Grasshopper

algorithmic modeling for Rhino

Well friends

This is still my very first attempt in GH (it's "part" of the code attached here: http://www.grasshopper3d.com/forum/topics/a-geodetic-exoskeleton-tower).

Using straight Loft and then decomposing the brep in quads and then getting the triangles is rather a very common way of thinking with regard similar solutions.

Problem is that the straight Loft doesn't work as expected (at least to me):

See 2 versions of the code (use Saved Views as well):

with profile rotation (OK):

and without (some profiles are not taken into account)

Spot the Scale values (the 2 lists with Input/Output) used on these profiles (values they appear "distinguishable" enough to me)

Is this a tolerance issue ? or what?

Other than that I can't get the gist of the Graph Mapper (how exactly works in edit mode? it's a mystery to me).

Best, Peter

Views: 2440

Attachments:

Replies to This Discussion

You're asking for a singular ruled surface over the height of these profiles?

As it's failing, I'd suggest it's not geometrically possible.  If you want each level to have a ruled surface, loft between adjacent curves (I didn't adjust or check your model after this, but I think it's the same effect as nominating a tight loft in your model).

Attachments:

Hi Jon

NO I'm not after a singular ruled surface - obviously impossible (thus the straight loft option...meaning a polysurface).

I only say that straight Loft doesn't "use" all the profiles and as a result the polysurface instead of "connecting" all adjacent curves ... omits some of them (doc.tolerance ?? who knows?).

See the very same problem on my latest posts here (v7) : ( http://www.grasshopper3d.com/forum/topics/a-very-strange-pline-actu... ).

See the solution as well (lofting pairs of lines):

But then use the tight option for loft instead of straight (if the loft must pass exactly through the profile curves).  If you read the rhino help maybe it can explain the difference better than me ;-)

No, tight doesn't address the problem, since we need to decompose the thing...anyway...forget it (for the moment) and wait for version v8 that address the puzzle and creates a nice(?) planar glazing array of cables, fasteners and other useless stuff.

PS: one issue brings the other :

There's some components that respect the tree and others that do whatever they want.

For instance imagine 23 line sets that belong to 3 branches (the facade profiles). If you apply Divide curv (say with 7 divisions) you expect to get back 23*7 points in 3 branches....not the case I'm afraid.

However if you apply Eval Curv with just one value for t > you get 23 points in 3 branches.

Moral: more red Russian vodka required (+ a decent cigar)

Forget my comments about components that don't "respect" the tree (blame ouzo with breakfast - NEVER again).

RSS

About

Translate

Search

Videos

  • Add Videos
  • View All

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service