Pleated Shell Structures 2 – Part 2 of 2.
An interview with Shajay Bhooshan of Zaha Hadid Architects, lead designer of the team that created the “Pleated Shell Structures” experiment/exhibiton at SCIArc.
The complete design team credits: Patrik Schumacher, Shajay Bhooshan, Suryansh Chandra, and Saman Saffarian Mostafa El Sayed of Zaha Hadid Architects; and Bill Kreysler, Josh Zabel, and Nick Colendich of Kreysler & Associates. The designers also give special thanks to Stephanie Atlan, and the students of SCIArc, CCA, and UC Berkeley who participated.
Composites & Architecture: Why is it called “Pleated Shell Structures?”
Shajay Bhooshan: The name was suggested before we made the design, based on a previous installation we had done in Venice. It’s a bit of a misnomer. In reality is just a shell structure, and not really pleated.
How did the project start?
I work as part of the design and research group at Zaha Hadid, about four people. We design tools and ideas that sometimes get used on other projects. We’ve been doing a series of prototypes, if you will. We’re trying to determine if the methodologies we’re exploring work, by trying them out. Pleated Shell Structures is the seventh or eighth in a series of design tools and methods, to test a collaborative workflow between architects, engineers and fabricators.
We’ve been in conversation with Bill Kreysler and Josh Zabel for a long time, They helped us make a prototype in Mexico City in concrete. We know they’ve been working extensively in plastic and composites, so we wanted to work with that material and see if we can develop a design based around the idea of their fabrication method.
We’ve been taking a look at historic pioneers like Frei Otto and Felix Candela and architect/engineers who’ve been investigating the idea of form finding, something that has, in its shape, a close relationship between the geometry of performance and the method of making. We tried to break from the traditional way of making a mold for plastics, to use fabrics for making a mold.
How does this fit into the question of collaborative work flow?
It was our intent. We had to figure out how to simulate stretching of fabric in a computer model. On the other side in terms of fabrication, we thought it would be interesting to try and see how it pans out. It would mean significant savings in terms of cost [as compared to fabrication of EPS molds using CNC routers, the usual method for translating a complex computer-generated shape into FRP – ed]. There is a natural way fabric takes a shape that has been proven to be more structurally efficient. In a way, it’s an old idea. People like Heinz Isler or Antonio Gaudi, hanging a fabric first, because it’s formed under natural gravity. If you invert that shape, it will work better than something not formed that way.
In this case, it’s stretching of fabric. Once put in tension, it will take the minimal surface area possible for that given boundary. In shell structures, most of the forces flow through the surface.
So then, using this method of design, the aesthetic development of a shape also inherently includes the basic engineering for it? Development of form and function completely unified?
That is exactly right. The idea of ‘finding’ the form based on the physical laws of gravity and force flow is a method pioneered and investigated with rigor and passion by the likes of Frei Otto, Gaudi, Isler, etc. We were trying to extend the method to include contemporary ambitions and pressures of design engineering and fabrication.
You mentioned just now that this fabrication method could potentially save money. How?
It takes time to mill EPS, and EPS is expensive. The plywood to make fabric forms has much less surface area, so it’s faster to cut and costs less.
How did you decide what forms you wanted?
The only way to control the surface is to control the boundaries. While designing them, we get an approximate idea of what the shape will be like.
What was the design intent?
It’s been a progression of design ideas. Previously, we made shells without apertures, continuous surfaces with no holes. One intent was to try and find how to place holes within the shells. Another was the ability to rest one shell within another, a compound shell. We went through a few iterations, aesthetically we liked this one.
Shells generally look like canopies, because they are thin structures that span a large area. The research group has been generally interested in curved and self-supporting structures. We’ve been investigating these kinds of geometries, self-supporting curved geometries.
Why is there wood edging on the structures?
We weren’t originally intending to put the wood in. Even now, we’re not sure it’s performing any structural purpose. There were some misalignments [of the cast materials], so the wood pieces are primarily guides to help place the panels in 3-D space.
What’s the story behind the wood-frame structures that stand behind the FRP shells in the exhibit?
The original design intent had five FRP sub-shells. Because we were making this ourselves along with Josh Zabel of Kreysler and some Sci-Arc students on a very short time-schedule, we realized we wouldn’t be able to make all of them. There were 44 shell panels when we started. We ditched about 12 of them, and almost three quarters of the way through the project, we ditched another six. We ended up with 22 panels. We thought we might make the others into a timber structure to highlight the difference between the two. They seem similar. They’re like empty crates, where FRP is a description of similar geometry.
We had the geometry already, and Kreysler has a lot of CNC milling machines, so it was fairly easy to provide information for the means to cut them out. They are cut entirely out of flat sheets of plywood, assembled into 4-sided shapes, then assembled into overall shape.
Does it have aesthetic value in and of itself?
We never tried to impose an aesthetic value to it, although subconsciously, we have an aesthetic language to most of our designs. We see these structures more as interim results or prototypes, before we can apply this into a full-fledged building. The aesthetic output is predominantly subconscious.
What did you learn from this exercise?
There are issues with fabric. We don’t yet know how much to stretch fabric to maintain the same shapes and geometries. And fabric can only take so much weight while we are pouring resin. It limits the size. We had some large panels, about 4 meters, and those deformed during pouring resin. Once we stretched fabric and started applying resin, the fabric deformed. When it is stretched, we don’t want that shape to change because there won’t be continuity between the panels. We learned that we need to improve the simulation: The boundaries come from a digitally simulated geometry, but the fabric is applied and forms according to natural rules. The digital simulation needs to come closer to the actual properties of the fabric.
The working life of ZHA, the driving force, has always been investigative, not just in materials, but also art and formal ideas. Design by research is integral to the working method of the office.