[Polarbear]

Hi all,
I have a little problem that hopefully is easy to solve. I'm trying to block mesh small hemisphere connected to a cylinder which in turn is connected to a larger hemisphere.

I start with initiating the blocking and see that some vertices on the side of the smaller hemisphere associates themselves with the larger hemisphere. I then make two splits plus an o-grid split and associate the faces to the surfaces to get right associations.
All that remains are the four corner vertices on the left side.
I've tried to associate them to a surface but nothing happens after selecting them and pressing done. I can dissociate them from the geometry but I have no luck in then associating them to the small hemisphere surface.

Any ideas folks?

Ps. When the hemispheres are of equal size, then the initial blocking associates the corner vertices correctly and I can make o-grid splits and so on without any problem(even though the ogrid becomes slightly assymetrical). Ds

[Polarbear]

The ICEM files attached

[siw]

I have found on some models of mine that vertices are associated to the surfaces that I don't want. However, when I then move the vertices closer the required surface ICEM re-associates them to the correct surface. So what happens if you just move those vertices towards the intended surface?

[Polarbear]

Hi siw,
Thanks for your suggestion, the corner vertices snaps onto the small hemisphere surface when I move the corner vertices. However, since I do it by hand they become skewed and mixed up, they don't become as beautiful as the edges on the right hemisphere when using autosnap. Any ideas?

[Polarbear]

Problem solved! I associated the faces to the surface by part and the associations became correct. Thanks!