土鍋の熱

What a donabe is actually doing

A donabe conducts heat very badly. That is not a defect to work around: it is the whole point of the pot, and it is also the reason it breaks. The figures below are from a prefectural ceramics laboratory that measured four pots taken off the market, which is a rarer thing to have than it sounds.

What was measured

Mie Prefecture’s ceramics laboratory measured four donabe bodies that are actually on sale — the white lithia body, the coarse chamotte body, the iron-bearing red body, and the cordierite body it describes as common among imported pots — to JIS test methods. Thermal conductivity came out at 0.7 to 0.9 W/m·K for all four, with no significant difference between them despite quite different porosities, which run from 16.7% to 32.4%. Specific heat capacity sat between 0.78 and 0.84 kJ/kg·K. The lithia body it examined in detail is 40% petalite with 60% clay and pottery stone, fired at 1200°C give or take 25.

What that buys you

That conductivity figure is the pot. Heat enters slowly and leaves slowly, so the contents change temperature gradually rather than tracking the flame, and they go on cooking after it is off. Nothing else about a donabe has to be invoked to get there. It is worth noticing what the laboratory does invoke, since it is explaining donabe behaviour in a paper about donabe: conduction, heat capacity, thermal expansion, porosity, elastic modulus, fracture toughness. Far infrared, which is how these pots are usually sold, is not among the properties it measures.

And what it costs

The same slowness is what breaks them. When one part of a pot is hot and another is not, the two want to be different sizes, and a material that equalises temperature slowly holds that disagreement for longer. The bodies resist it by barely expanding at all — the measured coefficients at 600°C run from 1.84 to 2.61 millionths per kelvin, which is very low, and the petalite is there for exactly that. What they cannot do is tolerate a crack once one starts: fracture toughness for all four came out at or below 1 MPa·m^0.5, which the paper calls quite low. A typical donabe wall is about 5 mm.

Why two donabe are not interchangeable

The finding a buyer can use is the one about how little the outside tells you. Combining the measurements into a thermal shock resistance figure, the four commercial bodies come out at 752, 478, 401 and 324 W/m — a spread of well over two to one between pots sold for the same job. The best of them on that measure was the plain white body, which was also the weakest in bending strength, so the ranking does not follow the property a buyer might guess from. The paper states outright that judging thermal shock resistance the traditional quick way, from thermal expansion and water absorption, can be inaccurate, and that resistance is not fixed by expansion and porosity alone but depends heavily on microstructure and composition. Its own conclusion is that an actual verdict needs the JIS S2400 water-quench test.

What follows from it

• ✓Slow is the product, not a limitation. 0.7 to 0.9 W/m·K is why the contents stop tracking the flame and go on cooking off the heat. A pot that did this faster would be a different tool.
• ✗Nothing sudden. Not an empty pot on a flame, not cold water into a hot one, not a hot pot onto a cold surface. All three are the same instruction: do not ask one part of the wall to be a different temperature from the part next to it.
• ✗“It is a donabe” is not a specification. Four bodies sold for the same job scored 752, 478, 401 and 324 on the same thermal shock measure. The ware name on the listing does not carry that number.
• ✓Treat the care instructions as the tolerance band. Seasoning and gentle heating are how you stay inside a limit whose actual figure is not printed anywhere you can read it.

What this does not tell you

None of that is on a listing, and this site is not going to pretend it can be inferred from one. What follows from it is narrower and more useful: the rules that come with a donabe are not folklore, and the fact that two pots look alike, or share a ware name, does not mean they tolerate the same treatment. Seasoning, gentle heating and keeping cold water away from a hot pot are how you stay inside the tolerance of a body whose actual figure you cannot see.

Where to look

No shop link on this page. The one thing the measurements establish about buying is that the figure which matters — how much thermal shock a particular body tolerates — is not something a listing states or a photograph shows, and the laboratory’s own answer is that finding it out takes a water-quench test. Recommending a pot on that basis would be inventing a judgement the source declines to make. The two guides below cover the pot once it is yours: the step before its first meal, and why an ordinary one does nothing on induction.