Stokes
|
February 10, 2006 13:46 |
About BC for turbulence k-omega models.
Hello, all, I'm testing the in-house k-omega/k-eps models by unstructured FEM method. I meet the negative k and convergence problems. It's well known that omega=(2~60)*dynamic visscous/(beta * y^2) depending on standard or SST or any other k-omega models etc, here, beta=0.075, y is the first grid point way the wall. For my testing case, visc is 10^-7 around, y 0.001, then the BC for omega is 10~250. However, if i impose 250.0 for example, then k is always negative. On the otherhand, if i impose 10.0, then there is no convergence of k/omega equaitons. Does anyone here have any comments about above problems? Thanks for your helps.
|