[Allankey]

How do I use radiation in a way that doesn't cause the temperature on my surfaces to rise to over 2000 degrees?

I would appreciate any help because this is driving me insane

[flotus1]

With a little bit more information about your problem, someone might be able to help.

[Allankey]

I am sorry for my vague description.

My model is made up of 3 domains. The first is the air domain, imagine a box where I have defined the top surface as a semi-transparent wall that allows 1000W/m2 of radiation to enter the model. The bottom surface of the air domain is the top surface of the solid domain (my aluminium plate). In the air domain the side surfaces have been defined as two velocity inlets and two pressure outlets to simulate air flow across the top of the collector. The collector is attached to a pipe which carries water and this is the third water domain. I have attached a fairly crude image to summarise what I am trying to model.

The problem I am having is that the irradiation enters the domain and has nowhere to escape resulting in a divergent energy residual which is only stopped when the solver prevents temperature increases above 5000 degrees. When I view the results my plate temperature is around 2500C and I have around 400000W of incident radiation on the surface. I have tried turning the emissivity on and off for the walls as well as turning toggling the option of having each domain participating in radiation however it is always the same problem.

It seems as though the radiation is getting trapped in the domain and not being able to escape. The only options fluent allows me is to change the emissivity of the boundaries but this seems to have little effect on the outcome.

I welcome any suggestions you have and can provide clarity on any of the points if needs be.

Many Thanks

James

[flotus1]

I also messed around with the radiation models in Fluent for a while.
My personal conclusion is: avoid these models whenever possible
I encountered some similar problems like the ones you described.

In your case, you could simply heat the surface of the aluminum plate with a heat source. The effect will be the same, with the benefit of not having to solve additional equations for the radiation.

[Allankey]

Thank you for your reply. Great advice. I have found a way to eliminate the use of radiation completely from my model and have decided to give up with Fluent and focus on CFX which seems to be a bit more user friendly.