COCONUTLAB bali

Interactive lab · filmed in Bali

The coconut that never escapes

A coconut circles behind a boulder in a Bali stream, while the chute beside it runs at ~2.5 m/s. Below is the same river from above, live. Push the river until it escapes.

slow <0.6 m/s0.6–1.8fast chutecoconutplan view, 12 × 5 m reach
filmed: ~2–3 m/s · escape ≳ 3.5 m/s
assumed, not weighed. 0.95 is probably too heavy: why
filmed: ~0.2–0.5 m/s

Break it on purpose

Four ways to free the coconut

Predict first, then press the button. Each one sets up the sim and scrolls you back to it.

01

Flood it

A surge more than doubles the jet speed for a few seconds. The eddy is drowned and the shear layer sweeps the coconut out.

02

Overtop the boulder

Push U past ~3.5 m/s. Water starts pouring over the boulder, its shelter collapses, and the eddy loses its grip.

03

Make it heavier

At ρ ≥ 0.98 the coconut barely floats. The water diving down at the shear line pulls it under, and the underflow carries it away.

04

Make it lighter

A fresh green coconut (ρ ≈ 0.8) rises faster than the water can push it down. It is even harder to free. Why? Buoyancy.

The real thing

What I actually filmed

A 13.8-second clip from a stream in Bali. The coconut never left the calm water behind the boulder.

Bali stream seen from above: a large boulder top left, white water rushing along the bottom, and a brown coconut floating in the calm dark water beside the boulder. STILL · t = 1.5 s
Frame at 1.5 s. The coconut sits in the slow lateral eddy, just inside the shear layer where the white-water jet races past. Labels are my reading of the footage; the eddy outline is approximate.
CLIP · 0–3 s, looped
Real footage, 3 s loop, no sound. Look for the small brown ball right of centre, drifting in the dark water.
About six seconds later, after the camera pans away and back, the same coconut is still floating beside the boulder. STILL · t ≈ 7.8 s
About 6 s later, after the camera pans upstream and back, it is still circling in the same eddy.

In 30 seconds

Why it can't get out

1 · A dead zone

The boulder blocks the near bank. Behind it the main flow separates, and the left-over slow water turns into a lateral eddy with closed streamlines. A floater just goes round and round.

The lateral eddy →

2 · A wall of shear

Leaving means crossing the shear layer, where speed jumps from ~0.3 to ~2.5 m/s. Exchange across it is slow: about 19 s for this eddy, longer than the clip.

Residence time →

3 · Float or dive

Water dives down along the shear line. A light coconut rises faster than it is pushed down and stays at the surface, trapped. A heavy one is dragged under and flushed away.

Buoyancy vs downwelling →
Read the full theoryDo it properly with free tools
Honest disclaimerThis is a simplified, educational model. The flow is prescribed, not solved: the coconut doesn't disturb the water, there is no free-surface or 3-D roller physics, and turbulence is a random walk. All parameters (D ≈ 17.5 cm, ρc/ρw ≈ 0.95, U ≈ 2.5 m/s, Uₑ ≈ 0.35 m/s, eddy 1.3 × 2.6 m, CD = 0.5) are rough estimates from video, and the assumed coconut mass (2.67 kg) looks too high. What the model gets wrong.