Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Free convection and conduction heat transfer - problem did not converge

Please login with a confirmed email address before reporting spam

Hello,
I am trying to build a model to simulate the heat transfer around a vertical pipe. Heat transfer will be due to two mechanisms, namely thermal conduction in the duct and free convection.

I have built a model with COMSOL Multiphysics 4.0a (CFD Module -> Non-isothermal Flow). I have attached the mph file.

My model consists of two subdomains, the first represents a duct with a cooling rectangular blade (structural steel), and the outside domain (air). As I would study both heat transfer modes conduction and free convection, I used the Non-isothermal flow module.
I fixed the temperature on the interior face of the duct. The objective of this study is to know the temperature variations in the domain that's why I used time-dependent solver.
My problem occurs when I solving the model. I obtain a error message:

Nonlinear solver did not converge.
Time : 2.86102294921875e-009
Last time step is not converged.

I'm beginner in Comsol's world and have some problem to understand all solver parameters.
So i tried to refine mesh and reduce the time step, but I encountered always the same problem.
Is anyone can help me and know what is wrong?

Best regards
Anthony


10 Replies Last Post Dec 16, 2010, 9:20 a.m. EST

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 15, 2010, 2:58 p.m. EST
Hi Anthony,

I loaded you model and noticed that it was missing at least one boundary. There was only one temperature boundary but no other where heat should come in or go out. Should it be like this?

Best regards

Tero
Hi Anthony, I loaded you model and noticed that it was missing at least one boundary. There was only one temperature boundary but no other where heat should come in or go out. Should it be like this? Best regards Tero

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 16, 2010, 3:22 a.m. EST
Thanks for your reply, effectively, I would like to study the cooling kinetic of the fluid in a closed and isolated system. That's why I think I can't impose another temperature boundary. So I don't know how to do differently
I think my problem is the solver choice, but as i said in my first message, i'm new user of Comsol and don't understand the solver working.
Best regards
Anthony
Thanks for your reply, effectively, I would like to study the cooling kinetic of the fluid in a closed and isolated system. That's why I think I can't impose another temperature boundary. So I don't know how to do differently I think my problem is the solver choice, but as i said in my first message, i'm new user of Comsol and don't understand the solver working. Best regards Anthony

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 16, 2010, 4:50 a.m. EST
So you want to study heat transfer and free convection with buoyancy force, right? I don't see any other way to do it that by temperature difference. If you don't have temperature difference, no buoyancy force is created.

Solver seems to work fine if at least one boundary is defined to temperature, open boundary or similar. I don't know what kind system we are dealing here so it is difficult me to say what is right one.

Best regards

Tero
So you want to study heat transfer and free convection with buoyancy force, right? I don't see any other way to do it that by temperature difference. If you don't have temperature difference, no buoyancy force is created. Solver seems to work fine if at least one boundary is defined to temperature, open boundary or similar. I don't know what kind system we are dealing here so it is difficult me to say what is right one. Best regards Tero

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 16, 2010, 5:43 a.m. EST
I understand that a gradient temperature is needed to create a buoyancy force. That's why i work in time dependent solver. I try to solve the problem with an outflow condition on the top of my geometry, the solver find the same issue.
Maybe i don't enough describe the system. I have a fluid domain. A cooling device go cross the domain (duct with rectangular blades). Trough this device flow a cold fluid, I would like to know the necessary time to cooling the entire domain at a fixed temperature. That's why i used an time dependent formulation, and why i don't fixed a temperature in the fluid domain.
If you have an other idea to solve my problem, they are welcome.
Best regards
Anthony
I understand that a gradient temperature is needed to create a buoyancy force. That's why i work in time dependent solver. I try to solve the problem with an outflow condition on the top of my geometry, the solver find the same issue. Maybe i don't enough describe the system. I have a fluid domain. A cooling device go cross the domain (duct with rectangular blades). Trough this device flow a cold fluid, I would like to know the necessary time to cooling the entire domain at a fixed temperature. That's why i used an time dependent formulation, and why i don't fixed a temperature in the fluid domain. If you have an other idea to solve my problem, they are welcome. Best regards Anthony

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 16, 2010, 7:37 a.m. EST
Hi Anthony,

I managed to solve problem. I made it with Comsol 4.1 and it is too large send here. I can send reset version without result if you want it? Changes what I made are following:

- selected domain to volume force (there was force but no active domain)
- top side wall => outlet
- side walls => symmetry

With these calculation was working. I hope this helps.

Best regards

Tero
Hi Anthony, I managed to solve problem. I made it with Comsol 4.1 and it is too large send here. I can send reset version without result if you want it? Changes what I made are following: - selected domain to volume force (there was force but no active domain) - top side wall => outlet - side walls => symmetry With these calculation was working. I hope this helps. Best regards Tero

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 16, 2010, 8:23 a.m. EST
Thanks for your help to solve my problem. If you would like to send me a reset version it will be magnificent.
I will try with your recommendations.
Again thank you.
Best regards
Anthony
Thanks for your help to solve my problem. If you would like to send me a reset version it will be magnificent. I will try with your recommendations. Again thank you. Best regards Anthony

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 16, 2010, 8:42 a.m. EST
Hi Anthony,

Here is the model. Side walls are here again walls (symmetry boundary had no influence).

Best regards

Tero
Hi Anthony, Here is the model. Side walls are here again walls (symmetry boundary had no influence). Best regards Tero


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 16, 2010, 8:46 a.m. EST
Sorry but i'm working with the 4.0a version and I haven't received yet the last version.
Is it possible you re-send me an compatible version?
Best regards
Anthony
Sorry but i'm working with the 4.0a version and I haven't received yet the last version. Is it possible you re-send me an compatible version? Best regards Anthony

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 16, 2010, 9:11 a.m. EST
Sorry, I cannot get you 4.0a compatible version since it is not possible to save it in earlier version (or if there is, I don't how to do it).

Best regards

Tero
Sorry, I cannot get you 4.0a compatible version since it is not possible to save it in earlier version (or if there is, I don't how to do it). Best regards Tero

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 16, 2010, 9:20 a.m. EST
It doesn't matter, I followed your advices, and it seem ok, the calcul is running.
Now I can adapt this method to the entire geometry.
Best regards
Anthony
It doesn't matter, I followed your advices, and it seem ok, the calcul is running. Now I can adapt this method to the entire geometry. Best regards Anthony

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.