Hmm. I’m trying to create a network srf from 3 continous longitudinal curves running in one direction, and a series of interpcrv s joining them in the perpendicular direction. I have confirmed intersections in all locations, and have experimented with the networksrf tolerances and “loose v position” radio buttons. Nothing seems to fix it. Any thoughts?
Hi Brad- please post the curves- thanks.
curves.3dm (1.1 MB)
I’m blaming the cyan curve- that is made of up of a zillion segments and has obvious kinks in it - I’d clean up that curve and see what happens. Note NetworkSrf cannot make kinked surfaces and will always be inaccurate at kinks.
Hm- still misses pretty badly with a cleaner curve there. Still poking…
Thanks Pascal! I rebuilt (300 pts, degree 3) but the condition mostly persists in the same location . . .
ok…it should not matter but apparently it does - your file tolerance is .01 - set that to .0001 and the NetworkSrf tols to .01 and .001 - then even your bad old curve make something decent. I’ll see what I can figure out about file tolerance coming into it, if that is happening.
Excellent! Thanks so much!
I was mistaken, I think - just set the Network tols tighter - .0001 for interior curves. Does that work better?
Much, much tighter. Thanks for all of your help!