Contour Section, Project bug - trimmed surface

Here is a somewhat strange bug. One area of this trimmed surface does not respond to Contour, Section or Project. The remainder of the surface is okay. See attached file. The red line is perpendicular to the contour planes. I also created a line in the location of the missing contour in the front view, and tried to Project it onto the surface - nothing. Also tried using Section in the same location in the front view - nothing.

However the untrimmed version works properly.

Contour Problem.3dm (109.9 KB)

Version 5 SR11 64-bit
(5.11.50226.17195, 2/26/2015)
Commercial

AMD graphics.

Hi David - thanks, I see that - for now, make a copy of the surface and InsertKnot in V with Automatic once or twice and Contour that.

-Pascal

I’ve seen imports from Maxsurf behave in this manner. For some reason it is OK to stack control points in Maxsurf to create chines and such but when imported to Rhino that technique can cause a problem. There is a term that refers to just that, ‘degenerate surface’ I think it is. That is usually referring to points forming a ‘singularity’ but think it can apply to the chines. I’ve split Maxsurf hull surfs along the chine isocurve as a workaround.

The surface was created in Rhino and as far as I know does not have any multiple knots, singularities, etc.

Solved the problem using RebuildEdges on the trimmed surface. No need to use InsertKnot. Turned on control points for the original trimmed surface and the one with rebuilt edges and there are no obvious differences.

Added: Relaxing the absolute tolerance from 0.001 to 0.002 also solved the problem with the original surface. RebuildEdges moved the surface by up to 0.0045 so that change in the surface may have solved the problem.

Good catch - retrimming also helps. I’ll add that to the bug report.

http://mcneel.myjetbrains.com/youtrack/issue/RH-30198

thanks,

-Pascal

@pascal You may have missed from my edit of a post above: [quote=“davidcockey, post:6, topic:19066”]
Relaxing the absolute tolerance from 0.001 to 0.002 also solved the problem with the original surface. RebuildEdges moved the surface by up to 0.0045 so that change in the surface may have solved the problem.
[/quote]

@davidcockey- do you remember how the surface was made?

thanks,

-Pascal

@pascal I finished that project almost two months ago. The surface was probably created with NetworkSrf, though Patch with a “starting surface” may have been used. It also may have been rebuilt, had control points deleted, or otherwise modified after being created. The example I posted was a trimmed and shrunk version of a larger surface which had the same problem.

OK- thanks - the developer was wondering…

-Pascal