Martensite in Steel — Diffusionless Transformation, BCT Structure and Hardness
Martensite is the defining microstructural constituent of hardened steel — a metastable, supersaturated solid solution of carbon in body-centred tetragonal iron produced by a diffusionless shear transformation during rapid quenching from the austenite phase field. Understanding martensite formation kinetics, crystal structure, morphological variants, and hardness response is foundational to every area of steel heat treatment, welding metallurgy, and structural integrity assessment.
- Martensite forms by a diffusionless, displacive (shear) mechanism when austenite is quenched below the martensite start temperature (Ms), trapping carbon in interstitial sites and distorting the BCC lattice into a body-centred tetragonal (BCT) structure.
- Tetragonality ratio c/a increases linearly with carbon content: c/a ≈ 1 + 0.045 × wt%C, reaching approximately 1.036 at 0.8%C.
- Two distinct morphologies exist: lath martensite (<0.6%C) and plate (acicular) martensite (>1.0%C), with a mixed region between 0.6 and 1.0%C.
- Martensite hardness rises approximately linearly from ~30 HRC at 0.1%C to ~66 HRC at 0.6%C, then plateaus due to increasing retained austenite.
- The Koistinen-Marburger equation quantifies martensite fraction: fM = 1 − exp[−0.011(Ms − T)], essential for predicting retained austenite.
- As-quenched martensite is highly brittle; tempering between 150 and 700°C progressively recovers toughness by precipitating carbides and relieving lattice distortion.
Formation Mechanism and Thermodynamics
Martensite formation is a first-order, diffusionless phase transformation driven by the difference in Gibbs free energy between austenite (γ) and martensite (α′) at temperatures below Ms. Unlike the pearlite or bainite transformations, no long-range diffusion of carbon or substitutional atoms occurs — instead, the transformation proceeds by a coordinated, military-style shear of the iron lattice, preserving the carbon distribution of the parent austenite in a frozen, highly distorted state.
Thermodynamic Driving Force
The transformation is thermodynamically spontaneous only when the product phase has lower free energy than the parent. At the equilibrium T0 temperature, γ and α′ have equal free energy; below T0, the growing driving force eventually exceeds the strain energy and surface energy penalties associated with the shape change, triggering nucleation at Ms:
Driving force: ΔG_chem = G(α') − G(γ) < 0 (below T₀)
Net driving force: ΔG_net = ΔG_chem − ΔG_strain − ΔG_surface
Transformation initiates when: |ΔG_chem| ≥ ΔG_strain + ΔG_surfaceThe strain energy associated with the martensitic shape change is substantial (typically 40–120 MJ/m³), which is why Ms lies significantly below T0 — typically 150–250°C below T0 for plain carbon steels.
Shear Mechanism and Habit Planes
The Bain correspondence provides the simplest geometric description: a compression of the austenite FCC lattice along one <100> direction and equal expansion along the two perpendicular <100> directions converts FCC into BCT with minimum atomic displacement. In practice, the observed transformation involves an invariant plane strain — a combination of the Bain distortion with additional lattice-invariant shear (slip or twinning) that preserves an undistorted, unrotated habit plane at the austenite-martensite interface.
Bain correspondence: [001]_γ → [001]_α'
[110]_γ → [100]_α'
[1̄10]_γ → [010]_α'
Habit planes (approx.):
Low-C lath martensite: {111}_γ (Kurdjumov-Sachs related)
High-C plate martensite: {259}_γ or {3 10 15}_γOrientation Relationships
Martensite maintains a specific crystallographic relationship with the parent austenite from which it grew. Two relationships are experimentally observed:
Kurdjumov-Sachs (K-S):
{111}_γ ∥ {110}_α' and ⟨1̄10⟩_γ ∥ ⟨111⟩_α'
→ 24 possible orientation variants per PAG
Nishiyama-Wassermann (N-W):
{111}_γ ∥ {110}_α' and ⟨112⟩_γ ∥ ⟨110⟩_α'
→ 12 possible orientation variants per PAGThese relationships are important for predicting texture evolution during thermomechanical processing and for interpreting EBSD orientation maps of martensitic steels. The prior austenite grain boundaries define the packets and blocks of lath martensite that control effective grain size for toughness.
Crystal Structure — Body-Centred Tetragonal (BCT)
In BCC iron (α-Fe), carbon atoms occupy octahedral interstitial sites with a radius of 0.019 nm — significantly smaller than the carbon atom (0.077 nm), creating lattice distortion even at very low concentrations. In martensite, the supersaturated carbon orders preferentially into one set of octahedral sites along the [001] direction (the c-axis), breaking the cubic symmetry and producing a tetragonal distortion.
Tetragonality and Carbon Content
Lattice parameters of martensite (nm):
a = 0.2861 + 0.0013 × wt%C
c = 0.2861 + 0.0116 × wt%C
Tetragonality ratio:
c/a ≈ 1 + 0.045 × wt%C
Examples:
0.2%C: c/a ≈ 1.009 (barely tetragonal)
0.4%C: c/a ≈ 1.018
0.8%C: c/a ≈ 1.036
1.2%C: c/a ≈ 1.054The tetragonality is directly measurable by X-ray diffraction as a splitting of the (110) and (011) peaks. At carbon contents below approximately 0.2%C the splitting is too small to resolve clearly, and the structure appears nearly cubic. The progressive increase in c/a with carbon content is the primary crystallographic evidence that carbon atoms are ordered along the c-axis rather than randomly distributed.
Dislocation Density and Substructure
As-quenched lath martensite contains an extremely high dislocation density of approximately 1014–1015 m-2, generated by the plastic accommodation of the transformation strain. This is comparable to severely cold-worked steel. The dislocations are tangled within laths and at lath boundaries, contributing significantly to yield strength via Taylor hardening:
Taylor hardening: Δσ = M × α × G × b × √ρ
where:
M = Taylor factor (~3.06 for random texture)
α = dislocation interaction constant (~0.3)
G = shear modulus (~80 GPa for Fe)
b = Burgers vector (~0.248 nm for α-Fe)
ρ = dislocation density (m⁻²)High-carbon plate martensite instead accommodates transformation strain primarily by twinning, producing the characteristic midrib structure visible at high magnification. The twins are approximately 2–5 nm thick, internally coherent, and bounded by {112} planes. Twinned martensite is significantly more brittle than dislocated lath martensite at equivalent carbon contents.
Morphological Variants — Lath vs Plate Martensite
The morphology of martensite is primarily determined by carbon content, which controls the driving force, habit plane, and accommodation mechanism. The transition from lath to plate morphology occurs progressively between approximately 0.6 and 1.0%C, with mixed morphologies common in this range. This is described in detail in the companion article on lath vs plate martensite types.
| Feature | Lath Martensite | Plate (Acicular) Martensite |
|---|---|---|
| Carbon range | <0.6 wt%C | >1.0 wt%C |
| Morphology | Fine parallel laths (0.1–0.3 µm wide) | Lens-shaped plates, random orientation |
| Habit plane | {111}γ (approx.) | {259}γ or {3 10 15}γ |
| Substructure | Tangled dislocations (~1014–1015 m-2) | Internal twins on {112} planes |
| Block/packet hierarchy | Present (lath → block → packet → PAG) | Absent — plates impinge independently |
| Toughness (as-quenched) | Moderate (CVN 10–40 J) | Very low (CVN <5 J); brittle fracture |
| Temper response | Good — effective carbide precipitation | Poor — temper embrittlement risk at 200–300°C |
| Typical steels | Low-alloy structural (SAE 4140, 4340); tool steels below eutectoid | High-carbon/hypereutectoid (SAE 52100, D2 tool steel) |
Martensite Start Temperature (Ms) and Koistinen-Marburger Kinetics
The Ms temperature marks the onset of martensite formation on continuous cooling. It is an intrinsic material parameter determined primarily by steel composition and, to a lesser degree, by austenite grain size and applied stress. Accurate prediction of Ms is critical for designing quenching schedules, predicting retained austenite, specifying cryogenic treatment, and avoiding quench cracking. Refer to the Ms Temperature Calculator for composition-based estimates.
Andrews Equation for Ms Prediction
Andrews (1965) empirical equation:
Ms (°C) = 539 − 423C − 30.4Mn − 17.7Ni − 12.1Cr − 7.5Mo
(All alloying elements in wt%C)
Example — SAE 4140 (0.40C, 0.90Mn, 0.15Ni, 0.95Cr, 0.20Mo):
Ms = 539 − 423(0.40) − 30.4(0.90) − 17.7(0.15) − 12.1(0.95) − 7.5(0.20)
Ms = 539 − 169 − 27 − 2.7 − 11.5 − 1.5
Ms ≈ 327°CCarbon has by far the largest depressing effect on Ms — more than an order of magnitude greater than any substitutional element on a per-wt% basis. Steels with more than approximately 1.2%C have Mf temperatures well below 0°C, meaning room-temperature quenching cannot complete the martensite transformation.
Koistinen-Marburger Equation
The fraction of martensite formed between Ms and any temperature T below Ms follows an athermal, temperature-dependent relationship described by the Koistinen-Marburger (K-M) equation:
Koistinen-Marburger equation:
f_M = 1 − exp[−α(Ms − T)]
where:
f_M = volume fraction of martensite (0 to 1)
Ms = martensite start temperature (°C)
T = quench temperature (°C), T < Ms
α ≈ 0.011 °C⁻¹ (empirical constant, valid for most steels)
Retained austenite fraction (f_RA) at room temperature (T = 20°C):
f_RA = 1 − f_M = exp[−0.011(Ms − 20)]
Example — SAE 4140 (Ms ≈ 327°C):
f_RA = exp[−0.011 × (327 − 20)] = exp[−3.377] ≈ 3.4% (negligible)For high-carbon steels with Ms closer to room temperature, the K-M equation predicts significant retained austenite. For example, if Ms = 150°C, fRA at 20°C ≈ exp(−0.011 × 130) ≈ 24% — enough to substantially reduce hardness and cause dimensional instability in precision components such as bearings and gauges.
Hardness of Martensite — Carbon Content Dependence
The hardness of fully martensitic steel is primarily controlled by carbon content through solid-solution strengthening and lattice distortion. The hardness testing methods most commonly applied to martensite are Rockwell C (HRC) for bulk specimens and Vickers (HV) for microhardness mapping of heat-affected zones or thin sections.
Empirical Hardness-Carbon Relationship
Approximate empirical relationships (fully martensitic microstructure):
HRC ≈ 30 + 60 × wt%C (valid for 0.1–0.6%C)
HV ≈ 127 + 949C + 27Si + 11Mn + 8Ni + 16Cr + 21log(Vr)
(Yurioka formula, where Vr = cooling rate in °C/s)
HRC to HV conversion (approx):
HRC 30 ≈ HV 297
HRC 40 ≈ HV 392
HRC 50 ≈ HV 513
HRC 60 ≈ HV 746
HRC 66 ≈ HV 940| Carbon Content (wt%C) | Approx. Hardness (HRC) | Approx. Hardness (HV) | UTS Estimate (MPa) | Dominant Morphology |
|---|---|---|---|---|
| 0.10 | ~32 | ~302 | ~1000 | Lath |
| 0.20 | ~38 | ~370 | ~1220 | Lath |
| 0.30 | ~47 | ~461 | ~1520 | Lath |
| 0.40 | ~54 | ~570 | ~1880 | Lath |
| 0.60 | ~63 | ~746 | — | Mixed |
| 0.80 | ~65 | ~832 | — | Plate |
| 1.00 | ~66 | ~870 | — | Plate |
| 1.20 | ~65* | ~850* | — | Plate |
* Hardness decreases slightly above ~1.0%C due to increasing retained austenite fraction. UTS estimates from hardness conversion apply only at <0.6%C where ductile failure governs.
Strengthening Mechanisms in Martensite
The extraordinary strength of as-quenched martensite arises from the superposition of several independent hardening mechanisms. In medium-carbon structural steels, the contributions can be ranked approximately as follows:
| Mechanism | Approximate Contribution to σy (MPa) | Physical Origin |
|---|---|---|
| Interstitial solid solution (C) | 500–1500 | Carbon atoms in BCT octahedral sites obstruct dislocation glide; tetragonal distortion creates asymmetric (Snoek-type) stress field |
| Dislocation forest hardening | 200–600 | Extremely high dislocation density (~1014–1015 m-2) from transformation strain accommodation |
| Boundary hardening (Hall-Petch) | 200–500 | Lath, block, and packet boundaries impede dislocation slip; finer prior austenite grain = finer packet size |
| Substitutional solid solution | 50–200 | Mn, Cr, Mo, Ni in iron lattice; modest individual contribution but cumulative in alloy steels |
| Precipitation hardening (auto-tempering) | 50–150 | Fine ε-carbide precipitates during quench in lath martensite with Ms >200°C |
Auto-Tempering
In steels with Ms temperatures above approximately 300°C — such as low-carbon, low-alloy grades — the martensitic laths formed early in the quench are exposed to elevated temperatures while the quench continues, allowing partial carbon diffusion and ε-carbide (Fe2.4C) precipitation within the laths before reaching ambient temperature. This auto-tempering reduces as-quenched hardness but can substantially improve toughness, making these steels less susceptible to quench cracking and more suitable for direct use without a separate tempering step.
Tempering of Martensite — Recovery of Toughness
As-quenched martensite is seldom used in structural applications because its fracture toughness is inadequate. Tempering — controlled reheating to temperatures between 150 and 700°C — allows carbon to diffuse from its supersaturated interstitial positions, progressively restoring lattice symmetry and relieving residual stresses. The tempering process occurs in overlapping stages:
| Stage | Temperature Range | Microstructural Change | Property Effect |
|---|---|---|---|
| Stage 1 | 100–200°C | Precipitation of ε-carbide (Fe2.4C, hexagonal); carbon content of matrix drops to ~0.25%C | Slight hardness increase in some steels; significant toughness recovery |
| Stage 2 | 200–300°C | Decomposition of retained austenite to bainite-like ferrite + cementite | Hardness may rise slightly; retained austenite eliminated |
| Stage 3 | 250–400°C | Transition carbides dissolve; coarse cementite (Fe3C) precipitates; tetragonality eliminated | Significant hardness decrease; large toughness improvement |
| Stage 4 | 400–700°C | Cementite spheroidises; subgrain and lath boundary recovery; recrystallisation above 600°C | Strength decreases; ductility and toughness increase substantially |
Hardenability and Martensite Formation in Section
Hardenability is the ability of a steel to form martensite at a given depth from the surface during quenching — it is distinct from hardness (which depends on carbon content). Hardenability is quantified by the Jominy end-quench test (ASTM A255 / ISO 642) and characterised by the ideal critical diameter (DI). High-hardenability steels (e.g., 4340) form martensite to the centre of large sections; low-hardenability steels (e.g., 1040) form a martensitic case over a pearlitic/bainitic core. Full details are in the annealing and normalising and related heat treatment articles.
Boron and Hardenability
Boron additions of 0.001–0.003 wt% to low-carbon steel dramatically increase hardenability by segregating to austenite grain boundaries and suppressing ferrite nucleation, without reducing Ms. The boron effect is lost if nitrogen is not tied up by Ti or Al (forming TiN or AlN), which would otherwise combine with boron to form BN. The boron hardenability multiplying factor can be as large as 3–5× depending on base composition.
Industrial Significance and Applications
Martensite formation underpins virtually every high-strength steel application. The practical implications span steel selection, heat treatment engineering, welding procedure qualification, and failure analysis.
| Application | Steel Grade | Target Microstructure | Key Martensite Parameter |
|---|---|---|---|
| Structural pressure vessels | SA-508 Gr.3, SA-533 Gr.B | Tempered bainite/martensite | Ms >300°C; Charpy USE >100 J at −10°C |
| Bearing rings (SAE 52100) | 1.0%C, 1.5%Cr | Tempered martensite + <10% γR | Cryogenic treatment to reduce fRA; dimensional stability |
| High-strength bolt shafts | SAE 4140, 4340 | Tempered martensite (40–52 HRC) | Avoid hydrogen embrittlement; PWHT if electroplated |
| Weld HAZ in P91 pipe | 9Cr-1Mo-V (P91) | Fully martensitic HAZ | PWHT at 730–780°C mandatory; Ms ~400°C requires preheat 200°C min |
| Armour plate | MIL-DTL-12560 RHA | Tempered martensite (~500 HB) | High Ms design for lath morphology; ballistic toughness critical |
| Cold-work tool steel (D2) | 1.5%C, 12%Cr | Plate martensite + primary carbides | Cryogenic treat to −80°C to minimise fRA |
In HAZ microstructure control, accurate prediction of Ms from weld metal or HAZ composition is essential for selecting preheat and interpass temperatures. Steels with Ms below approximately 300°C require preheat and PWHT to avoid hydrogen-assisted cold cracking, as described in the hydrogen-induced cracking guide. The Charpy impact test remains the primary industrial method for verifying that a tempered martensitic microstructure meets toughness requirements.
Metallographic Identification of Martensite
Reliable identification of martensite in as-quenched or tempered steel requires systematic metallographic preparation. Errors in etching or interpretation are common because bainite and heavily autotempered martensite can appear superficially similar at low magnification. Cross-reference with bainite microstructure and the pearlite colony growth guides to resolve ambiguous microstructures.
Preparation and Etching Protocol
| Step | Procedure | Critical Notes |
|---|---|---|
| Grinding | 120 → 240 → 400 → 600 grit SiC, wet | Remove each scratch layer completely before advancing; do not overheat — martensite is sensitive to re-tempering from friction heat |
| Polishing | 3 µm diamond; 1 µm diamond; 0.05 µm OPS (colloidal silica) | OPS step essential for revealing fine lath boundaries; ≥3 min OPS on vibratory polisher gives best results |
| Etching (standard) | 2–4% nital (HNO3 in ethanol), 3–15 s at room temperature | Over-etching obscures lath structure; wipe immediately with cotton swab after etching period; avoid picral which may obscure martensite |
| Etching (M vs B) | Klemm I reagent (aqueous sodium thiosulfate + potassium metabisulfite) | Martensite appears white/light; bainite appears brown to grey; more diagnostic than nital alone |
| Magnification | 200× for structure type; 500–1000× for lath/twin detail | Prior austenite grain boundaries visible at 100× after etchants containing picric acid (Bechet-Beaujard or saturated picral) |
Frequently Asked Questions
What is martensite and how does it form in steel?
What is the crystal structure of martensite?
What is the Koistinen-Marburger equation?
What is the difference between lath martensite and plate martensite?
How does carbon content affect martensite hardness?
What is the Ms temperature and what factors affect it?
Why is as-quenched martensite brittle and what does tempering do?
How is martensite identified in optical metallography?
What is retained austenite and why does it matter?
Recommended Reference Books
Disclosure: MetallurgyZone participates in the Amazon Associates programme. If you purchase through these links, we may earn a small commission at no extra cost to you. This helps support free technical content on this site.
Further Reading & Related Topics
