Dear all,

I have a question concerning incompressible flow with buoyancy.

If I have a vertical outlet plane (pressure boundary) in a case where gravity is included in the equations (e.g., to model natural convection), is it OK to set a constant relative pressure (e.g., 0 Pa) along the outlet or should I correct for the pressure gradient (rho*g*h)?

In my code, the pressure boundary is set by fixing the value of the pressure to a given value for each cell of the outlet. I have seen in commercial codes that you can specify an "average" pressure at an outlet; does anybody know how this is implemented?

Thanks,

Bouke

Bourke,

Define your pressure, p, in the Navier-Stokes equations, as the static pressure minus hydrostatic pressure. I.e. use the difference, density minus the reference density in the buoyant term. Then you can use the boundary condition p=0 on the outlet.

regards

DML

OK!

I thought that would be difficult to implement but the opposite is true & it works like a charm.

Thanks!