|
[Sponsors] |
November 1, 2004, 02:55 |
Boundary conditions in the proyection method
|
#1 |
Guest
Posts: n/a
|
Hi
Is it possible to apply the projection method in a system with combination of Neuman and Dirichlet boundary conditions for the velocity? sample: flow inside a pipe Boundary conditions (cylindrical 2D Navier-Stokes): inflow (z=0): Uz(0,r) = 3 , Ur(0,r) = 0 outflow (z=L): dUz/dz = 0 , Ur(0,r) = 0 center (r=0): dUz/dr = 0 , Ur(z,0) = 0 wall (r=R): Uz(z,R) = 0 , Ur (z,R) = 0 thanks |
|
November 1, 2004, 04:03 |
Re: Boundary conditions in the proyection method
|
#2 |
Guest
Posts: n/a
|
I think you can still use the projection method, by the way, you may want to change inflow (z=0): Uz(0,r) = 3 as someting like
Uz(0,r) = 3(R-r)(R+r)/R^2 |
|
November 4, 2004, 00:25 |
Re: Boundary conditions in the proyection method
|
#3 |
Guest
Posts: n/a
|
I'm trying program codes for projection method on staggerd grid. When simulate a problem like 'lid-driven cavity flow', the boudary conditions are easy to set, but when a outflow is included, the boudary condition is hard to impose. Most references I have found are concerned about the Dirichlet bondary conditions, who can provide me some more detailed papers on the boudary conditions for projection method? Thanks!
|
|
November 4, 2004, 04:23 |
Re: Boundary conditions in the proyection method
|
#4 |
Guest
Posts: n/a
|
Hi,
Your outflow condition is not right. If you use the velocity boundary condition at the exit (that means you do not use the pressure boundary condition at the exit), the mass conservation must be satisfied at the outlet. Try following way. (1) First impose at z=L, dUz/dz=0, dUr/dz=0 such that; Do J=2,JMAX-1 Uz(IMAX,J)=Uz(IMAX-1,J) Ur(IMAX,J)=Ur(IMAX-1,J) end do (2) Calculate mass flow rate at inlet Flowin=0. Do J=2,JMAX-1 Flowin=Flowin+density(1,J)*r(J)*delta_r(J)*Uz(1,J) end do where delta_r is the size of each control volume in r_direction. (3) Calculate mass flow rate at exit Flowout=0. Do J=2,JMAX-1 Flowout=Flowout+density(1,J)*r(J)*delta_r(J)*Uz(IM AX,J) end do (4) Calculate Factor=Flowin/Flowout (5) Imposed the velocity at exit such that Do J=2,JMAX-1 Uz(IMAX,J)=Factor*Uz(IMAX,J) end do Above method is what SIMPLE people are usually doing. I hope this helps. Halim Choi |
|
November 5, 2004, 09:28 |
Re: Boundary conditions in the proyection method
|
#5 |
Guest
Posts: n/a
|
HI! look at the papers by Philip Gresho & Co.:
(1984) Int. J. Num. Meth. in Fluids, Vol4, 557-598 (1987) Int. J. Num. Meth. in Fluids, Vol7, 1111-1145 (1990) Int. J. Num. Meth. in Fluids, Vol11, 587-620 I found there a lot of usefull tips, but the articles were too much theoretical oriented for me. Anyway, I think that it is a good start! Greetings, Oscar. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Water subcooled boiling | Attesz | CFX | 7 | January 5, 2013 03:32 |
Boundary Element Method = Boundary integral method? | aiqch | Main CFD Forum | 0 | April 3, 2010 16:29 |
Open Channel Boundary Conditions via journal | Matteo | FLUENT | 0 | January 21, 2008 11:05 |
compressible boundary conditions | vivian | Main CFD Forum | 8 | April 24, 2006 06:23 |
Convective Heat Transfer - Heat Exchanger | Mark | CFX | 6 | November 15, 2004 15:55 |