Latest Discussions - COMSOL Forums

laminar flow: a variable in a certain solid domain is undefined

Fri, 17 Apr 2026 13:57:03 +0000

I am conducting heat transfer simulations in a microchannel. The fluid domain is represented by a rectangular block, and the material is water, while the solid domain is made of copper. Using the laminar flow module, after specifying the inlet velocity, the solver returns the following error message:

Undefined variable. - Variable: comp1.spf.U - Geometry: geom1 - Domain: 80 Cannot compute expression. - Expression: comp1.spf.U - Plot: Volume 1

After that, I looked for domain: 80, which points to a fin, but I don't understand why this fin would cause the laminar flow module to fail. Moreover, the fluid properties and the fluid domain in the laminar flow module do not involve this fin. How should I troubleshoot this error?

Flow Boiling Simulation Convergence Problems

Fri, 17 Apr 2026 13:36:17 +0000

I am trying to adapt the extended RPI flow boiling model to my own expiramental setup where i have liquid nitrogen flowing through two channels in a metal test piece, with known heat flux applied to the bottom side. I have replicated the exact setup as the published application. The convergence has failed consistently. I have tried auxillary sweep on the boiling and heat flux, running fully coupled and segregated studies, putting a solver limit on the volume fraction, to no result.

This is how it has been failing:

Undefined value found. - Detail: Undefined value found in the stiffness matrix. There are 111 equations giving NaN/Inf in the matrix rows for the variable comp1.T. at coordinates: (0.0369032,0.029986,0.0268232), (0.0377371,0.0311923,0.026833), (0.0370727,0.0323953,0.0268382), (0.0348952,0.0336535,0.0268406), (0.0341695,0.0347622,0.0268414), ...

and similarly for the degrees of freedom, NaN/Inf in the matrix columns.

here is the application, and screenshots of my own model in an expirament that converged using water, as a reference: https://www.comsol.com/model/subcooled-nucleate-boiling-with-extended-rpi-model-113081

Any advice for convergence with phase change?

The file is too large to send here, but happy to provide more detail to anyone who is willing to help.

thanks!

Cut plane function for AC/DC/charged particle tracing

Thu, 16 Apr 2026 06:49:09 +0000

Hi, I am using the charged particle tracing module from the AC/DC module to simulate an Einzel lens. To get the beam shape i want to have the x and y positions of all my particles for several z values (propagation direction). When i export all data the particles are sorted after time, but do not feature the same z coordinates. I tried to do it with "cut plane", but that didnt work out. How could I do that? I would be very grateful for any help!

Exterior field sum

Wed, 15 Apr 2026 09:52:47 +0000

Hi,

I'm trying to optimize the far-field pattern of multiple acoustic sources (an array of sorts) by tweaking their phases. Rather than simulating the whole thing over and over, I think it could be smart to calculate the radiation of each source individually at first, and then simply calculating the exterior field based on the sum of these with various phases applied. Is this possible?

Carry Over of pre stress for parametric optimisation while using Livelink for Solidworks

Tue, 14 Apr 2026 17:58:50 +0000

I aim to perform a simulation that involves compliant mechanisms (mechanisms that achieve their motion due to their elasticity). Specifically, I am interested in orthoplanar mechanisms. These are flat sheets that have specific cuts to enable motion in perpendicular to plane direction. I have created a Livelink for Solidworks to allow ease of modelling and parametrization.

I have 4 layers of orthoplanar flexure mechanisms. I want to perform a simulation in 2 phases. Phase 1 involves the assembly of orthoplanar flexures. But the catch is the top most orthoplanar flexure and the bottom most orthoplanar flexure are to be pre deflected. The other 2 flexures in the middle are in neutral position. Phase 2 is analysis phase for which I want to connect all four orthoplanar flexures. That is the top most connects to the middle, middle connects to lower middle, lower middle flexure connects to lower most. But for me in phase 2 it is important that the top most and bottom most are pre deflected and pre stressed. So at end of this phase two of my flexures are deflected and two are not deflected. But they are all connected together at the end of this phase.

How can I do this in the most effective way. Keep in mind I am using Livelink for Solidworks and am aiming to perform optimisation on design parameters. I have tried to use remesh deformed configuration but realised that this would not contain the pre stress that I would need. It would be great if you could provide a simple easy-to- implement solution that would integrate in my current workflow with minimal changes.

complex band structure

Mon, 13 Apr 2026 17:32:16 +0000

I am studying complex band structure of a 1D two-layer periodic unit cell consisting of:

One elastic layer One viscoelastic layer (with damping)

My goal is to compute and plot:

Real part of wavenumber (Re(k)) vs frequency Imaginary part of wavenumber (Im(k)) vs frequency

to identify attenuation regions (band gaps).

state space extraction from heat transfer models

Mon, 13 Apr 2026 14:54:41 +0000

Hello

I am trying to extract state space models of a 2D heat transfer models with phase change and radiation boundary condition, my input is laser beam and output is a pyrometer defined as a Gaussian weighted integration over an oval area on the surface.

the issue is that while extracting state space model with pyrometer the deviation between state space and FEM is very large while when using a point probe for extraction the deviation is not much, can you please help me have a better state space for pyrometer?

I use mphstate for model extraction

Electromagnetic simulation on non reciprocal magneto plasmonic material

Mon, 13 Apr 2026 11:17:39 +0000

Hello everyone,

I am positing on this forum because I am new to COMSOL software and I am actually working on meta surfaces. The article I work on is about magneto plasmonic where they use a dielectric material in which they put a ring of MO material in which there is a silver disk inside.

The thing is even though I understand and know the physics concepts behind I have no idea on how to simulate it on COMSOL so I tried to be helped with some tutorial and GPT but it is actually way too difficult since I think it's already advanced simulations.

I tried to create the structure in 2D but, after that, I have absolutely no idea what type of physics, study and mesh I should use and why.

Can someone help me with that or give me some examples or link to tutorial to that type of study?

The article name is "Nonreciprocal isolator based on a plasmonic magneto-optical resonator" if that could help.

I also attach my model file.

Thanks.

simulation of a Steam Generator

Fri, 10 Apr 2026 18:55:35 +0000

Good Evening all,

Me and my team have been perfecting this geometry from a year, and have finally arrived at this model for a steam generator (SG).

The SG is a straight shell and tube type heat exchanger, where water enters into the SG via the central inlet and exits out as steam.

The water stream is pressurised to 80bars. The water is heated through natural convection with the help of lead.

While, the appropriate physics have been added, while trying to simulate it with stationary, and time-dependant studies, our workstation runs out of memory due to the high number of DOFs that are being created even at the coarsest level.

Kindly help us regarding this, on how to simplify this model, or simulate this. We need the heat transfer and fluid profiles for this SG.

The file for this SG model has been attached below:

the mph file is 29MB even with the physics removed (it was 1.9GB with the physics and extra-coarse mesh) ...... what do I do?

Pressure Field - CFD

Fri, 10 Apr 2026 02:38:54 +0000

Are the pressure field results from the CFD simulation static pressure or total pressure?

Compression plots

Thu, 09 Apr 2026 18:19:54 +0000

Hello,

I am simulating a model to see the interaction of electric field of a tip (like AFM tips) on a electrostriction substrate surface. In the model, I am trying to insert a defect in the material. I need a plot of the strain or compression the surface suffers when the tip is polarized, which characterizes a electrostriction effect. But in the plot I could not find a way to plot the strain in 2D, just the stress. How can I plot the plot equivalent a compression effect in th results?

I have another question regarding the electrostriction effect: how can I add this effect? I used the coupling for electrostriction and added the Q electrostriction tensor, is this enough? It seems to me that I am missing something.

I am sending a printscreen of the physics I am aplying to my model, as the geometry of my problem too.

Hope someone can help, thank you. Hugs, TMV

Message (plus 1 internal DOFs)

Thu, 09 Apr 2026 15:57:12 +0000

What is the meaning of the message "(plus 1 internal DOFs)" ? Is it something I should be concerned about?

Plasma time periodic

Thu, 09 Apr 2026 14:28:39 +0000

Hello,

I am trying to simulate a CCP plasma, I looked into the COMSOL examples and they use cross sections for evaluating the coefficient rates, but instead I want to use rate constants which are dependents from the electron temperature. I tried to put directly in the equation space for the rate constant as ptp.Te, but I am having problems with division by zero. I am trying to see where is the zero problem, but I am not finding, can someone help me?

Thank you,

TMV

Wrong Q factor

Thu, 09 Apr 2026 10:27:00 +0000

I am simulating 1/4 of a SiN membrane suspended with 4 thin arms and clamped to the surrounding substrate, the membrane is very thin (100nm) but very wide and long (high aspect ratio managed using Swept function in the mesh) and the entire structure has an isotropic damping equal to 1/Q where Q is set to 100.

Performing an eigenfrequency study, I am expecting to find for all the modes a value of Q (that COMSOL evaluates automatically) equal to the value I set (100). Unfortunately, for the fundamental mode this doesn't happen; precisely, the Q of the fundamental mode is sometimes higher (120,130,150) and sometimes smaller (70, 80) than 100. In addition, every time I launch a new simulation run (even though I do not change nothing) the value of Q (and consequently the Real and Imaginary parts of the eigenfrequency) changes.

Is it normal? Is it a big issue?

Installation problem COMSOL Multiphysics 6.3

Mon, 06 Apr 2026 13:50:32 +0000

When I am installing COMSOL Multiphysics 6.3, I face these kinds of error: for reference I am attaching the screenshoot of an error.

COMSOL 6.4 Error

Sun, 05 Apr 2026 10:47:39 +0000

Cannot set up file java error java.lang runtime.exception.invalidroot directory cd : c

component variables and global parameters for this .mph file

Thu, 02 Apr 2026 21:34:37 +0000

I have access to v6.2, and this simuulation is uploaded to the COMSOL application gallery only as 6.4. I can still use the .txt file containing all of the formatted equations for rpi model phase change at the wall and in the 3d space as well as corresponding heat transfer equations. The phase change is modeled in the formatted equations that are defined in the parameters and in the variables.

Can someone using v6.4 send a .txt export of all of the parameters and variables?

This is the simulation i am referencing from the application gallery: https://www.comsol.com/model/subcooled-nucleate-boiling-with-extended-rpi-model-113081

Laminar Flow Not Affecting Electroanalysis Results — Dean Flow Not Observed

Wed, 01 Apr 2026 14:58:52 +0000

I am modeling a microfluidic electrochemical system in COMSOL using a two-step study: (1) a stationary Laminar Flow (spf) solution to obtain the velocity field, followed by (2) a time-dependent Electroanalysis (tcd) study for cyclic voltammetry (CV). The intention is for the velocity field from the laminar flow solution to be used in the electroanalysis step so that convection influences species transport and, consequently, the electrochemical response.

However, I am not observing any change in the CV results when varying flow rate or channel curvature (including geometries where Dean flow is clearly present in the velocity field). The concentration profiles at the working electrode appear unchanged across different flow conditions, suggesting that convection is not being accounted for in the electroanalysis solution.

In the time-dependent study step, Laminar Flow is unchecked (not solved), and I have set “Values of variables not solved for” to use the solution from the stationary study. Despite this, the electroanalysis results remain insensitive to flow.

Is there something I may be missing in properly coupling the laminar flow solution to the electroanalysis physics (e.g., enabling convection in species transport, required multiphysics coupling, or solver configuration)? Any guidance would be greatly appreciated.

Explicitly Modeling Reference Electrode

Wed, 01 Apr 2026 14:43:12 +0000

Hello everyone,

I am attempting to model a three electrode Microfluidic Electrochemical Sensor (MES). I am able to model the working and counter electrodes and get some results, but I believe they are incorrect. One reason I am thinking is because I am not explicitly assigning a boundary condition to the reference electrode. Is there a way to explicitly assign a boundary for the reference electrode and assign the correct physics? Any help would be greatly appreciated. I am using the electroanalysis physics.

How to extract the stiffness matrix K in multi-body dynamics (mbd) Transient simulation using MATLAB?

Wed, 01 Apr 2026 08:49:15 +0000

As the title indicates, when I attempted to use the mphmatrix function, I encountered an error: "The solution does not contain an active Variables node"