CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Orthogonality Problem (http://www.cfd-online.com/Forums/fluent/108659-orthogonality-problem.html)

sadjad.s October 29, 2012 15:09

Orthogonality Problem
 
hi everybody,
i made a grid which due to Complexity, and making small elements with Y+=1 near wall, quality of grid is as follows:
maximum skewness: 0.87
orthogonal quality: 4.0e-4

i've heard that there is no serious problem with skewness below 0.9 in 3d.(although help suggests that maintain maximum skewness below 0.8).
but i couldn't find any suggestion for orthogonal quality!
my question is can i trust results?

davidwilcox October 30, 2012 03:19

Hi Sadjad,
There is a scheme in the Solver called Piso. It allows you to put it a skewness correction factor. You might want to play with that. Pertaining to your question as whether you can trust the results, I would say this is CFD. In other words, the results can only be validated if you compare them with experimental data. The skewness affects the Jacobian matrix ie the calculation of your gradients. So you might wanna play with the correction factor until you get reasonable results. If you're interested in things like the shear layer, it's gonna take a lil bit of an effort. Hope this helps. :)

sadjad.s October 31, 2012 14:31

thanks mate,
this is it! actually that's why some call it Colorful fluid dynamics, because you can not trust results 100% without validating.
but i already wonder how much those errors made by orthogonality problem (which was not a criteria in fluent 6.3) can affect results.
and also skewness is within safe range.
by the way i'm doing high reynolds modelling, and i have to use Density base, where there is no possibility of using PISO scheme.


All times are GMT -4. The time now is 02:16.