[W.T.S]

Dear Everybody， I'm now meshing a volume which has pointed corner as shown below, the skewness of tens of elements will always be higher than 0.97 after meshing first line then face then volume.No matter how i move the points on the face manually, it can not show a good quality. a guy told me since there're two connected lines ending up to a tangent profile, no way the quality of this face will be improved too much. Is that true?
Another problem is, after moving grids on the face, the equisize skewness of the face will be improved from 0.983 to 0.968, so gambit shows no error or warning. but i import the mesh file to the fluent and smooth and use case check, it told me the skewness exceed 0.98 in 5 elements thread id is 2. That's why?Haven't I lower the skewness to 0.96?
I'm waiting for your kind helps.



2011-05-19_101308.jpg

2011-05-19_101216.jpg

2011-05-19_101402.jpg

[-mAx-]

avoid having sharp angle.
In your case, I think you need to delete it.
Split your geometry short before the sharp angle, and delete the volume containing the sharp angle.
Sans titre.png

[W.T.S]

Originally Posted by -mAx-

avoid having sharp angle.
In your case, I think you need to delete it.
Split your geometry short before the sharp angle, and delete the volume containing the sharp angle.
Attachment 7775

Thank you very much! But will it affects the results too much since the model will not be the same one any longer?

[-mAx-]

I don't think it will affect the solution, since this is not an high velocity or pressure-gradient.
Anyway you won't be able to get an acceptable mesh if you don't fix it

[W.T.S]

Originally Posted by -mAx-

I don't think it will affect the solution, since this is not an high velocity or pressure-gradient.
Anyway you won't be able to get an acceptable mesh if you don't fix it

Thanks a lot for your quick reply and guidance. I'll try fixing it with this method. thank you again, -mAx-!

[W.T.S]

Originally Posted by W.T.S

Thanks a lot for your quick reply and guidance. I'll try fixing it with this method. thank you again, -mAx-!

what if i use stairstep function in hex meshing?

[W.T.S]

Originally Posted by -mAx-

I don't think it will affect the solution, since this is not an high velocity or pressure-gradient.
Anyway you won't be able to get an acceptable mesh if you don't fix it

what if i use stairstep function in hex meshing?

[-mAx-]

if the topology remains, the mesh has to follow it( especially at boundary egde/surface).
And it is indepedant from the kind of mesh you will use
In other words you will still have problem at the corner

[W.T.S]

Originally Posted by -mAx-

if the topology remains, the mesh has to follow it( especially at boundary egde/surface).
And it is indepedant from the kind of mesh you will use
In other words you will still have problem at the corner

thankyou!although i'm afraid if i cut the volume, the solution will be changed because it is connected to another entity. the entity i showed you is the inlet of a hydrocyclone. as shown below.thank you all the same,-Max-!You've helped me a lot!2011-05-25_213202.jpg

[-mAx-]

merge it with the volume, which is tangential to sharp angle.

[W.T.S]

Originally Posted by -mAx-

merge it with the volume, which is tangential to sharp angle.

well, merging them doesn't help anything, the sharp corner can't be avoided.2011-05-25_220245.jpg

[-mAx-]

For instance you can do an extrusion of your volume 8 in x-direction (<0).
More in details: sweep surface.6 with (-0.1 0 0).
Then unite volume 8 with the new one.
And you don't have your sharp corner

[W.T.S]

Originally Posted by -mAx-

For instance you can do an extrusion of your volume 8 in x-direction (<0).
More in details: sweep surface.6 with (-0.1 0 0).
Then unite volume 8 with the new one.
And you don't have your sharp corner

I know what you mean, thank you very much! Hope it will not affect the solution. You've helped me so much indeed.