convergence problem
 

Hi all,

I'm simulating a 2-D flow past flat plate in order to investigate the boundary layer (laminar/turbulent). The continuity residual is converging to high value (1.5e-1) whereas the x and y velocity is converging to 1e-7.

I'm pretty sure I have to change the under relaxation factors...anyone know what's the best choice to this kind of problem ?

Thanks, John

One more question...
 

One more thing...what are the best sub-models of k-w and k-e to use for a wall bounded flow ?

I know k-w is much more suitable for this case but I still want to check k-e.

Thanks john

Re: One more question...
 

In resononce to the first question about continuity residuals, depending on what you are simulating, changing an OUTFLOW B.C. for a PRESSURE OUTLET B.C. I would first try lowering pressure and then try lowering momentum.

Re: One more question...
 

Hi, a second possibility would be if you are simulating turbulent flows to change your simulation to an unsteady one!! You have also to look at your wall treatment methode and your grid resolution!!

regards

zachi

Re: One more question...
 

The single equation K-w model is best for boundary layer flow if there is no recirculation. However the K-w model does not allow for the transport of turbulence therefore if you are looking at the turbulent effect away from the wall then a K-e model should be used. If you are having problems with continuity try reducing the relaxation factor for momentum and pressure also check the grid on the boundary and your Y-plus values.