All structures are flexible, at least a little bit. Giant concrete and steel buildings sway in the wind. Nonetheless, the goal of structural engineering for most buildings and other elements of the built environment is rigidity.
Rigidity has been structurally important because, among other things, most of the major structural materials we have used tend to be most durable when they are not flexed, and undergo significant forms of deterioration when they are flexed.
Concrete has very little flexural strength, and cracks or disintegrates when flexed too much. That’s why it is reinforced with steel.
Steel and aluminum initially get more rigid (and brittle) from being flexed significantly, and if flexed repeatedly over a long period, can undergo metal fatigue and eventually fail. Steel structures are designed to resist excessive flexing. (One notorious steel-framed high-rise in New York City swayed so much in the wind that elevators would get stuck in their tracks, and it developed an inch-wide crack in the plaster of a stairwell that ran the entire height of the building.)
Wood is quite flexible, but too much flexing can may it split or splinter, creating weak areas that are less able to bear their loads and prone to more splitting.
Composites, by contrast, offer the possibility of durable structures that are more flexible. But are we psychologically ready for them?
Fabric-enclosed structures are becoming more common, but they’re psychologically easy to accept. They’re a lot like tents, and usually have a nice, comforting rigid frame supporting the fabric and defining its shape.
Using fiber reinforced polymer composites for structural applications presents the possibility of much more flexible buildings. Literally and figuratively.
The recently built Orb at Lawrence Technological University is a fiberglass ‘egg’ designed to have an elliptical cross section in its central area. It was cast with a circular cross section throughout, and then permanently flexed during erection. However, once installed, it doesn’t move.
A real step towards flexible structures is suggested by a modest project unveiled last fall called a Transformable Meeting Space. Designed for Google by MIT’s Self Assembly Lab in the collaboration with Michelle Kaufmann, and architect working for Google, it is a more-or-less cylindrical partition that can be lowered from the ceiling to quickly create an enclosed space where there was previously open plan space. It is presented as a work-around for the deficiencies of the open plan office (noise and distraction).
The enclosure is comprised of flexible wood boards attached to a woven lattice of 36 fiberglass rods. It operates sort of like a Chinese finger trap. It is attached to the ceiling. It its retracted position, the slats stack alongside each other in a short, thick annulus overhead, leaving the floor area open (or at least, unwalled). When extended, the slats descend in curving diagonals that reach to the floor and form a thin wall defining a private space suitable for seating 6-8 people. Lowering and raising is done manually, assisted by a system of counterweights that reduce the necessary human effort to easily manageable levels.
The design is scalable. The MIT website says, “This research proposes an alternative whereby structures can easily transform between private phone booths, lounge spaces or other quiet meeting spaces into open flexible areas.” Indeed, a series of prototypes were made in sizes ranging from 10 cm to 20m.
(And if you notice the resemblance to Maxwell Smart’s Cone of Silence, please feel free to take 15 seconds for laughing and reflecting on the absurdity of all human endeavor.)
The entire structure depends on a flexible, woven skeleton of FRP rods. Its durability depends on the ability of the materials to flex and keep flexing.
The new constructability of flexible structures presents architects with options that have probably not been deeply explored before. If flexibility is no longer a weakness, then what are its strengths? What does it allow us to do that we couldn’t do, or do well, before?
Images via MIT Self Assembly Lab