Shape-memory alloy

Shape-memory alloys can undergo a reversible solid-state phase transformation, commonly described as a martensitic transformation. In a typical thermally activated shape-memory cycle, an alloy is deformed in a lower-temperature phase and then heated through its transformation range. The material changes crystal structure and recovers a previously trained shape. The transformation is solid-state: the alloy does not need to melt to move. Specific transformation temperatures and behavior depend on the alloy composition and processing.

The shape-memory effect is often distinguished from superelastic behavior. In the former, heating can trigger recovery after deformation. In superelastic use, a stress-induced phase transformation can permit substantial deformation and recovery after the load is removed, under appropriate temperature conditions. These are related behaviors of some shape-memory alloys, but they are not interchangeable descriptions of every SMA component or operating cycle.

Nickel–titanium, often called NiTi or nitinol, is an important shape-memory alloy family. Research and engineering applications include medical devices, microelectromechanical systems and compact actuators. NASA describes SMA technologies investigated for actuators and deployable or adaptive engineering systems. The combination of recoverable movement and useful force from a small component can be advantageous where space is limited, but suitability depends on design requirements and the operating environment.

SMA behavior is sensitive to composition, heat treatment, fabrication and the temperatures and stresses experienced in service. Components must be designed around the relevant transformation range and expected loading. The same mechanical and chemical robustness that can be useful in service can make processing difficult: NIST research on NiTi thin films, for example, examines how focused-ion-beam machining affects near-surface mechanical properties. Reviews also identify manufacturing cost, machinability, fatigue and application-specific performance as considerations for wider adoption.