Help with a gradation of mesh reduction (or variable face density)

This is my first thread, and I hope I won't be racked across the coals.

At the start of this semester, I found an opportunity to use grasshopper

for an architecture project, Since then I've been trying to learn the software

on my own. That being said, my experience is at or below a novice level. While

I'm eager to learn this software, and hope to become proficient at it, please

understand, that I'm far from that goal. I'm throwing this out there in the hope that

someone will help me out, and to continue learning

If you have any insight, I am extremely thankful

In general (referencing the the attachment) I want to facet a form according to some point attractors, where one side is highly faceted and the other is not.

For my project, I'm not concerned on the facets being congruent shapes.

http://spacesymmetrystructure.files.wordpress.com/2012/09/varymesh.jpg" width="200" height="150"/>

My initial Thought on how to go about this:

- Reference surface geometry

- populate geometry with points (randomized)

Points would be populated to the highest desired resolution

-If I could then have some sort of gradient selection of points

-to my knowledge there isn't this type of function

-from points, create connecting lines,

-maybe with something like delaunay edges under the mesh tab.

-Create planes from lines

(I'm having trouble figuring out how to connect to the nearest vertices. I thought I might

have been on the right path with karambas: nearest neighbor tool.

_________________________

The other thought I had, that seemed to be cheating, was to reduce the meshes in rhino to

the lowest resolution, from there I was able to use Weaverbird's loop subdivision to have

somewhat of the reverse effect of creating more facets, and in essence smoothing it out.

I think one of the things i'm having trouble understanding is, if a function has a number as input

(say, how many times a surface is subdivided) can this number be variable?

My understanding is that if a number is the input, then it is asking for one number, where as

a mathematical function would yield multiple numbers.

Let me know if you can help out

Best,

Charles

Replies to This Discussion

Permalink Reply by cfox28 on April 3, 2014 at 3:47pm

so with some help, I've come up with a way to create a gradient fall off of points from

-referencing the "y" data of each points

-ordering the indices into sub lists

-a percentage of points is then deleted from each list

-the result is a gradient fall off of points initially generated by populating geometry

Permalink Reply by cfox28 on April 3, 2014 at 3:50pm

Permalink Reply by Daniel Piker on April 3, 2014 at 3:50pm

Sorry I missed this discussion earlier.

I actually posted a version of the script I used to generate that image here:

http://www.grasshopper3d.com/group/plankton/forum/topics/dynamic-re...

It can be tricky getting all the libraries set up correctly - feel free to ask if you have any trouble

Permalink Reply by cfox28 on April 3, 2014 at 3:53pm

What I want to do now is connect these points based on how they would have been meshed from the original surface.

Is there a way that I can reference the geometry to create a mesh?

Attachments:

script test.3dm, 4.5 MB
gradient points4-3.gh, 12 KB

Permalink Reply by Michael Pryor on April 3, 2014 at 7:31pm

Something like this? requires the mesh analysis and utility plug in from [uto] for the last component. http://utos.blogspot.com/2011/07/mesh-analysis-and-utility-componen...

Attachments:

surface mesh.3dm, 2.4 MB
surface mesh.gh, 15 KB

Permalink Reply by cfox28 on April 3, 2014 at 10:36pm

that is exactly what i'm looking for! Thank you! i'll have to wait till I get permission from uto to download the new utility plug-in, but otherwise it works great. Thanks alot

Permalink Reply by Charles Fox on April 6, 2014 at 4:18pm

This worked great! one thing I noticed is that this script works for one surface and the points on it, if you have multiple surfaces and a cloud of points it can be difficult to reference the right things.

how I solved this, was by using the populate geometry at the beginning of my script and referenced only one geometry. this gave me the points of the geometry before they were deleted in a gradient pattern

after having baked these points (and given that the gradient points from my script were baked) I then went into rhino, and selected duplicates (SelDup) this selection can then be referenced in the script above, and with reference to the same geometry it works well..

I know there is probably a way to do this in grasshopper, but for time sake i wasn't to concerned with figuring it out right away.

please note: by doing this step you are now making the parametric quality of the model, only semi parametric.

the other problem i ran into, was with poly surfaces. for this i made the polysurface a mesh, ran my script, in rhino did meshtonurbs selected new surfaces and referenced them. this seemed to work as well.

thanks again for helping out.

and thanks troy (the professor that helped me) for your help as well. I have certainly learned alot this semester, and the thanks can be given to people like you, who help out and contribute to the forum, as well as those generous enough to create online tutorials. can't tell you how much I appreciate the help and resources out there

RSS