I am working with CFD2000, and I try to resolve the convection problem in greenhouse so I don't know which type of boundary conditions must use in the outlet side. Any answers would be appreciated.

Hello,

I can say about only that I have experienced so far. In most of the computational methods the default boundary condition for outlet is the constant pressure. For example I'm using Fluent and when I make an outlet section, non of the options are allowed to be changed and it provides the second order boudary condition, that is the constant pressure. That is all I know so far. I hope it was useful for You.

Regards from Regert

Hi Although "cfd-"people used to set up zero pressure as outflow boundary conditions, implying zero traction normal to the outflow boundary for a fully developed flow, one must specify non-zero normal traction for the outlet when buoyancy forces are present as it concerns your problem. Notice that the non-zero normal traction attributed to a non-uniform pressure distribution. Normally, pressure gradient in vertical direction is directly related to the buoyancy force, as may be seen from the governing (here, momentum equation, vertical direction). Therefore, you could easily work out the relationship for the outflow pressure. As a rule, pressure is proportional to the temperature integral. If outflow temperature is specified, you can easily set the pressure on the outflow boundary. I'd recommend iterations though, since zero diffusive heat flux normal to the outflow boundary is usually specified. Hope this helps.

Hi, again. Sorry, I forgot to include a good reference for OBC: J.M.Leone Jr., "Open boundary condition symposium benchmark solution: stratified flow over a backward-facing step", Int. J. Num. Methods in Fluids, 1990, vol.8, pp 55-64. Results are quite good to verify your treatment of OBC, however, not the best ones. Regards

Dear Mr.ÑÎÂÎÊ and All, please answer on one question. I solve standard 3-D Navier-Stokes + Energy + Equation of State (incompressible). On an output I set pressure p = 0. All is O.K. Then I add in one of momentum equation a buoyancy term "g(rho-rho_r)", where rho_r=const. How it is necessary to change a condition "p=0"? Thank you. Grigory.

As long as you get the buoyancy forces involved, pressure distribution is no longer uniform. For the sake of simplicity, you may wish to set p=0 at your most favorite point on the outlet (i.e. the centre of the lower wall) though.

Hi,COBOK i didn`t find that reference you gave on Int.J.Num.Methods in Fluids,1990,Vol8,pp55-64. Are you sure about that. Regards

I apologize, I copied and pasted a wrong part of the reference. You should try: J. M. Leone Jr, `Open Boundary Condition Symposium Benchmark Solution: Stratified Flow Over a Backward-facing Step', International Journal for Numerical Methods in Fluids, 11(7), 969-984, 1990.

Thank you. I will try this.