[p_agoodboy]

Hi everybody,

I am trying to simulate convective heat transfer in a tube with cte wall heat flux. I use standard k-eps mpodel and SIMPLE scheme with first order discretization and Least Square Cell Bases gradient method.

My results shows that there is unacceptable temperature and velocity difference between the first node (on the wall) and the second. when I ask software to show cell values rather than node values this jump vanishes but the velocity on wall becomes non-zero.
It seems that the high gradient at the first cell is the cause of this problem.

Anybody has any idea? Please share with me.

PS: the boundary layer mesh is very fine since I need to have a number of nodes in the laminar sublayer.