[Alczem]

Hello everyone!


I am conducting tests with the externalWallHeatFluxTemperature boundary condition with buoyantSimpleFoam in OF v2212.


I use this condition with the mode coefficient and several layers and conductivities. There is an emissivity setting in this BC, and I noticed that depending on whether I would set emissivity at 0.9 or 0, the heat flux through the boundary would vary greatly (ten times more when set to 0.9).


I would like to be sure that the radiative heat transfer computed with this emissivity is considering the "outside" temperature, meaning the temperature at the end of all ther layers I set up in the BC, and not the temperature of my inner wall, if that makes sense.


Currently, my domain gets a bit too cold when I activate the emissivity towards the exterior ambient temperature.


Can anyone with experience with this BC help me on this? Thanks

[Alczem]

Hey!


A quick update:


I tested the boundary condition (OF2212 and chtMultiRegionSimpleFoam) on both sides of a solid at the same temperature, with virtual layers enabled only on one of the side:

boundaryField
{
    
    
    beton_to_woolrock
    {
         type            externalWallHeatFluxTemperature;
         mode            coefficient;
  
         Ta              constant 263.15;
         h               constant 5.0;
         
         kappaLayers    (1);
         thicknessLayers    (0.1);

         kappaMethod    solidThermo;
         emissivity     0.9;
         value           $internalField;
     }
    
     
     wall_int
    {
         type            externalWallHeatFluxTemperature;
         mode            coefficient;
  
         Ta              constant 263.15;
         h               constant 5.0;

         kappaMethod    solidThermo;
         emissivity     0.9;
         value           $internalField;
     }
     "adiabatic.*"
    {
         type            zeroGradient;
         value           $internalField;
     }

}

The heat flux should be 0, which it is when I disable kappaLayers and thicknessLayers, but when I enable them, the flux is around 2000 W.


I have been testing several configurations, and the only conclusion I can draw is that you can't use emissivity with the virtual layers and conductivities.


The case is attached for everybody interested to see for themselves. I am inclined to report this as a bug because I could not understand the source code of the BC, and I am out of ideas for pinpointing the errors I could have made. Can someone let me know if that would be okay?


Thanks

[dlahaye]

Greetings.

The bc externalWallHeatFluxTemperature works fine for us, see e.g. https://www.mdpi.com/1996-1073/15/24/9618

Can you please re-iterate on what you are trying to accomplish? Thx.

[Alczem]

Originally Posted by dlahaye

Greetings.

The bc externalWallHeatFluxTemperature works fine for us, see e.g. https://www.mdpi.com/1996-1073/15/24/9618

Can you please re-iterate on what you are trying to accomplish? Thx.

Hello and thank you for your paper. I took a glance at it and indeed, it looks like your wall heat fluxes are well converged and relevant.


The issue I have is that my wall heat flux at the boundary using externalWallHeatFluxTemperature is changing a lot depending on the value of emissivity in the boundary condition.


For instance, the case I posted is a simple solid with two opposite faces both using externalWallHeatFluxTemperature, with both of the same Ta and h and same virtual layers. The other faces are adiabatic. In theory, whatever emissivity I set won't matter for the total heat flux, it should be 0 since the temperature is the same everywhere.



But when emissivity is set to 0.9 for instance, my heat flux converges towards 2000 watts and the temperature is not identical in my domain, and that is what I don't get. I am probably understanding the BC wrong, but after quite some time I can't figure out my mistake.


Any help would be great ! Thanks

[dlahaye]

I find it hard to advice you. I am not able to examine your case in near future.

That said: the boundary condition essentially says that

partial T / partial n = h * epsilon * ( T - T_{inf})

this implies that

epsilon has definitely impact on heat transfer

and that

heat transfer is zero in case that T = T_{inf}.

Does this help?

[Alczem]

Originally Posted by dlahaye

I find it hard to advice you. I am not able to examine your case in near future.

That said: the boundary condition essentially says that

partial T / partial n = h * epsilon * ( T - T_{inf})

this implies that

epsilon has definitely impact on heat transfer

and that

heat transfer is zero in case that T = T_{inf}.

Does this help?

Hello,


Are you refering to the code source of the boundary condition? I can't find where the definition you mention appears, can you link it please?


Is epsilon the emissivity in this case? I am asking because I am confused by this whole formula that includes both the emissivity and h. I thought that the emissivity would be included only in the radiative heat flux: emissivity*sigma*(T^4-Tinf^4)


Thanks for the feedback and the help




Edit:


After some more tests, I can confirm that the settings that cause a non-zero flux are using a non-zero emissivity with virtual layers and thermal conductivities such as the following:

boundaryField
{
    
    
    wall_ext
    {
         type            externalWallHeatFluxTemperature;
         mode            coefficient;
  
         Ta              constant 263.15;
         h               constant 5.0;
         
         kappaLayers    (1);
         thicknessLayers    (0.1);

         kappaMethod    solidThermo;
         emissivity     0.9;
         value           $internalField;
     }
    
     
     wall_int
    {
         type            externalWallHeatFluxTemperature;
         mode            coefficient;
  
         Ta              constant 263.15;
         h               constant 5.0;
         
         //kappaLayers    (1);
         //thicknessLayers    (0.1);

         kappaMethod    solidThermo;
         //emissivity     0.9;
         value           $internalField;
     }
     "adiabatic.*"
    {
         type            zeroGradient;
         value           $internalField;
     }

}

Am I missing some basic physics understanding here?