An open-access study published on 29 October 2026 in Discover Metals maps, in unusual quantitative detail, how commercial rolled 42CrMo medium-carbon Cr-Mo alloy steel changes its microstructure when held isothermally anywhere between 350 and 725 degrees Celsius. Using a Gleeble-3800 thermomechanical simulator together with optical microscopy, scanning electron microscopy, transmission electron microscopy, electron back-scatter diffraction (EBSD) and X-ray diffraction, the team led by Yanning Liu captured the full sequence of phases that this widely used engineering steel can form under isothermal holding conditions.
The broad picture is a clean temperature ladder. At the warm end of the range, between 625 and 675 degrees Celsius, the steel forms a ferrite-pearlite microstructure: polygonal ferrite alongside pearlite colonies with lamellar internal contrast. Drop into the bainite regime between 350 and 475 degrees Celsius and the morphology flips in a characteristic way. Specimens held at 400 degrees Celsius developed fine acicular bainite with carbide-like features sitting inside the ferrite laths, the classic signature of lower bainite. Samples transformed at 450 degrees Celsius instead formed sheaf-type upper bainite, with carbide-like precipitates distributed along the lath boundaries rather than within them.
The reason for that flip lies in carbon diffusion kinetics, the authors argue. At 450 degrees Celsius carbon atoms migrate readily and can reach ferrite-lath boundaries, where the carbide-like features precipitate preferentially and produce the sheaf morphology. At 400 degrees Celsius diffusion is kinetically suppressed, carbon cannot fully reach the boundaries, and the carbide-like features instead form inside the laths. The researchers caution, however, that because TEM observation relies on representative local fields, the work confirms the location change between 400 and 450 degrees Celsius but cannot pinpoint a single exact critical temperature for the upper-to-lower bainite transition in this heat of steel.
The most striking quantitative result concerns how two different EBSD descriptors behave as temperature rises. The area-weighted equivalent circular diameter of BCC-indexed orientation domains, segmented at a primary 10-degree mis-orientation threshold, measured 4.43 micrometres at 350 degrees Celsius, 4.80 at 400, 7.67 at 425, 9.01 at 450, 7.65 at 475 and 6.40 micrometres at 500 degrees Celsius. Domain size therefore varies non-monotonically, peaking at 450 degrees Celsius. Meanwhile the fraction of 2-to-15-degree low-angle grain boundaries first rises and then falls with temperature, reaching a maximum of 58.4 percent at 425 degrees Celsius, where the high-angle boundary fraction correspondingly hits its minimum.
Those two maxima do not coincide, and the authors explain why. Orientation-domain size characterises the spatial scale of connected regions with similar crystal orientation, whereas boundary fraction counts pixel-level mis-orientation interfaces, so the two metrics need not reach their extremes at the same temperature. Physically, they attribute the pattern to the balance between bainite nucleation and growth under different degrees of under-cooling. At 350 to 400 degrees Celsius, large under-cooling drives a high nucleation rate, generating massive ferritic sub-structure while restricting atomic diffusion and interface migration, which keeps domains small and low-angle boundary fractions relatively low. Near 425 degrees Celsius, nucleation and growth reach a balance in which laths grow with preferred orientation relationships inside local austenite grains, generating abundant small-angle mis-orientations between adjacent laths. By 450 degrees Celsius reduced under-cooling lowers nucleation density while enhanced atomic migration promotes growth and merging of domains, pushing the domain size to its maximum; above that, at 475 to 500 degrees Celsius, the balance shifts again and domain size declines moderately.
The team is careful about what its EBSD metrics mean. The orientation domains are operational statistical objects obtained by segmentation, not equivalents of crystallographically reconstructed bainite packets, blocks or laths; precise packet-scale analysis would require parent-austenite reconstruction and orientation-relationship fitting. Even so, the trends proved robust. The non-monotonic domain-size pattern reproduced under 5-, 10- and 15-degree thresholds and remained stable when the analysis region, edge-domain treatment and indexing-mask settings were varied, with a full sensitivity analysis supplied in an appendix.
Transmission electron microscopy with energy-dispersive spectroscopy added a second layer of findings. Cubic TiN particles appeared in both the 400 and 450 degree specimens, and in the 400 degree lower-bainite sample spherical MnS inclusions were frequently found adjacent to TiN particles. Crucially, the authors conclude these are inherited inclusions formed during steel-making solidification, not new precipitates from the isothermal bainite transformation: the MnS-on-TiN proximity reflects heterogeneous nucleation of MnS on pre-existing TiN surfaces during solidification. They explicitly warn that the spatial proximity seen in local TEM fields cannot be interpreted as bainite nucleating on these inclusions. Carbide-like features, by contrast, did show temperature-controlled positioning, rod-shaped and intra-lath at 400 degrees Celsius but mostly inter-lath at 450 degrees Celsius, consistent with M3C cementite reported in the bainite literature, though selected-area electron diffraction phase identification was not performed for every feature.
X-ray diffraction with Co K-alpha radiation screened the phases in specimens treated at 400, 425 and 450 degrees Celsius. Diffraction peaks were dominated by the bcc/bct ferritic matrix, and no independent gamma-iron FCC peak could be distinguished within the experimental signal-to-noise ratio. EBSD phase statistics corroborated this, showing only sparse, isolated FCC-indexed pixels below 0.33 percent of indexed points at all test temperatures, mostly scattered single-pixel signals rather than continuous austenite regions. The paper therefore reports only a qualitative phase screening: if retained austenite exists in these specimens, it lies below the reliable quantitative detection limit of the current setup, and no volume fraction is claimed pending dedicated quantitative analysis with validated calibration standards.
The experimental design deserves note for what it does and does not support. Cylindrical specimens 6 millimetres in diameter and 80 millimetres long, machined parallel to the rolling direction from a single commercial heat of rolled bar, were heated at 10 degrees Celsius per second to 880 degrees Celsius, held five minutes, cooled at 30 degrees Celsius per second to one of thirteen isothermal temperatures, held for two to four hours depending on condition, and finally cooled to room temperature. One tested specimen represented each nominal temperature, and no independent repeat series was available, so each quantitative value rests on a single primary EBSD field per condition, albeit with extensive sensitivity checks on the analysis settings.
For a steel that ends up in automotive crankshafts, connecting rods and other heavy-load structural components, the practical payoff is a systematic baseline. Prior work on 42CrMo mostly addressed continuous-cooling transformation or tempered martensite; few studies had delivered multi-scale quantitative characterisation across the full bainite range, including orientation-domain statistics, grain-boundary mis-orientation distributions and TEM observation of inclusions and carbides under identical thermal cycles. By showing that bainite sub-structure evolution is genuinely multi-dimensional, with domain size and boundary fraction peaking at different temperatures, the study offers heat-treatment designers a richer set of descriptors than any single size parameter, and a quantitative reference for tuning austempering parameters in this workhorse alloy.
Subject of Research: Microstructural evolution of 42CrMo steel during isothermal phase transformation
Article Title: Microstructural evolution of 42CrMo steel during isothermal transformation
Article References: Liu, Y., Wang, C., Li, J., & Wu, Z. (2026). Microstructural evolution of 42CrMo steel during isothermal transformation. Discover Metals, 59(1), Article 9. https://doi.org/10.1007/s13404-026-00378-0
Image Credits: AI Generated
DOI: 10.1007/s13404-026-00378-0
Keywords: 42CrMo steel, bainite, isothermal transformation, EBSD, orientation domains, grain boundaries, TiN inclusions, MnS, carbides, retained austenite, XRD, heat treatment
News Source: Denise Maddox. (October 29, 2026). Bainite Shifts Shape as 42CrMo Steel Transforms Across a 375-Degree Window. Scienmag.
