Sketching elastic form

Attached is a schema model I created for my studio project, I would like to create this form using grasshopper if at all possible, the logic of this form would be created by emergent behaviour i.e. non hierarchical ordering system.

Attachments:

1.jpg, 2.1 MB

Replies to This Discussion

Permalink Reply by Christian Schmidts on November 13, 2012 at 5:18am

Thats tricky stuff. A good friend worked with resin-impregnated fiberglass. (I' wondering what material you using) He tried to simulate the heavier with different tools like kangaroo and processing but in the end it was more something that looked similar rather that a real simulation of the of the forces, etc

chris

Permalink Reply by shaun on November 13, 2012 at 6:09am

I am afraid I may have bitten more then I can chew :/ for my studio project. I started by casting a resin I can upload pics of that shortly. Eventually to get the effect I wanted I decidided to use chewing gum... I ve been looking all over the net to find tutorials using rhino to medic the effect to no avail which is a huge bummer

Permalink Reply by Anders Holden Deleuran on November 13, 2012 at 6:38am

You might want to have a look at some computational implementations of the classic Frei Otte wool path experiments. I know that several people have built simulations of such systems in both Processing and Grasshopper. Here's a thread (pardon the pun!) on the subject using Kangaroo.

Permalink Reply by Christian Schmidts on November 13, 2012 at 8:04am

I just found the links; here is a quick description:

there was also a paper published but unfortunately not public. maybe you can get access, if not i could ask my friend.

i appreciate the work of angel but when it comes to material behavior, stresses, surface tension i think that "our" tools are still no complex and powerful enough - and like i said i didn't really see the benefit in the work of my friend form the digital experiment.

so i think the question is is there a benefit from your digital experiment or do you rather stick to the physical experiment.

Permalink Reply by Anders Holden Deleuran on November 13, 2012 at 8:22am

I think that this part "..are still no complex and powerful enough" is pretty key. It all depends on what the intention of a simulation is. If one is trying to accurately simulate the physical process of fibres and resin interacting, than yes, it does indeed become very complex! Conversely, if the intention is to abstract a physical process as a generative mechanism for a spatial/structural/architectural schema, I think that we have seen plenty of work lately which suggests that this does not necessarily have to be "that" complex

(thanks in large part to physics based dynamics tools such as Daniels Kangaroo plugin, the Traer library for Processing and Nucleus in Maya etc).

Permalink Reply by Christian Schmidts on November 13, 2012 at 8:47am

"I think that this part "..are still no complex and powerful enough" is pretty key. It all depends on what the intention of a simulation is. If one is trying to accurately simulate the physical process of fibres and resin interacting, than yes, it does indeed become very complex! Conversely, if the intention is to abstract a physical process as a generative mechanism for a spatial/structural/architectural schema, I think that we have seen plenty of work lately which suggests that this does not necessarily have to be "that" complex"

Agree 100% on that.

I definitely see the benefit and progress that is made every day - not least here. But there is also (fortunately) a lot of space for improvement.

The other point is there no fundamental difference between digital and physical. The question is only what serves you more for your next task.. :-)

Permalink Reply by shaun on November 13, 2012 at 8:53am

My overall goal would be to take the digital model and get it 3d printed (stereolithograph) for my final...I am thinking...I've seen some provacrive forms via the web from students at Scri-Arc and the AA..I was thinking of paying to view some of the tutorials on Modelab.un ...have either of you used that as a resource?

Permalink Reply by Michael Pryor on November 13, 2012 at 9:04am

something like this I made very easily with random pipes and starlings fastmesh component plus wb smoothing. I imagine if it was very dense the look could come close.

Permalink Reply by shaun on November 13, 2012 at 10:16pm

Thanks for the information Michael.. I am not familiar with starling mesh I am intrigued with the form. I attempted to modify your grid pinching definition to create the effect i'm looking for I can't seem to figure out how to blend the surfaces together I know it is relatively simple using t splines however I will loose the parametric controls. I attached a image of what I've created so far. Its gunna be a long nite :/

Attachments: