Adding time to the picture
Guide 1 argued that cooling RATE, not the equilibrium diagram, decides the real microstructure; guide 2 gave the machinery of why — a nucleation barrier that must be overcome, then growth, producing the S-shaped fraction-transformed curve of the Avrami equation. Now we put those ideas to work on the most important alloy on Earth: steel. At 727 °C the iron-carbon diagram says austenite (face-centred cubic iron holding dissolved carbon) should decompose by the eutectoid reaction into a layered mix of ferrite and cementite called pearlite. What the phase diagram will NOT tell you is how long that takes, or what you get if you refuse to wait.
The natural measure of speed is the transformation rate — roughly one divided by the time to reach the halfway point. Here is the twist that runs this whole guide: that rate does not simply climb as you cool further below 727 °C. It rises, peaks, then falls again. Just below 727 °C the undercooling is tiny, so the driving force for nucleation is weak and almost nothing happens. Far below, the driving force is huge but diffusion has gone sluggish — atoms can barely shuffle into place. Only in between do a strong driving force and still-lively diffusion overlap, and the transformation races. That single competition is why the diagrams we are about to read are shaped like the letter C.
Reading a TTT (isothermal) diagram
The isothermal transformation diagram, or TTT diagram (Time-Temperature-Transformation), is built by an experiment that matches its name. Take many small samples of the same steel, make each fully austenite, then plunge each into a bath held at a fixed temperature and hold it there. At intervals you check how much austenite has turned into pearlite. Each temperature gives one Avrami S-curve; mark the moment transformation starts (say 1 percent done) and finishes (99 percent), and plot those two times against holding temperature. Repeat over a range of temperatures and the start and finish points trace out two nested C-curves.
T(C)
727 |===================================== A1 line
| (above: stable austenite)
650 | \ \ COARSE pearlite (soft)
| \ start \ finish
550 | ) <-- NOSE, ~1 s FINE pearlite (harder)
| / /
450 | / / upper bainite
350 | | | lower bainite
| | |
220 |----+--------+------------------- Ms (martensite START)
|
RT | (100% martensite if the cooling path clears the nose)
+----+----+----+----+----+----+----> log time
0.1s 1s 10s 100s 1e3s 1e4s
Miss the nose -> martensite. Land right of finish -> pearlite/bainite.Now the C-shape has a meaning you can read. The fast-transforming waist is the nose — for a plain-carbon eutectoid steel it sits near 540 °C, where transformation can begin in about a single second. Above the nose the reaction is slow because the driving force is small; below it, slow because diffusion is frozen; at the nose both cooperate. And crucially, the temperature at which you cross the diagram decides not just how fast, but WHAT you get: the same austenite becomes different microstructures depending on where on the C you let it transform.
The products of austenite
High on the diagram, just below 727 °C, atoms have plenty of time to diffuse, so pearlite grows with thick, widely-spaced layers of ferrite and cementite — coarse pearlite, soft and easily machined. Drop the holding temperature toward the nose (say 540-600 °C) and the same two phases form, but with far thinner, more tightly packed lamellae — fine pearlite. Why is fine harder? Every ferrite/cementite interface is a wall that a gliding dislocation must fight through, and fine pearlite crams in many more walls per millimetre, so it resists deformation more. Same two phases, same overall carbon; only the SCALE of the microstructure changed, and hardness climbed with it.
Below the nose, roughly 250-540 °C, diffusion is too sluggish for tidy layers, and you get bainite instead: fine needles or plates of ferrite studded with tiny cementite particles, feathery near the top of the range and finer, harder, and tougher near the bottom. Bainite is still a diffusional product — carbon must move — but it forms on the slow, cold side of the nose, so its structure is far finer than pearlite. For many steels it is a sweet spot: nearly as hard as fine pearlite yet noticeably tougher.
Cool fast enough to slip PAST the nose entirely, and something completely different happens. Below the flat Ms line (around 220 °C for eutectoid steel) austenite converts to martensite with no diffusion at all — the iron lattice simply shears over, trapping the carbon in place. That trapped carbon wedges the would-be body-centred cubic ferrite into a strained body-centred TETRAGONAL cell, and it is that distortion which makes martensite ferociously hard (up to about 65 on the Rockwell C scale) but brittle. Notice martensite has no C-curve: it needs no time, only a low enough temperature. It is the star of guide 4; here just fix its place on the map as the horizontal line at the very bottom.
From TTT to CCT: the diagram you actually cool along
The TTT diagram hides an assumption we should own up to: it pretends you can teleport the steel instantly to a holding temperature and freeze the clock there. Nobody quenches like that. A real part cools CONTINUOUSLY — its temperature slides down along a curve while it transforms. To describe that honestly we need the CCT diagram (continuous-cooling transformation), measured by cooling samples at controlled steady rates rather than holding them. The C-curves are still there, but both shift a little down and to the right compared with the TTT lines, because time spent drifting through the upper temperatures uses up some of the transformation before the part ever reaches the nose.
Now overlay a family of cooling curves, from a lazy furnace crawl to a violent water quench, and the diagram becomes a decision chart. A slow curve wanders through the pearlite field and comes out soft. A moderate curve clips the lower nose and gives fine pearlite. A fast curve that stays entirely to the LEFT of the nose never touches pearlite or bainite at all and arrives at Ms as pure austenite, transforming wholesale to martensite the instant it crosses. The particular curve that just grazes the nose is the critical cooling rate — cool any faster and you are guaranteed a fully hardened, all-martensite part.
For plain-carbon eutectoid steel that critical rate is brutal — of order 100+ °C per second — because its nose sits so far to the left. That is why a thick bar of it hardens only in a thin skin: the surface cools fast enough for martensite, but the core lags and turns to pearlite. The property that captures this is hardenability — the DEPTH to which martensite forms, measured with the Jominy end-quench test (Jominy test). Adding alloying elements like chromium, molybdenum or nickel shoves the CCT nose to the right, lowering the critical rate so martensite forms even on a gentle oil quench and deep into the section.
Watch a classic trap here: hardness is NOT hardenability. Hardness is how much a given spot resists indentation; hardenability is how DEEP you can get the hard martensite phase before the core reverts to something soft. Two steels can reach the same surface hardness yet differ wildly in hardenability — and for a thick shaft it is hardenability, not surface hardness, that decides whether the middle is any good.
Putting it to work — and a look ahead
Every common steel heat treatment is really a choice of cooling path across this diagram. Full annealing means a slow furnace cool that stays high on the C-curve, giving coarse pearlite — soft, ductile, and easy to machine. Normalizing is a brisker air cool that clips the nose region for finer pearlite and a stronger, more uniform grain structure. And the workhorse of hard steel, quench-and-temper, is exactly the two-step move you met as an analogy in an earlier rung: shock it glass-hard, then relax it just enough to be tough.
- Austenitize: heat the steel above the A1 line (about 730-780 °C for a low-alloy carbon steel) and hold until it is fully austenite with the carbon dissolved.
- Quench fast enough to clear the nose: cool faster than the critical rate (water, oil, or air, depending on the steel's hardenability) so the whole section arrives below Ms as martensite.
- Accept that as-quenched martensite is too brittle to trust — hard, yes, but liable to crack under shock or even from its own quench stresses.
- Temper: reheat to a moderate temperature (roughly 150-650 °C) so a little diffusion precipitates fine cementite, trading away some hardness for a large gain in toughness — the higher you temper, the softer and tougher the result.
One last idea, because it rides on the very same logic and closes this rung. Steels harden by racing austenite past a nose; many non-ferrous alloys, like aluminium-copper, harden by a cousin trick called age hardening. First solution treat — heat until all the copper dissolves into a single phase — then quench to trap a supersaturated solid solution that is desperate to precipitate. Age it (hold warm, around 150-190 °C) and the excess copper nucleates a dense rain of tiny GP zones and precipitates that pin dislocations, driving precipitation strengthening. Same philosophy as quench-and-temper: quench to a non-equilibrium state on purpose, then let a controlled, incomplete transformation build exactly the obstacles you want. Guide 5 takes both tempering and ageing all the way.