The Chladni laboratoryThe whole bench at once: every frequency, every edge, two drivers, and sand in the air

You know what a mode is, and you know why it forms. This page hands you the bench. John Tyndall's 1869 lectures printed page after page of Chladni figures — crosses, diamonds, nested rings, woven grids — and every one of them is hiding in a plate like this one, waiting at some frequency, some edge condition, some placement of the bow and the fingers. Hunt them.

1. The frequency explorer

One driver, one dial, the whole spectrum. The strip under the plate is the plate's response curve — how loudly it answers at each drive frequency. The peaks are the resonances; between them the plate barely replies, and the sand just smears. Two things to try beyond sightseeing. Free the edges: clamped edges force silence at the rim (sand piles there); free edges let the rim swing, and the figures become the curvier, X-shaped, diamond families Tyndall actually engraved. Use your finger: Chladni selected patterns by touching the plate while bowing — a touch forces a quiet point, and the plate must pick a figure that is silent there. Click the plate to place up to four fingers.

And keep turning the dial when the tone gets uncomfortable, and then when it disappears. The shaded end of the dial is beyond human hearing — the plate does not know that. The resonances keep coming, the figures keep getting finer, the physics does not blink. Only your ear gives out.

The full bench. The response strip is computed from the plate's actual mode ledger (peaks = resonances; the dot is you); snap to peak hops the dial along it. Click the plate to press a finger (up to four) — the figure must go quiet there, so the plate re-chooses its pattern. The dial's shaded end is past human hearing: sound on, the tone is the true pitch, and it fades out near 16 kHz while the plate plays on. For a small brass plate this dial spans roughly 1–24 kHz; ears stop near 16–20; the sand never does. —

2. Two drivers: the plate as a relationship-meter

Every figure so far came from one frequency. So here is a sharper question — the one this workshop exists to ask: what does the plate show when it is driven by two frequencies at once? Answer: it shows their relationship. Two drivers at exactly the same frequency make a fixed interference pattern set by their phase difference. Detune one slightly and that phase difference drifts — the pattern is never still; it breathes through its whole family at exactly the beat rate your ear knows from primer 00. And if you let the drivers couple — the same phase-pulling the harmonograph used — then past the locking threshold the drift stops and the pattern crystallizes.

Two drivers, left and right. With detune and no coupling, watch the interference pattern breathe — the sand can never quite commit, and the churn meter stays hot at the beat rate. Slide K past the threshold: the drive phases lock (beat rate 0.00), the pattern freezes mid-breath, and the sand finally gets a still figure to find. A Chladni plate can visualize coupled frequencies — not by drawing either tone, but by drawing whether they have come to an agreement. Sound on: both tones, real pitches — you hear the beats slow and stop as the pattern crystallizes. —

3. The plate in three dimensions

Finally, the honest picture of what the sand is actually doing — not walking, flying. A real grain on a real plate is thrown upward every time the surface accelerates beneath it, sails ballistically, lands somewhere slightly different, and is thrown again. Grains on loud regions are launched constantly and travel far; a grain that lands near a quiet line is barely disturbed. The figure assembles from thousands of tiny flights.

The same simulation, seen as a surface. The plate ripples (motion greatly exaggerated); the grains bounce ballistically — kicked by the surface where it moves, left in peace where it doesn’t — and hop their way to the nodal lines. Toggle the flashing off and the surface goes flat and dark while the grains keep hopping: the real lab, where the plate looks perfectly still and the leaping sand is the only witness that it isn’t. —
What's real here All three figures run the same verified wave engine as the previous page (96×96 finite differences; measured resonances match the operator's exact mode frequencies; sand rides a lock-in measurement of the steady response, which is how a laboratory would measure it too). Honest limits: this is a membrane, not a stiff plate — Tyndall's brass plates obey a stiffer equation whose frequencies climb faster and whose figures curve more, so hunt families and shapes here, not exact reproductions of his engravings. The kHz dial is a calibration choice (a coaster-sized plate; the simulation itself is scale-free), the response strip ignores your fingers (it is the bare plate's ledger), and at the dial's very top the grid resolves the finest figures with only a few cells per wave, so treat that end as a sketch. The 3D grains trade exact ballistics for a capped, stable cartoon of throw-fly-land — the mechanism, not the metallurgy.