FSI with OpenFOAM and Code-Aster

Hello Joubert!

I'm also trying to perform transient one-way coupled FSI simulations using OpenFOAM and Code_Aster. I'm familiar to the OF simulations but just at the beginning with Code_Aster.
Your last response was quite some time ago but hopefully you can answer my questions:

1) How do you export the pressure data with OF (I'm using OF 2.2.0) and which file format do you use (e.g. Ensight)?

2) How does your .comm file for CA look like for importing the data file (LIRE_RESU)?

3) How do you set up the CA simulation (using DYNA_LINE_TRAN) for the different time steps?

4) Is it possible to use different meshes (mesh type, number of nodes,...) for the interfaces on the fluid and the solid side to distribute the pressure field from the fluid patch to a group of faces on the solid side?


Thanks and BR
Markus

Hi Everyone

Sorry for my absence. I am currently occupied with CFD simulations and paper deadlines.
I will be actively working on this topic again from October 2013 (hopefully).

Markus

Since I have worked on this a while ago I do not remember exactly how it works but I'll do my best.

I wrote a utility for extracting the data the mesh and pressure info. It cycles through all the faces and writes out the face centres (ex: fCenter = mesh.Cf().boundaryField()[patchI][faceI]). The same utility will cycle through all time dumps and write out the pressures/forces at those faces (ex: fPress = p.boundaryField()[patchI][faceI]) . To apply the transient load in code aster I create a force list as was discussed in previous posts (I think) using LIRE_FONCTION (function of pressure over time for each node) which I then apply to shell elements. I use DYNA_LINE_TRAN for the transient analysis. In my case I have different meshes and achieve linking by searching for the nearest neighbour. Applying the pressure in this way introduces a error that can be reduced by refining your aster mesh (assuming its more coarse than the CFD mesh).

Sorry for the little info I can give you at the moment. As I said, working for a deadline. Will hopefully be able to answer better in October.

Regards
Eugene

Anyone is still interesting about this problem?