[shserg12]

Hello, I prepare the model of air moving in baloon of the inflated structure building. During modeling I faced with a problem: value definition of a reference density in the buoyancy settings.
I used a simple model and cfx help for understanding influence this value on the model results. I understood from the CFX help and this post Gravity/Buoyancy/Buoyancy reference density in two phase flow that "The exact value of this parameter should not matter", but in my case if I use value 1.41 [kg/m^3] (fig.1), I get a flow in the domain due convection with up-down direction. For used value equal 1.451 [kg/m^3] I get a flow with down-up direction(fig.2).

fig1.jpg

fig2.jpg

I used in both cases the Full buoyancy model with the Air Ideal Gas material; heat cylindrical surface with temperature: 22[C] and the Opening boundary with the Opening pressure 0 Pa for mass and momentum settings, and temperature -30[C]. Ref. pressure in domain equal to 101325 [Pa].

[evcelica]

I believe this is because you have an opening with some vertical height to it, so static head pressure will be included on that opening as a function of your reference density.

When you set you reference density too low, that static head is lower at the bottom than it should be, so the opening it the bottom it is a low pressure area, (or the upper opening is a high pressure area) and flow goes downwards due to that false pressure gradient you input with your reference density.


Or your gas is all heavier than reference density, so it flows downward.
These might be mathematically equivalent, and both cases would not happen if you did not have an opening with vertical height, as that pressure gradient would not have an effect without the opening.

[shserg12]

Thanks for the answer! Do I understand, that if I use the opening, the reference density is necessary to set how the air density outside the opening boundary (depending on the outside air temperature)? If I do so, then the result is the correct.