[bgp723]

Hello,

I'm using FLUENT 2022 R1

I was investigating adding radiation to my CFD model, and I wanted to do some simple checkout cases to make sure I understood how FLUENT does radiation calcs.

I set up a two 1x1 m2 plates (one held at 300 K, other at 400 K BCs) separated by 1 cm facing each other to make sure FLUENT outputs correct radiative exchange heat flux (essentially, I'm trying to simulate two infinite parallel plate problem to make sure it matches hand calc).

I'm assuming that the fluid medium has no influence on the radiative heat transfer, so I'm using Surface to surface radiation model.

I was very taken aback by the results though, and I think I'm doing something wrong, but I don't know what.

I tried the following:
* Hot and cold plates at 1 emissivity -> ~990 W heat transfer at cold and hot. This is good. Hand calc says it should be 992 W
* Hot and cold plates at 0.1 emissivity (energy iter per rad iter: 10 [default], max num of rad iter: 5 [default])-> -17.3 W for cold plate, 84 W for hot plate. Solution converged very quickly. This made no sense! System energy balance is not retained, and hand calc says it should be 52.2 W.
* I tried the 0.1 emissivity case again but with energy iter per rad iter at 2 and max num of rad iter at 50, and 500. I got -47.3 W for cold plate and 57.3 W for hot plate. This is closer to hand calc, but there's net system energy imbalance of 10 W that just won't go away. I tried playing around a bit with radiation residual conv criteria and # of clusters, but they didn't resolve the issue.

Hmm, I would upload the file here, but it looks like the file size limit is very small... so I'm instead just attaching a screen shot showing this.
I would like to see if anyone else had similar issues using radiation in FLUENT.

Thanks,
DB

[bgp723]

Ah, I have found an answer to my question.
I must have accidentally clicked off the radiation participation on the thin 1cm side walls enclosing the 1x1m plates participating in radiation.
Once I added that back in, the radiative heat transfer values now make sense and matches the analytical equation value.
I'm a little surprised that those thin side walls made that much difference, but I guess given the lower emittance, significant amount of rays were bouncing out of the domain.

One caution for others: it's not incredibly obvious to figure out how much radiative heat is leaking out the non-participating walls, which led me to believe that system level energy imbalance was improper.