h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

When you heat metal (perhaps specific to steel/ even a particular kind of steel) and it looks oily afterwards (shows a spectrum, sort of thing) , what's gone on? why does it do it? how does it work?

- I've done this to a knife, using it to weld plastic, and a needle, using a lighter to heat it.

I eagerly await your discussion!!

NH

It's a thin film interference effect, just as it is when oil produces its rainbow.
In the case of heated metals, the film is a thin layer of oxide. Its thickness and composition (both of which affect the colours you see) depends on the alloy composition, the temperature it's heated to and the atmosphere in which the heating takes place.
In some metals (notably carbon steels) these so-called 'temper colours' were once relied upon by heat-treatment experts. People learned to match the colours to the corresponding application properties after quenching.
Nowadays pyrometers are used to gauge heat-treatment temperatures. More accurate, less romantic.

Speaking as someone who has spent a considerable amount of time calibrating pyrometers, I can second that.

B

Shame about the obviously anticipated discussion.

This nail has been squarely hit on the head here I think

Pah - can't someone say something controversial or ludicrous so that I can string this out into a Post article..?

Oxide is transparent? ok.

Quartz is a crystal of silicon oxide and is very much transparent. High quality crystals of many solids can be very transparent.

Anyway, he didn't say it was transparent. What he said was that the thin layer of oxide is just the right thickness to create an interference pattern of light reflected from the surface.
Atoms and molecules and the bonds between them on a surface are a pretty good size to diffract light of visible wavelengths.

Any wave will be diffracted if it encounters an object of approximately the same size as its wavelength. Light diffracted from a regular array of such objects will start to interfere with other diffracted light beams. Also the differing wavelengths in white light will be diffracted to lesser and greater extents depending on their wavelenghts. This has the effect of 'spreading out' the light so that one sees a spectrum rather than just white light. This is the rainbow effect that you see. It's what you also see on the surface of a CD as you tilt it about and also what's happening with a thin film of oil on a water surface.

Diffraction and interference

http://electron9.phys.utk.edu/phys136d/modules/m9/diff.htm

I'm at uni studying physics, I have a reasonable idea of diffraction, but thanks! so the light gets through because it's simply so thin?

In a sense, probably the most similar to a diffraction ruler?

(something like this: http://www.mtholyoke.edu/courses/mpeterso/phys204/labs/Diffraction.htm )

Traveller in Time experimenting with light
"Or more like a grating on a mirror ?

Concidering the oxide will be one or a multiple layers thick and not a fractional number of molecules. The art of recognising the colour as relative to the temperature suggests it is not a spectrum what we see. Only some frequencies are produced. "

interesting selection these temper colours, aren't they

straw, brown, purple, dark blue

diffraction?

nothing to do with iron carbide being brown and magnetite being blue-black in thin section then

Traveller in Time on top
"You are suggesting the molecules of Fe3O and Fe3O4 have a resonant length of blue and brown light ? "

The use of the term diffraction was probably my fault as I saw the word interference and so shot diffraction straight from the hip.

According to this
http://dev.physicslab.org/Document.aspx?doctype=3&filename=PhysicalOptics_ThinFilmInterference.xml

Thin film intereference isa reflection property of the thin oxide film.

When you say something is 'brown' Seth, what else is a substance's colour other than how its chemical structure reflects absorbs and otherwise interferes with incident light?

not sure. Probably not.

What NH is doing is performing a crude version of blue annealing which in a particular temperature range coats the steel with a thin layer of magnetite

Typically this layer is 250 to 400 nm thick which admittedly does correspond to the wavelength of blue light, but the colour doesn't change if you view the surface from various angles (unlike oil film, CD etc). So my vote would be for an absorption effect rather than diffraction. Magnetite absorbs all wavelengths (even has a black streak) but maybe it absorbs blue just a little less than the rest

Given that NH was testing knives and needles, it's worth bearing in mind that these probably weren't carbon steel, and other metals (chrome, perhaps manganese, titanium and vanadium) could add to the colour effects.

Brown light?

Sorry Orcus, that reply was to (doing three jobs at once + nigerian internet connections = )

Actually, I've seen interference effects used to explain temper colours in more than one source, so not everything is clear cut.

The significance of a brown temper to me is that this has to be due to absorption. In my day, brown just wasn't part of the solar spectrum.

Traveller in Time mixing digital colours
"Also known as dark orange Brown "

It makes no difference whether a colouring effect is due to transmittance or absorbance (other than one is the inverse of the other).

Something is brown because it either aborbs or reflects light in such a way that the mixture of wavelenghts your eye detects is recognised by us as the colour brown.

Your eye is detecting the light rays that are *not* absorbed by a substance (and hence must have been refelected or otherwise emitted) so there must be a mixture of light that one could describe as brown. Just not a pure wavelength that is all - a shitty mixture if you like