Materials
Prince Rupert's Drop: A Bulletproof Head and an Explosive Tail
Prince Rupert's Drop is a tadpole-shaped piece of ordinary soda-lime glass, made by letting a blob of molten glass fall into cold water. The bulbous head is astonishingly tough — it shrugs off hammer blows that would pulverize a normal marble — yet if you so much as nick the wispy tail, the entire drop detonates into powder faster than the eye can follow. Both the toughness and the violence come from the same thing: a self-equilibrated field of residual stress frozen into the glass, with a compressed skin locked over a tensioned core.- Surface compression400–700 MPa
- Crack front speed1450–1900 m/s
- Full disintegration≈30 µs (5 cm drop)
- MaterialOrdinary soda-lime glass
- Peak stored strain energy≈1.8 MJ/m³
- First studiedRoyal Society, 1660s
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
A bulb you can't smash and a tail you dare not touch
Drip a pea-sized bead of ~1000 °C molten glass into a bucket of water and it freezes into a shape like a tadpole: a rounded head trailing a long, thin tail. The head is the party trick — historical demonstrators struck it with hammers and even fired pistols at it without breaking it. The tail is the punchline: snip, scratch, or snap it and the whole object disappears in an instant with a sharp crack, leaving only glass dust.
These were known across 17th-century Europe as Dutch tears or Batavian tears. They take their English name from Prince Rupert of the Rhine, who presented several to King Charles II in 1660; the King handed them to the newly founded Royal Society, where Robert Hooke and Robert Boyle puzzled over them. Hooke got remarkably close to the truth. The modern quantitative picture came from photoelastic stress measurements and high-speed photography — notably Chandrasekar & Chaudhri (Cambridge, 1994) and Aben et al. (2016), who pinned down the stresses and the crack speed.
The mechanism: freezing a stress field into glass
The whole effect is differential thermal contraction. Nothing is chemically special — it is the same glass as a jam jar. The magic is in how it cools:
- The outer skin hits cold water, plunges below the glass-transition temperature (~520–560 °C for soda-lime), and sets solid — while the interior is still a hot, viscous liquid that simply flows to relieve any stress.
- Now the trapped-hot core cools in turn. It wants to shrink, but the already-rigid shell will not let it contract.
- The shrinking core pulls the shell inward. Force balance leaves the skin in biaxial compression and the core in tension — a self-equilibrated field where
∫σ dA = 0over any cross-section.
An order-of-magnitude estimate for the frozen-in stress is σ ≈ E·α·ΔT/(1−ν). With E ≈ 70 GPa, α ≈ 9×10⁻⁶ °C⁻¹, ΔT ≈ 530 °C (transition-to-room), ν ≈ 0.22, this gives ≈430 MPa. The violent water quench adds transient gradient stresses that push the measured surface compression to 400–700 MPa — several times harsher than industrial tempering.
Why the head is "bulletproof"
Glass never really fails in compression — it fails in tension, by opening one of the microscopic Griffith flaws that cover every real surface. A flaw only grows when its tip is pulled apart. The Griffith condition is σ_f = √(2Eγ / πa), equivalently an energy-release rate G = πσ²a/E ≥ G_c.
Here is the point: before any applied load can put a surface flaw into tension, it must first cancel the 400–700 MPa of built-in compression. Ordinary annealed glass fractures at a practical tensile strength of only ~50 MPa (K_IC ≈ 0.7–0.8 MPa·√m). So the pre-stressed skin raises the bulb's effective surface strength by roughly a factor of ten. A hammer blow on the domed head is mostly compressive contact loading anyway — exactly the direction the skin is built to absorb. The head is not made of harder material; it is ordinary glass wearing armour made of its own locked-in stress.
Why the tail is the detonator
The compressive skin is a shield only as long as it is intact and thick. In the head it is deep and the tensile core is buried far inside. In the tail the glass is thin, so the protective skin is thin too and the tensile core sits just beneath the surface.
Break or nick the tail and a crack instantly reaches that tension zone. From that moment nothing external is needed — the crack is fed by the elastic strain energy already stored in the glass. The strain-energy density is u = σ²/2E; at σ = 500 MPa that is u ≈ (5×10⁸)²/(2·7×10¹⁰) ≈ 1.8 MJ/m³ packed into the core. The tensile core runs continuously from tail to head, so it acts as a single fuel line: light it at the tail and the flame has a clear path all the way to the bulb.
The explosive disintegration — the fast event
Once released, the crack races up the tensile core, and its speed is capped by the material's elastic wave speeds. The theoretical ceiling is the Rayleigh surface-wave speed (c_R ≈ 3100 m/s in glass), but a brittle crack becomes unstable above roughly 0.5–0.6·c_R and starts to branch. High-speed cameras (up to ~1 million frames/s) clock the fracture front at 1450–1900 m/s.
Every branch spawns more branches — an exponential cascade that dices the entire tensile core in one continuous run. For a 5 cm drop the timescale is t ≈ L/v ≈ 0.05 m ÷ 1700 m/s ≈ 30 µs: essentially instantaneous. All the kinetic energy of the flying fragments comes from the released residual strain energy, not from heat, trapped gas, or any stored pressure — the glass was stone-cold and solid the whole time.
Where this shows up in real engineering
The drop is the extreme, uncontrolled version of a workhorse technology: thermal tempering. Toughened glass is heated to ~620–650 °C and quenched with air jets to freeze in a gentler ~90–150 MPa surface compression. Codes such as EN 12150 and ASTM C1048 rely on it — tempered glass is ~4–5× stronger than annealed and, when it finally does break, dices into blunt cubes (the drop's disintegration, tamed) instead of dagger-like shards. That is why it is mandated for car side windows, shower doors, and balustrades. Chemically strengthened phone cover glass reaches 700–1000 MPa by ion exchange, but only 30–50 µm deep.
The dark side is the same mechanism running away: a deep scratch or a hidden nickel-sulfide inclusion that reaches the tension zone can make a tempered pane explode spontaneously — the reason facade glass is heat-soak tested. The identical trick appears in metals as shot peening, which hammers a compressive skin onto turbine blades and springs to fight fatigue cracking.
| Property | Prince Rupert's Drop | Thermally Tempered Glass | Chemically Strengthened Glass |
|---|---|---|---|
| How it's made | Molten glass drip-quenched in cold water | Reheat to ~620–650 °C, quench with air jets | Ion exchange (K⁺↔Na⁺) in a molten-salt bath |
| Surface compression | 400–700 MPa | 90–150 MPa | 700–1000 MPa |
| Compression depth | Thick — ~10–20% of the radius | ~20% of the thickness | Very thin — ~30–50 µm |
| Failure mode | Explosive disintegration into powder | Dices into blunt cubes | Fractures if a flaw beats the shallow layer |
| Stored elastic energy | Very high (violent quench) | Moderate, controlled | Low total (thin layer) |
| Typical use | Physics demonstration / curiosity | Car side glass, doors, balustrades | Phone & tablet cover glass |
Frequently asked questions
Why doesn't the head break when you hit it with a hammer?
Its outer skin is under 400–700 MPa of built-in compression. Glass only fractures when a surface flaw is pulled into tension, so an impact must first overcome all that compression before any crack can even begin to open. A blow on the rounded bulb is also mostly compressive contact loading, which the skin is designed to absorb.
Why does snipping the tail destroy the entire drop?
The tail is thin, so its protective compressive skin is thin and the tensile core lies just under the surface. A nick reaches the tension zone immediately, and from there the stored elastic strain energy (~1.8 MJ/m³) drives a self-propagating, repeatedly branching crack all the way up the continuous tensile core into the head — in about 30 microseconds.
Is it made of some special super-strong glass?
No — that is the most common misconception. It is ordinary soda-lime glass, chemically identical to a bottle or a marble. All of its remarkable behaviour comes from the residual-stress field frozen in during the fast water quench, not from any special composition or hardness. It also does not store heat or trapped gas; the energy is purely elastic.
How fast is the crack, and what limits its speed?
Measured fracture fronts travel about 1450–1900 m/s. The hard ceiling is the Rayleigh surface-wave speed of glass (~3100 m/s), but brittle cracks go unstable and branch above roughly half that speed, so the front cannot outrun ~0.6·c_R without splitting into a cascade of branches.
How is this related to the tempered glass in my car and shower?
Same physics, dialed down and controlled. Tempering heats glass and quenches it with air to lock in ~90–150 MPa of surface compression — far gentler than a water quench. That makes it several times stronger and, crucially, makes it dice into harmless cubes when it finally breaks, which is exactly the drop's disintegration held on a leash.
Who discovered Prince Rupert's Drops?
They were known in Europe as 'Dutch' or 'Batavian tears' before 1660, when Prince Rupert of the Rhine gave several to King Charles II. He passed them to the Royal Society, where Hooke and Boyle studied them. The quantitative modern explanation came much later from photoelasticity and high-speed photography (Chandrasekar & Chaudhri, 1994; Aben et al., 2016).