Here’s another physics phenomenon that, despite being fascinating, I still can’t quite get my head around. It’s Prince Rupert’s Drop, a phenomenon created when molten glass is dripped in a tadpole-like shape into ice-cold water. This object and its peculiar physical properties have been known since at least 1625, and are seen in this don’t-miss video by Destin of “Smarter Every Day,” famous for demonstrating how cats right themselves when dropped upside down. But this is not Cat Drop, but Prince Rupert’s Drop. Do watch it: the ultra-slow-motion videos (130,000 frames/sec) are amazing.
After watching this and reading about it, I still don’t understand it completely, so readers who do can weigh in. (Note: you have to explain it better than Wikipedia or Destin!)
Of course, as soon as I saw this, my first impulse was to go online and buy one of these drops, but for obvious reasons they’re not readily available. You can buy them at one place, but they can’t be sent to the U.S. Bummer!
My attempt at ‘splaining in non-technical-speak:
The whole drop exists under huge internal pressure. The drop is unable to shrink normally as it cools because the surface was hardened so fast in the ice water. That puts the entire internal structure into a sort of linked state of tension.
Once you snip the tail, all the links are disrupted in a chain reaction.
So the kicker is that the tail doesn’t have this internal pressure because it rapidly cools down and changes state and therefore breaks easily invoking a chain reaction?
No, I think the tail is just thin and therefor easier to snap, which triggers the chain reaction.
Shouldn’t the initial hit on the head of the drop then also start this chain reaction?
Small splinters of glass are breaking off before the tail when you see it in slow-mo…or maybe that just dust’n stuff…
You can start with the tail because there it is thin enough to snap. It isn’t broken by impact but by snapping. And then it is all just an explosion of released stresses.
I think I get it now. The splinters breaking off the head doesn’t break the outside layer and thus the energy from the intial hit causes the wole setup to wriggle and that wriggle is enough to snap the fragile tail.
Right.
So I was wrong about vibrations and the Beach Boys. The Wiggles are responsible:
Let’s Wiggle
Link fail by me 🙂
Rats. I was looking forward to the wiggles.
Go to Amazon.com and search for “Lets Wiggle”
Sorry, NewEnglandBob… link fail.
In techno-speak it is the constant fracture speed that tests that this is fed by a near constant stored energy density from the tension.
At least in the video they seem to purposefully leave the tail out of the tank, having ordinary fragile glass to set off the fracture process with.
I guess if they don’t do that they can still try to break the tail as it is the most fragile part. But the actual break would start at a scratch (or some other defect) anywhere along the drop, ruining the slow-mo effect.
At the tail the glass molecules don’t get the chance to line up as orderly with their neighbors, since they don’t have as many neighbors to actually line up with. The material strength of glass relies on orderly crystal structures that allow each molecule to contribute the most resilience (in a crystal each molecule has a bond with each of its neighbors). Metals derive their strength in a similar way, although those crystal structures are mostly achieved through the mixing of alloys.
If you could actually break the head and disrupt the crystal structure there it would also release the tension and cause a similar explosion, but the glass is very hard up there.
The best local structure of the glass varies as a function of temperature, but your discussion of the glass has some problems. The glass doesn’t consist of discrete molecules, glasses have silicate network structures. While they have local order (i.e., the Si–O–Si links have a narrow distribution of Si–O bond distances and the Si–O–Si angles have a somewhat broader distribution). Glasses don’t have “orderly crystal structures” – at least not beyond the first few neighbors (interatomic pair distribution functions quickly smooth out after several interatomic bond distances).
Note also that you can’t make Prince Rupert’s drops out of pure silica (melted quartz) because the thermal expansion coefficient of pure SiO₂ is too close to zero. You csan usually take silica glass right out of a glassblowing torch flame (we use an H₂/O₂ torch for silica) and plunge it into runing water without it cracking.
Another curious thing that we used to do with silica is to take a rod perhaps 10cm thick, melt a section in an oxy-hydrogen flame, and draw it out to 1-2mm diameter. When cool, you can then bend this section into a loop no more than 10-12mm diameter. Now stroke it once, very lightly, with a finger. Now when you try to bend it it snaps almost immediately. Like the glass drop, the slightest imperfection in the surface destroys the integrity of the whole structure.
My materials knowledge is pretty old and unused. Thank you for the enhancement. I looked at metals quite a bit, but it makes sense that glass would have some substantial differences.
The absolute amount of the internal tension in the tail would be smaller – because the tail is thinnner. But the relative amount of stored energy (joules/kg, joules/mole or joules/m³ ; I’m not sure if it’d be specific (per weight), molar, or volumetric) would probably be similar.
The fibres in “glass fibre” (of the rarely-named-in-full glass fibre reinforced plastic) are under sufficient stresses from their high cooling rates that they’re potentially subject to the same disintegration if the annealing is done wrongly.
Thanks, so rapid cool is key. That would explain why the end of the tail on the very first drop can break off without starting the reaction. It wasn’t cooled down as quickly.
Jesus can only get in through the stressed inner belly, once It gets in the whole thing blows.
You guys catch the little bible quote at the end? “Psalms 111:2”.
Otherwise, pretty damn cool.
Yes, I saw that too at the end of the video.
Yeah, I noticed that too, about a year ago when I first saw this (and a few of his other videos). Call me oversensitive, but it put me off my feed enough that I unsubbed.
Oh, and you just know he’s a big fan of Tim Tebow.
It seems to me that it’s related to the phenomenon of the candy bar wrapper. You can pull at that sucker and you can’t tear it but if you make the tiniest nick with your teeth it just glides apart with no effort.
When two adjacent molecules separate the bonding energy associated with their separation transfers to the next pair and on and on like a zipper.
It’s why hotrodders polish the connecting rods in the engines to make them stronger. A smooth surface is less likely for a crack to start and propagate.
I get why it’s called Prince Rupert’s Drop in English, because Prince Rupert took one to England to show Charles II (presumably while Charles still had his head on). But what is it called in Germany?
Bologneser Träne, apparently.
Correct.
in case somebody is interested there is something similar to Prince Rupert’s drop called the “Bologna bottle”. Strong enough that you can use the bottle as a hammer to drive nails into wood. But if you drop the nail into the bottle: it shatters.
This effect even has a practical application. The glass in your car very likely is special glass that is manufactured to have this kind of internal stress. If is breaks it is designed to shatter into lots of small fragments which are less dangerous than big chunks.
Charlie One was the one who got an object lesson in the power of democracy ; Charlie Two got to [ahemm] disport himself with orange sellers, charter the Royal Society, and die in his bed (or was it one of his mistresses? so many kings, so few novel stories) ; the presumptive Charlie Three talks to pot plants and may be more damaging for the institution of the Monarchy than his namesakes combined.
Viva Charlie Three!
What they did with them in the court was to get a bunch of court types together and then have a naïve person hold the bulb in their fist. There would have been much teasing from those in the know and assurances that it wouldn’t really hurt just to build the suspense, I imagine.
The near instantaneous expansion would be unexpected and pretty much guarantee the individual’s reaction that everyone else was expecting to see.
The particles that result are like sand or dust so there isn’t much chance of injury, apparently.
The Beach Boys are responsible:
“…Good, good, good, good vibrations…”
The best everyday analog I can come up with on short notice is a bicycle wheel. Suppose you were to slice the rim of the wheel into dozens of short arcs, each with one spoke attached. Then arrange the segments back into their original circular shape and start tightening up the spokes little by little. Make them tight enough (and do it carefully enough) and you end up with a perfectly serviceable wheel that’s just as strong as the original. But break just one spoke and the whole thing disintegrates catastrophically.
With a little creativity (and a lot of patience) you could probably model this with, say, napkin rings and rubber bands, making a pretty rigid and stable structure that flies apart if you snip one rubber band.
From the video, hopefully you get that somehow the glass is internally under tension, which is suddenly released in the explosion. But really “getting it” probably requires about three pieces of understanding:
(1) Presumably you know that most materials expand/contract as they are heated/cooled. Now imagine a meter long piece of steel cable heated up to about 1500 deg C, so by expansion is now about 1.02 meters long. Now anchor the ends somehow and let it cool. It wants to shrink down to a meter, but is not permitted to. Basically it is simultaneously contracting due to cooling and being stretched to maintain the same length. Now imagine the kind of force it would take to actually just stretch a meter long steel cable measurably, let alone 2cm. That same amount of tension would build up in the steel cable as cools with ends anchored. This is the first piece of understanding where the interal tension comes from. **
(2) The outside cools & solidifies first. The important thing to realize here is glass is not very compressible, so once you have an outer later solidified, it is not much going to want to change shape. ***
(3) The inside then cools & tries to shrink, but is anchored to the outside layer, so it is simultaneously stretched (like the cooling steel cable). Why doesn’t the outer layer just crack and get pulled in & wrinkled up? Because of the cirular cross-section & all parts of the solid outer layer being pulled in simultaneously — this means the molecules of the outer layers are crowding against each other, resisting the inward pull with the compression forces against each other — this is where the roman arch analogy comes in. Geing pulled inward simulataneously & crowding against each other they have nowhere to go.
So the inside of the cooled drop is like millions of stretch tiny springs, and the outside is like millions of compressed tiny springs.
They are in a fragile stability, because each “spring” is prevented from releasing by the neighboring springs — force one spring out of alignment and it pops, releasing its neighbors, which pop, releasing their neighbors, and so on.
A similar process will happen with any molten material suddenly cooled, but whether it will do anything quite so dramatic as Rupert’s Drop would depend on the specific properties of the material and the cooling process.
—
** I should note that I am not sure that this thought experiment about the steel cable would work as described (and if possible probably depends on the rate of cooling). The ductility of steel might mean the cable simultaneously contracts due to cooling, is stretched because it is not allowed to contract, but is permanently deformed due to the ductility of steel thus dissipating the tension that would otherwise build up.
*** A good question to ask on point (2) would be “shouldn’t the outer layer shrink when it cools?” And the answer would be that it does, it cools/shrinks/hardens and because the interior is still hot and not shrunken the shrinking outer layer probably cracks but those cracks are immediately filled in by more molten material from the interior which cools & hardens. This process of cracks forming & filling in will stabilize when you have a cool hard outer layer which exactly fits the remaining hot molten interior, so no more cracks form and the process described in (3) begins.
Moissan encountered a related phenomenon in his attempts at diamond manufacture in the 1880s. He’d dissolve carbon in iron, then drop the molten iron into liquid lead (not, contrary to Wikipedia, water ; it boils too easily and doesn’t have the thermal conductivity desired (what’s that spit-on-the-stove phenomenon called again?)), in the expectation of creating transient high pressures as the carbon was coming out of solution, and this crystallising to diamond. Logical ; he thought it worked ; but probably what he actually created were grains of silicon carbide (if his experiments weren’t “seeded by students, as has also been alleged). But as part of this work, he’d sometimes have his balls of iron explode from the trapped internal stresses – remembering that this was essentially a high-carbon cast iron, not steel, and intrinsically hard and brittle.
(I studied diamond formation for a thesis project many years ago – real geology, not this grubbing around in the crustal ephemera – there have been some real characters. I also discover that Robert Hazen has since written a book on the subject, “The Diamond Makers ; Author: Robert M. Hazen ; Published: August 1999, isbn: 9780521654746” ; on the basis of some of his other work on OOL – Origin Of Life, this has gone onto my wishlist!)
fascinating!
Booms, burns and flashy bits. What’s not to like?
” (what’s that spit-on-the-stove phenomenon called again?))”
That’s the Leidenfrost effect.
George
The very one!
Many an entertaining hour spent laying in the tent, cooking up a brew and pondering the interplay of feedback in the proper working of a Primus (paraffin pressure) stove.
“shouldn’t the outer layer shrink when it cools?”
No doubt it does, but I don’t see why that should be a problem. The outer layer doesn’t harden all at once as a closed shell. Rather, it forms progressively from head to tail as the stream of molten glass enters the water. So there’s always a soft trailing end through which the fluid interior can be squeezed out by the contracting surface.
Excellent point. I hadn’t thought of that. And now you point it out it seems somewhat more likely than the process of cracks forming and filling in. Although it does depend on the speed of sinking vs. the speed of hardening&shrinking. But I think your “squeezing out the tail” process is more likely to result in a sufficiently smooth, uniform (and therefore stable) hardening of the outer layer.
Did you note the video closes with a bible verse? Psalms 111:2.
Tail vs. Head
In principle you could start the chain reacion anywhere and sufficiently precise pressure applied to a single point of the head would probably do it. Also a blunt force, like the hammer, if fast and hard enough would probably start the chain reaction at the point of impact. But a weaker force — in sufficient to crack the head — will send a shockwave up the tail. Because the tail is tapered, the energy of the schockwave becomes more concentrated and at a sufficiently narrow point will be enough to crack the tail and start the chain reaction (see my previous explanation).
Maybe. But if you watch the slo-mo in the video, it looks like what happens is that the hammer blow knocks the head sideways, and that shear wave propagating up the tail is what snaps it.
I was using “shockwave” (incorrectly) to refer to the wave induced by the “shock” of the hammer strike. In any event, whether it is compression wave or shear wave or more likely a mixture of both, the power of the wave becomes increasingly concentrated as it travels down the tapering tip.
> You can buy them at one place, but they can’t be sent to the U.S. Bummer!
If you can get access to a muffle furnace, a crucible, tongs, goggles, smashed glass and a bucket of water, you can make them yourself, and as many as you want. Got any friends in the chemistry department?
I’d like to know if drilling into the bulb has the same effect – it should. Sadly, my muffle furnace broke down some years ago. I never worked enough with it to bother replacing it, and just for this one project? But now I wish I still had the old one. Bummer!
Where I live there are several glass artisans. I’m sure any of them would be willing to make some for a customer.
I’ve always wanted to ask for one, but haven’t yet.
So, if I accidentally break my little tow I could go poof.
I need to be more careful.
was your toe made of molten glass dropped in water? 😛
“Tow”. Sheeesh, dratted autospell.
You need to toe the line very carefully. It’s lucky you haven’t already gone off.
‘Tow the line’ is a common mistake known as an acorn where people mishear an expression.
Actually, it’s known as an eggcorn.
Well that’ll teach me to proofread more carefully. Pedantic is as pedantic does.
Surely that was on porpoise?
Don’t call me Shirley!
Sid Nagel in the UoC Physics department can probably make one for you if you ask nicely. He’s also got an ultra-high speed video to take a movie of it when it explodes.
It is basically the same reason you car windshield will break into little chunks rather than slivers. They take a formed piece of glass up past its annealing temperature, then cool the surface rapidly. The compressive forces in the surface layer are not as high as you get from a molten blob, but the principle is the same.
I’m afraid you’re wrong about car windshields. These consist of a plastic layer sandwiched between two normally annealed glass layers. The plastic layer is designed to have an optimum adhesion to the outer glass layers and when the windshield is struck and partially penetrated by an object, the inner plastic layer will partially adhere to the glass so that the stretched portion will adsorb energy and reduce the chance that the striking object will penetrate the windshield. I worked for the Dupont Company as a chemist for 29 years, and spent a little of that time working on the microstructure of Butacite, the plastic windshield interlayer made by Dupont.
On the other hand, the side windows of cars are made of tempered (rapidly cooled) glass. They are very strong but will shatter into small pieces if hit by a sharp object, such as a sharp stone. If windshields were also made of tempered glass they would shatter every time they were hit by a small stone kicked up by another vehicle, and wouldn’t last very long. Since cars don’t usually travel down highways sideways, tempered glass is OK for the side windows.
Cars from the sixties through to the 80s (roughly) used to have toughened windshields that would shatter as described. Later ones were ‘zone-toughened’ which meant that an area in front of the driver would shatter into larger pieces you could see through to steer the car rather than driving blind. The great thing about such windshields was, it took a hefty-sized rock to break them. There are Youtube videos of wannabe car thieves whacking toughened side windows enthusiastically and repeatedly with a hammer and the hammer just bouncing off.
The other thing about them was, the pieces they shattered into had less sharp edges; the small bits would break with roughly square edges rather than razor edges. Still not to be handled carelessly on account of tiny splinters, but at least no big sharp-edged guillotiney chunks.
Laminated windshields have two thin layers of glass and a tough plastic layer between. They won’t shatter, but on the other hand it takes depressingly little force to start a crack in one.
20 years ago I was at Monsanto as a hired gun… er… IT consultant. Part of the time I worked with the guys who made Saflex[tm], Monsanto’s version of polyvinyl butyral.
http://en.wikipedia.org/wiki/Polyvinyl_butyral
Isn’t this an analagous construction that anyone can make: the stick bomb?
The important point isn’t the amount of energy that you put into the tail – or wherever you choose to initiate the equilibration – it’s making that first defect in the surface, from which the internal stresses can tear the object apart.
An interesting experiment would be to put a drop of hydrofluoric acid (strongly not recommended to be anywhere near this stuff unless you’re appropriately petrified of it – it does really horrible things, starting off by destroying your nerves’ chemistry before getting to work on your flesh and bones) anywhere on the Drop’s surface, and letting it do it’s stuff. At some point it will either initiate a crack of sufficient magnitude, or excavate a (smooth) pit deep enough for the exposed stresses to overcome the limit of the glass’ (considerable) strength.
(I have a vague memory, from nefarious chemical experimentation as a youth, of hearing something similar used as a time-delay fuze. It sounded really dodgy then, and I explored electrical ignitions instead. Pardon?)
I wonder … how would the traditional soprano and wine glass trick go with one of these?
Hydrofluoric acid – or rather its gas HF – was the initial topic of Derek Lowe’s entertaining ‘Things I Won’t Work With’ column:
http://pipeline.corante.com/archives/2004/03/03/things_i_wont_touch_1.php
You won’t be surprised to learn that I know that column. Turning asbestos and brick into a raging inferno sounds … interesting.
I have a print out of that “Ignition” book. A veritable barrel of laughs it is too.
I feel sorry for poor Gigi…
Obvious reasons, yes. But possibly not insurmountable.
Try this : make a waterproof box of appropriate size, lined with something like “fleece” material ; fill with water (or possibly better, something syrupy) ; put drop in ; close, seal in several layers and label as “industrial ball bearings – please drop as often as possible” (we all know about baggage handlers and “Fragile” labels). Sure, you’d get an attrition rate, but it should be safe enough to ship.
Why the US in particular? Probably fear of company-destroying litigation. Get your online source to send it to a contact in a country to whom they do ship, and the contact slaps a new address label onto the box for Chicago.
Despite the dramatic video, the actual stored energy isn’t great. Given the packaging appropriate to delicate glassware (I had three Kleinsteins sent from the States a couple of years ago – no breakage in transit, and they’re not exactly easily annealed either), then the likelihood of anything exiting the packaging except a stream of powdered glass is minimal. Sure, there would be an attrition rate … but you’d hardly be surprised, would you.
Customs declaration : “gift – delicate glass artwork” would be entirely accurate.
If they ship them anywhere, then they’ve obviously got the physical aspects of the shipping sorted, so it would be that liability thing that is likely to be the [ehemmm] killer.
There’s a credible possibility that they have a half-life. So does fish, and you can get that sent by mail.
http://www.wikihow.com/Bag-and-Ship-Live-Fish
No big deal; it’s the same effect exploited in tempered glass. Rapid cooling = incredible stress in the glass. The glass will still be very tough, but once fractured it shatters into tiny pieces.
See the wikipedia entry for “toughened_glass”.
It is also fairly trivial to start the shattering from the large end of the drop. All you have to do is attempt to scribe the surface with a diamond (or any other thin sharp object which is much harder than glass). Traditionally of course you pinch the tail because that part is easily damaged without special tools – even fingers will do. Now if you want to be mean, create a drop and anneal it …
I read about these in the mid sixties, when I was in my mid teens, and had a home chemistry set. I made lots of them with a bunsen burner and soda glass rod, but they were small examples, only a few millimetres across and with short tails. My tap water at the time was very cold, about 6C.
A normal pair of engineering pliers would snap the tails and cause the whole thing to powder with an audible snap. I vividly remember the strange site of red hot glass under water with no boiling water bubbles around it.
In those days we had no safety goggles: I was probably just lucky.
Reblogged this on Mark Solock Blog.
I am NOT an expert on materials but…
It’s very similar to toughened car window glass. Now cracks need a lot more energy to get started than they do to propagate.
Toughened glass is under compression at the surface (so any surface micro-cracks starting will try to close up rather than extend) and tension in the middle. However, if damage occurs sufficient to penetrate the outer layer, starting a large enough crack in the inner tension zone, then that crack will race through the tension zone at extraordinary speed and the whole glass pane will ‘explode’.
J E Gordon’s fascinating book “The New Science of Strong Materials” and its companion volume “Structures” explains this pretty well, and by the way they’d also be an excellent read for any biologist who wants to understand how and why nature / evolution uses natural materials like wood and bone and tendon.
My doctor (M.D.) nephew once asked me if I knew why it hurts when you eat ice cream too fast. I relied, because your hard palate contracts. “No,” he replied, “because it gets cold!” I had to explain to him that most things contract when they get colder (and expand when they get warmer), including bones, and a cold palate that wants to contract will pull on the other, warmer parts of the skull with tremendous force, causing pain – simple engineering/physics stuff (thermal stress being one of the primary reasons things break, such as turbines). What is it with the people in biology-related fields that they don’t take any physics courses? (My nephew didn’t even take high school physics.)
I’m afraid this sounds like something someone who new something about physics thought sounded good, and started repeating, but which has never been experiementally verified & little basis in reality.
Given the general painless ease with which soft tissues stretch and deform, it seems highly unlikely that thermal contraction will build up painful tensions.
Some combination of sudden contractions and expansions of blood vessels (perhaps due in part to thermal contraction/expansion, but more due to automatic responses to either prevent heat loss/compensate for heat loss) along with quirks of our nervous system causing referred pain, seems the correct explanation. (see http://en.wikipedia.org/wiki/Ice-cream_headache)