Hi all

I have been simulating the flow of air over a symmetrical ship, and at 0 deg yaw (head-on to the wind) the flow should be symmetrical. It is not. I have tried many different meshes and adaptions, I have tried different turbulence models as well as inviscid flow, and I have re-scaled to reduce the Reynolds number. I cannot get a symmetrical flow. Any suggestions would be appreciated. Thanks.

Hi,

first I'd like to ask, whether the solutions you got are well converged. It might just be that the solution slighly oscillates.

Secondly - the solution will not be completely symmetrical on unstructured non-symmetrical grid due to numerical errors. So you need to think about better mesh or perhaps solution-based adaption.

Third option is - if you know that the solution must be symmetrical on a physical basis, you may enforce such solution by simply generating grid for only a half of the symmetrical geometry.

Hope this helps George

Thanks George

Yes the solutions are fully converged, I think the solution does oscillate slightly as a point velocity monitor shows this. The mesh is a combination of tet and hex cells, unstructured, and is not perfectly symmetrical. Solution adaptions have been done, but cannot achieve symmetry in the solution. I did wonder about the numerical errors. The solution is expected to be symmetrical and experimental validation has shown that it is, so I did think that a symmetry plane was my next move - just felt like cheating! Thanks for your help, Dave