[RicardoLB]

Hello everyone!
I am starting to work with overset meshes and taken the floatingObject tutorial as a starting point. I want to get a 2D floating object setup to carry out simple heave decay tests (and eventually to add wave generation and absorption boundary conditions). I have a half-working setup ready.



I am getting this error repeatedly:

--> FOAM FATAL ERROR:
Cell:2115 at0.42125 0 0.254285714286) zone:0 changed from hole to calculated but there is no donor

This happens in a cell which is in the last and second to last timesteps a hole... I attach an image. Can anybody give me hints on how to solve this problem?



I also observe some apparently strange movements of the floating object as if it where excited (I don't explain myself the initial acceleration from the initial position of non-equilibrium between the waterline and the submerged volume and the following steps). I think I have assigned the correct mesh groups and boundary conditions; are they correct?


I attach the case folder.


Greetings,
Ricardo

[romainping]

Hi Ricardo,

I am facing the same error message on OF 1812. I do not have this error on version 1712...
Have you found a solution ?
Thanks for your help,

Romain

[Jim1310]

This is a case that particularly interests me, so i searched what the pitfall had been. Eventually it must be the fact that although there should be motion only in the plane, for some reason (probably due to arithmetic error) there is some parasitic movement-rotation outside of the plane. It can be seen in the log-file of the solver that initially there is an off-plane rotation of about ~10^(-30), which is gradually build up until ~10^(-6) where the simulation cruses. This can be easily solved by imposing movement constraints (see wiki-page: https://openfoamwiki.net/index.php/P...on_and_Damping). I imposed a plane constraint (movement only on the z-x plane) and an axis constraint (rotation only in y-axis) and it worked without a problem (my Of version: v2106).

[sri_0548]

Hi Jim,


I am trying to simulate 2D floating cylinder in a numerical wave tank using interfoam solver (OpenFoam v2206). It is observed that the cylinder is sinking. I applied the constraints as you mentioned. Still no change in the result. Could you please check the case file and help me in solving this issue.


Thanks and regards
Srinivas

[Jim1310]

Hello Srinivas,


First of all this is a simulaiton of a box and i don't think that you can directly compare it with the 3D counterpart. In any case, from a look into your dynamicMeshDict i see that you have set the mass of the object to 21.6kg, which is too much given that the dimensions of the box are 0.5X0.01X0.2 (0.01 is the thickness of the mesh from extrudeMeshDict) and this leads you to a weight of 1kg if it had the same density with water. So i guess there is nothing wrong except that it crushes or slows down dramatically due to cells below the box beeing overly squished.


Best,
Dimitris

[sri_0548]

Thanks for the reply Jim.
Is it the correct technique to mimic a 3D cylinder in 2D? I constructed a 3D cylinder in STL file and converted it to 2D using snappyhexmesh and extrudemesh. Will the simulation work if I give mass according to 2D dimensions? Is there a way to simulate 3D geometry in 2D using a dynamic mesh?

[Jim1310]

Well in 2D you essentially simulate the 3D motion of a very long prism having the section you see in 2D. I dont think there is any way this motion can be an approximation lets say for the disk, since its sections change if you cut them using the x-z plane... (with the z-axis beeing in the cenere of the cylinder). If you changed the orientation of the cylinder, so that the y-axis is the axis of symmetry you could approximate the falling motion in 3D of a very long cylinder, but i guess that it would be much different than what you are trying to simulate now.

[sri_0548]

Here, I want to simulate the heave motion of a cylinder under wave interaction. Generally for flow past cylinder we use circle in 2D. In my case the orientation is different because as an extension of this I want to use cylinder with conical bottom.If I use circle I can't get the effect of conical bottom for the same orientation. Is there any better way to do this?

[sri_0548]

Hi Jim,


I tried the simulating the floating box with wave interaction. This time I gave the properties according to rectangle box in the dynamicmeshdict. The wave is generated and the box is oscillating up and down. But after 6.25 seconds the simulation is crashed. The time step and CFL number are with in limits.I am not able to find out the reason. Please find the attachments I have send the error.txt along with the case file. Please help me to solve this.




Thanks and regards
Srinivas