CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM (
-   -   Mixed Convection Turbulent Boundary Layer Problem (

deji March 4, 2013 18:52

Mixed Convection Turbulent Boundary Layer Problem
1 Attachment(s)
Hello Foamers... I am having a problem with my mixed convection turbulent boundary layer case. It seems the free stream velocity that I prescribe at the bottom of my computational domain is unable to "flow through" the domain. As you can see in the attached file, I prescribed a free stream of 0.2m/s at the southern boundary with a 10cm entrance length and the plate is at a higher temperature than the ambient. The more I look into it, I suspect that the BC at the east (opposite to the wall) might be the problem. The BC for U there is pressureInletOutletVelocity and pd is totalPressure. I am employing the pressureInletOutletVelocity to allow for entrainment due to the upflow of mass near the wall. Can anyone give me some feedback on what might be causing this. Two more things, I have expanded the domain and my simulation crashes, :confused: and when I prescribe the internal field to 0.2m/s, the BC on the east does not like it and it forces the streamwise velocity to zero. :confused::confused:

deji March 8, 2013 15:26

So no one can give me any feedback at all? :confused:

deji March 9, 2013 15:24

It seems like there's nothing in OpenFOAM that caters to combined convection flows. Am I correct? All flows aren't strictly forced or natural convection.

deji March 15, 2013 10:22

Problem solved!!! I had to make some changes to the modified pressure equation pd. It feels great. If anyone in the future encounters a problem in simulating the classical mixed convection turbulent boundary layer, I might be able to assist. ;)

cm_jubayer November 27, 2013 15:26

Hi deji,

Can you please let us know the way you solved this mixed convection problem? I am also trying to simulate a low Re flow where the natural convection cannot be neglected compared to the forced convection. Thanks.


deji November 27, 2013 15:32

Hey. It would be profoundly difficult for me to assist you without any specific questions.

cm_jubayer November 27, 2013 16:33

Hi deji,

You said you made some changes to the pressure. What are the changes? What are your pressure boundary conditions? or, if possible, Can you please upload a test case so that I can look into your boundary conditions? Thanks.


Thomas William March 13, 2014 05:32

Hi Deji,

I am trying to simulate water flow in a storage tank with forced and natural convection.

transient, compressible: variable density (T), turbulent

Im not having great results with buoyantPimpleFoam? What solver did you use and you mention some pressure modifications?

Thank you

deji March 13, 2014 10:56

I started with a solver called fireFoam, and I eventually had to modify the code to enable the computation of mixed convection turbulent wall flows. I called this code lowMachFoam. Regarding the pressure modification, I essentially took out the compressibility for stability. I think you may have success with the utilization of the fireFoam code, give it a try!


morard April 9, 2014 03:27


I'm trying to simulate a turbulent mixed convection case so that I have decided to post my question here instead of creating a new thread.

My case is turbulent mixed convection to water in a square duct. Heating is asymmetric (just one wall is heated while the other three are not). For
temperature I use fixedGradient at the heated wall while the other walls are adiabatic. fixedValue at the inlet and advective at the outlet are
use for both temperature and velocity. For p_rgh I use buoyantPressure at the walls, zeroGradient at the inlet and fixedValue at the oulet
(I have also tried totalPressure at the outlet without success). Solver is buoyantPimpleFoam (term dpdt is excluded from the energy equation).
SGS model is homogeneousDynOneEqEddy. OpenFOAM version is 2.1.x.

The problem is: as soon as I introduce the heating at the wall, it comes to the unphysical increase of the velocity at the outlet and to the back-flow
from the outlet boundary.

Unfortunately, I cannot upload an image at the moment. The geometry is alike the one form Deji with the difference that I have four walls in my case.

Does anybody know what might cause the problem and how to solve it?


morard April 15, 2014 05:24

2 Attachment(s)
Here are the images at the outlet boundary.

Kader May 26, 2014 15:20

Dear Morad,
I have similar difficulty, I think the problem in u, p, p_rgh boundary conditions at the inlet and outlet. I have tried differents combinaisons, but can't find the adequates boundary.
best regards

morard May 27, 2014 03:25

Dear Kader,

The only combination of boundary conditions that works for my case is:


type totalPressure;
U U;
phi phi;
rho rho;
psi none;
gamma 1.4;
p0 uniform 101325;
value uniform 101325;

type buoyantPressure;
gradient uniform 0;
value uniform 101325;


type pressureInletOutletVelocity;
phi phi;
value uniform (0 0 0.044);

type pressureInletOutletVelocity;
phi phi;
value uniform (0 0 0);

Four outlet velocity you can also apply inletOutlet or advective. But, I do not know how to specify inlet velocity at some specific value.


Kader May 27, 2014 06:24

Thank you Morad, I will try it

Kader May 29, 2014 09:51

Dear Dejan,
I have used the recomanded boundary condition, I have the flowing error message:
request for volScalarField rho from objectRegistry rgion0 failed availabe objects of type volScalarField are
nut ...
I think the error in p_rgh boundary condition
have you any idea

All times are GMT -4. The time now is 05:32.