vortex shedding
 

Hallo,

I am trying to solve a 2D flow around a rectangle in cartesian domain. I am using stream function vorticity formulation finite difference formulation. I have applied no-slip boundary conditions on the rectangle. My problem is I could not have vortex shedding when I increase the Re number. In my case the size of the vortexes at the back of the rectangle grow as I increase the Re number but shedding does not appear. Can I solve this problem if I increase the number of grids around the rectangle. I would be thankful if anybody can make me any suggestions.

Thank you very much

Altug

Re: vortex shedding
 

Usually you need to introduce some sort of disturbance to speed up the vortex shedding or else it takes a long time.Another possible reason might be as you said not enough points in the boundary layer.Use a stretching function if you want to reduce mesh size near cylinder surface and not increase the number of points.

Re: vortex shedding
 

try to apply a small disturbance on the grid, for example the upper contour of the rectangle, to break the symmetry of the problem. Alberto

Re: vortex shedding
 

http://www.titanalgorithms.com/works...der_sqr_fe.JPG

http://www.titanalgorithms.com/works...esh_coarse.JPG

http://www.titanalgorithms.com/works..._mesh_fine.JPG

http://www.titanalgorithms.com/works...transducer.JPG

Observe the second and third figures. In this simulation, which sounds to be very similar to yours, the coarse mesh does not produce the vortex street (second figure), while the refined mesh, same geometry and Reynolds number, will result in a vortex street. Links available at:

www.TITANAlgorithms.com/workspace_cfd

Kind Regards,

www.TITANAlgorithms.com

Re: vortex shedding
 

Thank you (Al, Harish and TITAN Algorithms) very much for your attention and suggestions.

Altug

Re: vortex shedding
 

It surprises me that no one has pointed this out, but it's probably because we assume you know this already: Vortex shedding is an unsteady phenomenon. You don't mention solving the unsteady equations, nor did you say anything about the temporal resolution you are using. Temporal resolution (small time steps) is just as important as spatial resolution (small grid cells). Coarseness in both space and time will have the effect of suppressing the instability of the steady flow, and hence suppressing vortex shedding.