Contouring surface produces arbitrary curve frame directions

I’m having issues with the contours of a surface. The contours themselves are nice and correct and even uniform in their directionality.
However, the internal frame directionality of the curves changes arbitrarily.

This image should explain the issue clearly. Contours and the frame Y vector.

This would not be an issue if I could use the frame directly (could always align or rotate) - but I’m using the curve as a guide to VisualARQ beam, which uses the curve internal frame directly. This causes the beam to be oriented differently, even though the curves themselves look quite ok.

Question is:
**Is there a way to recreate the curve internal frame direction **
/ or recreate the curve itself to a uniform frame
/ or force the contours somehow to follow uniform twist?

Hey !

Planes on these are somehow almost always flippping, at least in my case too.
If you say the directionnality of the curves are going in the same direction, you could use the Tangent vector out of evaluate curve to rebuild the planes.

Can you link the internalised file pls?

here’s the file. Should have added it from the get go. I added a small fix to it, in order to ensure unidirectionality of the contours.

The rebuild of the planes/frames is not the issue here - that would be easy to achieve. The issue is that I would need the curves themselves to have uniform torsion (if that is the correct term).

Or maybe an update from VisualARQ’s side that would allow to define the beam’s orientation through a vector.

If there is no solution here, then I would just have to use brep beams, created through sweep. But where’s the BIM in that? :slight_smile:

Contour (200.1 KB)

Like this?

Contour (199.1 KB)

Contour (199.3 KB)

Hi, thanks for the effort, but as I mentioned, fixing the planes themselves is not the issue - but fixing the curve.

Isn’t there any functionality to convert a solid(closed brep)to a beam instance in VisualARQ?

No, there is not. I can always tag it in IFC as ifcBeam type, but… I mean there are several workarounds to deal with this, but I would like to get this to work without relying on hacks.

The frame orientation seems to be some inherent property of the curve geometry, as even though I strip the curve to only its points, recreation of the nurbs-curve has the exact same frame direction.
Interestingly, the frame orientation differs if I create it as interpolated nurbs curve.

I found this to be an adequate solution.
I divide the original curve to points within ~100mm intervalls and create interpCrv from them. I get uniform frame direction and the deviation is at maximum 0.007mm, which is quite enough for this scale and accuracy of fabrication.

I’m still lacking the mathematical explanation for this phenomenon, but at least I have now uniform curves.