Looking for help with convergence of conjugate heat transfer case

[pkciuk]

Hello,
I am looking for help with conjugate heat transfer case.

SimScale

My case diverges although I have already meshed with nearly fine settings, with refinements near sharp objects. Also Settings are pretty straight forward as I only added natural convection inlet/outlet to air box and heat generation on one of the elements. All interfaces are predefined by simulation as “Coupled inteface thermal”. Lack of divergence might be also connected with solver settings and relaxation factors.
I will be glad for any guidance.

[jousefm]

Tagging the @cfd_squad here to give you some guidance on your project.

Best,

Jousef

[Ricardopg]

Hi, @pkciuk,

I had a look at the mesh & failed runs. Briefly, the runs are failing due to mesh quality (find some useful notes on mesh quality here).

Essentially you should follow the steps in that article to find where the bad cells are and then try to locally improve the quality.

Currently this is what we have for your mesh:

These 3 metrics are very critical to obtain a successful simulation. Unfortunately, there’s no solution that works for all cases here. Sometimes it may take some CAD changes, sometimes you can solve it by adding refinement.

Please inspect the mesh using the mesh quality feature (section 2 of the article) and spot where the bad cells are. They will most likely be located around these regions:

Zooming-up on the top-most circle:

The two faces approach in an asymptotic fashion. The small cells that capture this sharp corner tend to have a very low quality. These regions will definitely take some tweaks (most likely CAD-related, for your specific case) before you can get a good enough mesh for a simulation.

Cheers