Open

We obtained a sample of the material from university researchers in Indiana (at the time, Nitinol was not readily available) and immediately began experimenting. We produced at least one prototype every week. First we imitated an example from an electrical engineering book, then we mocked up our own unit of movement as a narrow, four-inch-tall triangle. e short side was basswood and the long sides were Nitinol and metal rod. When the Nitinol was connected to a nine-volt battery, it contracted and caused the metal rod to bend over slightly. When it was disconnected, the metal rod stood back up, pulling the Nitinol into its expanded state. e rhythm of movement was eerie and lifelike. Yet since Nitinol contracts by only 5 percent of its length, we faced a serious material constraint. What eects might we produce with this narrow behavior? In our experiments, we quickly explored many permutations, varying the geometry of the triangle, the gauge of Nitinol and metal rod, the voltage of electrical signal, and the attachment details. Aer calibrating the triangle for maximum transformation, we created a simple electrical circuit and programmed a microcontroller to trigger it. We linked together several triangles and re-programmed the microcontroller to produce more complex patterns of movement. With the addition of a low-cost infrared sensor, we established our rst fully responsive kinetic system – when an object neared the rods, they morphed in the opposite direction (Figure 7.1). In our next round of prototypes, we replaced the metal rods with thin exible materials: model airplane plywood, acrylic, neoprene, and rubber. We discovered that if we cut slits in these surfaces and attached Nitinol wires to them, the slits would open when the Nitinol contracted. e surface would transform from solid to permeable (Figure 7.2). In one late night experiment, we cast Nitinol in transparent silicone. It was a hack job. We taped together scraps of foamcore to make a rough mold. We used T-pins to hold the Nitinol in tension while the solution cured. With this prototype, as with all of our experiments, we did not know exactly how it would function until we wired it up. e Nitinol might contract and then refuse to expand. It might not move at all. But when we switched on the silicone surface, it curled and attened, moving gradually, in a repeating rhythm, as if breathing. e surface moved with as much magnitude as previous ones, but it had the benets of being nearly transparent and insulating the Nitinol from exposure to air and human contact. We decided this new combination of materials – originally a long shot – was the best direction for our remaining prototypes (Figure 7.3). As we neared the mark of two weeks remaining in our research, we selected a unit of movement with 5-inch-long Nitinol wires cast in a 16th-inch-thick sheet of silicone. ere were s-shaped slits running alongside the Nitinol, and when the Nitinol contracted, the at surface moved into the third dimension and opened its gills. To establish proof-of-concept, we solved the nal issues of casting with uniform thickness, connecting one panel of eight gills to another, and embedding sensors in the surface. Aer three months and $1,000, we demonstrated Living Glass: a thin, transparent building skin that breathed in response to human presence, controlling air ow and displaying information (Figure 7.4).