The Engineer Who Taught Steel to Remember Its Shape
William Buehler's 1962 discovery of Nitinol — a nickel-titanium alloy that returns to a preset shape after deformation — quietly transformed medicine, aerospace, and robotics, yet his name remains almost entirely absent from popular science history.

An Accident in a Naval Laboratory
In 1961, William J. Buehler, a metallurgist at the U.S. Naval Ordnance Laboratory in White Oak, Maryland, was searching for a heat-resistant alloy suitable for the nose cones of ballistic missiles. The Cold War had given materials scientists an unusual degree of institutional urgency and funding, and Buehler was one of many researchers systematically working through elemental combinations in search of properties that might serve the demands of supersonic flight and reentry physics. He had been testing combinations of nickel and titanium, elements that individually offered little obvious promise for the task at hand. What he stumbled upon instead was something far stranger and ultimately more consequential than any missile component.
Buehler noticed that a strip of nickel-titanium alloy he had bent and deformed would, upon heating, snap back to its original shape with surprising force. The behavior was unlike anything he had seen in conventional metallurgy, where deformation is considered permanent unless reversed by mechanical work. He called the material Nitinol, an acronym drawing on Nickel, Titanium, and Naval Ordnance Laboratory. The formal discovery is dated to 1962, when Buehler and his colleague Frederick Wang published their findings and demonstrated the effect to colleagues. The story most often told about that demonstration has become something close to legend in materials science circles: a folded Nitinol strip was passed around the room during a meeting, and when someone held a cigarette lighter beneath it, it unfolded itself as if alive. The room reportedly fell silent.
What Buehler had found was a shape-memory alloy, a material capable of undergoing a reversible solid-state phase transition between two crystalline structures: a high-temperature austenite phase and a low-temperature martensite phase. In its martensitic form, the alloy can be deformed plastically and will hold that new shape at room temperature. Heat it past a critical threshold, typically between 40 and 100 degrees Celsius depending on composition, and it reverts to its austenite structure, recovering its programmed shape with forces strong enough to move mechanical components, open blocked arteries, or deploy a satellite antenna. The implications of this behavior would take decades to fully unfold, and in some respects they are still unfolding today.
The Physics Hiding Inside the Crystal
To understand why Nitinol behaves the way it does requires a brief detour into crystallography. Most metals deform permanently when bent because their atoms are displaced into new positions that require no energy to maintain. Dislocations move through the crystal lattice and the metal stays where it is put. Nitinol is fundamentally different. In its martensitic state, the crystal lattice is distorted but internally stressed, like a compressed spring held in place by the crystal's geometry rather than by any external constraint. The deformation is accommodated not by dislocation motion but by the reorientation of crystallographic twins, a process that stores elastic energy within the material's structure at the atomic scale.
When the alloy is heated past its transformation temperature, the lattice undergoes what materials scientists call a diffusionless transformation. Atoms shift cooperatively, by tiny fractions of a nanometer, snapping the entire structure back to its memorized austenite geometry without any atoms being added, removed, or exchanged with neighboring grains. The transformation propagates through the material as a coordinated wave of atomic displacement rather than as the chaotic, thermally driven rearrangement that characterizes most phase changes in metals. It is, in a sense, a mechanical memory encoded in the geometry of the crystal itself.
This phase transition also occurs in a second, more sophisticated form called the two-way shape-memory effect, in which the material can be trained through repeated thermomechanical cycling to remember two different shapes, one for each temperature state, and to alternate between them indefinitely without any external deformation being applied. A related phenomenon, known as superelasticity or pseudoelasticity, allows Nitinol at body temperature to be stretched to strains of up to 8% and spring back completely without any heating. This is a recoverable deformation roughly four times the elastic limit of high-strength stainless steel, and it arises because, at body temperature, the alloy sits just above its transformation threshold, meaning that mechanical stress alone can temporarily induce the martensitic phase, which then reverts when the stress is removed.
The precise transition temperature of any given Nitinol sample depends on the exact nickel-to-titanium ratio, which can be tuned with extraordinary precision. A shift of just one atomic percent in nickel content changes the transition temperature by approximately 100 degrees Celsius, giving engineers a remarkable degree of compositional control over when and how the material responds. This tunability is a large part of what made Nitinol so commercially versatile across such different fields. The same fundamental material could be engineered to respond to body heat in a surgical implant, to warm tap water in a plumbing actuator, or to the elevated temperatures of an industrial process, with equal reliability and predictability.
From Missile Cones to Coronary Arteries
For roughly two decades after its discovery, Nitinol remained a scientific curiosity rather than a practical engineering material. Its unusual properties were well documented in academic literature, and researchers produced a steady stream of papers characterizing its behavior under various conditions. But manufacturing it consistently proved enormously difficult. Nitinol is notoriously reactive during melting, readily absorbing oxygen and carbon from the surrounding atmosphere, and even small amounts of these contaminants disrupt the phase-transition behavior that makes the alloy useful. Early production methods yielded brittle, unpredictable samples whose transformation temperatures varied from batch to batch, making reliable engineering design impossible. It was not until the late 1970s and 1980s that vacuum arc remelting and electron beam melting techniques matured enough to produce reliable commercial quantities with consistent properties.
The medical community was the first to exploit Nitinol at scale, and the results transformed interventional medicine. In 1990, the U.S. Food and Drug Administration cleared the first Nitinol-based endovascular stent. The device exploited superelasticity in a way that was elegantly matched to the demands of minimally invasive surgery. A stent could be crimped down to a fraction of its deployed diameter, loaded into a catheter, threaded through the vascular system to a narrowed or blocked artery, and then released, whereupon it expanded to its preset shape and held the vessel open without any inflation mechanism or mechanical deployment system. Unlike stainless steel stents, which are rigid once placed, Nitinol stents flex with the natural pulsatile movement of blood vessels rather than resisting it. This compliance dramatically reduces the mechanical stress concentrations that cause fatigue fracture in metallic implants over millions of cardiac cycles. Today, hundreds of millions of Nitinol stents, stent-grafts, inferior vena cava filters, and other endovascular devices are implanted worldwide each year, and the material has become as fundamental to interventional cardiology as the catheter itself.
Orthodontics adopted Nitinol archwires somewhat earlier, in the 1970s, following work by George Andreasen at the University of Iowa, who demonstrated that a superelastic wire could apply continuous, gentle force to teeth across a wide range of deflection without requiring frequent clinical adjustment. Conventional stainless steel archwires apply force that increases steeply as the wire is deflected and drops rapidly as the tooth moves, requiring orthodontists to replace wires frequently to maintain therapeutic force levels. A Nitinol wire, by contrast, operates on the flat plateau of its stress-strain curve, delivering nearly constant force regardless of how far it has been deflected. Patients with Nitinol braces experience lighter, more biologically appropriate forces across the full range of tooth movement, which both reduces discomfort and, in many cases, shortens overall treatment time.
Aerospace, Robotics, and the Quiet Future of Actuators
Outside medicine, Nitinol has found roles in aerospace that Buehler might have found gratifying given the original mission of his research. NASA has used Nitinol-based actuators in deployable structures across multiple programs, taking advantage of the material’s ability to replace electric motors or hydraulic systems with components that have no moving parts, no lubricants, and no gears or seals to wear out or leak in the vacuum of space. The Mars exploration rovers carried Nitinol components in their antenna and instrument deployment systems. In 2021, researchers at NASA’s Jet Propulsion Laboratory published work on shape-memory alloy tires designed for planetary rovers, which can deform substantially over rocks and obstacles and recover their shape completely without permanent damage. The concept addresses a genuine engineering problem: conventional pneumatic tires cannot function in the near-vacuum and extreme temperature swings of Mars, and rigid metal wheels are vulnerable to permanent deformation from sharp rocks. A Nitinol tire occupies a design space that neither technology can reach.
Soft robotics represents perhaps the most forward-looking application of the material. Because Nitinol wires contract when heated and relax when cooled, they can function as artificial muscles, producing smooth, silent, and continuous motion without the gears, linkages, or pistons that conventional actuators require. Research groups at MIT, ETH Zurich, Harvard’s Wyss Institute, and numerous other institutions have demonstrated Nitinol-driven robotic hands, grippers, and swimming robots that mimic biological motion with a fidelity that rigid-link mechanisms cannot approach. The principal engineering limitation remains actuation speed. Heating a Nitinol wire is straightforward using resistive electrical heating, but cooling it back to its low-temperature state requires either passive heat dissipation, which is slow, or active cooling systems, which add complexity and weight. Very thin wires cool quickly due to their high surface-area-to-volume ratio, but thin wires also generate less force. Balancing these constraints is an active area of research, with some groups exploring microfluidic cooling channels and others investigating alloy modifications that shift the transformation temperature closer to ambient conditions to reduce the thermal work required per cycle.
A Material Still Yielding Surprises
Decades after its discovery, Nitinol continues to generate new research that pushes the boundaries of what the material can reliably do. In 2023, a team at the University of Minnesota published findings in Nature Materials demonstrating that carefully controlled thermomechanical training protocols could produce Nitinol samples with fatigue lives exceeding ten million cycles under surgical loading conditions, a benchmark that had long been considered difficult to achieve with consistency. The work identified specific microstructural features that concentrate transformation-induced stress and developed processing routes to suppress them, offering a potential path to longer-lived implantable devices. Separately, researchers in South Korea reported in 2024 that doping Nitinol with small amounts of copper stabilized the transformation temperature against the gradual drift that occurs over repeated thermal cycling, a phenomenon that has historically limited the long-term predictability of Nitinol actuators and implants.
There is also growing interest in applying Nitinol to civil infrastructure. Researchers at several universities have embedded Nitinol wires in concrete beams and composite structural members, where the wires can be electrically activated after an earthquake to generate compressive stress that closes cracks and partially restores structural integrity. The concept, called active self-healing infrastructure, remains at the laboratory scale, but controlled experiments have demonstrated measurable crack closure in reinforced concrete specimens subjected to simulated seismic loading. Whether the cost and complexity of embedding and connecting Nitinol networks in real structures can be made practical remains an open question, but the underlying physics is sound, and the potential applications are significant given the scale of aging infrastructure worldwide.
William Buehler set out to protect a missile nose cone from heat. He ended up giving surgeons a tool to open blocked hearts, orthodontists a wire that moves teeth with biological gentleness, aerospace engineers a replacement for motors with no moving parts, and roboticists a material that blurs the boundary between mechanism and organism. Buehler himself received little of the recognition that eventually surrounded his discovery, continuing to work at the Naval Ordnance Laboratory until retirement and dying in 2021 at the age of 92, largely unknown outside specialist circles. The commercial and medical industries built on his work generated billions of dollars in annual revenue by the 2010s, yet his name remains unfamiliar even to many of the surgeons and engineers who depend on the material daily. That the alloy is better known by the acronym of a laboratory than by the name of its discoverer is one of the quieter ironies in the history of materials science, and perhaps a fitting one for a material whose most important quality is the ability to remember a shape that no longer appears to exist.
Sources & Further Reading
- Otsuka, K. and Wayman, C.M. (eds.). Shape Memory Materials. Cambridge University Press, 1998.
- Duerig, T., Pelton, A., and Stöckel, D. An overview of nitinol medical applications. Materials Science and Engineering A, 1999. https://doi.org/10.1016/S0921-5093(99)00294-4
- NASA Jet Propulsion Laboratory. Shape Memory Alloy Tire Development. NASA Technical Reports, 2021. https://www.nasa.gov/directorates/spacetech/game_changing_development/projects/ATHLETE
- Jani, J.M., Leary, M., Subic, A., and Gibson, M.A. A review of shape memory alloy research, applications and opportunities. Materials and Design, 2014. https://doi.org/10.1016/j.matdes.2013.11.084