CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Pre-Processing (
-   -   Numerical problem in buoyantBoussinesqPimpleFoam (

vtcampos July 3, 2012 08:43

Numerical problem in buoyantBoussinesqPimpleFoam

I am trying to simulate heat transfer in a small vertical cilindrical pipe. The pipe is established with its axis in the y direction, which is the direction of the acceleration g.
Air comes from the inlet which is the lower part of the pipe, flows through it and exits at the upper part. It looks similar to this:

| |
| | walls
| |
| |

I am using the buoyantBoussinesqPimpleFoam with kOmegaSST turbulence model.

The boundary conditions for time 0 are as following:
U: inlet fixedValue (0 8 0); outlet zeroGradient; wall fixedValue (0 0 0) (no slip condition)
T: inled fixedValue, 300; outlet zeroGradient; wall of the pipe fixedValue 450;
p: calculated uniform 0 in all the boundaries;
p_rgh: buoyantPressure uniform 0 in all the boundaries;
alphat: inlet fixedValue 0; outlet zeroGradient; wall alphatWallFunction uniform 0
kappat: inlet calculated uniform 0; outlet zeroGradient; wall kappatJayatillekeWallFunction value=uniform 0 and Prt 0 0.85;
nut: inlet calculated uniform 0; outlet zeroGradient; wall nutkWallFunction uniform 0;
k: inlet fixedValue uniform 0.3267; outlet zeroGradient: wall kqRWallFunction uniform 0.3267
omega: inlet fixedValue uniform 3760; outlet zeroGradient: wall omegaWallFunction uniform 3760

The problem appears when I try to run the solver and it asks for a reference point and value for the field p_rgh. I have set the reference value as 0, since my conditions are null pressure at outlet. Therefore I decided to set my reference point at some cell on the outlet boundary. However, no matter which cell or reference point I choose, in the outlet or not, it comes to a numerical error with unphysical velocity and pressure values (for example U = 500 m/s) at that exact point.
Can any of you help me finding my error on the simulation?

tomf July 4, 2012 05:12

Dear Vitor,

Try to put a

fixedValue uniform 0 for p_rgh on the outlet instead of buoyantpressure there, I think the solver then also does not need the reference cell or reference value, which it should only need in a closed volume.


AbdelkaderDZ July 7, 2013 07:55

Dear Tom,
I have a problem to simulate the natural convection in channel, I can't found the appropriate boundary conditions for p and p_rgh at the inlet and outlat channel.

All times are GMT -4. The time now is 14:20.