"Bicycle Mark" Perkins wrote:
> It has been my understanding (and I have studied metalurgy a
> little while at the University), that it's not so much the heating
> (unless one gets the metal really red-orange in color, and for a long
> period of time), but how quickly the metal is allowed to cool.
You're not far off, but it's more complicated than that. The peak temperature it got to is important. Above a certain temp, brass inclusions in the steel and intermetallic compounds can cause extreme brittleness. Time-at-temperature is also very important, because grain growth (which is, generally speaking, a bad thing) proceeds at a certain rate, so if it proceeds longer it has a larger effect. Cooling rate is, as you say, also important, but it isn't a simple case of "slower is better". Steel heated above its transformation temperature, then very slowly cooled, is *annealed* -- that is, as weak as it can be, though with high ductility. At the other extreme, steel (of a certain carbon content etc) quickly quenched from above its transformation temp, makes a hard but brittle steel, not suitable for bike frames but useful for other applications. Steel cooled from above it's transformation temperature at a moderate rate, around that rate which thin wall tubing will naturally cool at in still air, is called "normalized", and has a good balance between strength and ductility. Not the best, but that's how regular cheap bike steel and "aircraft" 4130 Cr-Mo is made.
Quenched-and-tempered ("heat-treated") steels like 753 can be a lot stronger
than normalized steel of the same alloy, but the trick is to not lose that
extra strength at the joint. It will lose a fair amount of it, that much is
inevitable, but your best hope is to keep the temperature below the
transformation temp, as well as keeping the time-at-temperature short as can
be. Cooling rate doesn't much matter in this case. Silver brazing usually
takes place completely below the transformation temp, in the tempering
range. A little tempering is OK, but it is weakening the steel, so one
should keep the time in the tempering range as short as possible, and the
peak temp as low as possible. Brass brazing takes place above the
transformation temp, at which point the heat-treatment "slate" is
essentially wiped clean, and you don't have heat-treated steel anymore, at
least in the heat-affected zone (HAZ).
> If [snip] the heat
> affected area cools too quickly, the crystal structure of the metal is
> locked (so to speak) in a heat altered state.
Yes, this is how heat-treating works. Quenching locks the steel in a structure that would not normally be stable at room temperature, a very strong but brittle state. Controlled tempering by the tube manufacturer is required to get some ductility back without losing all that extra strength. So you see it isn't simple.
I'd like to re-iterate: tempering means intentionally *weakening* steel that would otherwise be too brittle. (Well, ideally it's intentional!) Many people think tempered means strengthened, when it is just the opposite.
As I'm not a metallurgist, I'm sure I have made some errors and oversimplifications here, and welcome corrections.
Mark Bulgier
Seattle, Wa
USA