Alloys +

Useful information
Kit List: 

Cutlery, cheap stainless set more expensive stainless set (all spoons)
NiTinol springs
NiTinol Magic Tricks (Bending Paperclip and Heart Wire)
Normal steel spring
Heat gun and tongs
Other lumps of metal could be good (needs more thought)

Packing Away: 

This experiment now lives in the new small blue box called 'Periodic Table, Alloys and Carbon Allotropes', all three experiments may be done in combination. It's contained in a small Tupperware and a pencil case.
This experiment used to live in Misc box.

Frequency of use: 

EDIT: I (Yaron) am on tour at the moment but when I get back I will be working on a major overhaul for this experiment.

Also, I really want to take all the knives and forks out of the experiment since a kid threatened to STAB ME earlier. And we should replace the spoons with less rusty ones because it's difficult to demonstrate that stainless steel is oxidation resistant when all your SS samples have rusted.

Depending on how much steel is already in the demonstration, one could acquire a pearlite and martensite sample and compare their physical properties (strength, stiffness, hardness). This may have previously been a part of the demonstration but has since been taken out.

Finally, we should get a couple more samples to talk about because the demonstration is a little short. Perhaps an Al-Cu alloy to talk about age hardening? Can anyone convince the Materials Department to donate a single crystal turbine blade?
UPDATE: Yes, we may be able to get a defective sample from Rolls Royce. I discussed it briefly with Dr Catherine Rae (cr18) and she says it can be arranged. Will need to email at some point.

I've left Yaron's little rant in because he raises some excellent points about this demo, which we should work on when we have time. I've added to the explanation to make this into a PLUS version, but I reckon the experiment could do with more samples to make it more engaging. I've tried to write this in sections so when you, the lovely demonstrator, come to do this you can pick and choose what you'd like to talk about. Ideally not all of it because even I don't think alloys are that interesting. Grace, 17/06/19

Background/Intro to alloys

What is an alloy? A metal which is a metal mixed with something. More rigorously: A metallic solid or liquid that is composed of a mixture of two or more metals, or of metals and nonmetal or metalloid elements, usually for the purpose of imparting or increasing specific characteristics or properties.

Alloys may be homogenous or inhomogenous depending on how the different metals interact. This will have a large effect on the properties of the alloy, so alloying additions need to be carefully chosen to ensure you optimise the properties and don't ruin the stock.

The composition and manufacturing conditions of the alloy will determine which phases are present in the metal. A phase is the specific arrangement of atoms in the unit cell of a crystal lattice. For example in steels, the austenite phase is a face-centred cubic arrangement of iron atoms while the ferrite phase (typically more stable at room temperature) is body-centred cubic. Phases can have different compositions, mechanical and magnetic/electrical properties because of the differences in structure. For example, the austenite phase of stainless steel is not magnetic, but the ferrite phase is, whilst martensitic stainless steel is extremely hard - might be a good point to discuss hardness v. toughness.

Ask if they know the difference between hardness and toughness. Most won't: toughness is a measure of the amount of impact energy it can take before fracturing, whereas hardness is a measure of its how difficult it is to scratch. This is related to strength, which is a measure of how hard it is to permanently plastically deform.

If they're (understandably) glazing over, maybe move on to thinking about why we might want specific properties and what we can do to get them - composition, heat treatment, developing a particular microstructure etc.


There are (or should be) a few phase diagrams in the box, which could do with laminating, that can be used to illustrate the different regions and what phases you'd expect to find in under certain conditions. I think I put a steel one in there (because everyone loves steel) and some nice simple binary eutectic one (?) - all composition-temperature ones. Ask what they expect a material to look like under a microscope - all the same or different regions? Then introduce the micrographs and what they can see in them (also could do to be laminated, I can't remember exactly which ones I left in there...). They should notice the grains, some annealing twins in a brass, the grain boundaries and possibly notice the different appearances of the different phases. You can talk about how different phases form different grains, and discuss solidification (if they're really keen) using the phase diagram.

The point isn't to bore them to tears by discussing alloys at length - just to go beyond the A-level "alloys are a mixture of different metals" idea and show them that they're really complex and cool (if you like that kind of thing).

If we could get hold of some turbine blade, silicon wafers, copper alloys and similar materials stuff, that'd be great for discussing the specific applications of alloys (I am aware that single crystal pure Si isn't an alloy, but it's cool and shiny) and some materials processing methods (Czochralski process, quenching to get martensite, age hardening etc.), which would definitely make this section more engaging.

There's a model of some hcp layers (looks like a Christmas tree kinda) that you can use to illustrate the idea of bcc, hcp and fcc structures. You can use this to explain the twins you see in the brass micrograph - they occur due to a stacking fault. Stacking faults can be illustrated by organising the layers of "atoms" in the model out of sequence to show that they can stack in different ways. The "twin" comes from the symmetry that arises either side of the twin boundary:



Note the shape of the twins - they're squarish. Compare to the idea of deformation twins, which are lenticular. Annealing twins arise from growth accidents at high temperatures. Deformation twins need to be lenticular to minimise strain energy, whilst annealing twins don't need to do this because there's no strain energy associated with their formation.


Say that you can tell how expensive someone's cutlery is from whether it is magnetic. Good stainless steel contains Cr and Ni, the Ni stabilizes austenite phase, which is not magnetic. Bad stainless steel contains just Cr, this means the ferrite (magnetic) phase is stable and therefore cheap cutlery is magnetic. The proportions are usually 18:10, 18:8 or 18:0 Cr:Ni - the higher the Ni content the higher the quality. Show that the good John Lewis stainless steel is non-magnetic and the cheap Asda stainless steel is magnetic. The result of this means the Asda cutlery scratches more easily, which makes it look less shiny. They might say that the Asda cutlery is bendy (stop them before they break it) - this is to do with the manufacturing method and not the steel they've used!

Shape memory effect

Finally, demonstrate the shape memory alloys. Ask members of the audience to deform the NiTinol wire sample. After this, tell them you will return it to its original shape. Heat up using the heat gun, holding the wire using a pair of tongs (we will need to get tongs). You could ask an audience member to hold the sample in the tongs - they should wear gloves and the demonstrator should remain in control of the heat gun at all times.

If the wire has been tangled by an ambitious member of the audience, you may need to untangle it, as this may prevent the wire from uncoiling fully. The ideal geometry is to curl the wire into a spring. The heart magic trick is the best at springing back - the paperclip is a bit dead and doesn't look much like a paperclip any more. For curious audience members, there are some (slightly manky) normal springs that you could heat up for comparison. Ask what they expect to happen and then show them that it doesn't return to its original shape like the NiTinol ones do. There are a couple of springs that have been deformed permanently - show them that this isn't recoverable for a normal alloy, which they should ideally recall from work they've done on Hooke's law and elastic limits.

You'll also need to explain why it happens! I've uploaded an image that you could use/copy out. You'll need to know what a phase is for this to make any sense, and that phases can have different crystal structures (the Christmas tree model is useful here). The important thing about shape memory materials is that they can undergo a diffusionless transformation from a parent phase (austenite) into a low-symmetry metastable martensite phase. The names are taken from the phases of steel - shape memory materials are not steels - it's just that materials scientists aren't that inventive. Transformations between the austenite and martensite phases can be induces by temperature or stresses. You can expose the material to a temperature or apply a stress such that the parent phase becomes unstable with respect to the martensite phase, and the material transforms. This also works in reverse, when you remove the stress/temperature driving force, and you can get martensite -> austenite transformations as a result.

This is shown schematically in a v. poorly hand-drawn graph I've uploaded. I will leave this in the box :)

The shape memory effect depends on temperature. You can train your alloy in the austenite phase and recover deformation that occurs in the martensite phase.

1. Heat the alloy to above AF (i.e. above the austenite finish temperature, where the sample is entirely austenitic).

2. Rapidly cool the alloy into the martensitic region. The sample will transform such that minimal shape change occurs and the sample will look have approximately the same dimensions as when it was austenitic. This is because the martensite has adopted a twinned structure - you don't really need to worry about what that means for the purposes of the demo, only that it's an easy way for the martensite to deform by detwinning).

3. Shear the martensitic sample (i.e. what your volunteer is v. kindly doing for you). This detwins the martensite, which is important to us because the new detwinned structure is crystallographically related to the original austenite we had to begin with.

4. Heat it back up to above AF again (all austenitic). We said that the detwinned structure is related to the austenite we started off with, which means we can basically recover the original austenite grains we had in the first place - and the same shape! As the sample cools, it will return to the twinned martensite structure, but we can't see that because it will transform in order to generate the smallest dimensional change.

If they're still listening, then ask why you might want this effect at all:

1. Stents - nitinol/iron-platinum alloys. AF ~ 30 degrees (i.e. <37 degrees). Crimped while martensitic, then guided into the body (the heart), where it then expands as it transforms back into the austenite phase. No need for the balloon stuff you have to do with SS stents, and superelasticity means it recovers its shape with the high loads applied to it.

2. Morphing structures in planes - adapting chevrons in turbines. Change shape in response to temperature. Take-off and climb = high temperature, chevrons deployed to reduce noise and improve air mixing. Cruising = lower temperature, chevrons retract to maximise efficiency. I'll admit that this kinda needs a photo to be more interesting.

3. Braces - dental wires. Can be used to pull the teeth when the wire heats up (which I think sounds nasty, but y'know).

4. Fixing broken bones - kinda like the stent stuff but with bones. Plates used to apply a compressive force to bones to help knit them back together, which works because the shape memory alloy is heated when in the body.

5. Glasses - if you squish glasses that have shape-memory frames, you can just heat them back up and they should spring back into shape.

You could also compare this behaviour with a standard paperclip (you'll need one per demo because they're gonna break it). Get them to bend it out of shape and then ask them to bend it back into the original paperclip shape. They definitely shouldn't be able to do that. With an ordinary alloy, like the paperclip, bending introduces dislocations, which are awkward to explain. They are essentially defects that move through the material and allow it to shear - it's a mechanism by which a lot of metals tend to deform (the alternative is by deformation twinning, popular in hexagonal structures where there are very few slip systems). The problem is that you're introducing lots of dislocations in the paperclip that mean it can't just deform straight back into its original shape - the dislocations get in the way. When the shape memory alloy is deformed, you do introduce dislocations, but not as many. It will, eventually, stop cycling through the shape memory cycle, but it takes time. One of the ones we have (I think the paperclip) is kinda getting there now - best to use the heart.

Potential addition

If we need more shape memory samples, the Materials Department gives them away for free on open days and Rob said we could just ask him for them. Would also be nice to get a superelesticity demo if we can, but we'd need to do a Hooke's law F vs. extension style thing at a raised temperature and I'm not sure how feasible that would be - maybe a hairdryer, perspex tube, clamp stand and mass hanger would do the trick. Both shape memory and superelasticity exploit the same martensitic phase transformation, so they'd be good to do together.

Risk Assessment
Date risk assesment last checked: 
Sat, 11/01/2020
Risk assesment checked by: 
Emma Crickmore
Date risk assesment double checked: 
Mon, 20/01/2020
Risk assesment double-checked by: 
Beatrix Huissoon
Risk Assessment: 


Hazard Affected Person(s) Risk Likelihood Severity Overall Mitigation Likelihood Severity Overall
Heat gun All Fire risk and also the possibility of burns. 3 3 9 Demonstrator to control heat gun. Do not leave on. Keep flammable material away from the heat gun. Use stand instead of lying heat gun on a surface. Do not touch the heat gun.
In case of burns, run affected area under cold tap for 10 minutes. Call a first aider. Follow venue RA protocols in case of fire.
2 3 6
NiTinol wire and tongs hot when heated All Risk of burns. 3 3 9 Do not let anyone near the heated wire. The wire is thin and should cool within a few seconds, but care should be taken with the tongs. If possible, obtain a heat-resistant mat to lay the tongs/wire on after heating.
In case of burns, run affected area under cold tap for 10 minutes. Call a first aider. Follow venue RA protocols in case of fire.
2 3 6
Sharp ends of wire All Cuts 4 2 8 Double back sharp ends 3 2 6
Magnets Public Skin getting caught between cutlery and magnet 3 3 9 Use weak magnet so won’t cause harm if occurs 3 1 3
This experiment contains mains electrical parts, see separate risk assessment.
Experiment photos: