Is this an accurate result using ApplyCRV?

I’m not sure if I can trust ApplyCRV command.

I am making 3 versions of the same part, all scaled to different sizes, but the features on the part have to match. These features sit on a curved surface, so I have been using ApplyCRV to locate them.

The problem is that the width appears to change when applied on this specific location, and on this specific version. Measuring a curve length, it is .040" narrow. That will be noticeable.

But if I create a new UV curve from this wrong result, the dimensions comes up being correct in 2D.

Any ideas what is going on? Is there a better technique? I suppose I can modify the curves on the UV until the width ends up right…

the workflow
are based on the Surface s UV-parameters - and that means the commands are not meant for numeric (unit) precision, as parameters can be “distributed non-linear” …

see a simple example, black surface with uneven distributed CVs resulting in different “density” of parameters in the surface.
blue curves via _applyCrv

hope at least this explains your problem.
kind regards -tom

Is there another way? Or is this a situation for guess and check?

I also found some comments about UnRollSRF and FlowAlongSRF, but initial tests didn’t look promising for my drawing.

To get an useful answer, I think you would have to explain better what “features on the part have to match” means.

Is it that some other object that has a shape that needs to fit with these features? In other words are you planning to use the curves to cut a hole in the surfaces? Or maybe use the curves on the surface to create a feature that extends outward from the surface?

2023-11-14 09_17_32-2023-11-12 (Large) Molds re-drawn Final Versions.3dm (109 MB) - Rhino 7 Commerci

These are attachment points for straps. The width of the straps stays the same, even though the part’s overall dimensions are being scaled for different sizes.

for me this looks more like projecting
the Curves on the surface with a custom direction / cplane.