CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   Parameter for a good meshing (https://www.cfd-online.com/Forums/main/223028-parameter-good-meshing.html)

Aliosha86 December 18, 2019 07:42

Parameter for a good meshing
 
Hi guys,
my name is Alessio. I am struggling at deciding whether a mesh is good or not. I would not like to do a parametric study and I would learn to see if the mesh is looking good or not just looking at parameters such as Aspect Ratio, Jacobian, skewness etc..
Can someone give me rules of thumb to see if I get a good mesh?
Cheers,
Alessio

Stanley90 December 18, 2019 10:46

Quote:

Originally Posted by Aliosha86 (Post 752773)
Hi guys,
my name is Alessio. I am struggling at deciding whether a mesh is good or not. I would not like to do a parametric study and I would learn to see if the mesh is looking good or not just looking at parameters such as Aspect Ratio, Jacobian, skewness etc..
Can someone give me rules of thumb to see if I get a good mesh?
Cheers,
Alessio

Apart from what application and for which CFD solver this mesh is being created, I'd check skewness at first. You can do it easily in ANSYS Meshing for already created mesh. Just create named selection and choose worksheet instead of geometry. Then go to worksheet table -> right click in Action column -> Add -> Entity Type = Mesh Element ->Criterion = Skewness -> Operator = Greater Than or Equal -> Value = e.g. 0.95 -> Generate. You should see then mesh elements with skewness greater than desired value. By worksheet in named selections you can make many other check of mesh e.g. like stairsteps.

LuckyTran December 18, 2019 11:33

1) Resolution (in areas you care about)
2) Skewness (skewness and non-orthogonality cucks all computations)
3) Growth rates / stretch ratio / volume change. (i.e. smoothness)


That's plenty. Most grids don't even satisfy the first criteria.

sbaffini December 20, 2019 14:25

The more regular it is the better. Yet, resolution is really case dependent. Just changing viscosity would change your resolution requirements.


All times are GMT -4. The time now is 08:42.