Grasshopper

algorithmic modeling for Rhino

Non-twisted box or modules- applying to surface

Hi all,

I am also very new on Grasshopper. I am trying to apply basic modules on a curvature surface with 

1- dividing surface to boxes

2- applying my geometry to this boxes

3- Box Morph

The problem is I am going to build this model and we have just 3 axes cnc so we can not produce some bending modules. So I want to orient my geometry without any bending or twisting process. So the module should be flat in all areas of my surface. 

What should I do in order to achieve this? It is very urgent situation so I need help..

Views: 2457

Attachments:

Replies to This Discussion

You'll still have an issue with edge angles being merely square from a simple CNC cutting of planar stock, but Kangaroo can indeed planarize polygons:

https://youtu.be/MsbyfC2usUk

http://www.grasshopper3d.com/forum/topics/planar-hexagons-kangaroo

http://www.grasshopper3d.com/forum/topics/planar-hexagons-and-kangaroo

You didn't internalize data for this surface, so I can't view your output:

If I use a simple point edited surface, I get good output though. Your strategy of making a 3D object first ignores that your +Z CNC machine won't be able to cute many of the odd joint angles, so screw it, just use square edges and fill the joints with thick glue or something. I'll switch to starting with flat faces via Kangaroo and making those into solids later.

With slight modification, I used the same morph boxes, but merely morphed the outer hexagon curves:

Now separately I can planarize just the hexagons with Kangaroo, can't I?

Well, the square openings screw it up since they are not being dealt with:

Let's try flattening those too by manually adding them in Rhino, to a smaller portion of the overall array:

Something about not having a uniform system makes the indeed resulting planar polygons freak out, or else the rhomboids don't play nicely topologically compared to a hexagonal system. They are indeed all flat now though. For this, I even simplified the definition considerably too, since I was getting too many points on the polygons:

That's as far as I can get tonight with this system.

It may still be breaking the polygons wrong, for some reason. Too many damn points on them instead of natural breaks. I think it's because the morphing is giving curves instead of straight lines, so a different system for recreating your system morphed may be required. Yup, that's one of the core things to change. Too many break points when I make a polyline to feed Kangaroo.

Attachments:

If I manually rebuild the hexagons in Rhino as strict polylines between vertices, same problem:

If I also add in the "squares" as polyline polygons:

Nope, I'll have to add some angle retention goal too, since it's just being crazy.

Also, angle retention, even as a tiny force, flattens out the entire 3D curved surface!

One angle has been left out of each polygon due to my quick fix but the effect is pronounced in losing the original form, so I have to also add a very strong pull to the original surface. I must use a mesh, it seems, for Kangaroo2, but even a fine mesh with even a tiny force, gives chaos:

That's not local point retention enough, since they can slide around on the surface. Let's just add a goal to keep the points where they started:

BINGO!

...and I can remove my incomplete (misses last angle) angle force:

However, I must set the flatting force higher in order for the resulting curves to convert to surfaces directly by running the curves into a surface node, and that distorts the hexagons a bit more:

Let's add the angle goal back then. That's has little actual effect, since the point position retention already handle that.

Now let's remove the "squares":

Same thing, basically. Obviously with a finer grid of hexagons, there would be much less crazy distortion. As you tweak the force values some of the flat surfaces will no longer convert and so will disappear, as useful feedback.

Now you have the issue of thick plate dihedral angles between them and how that will cause fighting for space as they curve from concave over to convex so the side that hits the adjacent tiles switches from bottom side to top side or one top and the other bottom.

Attachments:

Hey Nik, great work!

I tried it with nearly the same surface. At the beginning the Kangaroo is working, but if i change any slider or panel the solver gets an error. If i slide back to the beginners input it doesnt work too! Do you now whats going on there?

greetings Hendrik

RSS

About

Translate

Search

Photos

  • Add Photos
  • View All

Videos

  • Add Videos
  • View All

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service