Path mapping question. relating coincident edges.

Replies to This Discussion

Permalink Reply by dani bas on April 28, 2015 at 10:34am

any idea?

Permalink Reply by Riccardo Majewski on April 28, 2015 at 1:43pm

Don't lose faith!

Good post, good description, interesting challenge.

I'll be back :D

Permalink Reply by ng5 Alex on April 28, 2015 at 2:44pm

challenging indeed. i am at the point where the curves are ready for loft but cant get them to orient correctly

Permalink Reply by Riccardo Majewski on April 28, 2015 at 3:39pm

Almost done... did you want something like this?

(I'm trying to semplify the definition, too big and abstract to be "useful")

Permalink Reply by ng5 Alex on April 28, 2015 at 3:49pm

ok this could get somewhere

two problems that derive from same origin. both list of curves (to be merged and lofted) need their direction to be sorted, for the first list clockwise and the second counterclockwise. my curve orientation solution fails at some offset values, propably flips the perp frames, and also the data structure due the sorting (green group) makes the data produce double surfaces for each loft.

cheers

alex

Attachments:

• questionpathmapping2.gh, 30 KB

Permalink Reply by Riccardo Majewski on April 28, 2015 at 4:58pm

I joined all voronoi faces to get a single brep.

After that all faces would have the same normal direction and brep topology comes in hand.

As you said, clockwise and counterclockwise.... if you debrep the total you will automatically get the first face boundary of an edge as normal, and the second flipped. (as every face boundary after a debrep is countercockwise)

Btw your definition is smarter than mine.

I've worked to create smooth joints between dani's lofts... but somehow I misunderstood :P

filling remaining triangles is more tricky than i thought...

I probably gone out of the way :D 

Attachments:

• questionpathmapping_V2.gh, 38 KB

Permalink Reply by ng5 Alex on April 28, 2015 at 5:12pm

nice!

thank you for your comment. i thought i misunderstood, working with planars.

yours is different beast, dealing with cases that need to have a "continuity"

but i cannot understand from dani's description.

cheers

alex

Permalink Reply by dani bas on April 29, 2015 at 1:29am

WOW guys this really is a great feedback! thank you.

this example is focused on a curve folding exercise, so the triangle surfaces in each node are suposed to be planar.

today i will give it a try to all your solutions and will post results.

Permalink Reply by Pieter Segeren on April 29, 2015 at 8:08am

Hi Dani - guys, here is my attempt at tackling this challenge. Maybe you can combine bits from each definition to suit your goal dani...
I must be missing something, besides sleep, because it has become quite a nasty monster... But this is where I stopped, because I need to be doing other things at the moment.
Oh, I used Aplha Shape component by Mateusz Zwierzycki in an attempt to make an alternate approach. Hope it somehow contributes.

Attachments:

• screencapture.jpg, 293 KB
• PathMappingQuestion_PSG.gh, 50 KB

Permalink Reply by ng5 Alex on May 3, 2015 at 12:26pm

+1