ALE method and Internal Coupling Problem

Questions and remarks about code_saturne usage
Forum rules
Please read the forum usage recommendations before posting.
Shayor
Posts: 24
Joined: Tue Jan 05, 2021 3:40 pm

ALE method and Internal Coupling Problem

Post by Shayor »

Hello,
I am doing a simulation of blood flow in artery and wanted to use the ALE(Deformable mesh) and Internal Coupling options for finding the deformation of flexible wall of artery.However the calculation is aborted while running ALE and says"The absolute value of the discriminant of the displacement matrix is 0."......Next line saying hence it is invertible.
Note:I am using only the fluid volume for the mesh.No solid part is included.Is that the reason?I mean do I need a separate mesh for solid part in contact with the mesh part for fluid domain only.
Please help.As an academic work I am quite in a dilemma and short on time(6 weeks or so :cry: )
Luciano Garelli
Posts: 280
Joined: Fri Dec 04, 2015 1:42 pm

Re: ALE method and Internal Coupling Problem

Post by Luciano Garelli »

Hello,

Do you need to solve a fluid-structure interaction problem? or Will you impose de mesh deformation?

The Internal Coupling option does not solve the elasticity equations.

Also, you can post the log file in order to help you with the error. Please check the Code_Saturne usage forum guidelines post.

Regards,
Luciano
daniele
Posts: 148
Joined: Wed Feb 01, 2017 11:42 am

Re: ALE method and Internal Coupling Problem

Post by daniele »

Hello,

As Luciano said, I think you have misunderstood the internal coupling capabilities: it allows to solve the motion of a rigid body under the fluid forces, it is not a CFD-FEM coupling tool.
Therefore, no solid mesh can be included.

Please provide more details about your case in order to discuss.

Best regards,

Daniele
Shayor
Posts: 24
Joined: Tue Jan 05, 2021 3:40 pm

Re: ALE method and Internal Coupling Problem

Post by Shayor »

Luciano Garelli wrote: Mon Nov 29, 2021 6:34 pm Hello,

Do you need to solve a fluid-structure interaction problem? or Will you impose de mesh deformation?

The Internal Coupling option does not solve the elasticity equations.

Also, you can post the log file in order to help you with the error. Please check the Code_Saturne usage forum guidelines post.

Regards,
Luciano
Yes I want to solve a FSI problem.It's a problem where I would want to find the deformations of the artery wall(much like any flexible pipe flow problem)and the effect of the flexible wall on the pressure,velocity,etc.I have created only the geometry for the fluid domain only.So is that not possible using the internal coupling and ALE options(deformable mesh option)?
Shayor
Posts: 24
Joined: Tue Jan 05, 2021 3:40 pm

Re: ALE method and Internal Coupling Problem

Post by Shayor »

daniele wrote: Mon Nov 29, 2021 9:33 pm Hello,

As Luciano said, I think you have misunderstood the internal coupling capabilities: it allows to solve the motion of a rigid body under the fluid forces, it is not a CFD-FEM coupling tool.
Therefore, no solid mesh can be included.

Please provide more details about your case in order to discuss.

Best regards,

Daniele
So will it not give me deformations of the wall at points for the case as mentioned in Luciano Sir's comments reply?If not how can I do it then?
Shayor
Posts: 24
Joined: Tue Jan 05, 2021 3:40 pm

Re: ALE method and Internal Coupling Problem

Post by Shayor »

Sorry.Forgot to add the log file.Here are the log files and setup file.
Attachments
preprocessor.log
(5.39 KiB) Downloaded 85 times
setup.xml
(11 KiB) Downloaded 87 times
setup.log
(31.17 KiB) Downloaded 86 times
Luciano Garelli
Posts: 280
Joined: Fri Dec 04, 2015 1:42 pm

Re: ALE method and Internal Coupling Problem

Post by Luciano Garelli »

Hello,

In order to solve a FSI simulation, you have to couple Code Saturne(fluid-mesh movement) and Code Aster(Solid). This coupling (by the moment) is not straightforward.

Regards,
Luciano
Shayor
Posts: 24
Joined: Tue Jan 05, 2021 3:40 pm

Re: ALE method and Internal Coupling Problem

Post by Shayor »

Could you tell me the steps please?Any tutorial or guide to carry it out.Also is it too hard?I mean would it require extensive knowledge of Code Aster?
Yvan Fournier
Posts: 4070
Joined: Mon Feb 20, 2012 3:25 pm

Re: ALE method and Internal Coupling Problem

Post by Yvan Fournier »

Hello,

Worse than that, thou coupling with code_aster from salome_meca 2017 and alder with code_saturne 6.0 is possible but not easy to set up (i.e. it requires a tricky installation).

With salome_meca 2019, this was broken on the code_aster side. An I never tried with the intermediate salome_meca 2018.
With code_saturne 7.0 and current code_aster versions, the coupling method has been changed to use pure MPI. The ground-work is done on the code_saturne side, but work remains on the code_aster side. It might be possible to get things working for an advanced user of code_aster and code_aster user python scripts (I could help on the code_saturne side), but a few minor changes in code_aster itself might be needed (this needs verification). Unless you are a very advanced code_aster user and ready to invest time on this (in which case i would be happy to work with you to get this coupling back in shape), you should probably forget this approach.

What might be more realistic is using a simplified model for the movement of the boundaries (for example an expansion and contraction around a given axis following a given time law), in which case you can specify the wall displacement or velocity with user-defined functions, without any coupling required. In this case you directly define the conditions for wall deformation in the matching user-defined function (or even in the equation editor in the GUI), without using the "mass-spring" model type boundary conditions.

Best regards,

Yvan
Shayor
Posts: 24
Joined: Tue Jan 05, 2021 3:40 pm

Re: ALE method and Internal Coupling Problem

Post by Shayor »

Thanks a lot for all the guidance uptil now @Yvan Fournier Sir and  @Luciano Garelli Sir.I did not quite understand how I can define or specify the wall deformation(I did not find the option but yes the velocity I did find ,are you refering to using the veolicty to define displacements?)?.Also can it be defined momentarily at points in small regions?

Lastly I know I asked a lot of questions but I really love the software and want to learn as much of it as I can so my question is could you guide me as to how to couple the code?I have a test case and I do know python basics.If you tell just guide me how to use the main.py script and what to modify in my code/run_case or others I would really appreciate it.Thank you very much again.  
Post Reply