For me, gridshells culminated about 5 years of hands-on research between complex geometry, digital fabrication, and material constraints. There is also a wealth of information on this in the Grasshopper forum, you can start with a long-thread on gridshells and GH here. Do note that since the time that I contributed to that thread, there is now a Geodesic component in GH (been there for a while now). More recently, Grasshopper has developed through new form-finding gravity-based plug-ins to Grasshopper such as Kangaroo and Geometry Gym. Before I get to details, what are we after and why?
Gridshells offer an elegant integration between form, structure, and material, and yet without a more flexible topological design tool, they are very difficult to implement more than the typical straight long span approach. Certainly Otto’s Mannheim Gridshell was not a straight space, but truly more of a blob (and note Greg Lynn’s reference to this in what I’d like to call “groovy tectonics” in his “Why Tectonics is Square and Topology is Groovy”, in Folds, Blobs, and Bodies). And, as I shared on Wednesday, the most recent and compelling example in Shigeru Ban’s Pompidou Metz. But Metz points to the challenge of gridshells and free-form – the Metz is a heavy timber gridshell which departs from the very “lightweight principle” that Otto was after. Therefore, I would like the research focus groups to explore the more restrained lattice approach to gridshells, which develops from straight laths. Even if at a certain point we decide to depart from this more restricted case of gridshells, going down this road will be instructive.
In summary, the desire is for a parametric tool that can help a designer integrate form and material through a simultaneous top-down (form driven) and bottom-up (material driven) approach. Strictly speaking we won’t be able to develop a true computational bottom-up approach in GH, but we can use GH to test solutions in real-time through geodesics. As the geodesic is the shortest distance between two points on a surface, when it is unrolled it will be a straight lath. (Keep in mind that this is geometry only, and does not include a material’s capacity to bend (true material constraints). However, the rainscreen group will be studying a way to do this through curvature analysis, so for the gridshell group, the geometrical problem is difficult enough!) For all this, the analog computers of Otto may still make great sense!
Near term goals over next couple of weeks:
- Precedent analysis and review with images, what’s been done (Otto’s Mannheim, Weald and Downand, Ban’s Hannover Pavilion, Savill Building, Ban’s Pompidou Metz, and others for you to discover!), and who has done it (Frei Otto, Buro Happold among others) and also who is constructing these things? (be smart and coordinate these efforts across the gridshell research groups, more depth, less overlap).
- Get comfortable with Geodesics by playing with them “manually” in Rhino through the “shortestpath” command, as well as get into them in GH.
- Close the damn computer and play with material, developing lath-like gridshells, could also then use the digitizer to evaluate these forms.
- Read my “Gridshell Tectonics: Material Values and Digital Parameters” to understand where I am coming from, and why dynamic surface relaxation matters. A non-formatted version can be found here, warning it is a slow download.
- Get into form-finding through GH with Kangaroo and here and Geometry Gym, GG has a very simply surface relaxation component in it that I have had some limited success with.
Long-Term Three-Part Development for GH:
- Surface relaxation in GH
- Developing Geodesic Pattern (grid) in GH
- Developing lath geometry from Geodesic Pattern
Only three goals, but not easy, and likely won’t be possible to do all three in a single GH definition. I will work with both gridshell research groups along the way to see what your interests are in full-scale prototypes etc. But this should be plenty to get you busy now.