Resonance: How a Tiny Push, Perfectly Timed, Can Shatter Glass and Topple Bridges

You already know the secret. You learned it on a playground.

To get a swing going high, you don't push hard — you push on time. A toddler's gentle shove, delivered at exactly the right moment in every cycle, beats a single mighty heave delivered at random. Each small push arrives just as the swing is moving away from you, so the energy adds instead of canceling. Twenty tiny pushes later, the swing is soaring.

That trick — small inputs, perfectly timed, accumulating into enormous motion — is called resonance. It is why opera singers can (sometimes) shatter wine glasses, why your radio plucks one station out of thousands of overlapping signals, why a microwave oven heats food, why MRI machines can see inside your body, and why engineers lose sleep over bridges and skyscrapers.

This fable gives you the resonance curve to play with directly. By the end, you'll know exactly where the danger zone is — and how engineers tame it.

Every object has a favorite frequency

Pluck a guitar string, flick a wine glass, bonk a tuning fork: each one rings at its own pitch, no matter how you strike it. That pitch is the object's natural frequency, written $\omega_0$ . It's set by the object's stiffness and mass — stiffer means faster, heavier means slower:

$\omega_0 = \sqrt{\frac{k}{m}}$

Left alone, a struck object oscillates at $\omega_0$ while friction slowly bleeds the energy away. That bleed rate is the damping, $\gamma$ . Drag the sliders and watch a plucked object ring down:

Try it — drag the values

Natural frequency ω₀3.0Damping γ0.40Driving frequency ω1.20

free ringing (plucked, then released)decay envelope

With damping γ = damping γ = 0.40, the vibration's envelope shrinks by a factor of $e$ every $2/\gamma$ time units. Crank γ all the way up and the ringing dies almost instantly — that's what shock absorbers in a car are for. Drop γ near zero and the object rings nearly forever, like a high-quality bell.

The equation behind the ringing◆◆◆◆◆go deeper +

The displacement $x(t)$ of a mass on a spring with friction obeys

$m\ddot{x} + b\dot{x} + kx = 0.$

Dividing by $m$ and defining $\gamma = b/m$ and $\omega_0^2 = k/m$ gives $\ddot{x} + \gamma\dot{x} + \omega_0^2 x = 0$ . Trying $x = e^{\lambda t}$ yields $\lambda = -\tfrac{\gamma}{2} \pm \sqrt{\tfrac{\gamma^2}{4} - \omega_0^2}$ . When damping is light ( $\gamma < 2\omega_0$ ), the square root is imaginary and the solution is a decaying cosine:

$x(t) = A\, e^{-\gamma t/2}\cos\!\left(\omega_d t + \phi\right), \qquad \omega_d = \sqrt{\omega_0^2 - \tfrac{\gamma^2}{4}}.$

Notice damping slightly lowers the ringing frequency — a heavily damped bell rings flat. At $\gamma = 2\omega_0$ ("critical damping") oscillation disappears entirely; the system just oozes back to rest. Car suspensions are tuned close to this point.

Now push it, over and over

Free ringing is the warm-up. The real story begins when you drive the object continuously — push the swing every cycle, sing at the glass, let the wind buffet the bridge. Add a periodic force $F\cos(\omega t)$ and ask: how big does the steady motion get, as a function of the driving frequency $\omega$ ?

Here is the answer — the most important picture in this fable. The horizontal axis is the frequency you choose; the curve is how violently the object responds:

response amplitude (your damping γ)same object, 4× more damping

Drag ω₀ and the peak slides with it. Drag γ down toward 0.05 and watch the peak erupt into a needle-thin tower. That tower is resonance: driving at the object's favorite frequency, where amplitude is roughly $1/(\gamma\omega_0)$ — small damping, monstrous response.

1
Far below ω₀: the object just follows
Push much slower than the natural frequency and the object placidly tracks your hand. Amplitude levels off at $1/\omega_0^2$ — it's basically a spring being slowly stretched. Boring, safe.
2
Near ω₀: every push lands on time
Now your pushes arrive in sync with the motion. Each cycle deposits energy on top of the last cycle's energy. Amplitude grows until friction burns energy as fast as you add it. With weak friction, that balance point is enormous.
3
Far above ω₀: pushes cancel themselves
Drive too fast and the object can't keep up — by the time it starts moving one way, you're already shoving the other way. Amplitude collapses like $1/\omega^2$ . A jackhammer's high-frequency buzz barely sways a building.

Deriving the resonance curve◆◆◆◆◆go deeper +

The driven equation is $\ddot{x} + \gamma\dot{x} + \omega_0^2 x = \tfrac{F}{m}\cos(\omega t)$ . After transients die out, the steady solution oscillates at the driving frequency: $x(t) = A\cos(\omega t - \delta)$ . Substituting and matching the sine and cosine parts gives

$A(\omega) = \frac{F/m}{\sqrt{(\omega_0^2 - \omega^2)^2 + (\gamma\omega)^2}}, \qquad \tan\delta = \frac{\gamma\omega}{\omega_0^2 - \omega^2}.$

The plot above is exactly $A(\omega)$ with $F/m = 1$ . Setting $dA/d\omega = 0$ shows the peak actually sits at $\omega_{\text{peak}} = \sqrt{\omega_0^2 - \gamma^2/2}$ , a hair below $\omega_0$ — another way damping flattens the music.

Q: one number that rates a resonator◆◆◆◆◆go deeper +

Engineers compress the whole curve into the quality factor $Q = \omega_0/\gamma$ . High Q means a tall, narrow peak: the object is picky about frequency and rings for a long time (roughly $Q/\pi$ cycles before the amplitude falls by $e$ ). A wine glass might have $Q \approx 1{,}000$ ; a quartz watch crystal $\approx 100{,}000$ ; the cesium transition in an atomic clock, upward of $10^{10}$ . Your car's suspension, deliberately, has $Q \approx 1$ . High Q is wonderful when you want selectivity (radios, clocks, lasers) and terrifying when you don't (bridges, aircraft wings).

The hidden half of the story: timing

Amplitude is only half the resonance picture. The other half is when the object moves relative to your push — the phase lag. Set the driving frequency ω with the slider and compare the push (gray-ish first series) to the response:

your push F·cos(ωt), scaledthe object's response

Try three experiments with the ω slider:

ω far below ω₀ (driving frequency ω = 1.20 vs natural frequency ω₀ = 3.0 right now): the response moves with the push, in lockstep.
ω exactly at ω₀: the response lags by a quarter cycle (90°). This is the magic alignment — the force is always pointing the same direction the object is moving, so it's always doing positive work. That's the playground-swing timing.
ω far above ω₀: the response is fully opposite the push (180°), which is why fast shaking accomplishes nothing.

Power delivered to the oscillator is force × velocity. At resonance, force and velocity are perfectly in phase, so every instant of every cycle pumps energy in. Off resonance, force and velocity spend part of each cycle opposed, and the driver actually sucks energy back out. Resonance is simply the frequency of one hundred percent efficient pushing.

A short history of unwanted resonance

1831
Broughton Suspension Bridge collapses, England
Soldiers marching in step set the bridge swaying until a bolt failed. The British Army issued the famous order to 'break step' when crossing bridges.
1850
Angers Bridge disaster, France
A battalion crossing in a storm; the combination of wind and marching destroyed the bridge, killing over 200 soldiers — the deadliest resonance-linked failure on record.
1940
Tacoma Narrows Bridge ('Galloping Gertie') tears itself apart
Filmed in spectacular detail; the footage became physics-class legend — though the true mechanism is subtler than textbook resonance (see caveat below).
1985
Mexico City earthquake
The quake's energy peaked near a 2-second period — matching the natural sway of buildings roughly 6–15 stories tall. Many mid-rise buildings collapsed while shorter and taller neighbors stood.
2000
London Millennium Bridge wobbles on opening day
Pedestrians unconsciously synchronized their steps with the bridge's lateral sway, feeding it — a feedback-flavored cousin of resonance. Closed for two years; fixed with 37 fluid dampers.

That Mexico City story deserves a picture. A building's sway period grows with height — roughly one second per ten stories — so a quake with a strong 2-second component selectively hunts buildings of a particular size:

natural sway period1985 quake's dominant period

Where the curves cross is where the damage concentrated. Modern skyscrapers in seismic zones carry tuned mass dampers — enormous pendulums (Taipei 101's weighs 660 tonnes) deliberately tuned to the building's $\omega_0$ , swinging out of phase to cancel sway. It is resonance, weaponized against itself.

Resonance you actually want

The same needle-thin peak that menaces bridges is a precision instrument everywhere else:

System	What resonates	Rough Q	Why the sharp peak helps
Playground swing	child + chains (pendulum)	~20	small pushes accumulate into big rides
Radio tuner	electrical LC circuit	~100	amplifies one station's frequency, ignores all others
Wine glass + singer	glass rim flexing	~1,000	a loud, sustained, exactly-pitched note can exceed the glass's strain limit
Quartz watch	vibrating quartz crystal	~100,000	rings so purely it loses under a second a day
MRI scanner	hydrogen nuclei in a magnetic field	very high	nuclei absorb radio waves only at their exact resonant frequency, mapping tissue
Atomic clock	cesium electron transition	~10¹⁰	the sharpest resonance we can build defines the second itself

Why selectivity follows from sharpness◆◆◆◆◆go deeper +

A resonator with quality factor $Q$ responds strongly only within a frequency band of width $\Delta\omega \approx \omega_0/Q$ around its peak (the "full width at half maximum" of the power curve). A radio circuit with $\omega_0 = 2\pi \times 1\,\text{MHz}$ and $Q = 100$ accepts a ~10 kHz slice — neatly one AM station — while signals a few slices away are suppressed by orders of magnitude. The same bandwidth relation, read in time instead of frequency, says a high-Q resonator takes about $Q$ cycles to build up or release its energy: sharpness in frequency is slowness in time. This tradeoff — a disguised uncertainty principle — governs everything from laser linewidths to why a piano's low strings take longer to speak.

Test the timing in your head

Check yourself

You're pushing a friend on a swing with a natural period of 4 seconds. Which strategy builds the biggest ride?

One open question to carry with you: the Millennium Bridge wobbled because people adjusted to the bridge, closing a feedback loop the simple driven-oscillator equation doesn't contain. Tacoma Narrows fell to a similar self-feeding twist. How many "resonance disasters," in markets, ecosystems, or social networks, are really feedback wearing resonance's clothes — systems that learned to push themselves at exactly the right moment?

Find a swing this week. You now know precisely why your timing works.