What is Hysteresis in Nitinol – Kellogg’s Research Labs

Hysteresis in Nitinol is one of the poorly understood aspects of nitinol.  We often get customers who expect the nitinol to reverse transform once it cools below the transition temperature.  Unfortunately, nitinol doesn’t work that way, so, let’s take a moment to look at some of the nuances of hysteresis and how it affects projects.

Hysteresis, by definition, is a lag or delay between input and output.  Think of a car sitting at a red light, preparing to drag race.  There is a delay from the moment the light turns green to the moment the car begins to accelerate.  Part of this delay is due to the response time of the driver, while some is a result of the mechanical systems in the car loading up.

Nitinol exhibits thermal hysteresis.  This means that the temperature at which the martensitic nitinol transforms into austenitic nitinol is not the same temperature at which austenitic nitinol transforms into martensitic nitinol. 

Critical Temperatures in Nitinol: While people often talk about the transition temperature of nitinol, at Kellogg’s Research Labs, we really track eleven critical temperatures to fully characterize the nitinol.  For each transformation, there are three critical temperatures: start, peak, and finish.  Since there are three phases (martensite, austenite, and rhombohedral phase), this makes up nine of the critical temperatures.  It is important to note that the peak temperature is the temperature at which the transformation is happening most rapidly.  Next is the equilibrium temperature, denoted as M₀, which is calculated as being half the distance between the martensite start (M_s) and austenite start (A_s) temperatures.  The final critical temperature is the martensite difficult (M_d) temperature.  Above this temperature, it is difficult for martensite to form, causing nitinol to behave linear-elastic across most of its range.

There are multiple methods of measuring hysteresis and choosing which one generally depends on the application.

1. The most conservative method is to measure M_f to A_f.  This would be a good choice if your application requires a full transformation in order to properly work.  Cases of this would be if you were using the nitinol as a binary actuator and had the power very closely matched to the needs of the application.
2. M_f to A_s: This measurement displays the amount of temperature change required to begin motion.  If the start of motion is critical to your application, then this measurement is going to be the one to choose.
3. M_p to A_p: We call this the functional hysteresis because if you pass the peak temperatures in each direction, then the bulk of the material has transformed and the nitinol can be operated through much of its range.
4. M_s to A_s: Measuring the gap between the start of each transformation can help design applications that trigger on slight movement.  However, this method can get you in trouble if you are not careful because nitinol can often have overlapping peaks (resulting in a negative measurement).  Use it only if it is critical to your application.

So, is hysteresis good or bad?  That largely depends on your application.  For applications requiring speed, hysteresis is most definitely bad and measures need to be taken to reduce the hysteresis as much as is feasible within the project budget.  However, for steady state applications, where the nitinol needs to remain in one state for prolonged periods of time, then hysteresis is helpful because it prevents the reverse transformation.  This can reduce power consumption for electrically activated products or it can widen operating temperature ranges for thermally activated products.