[TaPantaRei]

Dear all cfd-online members,

my question involves two parts. First of all I am using buoyantSimpleFoam solver and I would like:
1) to include a fixed heat flux wall boundary condition. Is that possible in OF and if yes how?
2) in case that is possible to have a fixed wall heat flux boundary condition, I would like that I dont have to specify the heat transfer coefficient, so that I calculate it in post-processing using the temperature difference solution field. Is it possible?

thanks and cheers,
TaPantaRei

[thiagopl]

Hi,

Yes, it is possible. I think what you want something is a boundary condition of type fixedGradient (see sec. 5.2 of OF Users Guide).
What you want then is to calculate a convection heat transfer coeficient h [W/m²K]?

[TaPantaRei]

So thiagopl you mean that in the file T for the temperature, for this wall, I should set a fixedGradient boundary condition? Yes my aim is to calculate a heat transfer coefficient, assuming a heat flux value.

[thiagopl]

Yes. You'll need to define your Nusselt number and then calculate the h coeficient.
I don't know how to do this in OF. I usually export data and do all the post processing outside OF.

[TaPantaRei]

Thiago there is a misunderstanding, I didn't clearly mentioned it, but I want to calculate the material's conductivity k [W/mK]. So, from the equation: q = -k*dT/dx, I need k. I was thinking if it was possible to fix a value q, and then from the temperature solution field I compute dT/dx and then I just find k. In Ansys i know thats possible. I don't know if I fix a gradient dT/dx how can i find k?

[thiagopl]

You already gave the answer
You know the flux and you calculated the temperature field, so, use some discretisation scheme for dTdx and then you can find k.
It is still (for me) a post processing issue.

[TaPantaRei]

hmm i don't see how exactly I gave the answer because you say that I have to use fixedGradient in the T file, so this means I will give a value for dT/dx which is -q/k. Then from the temperature field how I get k? what I wanted is that I give in T file a fixedValue for q if that's possible? and then from the temperature field I can solve for k.

[kingjewel1]

Originally Posted by TaPantaRei

hmm i don't see how exactly I gave the answer because you say that I have to use fixedGradient in the T file, so this means I will give a value for dT/dx which is -q/k. Then from the temperature field how I get k? what I wanted is that I give in T file a fixedValue for q if that's possible? and then from the temperature field I can solve for k.

Dear TaPantaRei,

Did you ever solve this problem? Like you I just want to specify q but don't know k or T... What do you think?

Cheers,

[zfaraday]

Maybe what you need is externalWallHeatFluxTemperature or turbulentHeatFluxTemperature BC's.

Hope it helps.

Best regards

Alex

[gtarang]

Hi,
I have to use heat flux at the boundary and convective (robin) boundary condition at the other one as heat loss.
I tried using externalWallHeatFluxTemperature, but it doesn't seem to work with buoyantPimpleFoam but turbulentHeatFluxTemperature works. Can anyone explain the reason. In principle both are same and during debugging I found both are getting solved and both are giving same output, but externalWallHeatFluxTemperature fails with some library error message:

diagonal:  Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0
PIMPLE: iteration 1
DILUPBiCG:  Solving for Ux, Initial residual = 0.99845272, Final residual = 4.411321e-07, No Iterations 11
DILUPBiCG:  Solving for Uy, Initial residual = 1, Final residual = 3.8992528e-07, No Iterations 11
#0  Foam::error::printStack(Foam::Ostream&) at ??:?
#1  Foam::sigFpe::sigHandler(int) at ??:?
#2  ? in "/lib/x86_64-linux-gnu/libc.so.6"
#3  Foam::heThermo<Foam::rhoThermo, Foam::pureMixture<Foam::constTransport<Foam::species::thermo<Foam::hConstThermo<Foam::perfectFluid<Foam::specie> >, Foam::sensibleInternalEnergy> > > >::he(Foam::Field<double> const&, Foam::Field<double> const&, int) const at ??:?
#4  Foam::mixedEnergyFvPatchScalarField::updateCoeffs() at ??:?
#5  Foam::fvMatrix<double>::fvMatrix(Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> const&, Foam::dimensionSet const&) at ??:?
#6  Foam::tmp<Foam::fvMatrix<double> > Foam::fv::optionList::operator()<double>(Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> const&, Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh>&, Foam::word const&) at ??:?
#7  ? at ??:?
#8  __libc_start_main in "/lib/x86_64-linux-gnu/libc.so.6"
#9  ? at ??:?
Floating point exception (core dumped)

It shows that Ux and Uy are getting solved but during T solution there is some mistake.
Kindly suggest

--
Tarang

[cfdsolver1]

I am little bit confused what the problem is but, using fixedGradient you can define constant heat flux. So, as I know if you define q = 1000 then you need to give OpenFoam the result of dT/dx by dividing your q into k and you will define k in your thermophysical properties file.

[gtarang]

That is a way I am solving the problem right now but further ahead, my k will change with temperature and also during turbulence, k will be replaced by k+keff.

[calf.Z]

Originally Posted by cfdsolver1

I am little bit confused what the problem is but, using fixedGradient you can define constant heat flux. So, as I know if you define q = 1000 then you need to give OpenFoam the result of dT/dx by dividing your q into k and you will define k in your thermophysical properties file.

I need the fixed wall flux q in wall BC. If kappa is changing with T, and now I know the value of fixed wall flux q, but kappa is not constant, how should the fixedGradient be set?

[gtarang]

Originally Posted by calf.Z

I need the fixed wall flux q in wall BC. If kappa is changing with T, and now I know the value of fixed wall flux q, but kappa is not constant, how should the fixedGradient be set?

This will work with buoyantSimpleFoam or buoyantPimpleFoam because they use thermophysical properties. *BossinesqFoam do not use thermophysical properties.

[calf.Z]

Originally Posted by gtarang

This will work with buoyantSimpleFoam or buoyantPimpleFoam because they use thermophysical properties. *BossinesqFoam do not use thermophysical properties.

Thank you, I have noticed the boundary BC. But I am confused with some parameters in it, such as: thicknessLayers 、kappaLayers and kappaMethod. Do you have a good understanding of it?

[calf.Z]

Originally Posted by gtarang

This will work with buoyantSimpleFoam or buoyantPimpleFoam because they use thermophysical properties. *BossinesqFoam do not use thermophysical properties.

Thank you, I have noticed the boundary BC. But I am confused with some parameters in it, such as: thicknessLayers 、kappaLayers and kappaMethod. Do you have a good understanding of it?

[gtarang]

Leave the thicknessLayers, kappaLayers as it is (0 and empty). These are use if you want to put liner(s).
kappaMethod denotes tells the way to calculate kappa (thermal conductivity). I generally leave it as fluidThermo. If you spend some time with the code and go layer by layer, you will get an idea how to change it. For my incompressible heat transfer simulations, I leave it as fluidThermo. For compressible flows, thermodynamics is slightly different, so some different method may be used.
Check thermophysical properties in constant folder, you may get some hint.

[peterhess]

https://drive.google.com/file/d/1d-e...ew?usp=sharing

see 4.7. - 4.9.

Regards

Peter

[calf.Z]

Originally Posted by gtarang

Leave the thicknessLayers, kappaLayers as it is (0 and empty). These are use if you want to put liner(s).
kappaMethod denotes tells the way to calculate kappa (thermal conductivity). I generally leave it as fluidThermo. If you spend some time with the code and go layer by layer, you will get an idea how to change it. For my incompressible heat transfer simulations, I leave it as fluidThermo. For compressible flows, thermodynamics is slightly different, so some different method may be used.
Check thermophysical properties in constant folder, you may get some hint.

Thank you for the hints. I am now using compressible solver, so should I use other method instead of fluidThermo?

Kappa in my simulation in a little different. I use tabulated tables to provide thermal properties. So Kappa is read from "Kappa" table. In thermophysicalProperties:
.......
transport
{
mu
{
file "constant/mu";
outOfBounds clamp;
}
kappa
{
file "constant/kappa";
outOfBounds clamp;
}