|
[Sponsors] |
January 20, 2008, 11:50 |
CFL number
|
#1 |
Guest
Posts: n/a
|
Hi all,
I am using "density based implicit " solver with steady state.For Turbulene: RNG k-epsilon model with volumetric chemical reactions. For this problem what should be the maximum CFL number? with regards sisir |
|
January 20, 2008, 16:35 |
Re: CFL number
|
#2 |
Guest
Posts: n/a
|
for most of the problems, the solution will be stable if its close to 1.. exceptions do exist...if your soln is converging slowly, try increasing d Courant no.
|
|
January 21, 2008, 00:04 |
Re: CFL number
|
#3 |
Guest
Posts: n/a
|
Thanks Deepak for ur reply. But in my problem, the residuals start oscillating after few iterations. So what shall i do for this? In the mannual it is mentioned i can go for higher CFL number .
with regards sisir |
|
January 22, 2008, 07:51 |
Re: CFL number
|
#4 |
Guest
Posts: n/a
|
there are so many factors cencerning oscillation, it dostnt necessary have something to do with CFL;in my understanding usually CFL smaller than 1;for LES it might between 3-1, but at end it will be smaller than 1; for normal turbulence Model, CFL has to be smaller than 1 to converge
|
|
January 22, 2008, 20:15 |
Re: CFL number
|
#5 |
Guest
Posts: n/a
|
Thanks Jesse for your information. CFL number will be 1 even for "density based implicit " solver with steady state! But FLUENT manual is suggesting for increase of CFL number beyond 1 even more than 5 also. What shall i do ?
with regards sisir |
|
January 27, 2008, 02:57 |
Re: CFL number
|
#6 |
Guest
Posts: n/a
|
See, manual suggests to increase the CFL, if your residual plot is going smoothly but slowly. Increasing the CFL no will accelerate the convergence (Correct me if iam wrong).
But to start with, go ahead with 1, if it diverges, fluent automatically decreases to .1,.001 .....if it converges try increasing. Oscillating residual plots occur either if your bcs are not correct or your mesh is not as per the requirement of the turbulence model. make sure that ur bcs are as per the models and if thats fine, try a finer mesh. i hope this is of some use. |
|
January 31, 2008, 07:46 |
Re: CFL number
|
#7 |
Guest
Posts: n/a
|
Thanks for the suggestion. I will try with finer mesh.
sisir |
|
February 2, 2008, 08:30 |
Re: CFL number
|
#8 |
Guest
Posts: n/a
|
Hi Sisir
For an implicit scheme your courant number do not have to be equal or less than 1. Infact in order to obtain more stable solution you should increase your courant number as recommended in documentation. Osciallating residuals is more complicated issue. Residuals can osciallate due to bad mesh or there can be other reasons. As a rule of thumb you need to understand that oscillating residuals indicate unsteady flow. Try running your case as unsteady and see what happens. |
|
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[snappyHexMesh] SnappyHexMesh for internal Flow | vishwa | OpenFOAM Meshing & Mesh Conversion | 24 | June 27, 2016 09:54 |
[snappyHexMesh] external flow with snappyHexMesh | chelvistero | OpenFOAM Meshing & Mesh Conversion | 11 | January 15, 2010 20:43 |
CFL number specification sonic/supersonic flow | gr | Main CFD Forum | 0 | January 16, 2009 12:14 |
high cfl number and discretization scheme | Fab | Main CFD Forum | 0 | March 2, 2008 13:19 |
About Courant (CFL) number | Jason | Main CFD Forum | 2 | March 17, 2003 12:11 |