Latest Discussions - COMSOL Forums

Freezing through convergence Plots

Wed, 23 May 2018 12:35:24 +0000

According to

http://cn.comsol.com/forum/thread/105901/comsol-5-2-gets-stuck-after-some-time?last=2017-04-26T18:39:50Z

Comsol freezes occasionally when plots are created while solving, including the convergence plots. For my model (SST Flow, around 1,500,000 elements, 128 GB RAM), it happend three times in a row. After disabling the plots, it works pretty good.

Is there a possibility to see the convergence plots after the simulation is finished? Or see at least the Errors? They are not even shown in the progress when the convergence plots are disabled.

It´s pretty important for me, because I use two different rampings (viscosity and and a wire gauze), and i often only find a soluation when i use the interation number as a criteria.

Please don´t suggest to upgrade the version to 5.3, I use Comsol for my Master Thesis, and the Prof is not willing to pay for the new version...

how to plot circularly polarized light

Wed, 23 May 2018 09:04:09 +0000

I recently simulate phase metasurface : when a** left** circularly polarized light reflect on this metasurface, the reflected light would be a ** right** circularly polarized light. Here are some question

1) .How to describe left and right circularly polarized light the like "ewfd.Ex" and "ewfd.Ey" in COMSOL？ (I want to know the reflected angle/reflection direction of the light beam.)

2)In "far field" of comsol ,we only get ewfd.normEfar. Could I plot right circularly polarized component in far field?

Pressure drop calculation

Wed, 23 May 2018 08:04:21 +0000

How to calculate the change in pressure in a closed cavity due to volume change?

Error in elPlastic element.

Wed, 23 May 2018 03:17:23 +0000

I ran a geomechanical model and I got this Error "Error in elPlastic element. - Detail: Lambda non positive for plastic point" , I decreased the time step and tighten the convergence tolerances acccording to an apply in this fourm. However, it didn't work.

Could anyone tell me why this error happens. I used plasticity model.

Basic Resistance measurement at EC with probes

Tue, 22 May 2018 21:27:49 +0000

Hi there, Everybody keeps asking this question but still no any simple answer here or anywhere alse, How such powerful software can't measure resistance between two points on the wire like simple multi meter can do. If this is EC simulation stationary or time dependent. Can somebody show how to set up resistance measurement between two points in few clicks, without formulas or manual calculation, just want to pick up two surfaces and click "measure resistance". Thanks!

Discretization tool not available

Tue, 22 May 2018 14:57:37 +0000

Hi, I'm attempting to switch the discretization in a solid mechanics beam problem from linear to quadratic lagrange elements (or back again.) From what I understand most people can see the discretization option under the solid mechanics dropdown menu, but I do not. I have tried restarting the geometry etc from scratch several times, and I could not find a similar issue in previous forum articles. Please advise!!

I have linked screenshots of my COMSOL console, vs. a friends' who can see the discretization option.

Energy loss in symmetric heated turbulent flow

Tue, 22 May 2018 11:56:58 +0000

Hello,

I created a simplified model which just simulates the flow of heated air through a pipe, heating up a small piece of the pipe-wall.

It´s completley symmetric, so I decided to use the symmetrie line in the center. For the further application of the Model, I cant reduce the problem to a 2D-Axiasymmetric model.

The transferred heat from the gas to the pipe wall seemed to be way to low. My Calculations showed that the energy which is lost by the hot air is several times bigger than the energy which heated up the pipe.

Checking the heat fluxes of the system revealed that energy goes through the symmetry and "disappears".

Removing the cut-line and simulating with the whole cross-section gives a correct result ( less than 3% difference between theory and simulation).

Are there any wrong boundary conditions? Why do those heat-fluxes out of the System appear? Or is it a problem in Comsol?

About the Simulation:

Run Study 2 (solves the flow)

Enable the 3 Multiphysics (Flow does not converge while those physics are enabled, though they are deactivated in the solver for this solution is not activated)

Run Study 4 (solves the temperature for 20 sec, pipe should be heated to around 22 degC)

Changes done in full cylinder: - Disabling workplane 1, Partition Domain 1, Delete Entities 1 - Adding the new Boundaries to the Boundary Layer Mesh

Same procedure to run like the first model. Now the pipe heats up to nearly 27 degC.

Thanks for any help!

If you have any further remarks or suggestions about the model, feel free to answer :-)

Coupling between rf and electric currents

Tue, 22 May 2018 11:30:22 +0000

I would like to set a fixed potential in the rf module. If I use the lumped port I can set a potential, but this only applies if wave exication is enabled. When I activate this, my model becomes very complicated and the solver takes a very long time. To avoid this, I wanted to define the potential with the electrica currents physic. The problem is that the defined potential in electrican current module does not exist in the rf module. Because the two modules are not coupled. The question is can the two modules be coupled? And if so, could someone show me how? In the appendix is a model example.

I would be very grateful for any idea!!!

Edge rotation for shell with prestressed eigenfrequency analysis always gives 0

Tue, 22 May 2018 10:01:30 +0000

Hello! I'm having troubles trying to evaluate the rotation of one edge of a shell element around the axis parallel to it (in my case, the z axis). I'm running a prestressed eigenfrequency study, i.e., a two-step study with stationary and eigenfrequency steps. Trying to plot shell.thZ (or any other shell.th) gives a clean 0.0 over the whole shell surface.

Funny thing is, if I use a different study with only an eigenfrequency analysis (no stationary one), I get reasonable values for the same variable .

Am I forgetting or overlooking something obvious? Thank you in advance to anyone who will shed a light on this!

Far-field calculation of a duct outlet

Tue, 22 May 2018 08:30:57 +0000

Hello,

I'm working on a model of a duct that has an outlet into a room, which is modeled as a semi-infinite space. A diagram of the 3D model is attached. The black part is the duct, the grey part is the room, which is surrounded by a PML (blue area). Both domains are filled with air. At the far end of the duct, I have added an incident plane wave.

I want to know the far field pressure in the room. However, I haven't been able to get good results from the far-field calculation. I just want to know the far field pressure in the room, but in my results all I see is a peak in the direction where the source is located.

Obviously I'm making a mistake in setting up my far-field calculation. But I don't have the faintest idea what I'm doing wrong. I'm hoping you can help me out. I've attached my COMSOL model. In this model the domain modeling the room has a larger radius than the duct, which is slightly different than in the attached diagram.

I hope someone can help me. Thanks in advance.

Kind regards,

Thijs Hekman

Convergence criteria of the nonlinear solver are met but not all linear solver conditions are fulfilled.

Mon, 21 May 2018 15:53:39 +0000

I was running Semiconductor module. The solver is fully coupled (constant Newton).

The simulation shows the following warning:

"Convergence criteria of the nonlinear solver are met but not all linear solver conditions are fulfilled."

Anyone has the same issue? Please help.

Thanks

Magnetic field and Thermal study Combined

Mon, 21 May 2018 14:46:54 +0000

Hi,

I am currently simulating a model (magentic field study) to evaluate the heat source distribution surrounding an copper inductor cavity and wanted to consider these heat sources asinput for my thermal model .

There is no problem in magnetic field study , can see the surface loss density plot (mf.Qsrh) but when i assign this mf.Qsrh value in thermal model as assigned heat flux I am getting some errors while comuting thermal study ,

I can see same intial temperature at all time steps
If not the above error (1) , getting error as undefined variable on boundaries (boundaries where heat flux is assigned)

I am attaching my model here , kindly let me know what is the nature of this error.

P.S - I have also included a PID study in this , however dont feel PID is reason for this error . Correct me if I am wrong

Supercapacitor

Mon, 21 May 2018 04:12:18 +0000

Hey.I am trying to model supercapacitor but get some difficulties.Could someone who have already succeed in it kindly give me some guidance or share some details with me.I would very appreciate it.

Conjugate Heat Transfer for Turbulent Flow problem

Mon, 21 May 2018 03:47:04 +0000

Hello,

I have seen that you have helped a lot of people many times about different types of problems in COMSOL, that's why I'm contacting you.

I study the aerothermal effects of a CubeSat which comes back to earth. First, I made lot of assumptions to simplify my model.

I began to learn COMSOL few weeks ago. First I made a model with laminar flow, I imposed a velocity for the inlet and open boundaries for the outlet. It worked.

Them, I add conjugate heat transfert for laminar and it worked too.

But when I study the turbulent flow I don't understand why it doesn't work ...

I have the message below :

Failed to find a solution. Segregated Step 2 Singular matrix.

There are 409 void equations (empty rows in matrix) for the variable comp1.T. at coordinates: (0.221669,0.25,0.551122), (0.227749,0.25,0.542537), (0.225979,0.241028,0.546981), (0.220487,0.240859,0.555651), (0.225796,0.258812,0.547148), ... There are 409 void equations (empty rows in matrix) for the variable comp1.nitf1.TWall. at coordinates: (0.233762,0.240739,0.538702), (0.234585,0.25,0.535845), (0.227749,0.25,0.542537), (0.24I have seen that you have helped a lot of people many times about different types of problems in COMSOL, that's why I'm contacting you.

I study the aerothermal effects of a CubeSat which comes back to earth. First, I made lot of assumptions to simplify my model.

I began to learn COMSOL few weeks ago. First I made a model with laminar flow, I imposed a velocity for the inlet and open boundaries for the outlet. It worked.

Them, I add conjugate heat transfert for laminar and it worked too.

But when I study the turbulent flow I don't understand why it doesn't work ...

I have the message below :

Failed to find a solution. Segregated Step 2 Singular matrix.

There are 409 void equations (empty rows in matrix) for the variable comp1.T. at coordinates: (0.221669,0.25,0.551122), (0.227749,0.25,0.542537), (0.225979,0.241028,0.546981), (0.220487,0.240859,0.555651), (0.225796,0.258812,0.547148), ... There are 409 void equations (empty rows in matrix) for the variable comp1.nitf1.TWall. at coordinates: (0.233762,0.240739,0.538702), (0.234585,0.25,0.535845), (0.227749,0.25,0.542537), (0.242123,0.25,0.531499), (0.25,0.25,0.53), ... In Segregated Step 2: Returned solution is not converged.

Please can you help me.

You will find my model attached.2123,0.25,0.531499), (0.25,0.25,0.53), ... In Segregated Step 2: Returned solution is not converged.

Please can you help me.

You will find my model attached.

temperature gradient boundary condition

Sat, 19 May 2018 14:21:52 +0000

I am trying to implement a boundary condition where the temperature gradient in the normal direction to the wall is -1. how do i do it. i am using heat transfer in fluid physics and laminar flow physics. and in a wall i have to set temperature gradient in the normal direction to the wall to -1. is there any way i can do that?

Sound transmission loss and impedence

Sat, 19 May 2018 09:17:28 +0000

Hello everyone! How to simulate for the sound transmission loss and impedence for the thick rectangular plate. Please help me out with step by step procedure to simulate. And matlab program for it. Thank you

Galvanic Cell with same reaction for the anode and for the cathode

Fri, 18 May 2018 20:10:29 +0000

How to model a galvanic cell that has the same reaction for the anode and for the cathode? We have a Galvanic Cell with only two species Ox (oxidized species) and Red (reduced specie) and these species react as follow:

*cathode: Ox + e = Red E= 0 V
* anode: Red = Ox + e E= -90mV

The different potential could be due to the temperature difference between the cathode and the anode. If the initial concentration for Ox and Red are the same and regardless the fast kinetic for the anode (anode has a major temperature that cathode). It is supposed that these concentrations are invariant with the time, because one specie is consumed in one electrode and is produced in the other electrode. I am trying to solve this, but the system do not reach the steady state. The species change their concentration with the time. Does anyone know why? Thank you, for your time!

Resuls and Plots

Fri, 18 May 2018 20:10:01 +0000

Hello! every one. I have some troubles to Plot and reproduce my results, because each Plot appear one over the other and I can´t see properly the curves and 3D plots . I would like to know how can I fix this, please.

Nonlinear solver did not converge.

Fri, 18 May 2018 17:56:32 +0000

I'm trying to use the Particle Tracing for Fluid Flow. However, I keep getting an error saying the nonlinear solver did not converge, last time step did not converge. I've looked to make sure all of the boundary and initial conditions are defined but am still gettting the error for time dependent studies. What parameters need to be altered so the solution converges?

For another part of the simulation, I need to mix two substances together. Each substance would come from a different inlet, and the sample would leave as a mixture. What would be the best physics module to use?

Thanks!

Heat Transfer Boundary Condition

Fri, 18 May 2018 17:20:41 +0000

Hi, I'm trying to model a heat transfer physics with bottom boundary condtion: dT/dz = constant independent on depth (when the depth is more than 5m)

how can i do this in comsol? I'm slightly new to comsol

Thank you