CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM (
-   -   wallHeatTransfer & Neumann BC (

giordy April 13, 2010 17:27

wallHeatTransfer & Neumann BC
Dear All.

Can someone tell me how to use wallHeatTranfser type to setup Neumann boundary condition for unsteady diffusion equation? Can someone post a case for that?

thank you in advance.

PeterNaa June 3, 2010 07:35

wallHeatTransfer, Neumann or Newton boundary condition for the temperature in OF-1.6
Dear all

I do not yet know how to the BC wallHeatTransfer should be set in OF-1.6. I am trying to find out how I should
make OF-1.6 use the BC wallHeatTransfer on the T-field for my solver called forchFoam. I can compile and run my
solver with this BC, but it does not produce the correct results on the T-field :mad:. The way I have set it in
the 0/T-file is:

type wallHeatTransfer;
alphaWall uniform 20;
Tinf uniform 400;
value uniform 300;

I have included the file thermophysicalProperties under constant/. This file contains:

version 2.0;
format ascii;
class dictionary;
location "constant";
object thermophysicalProperties;
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //


mixture air 1 28.9 1000 0 1.8e-05 0.7;

pRef 100000;

Instead of hPsiThermo, I have also tried using ePsiThermo or hRhoThermo, which works just as well (that is; it runs but does not
give the correct results).

The file createFields.H in the solver-directory contains the following extra lines:

Info<< "Reading thermophysical properties\n" << endl;

autoPtr<basicPsiThermo> pThermo
basicPsiThermo& thermo = pThermo();

My file forchFoam.C in the solver-directory contains the following included files:
#include "fvCFD.H"
#include "volFields.H"
#include "basicPsiThermo.H"
#include "basicThermo.H"
#include "fixedGradientFvPatchFields.H"

and my Make/options contains:

-I$(LIB_SRC)/thermophysicalModels/basic/lnInclude \
-I$(LIB_SRC)/thermophysicalModels/specie/lnInclude \
-I$(LIB_SRC)/finiteVolume/cfdTools \

-lbasicThermophysicalModels \
-lspecie \
-lmeshTools \

As I said, I can compile and run everything without error-messages, but the BC wallHeatTransfer does not produce correct results on T. What I have done in the files createFields.H and in thermopysicalProperties is done in order to make the solver forchFoam run, but what is written there has nothing to do with my actual fluid. I know that the thermophysical model has its own pressure, temperature, viscosity, thermal diffusivity, enthalpy, etc.. built in, so what I am trying to do is essentially nonsense, but somehow I feel it should be possible to use the BC wallHeatTransfer without having to specify an entire model for the thermodynamics. Am I wrong about that ?:confused:

I need to do something that will allow me to run and correctly include the BC. As far I can tell, I need to write autoPtr<...
in createFields.H and include some thermo-stuff written in thermopysicalProperties or a similar dictionary. Does any of you have
any idea, what I should write in createFields.H and in thermopysicalProperties or a similar dictionary instead, in order to
make wallHeatTransfer work correctly?


thekay June 3, 2010 08:11

Can you give a brief description of your model (what you want it to calculate, how it works, phases, etc)? At which part it gives wrong results?

PeterNaa June 3, 2010 08:49

wallHeatTransfer, Neumann or Newton boundary condition for temperature in OF-1.6
Dear All :)

I am trying to make a CFD-model with OF-1.6, which can simulate drying of corn by the use of forced convection
of hot air through this fixed porous material. I am at the beginning phase of the project and I am testing the
various components (the BC's of course play an important role) that should be part of the OF-1.6-files.

My final CFD-model will contain fields of air, vapor, pressure, velocity, temperature and huimidity of the solid phase. Although the
density of the gases may vary somewhat, the flow is essentially incompressible (no compressibility effects). I have not yet
come so far as to implement all these fields since I am now testing the BC's for the temperature. I can define BC's for T on
walls that are adiabatic (zeroGradient BC) or with a fixed rate of heat inflow or outflow (fixedGradient BC), but I have
not been able to define a Newton (also called Neumann) BC, where flux = -const*(T-T0). The BC wallHeatTransfer
in OF-1.6 should do exactly that. It is part of the thermodynamical models in OF-1.6 but I have not yet been able to
understand the full use of these and I feel it should be possible to use wallHeatTransfer separately without having to
specify a full and sophisticated thermodynamic model.


nishant_hull June 4, 2010 11:58

Hi peter,

Now I am getting the problem. Can you tell me from which version of the openFoam you started you forchFoam solver? It doesn't look like the OF-1.6 version. or may be it is icoFoam solver. is it?

your solver is asking for the thermophysicalPreoperty file in the constant folder, which might have initiated because of the wallHeatTranfer BC. you should look for some other solver for this purpose, where the solver read the thermophysicalproperty file.



PeterNaa August 26, 2010 09:51

Dear all

Installing groovyBC does the same job as wallHeatTransfer (


All times are GMT -4. The time now is 05:57.