Meshing issue
All  1  2-8

Previous
Next
 From:  OSTexo
7710.2 
Hello,

It looks like it could be an actual geometry difference, here is the deviation analysis between the two parts.

Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  BurrMan
7710.3 In reply to 7710.1 
Hey PaQ,
The red surface is slightly twisted. That's causing the mesher to break up in the "Y" direction and required a "divide larger than" for what it reads as a "Long, slivery, curved" surface. The green one is straight and so the mesher does a very clean, unbroken mesh there.
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
7710.4 In reply to 7710.1 
Hi PaQ, I guess it's just due to the particular way this surface is straight in one direction and has a particular amount of curvature in its other direction, that just meshing by an angle parameter alone makes some pieces of it to be subdivided and another area right next to it to not be subdivided, making for long and skinny polygons sitting right next to some subdivided ones. When you get a situation like that it's pretty easy to get shading glitches.

Like Burr says above if you use the "Divide larger than" meshing parameter, it should help to solve this - put in for example 1.0 or 0.5 in the "Divide larger than" meshing option and then look at the mesh wireframe, you should see that it doesn't have any long skinny triangles spanning the entire length of the surface anymore and should work better.

When you have long skinny triangles, any slight shape changes at the ends of the triangles tends to make for low-poly type shading glitches.

I don't think there's anything really wrong with the red surface, it's just that due to its particular shape you'll get low poly like shading glitches if it gets long skinny polygons being made right next to some subdivided polys, and meshing by the angle parameter alone just happens to give that kind of result.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  PaQ
7710.5 In reply to 7710.4 
Hello there,

Indeed the 'divide larger than' did the trick.
Now I remember having this meshing issue with cones shape, but well, I forget about the solution :)

Thanks for you help guys, love this place.

(My co-worker seems to appreciate Moideling a lot, one converted ... 299 lefts ... )
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  MarkRG (MARKRG0)
7710.6 In reply to 7710.2 
Hi, how was that deviation analysis done?

M
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  BurrMan
7710.7 In reply to 7710.6 
Hey M,
Looks like rhino.
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
 From:  OSTexo
7710.8 
Hello,

It's Rhino using the Autodesk / VSR Shape Modeling Plugin.
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged
 

Reply to All Reply to All

 

 
 
Show messages: All  1  2-8