Rice cooker cutoff

Why does a rice cooker with no timer or sensor switch itself off exactly when the rice is done?

A basic rice cooker has one moving part the user can see: a lever that clicks down when you start it and springs up when the rice is done. There is no timer, no probe in the rice, nothing weighing the pot. It does not know how much rice you put in, or how much water, or whether the kitchen is cold.

And yet it works, across a wide range of quantities, for decades, with no adjustment. Any device that reliably gets the right answer without measuring the thing you care about deserves suspicion — it must be measuring something else that happens to coincide. The question is what.

Start by asking what "done" means physically. Rice is cooked by absorbing water and being held hot enough for the starch to gelatinise. The cook's instruction is a ratio — so much rice, so much water — and the rice is done at about the moment the water is gone: absorbed into the grains, with the remainder driven off as steam.

So "done" and "out of free water" are very nearly the same event. That is the first half of the trick. The cooker does not need to detect doneness if it can detect dryness.

Now, how would you detect the disappearance of water without a moisture sensor? Here the physics does the work. While liquid water remains in contact with the pot at atmospheric pressure, the temperature of the pot cannot rise above the boiling point. Add heat and it does not raise the temperature; it converts water to steam. The temperature sits on a plateau at very close to 100 °C, and it sits there regardless of how vigorously you heat, how much water there was to begin with, or how long it has been going.

That plateau is the invariant the cooker exploits. And the moment the last free water is gone, the plateau ends — there is nothing left to absorb the incoming heat as latent heat, so the pot's temperature begins to climb, and climbs quickly, because a dry metal pot with a heating element under it has very little to slow it down.

So the signal the cooker needs is not "is the rice cooked" but "has the temperature just left the boiling plateau". That is a sharp, unambiguous, self-calibrating event. It happens at the same temperature every time, in every kitchen, for every quantity.

The last piece is the switch. The classic mechanism is a small permanent magnet held against a ferromagnetic plate in contact with the pot base, holding the lever down against a spring. Every ferromagnetic material loses its magnetism above a particular temperature — its Curie point — and the material here is chosen so that its Curie point sits a few degrees above the boiling point of water. While the pot is on the plateau at 100 °C, the magnet holds. When the water goes and the base runs up past that threshold, the plate abruptly stops being magnetic, the magnet lets go, the spring lifts the lever, and the element is switched off.

Look at what has been achieved. There is no measurement of rice, water, time or temperature in any readable sense. There is a physical constant — the Curie point of a chosen alloy — placed just above another physical constant — the boiling point of water — so that the transition between them coincides with the event the cook cares about.

The cutoff temperature is not 100 °C, and it matters that it is above it. If the trigger sat at the boiling point, the cooker would switch off as soon as the water boiled — at the beginning of cooking rather than the end. The design depends on the trigger being unreachable while water remains. The gap of a few degrees is the whole margin, and it is why these devices are not adjustable: moving the threshold either breaks the logic or does nothing.

It works by proxy, and the proxy can be wrong. The cooker detects the end of free water, not the state of the starch. Where those come apart, so does the result — a badly wrong water ratio produces confidently switched-off rice that is undercooked or scorched, and the machine has no way to know. Fancier cookers add temperature profiles, pressure, or fuzzy control precisely to break this dependence, and they achieve better results for awkward grains at the cost of the elegant simplicity described here.

Altitude shifts the plateau but not the outcome. At reduced pressure water boils below 100 °C, so the plateau sits lower and the gap to the Curie point widens. The cutoff still fires at the same point in the process, because the mechanism keys on the end of the plateau, not its height. What altitude changes is the cooking, since the rice is held at a lower temperature.

The keep-warm function is a separate system. On models that stay warm afterwards, a second, lower-temperature thermostat takes over once the main switch releases — ordinary closed-loop control around a setpoint, of the kind the cooking phase deliberately avoids.

The falsification test. If the mechanism keys on the end of the boiling plateau rather than on elapsed time or temperature as such, then the cutoff should occur later with more water and earlier with less, at constant rice — and adding water partway through should postpone it. If the lever popped after a fixed interval regardless of how much water was in the pot, the account would be wrong and there would have to be a timer somewhere.

• Why the Curie point of an alloy can be tuned by composition, and how that is used in other thermal switches.
• What fuzzy-logic cookers add, and which grains actually need it.
• Why the same plateau logic underlies double boilers and why a pan of water cannot scorch.