|
[Sponsors] |
July 3, 2008, 09:51 |
Y + in the k-e turbulent modeling
|
#1 |
Guest
Posts: n/a
|
Dear CFD masters, I would like to know how to check the y + value. In the k-e low reynolds number turbulent modeling, that the mesh spacing in the wall-normal direction should be chosen such that the value of y+ corresponding to the near-wall cell centroid is ~1. How to do that? and for k-e high reynolds number turbulent modeling, is there a limitation for y+ value?
Thanks |
|
July 3, 2008, 12:31 |
Re: Y + in the k-e turbulent modeling
|
#2 |
Guest
Posts: n/a
|
getw yplus
popt cont wplo For high Re, 30-100 is an optimal range. Some problems can tolerate higher values. |
|
July 3, 2008, 22:05 |
Re: Y + in the k-e turbulent modeling
|
#3 |
Guest
Posts: n/a
|
Hi Thanks Pauli,
Btw, if the value of Y+ lets say around 10, so to use k-e low reynold number, we should reduce the size of the cell at the wall boundary until we can get Y+ ~1? or to increase the cell size at the wall boundary until we get the range of 30 - 100 to use the k-e high reynolds number model? regards |
|
July 4, 2008, 07:13 |
Re: Y + in the k-e turbulent modeling
|
#4 |
Guest
Posts: n/a
|
you could go either way, depending how much time and computer power you have and what degree of accuracy you need.
|
|
July 4, 2008, 08:38 |
Re: Y + in the k-e turbulent modeling
|
#5 |
Guest
Posts: n/a
|
Hi, I manage to get the y+ <1, and I try to use the low reynold number, but there it is error from the windows stating that "star402.exe has encountered a problem and needs to close".
Anyone has this experience and know why, please let me know Thanks |
|
July 6, 2008, 12:22 |
Re: Y + in the k-e turbulent modeling
|
#6 |
Guest
Posts: n/a
|
Hi, I use to do Low Re sim but I've never ecountered that problem...maybe there is something else that cause that problem, not the low Re approach.
Regarding your previous question: you know High Re models use wall functions ad so if you use that models you'll get an approximation about the boundary layer dynamics... If you use low Re approach you can describe with less approximation the boundary layer, but you need a very fine mesh as you see, so the computation time rise a lot... In some problems, by the way, you can get right results only by means of Low Re models (where is important the boundary layer, where you see separating regions, etc.) Regards, Andrea |
|
July 8, 2008, 05:27 |
Re: Y + in the k-e turbulent modeling
|
#7 |
Guest
Posts: n/a
|
Hi, thanks for the insight, I noticed that if I increase the cell number the error comes, I dont know why, anyone know why? memory?
Thanks in advanced |
|
July 21, 2008, 06:07 |
Re: Y + in the k-e turbulent modeling
|
#8 |
Guest
Posts: n/a
|
you might want to check the ff: 1. negative volumes by performing grid checks or "everything" under "check setup". 2.double check if your inputs are corrects.
PS:why do you get y+ ~1 without not able to run your sim. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Using different turbulent viscosities for turbulent stress components | braennstroem | OpenFOAM Running, Solving & CFD | 13 | October 15, 2010 09:59 |
turbulent modeling 2D cylinder Re10^5 Compressible | Ehsan | Main CFD Forum | 0 | August 23, 2008 04:56 |
turbulent modeling with wall functon(or modeling) | yoo seon oh | Main CFD Forum | 5 | July 4, 2007 15:06 |
Modeling Turbulent Horizontal Buoyant Jets | Jack Travis | Main CFD Forum | 2 | September 11, 2006 07:36 |
Modeling Atmospheric Turbulent Diffusive Transport | wind_flow | FLUENT | 11 | June 6, 2003 14:59 |