|
[Sponsors] |
September 12, 2005, 01:31 |
upwinding in the convective term
|
#1 |
Guest
Posts: n/a
|
Hi
In the convective term of the discretized momentum equation you have the sum over all the faces of rho_f*u_f*phi_f . A_f and when upwinding is discussed they say that the face value of phi_f is set equal to the value at the centre of the upstream cell (for a first order upwinding scheme that is). I have two questions regarding this. First of all, phi_f IS u_f, so you don't know it in advance. So when the matrix equation a_p phi_p = sum a_n phi_n + b is formed u_f is in the a_p and a_n coefficients... so you need the value for the thing you're trying to solve for! Does fluent get around this by using the value for u_f from the previous iteration? Secondly is the value of u_f upwinded the same as phi_f? Steve |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
RAS Turbulence Models - Convective Term | john_w | OpenFOAM | 3 | May 29, 2010 22:31 |
ATTENTION! Reliability problems in CFX 5.7 | Joseph | CFX | 14 | April 20, 2010 16:45 |
How to suppress the convective term | SAKTI | Siemens | 1 | August 28, 2008 02:38 |
Convective term evaluated at time n+1/2 | Lionel S. | Main CFD Forum | 0 | February 12, 2007 15:22 |
Approximation Schemes for convective term | Michail | CFD-Wiki | 10 | July 31, 2006 15:33 |