Pressure drop not converging with mesh
 

Hello,

I am simulating a laminar (actually with Re<<1) flow between two parallel plates (2D model).
I am trying a convergence study by computing the pressure drop as function of the mesh size: well, refining the mesh I get lower and lower pressure drops, and an increasing difference with the theoretical value which is higher. Surprisingly only a coarse mesh gives me a value close to the theoretical one.

I am using:
- velocity inlet, pressure outlet BCs,
- a quadrilateral mesh refined at the boundaries (bias factor),
- fully developed velocity profile extracted from a previous calculation on the same mesh,
- 1e-6 residuals, QUICK scheme for momentum, second order for pressure (checked may schemes, no differences).

Can anybody help me please?
Thanks a lot

the velocity profile is correct...I don't know the problem...