algorithmic modeling for Rhino
Linking Excel API with ROBOT API using VB.NET to facilitate form finding of MULTIPLE hanging chains. With care, the engineering platform ROBOT can be used to facilitate the form-finding of multiple shapes. Inverting the form will theoretically give the optimum shape for a purely compressive action. By varying the catenary profile lengths in Excel we can arrive at an infinite range of possible shapes. There is no explicit mathematical solution for multiple hanging chains. Geometric non linear solutions are needed or physical models a la Gaudi.
Robot has a geometric non-linear solver. Typically for a linear analysis the stiffness matrix is not updated when the geometry changes. For typical rigid structures this assumption is fine. Obviously for chains, tensile structures, soap films this is not the case. Furthermore, we have to find the shape that is in a static equilibrium - we can’t simply dictate the shape of a soap film tensile structure for instance.
So basically robot updates the matrix as the geometry changes. It still uses a stiffness matrix approach hence can be very unstable and you have to "nudge" teh model along by not applying too much load at once. I basically start with a network of flat chains at the start of the analysis. Their connectivity, geometry of end points, final hung lengths and other attributes are defined in excel. Hence we can quickly vary the hung length to get a new profile. I then apply a nominal vertical load on each chain of the network to "nudge" it to start the analysis process of finding the stable equilibrium shape of all chains. We then invert teh chains and rotate to get the inverted cupola form.
Unfortunately a stiffness matrix approach is not the best or most stable solution for geometrically non-linear structures. The best approach is to use Dynamic Relaxation - see my images where I form-found the shape of a soap film from an initially flat shape.
Tags:
Albums: Work and Research
Comment
I was not able to write more than 200 word in the message I have just sent you. Here comes the rest: you will be cited and will give access to an online version of the book. This is a research book and non profit.
Best Regards
About me you can find more here : http://www.ltu.se/staff/m/mania-1.54092
Hi Frank
Yep Ive seen the Stuttgart Example and spent time at the IL several years ago. Its such an inspiring place!
Yes physical modelling is still a very valid proof of any numerical solution for form-finding, a lot of the minimal form-found solutions claimed are not in fact correct.
What form-finding studies are you currently exploring in Grasshopper?
Im hoping to try it out when I get a chance. The problem with the catenary component currently in GH is that it cannot solve the form-found shape for the case when one catanry is connected to another catenary. It is a geometrically non-lienar problem and requires iterative methods of analysis.
The best bet woudl be to look at Daniel Pikers Kangaroo plug in. That should work. I havent tested it yet. Let me know how you get on.
Welcome to
Grasshopper
Added by Parametric House 0 Comments 0 Likes
Added by Parametric House 0 Comments 0 Likes
© 2025 Created by Scott Davidson. Powered by
You need to be a member of Grasshopper to add comments!