|
[Sponsors] |
March 11, 2005, 07:34 |
On the convergence of unsteady problems.
|
#1 |
Guest
Posts: n/a
|
The time must be 5s or 10s before the convergent solution obtained? If the computation time is very long, what should I do? Thanks.
|
|
March 11, 2005, 14:37 |
Re: On the convergence of unsteady problems.
|
#2 |
Guest
Posts: n/a
|
Hi,
Unsteady problems (in general) tend to take more time than the same geometry and flow but steady state flows. This is because you have additional parameter (namely time,t) in addition to velocity, pressure, temp..etc. The computation time also depends on your particluar flow geometry (mesh size) and flow type (e.g natural convection problems tend to take more time). I don't know your particlular problem but there is a general rule that I follow i.e your time step should be less or equal to the smallest time-scale of the flow. So, first calculate this time-scale and set your time step based on that....this could be an eastimate not an exact calculation since you don't know the flow before hand. But you still have a feel what to expect etc.. Also set the time step to be as small as possible during the initial step then keep on increasing or let Fluent do this for you by switching to "adaptive time steping". Last, your mesh size directly affects your time scale. regards Melaku Habte |
|
March 12, 2005, 02:37 |
Re: On the convergence of unsteady problems.
|
#3 |
Guest
Posts: n/a
|
Thanks for your help. But I don¡¯t know how long the computation will take before the convergent solution obtained, in general, 5s or 10s? I don¡¯t know.
|
|
March 12, 2005, 03:16 |
Re: On the convergence of unsteady problems.
|
#4 |
Guest
Posts: n/a
|
Hi,
I still didn't get your question. How did you pick 5s or 10 sec - is there any physics behind. you must have a reason to pick a certain time step. I also whant to know how long is your actual flow time and mesh size. People would be able to help you when u give a little more clarification of ur problem. If for some reason you have to pick either 5 or 10 s then I would say pick 5s. regards Melaku Habte |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
conditions for unsteady problems | dvdromnu | OpenFOAM Running, Solving & CFD | 5 | November 9, 2009 03:15 |
Convergence of CFX field in FSI analysis | nasdak | CFX | 2 | June 29, 2009 02:17 |
unsteady flow problems | Daniel | FLUENT | 0 | January 26, 2009 00:54 |
convergence for an Unsteady simulation | Palani Velladurai | FLUENT | 1 | March 19, 2007 11:56 |
Unsteady simulation convergence | Tomislav | Main CFD Forum | 1 | December 6, 2006 08:53 |