[kyubi]

Hi, I'm a new OpenFoam user.
My first openFoam experience is the study of turbolent flow in a circular pipe with a sudden enlargement.
I'm interested about steady state, and so I choose simpleFoam with standard k-epsilon mode.

The geometry is axi-symmetric, so for the mesh I use the blockMesh and makeAxialMesh utilities.

About the boundary conditions I set:
• inlet BC for the pipe inlet, so fixed value for U (1m/s), k and epsilon. About k and epsilon I extimated the values following the Programmer's Guide, page 58.
• outlet BC for pipe outlet: fixed value (0) for pressure; (but in fvSolution I change the pRefValue to 101.325 = 101325/1000)
• wall BC for pipe surface: fixed value (0 0 0) for U (no-slip condition);
• wedge BC for the front and back planes.

In the transportProperties I set 1e-6 for nu value (the fluid is water).
In this conditions the Reynold's number is 10000.

I use upwind in the fvScheme to ensure boundedness:
divSchemes
{
default none;
div(phi,U) Gauss upwind;
div(phi,k) Gauss upwind;
div(phi,epsilon) Gauss upwind;
div(phi,R) Gauss upwind;
div(R) Gauss linear;
div(phi,nuTilda) Gauss upwind;
div((nuEff*dev(grad(U).T()))) Gauss linear;
}

Everything seems to work fine, but when I try to compute the head loss coefficient xi I don't get the correct value: the areas ratio is A₀/A₂ = 0.2; hence, according with theory, I should find 0.64 = (1 - A₀/A₂)².
In order to calculate xi I use the mean velocity in the smaller pipe (U₀) and the pressure difference between section where streamlines return parallels to pipe axis (p₂) and section immediately after the enlargement (p₁):

csi = 2(p₂-p₁)/U₀²

my result is 0.28...

Have you got an idea about which is my mistake?

Thank you!

Marco