Loft not behaving as expected

Is it possible to use loft to produce the geometry of B but using the curves of A ?

I’m trying to achieve A, where the individual surfaces of the loft correspond to the input arc segments.

The curves are all arcs, but joining them appears to do something strange with their parameter space.
all curves are similar in that each has the same number of control points, discontinuities, direction and the seams are all aligned.

Is this something to do with irrational weights of the arcs? Does loft convert them to degree 3 or something?
(i am constructing this in grasshopper but the behavior is the same in Rhino)

*edit: added rhino file
arc_lofts.3dm (1.5 MB)

Hi Dharman -

I suppose it could be the kink-splitter being disabled but you haven’t attached the gh or 3dm file…
Since you are doing this in Grasshopper, why not just create the 12 lofts and join these?
-wim

Oops sorry @wim. I forgot to attach and I’m not at the computer. Yes I’m am going with joining after the loft as you suggest, but wanted to understand why the other option was so strange

Yes, this does work for most cases, but i kept getting some errors in my workflow. I traced it to this example here. When lofting this circular profile, the edges overlap and will not join.


sweep_profiles.gh (25.5 KB)

Anyway i think it turns out Sweep1 is what i was really after!

I’ve made yet another discover here.


in this case sweep produces some strange wavy artifacts in the surface


network surface produces a much smoother result which i am very pleased with.

sweep vs networksrf.3dm (1.5 MB)

The Rhino file contains a set of profiles which i am using to create the surface (purple),

and another verification set which the surface should pass through closely (red)

If anyone else comes across this consider using NetworkSrf. It produces the kind of surface I mistakenly expected from Loft

1 Like