I have managed to create a simple CFD model to integrate in 3D the heat/diffusion equation, namely dT/dt = k d2T/dx2. This works fine, but I would now like to add the effect of 'less dense rising' assuming the gas is operating in air -- I think I'm right in saying that's called buoyancy (I'm not a physicist). SO my question is, can someone post a (hopefully) simple extension to this nice little equation that includes the buoyancy effect. I would assume it involves divT dot g, (where g is a vector of gravity), but thats as far as I can fathom. Any help much appreciated, Dave.