaluationGroup("eg10").feature("int1").selection().named("sel15"); model.result().evaluationGroup("eg10").feature("int1").set("expr", new String[]{"rmm.Qh"}); model.result().evaluationGroup("eg10").feature("int1").set("descr", new String[]{"Volumetric loss density, electromagnetic"}); model.result().evaluationGroup("eg10").feature("int1").set("unit", new String[]{"W/m"}); model.result().evaluationGroup("eg10").feature("int1").setIndex("expr", "rmm.Qh*L", 0); model.result().evaluationGroup("eg10").feature("int1").setIndex("descr", "P_iron_MOD", 0); model.result().evaluationGroup("eg10").run();
//model.result().evaluationGroup("eg10").save("Z:\Simulationsergebnisse\Iron_losses.txt");
//model.result().evaluationGroup("eg10").save("C:\Users\kzt\Documents\Iron_losses.txt");
//model.result().evaluationGroup("eg10").save("C:\Temp\Iron_losses.txt");

1.) Everything works fine up to the last commented-out line. Unfortunately, I can’t save to any of the specified directories and I always get the following error:

"Cannot open file for writing. - Filename: Z:\Simulationsergebnisse\Iron_losses.txt (with all the other paths as well) - Table: Iron Losses"

However, I can manually save to the paths via mouse click.

2.) Is there an alternative to the following line?

"model.result().evaluationGroup("eg10").feature("int1").selection().named("sel15");"

I’d like to use the name of the selection instead of the index number 15. When I add or remove selections, the index changes, so "15" may no longer refer to the correct selection.

Best regards and many thanks in advance for your help!

Tobias Zeller

Modelling charge transfer in a slag-crucible (solid-liquid) interface

Fri, 25 Apr 2025 12:51:31 +0000

I want to model the electronic conduction of my slag (essentially a mixture of metal oxides, containing FeO/Fe2O3 which allow for electronic conduction (i.e. electron transfer)). However in my tests I see that my platina crucible is interferring with my actual slag conduction. I want to model this test but I don't know exactly how to model the (possible) charge transfer at the slag-crucible interface. Anyone has any ideas/experience in how to model this? Thanks in advance.

Modeling induced stresses

Fri, 25 Apr 2025 12:06:00 +0000

Hi everyone, I am trying to do a 1D simulation of induced stresses in the carbon fiber anode and LiFePO4 composite matrix cathode of a structural composite battery due to Li-ion intercalation/deintercalation using COMSOL. Carbon fiber acts as the negative electrode and is transversely isotropic. I have a value of intercalation-expansion coefficient B_j and each component's electrochemical and mechanical properties. I have not been able to solve the following points. 1. I have not been able to define transversely isotropic materials of carbon fiber anode. 2. I have not been able to assign swelling strain to the console. 3. I'm not able to couple lithium battery physics with solid mechanics.

The displacement and stress solutions obtained for each phase must satisfy the following. Interface conditions: (i)Radial displacement must be zero on the symmetry axis. (ii)Radial displacement and radial stress continuity at all interfaces. (iii) The outer boundary of the cylinder assembly is stress-free. In the equation, where ck is the time-dependent Li concentration in the active material, ck, max is the maximum concentration corresponding to Li intercalated into all available sites in the host materials. index k-m for the matrix, k-c for the coating(solid electrolyte), and k-f for the fiber in the cylindrical coordinates.

I need to follow the variable plot during charging and discharging. 1. Li distribution in fiber radial position. 2. Lithium concentration in the carbon fiber center and at the surface. 3. Li distribution in the matrix during loading . 4. Stresses at fiber boundary. 5. Stress distribution along fiber radial direction

I request to you please guide me how can i perform this simulation.

Eigenvalue solver 1 (sol1/e1) No eigenfrequencis found

Fri, 25 Apr 2025 11:34:02 +0000

Hello we are trying to complete a lab moment in our course in mechanical physics. In this course we are trying to build a cube sat and try to find eigenfrequencis but our lab instructions are very unclear anyone got an idea?.

Build the model

Fri, 25 Apr 2025 10:13:23 +0000

Hello, I have some questions to ask: While reading an article, I saw a 2D model built using COMSOL software, as shown in the figure. How can I construct such a 2D dipole antenna in COMSOL? Are there any related tutorial cases available? Thank you very much for your help!!

How to implement Hoffmann & Lauritzen's crystallization model and integrate the absorption coefficient over the thickness

Thu, 24 Apr 2025 11:03:38 +0000

I am working on laser welding two blocks of the same measurements placed on top of each other, one being semi-transparent and the other absorbing. The absorption coefficient depends on the crystallinity rate, and my goal is to see how the source term evolves over the thickness by integrating the absorption coefficient over the thickness and implementing crystallization in COMSOL and coupling the models.

Topology optimization Warning

Thu, 24 Apr 2025 02:36:15 +0000

Hi Everyone, I am working on a Topology Optimization setup of model (that consists of two materials) that undergoes Induction Heating(Physics involved: electromagnetic heating) in two step study ( Step1: Coil Geometry analysis, Step2: Frequency-Transient Solver, the optimization is applied for step 2). I have setup the settings for the required density model for topology optimization as following: theta_min: 0.01 theta_initial: 0.5

Control Variable Constraints: 0.9>=theta_average>=0.5 The optimization(MMA solver) starts properly at first iteration, but after that depending on the initial value given, after some iterations, I start getting this "Warning: Forward solution failed, requesting reduced step." This persists continuously and eventually I have to cancel the compute.

I would really grateful any help and please let me know if you need any further information regarding this.

Best, Ashok

An Additional Matrix "Constraint Jacobian" In Shell Module

Wed, 23 Apr 2025 17:40:53 +0000

Hi,

Recently, when I checked the process of "Elimination Constraint Handing" in Shell Module, I found something confused.

As a toy model, I just built a plate in the Shell Module and no boundary conditions are applied. However, in the matrix "Constraint Jacobian", an additional constraint was used, i.e. . What's the meaning for this constraint?

In my knowledge, it implies the displacement component u_3 should be equal to 0 for all nodes. Actually, the surface plot in "Results" tells me this is misunderstanding.

Any guidance would be very appreciated. Thanks!

Time-Dependent Solver doesn't progress from 0%

Wed, 23 Apr 2025 17:40:52 +0000

Hi everyone,

I'm currently working on a 3D simulation (+Magnetic Fields (mf) interface) involving an oscillating object along the z-axis, surrounded by deforming air domains. The object motion is prescribed using a sin(2pift) deformation, also for the upper and lower "air slugs" with (I hope) the correct deformation functions. The simulation needs to be in 3D since I'm gonna add asymmetric coils to the geometry to simulate induction voltages in pickup coils.

I've added a stationary solver at the beginning, and producing a stationary magnetic field plot is fine. Everything also seems fine geometrically and mesh-wise, but my time-dependent solver fails immediately with the following error:

Failed to find consistent initial values.Singular matrix.For mesh case 0 there are 72 void equations (empty rows in matrix) for the variable comp1.mf.mag1.Vc_d0...

I correctly added all the magnetic insulations boundaries, continuity bound pairs and right now I am at a loss for solutions. I would be happy if someone with experience helped me out.

Best regards.

How to pass the temperature when use the external material interfaces for structure module?

Wed, 23 Apr 2025 13:23:25 +0000

I want to implement a new shape memory alloy material model, where the parameters denpend on temperature. Is there any way to pass the solved temperature to the general strain-stress interface?

Antenna calculation Frequency combination

Tue, 22 Apr 2025 13:07:01 +0000

I would like to consult regarding a technical issue concerning dual-band antenna simulation using COMSOL's RF Module. Currently, I'm modeling an antenna that needs to operate simultaneously across two frequency bands: 0.6-1.0GHz and 1.6-2.0GHz, with the goal of obtaining composite simulation results.

Specifically, I need to:

Apply a single excitation source that simultaneously covers both frequency bands

Analyze the antenna's radiation characteristics under this dual-band operation

Acquire the superimposed field distribution results

Could you please advise on the proper configuration for implementing this multi-frequency analysis in COMSOL? Your guidance on the setup methodology would be greatly appreciated.

Thank you very much for your assistance!

Polypropylene PP material library

Tue, 22 Apr 2025 07:29:01 +0000

Hello, I am looking for polypropylene in the material library but can not find it. How to have it ? Thanks

How to Export Stiffness and Mass Matrices in COMSOL Eigenfrequency Study

Tue, 22 Apr 2025 01:52:51 +0000

Hello,

I'm using COMSOL Multiphysics to perform an eigenfrequency analysis of a cryogenic suspension beam for a gravitational-wave detector project. I want to export the stiffness matrix (K) and mass matrix (M) for each mode, so I can compute the energy using the quadratic form:

externally.

Issue: I followed these steps:

Created an Eigenfrequency study (with or without prestress).
Added a Matrix Export node under "Derived Values."
Selected “Stiffness Matrix” and “Mass Matrix” with options "Global" and "Sparse."

However, when I try to export, COMSOL gives the error:

"K is not stored."

I’ve also tried switching to frequency-domain or stationary studies and adding Input Matrix nodes under the solver configuration, but nothing has worked so far.

My Questions:

Is there a correct way to configure the eigenfrequency study so that the stiffness and mass matrices are stored and exportable?
Does this feature require enabling something before solving?
Is there any workaround if I don’t have LiveLink for MATLAB or Java API access?

Any guidance would be very appreciated

Can COMSOL prestressed Eigenfrequency Analysis Provide Valid Elastic Energy Estimates for Mode Shapes?

Tue, 22 Apr 2025 01:40:56 +0000

Hello,

I'm working on a cryogenic suspension design for a gravitational wave detector and using COMSOL Multiphysics to study internal vibration modes. I'm trying to determine whether the elastic strain energy reported in an Prestress Eigenfrequency Study can be reliably used to estimate the thermal motion of modes at cryogenic temperatures using energy scaling — for example, by scaling mode shapes such that the total energy equals

A senior collaborator raised this concern:

“Eigenvalue analysis is a frequency-domain calculation, and energy is not properly defined because initial conditions are not included. Therefore, the energy returned by COMSOL may not be meaningful.”

I understand that eigenfrequency mode shapes in COMSOL are arbitrarily normalized. However, COMSOL does internally compute the strain energy density and integrates it over the domain to report a total elastic energy per mode. I have already exported this using a volume integral of:

My questions:

Does this computed strain energy in COMSOL prestress eigenfrequency studies represent the actual mechanical energy stored in the mode shape (up to normalization)?
Is it valid to rescale the eigenmode displacements using:

If not, what is the rigorous way to obtain mechanical energy or displacement amplitudes from eigenmode simulations?
Are there COMSOL resources, papers, or documentation that clarify the assumptions and limits of using eigenmode strain energy for physical interpretations?

I understand that damping, Q-factor, and time-domain dynamics are not included in the eigenfrequency solver, but I’m only interested in using the elastic energy to scale the displacement fields to match thermal energy at 4 K.

Thank you for any help or guidance!

power extraction from PN junction using Electric Circuit interface

Mon, 21 Apr 2025 11:22:03 +0000

Hello everyone,

I’m currently working on a project in COMSOL where I simulate multiple PN junctions connected in a serpentine layout. I apply a thermal gradient (Seebeck effect) and successfully generate a voltage (~70 mV), but I’m struggling to calculate the extracted power.

I’ve tried adding the Electrical Circuit interface and connected a resistive load, but COMSOL won’t compile. I suspect that the circuit isn’t properly linked to the 3D physics, or I may be missing a terminal or an external connection setup.

Has anyone worked on a similar model or could point me to a tutorial/example showing how to connect a terminal from a thermo-electric model to an external circuit and extract power?

Any help or guidance would be greatly appreciated.

Thank you!

LiveLink for MATLAB: mphplot produces no legend

Mon, 21 Apr 2025 10:46:32 +0000

The following code with mphplot(model,'pg1') plots 2 lines in a figure properly but doesn't show the legend. What's the problem?

model.result.create('pg1', 'PlotGroup1D'); model.result('pg1').create('lngr1', 'LineGraph'); model.result('pg1').feature('lngr1').set('xdata', 'expr'); model.result('pg1').feature('lngr1').selection.set([1]);

model.result('pg1').set('looplevelinput', {'manual'}); model.result('pg1').set('looplevel', {'1, 101'}); model.result('pg1').set('xlabel', 'x-coordinate (m)'); model.result('pg1').set('ylabel', 'Dependent variable u (1)'); model.result('pg1').set('legendpos', 'lowermiddle'); model.result('pg1').set('xlabelactive', false); model.result('pg1').set('ylabelactive', false); model.result('pg1').feature('lngr1').label('Line Graph'); model.result('pg1').feature('lngr1').set('xdataexpr', 'x'); model.result('pg1').feature('lngr1').set('xdataunit', 'm'); model.result('pg1').feature('lngr1').set('xdatadescr', 'x-coordinate'); model.result('pg1').feature('lngr1').set('linemarker', 'point'); model.result('pg1').feature('lngr1').set('markerpos', 'datapoints'); model.result('pg1').feature('lngr1').set('legend', true); model.result('pg1').feature('lngr1').set('resolution', 'norefine');

Non linear solver does not converge, and simulation is very very slow

Thu, 17 Apr 2025 11:48:26 +0000

Simulation is very very slow also non linear solver convergence error. When i try to increase the time for simulation, then i often encounter "non linear solver does not converge" error. I modified some things in segregated solver but that is making simulation very very slow. How can i solve that problem.

Handling Geometry or Meshing Failures in MATLAB-COMSOL Model

Wed, 16 Apr 2025 20:42:02 +0000

Hi all,

I’m running a MATLAB-COMSOL model where I update my design variables using a command like:

% Set the variables nPoints = size(xy, 1); % Number of point pairs for i = 1:nPoints xi = xy(i,1); % x-coordinate yi = xy(i,2); % y-coordinate

model.param.set(sprintf('x%d', i), sprintf('%0.4f [mm]', xi));
model.param.set(sprintf('y%d', i), sprintf('%0.4f [mm]', yi));

end

However, in some cases, certain combinations of design variables lead to non-viable geometries. As a result, the model fails either during geometry generation or meshing, and I get an error when trying to evaluate:'

mphglobal(model, 'solid.LFcrit');

To handle this, I would like to check whether the geometry and mesh were successfully generated before calling mphglobal. If they weren’t, I want to skip that part so that MATLAB can continue iterating without crashing the whole optimization loop.

Is there a recommended way to check the status of geometry and mesh generation in COMSOL via the MATLAB API?

Any suggestions would be much appreciated!

Best regards, Mirhan