syrthes coupling not updating

[filip2109]

Hello everyone,

sorry for bothering you with what is probably a trivial question but I really cannot find the answer to it.

I'm trying to replicate case6 from the tutorials using CS_4.1.1 and Syrthes 4.3 (latest available atm).
I'm having the problem that CS seems to get the initial boundary conditions from syrthes and then the two simulations seem to proceed independently: CS walls are constantly at the same temperature of Syrthes's initial conditions while the temperature of solid Syrthes domain keeps increasing. This goes on until the simulation crashes after ~300 iterations.

I certainly missing a setting somewhere but I cannot seem to find the problem. Any help would be very welcome.

Thank you very much in advance for all help,
Kind Regards

Filip

Ps: I'm attaching the CS case and the syrthes case, if any file could help please let me know. Thank you very much

[Yvan Fournier]

Hello,

Could you post all the other recommended troubleshooting info (except meshes, but especially logs) ?

Regards,

Yvan

[filip2109]

Hello Yvan,

thank you for your reply.
I'm attaching the compressed logs.
If I'm forgetting anything else please let me know, I'm still very new to CS and I'm really grateful for your help.
CS version 4.1.1
Syrthes version 4.3.0

Thank you very much,
Filip


(Edit: sorry I meant to click the upload file button but instead clicked submit, my apologies).

[Yvan Fournier]

Hello,

I just dowloaded your logs, and will look at them within a day or 2.

Best regards,

Yvan

[Yvan Fournier]

Hello,

I am not sure which tutorial you call "tutorial 63. In addition, your logs contain a dp_FCP.xml file, which seems not to be used, but should not be there (it is confusing, as this is for coal combustion only).

Your computation is diverging, but probably not due to the temperature: temperature values stay reasonable, but from the start, the CFL values seem large, and quickly become worst.

I would recommend dividing the time step at least by 5, the checking the evolution of the CFL (CouranftNB) in the "listing" file.

Temperature values in Code_Saturne are evolving, so coupling seems OK. Similarly, in Syrthes, the minimum is decreasing at first, with, the maximum increasing (then both increase), so I assume you have both coupling with a cooler fluid and a source term (but again, I am not sure which is tutorial 6).

Did you visualize the Syrthes coupling info on the associated submesh (use "extract blocks" unde ParaView, or check the appropriate Syrthes part under EnSight to check the wall temperature provided by Syrthes and the heat flux as viewed on the Code_Saturne size; Your output frequency seems to be every 25 time steps, so that should be quite enough to visualize how both evolve.

Regards,

Yvan

[filip2109]

Good morning Yvan,

thank you very much for looking into this.
I've tried to replicate http://code-saturne.org/cms/sites/defau ... -disks.pdf which I've seen is called "Case 6" or Three 2D disks. I'm sorry for creating this confusion.

As I didn't know where to find the related mesh, I've created a 3D mesh for both the disks and the fluid, in Salome trying to replicate the shown geometry (Netgen mesh, with viscous bl on the fluid part).

You are absolutely right: in creating the geometry in salome, I forgot salome works with m and not mm, so looking at the syrthers paraview images I've seen that the heat does not get at the interface by the time the simulation crashes, thus the interface temperature does not change.
I'm really ashamed I didn't see this before and that I wasted your time.

Please accept my apologies for it.

About the other cause of divergence, I will try modify it as you suggest after I resized the whole geometry to bring it in mm.

Thank you very much for you help and again I'm very sorry for wasting your time.
Kind regards,
Filip

[Yvan Fournier]

Hello,

No problem. Some tutorial meshes are on the website, others are missing, so we need to check which ones are missing, and make them available (3disks is probably not the only missing one).

Best regards,

Yvan