[JinBiao]

Dear All,

I am calculating the flow in a tube with the wall heated. The fluid property, including density, varies strongly with temperature change. So the flow at the outlet apparently deviates from the fully developed condition. The calculated radial velocity at outlet is given in the attached figure. With non-zero radial gradient of radial velocity, the zerogradient boundary condition at the outlet leads to poor continuity convergence. Also the streamwise velocity (see attached file) is distorted at the outlet compared with the upstream profiles due to the application of zerogradient boundary condition for velocity.

Is there any boundary condition type in OpenFOAM can avoid this problem?

Thanks for any discussion.