It’s hard to resist the temptation to tear apart a shiny new gadget, but fortunately, iFixIt often does it for us. This helps to keep our credit cards safe, and reveal the inner workings of new stuff. That is definitely the case with the Microsoft Surface Book teardown that they have just published. Apart from revealing that it is pretty much impossible to repair yourself, the teardown reveals the mechanism for the innovative hinge and lock mechanism. The lock that keeps the tablet part in place when in laptop mode is held in place by a spring, with the mechanism being unlocked by a piece of muscle wire.
We are no strangers to muscle wire (AKA Nitinol wire or Shape Metal Alloy, as it is sometimes called) here: we have posted on its use in making strange robots, robotic worms and walls that breathe. Whatever you call it, it is fun stuff. It is normally a flexible wire, but when you apply a voltage, it heats up and contracts, much like the muscles in your body. Remove the voltage, and the wire cools and reverts to its former shape. In the Microsoft Surface Book, a single loop of this wire is used to retract the lock mechanism, releasing the tablet part.
Unfortunately, the teardown doesn’t go into much detail on how the impressive hinge of the Surface Book works. We would like to see more detail on how Microsoft engineered this into the small space that it occupies. The Verge offered some details in a post at launch, but not much in the way of specifics beyond calling it an “articulated hinge”.
UPDATE: This post was edited to clarify the way that muscle wire works. 11/4/15.
“It is normally stiff, inflexible wire, but it contracts when you apply a voltage across it,”
That’s completely wrong. Nitinol is normally bendy, but becomes rigid and springlike when you apply HEAT to it. Applying a voltage isn’t what triggers it, it’s the heat of the current going through the wire.
To understand how Nitinol behaves, imagine that you have a normal coiled spring and you stretch the spring to twice its length. You need to apply force because the spring is trying to pull back on you. Now, if this spring was made of Nitinol, and you lowered its temperature below the phase transition point, it would suddenly just give up and stop pulling at your fingers. It would become floppy and bendy, but, on increasing the temperature it would become the original spring again and start pulling.
And judging by the looks of it, this is a simple lever.
The hinge of the lever is down on the left from SW1. The nitinol wire is being stretched by the spring on the right, which keeps it extended in the relaxed state. Nitinol wire can be stretched up to 10% without permanent change, so the wire can pull something less than 1/10th its length back against the spring when a current is applied. It’ll probably only move a couple millimeterts, but it’s quite strong. A wire like that can lift a 1 kg dumbbell off the floor easily.
However, when a force is being applied while the wire is hot, the nitinol wire starts to creep and slowly forget its original shape, so after thousands of pulls it stops moving.
1/20ths its length, as it’s doubled back around a pulley. Although that means the stress on the wire is halved as well.
It’s not acting as a pulley. The other ends of the wires are both fixed, so the two lenghts of wire pull the same distance.
It’s wrapped around a post because both lenghts of wire need the same current through them, so they might as well be the same wire. It’s doubled up because as you said it halves the stress per wire.
Another property of nitinol is that its still superelastic in the activated state as well.
https://www.youtube.com/watch?v=bb8MJgd1MGY
That means it deforms with a nearly constant force. Unlike regular spring material that would increase in force the more you stretch it, nitinol has a constant pull.
That’s why it’s useful for things like tooth braces that would otherwise need re-tightening as the teeth shift. You take a nitinol alloy with a transition temperature below room temperature, and pre-tighten it to the constant force region, and it will keep that constant pull on the teeth without adjustments.
Of course it does have a hysteresis effect, as seen on the video, where going one way you get more force than going the other way, which is what normal springs don’t do.
That’s because the material is technically in the plastic deformation region, and the force you see is the crystal structure trying to align itself back to a lower energy state. It operates under a kind of thermal diffusion random walk kind of mechanism, so it takes some time.
In other words, if you made a coiled spring out of low temperature nitinol and stretched it to the constant force region, then let it back a few mm it would decrease in force, and then slowly increase in force as the atoms reconfigure. That makes it ill suited for fast dynamic applications like shock absorbers, but good for structural stuff like bra underwires or the spring in the hinge of your sunglasses.
Yup, I cursed my orthodontist every time he showed up at the chair with a wire sitting on a block of ice–it meant there was pain to come.
Dax, thanks for the note. I got my states mixed up in the post: I have updated it.
Judging by your note, you have used this stuff, so let me ask you this: if this wire has a tendency over time to forget its original shape, is this mechanism likely to stop working over time?
The mechanism is only operating for a few seconds at a time. With that sort of duty cycle, it’s going to take years and years to break it.
In my experiments with the wire under PID control, the range of motion shifted around 20% over several hours on several days. The wire became longer and longer.
The wire was loaded with a spring similiar to this setup, and the heating current was adjusted to adjust how much the wire pulled on the spring, so as to get it to settle into different positions. The PID loop was measuring the position, which in the end was all it would do because the wire was basically just driven on/off. It was continuously driven as far in and as far out as it would go without overheating the wire (current limiter), and the range was measured to see how it would shift.
The wire we used had a transition temperature around 90 C. If you got it hot enough that the wire went blue, it would remember the stretched out configuration it was in, and you couldn’t get the genie back in the bottle again because there’s no way to squeeze a wire from the ends to push it short again.
Thing was, it apparently kept doing that a tiny bit at any temperature above the transition point. That meant, the harder and further you pulled it in, the more quickly it would “get tired”.
This is the type of HAD article I like the most. The article is based on a good teardown and analysis and the comments are informative and thought provoking rather than snarky.
Thanks!
“Remove the voltage, and the wire cools and reverts to its former shape”
That isn’t accurate either. When it cools, it doesn’t do anything – it just becomes soft again. The spring is what returns it back to the stretched state.
Another way to understand intuitively how the SMA acts is to think of the collapsing bones toy:
http://www.mywoodentoys.com.au/images/collapsing-giraffe-yellow.jpg
There’s a thread going through all the pieces, and when the thread is loose the toy is floppy. When you tension the thread, all the parts try to align back to their original positions and the toy becomes springy. That’s the basic behaviour of the SMA.
In the SMA, a temperature driven re-ordering of the crystal structure is responsible for the effect. When the material is cold, the atoms can slide around when pushed but they become stuck in their dislocated positions. The crystal lattice becomes skewed and squished and deformed. The material is plastic, not springy.
When the temperature goes up, the atoms start to jostle around, become unstuck, and the crystal lattice starts to pull back towards its original shape – the material becomes springy and jumps like the wooden toy.
Then there’s some critical temperature where the thermal motion is so strong that the lattice is just randomized as if you just cut the string holding the wooden toy together – the material becomes soft and malleable – and when it cools back down it assumes a new ordered structure, and that’s when it memorizes a new shape.
In the braces example, Nitinol would be applying a constant force when it is BELOW the transformation temperature. In contrast, the Surface latch mechanism essentially requires zero force from the wire when cold, and then a strong force when heated. I’m unclear on whether the negatives of hysteresis, permanent stretching, etc. are relevant in this case, because the wire is not subjected to varying (or any) stresses when cold.
Afaik Nitinol can be tuned for different transition temperatures. It may be that it’s used below transition in braces. It is however used above the transition any time you need it to be actually springy – otherwise it’s kinda like solder wire – it just bends. There are SMAs with two transition temperatures that can remember two shapes, which are springy in the temperatures between.
But the hysteresis etc. was a bit off the point anyways.
The main concern here is that when you power cycle this kind of Nitinol actuator, it very slowly shifts from remembering the pulled in state to remembering the stretched out state, which would make it stop pulling in.
This is really neat. I’m not aware of SMAs getting used in any other mass-produced electronic device–anyone know of any examples?
Also, fun fact: NiTiNOL is short for Nickel Titanium Naval Ordinance Laboratory, named after the alloy constituents and the lab that discovered the effect.
Nitinol is apparently used in F-14 fighters and others in hydraulic line connectors, because it’s a hard metal that takes wear and tear, but it isn’t brittle, and it’s malleable so it makes a tight seal, and yet it springs back so you can wiggle the line without causing a leak.
Another use mentioned in the literature was cellphone antennas.
Think, those little pull-out whips you used to have in old phones were made of nitinol, because nitinol wouldn’t break when bent repeatedly.
Except on the Motorola Micro TAC the antenna was just a piece of plastic that did nothing.
The Air Hogs Switchblade RC toy uses it. I scavenged some SMA wire from a couple broken ones a few years ago. It has a part that moves under traditional spring power, and the SMA pulls a pin out to release it.
Ooh, I forgot! The Air Hogs helicopters that fire missiles use Nitinol too. Same thing–loop of SMA contracts and pulls the trigger on a spring-loaded device. It’s the perfect mechanism for this purpose, as it weighs almost nothing and doesn’t need long stroke or any precise control.
Low-temp stuff is used in orthodontia with a transition temperature just below body temp so that the wire is soft and flexible when put in and then springs into shape as it heats up.
I thought Nitonol used a lot of power as its geat that ‘activates’ it. I wonder how much current this uses and how efficient it is vs a simple solenoid, or a worm gear motor drive…
A simple solenoid is not as simple as a short piece of wire. A worm gear motor is even less simple. So, even if the current is large it’s only for a very short time, and it’s a massive tradeoff compared to the complexity of any other mechanism.
Well, I’ve done the calculations once for an electrically driven bicycle lock, and it’s basically impossible to do with a solenoid because the return spring has to be very strong to prevent anyone from bumping the lock open.
I basically went with a linear screw drive, and calculated the friction of a badly lubricated, but otherwise undamaged screw (friction cf = 0.5), and came up with a figure that the motor needs about 1.4 Watts for 2 seconds to pull the lock open.
That’s plenty power to heat a wire with. Most of the energy is lost in friction in the mechanism anyways. The muscle wire mechanism is so simple that it avoids all the friction losses in the reduction gear and screw, which compensates for the low efficiency of the wire itself and as a bonus the whole thing is incredibly robust. It doesn’t mind how you bang it around, and it can’t seize up from dirt.
Hey Dax:
Your comments about replacing linear screw mechanisms with Nitinol have me thinking: what are some common figures for length when contracted vs. relaxed? I’m wondering how much wire it would take to produce a range of motion of 2-3 inches? For bonus points, about how many watts would it take to heat that much wire?
The superplasticity gives you a stretch of about 10%, so less than that. You don’t want to use the full range of motion.
If you want large ranges of motion, you’re best off using a lever or other mechanical amplification. The wire itself doesn’t need to be very long, but if you want to use a plain wire then about a foot of it can be stretched a couple inches, and pull back almost the same.
The amount of power to heat the wire varies by how well the wire is cooled and how fast you want it to react. If you surround it with insulating mineral wool, it may use very little power, but there’s between 15-30 C hysteresis between pulling and releasing, so it might take a considerable time to let go. On the other hand, that means you don’t need to heat it quite so much to keep it pulling.
I mean, a foot can be stretched a couple cm. Wouldn’t want to stretch it more than 1″