Kangaroo Frei Otto's Wet String Experiment

Hi everyone,

I've been trying my hand with a small kangaroo simulation to recreate Frei Otto's wet strings experiment. However, it seems to crash as there are apparently 10,000+ lines being calculated, even when there are relatively few to start with. Is there any way to streamline this a little bit? Also, is there any way to keep the lines on the xz axis? As at the moment they are attracting lines parallel to them as well.

Thanks

Laurens

2015_1109_2140a.gh

2015_1109_2140a.3dm

2015_1109_2140a.3dmbak

Attachments:

2015_1109_2140a.3dm, 812 KB
2015_1109_2140a.gh, 93 KB

Replies to This Discussion

Permalink Reply by David Rutten on November 9, 2015 at 3:53pm

Can't help you with the Kangaroo bit, but you don't have to share the *.3dmbak files, they are files Rhino uses to prevent file corruption in case a crash happens while a file is being saved. Just the *.3dm and the *.gh should suffice.

Permalink Reply by Laurens on November 10, 2015 at 7:10am

Thanks David! They've saved me a few times, long day... (UK)

Permalink Reply by David Stasiuk on November 9, 2015 at 4:33pm

Just from looking at the image I'd guess that your anchor points need to be flattened.

Permalink Reply by Laurens on November 10, 2015 at 7:14am

Yeah I thought it might have been something along those lines, managed to reduce it to only 450 liens after doing that. Do you know how I might make sure the curves remain planar?

Permalink Reply by Daniel Piker on November 10, 2015 at 7:31am

Hi Laurens,

there's a lot going on in the definition you posted, but for the actual simulation part there appears to be just several polylines of only 2 segments each, in separate planes, and there are no forces between them because they are in separate branches. I don't understand what you mean about the curves remaining planar.

Could you make a simple sketch/description of the result you are after?

Permalink Reply by Laurens on November 10, 2015 at 7:57am

I simplified a lot of the parameters to stop my computer from freezing, so that's the output itself is a bit more simplified than I would like.

This is what I'm going for. The lines will eventually form the fenestration for the windows in a series of vertical planes. I hope this clears it up a little.

Permalink Reply by Laurens on November 12, 2015 at 8:57am

Hi all,

I've updated the algorithm to separate each panel, this has meant I've been able to increase the complexity, however after tweaking it for a few days the outcome still doesn't resemble the Wet Strings experiment. Can anyone see anything in my settings that I could change?

My Rhino file is quite large but I can upload if anyone would like to look at it first-hand.

Laurens