Grasshopper

algorithmic modeling for Rhino

# Kulpa Gridshell challenge

hi everyone,

I'm a french Arch student and for one of my school class me and my friends have to study the project from Ville Hara and Anu Puustinen, named Kupla, in Helsinki

We're trying to create the same kind of script used to build this structure. The idea is to be able to have full control of the mesh and the grid nodes needs to be align on the same line in order to support the different levels of the building.

we tryed to create a loft from 3 curves and than appli geodesic curves from top to bottom with a shift item, but it didnt work well.

Any help is greatly appreciated :)

Views: 356

Attachments:

### Replies to This Discussion

You need to internalize your curves (one at a time):

Hello

the solution is not as simple. There was certainly optimization. I looked a bit to that and it appeared that point are not linearly placed on curve so "divide curve" is not optimal except for revolution shapes. For non symetric shapes you will have to find optimal...

thanks for the reply, do you know any GH plugin or tool or exemple that could help me achieve this ?

Apparently physical form-finding with a model was used to arrive at this geometry:

When the drawing of the irregular form proved difficult Hara moulded a plastoline model. Digital images of the model then functioned as a basis for the AutoCAD drawings.

Using the level drawings he constructed a 3D-computer model onto which he “taped” the curved battens to form the grid shell.

from here

Also interesting:

To test if  the pre-bent battens, that form the  basket-like grid shell, tolerated bending and twisting Hara laminated full-scale pieces of the battens. When the laminated timber resisted twisting he resorted to steaming, a traditional method in boat building.

So i sliced the model in contours and divided the  curves in points, How can i connect the first point of the first curve to the 2nd point of a the 2nd curves, the 2nd point of the 2nd curve to the 3rd point of the 3rd curve... ?

Attachments:

Surely too complex but it seems to work. Perphas a duplicate curve somewhere.

Attachments:

thanks a lot !!!!

• View All