3.1 — (i) Experimental construction of a CCT curve
A CCT (continuous-cooling-transformation) curve is built from a batch of identical specimens, all austenitized together (same temperature above A3, same soak time), but — unlike a TTT curve — each specimen is then cooled CONTINUOUSLY rather than held isothermally:
Austenitize every specimen identically, then cool ONE specimen continuously, at a chosen constant rate, from the austenitizing temperature down through room temperature — different rates are obtained with different quench media (furnace, still air, oil, water) or, for a precisely controlled rate, a dilatometer/programmable furnace.
Continuously record a property that changes at transformation throughout that one cooling run — dilatometric length/volume change is standard, since austenite, ferrite, pearlite, bainite and martensite all have different specific volumes, so a plot of dilation vs. temperature (equivalently vs. time, since the rate is constant and known) shows a clear departure from the pure thermal-contraction baseline where transformation begins, and a return to a new baseline slope where it ends.
Confirm the transformation product for that cooling rate by metallography and/or hardness measurement on the same (or a companion) specimen once it has fully cooled — pearlite, bainite and martensite are readily distinguished this way.
Repeat steps 1–3 with fresh specimens at a SERIES of different constant cooling rates, spanning from very slow (furnace cool) to very fast (severe water quench).
Plot every individual cooling curve as its own line on temperature (vertical) vs. log(time) (horizontal) axes, and mark the transformation-start and transformation-finish point directly ON each cooling curve, from the dilatometry of step 2.
Join the start points across all the cooling-rate curves into one locus, and the finish points into a second locus — these two joined loci (together with $M_s$/$M_f$, read from the fastest-cooled curves, which cross straight down into the martensite range) are the CCT diagram.
Read the critical cooling rate directly off the family of curves: it is the steepest curve that just grazes the pearlite "nose" of the start locus — any faster curve produces 100% martensite with no prior diffusional transformation.
3.2 — (ii) Why the lower ("bainite") portion of a CCT curve is usually incomplete
Fig. 3.1 — the same steel's complete TTT "C" (grey dashed) compared with its CCT curves (blue), shifted to longer times/lower temperatures. A cooling path either crosses the pearlite nose (green, becomes part pearlite) or misses it (purple, falls straight through to $M_s$) — neither path lingers at bainite-range temperatures, so the CCT diagram's own bainite "nose" is left incomplete.
On a CCT diagram, unlike a TTT diagram, the specimen is continuously falling through temperature the whole time, so time and temperature are locked together along ONE sloped cooling path rather than independently chosen as in an isothermal hold. Any cooling path slow enough to intersect the pearlite "nose" region begins transforming to pearlite there and continues consuming austenite as it keeps cooling past that temperature. A path fast enough to bypass the pearlite nose (miss it to the left) proceeds, still cooling, essentially straight down to $M_s$ — but it does NOT dwell at the intermediate (bainite-range) temperatures long enough for bainite to nucleate and grow, because the available "clock time" at those temperatures along a fast-falling continuous path is far shorter than the isothermal bainite start-time at that same temperature. Bainite formation simply cannot keep pace with the falling temperature. The practical result, for plain-carbon and many low-alloy steels, is that there is effectively NO continuous-cooling path that both avoids the pearlite nose (survives past the upper "C") AND then also dwells long enough in the bainite range to actually form bainite before crossing $M_s$ — a given cooling curve either produces pearlite (if it crosses the nose) or skips straight through to martensite (if it misses the nose), bypassing bainite either way. The CCT diagram's bainite "nose" is therefore drawn incomplete, or omitted outright, because the family of continuous-cooling paths that would populate it is essentially unreachable — even though the SAME steel's TTT (isothermal) diagram shows a full, complete bainite "C," since an isothermal hold can sit at a bainite temperature indefinitely with no such time constraint.
