[MFWilliams]

Hello,


In running an FSI case similar to the 3dTube case, would there be any reason to change the default entries for the dynamicMeshDict for solid and fluid?



And why are the two dynamicMeshDicts different for solid and fluid?


Thanks
Merlin

[Hgholami]

Hello,
the dynamicMesh try to move mesh relative to moving boundary to keep the quality of mesh. So in fluid usually boundary moves and the dynamicMesh try to modify internal nodes of mesh via dynamicMeshDict parameters.
Default dynamicMeshDict can use for general cases.
Base on your geometry, the moving patch must change (wall) and to prevent negative mesh, may you need to change parameters.
The dynamicMeshDicts of fluid and solid must be different, each applied to related region. For solid usually, the staticFv uses because the movement occur on whole mesh and you don't need to modify internal nodes of solid mesh.

Originally Posted by MFWilliams

Hello,


In running an FSI case similar to the 3dTube case, would there be any reason to change the default entries for the dynamicMeshDict for solid and fluid?



And why are the two dynamicMeshDicts different for solid and fluid?


Thanks
Merlin

[MFWilliams]

Dear Hojatollah,


Thank you very much for that explanation.


Just to clarify, for my geometry the default 'dynamicMotionSolverFvMesh' will suffice but I might need to change the parameters?


What parameters do you refer to becuase there are not many entries in the dynamicMeshDict. The default one is seen below:

dynamicFvMesh   dynamicMotionSolverFvMesh;

solver velocityLaplacian;
diffusivity quadratic inverseDistance 1(wall);

Also what does the '1' mean after inverseDistance because from what I found this only appears in the 3dTube tutorial and none of the other FSI tutorials.

Also what is 'wall' referring to? Is it referring to the fluid-solid interface patch?


Thank you in advance.
Merlin

[Hgholami]

Hi Merlin,
you should explain your geometry, but as default dynamicMotionSolverFvMesh is OK.
You can change the "solver". Two solvers available, displacement base, velocity base. In the

solver velocityLaplacian;

the velocity base selected. you can also refVelocityLaplacian.
This solver, use laplacian to calculate internal node motion, so need diffusivity that be in

diffusivity quadratic inverseDistance 1(wall);

The "quadratic inverseDistance" is the diffusivity that uses quadratic inverse distance from (patches), "1/L^2". The "1" number of patches that be in "(****)". In this file, "1" means one patch in parenthesis. The patch "(wall)" usually is FSI Interface, because it moves, the 1/L^2 does higher mesh deflection in near of interface and smaller mesh deflection in further of interface.

[MFWilliams]

Ok I understand now. Thank you very much for explaining.


Have you had a chance to look at my questions regarding solidProperties?