ALE method and Internal Coupling Problem

[Yvan Fournier]

Hello,

if you are prescribing wall deformation with an explicit formula, you can probable define that directly with the GUI.

Otherwise, check Example 2 here: https://www.code-saturne.org/documentat ... _ale2.html

Regarding coupling the code with code_aster, as mentioned, I ma not 100% sure it can be done without minor changes in the code_aster side, but if you are familiar with code_aster user-defined functions, we could try. On the code_aster side, would you know how to define boundary conditions (forces) on a given set of boundary faces, based on a array of values (one value per face), and retrieve an array of displacement values at the vertices of those faces ? If this is the case, we could try the coupling (with an additional thing to check regarding MPI initialization that I can explain separately).

Best regards,

Yvan

[Shayor]

I don't understand completely what you mean but from whatever I understood I think you meant outputing the values of fluid forces from Code Saturne as an array ofcourse and reading them,run Code Aster with it,The script for such maybe the one given below(I got it from Code Aster forum,someone who chained these 2 and shared it).I also got a slide to do it using ASTK on Code Aster side.Attached it below.

[Yvan Fournier]

Hello,

Yes, that document is based on old versions of the codes, but you could still apply the same technique with more recent versions. The suggested solution should work, though using files to exchange data at each time step could slow down the computation a bit.

The solution I was mentioning would use a finer-grained coupling so as to use arrays exchanged by MPI instead of files, but a good approach would to to start with the simplest solution (files, as in the document you found), and possibly improve it with the elements I suggested as a second step, if necessary

Best regards,

Yvan

[Shayor]

Hello Yvan.I did not understand with each "tile step" means.Could you please clarify that?Thank you.

[Yvan Fournier]

Hello,

Sorry, this was a typo : read "time step".

I fixed it in the message above for clarity.

Best regards,

Yvan