Hi,guys:

I am modelling single phase flow in porous media region and want to get the pressure drop. The case is 2d, 0.5x60 mm dimension with water liquid as the fluid media. The inlet is uniform velocity 0.0667 m/s and outlet is zero pressure. Since it is laminar, i only set the viscous resistance 1.5e11 for each direction. I found an interesting phenomenon: the pressure drop depends on the mesh fineness dramatically (several order of the magnitude)! For example, i have used uniform mesh with size 50, and got 6.0e5 Pa pressure drop. However, if i set the boundary layer with starting size 10 or 20, and 4 ~ 5 grouth layers, the pressure drop is 2 or 3 order of magnitude bigger. Beside, the velocity field is flat, same almost everywhere except the boundary point. Anybody have an idea? If somebody has interest, you can test it very quickly. Please help me out. Thanks.

Janice

Do you think a pressure loss of 6 bar over a length of just 6 cm with laminar flow is realistic? What kind of porous media is this supposed to be?

At least it explains why the velocity field is completely flat, don't you think?

The unit is Pa, not bar.

6.0e5 Pa = 6 bar

Hi Janice

i have 2 suggestions:

1) Could you check your grid when you use a different BL resolution - make sure that you do not accidentally change the length of the porous region (change the physical length of the porous zone).

2) What are your convergence criteria - do you run the solution to machine accuracy, or do you use the standard 1e-3 convergence limit. For such small geometry you should check the influence of the convergence criteria on the pressure drop. and use 2nd order discretization.

cheers