primitive variables vs vorticity-streamline results
 

Hi,

I am predicting the velocity field for a droplet of water moving between two parallel plates. I've done this by taking a lid driven cavity code in primitive variables and adding velocity to the lower wall. Also I have considered different droplet aspect ratios (H/L) and nondimensionalized using v*=v/(UH/L) with standard nondimensionalization for the other variables.

I am comparing to with some results that are computed using the vorticity stream function formulation and I am getting the correct behavior, but slightly different magnitudes. In the center of the droplet I see a parabolic profile, but according to the results from vorticity stream function, the parabolic profile should be the same as the poiseuille profile. Near the ends, I am seeing slug like velocity profile which corresponds to other results.

Does anyone know what could cause a quantitative difference in the center, but not have significant qualitative differences? Thanks