How does a blacksmith know the iron's temperature without a thermometer?
― Hot iron glows a colour fixed by its temperature
Pulling iron from the forge, a smith can tell "that's about 800°C" at a glance. It looks like intuition, but what they're really reading is a law of nature. The colour of light given off by something hot is fixed almost entirely by temperature, not by what it's made of. And before it starts to glow at all, iron has a second "thermometer" built in.
On TV, a swordsmith's workshop appears. In a dim room with the windows shut, iron glows red inside the forge.
The smith never looks at a thermometer. They stare at the colour of the iron, and the instant it looks "right," they plunge it into water.
Why can colour alone reveal the temperature? And why bother darkening the room?
The smith is reading two different kinds of "colour"
Above roughly 500°C, an object starts to glow on its own. As the temperature climbs, the colour shifts from dark red to orange, yellow, and then white. This order is nearly the same whether the object is iron or stone.
Around 200–300°C, iron still doesn't glow. Instead, an extremely thin layer of oxide forms on the surface, taking on yellow, brown, purple, or blue depending on its thickness. And that thickness is set by temperature.
In other words, iron carries two "scales" — one for high temperatures, one for low. Let's look at each in turn.
Why does temperature alone decide the colour of something hot?
Everything gives off invisible light (infrared) according to its temperature. Our bodies, the desk in front of us — everything is emitting it right now. As temperature rises, the light gets stronger, and shorter wavelengths start to mix in too.
Look at Figure 1. The horizontal axis is wavelength, the vertical axis is brightness. The 1100°C curve is tall, but its peak sits far to the right of the "visible light" band — deep in the infrared. Only a sliver at the foot of the curve reaches our eyes.
At around 500°C, only a tiny sliver of red light falls in the visible band. So in a bright room, you can't tell it's glowing at all. Swordsmiths are said to darken their workshop precisely so they can catch that faint red. In daylight, the same iron would look a full step cooler than it really is.
As the temperature climbs further, the light entering the band spreads toward orange and yellow — shorter wavelengths. By the time it passes 1300°C, the visible colours are said to blend together and the iron looks white.
How do you read the temperature of iron before it glows?
In "tempering" a blade, the smith must hold a low temperature of around 200–300°C precisely. At this temperature iron doesn't glow. So instead, the smith watches the colour of the polished iron's surface.
When iron is heated, an extremely thin film forms on its surface where the metal bonds with oxygen in the air. Light reflecting off the top and bottom of that film overlaps, and only certain colours reinforce each other strongly. It's the same mechanism that makes soap bubbles look rainbow-coloured. The higher the temperature, the thicker the film grows, so the colour progresses through pale yellow, brown, purple, and blue.
Figure 2 lines up both scales along a single temperature band. The left side is film colour, the right side is glow colour. Notice the band in between where neither scale is easy to read.
Iron in band ③ of Figure 2 looks ordinary grey. Yet it can still be around 400°C. The same goes for a frying pan right after cooking, or a metal skewer just pulled from a campfire. Among craftspeople, the rule of thumb is: don't touch set-down iron bare-handed, glowing or not.
A reddish star in the night sky has a low surface temperature; a bluish-white one is hot. Astronomers estimating a star's temperature are reading the very same "colour-of-light scale" as a blacksmith.
Summary
A blacksmith can read temperature from colour because the colour of light given off by something hot is fixed almost entirely by temperature, not material. Below the glowing range, the colour of a thin surface film serves as a substitute scale. Working in a dark room is a trick for catching that faint red light before it's missed.
A craftsman's "intuition" is really a scale nature carved into iron.
But some temperature ranges have no scale at all.
For more on heat arriving as light, see Why does only your face get hot by a campfire, while your back stays cold?; for light made by a different mechanism entirely, see Why do LEDs use so little electricity?. For how thin films produce colour, see also Why do soap bubbles look rainbow-coloured?
- Compare an electric stove or toaster's heating element in a bright room and a dark one. In the dark, you'll spot the low-temperature red sooner (view from a safe distance, don't touch).
- Look at the bottom of a stainless steel pot, or a gas stove's burner grate. Any rainbow or blue-tinted scorch marks are "film colour."
- On a clear night, look for a reddish star and a bluish-white one. The colour difference reflects a difference in surface temperature.
When using any heat source, always work with an adult, and confirm it has been switched off and has cooled before putting it away.
Want to go deeper? ― Terms, equations, and textbook linksWe flag which level each point belongs to, from middle-school science to university specialist courses
- MSCovered in middle-school science
- HSCovered in high-school "Physics"
- HS+High-school enrichment, or textbook sidebar content
- UnivNot covered in high school — university specialist courses (statistical mechanics, quantum theory, metallurgy)
- ResearchNot yet settled even at university — an open question researchers are still working on
MSTerminology: this phenomenon has a name
- Thermal radiation: the light (mainly infrared) that an object gives off as energy, according to its temperature.
- Blackbody radiation: the thermal radiation given off by an ideal object that absorbs all light. Its colour and intensity are fixed by temperature alone, and real objects behave close to this.
- Quenching and tempering: quenching is rapidly cooling hot steel to harden it. Tempering is reheating it afterward at a lower temperature to restore toughness.
- Draper point: the temperature at which an object starts to glow visibly dark red. In 1847, Draper reported this as roughly 525°C, regardless of material.
MSHSChecking with an equation: where does the peak of the curve sit?
The wavelength at the peak of a thermal-radiation curve can be found just by dividing a constant by the absolute temperature (Wien's displacement law). Here λ (lambda) is the peak wavelength in micrometres, and T is the absolute temperature in kelvin.
| Wien's constant | about 2898 (micrometre·kelvin) |
| Visible light range | about 0.38–0.78 micrometres |
| Difference between absolute and Celsius temperature | 273 |
| Symbol λ | peak wavelength (micrometres) = 2898 divided by T |
| Convert 800°C to absolute temperature | 800 + 273 = 1073 |
| Peak wavelength | 2898 ÷ 1073 ≒ 2.7 |
| How many times the wavelength of red light (about 0.7) | 2.7 ÷ 0.7 ≒ 3.9 |
The light most strongly emitted by iron at 800°C is infrared with a wavelength about 4 times longer than red light. What we see is only the foot of the curve.
| Absolute temperature at 1000°C | 1000 + 273 = 1273 |
| Absolute temperature at 500°C | 500 + 273 = 773 |
| Temperature ratio | 1273 ÷ 773 ≒ 1.65 |
| Ratio squared | 1.65 × 1.65 ≒ 2.72 |
| Ratio to the 4th power (ratio of total light emitted) | 2.72 × 2.72 ≒ 7.4 |
The total light emitted is proportional to the 4th power of absolute temperature (the Stefan–Boltzmann law). A temperature increase of less than double makes the thermal glow over 7 times brighter. Restricted to just visible light, the increase is even steeper.
HSHS+Why is film colour set by thickness?
HSLight reflecting off the front and back of a thin film travels slightly different path lengths. Colours where that difference is an exact whole number of wavelengths reinforce each other; colours where it's off by half a wavelength cancel out. As the film thickens, the reinforced colour shifts steadily along.
HS+The oxide film that produces steel's tempering colours is said to be only tens of nanometres thick. Because the film's growth depends on heating time as well as temperature, the same colour can sometimes mean either "brief and hot" or "long and cool" — the two can't always be told apart.
UnivWhy is colour "independent of material," fixed by temperature?
The distribution of thermal radiation is described by Planck's radiation formula. Atoms and electrons inside an object vibrate randomly according to temperature, and when that energy balances against the exchange of light quanta, the shape of the distribution is fixed by temperature alone. Real objects have an "emissivity" — how readily they radiate light — that's less than 1, so brightness does vary by material. Still, if emissivity in the visible range doesn't change much with wavelength, the perceived colour is fixed almost entirely by temperature. A polished metal surface has low emissivity, so it's said to look dimmer, and therefore cooler, than an oxidised black surface at the same temperature.
ResearchWhat's still not fully understood
- How accurate is a craftsman's eye? There are reports comparing skilled workers' colour-based temperature judgements with non-contact pyrometer readings, but how to account for individual differences and lighting conditions isn't settled.
- Predicting tempering colours. How the film grows depends on the steel's composition and surface finish, and models that accurately predict colour from temperature and time are still being refined.
- Changing emissivity. As hot metal oxidises during heating, its emissivity itself changes. Finding accurate non-contact ways to measure temperature is still an active research topic in steelmaking too.
In other words, everything in this article is "the best explanation we have so far." Treat the colour-to-temperature charts as rough guides only.
Links to textbooks (by level)
| Level | Subject/unit | Where in this article |
|---|---|---|
| MS | Science: properties of light, energy transformation | Heat leaving as light |
| HS | Physics: interference of light, absolute temperature | Film colour, conversion to kelvin |
| HS+ | Physics: thermal radiation (enrichment) | Wien's displacement law, the 4th-power law |
| Univ | Statistical mechanics, quantum theory, metallurgy | Planck's radiation formula, emissivity, heat treatment |
| Research | Non-contact thermometry, surface oxidation | Accuracy of craftsmen's judgement, film growth models |
| ― | Everyday connections | Metal can be hot even when not glowing |
- National Astronomical Observatory of Japan (ed.), Rika Nenpyo (Chronological Scientific Tables), Maruzen Publishing (Physics/Chemistry section, thermal radiation entry) — 国立天文台編『理科年表』丸善出版
- Draper, J. W. (1847). On the production of light by heat. Philosophical Magazine, 30, 345–360.
- ASM International, ASM Handbook, Volume 4: Heat Treating
- Halliday, D., Resnick, R., Walker, J. Fundamentals of Physics (chapters on thermal radiation and thin-film interference)
※This article is a general-audience science explainer. Figures given are approximate, meant to aid understanding of the underlying mechanism. The link between colour and temperature varies with steel type, heating time, and lighting. Follow product manuals and workplace safety rules when handling heated metal or electric heating equipment.