The Physics of Elasticity: Why Rubber Bounces, Glass Shatters, and Steel Springs Back

Everything Deforms

Pick up a pencil and bend it. You can feel it flex. Release it, and it springs back to its original shape. Now bend it further — past a certain point, it snaps.

That sequence — flex, spring back, break — captures the three regimes of mechanical behaviour that govern every solid object in the universe. Elastic deformation (reversible), plastic deformation (permanent), and fracture (catastrophic failure). Every material goes through some or all of these stages when you push it hard enough. The details — how much force, how much bending, how sudden the failure — depend on what the material is made of at the atomic level.

This is the physics of elasticity. It’s the reason bridges stand, springs work, trampolines bounce, and your bones don’t shatter every time you take a step. It’s also the reason buildings collapse in earthquakes, aeroplane wings flex in turbulence (by design), and you can’t unbend a paperclip back to its original straightness no matter how carefully you try.

Hooke’s Law: The Spring of All Things

In 1678, Robert Hooke published one of the simplest and most useful equations in physics:

F = kx

The force needed to stretch or compress an elastic object is proportional to the displacement. Double the stretch, double the force. Triple the stretch, triple the force. The constant k — the spring constant — measures how stiff the object is. A car suspension spring has a high k. A slinky has a low k.

Hooke originally published this as an anagram — “ceiiinosssttuv” — which rearranges to “ut tensio, sic vis” (as the extension, so the force). Physicists in the 17th century were fond of priority games.

The law is an approximation, and Hooke knew it. It works only for small deformations — the linear elastic regime. If you stretch a spring to three times its length, the force is no longer proportional to displacement. But for the kinds of deformations that matter in most engineering applications — beams flexing under load, wires stretching under tension, columns compressing under weight — Hooke’s law is extraordinarily accurate.

The physical reason is straightforward. Atoms in a solid sit in potential energy wells — each atom is at an equilibrium position where the attractive and repulsive forces from its neighbours balance. Displace an atom slightly, and it experiences a restoring force proportional to the displacement. This is exactly a spring-like (harmonic) force. For small displacements, every atomic bond behaves like a tiny Hookean spring. Hooke’s law is the macroscopic manifestation of this atomic-scale elasticity.

Push atoms too far from equilibrium, and the potential energy well becomes asymmetric — the restoring force is no longer proportional to displacement. That’s where Hooke’s law breaks down and more complex behaviour begins.

Stress, Strain, and Young’s Modulus

Hooke’s law as F = kx is fine for individual springs, but it’s useless for comparing materials. A thick steel rod and a thin steel wire have very different spring constants, but they’re made of the same stuff. We need a property that depends on the material, not the geometry.

This is where stress and strain come in.

Stress (σ) is force per unit area: σ = F/A. It measures how hard the material is being pushed, normalised by how much cross-section is sharing the load. Units: pascals (N/m²).

Strain (ε) is the fractional change in length: ε = ΔL/L. It measures how much the material has deformed, normalised by its original size. Dimensionless.

Young’s modulus (E) is the ratio: E = σ/ε. It’s the fundamental stiffness constant of a material — how much stress is needed to produce a given strain. A high Young’s modulus means the material is stiff. A low one means it’s flexible.

Some numbers to build intuition. Diamond has a Young’s modulus of about 1,200 GPa — the stiffest natural material. Steel is about 200 GPa. Aluminium about 70 GPa. Glass about 70 GPa (stiff, but brittle). Wood about 10 GPa along the grain. Rubber about 0.01–0.1 GPa. Human tendon about 1.5 GPa. Human bone about 17 GPa.

These numbers tell you immediately useful things. Steel is about three times stiffer than aluminium, so a steel beam deflects a third as much as an aluminium beam of the same dimensions under the same load. This is why skyscrapers use steel, and why aeroplane designers who use aluminium must make structural members correspondingly thicker.

Young’s modulus was named after Thomas Young, who described it in 1807, though the concept had been explored earlier by Euler and others. It remains one of the most important material properties in engineering — arguably second only to density.

The Stress-Strain Curve: A Material’s Biography

If you take a sample of material — say, a steel wire — clamp one end, and pull the other end with gradually increasing force while measuring the elongation, you get a stress-strain curve. This curve is, in a sense, the material’s biography. It tells you everything about its mechanical behaviour.

For a typical metal like mild steel, the curve has several distinct regions:

Linear elastic region. At low stresses, stress and strain are proportional. Hooke’s law holds. The slope of this line is Young’s modulus. Remove the load, and the material returns to its original length. Nothing permanent has happened.

Yield point. At a critical stress — the yield stress — the material begins to deform permanently. For mild steel, this is about 250 MPa. Below this stress, the deformation is elastic. Above it, it’s plastic. The yield stress is the most important design parameter in structural engineering: you need to keep stresses below this value with a comfortable safety margin.

Plastic region. Beyond the yield point, the material deforms permanently. It gets longer but doesn’t spring back. In metals, this happens through the motion of dislocations — line defects in the crystal structure that allow planes of atoms to slip past each other. The material may work-harden (becoming stronger as dislocations tangle and block each other) or remain at roughly constant stress, depending on the metal.

Necking and fracture. Eventually, the deformation localises in one region — the material “necks” — and the cross-sectional area drops rapidly. The stress in the necked region skyrockets, and the material fractures. For mild steel, this happens at about 25–30% total elongation.

Different materials have dramatically different curves. Rubber has an enormous elastic region (hundreds of percent strain) but a very different shape — an S-curve rather than a straight line. Ceramics and glass have a straight line (elastic) followed by immediate fracture — no plastic region at all. Polymers can show complex behaviour with multiple transitions depending on temperature.

The area under the stress-strain curve up to fracture represents the total energy the material can absorb before breaking — its toughness. Steel has high toughness (stiff and deforms a lot before breaking). Glass has low toughness (stiff but breaks with almost no plastic deformation). Rubber has moderate toughness (flexible and stretches a lot, but the stresses are low).

Why Metals Bend and Ceramics Crack

The fundamental question of materials science is: why can you bend a copper pipe but not a ceramic mug?

The answer is dislocations — or rather, the presence or absence of them.

In a metal, atoms are arranged in a crystal lattice. Within this lattice, there are always defects — places where atoms are missing, extra, or misaligned. The most important defects for mechanical behaviour are dislocations: line defects where one plane of atoms ends abruptly within the crystal.

When you stress a metal, dislocations move. They glide along crystal planes, allowing rows of atoms to shift by one atomic spacing at a time. The net effect is macroscopic plastic deformation — the metal changes shape permanently without breaking. It’s like moving a heavy rug by sending a ripple through it rather than trying to slide the whole thing at once. The ripple (the dislocation) requires far less force to move than the entire rug (the whole crystal plane).

Ceramics and glasses don’t have this option. In ceramics, the bonds are ionic or covalent — directional and rigid, unlike the non-directional metallic bonds in metals. Dislocations exist, but they can’t move easily because the bonds resist the rearrangement. There’s no low-energy pathway for atoms to slip past each other. The material deforms elastically until the stress at some flaw (a crack, a pore, a grain boundary) exceeds the bond strength, and then it cracks.

This is why a ceramic coffee mug shatters when you drop it on a tile floor. The impact creates a stress that exceeds the fracture strength at the weakest point, a crack nucleates and propagates at the speed of sound (about 4,000 m/s in most ceramics), and the mug is destroyed. No warning, no bending, no denting. Just shattering.

Metals, by contrast, dent. Drop a steel pot on the same floor, and it acquires a dent — a permanent plastic deformation where dislocations moved. The pot absorbed the impact energy through plastic work rather than crack propagation. This is the fundamental reason metals are used for structures that must survive impacts: car bodies, helmets, ship hulls.

Rubber: The Weird One

Rubber doesn’t behave like metals or ceramics, and its elasticity has a completely different origin.

A rubber band is made of long polymer chains — thousands of carbon atoms linked in a chain, with cross-links between chains (introduced by vulcanisation, a process invented by Charles Goodyear in 1839). In their relaxed state, these chains are tangled and coiled randomly, like a bowl of spaghetti.

When you stretch a rubber band, you’re not stretching the chemical bonds within the chains (that would require far more force). You’re straightening out the coiled chains — pulling the spaghetti into parallel lines. The chains resist this straightening not because of bond stretching but because of entropy.

Wait, entropy? In elasticity?

Yes. A tangled, coiled chain has far more possible configurations than a straightened one. Thermodynamics tells us that systems prefer high-entropy (many configurations) states. Stretching a polymer chain reduces the number of configurations — it reduces entropy — and the system resists. The restoring force in rubber is fundamentally entropic, not energetic.

This has a bizarre consequence: if you heat a stretched rubber band, it contracts. In a normal material, heating causes thermal expansion. In rubber, heating increases the thermal motion of the chains, which favours the high-entropy (coiled) state, pulling the material back to its relaxed length. You can demonstrate this yourself: hang a weight from a rubber band and heat it with a hair dryer. The weight rises.

Rubber can undergo enormous elastic strains — several hundred percent — because you’re uncoiling chains, not stretching bonds. But there’s a limit. Stretch too far and the chains reach their maximum extension. Stretch further and the covalent bonds begin to break. The rubber tears.

The coefficient of restitution — the fraction of kinetic energy returned after a bounce — depends on hysteresis: energy lost to internal friction during the deformation-recovery cycle. A superball (made from a highly elastic polybutadiene compound) has a coefficient around 0.9 — it returns 81% of the impact energy. A tennis ball returns about 55%. A ball of putty returns almost nothing.

Fatigue: Death by a Thousand Bends

Materials can fail not just from a single excessive load but from many small ones. Bend a paperclip back and forth twenty times and it snaps — even though each individual bend is well below the force needed to break it in one go.

This is fatigue failure, and it’s one of the most dangerous modes of structural failure because it occurs at stresses below the yield stress — stresses that would seem perfectly safe based on a static stress-strain test.

The mechanism is crack growth. Each loading cycle opens tiny cracks — often initiated at surface scratches, machining marks, or material defects — by a microscopic amount. Over thousands or millions of cycles, the crack grows across the cross-section. When the remaining intact material can no longer support the load, the final fracture is sudden and catastrophic.

Fatigue failure has caused some of the worst engineering disasters in history. The de Havilland Comet, the world’s first commercial jet airliner, suffered three catastrophic in-flight break-ups in 1953–1954. Investigation revealed that the square windows of the fuselage concentrated stress at their corners, and repeated pressurisation cycles (each flight) grew fatigue cracks until the fuselage burst. The Comet’s square windows were replaced with round ones in subsequent aircraft designs — round windows distribute stress more evenly, eliminating the stress concentration. Every commercial aircraft since has had round or rounded-rectangle windows.

Modern engineering accounts for fatigue through S-N curves (stress versus number of cycles to failure) and through inspection protocols that look for cracks before they reach critical size. The discipline of fracture mechanics — developed largely by George Irwin in the 1950s — provides the mathematical tools to predict how fast a crack will grow under cyclic loading, and therefore how often to inspect and when to retire a component.

What Elasticity Teaches Us

The physics of elasticity is, at its heart, the physics of atomic bonds. Hooke’s law is a macroscopic shadow of the harmonic potential well that every atom sits in. Young’s modulus is a bulk measurement of bond stiffness. Yield stress tells you when bonds start to rearrange. Fracture strength tells you when bonds break.

What I find compelling is how the same basic physics — atoms connected by bonds, resisting displacement — produces such radically different behaviour depending on the arrangement. Metallic bonds allow dislocation motion, giving metals their ductility. Ionic and covalent bonds resist dislocation motion, making ceramics brittle. Polymer chains introduce entropy-driven elasticity, giving rubber its stretch. The bonds themselves aren’t so different. It’s the architecture — the crystal structure, the chain topology, the defect population — that determines whether a material bounces, bends, or breaks.

Engineers spend their careers navigating this landscape. The right material for a bridge is not the right material for a trampoline, and neither is right for a phone screen. Every material choice is a trade-off between stiffness, strength, toughness, weight, cost, and a dozen other properties — all of which trace back to what happens when you push atoms out of their equilibrium positions.

Hooke figured out the proportionality in 1678. We’ve been working out the consequences ever since.