CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   CFD-number-segregated solver (http://www.cfd-online.com/Forums/fluent/101355-cfd-number-segregated-solver.html)

Lilly May 2, 2012 05:12

CFD-number-segregated solver
 
Dear all,

I want to simulate a transient case for an incompressible fluid (mixture: contrast agent bolus in blood) using a segregated solver and have some general questions concerning the CFL-number:

I read in some threads that the CFL-number ist important for stability reasons at coupled solvers. But do I have to take care of the CFL-number at segregated solvers as well? And is it just a stability reason or may I also get wrong results if I choose a time step which is too large for my grid or a grid size which is too small (at using a segregated solver as well)? I just thought I might get wrong results since my contrast agent bolus would "jump over" some grid-elements then...

Thank you for all your help and ideas,
Lilly

Lilly May 2, 2012 05:21

CFL-number
 
Sorry, I meant CFL-number, of course!

LuckyTran May 2, 2012 09:50

Quote:

Originally Posted by Lilly (Post 358788)
Dear all,

I want to simulate a transient case for an incompressible fluid (mixture: contrast agent bolus in blood) using a segregated solver and have some general questions concerning the CFL-number:

I read in some threads that the CFL-number ist important for stability reasons at coupled solvers. But do I have to take care of the CFL-number at segregated solvers as well? And is it just a stability reason or may I also get wrong results if I choose a time step which is too large for my grid or a grid size which is too small (at using a segregated solver as well)? I just thought I might get wrong results since my contrast agent bolus would "jump over" some grid-elements then...

Thank you for all your help and ideas,
Lilly

Hi Lilly, it is the mostly same for the segregated or coupled solver.

You only need to focus on CFL number for the explicit time-stepping. For implicit time-stepping (stability is guaranteed). The way implicit time-stepping scheme is handled in Fluent it is dummyproof (as there are iterations in-between) and you can do iterations as large as you like. For explicit time-stepping you will have both of the problems you mentioned: 1) stability and 2) the case of grid jumping so CFL number is very important. That said, maintaining the proper CFL number even for the implicit time-stepping helps drastically to improve its convergence at each time-step.

Lilly May 4, 2012 10:54

Thank you for your helpful explanation LuckyTran!


All times are GMT -4. The time now is 21:46.