Published in Metallurgy & Materials Science Insights • Technical Review

Bainite formation below the martensite-start temperature

H. K. D. H. Bhadeshia

There is a nice paper recently published on this subject: Liang YL, Zhou MX, Gan XL, Tian JY, Chen ZY, Xu G. Journal of Iron and Steel Research International 33 (2026) 188. It is shown here that the published data on the carbon concentration of the retained austenite entirely support the mechanism of the bainite transformation, even when the bainite forms below the martensite-start temperature.

Audio podcastAudio podcast     Study guidestudy guide

A long-standing point of discussion in physical metallurgy is how phase transformation kinetics change when moving across key thermal thresholds. A classic example is exploring what happens when the bainite reaction is forced to take place below the traditional martensite-start temperature (MS).

Recent research confirms that even when bainite forms below the MS temperature, the thermodynamic rules governing the transformation remain remarkably resilient. Specifically, the terminal carbon concentration of the residual austenite cannot exceed the T0 phase boundary limit. At this boundary, the free energies of the ferrite and austenite phases of identical composition become equal, acting as a strict barrier to further displacive growth.

Thermodynamic consistency: It is evident that the formation of bainite below MS is not fundamentally different from a thermodynamic perspective. The incomplete reaction phenomenon still dictates the limits of phase transformation.

This thermodynamic consistency is clearly illustrated by experimental data gathered from a high-strength alloy steel (Fe–0.25C–1.68Si–1.82Mn–1.45Cr–0.27Ni–0.054V wt%). By using either direct isothermal holding or a distinct two-step heat treatment involving supercooling below the MS point, carbide-free bainite was successfully generated within a pre-existing martensitic matrix.

Thermodynamic diagram showing carbon concentration in mole fraction versus temperature in degrees Celsius, plotting the T0 and Ae3 prime boundaries relative to the Ms temperature threshold for a Fe-0.25C-1.68Si-1.82Mn-1.45Cr-0.27Ni-0.054V wt% steel.
Figure 1: Experimental data published by Liang et al. for a steel composition of Fe–0.25C–1.68Si–1.82Mn–1.45Cr–0.27Ni–0.054V wt% featuring an MS temperature of 365 °C. The data points track the ultimate terminal carbon enrichment of the parent austenite when the bainite reaction is allowed to progress to its maximum possible transformation fraction. Note that although six distinct data points are recorded on this system, the two measured parameters at 300 °C superimpose identically. The curves are calculated using MUCG46 The Ae3' curve represents paraequilibrium.

Microstructural consequences of low-temperature undercooling

While the overall thermodynamic framework remains unaltered, shifting the transformation below the MS threshold drastically scales the structural dimensions of the final microstructural features: