CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   TRANSIENT SOLUTION CONVERGE (http://www.cfd-online.com/Forums/main/12478-transient-solution-converge.html)

MARK October 30, 2006 09:16

TRANSIENT SOLUTION CONVERGE
 
I will ask a simple question related transiet analysis. How can I understand a transiet solution is converged or not? Thanks already for your replies.

ganesh October 31, 2006 00:25

Re: TRANSIENT SOLUTION CONVERGE
 
Dear Mark,

A "transient solution" does not "converge" in the same sense as a "steady-state" soltuion. To terminate a transient problem, you would need to look at the response of suitable integrated coeffcients(eg. Cl or Cm for an airfoil) with time, and if they have achieved a periodicity with nearly constant amplitude then transient response is considered to be "converged"

Hope this helps

Regards,

Ganesh

m.suresh kumar October 31, 2006 02:04

Re: TRANSIENT SOLUTION CONVERGE
 
In the software CFDesign, we will see the last ten to twenty iterations in the convergence monitor. If the same pattern has been repeating and values are more or less similar, we will judge as converged.

Mani November 2, 2006 11:08

Re: TRANSIENT SOLUTION CONVERGE
 
ganesh, that's useful only if there is a periodic solution.

I always thought the usage of the term "transient" is arbitrary. The real question is "how do I know that an unsteady solution is converged?", and the answer is: by observing convergence within each time step (in a dual-time approach) and by performing a temporal resolution study (decrease the time step and see if the solution changes significantly).


All times are GMT -4. The time now is 04:51.