CFD Online Discussion Forums - Low Speed Flow Boundary Conditions

Hi,

This is a question from a newbie CFD user regarding a problem that really has me stumped.

I have successfully used FLUENT in the past for transonic aerofoil and wing analysis using a compressible flow setup and the coupled solver. However, I'm now trying to simulate the flow around various 2D bodies at low speeds (e.g. 10 to 30 m/s at sea level) and small angles of attack. No matter what I try I seem to get x and y velocities that converge, but mass continuity gets "stuck" at about 10e-1 or 10e-2.

The flow field looks sensible, there are no "hot spots" in terms of velocity in the domain, and lift coefficient seems sensible(ish). Mass fluxes into the domain = mass fluxes out of the domain.

Mesh:

"Extended D" shaped domain with aerofoil at centre.

Boundaries are at least 50 chord's away from aerofoil. ~30000 elements.

I have tried:

1) unstructured tet mesh (with high quality elements)

2) Structured quad mesh

Boundary conditions:

Front of D & bottom of D - set to velocity inlet at 30 m/s with velocity components to give 2 degree AOA.

Top of D and back of D - set to pressure outlet with a pressure of 101325 Pa normal to boundary.

Aerofoil - set as a wall.

Solution set up:

Segregated Solver.

Incompressible flow.

Energy equation disabled.

Operating pressure set to 101325 Pa.

I have tried both first and second order upwind discretisation.

I have tried inviscid, laminar and turbulent solutions.

I have tried changing under relaxation of momentum with no improvement.

I don't think the problem is to do with grid resolution because it won't run inviscid. Plus, if I set up the same mesh as a transonic problem with the coupled solver and pressure far-field boundaries it converges nicely. I think it might be my selection of boundary conditions, but I don't have enough experience to know better.

If anyone has any pearls of wisdom to offer I will happily try anything. Thanks in advance for any help offered.

Paul