|
[Sponsors] |
February 8, 2012, 07:46 |
skewnesss
|
#1 |
Senior Member
hamid
Join Date: Nov 2010
Posts: 185
Rep Power: 15 |
Hi
I have a question regarding skewness, how much is the maximum allowed skewness, how can i improve the mesh by that? i can not make it finer than what it si, on the other hand i even fined it much but it became even worst |
|
February 8, 2012, 10:22 |
|
#3 |
Senior Member
hamid
Join Date: Nov 2010
Posts: 185
Rep Power: 15 |
thanks for reply,
but convergence is a sort of self-defined object, if i have a reasonable-defined-residual convergence and the results dont change with the finer mesh (OK grid independence study), then can i rely on my mesh regardless of how much the skewness is, another question, how can i fix skewness issue except fining the mesh? thank u |
|
February 8, 2012, 10:24 |
|
#4 |
Senior Member
hamid
Join Date: Nov 2010
Posts: 185
Rep Power: 15 |
also another question plz,
how do u measure skewness in icem, for example by fluent for my model it says .65 but in icem it says something else... |
|
February 8, 2012, 10:26 |
|
#5 |
Super Moderator
|
If results are validated then no problem for your particular case. However we cannot take this as general guideline.
Refinment and skewness are two different things. You can reduce skewness without refining the grid, you just need to locate the bad cell and see how you can fix it. |
|
February 8, 2012, 10:30 |
|
#6 | |
Super Moderator
|
Quote:
|
||
February 8, 2012, 10:43 |
|
#7 |
Senior Member
hamid
Join Date: Nov 2010
Posts: 185
Rep Power: 15 |
if the questions dont bother u, i would like to ask how do u find the bad cells in icem and more important, how do u fix them,
are there any general guide lines or "what to do" to fix the problem thanks again |
|
February 8, 2012, 10:46 |
|
#8 |
Super Moderator
|
Check quality and click on bar of quality histogram with min quality, bad elements shall be highlighted. For unstructured mesh are many options to rectify them go to Edit mesh. In Hexa you need to change the blocking, adjust vertices, few more splits and may be some o-grid
Last edited by Far; February 8, 2012 at 11:10. |
|
|
|