Latest Discussions - COMSOL Forums

S parameter extraction from 3D model + external circuits

Mon, 20 Oct 2025 19:00:11 +0000

Hi everyone,

I’m modeling a two-port piezoelectric resonator that functions as a transformer (per the mBVD representation). I’ve connected the device to an external matching network with a series inductor at the output to cancel Cout, and I’d like to extract the S-parameters of the combined device.

So far: Using Terminal (power) with Terminated boundaries on the 3D structure gives me S-parameters for the bare 3D model only, not the coupled device-and-circuit.

I also tried an Electromagnetic Waves, Frequency Domain (EMW) study, but I’m not confident this is appropriate for a primarily lumped, electromechanical two-port with an external circuit.

Question: What is the recommended workflow to obtain S-parameters when a 3D piezoelectric resonator is connected to external circuitry?

Thank you for any guidance.

induction heating

Sun, 19 Oct 2025 06:52:30 +0000

Why, during the process of conducting the simulation of workpiece induction heating, as the workpiece moves, the temperature of the induction heating part of the coil keeps rising continuously, and it fails to reach a state where the heat input and output are balanced, thus preventing the temperature at the heating location from remaining constant.The 2-D rotational modeling that I use.Set up heat radiation and heat flux for heat dissipation. Thanks for help!

Electric field not reaching material across air gap (Electric Currents, Frequency Domain)

Sat, 18 Oct 2025 04:01:01 +0000

Hello everyone,

I’m currently working on a model in COMSOL Multiphysics (AC/DC Module – Electric Currents, Frequency Domain) and I’m facing an issue where the electric field does not seem to penetrate the material located across a small air gap.

Here is the setup: -Two planar electrodes (terminal and ground) placed side by side on a PCB substrate. -Above the electrodes, there is a thin polymer layer, then a 2 mm air layer, and finally a block of bone material. -Everything is enclosed by a cylindrical air domain. -Excitation: 3.3 V at 32 kHz (frequency domain study).

When the air layer is present between the electrodes and the bone, the electric field and current density completely vanish inside the bone, only the region in the air shows nonzero values. If I replace the air layer by the bone (so there’s direct contact), then I see the expected electric field inside the bone and i can see changes in capacitance, even when changing bone properties and material.

The strange part is that if I artificially set the air conductivity to something like 0.001 S/m, the field appears inside the bone again. But with σ_air = 0 S/m, it disappears entirely and capacitance dont change.

Thanks in advance for any advice or insight, I’ve been stuck on this for a while and would really appreciate any suggestions on how to make the field correctly propagate through the air and into the bone layer.

Best regards, Diogo Pires

Shape changes with Temperature in COMSOL using Joule heating

Thu, 16 Oct 2025 22:02:26 +0000

The below post is related to an archived discussion

[start here] Hi, Sorry to follow up now after a long time. I could not work with the model. I am not actually looking for thermal expansion. My problem is a little different. Let me explain- My model starts with a regular octagonal shape, with side marked as 1,2,3,4,5,6, 7 and 8. After modeling, I am expecting the shape to convert to a rectangular shape. Considering both shapes will have same volume, after modeling side 3 and 7 will move to -x and +x respectively. The change in length of the sides will be function of domain temperature. I am attaching an image here. Can anyone please guide where to start with this problem?

Transient study - Convergence issues

Thu, 16 Oct 2025 11:47:48 +0000

Dear all,

I am experiencing an issue with the model attached to this request. My goal is to analyze the behavior and response of a superconducting tape under the action of an external field generated by appropriately applying currents in the system’s exciter.

When I modeled the superconducting region by approximating it as a line, I did not encounter any problems. However, now that I am treating it as a rectangular domain (also because I plan to include additional non-superconducting layers in the future), the model shows convergence issues.

I am fairly confident that the formulation is correct as implemented, since I have verified it against several references in the literature, and it is consistent with those sources.

Could you please help me resolve the issue related to Study 2 – Transient? I have tried setting the stationary Study 1 as the initial condition and tested various solvers, but without any positive results.

Thank you very much for your support.

Geometry Repair

Wed, 15 Oct 2025 14:13:32 +0000

Hi all

Could you explain an effective procedure to repair a geometry (files: *.stl or step) in order not to produce errors such as "intersection faces", and reduce elements or size of the geometry to improve compilation time ?. I've tried already several strategies as reparation in Meshlab and Freecad platforms, but not succefull so far. I attached the geometry file for consideration... Thanks

Reg- Microbalance sensor

Tue, 14 Oct 2025 03:05:34 +0000

Dear Staff, Good morning, I would like to simulate Mass sensor based on graphene. Could you assist me or send any related model to start.

thank you

TFLN rib waveguide - optical field results not showing

Mon, 13 Oct 2025 10:08:50 +0000

Hello everyone, I am trying to simulate the optical field of a rib waveguide in lithium niobate. I attached an image of the cross section, together with the mph file. I first defined a work plane and extruded it to create the structure. I defined the materials and inserted a diagonal refractive index for the lithium niobate since it is an asymmetrical material. In the end I defined the ports and inserted the mesh, and launched everything.

The problem is that the program runs and ends the simulation, without any shown error, but at the same time it doesn't compute the electrical field. In fact, the electrical field section under "Results" doesn't exist. Why is that?

Solver Convergence Issue in 2D Brinkman Flow through Anisotropic Porous Medium (Wavy Channel)

Mon, 13 Oct 2025 05:07:18 +0000

Hello,

I’m simulating steady-state flow in a 2D wavy channel using the Brinkman Equations interface (COMSOL Multiphysics 6.0). The flow domain is an anisotropic porous medium confined between two impermeable wavy walls.

MODEL SUMMARY: * consists of a wavy bottom that i defined using parametric curve * upper top plate is a moving plate with velocity lets say U0 in x direction and 0 in y * the left and right walls are not moving * No slip condition is applied for all * Physics used is brinkaman equation where the porous matrix is set to be a diagonal matrix where the diagonal entries were already defined in the parameters under global definitions. * Physics controlled mesh is used. * stationary study is added.

ERROR:

Failed to find a solution. Maximum number of Newton iterations reached. There was an error message from the linear solver. The relative residual (0.041) is greater than the relative tolerance. Returned solution is not converged. Not all parameter steps returned.

TRIED SO FAR: * Adjusted tolerances and damping * switched between direct solvers * increased iteration limits

QUESTIONS: * Recommended solver setup (Fully Coupled vs Segregated) for Brinkman flow? * Tips for stabilizing convergence in anisotropic, wavy porous domains? * Any scaling or nondimensionalization suggestions?

Any help to stabilize or improve convergence for this steady Brinkman model would be appreciated.

Regards!

Regarding the issue of the presence of both the paranoid coil and the incentive coil simultaneously

Sat, 11 Oct 2025 09:30:41 +0000

I want to conduct a simulation of a transducer. The overall structure is a cylinder. The coil part consists of an inner bias coil and an outer excitation coil. The bias coil provides a constant bias magnetic field, while the excitation coil provides an alternating magnetic field. How should I set up the bias coil to be energized by direct current, and the excitation coil to be energized by alternating current of a certain frequency?

How to add multiphase material in COMSOL 6.0

Sat, 11 Oct 2025 08:30:07 +0000

I am trying to replicate the capillary filling model for two phase field in COMSOL 6.0. How to add a multiphase material? If we add from material library, how to specify the phases?

Potential Field in Electro-osmotic Flow

Fri, 10 Oct 2025 09:19:25 +0000

Hi all,

Currently I was trying to reproduce results from a research paper titled from Choi et al. (2011), (attached below). In that paper, two boundary conditions are used at the interface for electric potential. These are (i) potential jump and (ii) simplified gauss law at the interface. Now, while modeling using the 'es' physics in COMSOL, I am not able to see how to model both the conditions at the interface.

I have attached my trial file, in which when I tried putting these conditions at the identity pair, the surface charge condition (gauss law) was getting overridden or fallback-ed due to the potential jump condition. From my understanding, COMSOL isn't letting me put both at the same time. I tried using a coefficient-form PDE to solve it too but it had the same issues with the interface conditions.

What could be a potential fix to this? Is there another way to model the system as given in the paper?

Dual gate MOSFET design

Wed, 08 Oct 2025 05:08:55 +0000

Can anyone give any guideline regarding the design of dual gate MOSFET in COMSOL?

Gaussian Beam Plane Wave Expansion

Tue, 07 Oct 2025 21:21:26 +0000

Hi, I just recently learnt that the paraxial gaussian beam equation is not really effective for a scatterer surrounded by inhomogeneous domains when using the scattered field formulation. it was proposed I use the plane wave expansion approach. for a 3D model, has anyone used the plane wave expansion approach using equations. I have attached my model of the paraxial approach. you can edit my model to suit the plane wave expansion method or if you have any example. That will be great.

NaN/inf value error

Mon, 06 Oct 2025 08:40:48 +0000

I'm trying to simulate a cylindrical gas bearing with a sort of foil "floating" on it but I keep getting the same error: Undefined value found. - Detail: Undefined value found in the equation residual vector. There are 7 degrees of freedom giving NaN/Inf in the vector for the variable comp1.spatial.u. at coordinates: (1.94255,-0.112035,0.0920496), (1.94284,-0.111364,0.09286), (1.94356,-0.111577,0.092605), (1.94329,-0.111264,0.0930404), (1.9433,-0.111842,0.0923323), ...

I don't think my mesh is the problem as well as my boundary conditions. I suspect it is something to do with my geometry although I also have the feeling that is correct as well. The model was working when I used the fluid-structure interaction model but I wanted to model smaller thickness of the foil and decided to change to fluid-shell interaction but then it did not work anymore whilst the boundary conditions basically have not changed compared to the previous model that did work. To get a surface from my previous model I have used the tool "Convert to surface" where I had a finite thickness ring around the gas bearing. What could be the problem here and where should I be looking to solve this problem?

Issue With Object Contact

Sun, 05 Oct 2025 22:49:07 +0000

Hi everyone,

I have been trying to model a basic piezoelectric compression model involving two domains. I have one block on top which represents a steel mass and another block on the bottom which is meant to model piezoelectric paper.

I am very new to using COMSOL and am currently having trouble getting the two blocks to connect. When I apply a boundary load to the piezoelectric paper block and perform a stationary study, it works just as intended. However, when I place the block on top (in the assembly) and apply a Continuity to both domains of the Identity Boundary Pair (the two touching faces), I can't seem to get a solution that converges. I also apply gravity to both objects and put a fixed constraint of the bottom face on the paper.

I can imagine this issue is likely due to the meshing as I am currently trying to use the Normal, Physics-controlled mesh. I wanted to try to build a User-controlled mesh in which the mesh gets finer when it is close to the point of contact for both objects. However, when ever I try to build a User-controlled mesh it never seems to work (the mesh is not physically displayed).

I can't imagine that the fix is too challenging, but I have been stuck on this problem for quite some time. If anyone has any suggestions or ideas, that would be greatly appreciated!

Implementation of Weak Form PDE from PDE interface for battery charging discharging cycle

Thu, 02 Oct 2025 14:00:04 +0000

Hello,

I'm trying to build a Lithium-ion Battery model using Weak Form PDE from PDE interface in mathematics module, not using the lithium ion battery physics in Batteries module.

I want to implement a charge-discharge cycling condition using Weak Form PDE from PDE interface in mathematics module, but I don't know how to do this.

I know the way to applying the charge-discharge cycle in Lithium Ion Battery Physics, but I want to know the way to applying the charge-discharge cycle using Weak Form PDE from PDE interface Physics.

Please provide the way, how can i Implement this?

Thanks for reading,

Prakash

Plot of temperature along a line inside of a thin layer

Wed, 01 Oct 2025 16:20:37 +0000

Hello

How can we plot the temperature along a line inside a thin layer? The line is parallel to the thin layer surface. I manage to do the surface graph of the temperature on the layer, but I would like to do the plot of the temperature only on a line. I've tried several approaches with the command atxd but they didn't succeed.

Thank you

Mass conservation in COMSOL

Tue, 30 Sep 2025 20:32:41 +0000

Hi! I am using Heat transfer in solid and fluids module where phase changes from solid to fluid with larger density difference. I have integrated Laminar flow too to account for fluid movement. The phase change is working and the mesh size remains the same. How is it possible as the Laminar Flow module uses mass conservation equation too without any change in mesh shape? Thanks in advance.

Conditional parametric sweep

Tue, 30 Sep 2025 16:03:11 +0000

I'd like to run a parametric sweep with a couple of conditions. Here is an example of what I'm trying to do.

a = range(1:1:10) b = range(1:1:10)

The condition I want to add is a+b<15

Can I do this directly in COMSOL?

Alternatively I generated the combination on MATLAB but when I am having the file read by COMSOL through a saved text file. It is saying Failure reading file. '1' is not a valid variable name.

Could you please help me troubleshoot either of my methods. I'm wondering if the way the data is output from MATLAB is not compatible with COMSOL.

Thanks and regards VK