CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)

 sunnysun October 29, 2008 12:04

Hi, everyone, I am sorry if

Hi, everyone,

I am sorry if this question is too simple...

It seems that explicit method is suitable for steady flow and implicit is for unsteady flow. But why icofoam can use implicit method for momentum? Also i am confused that why in icofoam,if courant number above 1, solution diverge? As I undertood, with implicit method, convered solution can be found even for large time step/courant number...

Thanks a lot!

vivien

 sunnysun November 3, 2008 06:54

Can anybody give me some ideas

Can anybody give me some ideas?

Regards,

vivien

 ep4 November 3, 2008 10:22

Hi, an explicit method expr

Hi,

an explicit method express the value of a varaiable at the iteration n as a function of values obtained at n-1.
An implicit methods express the value of a variable at the iteration n as a function of values obtained at n-1 and n. Therefore the resolution of a sytem of equations is necessary at each iteration. This takes more time but there is no problem of stability while an explicit method will be stable only if the courant number is smaller than 1.

A way to limit the courant number is to reduce the time step. But, for long unsteady simulations, this becomes rapidly too computing power consuming. It can then be interesting to use an implicit method to use bigger time steps even if the resolution of at each time step is longer.

Note that the fact that an implicit method is stable does not ensure a good solution quality, just that the solution won't diverge.

 sharonyue December 23, 2012 20:12

Quote:
 Originally Posted by ep4 (Post 194796) Hi, an explicit method express the value of a varaiable at the iteration n as a function of values obtained at n-1. An implicit methods express the value of a variable at the iteration n as a function of values obtained at n-1 and n. Therefore the resolution of a sytem of equations is necessary at each iteration. This takes more time but there is no problem of stability while an explicit method will be stable only if the courant number is smaller than 1. A way to limit the courant number is to reduce the time step. But, for long unsteady simulations, this becomes rapidly too computing power consuming. It can then be interesting to use an implicit method to use bigger time steps even if the resolution of at each time step is longer. Note that the fact that an implicit method is stable does not ensure a good solution quality, just that the solution won't diverge.
That helps me alot! thanks!

 All times are GMT -4. The time now is 07:10.