Equilateral trianglar mesh with kangaroo

Hello,

I am designer attempting to apply an equilateral triangular grid onto a mesh shell (walls/ceiling).

If the overall shape gets modified as a result of the the sequence that is fine but I would like to preserve the general form.

I have been experimenting with Kangaroo the last few days and am at a bit of a loss.

Unfortunately most of the GH/Kangaroo scripts are for older versions and therefore are missing elements when I import them.

Any assistance would be greatly appreciated!

Thanks for your time.

Sean

Attachments:

Hall_ST.3dm, 130 KB

Replies to This Discussion

Permalink Reply by Moonbeast on May 25, 2016 at 9:38am

Permalink Reply by Moonbeast on May 25, 2016 at 9:39am

try using mesh machine

Permalink Reply by SEAN TITTLE on May 25, 2016 at 9:46am

Thanks Moonbeast.

I will give it a try. You wouldn't happen to know a working GH script for this, do you?

Thanks again for your speedy response!

...actually, this looks promising

http://storage.ning.com/topology/rest/1.0/file/get/2769000453?profi...

from

http://www.grasshopper3d.com/forum/topics/equilateral-trianlular-pa...

Permalink Reply by Arthur Lara on May 25, 2016 at 10:09am

Sean

Try the newts Kangaroo2

At that time of the year the new Kangaroos are
leaving the field ... have patience!

Greatings

Art

Attachments:

mesh_relax.gh, 19 KB

Permalink Reply by SEAN TITTLE on May 25, 2016 at 10:43am

Thanks Art Agent!

Permalink Reply by Joshua Jordan on May 25, 2016 at 10:41am

Sean,

FYI, an "equilateral triangle grid" (i.e. triangles of all 60-60-60 degrees) is not possible to tile 3D over surfaces that do not conform to a very rigid set of properties.   Even the triangles you see in a geodesic dome are not equilateral (when flattened into cartesian coordinates).

Are you looking for a mesh that *looks* equilateral?   In other words, a meshing of triangles that comes as close as possible to having equally sized and angled triangles?   Be aware that over complex surfaces like yours, the triangles will (almost certainly) not be equilateral and each one will (almost certainly) be unique.

If that's the case, meshmachine is your best and quickest bet.   It will allow you to make a mesh that conforms to your starting geomtery, allow you to pick points and edges to constrain it to, and allow you to pick a target size for each one of your triangles.

If you're having trouble getting it running, try doing a fresh install of Kangaroo. Make sure you right click "properties" on the files it installs into your Grasshopper libraries folder and check "unlock".

Permalink Reply by SEAN TITTLE on May 25, 2016 at 11:08am

Thanks Joshua,

There are limitations, for sure, but there are ways of using equilateral triangles (or nearly equilateral triangles) to conform to a shape (https://vimeo.com/64244306).

As you said, Meshmachine looks like the best bet. I am attempting it now.

Thanks again!

Permalink Reply by Joshua Jordan on May 25, 2016 at 2:16pm

Good point!

Although I do think that the pliability of that tiled surface comes from the extra degrees of freedom that you get from the tiny gaps in between pieces, and the tiny stretchability of the laminating fabric.

I only know this, because I've tried it. And it is awesome.

Permalink Reply by Arthur Lara on May 25, 2016 at 2:30pm

Joshua & Sean

A grid could be hexagonal take a look

https://youtu.be/E3TI5HFsxGc?t=34

Cool the Mash Machine component!

Thanx

Attachments:

mesh_relax1.3dm, 94 KB
mesh_relax.1gh.gh, 12 KB

Permalink Reply by Joshua Jordan on May 26, 2016 at 7:46am

Meshmachine is certainly one of the coolest, most powerful off-the-shelf Grasshopper tools Mr. Piker has graced us with in the last 2 years.

RE: the hexagonal grid pavilion-- This geometry is in fact taken directly from a triangle mesh like you would produce in meshmachine. If you also use weaverbird, there is a 'Mesh Dual' component that will give you those shapes

The usual problems (at least for construction) that arise with a 4,5,6+-gon tiling is that while a triangle is necessarily planar, the higher polygons are in most cases not planar when tiling more complex surfaces.

What is required is ANOTHER rationalization step as outlined in this paper:

http://research.microsoft.com/en-us/um/people/yangliu/publication/p...

Which is why you see the convex hexagons in one part of the pavilion, and the bowtie-shapes on another. Very cool.

Permalink Reply by Arthur Lara on May 26, 2016 at 8:26am

Joshua

I´m busy with a octogonal mesh...

trying to rotate the octogonal mesh to built a star inside...

to do somethig like a Ron Resh mesh!

The idea came from these Wang & Liu paper

http://research.microsoft.com/en-us/UM/people/yangliu/publication/A...

I´ll try your tips

thanx a lot a long road is open...

Coll

Arthur