Treatment of open radiation boundaries

[carstenp]

Hi,

currently I am working with the heat solver and I am not sure how open diffuse radiation boundaries are handled in Elmer. According to the Models Manual eqn. 1.14 describes the coupling of the surface elements. But this is only valid if the radiation surface is closed, otherwise there should be also a factor representing the coupling with the ambient background radiation. Where is this factor and where is this coupling modeled?

carsten

[mzenker]

Hi,

for radiation exchange with the environment, you can use something like

Code: Select all

  Radiation = Idealized
  Emissivity = 0.98
  External Temperature = 298

as boundary condition (under Heat Equation in ElmerGUI).

HTH,

Matthias

[cemg]

Hi carsten,

when the radiation exchange takes place into an open enclosure you can use the keywords:

Code: Select all

  Radiation Boundary Open = Logical True
  Radiation External Temperature = Real xxxx

The first one avoids the normalization of the view factors (as long as an angle fraction is lost for some or all of the surfaces). The second assigns a temperature to those lost angle fractions so that idealized radiation can be applied.

Regards,

Cesar

[carstenp]

Hi Mathias, hi Cesar,

thank you for your quick reply. But the question is slightly different. I already know about the flag "open radiation boundary". The question for me is how the open radiation boundary is handled. I often deal with the problem that I have models with a view factor sum larger than one. Now the question is what happens with the external contributions? Is the external contribution in this cases equal or smaller than zero to get in sum a factor of one?

carsten

[raback]

Hi

The sum of view factors should not be larger than one. That could of course happen because numerical integration is not precise. In that case, I think, the result would be a negative flux to external radition.

Ideally for such cases there would be some better solution. For example, use the average temperature as the temperature for the negative flux. Or have a normalization that will normalize only such that sum(f)<=1, that could be tricky though. However, I cannot promise anything. The real problem seems to be in the view factor compuation.

-Peter