CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Solution converged, and then diverge again. (http://www.cfd-online.com/Forums/fluent/45993-solution-converged-then-diverge-again.html)

jrg September 11, 2007 08:01

Solution converged, and then diverge again.
 
Hi. I am simulating a rectangular duct flow with 90 degrees inlet and outlet. Flow is laminar and I use the pressure solver, simpleC, standard pressure and 2nd order upwind momentum discretization.

To assess convergence, I monitor the average velocity at the outlet. After a few hundred iterations (say 300), the value becomes very constant and one would think the solution has converged. But if I keep iterating, oscillations appear and I would need to iterate another few hundred before it becomes constant again. I did in total a few thousand iterations and I could see this cycle repeating.

Does this imply that the solution at 300 iterations is not the converged solution? Is there any possibility that a converged solution would diverge with more iterations?

Thanks.

AAA September 11, 2007 16:37

Re: Solution converged, and then diverge again.
 
try adapting velocity gradient

mohitmmb April 2, 2013 00:10

hi .... if u solve this problem then pls send me copy of solution of this.... i am also doing same and facing problem ... my e mail id is mohitmmb@gmail.com



;145276]Hi. I am simulating a rectangular duct flow with 90 degrees inlet and outlet. Flow is laminar and I use the pressure solver, simpleC, standard pressure and 2nd order upwind momentum discretization.

To assess convergence, I monitor the average velocity at the outlet. After a few hundred iterations (say 300), the value becomes very constant and one would think the solution has converged. But if I keep iterating, oscillations appear and I would need to iterate another few hundred before it becomes constant again. I did in total a few thousand iterations and I could see this cycle repeating.

Does this imply that the solution at 300 iterations is not the converged solution? Is there any possibility that a converged solution would diverge with more iterations?

Thanks.[/QUOTE]


All times are GMT -4. The time now is 11:08.