Something to do with Branches?

Hi Everyone,

I'm completely stuck! I am trying to join control points on curve (grasshopper generated) to create panel like surfaces. Although the number of curves are changeable through a slider, the pattern of points to make these surfaces remain constant.

I'm using the true/false cull procedure to get the order of the curves/points however i need some help working out how to order the points so that i can connect them to the srf4pt component. I am sure my explanation is quite unclear so i've attached a screen grab to help. The black lines i have drawn in paintshop to demonstrate which points need connecting.

I'm pretty sure this is something to do with sorting my point list and putting them into the correct branches but I'm not sure how that works exactly.... can someone help?

Attachments:

screen grab.jpg, 72 KB

Replies to This Discussion

Permalink Reply by David Rutten on March 20, 2010 at 1:18pm

I think my first approach would be to use the [Tree Item] Component to get the corner points of each triangle. This probably involves using three [Tree Item] objects (one for each corner) and a whole lot of [Series] components to generate the required indices.

I'll give it a whirl, see what I can come up with.

--
David Rutten
david@mcneel.com
Poprad, Slovakia

Permalink Reply by Dean Foskett on March 20, 2010 at 2:12pm

Thanks David,

To help explain, I'm trying to rebuild some geometry i originally generated using a slightly different process (see screen grab). To build it I used a series of evaluate curve components to get my triangle corners , however I wanted to adjust the amount of panels efficiently through a slider as opposed to keep adding more points. I just wondered whether there was a way of generating a tree that would join the various three points, perhaps using a couple of shift components or something... I'm not sure..?

Attachments:

Screen grab 4.png, 61 KB

Permalink Reply by taz on March 20, 2010 at 3:32pm

If you can get all the points that define your surface you could create a Delauney mesh and then convert that to surfaces.

Then you can cull any surfaces you don't want (instead of creating only specific surfaces).

Permalink Reply by taz on March 20, 2010 at 3:48pm

Oops... Sorry, that doesn't exactly work since the orientation of the triangles isn't consistent.

Permalink Reply by David Rutten on March 20, 2010 at 2:25pm

Humpf... nothing so far. I get stranded with even more Series and ParamViewers.... surely there must be an easier way to solve this!

--
David Rutten
david@mcneel.com
Poprad, Slovakia

Permalink Reply by Arthur Mamou-Mani on March 21, 2010 at 9:23am

Hey Dean,
Do you mind attaching your gh file?
cheers

Permalink Reply by Dean Foskett on March 21, 2010 at 11:20am

Hey Arthur,

I think I may have cracked it!

You were right David it took a few series functions to get it to work but it appears to be doing the Job!

I've a attached a couple of screens grabs showing how it works.

Thanks Guys for taking a look.

Dean

Attachments:

Iterations.png, 195 KB

Permalink Reply by Arthur Mamou-Mani on March 21, 2010 at 11:27am

Cool could you post a print screen of your definition?

Permalink Reply by Dean Foskett on March 21, 2010 at 12:56pm

Yeah I will do I just need to clean it up a bit first, its a complete mess at the mo and unclear to follow.

I've just noticed something which you may be able to help with. Basically the logic of the definition is that the as the span of the structure increases the depth and angle of the folded panels update to suit. At the moment this 'offset' is being determined by a line SDL component, I've used perpendicular frame on sections of the structure to get the direction of the offset. Now there appears to be an issue with the two end sections, it looks as though their perpendicular direction is different to the others, why would this be and would you suggest i do something different? Once again take a look at the image to help explain.

Attachments: