Hello,

Has anyone had success phase unwrapping within COMSOL? We can do so through livelink, but we would like to extract the adjoint sensitivities using a phase front profile as the goal, so unwrapping within COMSOL would be best. With it wrapped, this can play havoc with sensitivities near the phase jump.

In the process of using COMSOL to simulate electron acceleration, after I imported the electric field defined by myself, checking the relativistic correction will cause some electrons to stop after a certain distance. If I do not check the relativistic correction, everything will be normal. What is the reason for this problem?

Hello I'm doing an Eigenfrequency Prestressed study of a rod and it is giving me this error

**Failed to find a solution. The relative residual (1.9) is greater than the relative tolerance. There was a warning message from the linear solver. Iterative refinement triggered. Returned solution is not converged. Not all parameter steps returned. **

here are the boundary condition that is given 1. The rod has a rigid connector boundary that constrained rotation in x y and z direction place at the bottom of the rod 2. the rod has an added mass boundary of 50 kg at the bottom of the rod 3. gravity is used, set at -9.82 m/s^2 in y direction 4. the material used is anisotropic silicon

The mesh is set at extra fine

attached is the rod mph.file that is used.

The below post is related to an archived discussion

[start here] Undefined variable. - Variable: freq - Geometry: geom1 - Domain: 1 Failed to evaluate variable. - Variable: comp1.mat1.rfi.n11 - Defined as: sqrt(((1+((comp1.mat1.DispersionModelSellmeierStandard.ODsma1((c_const/(frequnit_um_cf))^2))/(((c_const/(frequnit_um_cf))^2)-comp1.mat1.DispersionModelSellmeierStandard.ODsma4)))+((((c_const/(frequnit_um_cf))^2)comp1.mat1.DispersionModelSellmeierStandard.ODsma2)/(((c_const/(frequnit_um_cf))^2)-comp1.mat1.DispersionModelSellmeierStandard.ODsma5)))+((((c_const/(frequnit_um_cf))^2)comp1.mat1.DispersionModelSellmeierStandard.ODsma3)/(((c_const/(freq*unit_um_cf))^2)-comp1.mat1.DispersionModelSellmeierStandard.ODsma6))) Failed to evaluate variable. - Variable: comp1.gop.N_localxx - Defined as: comp1.mat1.rfi.n11 Failed to evaluate variable. - Variable: comp1.gop.nabs_local_real - Defined as: ((comp1.gop.N_localxx+comp1.gop.N_localyy)+comp1.gop.N_localzz)/3 Failed to evaluate variable. - Variable: comp1.gop.nabs_local - Defined as: comp1.gop.nabs_local_real Failed to evaluate variable. - Variable: comp1.gop.nref_local - Defined as: comp1.gop.nabs_local Failed to evaluate expression. - Expression: real(comp1.gop.nref_local) Failed to evaluate operator. - Operator: env_out_pgeom_gop - Geometry: Rays Failed to evaluate expression. - Expression: env_out(real(comp1.gop.nref_local))

Please help me to solve this

I would like to numerically calculate the scattered intensity field inside a suspension when a 385nm UV laser is applied to the surface of a particle-filled suspension with a Gaussian beam intensity profile. The particles are ceramic particles like alumina with a d50 size ranging from 150nm to 500nm. Between the Wave Optics Module and the RF Module, which optics module would be more suitable? Is the Electromagnetic Waves, Frequency Domain (ewfd) interface alone sufficient for this simulation?

Hi everyone,

I'm working on building a model that simulates a confined space with a custom, arbitrary time-varying waveform applied from an electrode. I'm looking for advice or examples from anyone who has experience using a time-varying interpolated waveform as an electrode voltage in COMSOL.

Any guidance or references to similar models would be greatly appreciated.

Thank you in advance for your help!

Hi all,

I have a model that I've just posted about re AC losses, but I have a more major, pressing issue with the simulation that I cannot solve and wondering if anyone could give some advice. It is a superconducting cable that was downloaded from this source: https://htsmodelling.com/model-files/3d-comsol-model-for-corc-cables which simulates redistribution of current after an external heat pulse.

The issue I'm having is that, after disabling the heat source (I'm only concerned with the electromagnetic cable heating not external), there is no heating internally of the cable when I run the model. The external boundary is maintained at 77K and designed to keep the cable cool, so there should not be much internal temperature change, but even when I drastically increase the current input to way over critical temperature (when resistance should become very large) there is zero internal heating. The heat transfer node is not the issue, because in the original model with external heat, the 77K boundary condition cooled the cable back down over time with the correct heat flux.

I would like to know how much cooling power is required by the external boundary without a heat pulse but with only internal electromagnetic heating (which should be present even if very small as far as I'm aware), but no matter the parameters that I change I cannot get the cable to heat up.

For reference, the T-A formulation is used for current vector generation, along with Magnetic Field node with electrical conductivity defined from the material, Heat Transfer in Shells node that seems to be all OK, and a Multiphysics node for Electromagnetic Heating. Any other model information I'm happy to provide! The temperature variation within the superconducting tape is in the order of 10e-13 over the sim time, which would be purely calculation error.

Thank you massively in advance if anyone can help shed any light on why I'm not getting any internal cable heating. The top-level model builder is shown in the image attached.

Hi all,

First time posting here but with a quite urgent issue. I'm very new to COMSOL (and electrical engineering in general, I'm a mechanical engineer) and I'm doing a project simulating superconducting cable with high current flow. I am using a model that has already been made, changed some parameters and have a bunch of plots and results that were predefined from the download. For reference, this is the model https://htsmodelling.com/model-files/3d-comsol-model-for-corc-cables

I am wondering, could anyone shed some light on how to calculate or plot AC losses in my cable, and where these losses actually originate? I have current density outputs and various others, but no temperature rise in the cable (this is a separate issue I'm having) and can share more info if required.

I really appreciate any help and please do assume that my knowledge around this area is very limited! Thank you.

Mark

Hello, I want to modify a pressure value at a specified node and use it as the initial value for the stabilized Convection-Diffusion Equation. I can access the field data in Livelink for MATLAB through: pd = mpheval(model, 'p0', 'dataset', 'dset1');

I changed the pressure value as follows: p0_modi = pd;
p0_modi.d1(1,1) = 3;

Since I want to use the modified p0 ("p0_modi.d1"), I tried to set it again through: model.variable('var17').set('p0', p0_modi.d1);

However, due to the unacceptable data type for the .set function, it doesn't work. How can I use "p0_modi" for the p0 variable?

Dear community, I am running Optimization Module for the first time in order to maximize the current density in an Electromagnetic Study (ACDC module) of a coil detecting a defect over a metal cylinder; for now i am just changing the height of the coil ("altbob"). I have found message below, tried several changes in tolerance in Optimization Module but nothing changes. I appreciate any advice.

• Parameters: "altbob","4.3" The following feature has encountered a problem:
  • Feature: Stationary Solver 2 (sol1/s2) Failed to find a solution. The relative error (0.027) is greater than the relative tolerance. There was a warning message from the linear solver. Iterative refinement triggered. Returned solution is not converged. Not all parameter steps returned.

Thanks in advance, Thiago

Dear Forum:

Specifically, I used a two-dimensional symmetrical fluid structure coupling module. Due to the strong nonlinearity of the structure, I attempted to use resonant transient analysis for multiple periods.

I derived the sine displacement function twice from t and used it as the acceleration of the fixed direction base vibration. However, the calculated displacement of the base, expressed as solid.u_maseZ, shows a curve of increasing sine function plus linear function.

After searching, the help documentation for the "base excitation" module shows that its acceleration, velocity, and displacement are all relative to a fixed coordinate system. How to make the base displacement become a sine curve with only acceleration input? What is the relationship between the incentive application of "base excitation" and fixed boundary conditions?

I'm simulating a nonlinear phononic crystal using time dependent study. My goal is to get the transmission loss for the structure and compare it with frequency domain study.

So I defined my study following the Loudspeaker Driver — Transient Analysis defining study 1 as time dependent and applying the sin displacement for a certain frequency and then collecting the input and output displacement history for last cycle. Then I defined another study as time dependent + fft to convert my results into frequency domain defining ODE & DAEs. I can successfully do this. However, I'm not clear how can we calculate the transmission loss out of that since I'm getting a different peaks for various frequencies.

also if I apply a range of frequencies as my source frequency, how can I integrate all of these results into a unique general transmission loss spectrum.

Thanks

Hi, I am using the scattering field formulation for my model. I am using an x-polarized negative z-propagated plane wave as my background electric field. My structure is a box consisting of an air domain(top) and cancer domain(below). I used scattering boundary conditions. After the simulation, I realized that 1. emw.normE = abs(emw.Ex) 2. but abs(emw.Ez) is not = 0. if 1 is true, then how is 2 possible(since i am propagating the z direction anyways). or i am not plotting it correctly. I have attached my file. Thanks

Hello all,

I am modeling a heating device placed inside a room using conjugate heat transfer module with k-ε turbulence model.

The simplified geometry consists of three cubes (figure attached), the smallest cube represents the device, which is placed inside the room very close to two walls. The air inside the room is defined using the second cube, and the difference between the biggest cube and the second cube is defined as a brick wall. The heating device is at a constant temperature of 100 degC, and there is a constant air inlet (mass flow rate defined) from the top face of the device. To make it a stable system, an opening is made in the roof (as seen in the attached picture).

Mentioned below are the specified boundary conditions. Heat transfer in solids and fluids: - Air inside the room and the brick wall are initially at 25 degC - The device is initially at 100 degC - Constant temperature of the domain is defined using a pointwise constraint - The exterior surfaces of the brick walls are convecting with the external air at 25 degC - An "outflow" condition is specified at the exterior surface of the window

Turbulent flow: - Initial and reference pressure levels at 1 atm for the whole model - Weakly compressible flow with no gravity effects - Inlet at the top surface of the device (mass flow rate defined) - The surfaces of the device (other than the inlet) are defined as walls - The internal surfaces of the brick wall are defined as walls - An "outlet" condition is specified at the exterior surface of the window - Used a pressure point constraint set at 1 atm at one of the point on the external surface of the window

For the multiphysics coupling, viscous dissipation is enabled.

With the above set of boundary conditions, a stationary model is working well and is giving reasonable results. But the same model when run using a time-dependent study is giving several errors. It initially gave an error of inconsistent initial values, following which I used step function to ramp up the transition of the device's temperature, inlet. and convective coefficient. This time, I am getting an error that says

"Time-Dependent Solver 1 (sol3/t1) Repeated error test failures. May have reached a singularity. Time: 3.2249932802768407e-06 s. Last time step is not converged".

I have tried changing the solvers and mesh resolution too, but nothing seems to work. Would be great if anyone could help me with this.

Is it possible to use the penalty method to eliminate the pressure term from the momentum equation in COMSOL?

Dear fellows, I hope you are doing well.

Could you please guide how to define components for S-Wave and P-wave of Electric field incident at specific incident angle on 3D geometry in COMSOL 6.1.
It would be appreciated if some tutorial or example can be provided.

Thanks in anticipation.

While I am running an optimization study, I get the probe results for each iteration in a table. Is it possible to write these data, automatically after each iteration, to a file, while the study is ongoing?

Motivation: I want to check the intermediate results in batch mode without graphical interface.

When studying the calculations, there were errors in compiling the equations, and the differential algebraic equations were structurally inconsistent。 so how to deal with this question?

Hello, I was wondering what is the right way of using 1D extrusion/projection to a 3D surface. I understand projection decreases the number of dimensions.

What I am trying to do is to project temperature from a 1D pipe to a 3D surface, where x and y will be the pipe wall temperature as a function of z.

Since, the pipe is in non-isothermal conditions, I dont want to average my values. I tried linearizing it to 1D line on my surface, it didnt work.

I did a stationary study on comsol to get the displacement on the bottom flexure top face here was the set up 1. Upper and lower flexure with the beam 2. Anisotropic silicon [100] 3. Added mass at Top of the beam: 0.9 kg 4. Added mass at the top face bottom flexure: 50kg 5. Fixed constant at the top flexure bottom face 6. No-rotation constant at the bottom flexure top face 7. Boundary load: first test 0.5 N then second test 0.005 N both test at bottom flexure top face both in the x-direction 8. Gravity is on first test. Gravity is off on the second test 9. Study is static study 10. Use a probe at the bottom flexure top face to grab this displacement value.

i got the displacement on the bottom flexure top face and it 0.323 m is this correct?