The trailing edge velocity

maximus

Hi,everyone! I am conducting the calculation of flow around a NACA0012 airfoil with a sharp trailing edge. I use the RANS, and turbulence model is S-A one equation model. The results show that the velocity is reverse at the nodes near trailing edge. I don't know why, and has anyone meet this problem ever? There isn't this problem when I use the k-e model, I wonder if it is because the the length scale in S-A model. It needs to calculate the distance to the wall, and there is a free shear flow after the trailing edge. how to deal with the length scale at the nodes after the trailing edge? Thanks!!!