For my amusement I am trying to understand Nitinol, the metal with a memory. I read that the memory is related to state changes from martensite to austentite, but do not really understand why such a slight phase change can cause such significant shape change. I further understand that the change from matensite to austentie is similar to what is used when forging steel, but do not understand why this change occurs in nitinol is such a narrow temperature range. Can anyone help me out?
Well, it just so happens I wrote a Master’s Thesis on models to use for Nitinol and other SMAs when incorporating them into dynamic systems. It has been a while though, so let’s see how much I remember.
First, for the rest of those listening in, Nitinol is what is known as a Shape Memory Alloy, or SMA. It was discovered, by accident really, that certain metal alloys (Nitinol is a Nickel Titanium alloy) exhibit very strange properties. You can forge them into whatever shape you want (say Mickey Mouse) and then plastically deform them (bend them so they stay that way). So far, so normal. With SMAs though, when you heat them they return their original shape (Mickey jumps back up out of a crumpled pile of metal), hence shape memory alloys.
This effect can also be triggered at the right temperatures by cyclically stressing the material around a preload in order to produce a hysteresis effect and therefore damping. In other words, at the right temperature it will change phase when you stretch it and change back when you release it. This process eats up a lot of energy and can be used to damp out vibrations. In fact, some washing machines use them to keep them from shaking all over the place.
Now we get into the material science part. I warn you I am a little fuzzy on it myself. I am an engineer so it wasn’t particularly relevant and I only really delved into it for my own edification. I am working off of the paper, “An Introduction to Martensite and Shape Memory” by Wayman and Duerig to provide the basic introduction.
You start in the Austenite phase. Picture this as a crystal lattice that is a perfect grid. This perfect grid is a stable crystalline configuration, but only at higher temperatures. As the material cools it transforms into the lower energy configuration of Martensite. Picture martensite as the same grid but sheared several degrees. Every row does not need to shear in the same direction. You can have one shear right, the next left, etc., etc. This is called twinning. When the material twins like this there is no macroscopic change in shape, so you see no change when the material cools.
Next, you stress the material. By stressing you can cause the structure to shear all in the same direction, rather than alternating. This can result in very large strains (3-8%) that do not immediately spring back when the material is unstressed. So now you have a deformed cool material.
Next you heat it. Heating the SMA causes it to form back to the perfect grid of Austenite. What you see is it spring back to whatever shape you originally formed the material to (Mickey Mouse). Pretty cool, eh? Now, if you want to know the molecular level mechanisms that make Austenite the more stable configurations within certain temperature bands you better prepare for some in depth reading.
Here is some trivia since I never get to talk about this kind of thing anymore. Did you know it is used in some underwire bras? Not for the phase change, but for another related property known as superelasticity. It was also used as an actuator on a Mars probe. You apply heat (by running a current through it), the phase changes, and it seeks to return to its original shape. This provides an actuation force. When it cools it stays in this shape, but if there is a spring or something that pushes back against it, you can force it move back again.
Oh, and as for the relation to steel, this is tricky. Originally it was thought that it was very similar, thus the names Austenite and Martensite (which originally related to steel states) were borrowed. This has come under scrutiny though and last I checked they were not sure.
Hope this helped.
Thank you. That helped alot. And since IANAE I think I now realize that it is fairly hopeless for me to comprehend SMA’s. Too many terms that I truly do not undersatnd.
It’s not exactly easy, that is for sure. Don’t worry too much about the terms though. I have a couple new ones to figure out every time I read a paper. Thank god for the internet.
Years ago in college, when I first heard about Nitinol, I had the idea for a mystery story where the murderer presents the victim with a jewelery neck bracelet made of Nitinol. But when it warms up to 98.6º (body temp), it contracts, thereby choking the victim to death. After they died, their body would cool, and the neck bracelet returns to it’s original shape, leaving an unsolveable mystery for the police.
Probably several technical inaccuracies in that, but that wouldn’t matter just for a story.
If I had some writing talent, I might have tried to write such a story.
That is a pretty cool idea. To make it work though you need a restoring force. The necklace would not normally re-elongate on its own, it just has the potential to do it, so you need to incorporate some kind of spring-like material into the necklace that would force it re-widen. Also, it would have to be a choker (ha!) to start with as current materials only give you about an 8% strain. Even then you are pushing it.
Can’t wait until Lion’s Gate turns this one into a major motion picture. 
We had an earlier thread on NiTiNOL in which I scrounged up some links and other information–nothing much different than flight has already posted, but some of the links have cool pictures and so forth that might aid in understanding.
Zut, thanks, some really good stuff there. NiTiNOL really is a very interesting material.
8% is hugely more than you need. You wouldn’t achieve the strangulation by the shrinking of the metal itself. What you’d do is make a bimetallic strip with it and some other metal. In its safe configuration, it’d have one radius of curvature, but in its strangling configuration, it could have some radically different radius of curvature (so in the strangling configuration, it’d be wrapped, say, one and a half times around the victim’s neck, rather than just one time). An 8% strain just means that you could make your bimetallic strip as wide as 8% of the size of the smaller radius of curvature, which is well more than you need for the thickness of a piece of neck jewelry.
Very interesting idea. The bimetallic strip would also fulfill the requirement for a restoring force. Now you just need a snazy design so she puts it on in the first place.