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.

Localized divergence in heat transfer problem?

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Hi,
Hopefully someone can help me with a problem I'm having with localized divergence in a time dependent heat transfer problem. I've attached a picture of one of the solution steps which show the behavior I'm talking about.

The geometry consists of two solid pellets separated by a gas filled gap. Solid mechanics is used in the two pellets to calculate deformation due to thermal expansion and heat transfer is active in all of the domains. The two pellets will eventually expand and come into contact with each other (save a small gap to prevent topology issues for the domain in the middle). Heating occurs only in the solid region, not in the gas.

The picture shows one of the temperature profile from one of the solver steps. All materials were started at the same temperature and the solid region was heated volumetrically. I've included the mesh in the output, although I do not believe this to be a mesh dependent problem as I have encountered the same problem with a much finer mesh.

Any ideas?



4 Replies Last Post 28.06.2010, 16:06 GMT-4
Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 25.06.2010, 16:05 GMT-4
Opps. I realized I didn't attach the pictures I intended. Here they are.
Opps. I realized I didn't attach the pictures I intended. Here they are.


Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26.06.2010, 10:17 GMT-4
Hi

It looks indeed funny.

But what are the relative scale of these values, I have "burned" myself sevral times loking at beautifull rainbows of funny effects, then to see that the relative scale was down in the 1E-6 or 1E-9 and could easily be expalined by numerical instabilities due to the binary representation of numbers and the limited resolution from our numerical CPU's.

Have fun Comsoling
Ivar
Hi It looks indeed funny. But what are the relative scale of these values, I have "burned" myself sevral times loking at beautifull rainbows of funny effects, then to see that the relative scale was down in the 1E-6 or 1E-9 and could easily be expalined by numerical instabilities due to the binary representation of numbers and the limited resolution from our numerical CPU's. Have fun Comsoling Ivar

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 27.06.2010, 13:14 GMT-4
Thanks for the reply.

The colour range represents a difference of about 3 k, from about 499 to 502. I've attached a simplified version of the model I'm working with if it helps.
Thanks for the reply. The colour range represents a difference of about 3 k, from about 499 to 502. I've attached a simplified version of the model I'm working with if it helps.


Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 28.06.2010, 16:06 GMT-4
I think I have made some progress on determining the source of this issue. I believe the issue may be caused by bad continuity mapping caused by incorrect deformation coupling of the model.

The problem involves contact, therefore I was forced to work with an assembly. (Alternatively if contact pairs worked with unions I would also be ok however they do not. ) In solid mechanics I cannot simultaneously have a contact pair and identity pair on the same boundary. Therefore I must find some other way of coupling the deformation on one side of the boundary with the other.

I believe that I need to use an extrusion coupling to achieve this however I don't know how to implement this in C4. Can anyone provide a model where extrusion coupling has been used in C4?

I think I have made some progress on determining the source of this issue. I believe the issue may be caused by bad continuity mapping caused by incorrect deformation coupling of the model. The problem involves contact, therefore I was forced to work with an assembly. (Alternatively if contact pairs worked with unions I would also be ok however they do not. ) In solid mechanics I cannot simultaneously have a contact pair and identity pair on the same boundary. Therefore I must find some other way of coupling the deformation on one side of the boundary with the other. I believe that I need to use an extrusion coupling to achieve this however I don't know how to implement this in C4. Can anyone provide a model where extrusion coupling has been used in C4?

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.