[Aben]

Hi,

I am a new user of OpenFoam. I've just installed foam-extend 4.0 in order to work with the interDyMFoam solver. The issue is that when I run the floating object tutorial, the object doesn't appear. However, the movement of the free surface is normal, the problem is just with the floating object that isn't in the tank. From the log file, it seems that the solver isn't recognised. In addition, when I go to foam-extend-4.0/src, there isn't as in openfoam 4.0 the file sixDoFRigidBodyMotion.

Has somebody ever faced the same issue? If the problem is indeed with sixDoFRigidBodyMotion, is it possible to add it to foam-extend?

(I've already tried to put the file of SixDoFRigidBodyMotion from OF 4.0 in the src file of FE 4.0 and to run the make files but it didn't seen to function.)

Thank you for your help,
Best Regards,
Aben

[wyldckat]

Quick answers:

1. The source code for this feature from OpenFOAM 4 is not compatible with the one in foam-extend 4, so it is not straight forward to either bring the tutorial case from OpenFOAM to foam-extend, nor are the libraries easy to directly port.
2. Viewing the object can be done by following the instructions given here: https://openfoamwiki.net/index.php/H...rocMultiregion
3. This tutorial case is known for being problematic for several years. See the following thread for the known issues back then: The floating object tutorial
   1. Side-note: This is meant to have already been fixed as of OpenFOAM 4.0: https://github.com/OpenFOAM/OpenFOAM...de6fa468dd7148
4. As for specific details for foam-extend, see this old bug report: https://sourceforge.net/p/openfoam-e...ndrelease/255/

[Aben]

Hi,



Thank you for your answer. The issue I have is more with the solver that isn't available on foam-extend. There is no sixDoFRigidBodyMotion in the src file as in OpenFoam4. Openfoam doesn't recognize sixDoFRigidBodyMotion and therefore there is no object in paraview. For the bug reported on: https://sourceforge.net/p/openfoam-e...drelease/255/, the object appears but doesn't move. In my case, there is no object at all and the issue, I think, isn't with Paraview/multiprocessor post processing. However, I have the same issue with the displacement vector that takes large negative values.

Here is the error message diplayed:


--> FOAM FATAL ERROR:
Unknown solver type sixDoFRigidBodyMotion

Valid solver types are:

11
(
RBFMotionSolver
displacementComponentLaplacian
displacementInterpolation
displacementLaplacian
displacementSBRStress
laplace
mesquiteMotionSolver
pseudoSolid
refVelocityLaplacian
velocityComponentLaplacian
velocityLaplacian
)

[wyldckat]

Quick answer: I'm sorry, but I'm unable to understand how exactly you've reached the current error message.

Please provide a step-by-step description of the steps you've taken to reach this error message and how you are trying to see the geometries in ParaView.

The other detail is that it would be helpful to have the log files "log.*" that are in the case folder, in order to see what error messages you've gotten during the mesh generation.

Furthermore, are you running the Allrun script? Or are you doing each command manually?
I ask this because I suspect that if you are doing the commands manually, that would explain why the floating object is not appearing in ParaView.

Run the Allrun script is done with this command within the case folder:

./Allrun

[danib1802]

Hi Aben and wyldckat.

Aben, I don't know if this is the problem, but for me in order to see the object I turned of all the mesh in the Paraview interface and just activate the stationaryWalls, floatingObject and setSubset/floatingObject and you can see the object from below the box. The thing is that otherwise it is covered.

Wyldckat, on the matter that Aben is saying about the sixDoFRigidBodyMotion, I also encountered the same situation trying to solve another problem. I don't find how foam-extend uses this for dynamicMotionSolverFvMesh if it doesn't exist in foam-extend (as in the file is there but it is not linked, or at least that is what I think). I am trying to change the tutorial for the pitchingPlate (that uses IBM and the plate is under forced motion) for the plate being under flow-induced motion.

[Aben]

Hi wyldckat and danib1802,

I figured out why the error message appears. It's because I changed the DynamicMesh from:

//
dynamicFvMesh dynamicMotionSolverFvMesh;

motionSolverLibs ( "libfvMotionSolver.so");

solver displacementLaplacian;

diffusivity inverseDistance (floatingObject);
//

to

//

dynamicFvMesh dynamicMotionSolverFvMesh;

motionSolverLibs ("libsixDoFRigidBodyMotion.so");

solver sixDoFRigidBodyMotion;

sixDoFRigidBodyMotionCoeffs
{
patches (WEC);
innerDistance 0.05;
outerDistance 20;

centreOfMass (0 0 0.005);
momentOfInertia (99999999 200 99999999);
mass 15.672;
report on;
accelerationRelaxation 0.7;
}
//

However, the problem remains. When I put the first command lines in the DynamicMeshDict the error message disappears but there is still no body in the tank and in the log file there is a centre of mass but it takes large negative values after a short period of time (-200 in the z direction).

To see the case in ParaView, I use the ParaFoam command. I use the ./Allrun command to run the case.

Thank you danib1802 for your answer. Unfortunately, the problem is that there is no body in the tank. Actually I wasn't sufficiently precise while describing my problem. The issue is that there is no body at all in the tank.

I tried to run the floating object tutorial on another computer with OF4 and it worked. It workes with RigidBodyMotion. So the problem seems related to FE4.

You will find enclose the log files for the floating object tutorial (I put only an extract of the log.interDyMFoam).

[danib1802]

Originally Posted by Aben

Hi wyldckat and danib1802,

I figured out why the error message appears. It's because I changed the DynamicMesh from:

//
dynamicFvMesh dynamicMotionSolverFvMesh;

motionSolverLibs ( "libfvMotionSolver.so");

solver displacementLaplacian;

diffusivity inverseDistance (floatingObject);
//

to

//

dynamicFvMesh dynamicMotionSolverFvMesh;

motionSolverLibs ("libsixDoFRigidBodyMotion.so");

solver sixDoFRigidBodyMotion;

sixDoFRigidBodyMotionCoeffs
{
patches (WEC);
innerDistance 0.05;
outerDistance 20;

centreOfMass (0 0 0.005);
momentOfInertia (99999999 200 99999999);
mass 15.672;
report on;
accelerationRelaxation 0.7;
}
//

However, the problem remains. When I put the first command lines in the DynamicMeshDict the error message disappears but there is still no body in the tank and in the log file there is a centre of mass but it takes large negative values after a short period of time (-200 in the z direction).

To see the case in ParaView, I use the ParaFoam command. I use the ./Allrun command to run the case.

Thank you danib1802 for your answer. Unfortunately, the problem is that there is no body in the tank. Actually I wasn't sufficiently precise while describing my problem. The issue is that there is no body at all in the tank.

I tried to run the floating object tutorial on another computer with OF4 and it worked. It workes with RigidBodyMotion. So the problem seems related to FE4.

You will find enclose the log files for the floating object tutorial (I put only an extract of the log.interDyMFoam).

Hey Aben,

One question, did you manage to include the sixDoFRigidBodyMotion somehow into foam-extend? Checking back this; that was the main problem, it was not included into the motionSolvers... Please let me know. Thanks