UV coordinates on flat surface

Replies to This Discussion

Permalink Reply by David Rutten on February 24, 2014 at 4:37am

Hi quishi,

you'll have to create a surface underneath your pattern. Make sure it's bigger than the pattern, otherwise you cannot project points to UV space losslessly. Then you can use the [Surface Closest Point] component to find the UV values of XYZ points.

--

David Rutten

david@mcneel.com

Permalink Reply by qiushi on February 24, 2014 at 4:51am

Thanks!David!

Actually, to create a correspond surface and a bit bigger than the grids is part of my question!

I find an answer you gave before using "surface from points" to get the correspond surface (which changes when I change the UV count of the grids), but it is smaller than the rotated grids!

Permalink Reply by David Rutten on February 24, 2014 at 5:01am

I'd use the BoundingBox of all the line segments to create the base surface.

In this file I grow the box by 1% in all direction (-0.01 instead of 0, and +1.01 instead of +1), which is actually a redundant step. The boundingbox will be exactly big enough to handle all lines.

--

David Rutten

david@mcneel.com

Attachments:

• boxeval.gh, 50 KB

Permalink Reply by qiushi on February 24, 2014 at 5:14am

Great! Thank you! David!

Permalink Reply by qiushi on February 24, 2014 at 5:38am

Hi David,

How did you get the line segments?

I used the same components as you did, but I didn't get the right results.

Permalink Reply by Danny Boyes on February 24, 2014 at 5:49am

It looks like you have multiple Boxes, where as David's has a Union Box option displayed. Try changing this in the Context Menu of the BBox component.

Permalink Reply by qiushi on February 24, 2014 at 6:24am

I make it with your help!

Thanks! Danny!

Permalink Reply by Daniel Piker on February 24, 2014 at 5:02am

Hi qiushi,

From the pattern there it looks like you might be working with reciprocal structures, in which case you might be interested in this alternative approach: http://www.grasshopper3d.com/group/kangaroo/forum/topics/reciprocal...

Instead of UV mapping the pattern it is generated from the edges of a mesh (which can be quads or triangles)

Permalink Reply by qiushi on February 24, 2014 at 5:19am

Thank you! Daniel!

It will be helpful for me!

In my research, I am going to map the 2d pattern on a 3d double curved surface, finally materialize it with a interlocking connection(which is similar with reciprocal structures). So you study is really helpful for me!

You'd better know this: MIT Master thesis:Design for assembly:a computational approach to construct interlocking wooden frames, 2012.

Permalink Reply by Daniel Piker on February 24, 2014 at 5:46am

Thanks, I didn't actually know that MIT thesis - it looks interesting work.

This paper is also worth looking at (they use a surface mapping approach closer to what you described):

http://vecg.cs.ucl.ac.uk/Projects/SmartGeometry/reciprocal_frame/re...

Permalink Reply by qiushi on February 24, 2014 at 6:25am

Cool!