code_saturne User's Forum

Thank you Antech for the tutorial...
This is what I can produce in the middle of the outlet pipe

Ccaccia73
Thanks, can you make the same picture for Saturne v2 results? IMHO, it's a good idea to set uniform scale for vectors (AFAIR, ParaView can do this).

The problem is that I don't now your case (what is physically in the domain, what are BCs, what flow patterns/features are typical for this device). I understand that the details are confidential, but only you can treat results right. I can only guess some things. I cannot do even a basic analysis of your case, but for my work tasks I always perform at least visual vector field analysis to check wether the results are "physical" and expected qualitatively.

Here is what I get with saturne 2:
streamlines seem better, as posted before.

Ccaccia73
Thanks. Vectors look similar to Saturne-4 so they doesn't explain why streamlines are strange in version 4 results... In your place I would prefere to analyze the flow details with vectors in various slices, and, maybe, with streamlines from different cross-sections (near the problematic area). It's also often usable to set constant vector length (don't remember if ParaView can do this). I can't say more definitely because I don't know you case specifics.

[Sorry for language issues, english is not my native]

[I have holydays for month so I will not be online until 3'th Aug]

Antech
have nice holidays, hope to hear from you when you're back.

Yvan
I've talked here and I can publish the geometry, mesh and or results of this particular case. Do you believe that it can be interesting? Should I start a new thread?

Hello,

Regarding the result, I'll let you choose whether you publish results or not. In any case, if you still have unexplained differences between 2.0 and 4.0 results, I'll try to take a look if you can publish the mesh and setup, though I'll do that as a "background task", so I probably won't have results before a few weeks.

I'll also check with colleagues whether we should reopen an "examples" section in this forum, as it would be a good place to showcase "working" setups.

Regards,

Yvan

Ccaccia73
Hello.
Have you solved your problem with different results?
I found some inconsistency in Saturne 4.0.0 results in pressure drop across a tube bundle, but there is a "bug" in one of two compared cases because of non-orthogonal mesh at the pressure outlet (it was the second time I've encountered this "bug" and now I'm sure that it's due to incorrect mesh). Therefore the difference in my case is likely to arise from this incorrect mesh issue rather than from Saturne solver.