Help Please: Reversing loft direction/ filling gaps in pattern

Hey guys, Im a long time follower on these boards but have never posted myself.

I am designing a modular screen prototype and am having trouble getting my grasshopper definition to loft the way I want. As you can see in the files attached I have created lofts that are essentially trapezoids on an angle. However, you'll notice there are gaps between each trapezoid. I need to fill those gaps with an identical loft as its adjacent counterpart (or surface created somehow) so that from a top view it will look pretty much like an overlapped pattern of hexagons.

I have tried reversing my data and lists, shifting lists and paths, but no matter what I keep getting the same loft. I also tried messing around with the path mapper, but Im pretty lost with that thing.

**For some reason I cant attach files right now, I know that makes any help difficult but I'm open to any input.

Thanks for your help!

Dan

Replies to This Discussion

Permalink Reply by davisdc on February 8, 2012 at 2:10am

Here is the grasshopper definition I've written so far

Attachments:

hexagon panels.3dm, 45 KB
hexagon panels.gh, 16 KB

Permalink Reply by Peter Kluck on February 8, 2012 at 4:27am

Is this what you are looking for?

Best, Peter

Attachments:

hexagonpanels_REV01.gh, 17 KB

Permalink Reply by davisdc on February 8, 2012 at 12:06pm

Yes that is it! Thanks so much

Permalink Reply by Dedackelzucht on February 8, 2012 at 4:28am

Hey,

I could not manage to get your version work for a division in u and v direction, and only with a odd number of divisions.

therfore I added a definition i would do it. Maybe this helps

Best Regards

DeDackel

Attachments:

hexagonpanels_DeDackel_1.gh, 49 KB

Permalink Reply by davisdc on February 8, 2012 at 1:33pm

I see what you did there, that works very well. Also gets rid of the random extra "zig zag" line you see in the above photo.

Thanks!

Permalink Reply by davisdc on February 9, 2012 at 12:22am

I've been trying to apply the above lofted pattern to an elipse form but as expected, it does not wrap. Thus I've changed my original script because I realized I need this screen to wrap a circular building facade.

Anyways, I am now back to square one with the same problem as before because the additions you guys made to my last definition don't work when wrapping the pattern full circle to reconnect with itself. I think the problem was that the definition sorted by lengths of the trapezoid's edges that remained constant, but now they all vary based on the input geometry.

So, any help with the new definition will be greatly appreciated!

Thanks!

Attachments:

hexagon panels- ellipse.3dm, 62 KB
hexagon panels- circular repeat v2.gh, 16 KB

Permalink Reply by Peter Kluck on February 9, 2012 at 5:20am

it´s a bit buggy, but have a look at this,

Best, Peter

Attachments:

hexagonpanelscircularrepeatv3.gh, 18 KB

Permalink Reply by davisdc on February 9, 2012 at 12:09pm

Thanks Peter

That kind of works.. I was messing around with the mirror component as well. The pattern isn't seamless though, which will be a problem.

Is the problem perhaps within my original definition? Am I going about wrapping the pattern to the geometry in the right way?

Permalink Reply by Pieter Segeren on February 12, 2012 at 8:02am

Hi davisdc, here is my attempt. I think it could be simplyfied some more though..

Attachments:

Hexalt_PSG.gh, 21 KB

Permalink Reply by davisdc on February 12, 2012 at 3:22pm

This is fantastic Pieter, I can't thank you enough!

From what I can see, you were able to achieve what I needed in half the components that I was already using- to simplify more would be very impressive.

How were you able to wrap the pattern seamlessly? Is it the reparamaterize?

Permalink Reply by Pieter Segeren on February 13, 2012 at 11:30pm

Glad you like it. But I don't understand what you mean asking how it's 'wrapped seamlessly'. Maybe you mean that the slider is set to allow even numbers only?

Actually I simply forgot to un-reparameterize the referenced surface in the process, but you can just uncheck that.
I made this so that I myself could follow the logic, but I have a feeling there must be a smarter way to achieve the yesnoing I'm doing here. I kind of added the simplify remark to try and challenge some people... ;)

RSS