Radiation ("warning: property request") + Boundary Treatment

[Saint]

Hi !

A
I am trying to gradually advance my simulation to a multiphysic, natural convection with radiation; but after adding the radiation I have faced a warning:

"WARNING:: ListFind: Requested property: [External Temperature], not found"

This error has occurred after adding the radiation property to the boundary condition of my boundaries.
I have used the "Idealized" with emissivity of 0.9 - 0.91 for different boundaries.

The main medium is "air" with emissivity of "0", but I thought, the boundaries are in contact with different materials and the emissivity will be applicable.
Side note, which might be cause of the warning: I have only one material in the simulation and one shape.
Sif and log file of this simualtion is attached.

22 sif.txt

(4.48 KiB) Downloaded 5 times

22 log.rar

the log file was filled with so much warning that has caused the file size to exceed 2Mb !

(8.72 KiB) Downloaded 3 times

https://uploadly.me/34cbc (entire simulation)



B
My second question is regarding the upcoming steps:
Assuming that I have a box filled with air and only 1 hot surface:
How should I introduce the radiation of boundaries?

1- Please confirm: with radiation with "idealized" mode, I don't/can't need to define the "radiation target body" or "radiation boundary" but with the "diffuse gray" I need to define them (boundary is needed if the surfaces do not see each other directly)

2- If the temperature of my surface is fixed, is the dirichlet temperature enough ?
how about checking the box for the "open boundary" to define the T as fixed ?

3- when I have two materials, like here, a box and an air; to simulate the heat transfer between two materials, I must check the heat gap box; and without it, Elmer will assume there is a chemical/physical continuity, and will simulate the heat transfer as an internal property ?




4- If I replace one of the box surface with a heating element (e.g. if the red boundary of the image was a heating element as a third body) how should I introduce its generated heat:
- Bodyforce: Heat source
- Bodyforce: Bodywise dirichlet conditions
- Manually setting different boundarie's temperature and heat fluxes

5- assuming that I have selected either internal heat generation or manually sat the heat flux at each boundary, will the Elmer automatically calculate the back reflections?
I mean, my element is radiating heat, but the same radiation will be reflected out of nearby surfaces and reach back; should I do something special to compensate for this effect ?

[kevinarden]

If you are radiating heat out of the boundary than Elmer wants an external temperature not a fixed temperature.

Boundary Condition 1
Target Boundaries(1) = 7
Name = "400 ns"
External Temperature = 400
Emissivity = 0.9
Noslip wall BC = True
Radiation = Idealized
End

Boundary Condition 2
Target Boundaries(7) = 1 2 3 4 5 6 8
Name = "ns"
Emissivity = 0.91
Radiation = Idealized
Noslip wall BC = True
External Temperature = 300
End

[Saint]

Thank you so much as always.

how about if I introduce a second body to the system:
if the second body is generating heat (but it exert the heat only from one surface, and yet it can absorb heat from surrounding)

I don't think that I should use "bodyforce heat", since it can escape from all of the surfaces.

which leads me to use dirichlet/heat flux or external heat at boundaries.

IF, this scenario is correct:
"using a heat flux (or fixed dirichlet temp) in boundary of my hot-plate"
Then, should I define another flux/dirichlet value for the corresponding boundary of my air ?
(I mean the air boundary which is directly overlap the hot-surface of my hot plate)

or it will be redundant, and Elmer can detect it itself?

what if I want to fix the temperature of a box, containing my air and hot-plate,
- Should I use the dirichlet T, at its boundaries, or is there something similar to the "boundary open" (in radiation) section?
Correction, I belive, the bodywise drichlet conditions is the better choice

[kevinarden]

The body force creates heat in the entire body. If heat is just one surface you should use boundary condition. The heat at a boundary will conduct to any body that is attached to that boundary.

You can fix the temperature of the box at its boundaries using the boundary conditions. But if there is heat input into the system the temperature of the body will not be fixed.

[Saint]

The body force creates heat in the entire body. If heat is just one surface you should use boundary conditions. The heat at a boundary will conduct to any body that is attached to that boundary.

You can fix the temperature of the box at its boundaries using the boundary conditions. But if there is heat input into the system the temperature of the body will not be fixed.

Thanks !
-So if I have only a single body (air) in my simulation, if I want to introduce heat to this system, I should use "External Heat" and not a "dirichlet temperature"
(the yellow marked, is the correct way?)




-If I have a hot-plate and air surrounding (2 bodies):
Should I use one "dirichlet temperature" as for my hot-plate + one "external temperature" as for my air boundary
or
Only having one "dirichlet temperature" is enough and Elmer, will understand that what is exiting form one boundary, will enter the corresponding/overlapping boundary of the other body (air)

[raback]

Hi

To original questions
1) yes
2) Dirichlet fixes the temperature and overruns everything else. Open boundary means that view factors will not be normalized to 2Pi. As they sum of to less the rest is assumed to be idealized radiation.
3) No. Heat gap is intended to jumps. All PDEs are by default continuous over material interfaces.
4) Depends on the physics but typically "Heat Source".
5) For "diffuse grey" radiation yes. It has nothing to do with other settings.

-Peter

[Saint]

Hi

To original questions
1) yes
2) Dirichlet fixes the temperature and overruns everything else. Open boundary means that view factors will not be normalized to 2Pi. As they sum of to less the rest is assumed to be idealized radiation.
3) No. Heat gap is intended to jumps. All PDEs are by default continuous over material interfaces.
4) Depends on the physics but typically "Heat Source".
5) For "diffuse grey" radiation yes. It has nothing to do with other settings.

-Peter

Thanks !

how about this:

-If I have a hot-plate and air surrounding (2 bodies):
Should I use one "dirichlet temperature" as for my hot-plate + one "external temperature" as for my air boundary (1+1)
or
Only having one "dirichlet temperature" in my hot surface is enough and Elmer, will understand that what is exiting form one boundary, will enter the corresponding/overlapping boundary of the other body (air) (1)