Exoskeleton lines to origin

Replies to This Discussion

Permalink Reply by Mario A. Benavides on January 17, 2013 at 5:00pm

Permalink Reply by Jonny Blackmore on January 23, 2013 at 2:41pm

Having the same origin (0,0,0) issue with Exoskeleton.

Any one find a fix? 

Cheers, 

Jonny

Permalink Reply by Mario A. Benavides on January 23, 2013 at 3:15pm

It has something to do with the node size, the only workaround I've been able to come up with is to Cull mesh faces by area

Permalink Reply by Sharon Lam on June 11, 2015 at 4:15pm

Hi Mario!

Would you be able to share a screencapt or orginal file as to how one can proceed to "cull mesh faces by area"?

Much thanks in advance from a fellow grasshopper user stuck at this same problem

Permalink Reply by Daniel Piker on January 23, 2013 at 3:08pm

Hi guys,

We are aware of the issue, and have some ideas for a more robust approach, but simply haven't had the time recently.

As general pointers, using 3 sided tubes causes more problems than with higher numbers. Very short segments, duplicate or overlapping input, and tight angles also cause problems.

and I think it will always be problematic to apply it recursively as you have done - skeletonizing a wireframe, then skeletonizing the wireframe of the skeleton, and subdividing the result is going to result in meshes with tens of millions of vertices, which will probably push the limits of rhino/gh.

Permalink Reply by Mario A. Benavides on January 23, 2013 at 3:22pm

Ok, I will be watching out for new releases of this component.

Thank you

Mario

Permalink Reply by Jonny Blackmore on January 24, 2013 at 4:59am

Cheers Daniel. 

I was trying to do it on a voronoi wireframe cube so probably to complex. 

Thanks I'll keep an eye out for updates anyway keep up the good work!

Jonny

Permalink Reply by David Stasiuk on January 24, 2013 at 6:28am

Dense 2D and 3D Voronoi diagrams often have extremely small edges embedded on the interior here and there...and that creates problems for exoskeleton.  I've been thinking about trying to implement some means for super small edges to combine into "super nodes" so that they get effectively eliminated before being processed.  Basically what happens is when the hulls get built around the connection nodes, if either the wireframe thickness, the node offset or minimum angle results in a radius greater than any of the lines the node is connected to, the mesh breaks.  So if you can cull out any input lines with too tight angles between or which are shorter than your node size, you should be able to eliminate a lot of these artifacts.