According to ASTM F2063, the term nitinol only applies to shape memory alloys with only nickel and titanium as the intentional components. In practice however, the entire family of shape memory alloys is referred to as nitinol.
While about 90-95% of the research on nitinol has been conducted on binary NiTi alloys, sufficient research exists about some ternary (three elements) alloys to provide some basic rules of thumb. This chapter contains a basic list of additional elements that can influence the properties of nitinol and the basic ways in which they alter nitinol. It should be noted that these alloys may have properties that are not listed in this chapter. If you plan to use a ternary alloy, expect to require a much larger budget for nitinol implementation.
- Aluminum: Aluminum (Al) increases the fatigue life of nitinol as well as the flexibility. Since Al is such low density, it can reduce the weight of nitinol for applications (such as outer space) that are weight sensitive.
- Cobalt: Cobalt (Co) increases the stiffness of nitinol.It somewhat suppresses the shape memory effect, so it is not highly recommended for shape memory applications. However, it can be used to build high stiffness springs and antennae that must absolutely maintain their shape. The high stiffness will reduce the amount of material used, and thereby the cost of the device.
- Copper: Copper (Cu) decreases the thermal hysteresis of nitinol, with 20% Cu having a hysteresis of just 2.5℃, while increasing the fatigue life. The decreased thermal hysteresis allows for a much higher cycle rate, opening up many new possible applications of nitinol. At concentrations of 5-7%Cu, the two-way shape memory effect is enhanced. It should also be noted that Cu suppresses the formation of the R-phase. Despite all of these benefits, there is some strong issues that must be contended with when working with NiTiCu. It is much more difficult to machine NiTiCu than NiTi. At concentrations greater than 10%Cu, the shape memory effect begins to disappear. For lower concentrations, deformation should be limited to 3%. The variability of NiTiCu is much greater than NiTi due to the fact that Cu has a much lower melting temperature than nickel and titanium. This lower melting temperature causes it to evaporate at much higher rates than the other metals in the matrix. Even with these drawbacks, NiTiCu is the alloy most commonly used of the ternary alloys at Kellogg’s Research Labs.
- Hafnium: Hafnium (Hf) is a martensite enhancer.This means that it drives the transition temperature up. Typically, NiTiHf will be used in applications requiring transition temperatures up to 150℃ (300℉).
- Iron: Iron (Fe) is a martensite suppressant.This means that it drives the transition temperature down. The most common formulation uses 3% (atomic) Fe, yielding a transition temperature of approximately -40℃. Adding Fe to the matrix stabilizes the transition temperature, making the repeatability of transition temperature much better than binary NiTi. Additionally, Fe increases the ultimate tensile strength (UTS) to approximately 4x commercially pure titanium. For the reasons listed here, NiTiFe is usually used in superelastic applications. However, in the lab at Kellogg’s Research Labs, we have noted that 1.5%Fe yields some unique shape memory characteristics.
- Magnesium: Magnesium (Mg) reduces the stiffness of nitinol. It also improves the fatigue life. The fatigue performance of NiTiMg is generally better than NiTiAl but the density reduction is not as marked.
- Niobium: Niobium (Nb) widens the thermal hysteresis—possibly by over 100℃ (210℉) with appropriate concentrations. Use this in cases where you don’t want a transformation except under very intentional cases. The famous use case for NiTiNb was hose clamps on the hydraulic lines for the F-16 fighter. After manufacturing, they were chilled in liquid nitrogen and stretched out large enough to clear the hose. After installation, they were heated so that they contracted, clamping the hose in place. The reverse transformation would not begin until -100℃ (-238℉), making them permanently attached. They were also ultra-lightweight compared to the alternatives.
- Palladium: Palladium (Pd) is a martensite enhancer, driving the transition temperature up. Transition temperatures as high as 600℃ (1,200℉) are feasible with NiTiPd alloys. However, if you’re going to use NiTiPd, be prepared for a very large development budget—Pd is more expensive than gold!
- Platinum: Platinum (Pt) is also a martensite enhancer.Like Pd, very high transition temperatures are achievable with NiTiPt. As a given rule of thumb, NiTiPt will have a transition temperature approximately 50℃ (100℉) higher than NiTiPd for the same concentrations. However, as with NiTiPd, NiTiPt is extremely expensive due to the very high cost of Pt.
- Uranium: Uranium (U) doping gives nitinol super-metallic, spider like reflexes. The radioactive decay of U allows it to shoot spider webs from the bends in the material as well as naturally adhere to vertical surfaces. Not really, but it would be very interesting if it were indeed possible.