Questions about CS & Syrthes coupling

Questions and remarks about code_saturne usage
Forum rules
Please read the forum usage recommendations before posting.
Post Reply
Emonet_L
Posts: 7
Joined: Fri Mar 01, 2019 10:58 am

Questions about CS & Syrthes coupling

Post by Emonet_L »

Hi,

I am new to these software, I have to work on an autoclave's project so I have chosen to simulate the device curing a composite slab. Is it possible to have some precisions about the CS/Syrthes coupling?

- First, I wonder if it is a strong coupling? I mean, are CS and Syrthes exchanging informations together? Is it just one of them giving infos to the other, and is this case which one?

- Then, is it possible to simulate a forced convection steady flow around the composite but an unsteady conduction inside?

- I saw on EDF's website that Syrthes 4.3 is compatible with CS 4.x but is it compatible with 5.0 version too?

- My last question concerns the GUI, is it possible to use only this GUI to achieve the coupling (saw something like that on CS 5.0.0 practicle user's guide part 7.5.2) or do I need to use commands as the topic "Guide to couple Saturne & SYRTHES" ( viewtopic.php?t=1465#p7724 ) explains?

I am kind of a beginner, I have created a VM using Ubuntu 18.10 but it is the first time I will use Linux OS... I know I have to learn a lot, mainly the "technical terms" used to describe the installation or coupling process.

Thanks you for the future answers.

Regards,


Luc
Yvan Fournier
Posts: 4070
Joined: Mon Feb 20, 2012 3:25 pm

Re: Questions about CS & Syrthes coupling

Post by Yvan Fournier »

Hello,

Yes, Syrthes 4.3 is compatible with Code_Saturne versions 4.x, 5.x, and 6.x (upcoming).

The coupling is a "true" coupling, as both codes exchange data, but is explicit in time, simply coupling boundary layers. For a stronger coupling, a new "internal" coupling feature is available in Code_Saturne. It is not yet usage with the graphical interface, but should be usable with that interface within one or 2 months (before v6.0 release around May or June).

A mix of steady flow and unsteady thermal simulation will be considered an unsteady computation, but if you are sure the temperature does not influence the flow (not buoyancy/gravity effects, quasi-constant fluid preperties in the temperature range), you can run an initial fluid computation, and do the thermal computation as a restart using the "frozen flow field" option of Code_Saturne (which wil be much faster as only the temperature field will be solved).

The GUI is mostly usable for the coupled calculation, though one step requires manual launch. Using the Salome_CFD distribution (https://www.code-saturne.org/cms/salome-cfd), you should be able to set up the coupling purely in graphical mode. This also avoids installing Syrthes from sources, which is a pain (it also avoids installing Code_Saturne from sources). We currently have builds for Debian 8 or Ubuntu 16.04. We need to build for Ubuntu 18.04 but this has not been done yet to my knowledge. So if you are using a virtual machine anyways, using a Debian 8 or Ubuntu 16.04 VM might be a better option, since the binaries should work on that.

Best regards,

Yvan
Emonet_L
Posts: 7
Joined: Fri Mar 01, 2019 10:58 am

Re: Questions about CS & Syrthes coupling

Post by Emonet_L »

Hi Yvan,

Thank you very much for your quick answer.

Do you have some documentation about this new "internal" coupling please? Maybe it is already in Code_Saturne documentation? I will also watch out news about usage with the graphical interface and Code_Saturne's upcoming version.

For now, I will try to use Salome_CFD distribution as you told me so. It might be sufficient for my project.

Best regards,

Luc
Yvan Fournier
Posts: 4070
Joined: Mon Feb 20, 2012 3:25 pm

Re: Questions about CS & Syrthes coupling

Post by Yvan Fournier »

Hello,

The documentation for the new feature is limited, though a few others on this forum have used it (help welcome).

Check the cs_user_internal_coupling function in src/user_examples/cs_user_parameters-base.c (the examples do not seem to be linked to Doxygen yet). The volume-based approach is usually simpler.

You also need to adjust the physical properties in each region (not zone-based yet in the GUI, so you need to Fortran routines in cs_user_physical_properties.f90 for this).

To avoid fluid flow in the solid zone, zero-velocity on all boundaries of that zone should be enough (though we need to improve the robustness of that part; using a high head loss in this zone may also help, tough is not required).

That's about it.

Best regards,

Yvan
Post Reply