[ThatGuy]

Hi all,

I've been trying to run some simulations on a full-scale geometry which is about 20mx10m with some things inside. The models I chose are steady, ideal gas, coupled flow, gravity, turbulent (k-epsilon), axisimmetric. The kind of sim I'm trying to run is a 2D natural convection with a volumetric heat source with gravity vector pointing towards -x.
If I try with this model, either I get some error of overfloat, or it simply diverges. If I try the same with a really downscaled geometry, everything works fine.
I thought it could be a mesh problem, but all me attempts failed as well.
Any ideas?

Since I have to study some heat transfer, I'd like to avoid any reynolds similarity etc, since I'm not very used to that kind of thing.