[enjo]

Besides my previous post about the 3D cavity, I'm also trying to set up a simple 2D Karman vortex street example, still using OpenFOAM 10 and icoFoam. I do get the expected vortex shedding behind the cylinder, but the vortices quickly dissipate with distance into not much more than waves, and I also need to set the viscosity unexpectedly low for that (=> Re = 1000 or so with 1 m/s inflow and 0.2 sqrt(2) cylinder diameter).



Already tried refining the mesh some more (double cells in each direction) and a bunch of viscosity values but no visible difference. I'm pretty new to all this so not sure if a specific solver algorithm or so could help?


Attached: Case directory, mesh view, result after 20 seconds (glyph view). The fvSchemes/Solution is originally from the OpenFOAM tutorial pisoFoam/LES/pitzDaily by the way.

[Yann]

Hello,

First I would try to increase the height of the domain. I guess the top and bottom boundaries are walls and I doubt you can develop bigger vortex structures with such narrow domain.

Let us know how it goes,
Yann

[enjo]

Hi Yann, thanks for the tip, but increasing the channel height didn't change anything either. There aren't even any waves in the newly added area, let alone vortices.


I still suspect some algorithm tweak is needed... I tried using the explicit Euler scheme instead of backward also, but that needs very small time steps and doesn't add anything besides making the result look more chaotic, in fact the structures are actually less visible then!


Btw I did find an OpenFOAM Karman street simulation on GitHub, will check if any parameters from there help.


Edit: If I double the space in front of the cylinder as well, as in that simulation, I do get some broader wave/semi-vortice action, but only briefly after the shedding begins.

[enjo]

Image of that as I can't add attachments while editing apparently. Time is 13 seconds of 20 simulated.