Hi all,
I am working on a spiral stair with an oval footrprint. I am looking to create a sweep path for a recessed handrail profile. I am having trouble extending the path from the top of the nosings to the landing while maintaining the curvature of the oval footprint.
I understand this is due to the NURBS curves and control points being outside the footprint which I am struggling to accurately generate.
I have tried a couple of methods, I started by simply extending, finding intersection, then plugging distance back into the extension again, however the end point had missed the footprint.
I am then pulling this point to the curve using this as an approximate start point. (as annotated i was also making a tween point). I then ran into the issue that having 2no points to insert to the beginning of the list of points to create a new curve, depending on whether i interpolated or fed into NURBS, the new curve never exactly matched the original nosings curve.
- Tangents / BiArcs
This all felt quite attractional, so i’ve tried a second method (attached),
I understand it is very difficult with an oval with an ever changing radius to specify to the extend curve how to accurately construct.
I just wonder if someone has a totally different approach. I keep wondering if i could use a sub curve of the original oval, which would be the exact right curvature. But i dont know how to make the control points sit exactly on the nosings and transform this new curve.
I think my approximation is quite close and i can just oversize the void, but if someone has a super clean accurate method i’d be interested to hear it.
Thanks everyone,
Also this topic seems a bit similar, discussing the problem of NURBS control points and constructing curves when points are not on the footprint of the curve.
(I also appreciate I’ve been a bit lazy by repeating groups and not grafting)
260820_Oval Stair Handrail Path Query.gh (72.9 KB)








