Material systems in nature generate movement and force through the interaction of materials, structures, energy sources, and sensors,² but do not distinguish between structural and functional properties within material itself. The hierarchical structure of natural material systems span several orders of magnitude which support the change of properties from one hierarchical level to the next,³ promoting variability and active adjustment to the ever-changing environment. Engineered systems, on the other hand, are designed to perform specific tasks under predictable external conditions and have their performance goals adjusted immediately if variation in performance is observed.⁴ There is a need for a new class of adaptive systems able to operate in changing external or internal conditions, capable of variation in performance with performance goals adjusted to interface effectively with the unpredictability of indeterminate occurrences. Architects are now beginning to recognize the role active matter can play in the interchange between the built environment and its larger material, social, and ecological context. This shift from “what the building is to what it does” ⁵ gives rise to a new attitude where architecture can become an active participant in its environment. 

To achieve the adaptivity that our current climate predicament demands of us, we need to look beyond traditional architectural materials and assemblies. Technology transfers from fields such as material science, biomimetics, autonomous robotics, soft robotics, interface design, and computation are influencing not only the range of the materials used in architecture, but even the very scale at which they are deployed. In this context, smart materials present an opportunity to augment the capabilities of architectural assemblies. Architectural components conventionally rely on engineering criteria that promote constancy in performance.⁶ Smart materials, though, are technologies of motion, energy, and exchange.⁷ Their activation is triggered by chemical, electrical, magnetic, or other interactions between molecular layers. Radically different from normative building materials, they are designed to dynamically respond to energy fields. 

Shape Memory Alloy is successfully used as an actuator in robotics, aircraft hydraulic systems, microcircuit breakers, temperature controls, electronic locks, and even medically as reinforcement for arteries and veins (stents) or in dental braces. In the building industry, superelastic SMA is used in seismic devices, to restrain movement during an earthquake, and for post-tensioning concrete structures that have fractured from insufficient shear resistance. In all these instances, the superelasticity of SMA is utilized to prevent structures from excessive deformation by yielding to, and then recovering from, deformation.¹² In architecture, the use of shape memory wire is particularly interesting for the activation of dynamic skins, even though its use is still relatively new and limited. Similar to other thermally sensitive smart materials SMA reacts to temperature changes in its environment with a significant material response at the molecular level. This is possible due to a phase change that occurs within its crystalline structure, transitioning between a high-temperature phase (austenite) and a low-temperature phase (martensite). This internal molecular elasticity enables SMA to deform significantly without residual strain,¹³ and recover its initial shape after deformation through a reversible thermo-elastic phase transformation. 

There are several types of SMAs. The most common are based on a combination of nickel and titanium (Ni-Ti), with roughly 55–56% nickel and 44–45% titanium. The temperature of transition can be calibrated through adjustments to this ratio, producing SMAs that contract at room temperature, body temperature, etc. When SMA changes from its martensite to austenite phase, the resulting stress within the material causes a 4–5% wire contraction, producing motion.¹⁴ This shape memory effect has been found in many types of alloys: Ni-Ti, CuAlNi, CuZnAl, AuCd, FeMn, etc. However, Ni-Ti (also known as “Nitinol”) is used most often for its corrosion resistance, ductility, high deformation recovery, and biocompatibility properties.¹⁵ 

When SMA wire is “trained” by baking the material into a particular shape, it remembers that geometry and, if deformed, returns to it when exposed to heat. It can also have superelastic properties. Superelastic SMA wire is in austenite state at room temperature and, when deformed, changes its state to martensite. After the load is removed, the superelastic SMA wire returns either to its austenite state or its original shape. Shape memory wire is in martensite state at room temperature and will change its state to austenite phase when exposed to heat.¹⁶ In the architectural examples discussed in this paper, there are two criteria that provide a framework for analysis and indexing: 

(1)the nature of SMA kinetic action on the assembly (length change, geometry change), and

(2)the way SMA is integrated into a dynamic material assembly (embedded into material substrate, using the action of a lever, combined with a mechanical mechanism). 

These systems of classification delineate the way SMA is utilized as an actuator and highlight the specific opportunities and challenges of the application. The kinetic action of the SMA is most often used to activate a lever or a system of pulleys that, in turn, can animate the surface or construct, as shown in the work of Philip Beesley and his Hylozoic Ground project. If embedded in the surface, the force of the activated SMA wire has the capacity to change the surface topography. David Benjamin and Soo-in Yang utilize this process in their Living Glass project to open their silicone surface’s “gills.” As these examples show, the force produced by the action of the SMA can:

(1)act indirectly, by producing motion in another connected component that could in turn move a larger construct, or 

(2)activate the surface directly, by producing a tension within its material layers. 

Because SMA wire can also be trained to return to a specific shape after deformation, a prescribed change in the construct can be easily triggered, as demonstrated by the SKiN and Lattice projects by the author, described later in this paper. This aspect of SMA remains currently underexplored at the scale of architectural assembly and this research, particularly the Lattice project, presents a novel way of amplifying the force of SMA at a larger scale. 

According to Payne, Air Flow(er) can provide an automatic response in summer to temperatures in the cavity between the inner and outer skin, allowing hot air to vent out of the building, and in so doing mitigate solar gain and decrease the cooling load on the building’s mechanical equipment. In the winter, absorbed radiation can be retained within the double-skin façade’s cavity, using the absorbed radiation to minimize the façade heat loss.¹⁷ 

This materially actuated device is energy independent and, without mechanical parts or motors, the system’s movement is completely silent. In addition to its environmental contribution the system could generate dynamic material effects across the entire building façade. With devices heating at different rates, a building skin would reflect the exchange of air and make visible the subtle processes of energy exchange between building skins and their surroundings. One of the challenges associated with the use of ambient temperature activated SMA, though, is that its calibrated response means the system cannot be user operated. This absence of control can be problematic if the system of air exchange could benefit from opening or closing outside of the designed SMA wire austenite (active) phase. 

The power of a silent dynamic environment is best experienced in the Hylozoic Ground experimental architecture series by Philip Beesley, an architect and educator at the University of Waterloo in Canada. Featured at the 2010 Venice Biennale, the project is made up of an intricate lattice of acrylic meshwork and covered with interactive fronds, filters, and whiskers.¹⁸ Equipped with microprocessors and proximity sensors, the installation offers an immersive experience that reacts to human presence and movement. The environment includes several kinds of actuating elements: breathing pores, sensor lashes, filters, and crickets, all actuated by SMA wires.¹⁹ The simple dimensional contraction of SMA is amplified by the use of levers or systems of pulleys. When used in this manner, the length of the SMA wire plays an important role in the amplitude of movement. The breathing pores and lashes are driven by 10 in. [254 mm] long Flexinol wire that is only 300 microns [0.01 in.] in diameter, so the contraction of this long wire, amplified by mechanical leverage, translates into a curling motion of the mylar frond. 

Filters and crickets, meanwhile, use shorter lengths of wire in series to provide more subtle kinetic responses. The wire diameter plays an important role since it determines the weight the SMA wire can lift (e.g., an SMA “muscle” wire with a 0.004 in. [0.1016 mm] diameter can pull 150 g [5.29 oz] per one foot [30.48 cm] of wire while a 0.012 in. [0.31 mm] thick wire can pull 1,250 g [2.76 lb.] per foot [30.48 cm] of wire).²⁰ 

Embedding the SMA wire into a surface presents another way to activate a surface or an architectural element, but this approach requires experimentation and close observation of the surface movement caused by the wire, as shown by the rigorous studies for Living Glass. This approach also required a surface edge restriction to leverage the internal force generated by the embedded wire’s movement. The composite surface demonstrated that a certain level of fluidity is possible when smart materials, embedded electronics, and a material matrix are integrated. As Living Glass demonstrates, embedding SMA into a surface has the power to change its topography, but the surface does not have to be light and soft. 

Marco Formentini and Stefano Lenci, engineers from the Department of Civil and Building Engineering and Architecture at the Polytechnic University of Marche in Italy, proposed a ventilated façade system with aluminum panels actuated by SMA wire. In this project, SMA is used as a thermal sensor and as an actuator, similar to the Air Flow(er) device, making the operation of the panels energy independent, i.e., responsive to the ambient heat they absorb. When the SMA wire affixed along each panel’s lower edge reaches 86ºF [30ºC], it contracts just enough to slightly buckle the panel. These openings then allow air to enter the wall cavity, producing a chimney effect and promoting ventilation to cool the cavity. As the temperature falls again to 68°F [20ºC], the cooling wire closes the panels, allowing a wall cavity to retain the heat.²⁴ What makes this project particularly interesting is its focus on making an otherwise static, hard wall paneling system active and responsive to the dynamics of the temperature changes in its surroundings. 

SKiN’s second experiment examined the capacity of SMA wire to act as an embedded linear source of actuation. The “long” (45 cm) [17.72 in.] pieces of SMA wire “trained” into large amplitude (15 cm) [5.91 in.] waves were threaded through the silicone tubing grid and then integrated into a silicone surface. The fusion of the grid and silicone matrix created a flexible surface that when activated achieved a level of material equilibrium; the SMA wire pulled the surface into a particular shape dictated by its large amplitude wave shape, while the weight of the silicone layer pulled the material system back to its original shape, deforming the SMA wire in the process and returning it into its original dormant configuration, ready to be activated again. In this experiment, the local movements of each wire resulted in a complex global movement of the entire surface where each shift of a cell depended on the movement of adjacent cells or regions (Fig. 5). 

In contrast to theSKiN project, where surface movement was at the center of the exploration, the focus of the Lattice project, also prototyped by the author, is an actuated structure with structural hierarchy that can facilitate movement. The system is based on interconnected rib elements, each consisting of three layers of flexible plastic strips connected sectionally to form cells. 

Combining structural behavior of the gridshell lattice with dynamics of the SMA spring is what differentiates the use of shape-trained SMA in the Lattice project from the use of SMA in other mentioned projects. This could be a direction for further development of SMA actuation in architecture where structural behavior and dynamics reinforce each other’s effects to form a dynamic system (Fig. 9). 

In combining the performance of the SMA spring with the DC motor, their project demonstrates that hybrid use is possible and even desirable when the temperature is outside of the shape-programming range of the wire. This approach can augment SMA performance by using discrete motor power and adding the possibility of user control for times when temperature ranges render SMA inactive.²⁷ 

Shape Memory Alloy can be combined with other active structures to produce a composite structure capable of retaining multiple shape configurations. Wei Wanga and his team developed SMA-based composite actuators that consist of fusible alloy (FA) material, Ni-chrome (Ni-Cr) wires, and SMA wires brought together in a smart soft composite (SSC) structure. The matrix of this composite material is highly flexible Polydimethylsiloxane; SMA wires are positioned in this matrix in the upper and lower zone, and FA with Ni-Cr wire is positioned at the edges of the actuator. When activated, the embedded FA elements are melted by applying current to the Ni-Cr wires. The SMA wire is then activated to bend the actuator. Electrical current is applied to produce shape change of the SMA wire until the FA elements solidify to retain the produced deformation. 

SMA-based SSC structures have soft morphing capabilities that allow the actuator to produce a smooth continuous deformation and retain multiple shape configurations. The composite structure of the actuator can be tailored to its application by modifying its geometry, volume, position, and composition of the embedded FA structures. Like an elephant trunk or an octopus arm, they can retain shape by changing locally between a high and low stiffness state when electrically stimulated. This research in SSC could be a step toward a new kind of material that can morph into complex forms. This lightweight morphing structure with variable stiffness could find an application in the auto or aerospace industries (morphing wings) but could also be used in adaptive building skins where change of geometry of the skin can influence its performance.²⁸ 

Integration of new material technologies and kinetics into architectural assemblies can have a transformative effect both on people and the built environment itself. Whether ambient activated or electrically stimulated, architectural components using material actuation could provide not only a new aesthetic for architecture, but also a new level of buildings’ sensitivity to environment changes, producing surfaces more fluidly synchronized with users than their mechanical counterparts. Smart materials are usually designed for a specific behavior, though, and this presents a limitation in their degree of responsiveness. While Hensel and Menges ³³ suggest that smart materials should be able to evolve so that they can respond to varying stimuli and stay active, Joanna Aizenberg describes difficulties in developing multifunctional materials. In biological organisms, change between the optical, electrical, or magnetic functions of the tissue happens seamlessly; in artificial materials, this requires specific interfaces that are difficult to replicate.³⁴ 

Shape Memory Alloy is a functional material: it can serve as a sensor and actuator and, to some extent, mimic a biological organism by synchronizing with environmental changes. However, for any material to respond to varying stimuli, a complex molecular communication, akin to a living cell mechanism, would have to be established. This remains a limitation for SMA actuators and many other smart materials. The future of dynamic building skins will likely belong to low-energy systems that can harvest heat from the sun or kinetic energy of the wind. In the projects described in this paper, the “sensing” and “actuating” capacities are built into the material, eliminating the need for complex mechatronic assemblies. Such systems of dynamic activation that rely on innate, albeit designed, properties of materials are perhaps the most promising direction for developing adaptive building envelopes.