I read from one topic, Robin said that 'Pressure is the relative static pressure (relative to the reference pressure you input in the Domains form). If you are running 5.5.1, the hydrostatic pressure contribution will also be included.' I use CFX 5.5.1 to compute the problem with one inlet and one outlet. It is the problem about natural convection with significant hydrostatic pressure. I specify 'Openning' with zero total pressre at inlet, and 'Outlet' with zero average static pressure at outlet. I got the reasonable results. I have to add the hydrostatic pressure to 'pressure' in CFX-Post, then I get the value of pressure almost the same as the value from my own Fortran program. What is the problem of mine? Thank you very much.

If you want to see the hydrostatic pressure, set your reference density to 0 (or at least a very small value). Remember that even when you physically measure the hydrostatic pressure, it is usually relative to the hydrostatic pressure difference in air, this is equivalent to using a reference density of 1.2 [kg/m^3] (i.e. the density of air).

If you do use a different reference denstiy, you will have to accomodate this in your pressure boundary conditions. The advantage to using a reference density equal to your fluid density is that the pressure at a boundary condition is constant for different hights (ie you don't NEED to add the hydrostatic contribution), at least for boundaries that are at the same temperature as your reference.

-Josh

Thank you for your answers, but I am sorry I still have the questions. 1. Comparing with the result with my own Fortran program, I have to add the hydrostatic pressure to 'pressure' in CFX-Post Why? 2. In my case, I do not have to add the hydrostatic pressure to the total and static pressure that I specify as the boundary conditions, do I? Thank you very much.