Hi all,

I made 3 models of a 200m long duct with two 1m^2 wide openings. They were angled at 30 degrees, 45 degrees, 60 degrees and 90 degrees.

The first 3 respective ducts worked fine. They had one side with a 500W heat flux, and the other sides ambient... and the air flowed out of the top side at speeds increasing with angle as expected.

However, the 90 degree (vertical) duct gave me some problems. The solver would not converge, and showed a circulation pattern inside the duct, with the downdraft side actually running along the heated side... clearly not realistic. Any suggestions on how to model convection off of a vertical heated side along the length of a duct?

have u switched off the radiation solver option ?

Actually I didn't. I'll give it a try, since 1 meter is definitely a small enough length to cause some heating on all sides.... however, it didn't happen in the other 3 angles.