2D flow dispersion

Stephen

I'm wondering is anyone could give me any pointers on a problem I am facing. I am trying to model flow in a 2D duct which widens at an angle of 45degrees causing separation of the flow. The problem I am having is that, when compared to a velocity profile produced using PIV experimental techniques at the point where the pipe straightens again, it seems that the spreading of the separated flow is greater in my CFD calculations. I have checked turbulence parameters to ensure reasonable figures. Are there any other variables within Fluent which would suppress this spreading?

Thanks