I am setting up a vertical duct natural convection problem. It is starting from static conditions so I am trying to use pressure boundries at the inlet and outlet. It seems to just let all the fluid fall out of the bottom. I am wondering if anybody had any experience with pressure boundries or natural convection and could help. Thanks Brandon

I think pressure boundary is used to solve momentum eq. instead of pressure eq. (which is from continuity eq. and does not need pressure boundary). If using pressure boundary on both in/out, no velocity boundary is imposed except wall. The code will do "anything" to make up the flow field and sometimes it fails. I would suggest extend either in or out to some extent and close the far end, if possible. Some codes may do it better for make-up for this kind of boundary, even it is not a reasonable setup. Have you ask ADAPCO technical support. They might have more experience to help you get around.

You may try to set the pressure at the inlet (at the bottom of the duct) little bit higher than the pressure at the outlet (the top) and see what happen.

Have you tried some other boundary configration? For example, stagnation boundary (at inlet)--- pressure boundary (at outlet) or inlet boundary (at inlet) -- outlet (at outlet). The former one considers the velocity distribuiton while the latter one included the flow rate influence explicitely.

In some cases, they works well.