[Tobi]

OpenFOAM does not support 2D adaptive mesh refinement. So you cannot do it. So you have to develop your own library.

[saddy]

as far as i know openfoam refinement works only for 3d cases. there is no refinement for 2d cases

[indy07cz]

Hello,
i have a question. I'm solving a spillway simulation with interFoam solver and i solve 59 seconds of simulation. After that (in 60th second) I would like to refine surface to obtain better (smoother) results but cells I want to refine are protected (300k cells) and refine doesn't work. Should I solve entire simulation with dynamicrefinement or is this problem somehow connected with mesh? lowerLimit for alpha is 0.2 and upper 0.6, maxCell refinement is 2000000. Thank you very much.

Edit: Sorry, there was something wrong with simulation running. Refinement works, protected cells are near geometry surface and they are not hexahedral so protecting is on purpose.

[dzordz]

Originally Posted by Lorena2fdez

Hi all,

I use this library to study the evolution of a droplet at rest. This is a very interesting tool for me, but the problem is that the volume of the drop increases progressively.

I check it and this problem only appears when I use this library and an axi-symmetric mesh (with an empty border and two patch wedges). And more exactly, I think, the reason is that the width of the cells is divided (see attached photo). Width (z-direction) should be 1 cell.

Do you know how to avoid this problem?

Thank you in advance

Lorena


Attachment 58399

Hi Lorena,

the refinement work in 3D, but I have been able to use it also in 2D axis-symmetry. I do not mind if the refinement is also in the third directions, since it still greatly reduces the amount of cells that need to be calculated.

I can see that near the symmetry line you do not get the refinement. This is due to the fact that the cells there are not hexahedrons and are not set to be refined. Simplest way this can be resolved is by doing a small cut of the bottom cells (e.g. 1% of length) which can be done with extrudeMeshDict. In this way you basically turn all the bottom cells to hexahedrons. (Just be careful that this cut does not change your case in any significant way)

I did a "damBreak" test, to test if Tobias's interface refinement behaves properly also in wedge and it does (at least for a simple case like this). See the attached picture.


The only problem with doing this kind of refinement in wedge is that you get nonAlignedEdges where the refinement happens, which can be problematic. If anybody knows how to resolve this I would much appreciate it (2D refinement would ?).

Hope this helped,
cheers