Bug: Contour creates corrupt closed curve

Hello!
Can you please open this file…
test contour.zip (73.0 KB)

… and select this curve:
image

‘What’ says it’s closed, valid, and has 20 segments, but the ‘Area’ command does not compute.
Interestingly, just by moving the curve, or creating a copy, it becomes valid! ‘What’ will then say it has 16 segments.

Thanks for checking!

Please use normal .zip, many people cannot open .7z without downloading an external app.

Without opening the file, I suspect you might have a micro sliver surface in the object you’re slicing which will translate to invalid microsegments in the contour curve(s).

Fixed!

Quite possibly. Yet what’s interesting is that just moving the curve obviously fixes the problem. Strange.

With your problem curve selected, run the command _IntersectSelf

It will report “Found 4 self-intersections.” and mark these locations:

Also, if you look at the output from the _What command you will see that 4 of the line segments show “line length = 0.0000”.

You can fix your curve by deleting one of the 2 edit points at each of these 4 locations.

Strange that you can also fix it by moving or copying it - I don’t understand that.

-Kevin

1 Like

_SimplifyCrv
will solve your problem
just run it after _contour to get proper polylines

(seams like _contour is designed to output polyCurves, one segment per Surface - and does not focus on outputting something special if everything is planar. But of course it would be nice if the output is filtered for small zero-length line-segments, therefor you might call it bug - @pascal has to decide this)

kind regards -tom

1 Like

My guess is anything that replaces the object also does some cleanup - Explode/Join also sorts it out.

I’ll see if we can get this fixed, thanks for the report. RH-67779 Contour messy output
(If I contour this object, the one at that level is in two pieces)

-Pascal

1 Like