CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   Surface-mounted cube in turbulent regime using k-epsilon (

harly August 14, 2009 06:49

Surface-mounted cube in turbulent regime using k-epsilon
3 Attachment(s)

I am using a 3D Mesh created in Gambit and I am trying to visualize the flow past a cube at Re=20.000. I have the mesh running for a couple of days now and my problem is, that the side force coefficients (which are computed by OF using the appropriate libs) are decreasing (see attachment), so that I am afraid of getting a stationary result in the end.

What could be the reson for that ?

Here my config. and some pictures:

I am using turbFoam with a DeltaT of 0.0078 which results in a theoretical Co. No. of about 0.75.

I use a velocity inlet with U = 1 m/s (resulting in a viscosity of 0.00005)

I use a pressure outlet with p = 0

So far I've been using Euler as time and upwind as convective discretization scheme. I wanted to wait until I have a constatn amplitude for the side forces before switching to higher order schemes.

The lower wall is a wall

and the upper boundary is set to symmetry right now.

My grid has the dimensions 42h x 24h x 12.5h where h is the cubes height.

I will assist with all my resources if asked for - I just didn't want to overload the post.

Thanks in advance.

I had a similar Problem with a sphere in a similar setup and it also tended to become stationary despite the high reynolds numbers - maybe there is a general bug in my setup ?

BTW: When I am using the same BCs and calcualte a 2D cylinder everything works fine - so it works on some level ...

All times are GMT -4. The time now is 18:21.