Abeille's Vault

Hi Everyone,

I am trying to recreate the abeille's vault onto a rotated geodesic sphere in grasshopper as per the work in this paper: http://www.paris-malaquais.archi.fr/brocato/papers/Abgeo.pdf

The image above is my progress so far however I am having problems ensuring each of the stones meets its neighbours cleanly and then how to cut each of the stones so that they don't intersect - essentially a kind of mass boolean.

See the attached script so far. It does use weaverbird to generate the initial icosahedron to project for the geodesic sphere.

The model links to work on reciprocal frame structures which can be seen for those that are interested on another discussion topic: http://www.grasshopper3d.com/forum/topics/reciprocal-systems

If anyone can help me or is working on anything similar and has something to share, it would be great to hear from you,

Regards,

Mike

Attachments:

Abeille's Vault 02.gh, 20 KB

Replies to This Discussion

Permalink Reply by Michael Clarke on October 10, 2012 at 4:57pm

To clarify:

After defining the trapezoids, the stones need to be cut so that hey rest on top of each other and not intersecting each other as currently. This means the flat top of each trapezoid should meet the next at a chamfer but touching and the base of each should be cut away.

My main problem is calculating intersection events for meshes/planes/curves between items that are all in the same list. Originally I thought that this was because the 'nexor' axis / centre line for each stone would never intersect it's neighbours due to the rotation but I can't get it to work with planes intersecting either.

Here is an image from inside the sphere showing the problem with a transparent mesh:

Thanks for any help,

Mike

Permalink Reply by Arthur Mamou-Mani on October 10, 2012 at 5:13pm

Hi Mike,

Wow that's a tough one! A "multiple brep intersection" component might really help there.

It seems like you need to change the design of the component depending on its rotation angle. This will only work on a surface with regular curvature where components are repetitive though.

A better solution would be to write a separate script/hoopsnake that finds the 4 closest neighbour of each breps and creates a boolean difference with it? This could then be applied to any surface.

hope that helps somehow.

Arthur

RSS