CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Outlet Boundary Condition (http://www.cfd-online.com/Forums/main/4245-outlet-boundary-condition.html)

 Arash December 30, 2001 12:41

Outlet Boundary Condition

i want to know in defining at oulet boundary condition which parameter must be defined if we know nothing about it. i use finite volume simple method. thanks arash

 Peter December 31, 2001 09:32

Re: Outlet Boundary Condition

If you are running incompressible flow, you don't need to prescribe a pressure profile at the outlet. Usually one specifies that velocities and pressure do not vary in streamwise direction, du/dx=dv/dx=dp/dx=0. If you are using staggered structured mesh, this can be easily implemented in your momentum equation system by adding the term where the velocity is known to the coefficient vector. For the pressure correction equation, no information for the correction term is needed in the boundary, so the corresponding matrix coefficient is equal to zero. Check the book of Patankar for details: Numerical Heat Transfer and Fluid Flow. Page 129.

 John December 31, 2001 13:56

Re: Outlet Boundary Condition

dp/dx ca not be zero - just look at simple channel or pipe flow where dp/dx is constant even at outlet boundary

 Thomas P. Abraham January 1, 2002 20:41

Re: Outlet Boundary Condition

Hello Arash,

You could apply a constant pressure at the outlet. This boundary condition holds good in the region where the flow is fully developed. So, you need to make sure that there is no recirculation in the region where you intend to apply this constant pressure boundary condition.

Even in the fully developed regime, dp/dx is not zero but a constant. Static pressure continues to go down as the flow tries to overcome the viscous forces.

Thanks,

Thomas

 Steven Jachec January 4, 2002 09:41

Re: Outlet Boundary Condition

In modeling shallow water waves in the marine environment, there are times when one does not know the pressure head nor the velocity at an open (unforced) bonundary. In these cases, I use a radiating BC that allows the solutions to be determined on the boundary, as well as allows the waves (including flow) to propagate out of the domain without reflection (Orlanski, 1976. "A Simple Boundary Condition for Unbounded Hyperbolic Flows." Jrn. of Comp. Physics v. 21 (251-269)). Hope this gives you some answers.

Steven Jachec

 All times are GMT -4. The time now is 21:47.