represent loftet curves by a spiral

hi folks,

first of all I hope my request gets the same result like sooo many others - amazing forum.

here the thing:

I have a series of lofted curves and want to wrap a spiral around it, as easy as it seems, but...

(I know there's already a a request like this with a pretty good result - unf. this result is not working for me.)

So I took the definition of this former discussion, but when it comes to the part of pulling the points, the component is not respecting all of the faces of the loft and the result is super unprecise and vage.

I can switch from xz-plane to yz-plane - there's always a part that is not pulling the points. but anyway still super vage.

maybe there's another way instead of pulling points and interpolating them.

for me it doesnt matter how - my goal is to represent the loft by a spiral (the exact loft)

maybe someone can have a look - I would highly appreciate it.

thanks in advance, ©.

Attachments:

Lofted Surface t.b. repr. by Spiral.3dm, 864 KB
Lofted Surface t.b. repr. by Spiral.gh, 20 KB

Replies to This Discussion

Permalink Reply by Daniel González Abalde on July 26, 2015 at 9:23am

This user photo deserves a good answer! X)
I would do it by this way, tell me if what you wanted.

Attachments:

LoftedSurfacet.b.repr.bySpiralDGA.gh, 21 KB

Permalink Reply by Costa on July 26, 2015 at 10:28am

thanx Daniel for your reply,

what I want to achieve is a perfect alignment of the spiral to the surface..

the result I have is a start, but there's is no alignment yet..

Permalink Reply by Daniel González Abalde on July 26, 2015 at 11:04am

Ups, here is fixed.

Attachments:

LoftedSurfacet.b.repr.bySpiralDGA.gh, 23 KB

Permalink Reply by Laurent DELRIEU on July 26, 2015 at 10:30am

Here is another way of doing, loft surface has a UV of 8901 in U and 200 in V. So helix coordinates is put as UV coordinates on curve. Constant progression on U, and constant on V but nturn times.

Attachments:

LoftedSurfacet.b.repr.bySpiral_LD.gh, 15 KB

Permalink Reply by Pieter Segeren on July 26, 2015 at 1:37pm

Hi Costa, hi guys,
here's my approach at this, using the CurveOnSurface component.

Attachments:

SpiralOnLoft_PSG.gh, 24 KB

Permalink Reply by Costa on July 26, 2015 at 2:30pm

hey daniel, laurent and pieter,

tested all your solutions and its precisely what I was hoping for - all of them, thank you guys very very much. confirms the impression I have bout the forum.. great place.

thanx, ©.

Permalink Reply by Joseph Oster on July 26, 2015 at 2:44pm

Pieter, what is going on there?

I see 369 points between X=0 and X=1; their Y values range from 0 to 46, yet the X-dimension appears to be far greater than Y?

Then it seems that 'Curve On Surface (CrvSrf)' is "interpreting" the X,Y values of the points as 'uv' (ignoring Z?) and, because the lofted surface has been reparametrized, the Y values greater than one are being wrapped off the "far edge" and back onto the "near edge" (as many times as needed since the surface is effectively only one unit in each direction).

Is that it? Why do the Y point values (max = 46) appear to be smaller than the X point values (max = 1)?

Permalink Reply by Pieter Segeren on July 26, 2015 at 3:43pm

This autoconversion/limiting behaviour was new to me too, but as you said, the uv input is apparently limited to 0 to 1 range when you reparameterize the surface (v=v\1). I thought it was clever of David to tackle it like this, and quite handy to know:)

Permalink Reply by Pieter Segeren on July 26, 2015 at 4:48pm

Ehr, reading it again, I'd think your understanding of it is correct and well explained. Better than I could. Thanks. Maybe David can elaborate on it a bit further ... because
I admit this is a bit of a hacky definition. But hey - I think it works pretty good.
From my point of view, the key was rebuilding the loft to have a surface for a result rather than a brep/polysurface, and of coarse (the real issue of the discussion:) the CurveOnSurface component.

Permalink Reply by Joseph Oster on July 26, 2015 at 5:43pm

Well, I missed the significance of the 'Rebuild (Rbd)' loft option - and I mis-read the layout of the points; the long direction is 0..46, the short direction is 0..1 (duh). Was confused by the fact that they are layed out at ninety degrees to how they get "plotted", due to the 'uv' direction? And the radical difference in scale between the points and the surface. It's a bizarre and elegant solution really, thanks again for sharing it.

P.S. I realize now that Daniel and Laurent are also playing with 'uv' mapping in their solutions... I hadn't paid much attention to this thread because wrapping a spiral around a surface like this didn't make a lot of sense to me, but it's amazing what one can learn from it.

RSS