pipe turbulent flow

Hao

i'm using RSM to simulate a 5 m/s turbulent flow through a pipe. The geometry is 0.02*0.4 (width*length). I couldn't get a fully developed flow by one time. So I ran the simulation first time and got the profiles at the outlet. Then i imported the profiles as the inlet conditions in the second simulation. Is this method right? It seems that all the turbulence properties varies little along the pipe after the second one. However, one thing is just weird. The wall shear stress has a sudden rise at the very outlet when the outlet type is pressure outlet. With a outflow-oulet, the wall shear stress has a sudden drop. could anyone give me some suggestion? Thanks in advance.