Martensite and austenite are two distinct structural forms that the same metal can take, and the differences between them drive much of what makes steel useful. Austenite is the high-temperature, softer, more ductile phase with a face-centered cubic crystal structure, while martensite is the hard, strong, and relatively brittle phase that forms when austenite is cooled rapidly. The transformation between them is one of the most important phenomena in metallurgy, underlying everything from the hardening of a kitchen knife to the superelastic behavior of medical stents.
Crystal Structure Sets the Stage
The fundamental difference between austenite and martensite is how their atoms are arranged. Austenite has a face-centered cubic structure, meaning atoms sit at the corners and the center of each face of a cube-shaped unit cell. This arrangement is relatively open and symmetrical, which allows atoms to slide past each other more easily. That is why austenite tends to be ductile and relatively soft.
When austenite transforms into martensite, the lattice distorts dramatically. In carbon steels, the resulting structure is body-centered tetragonal: essentially a stretched-out cube. Carbon atoms get trapped inside the lattice during the rapid transformation and force it out of its natural cubic shape, creating internal strain. The degree of stretching scales directly with carbon content. In iron-carbon alloys with less than about 0.6% carbon by weight, the ratio between the long and short sides of the tetragonal cell follows a predictable relationship.1Materials Science and Engineering: A. The effect of carbon content on the c/a ratio of as-quenched martensite in Fe-C alloys That trapped carbon is the source of martensite’s hardness: it pins the lattice and resists deformation.
In other alloy systems the symmetry differs. Shape-memory alloys like nickel-titanium (Nitinol) form martensite with a monoclinic structure, while some copper-based alloys produce orthorhombic martensite. A study of Heusler-type alloys confirmed that the derived martensitic structure from a face-centered cubic austenite parent had orthorhombic symmetry, with lattice dimensions that could index complex X-ray diffraction patterns.2Europe PMC. Derived crystal structure of martensitic materials by solid-solid phase transformation The point is that “martensite” is not one specific crystal structure; it is whatever lower-symmetry structure results from the rapid, shear-driven transformation of a higher-symmetry parent phase.
How One Becomes the Other
The transformation from austenite to martensite is fundamentally different from most phase changes you encounter in everyday life. When water freezes, molecules have time to rearrange by diffusing into their preferred positions. Martensitic transformations skip that step entirely. They are rapid, diffusionless transitions in which the parent crystal lattice reconfigures through coordinated shear and atomic shuffles into the new martensite lattice.3Nature Index. Martensitic Phase Transformations in Metallic Alloys Atoms do not wander through the lattice looking for new homes. Instead, whole blocks of atoms shift in concert, almost like a deck of cards being sheared sideways.
This is why cooling rate matters so much. If you cool austenite slowly, carbon atoms have time to diffuse out and form separate carbide particles, producing softer structures like pearlite or bainite. Quench it fast enough and the carbon gets locked in place, and martensite forms instead. In many alloys, this transformation is “athermal,” meaning it depends on temperature but not on time: once you hit a specific temperature threshold (called the martensite start temperature), the transformation proceeds in sudden bursts rather than gradually.4PubMed Central. What is the speed limit of martensitic transformations? You can hold the steel at a temperature between the start and finish points, and the transformation simply stalls until the temperature drops further.
Atomic-scale imaging has confirmed the detailed steps of this process. High-resolution electron microscopy has revealed that partial dislocations play a central role, creating intermediate hexagonal epsilon-martensite before the final body-centered-tetragonal alpha-prime martensite forms. These observations verified for the first time a model of the transformation mechanism that had been proposed half a century earlier.5PubMed Central. Dissecting the mechanism of martensitic transformation via atomic-scale observations The transformation proceeds through a specific cascade of shearing steps at specific crystallographic planes, and the transition zones between the old and new phases act as bridges that accommodate the strain.
Why Martensite Is Hard and Austenite Is Tough
The mechanical contrast between the two phases is stark. Austenite is relatively soft, formable, and ductile. You can bend it, stretch it, and draw it into wire without it cracking. Martensite, by contrast, is extremely hard and strong but also brittle in its as-quenched state. A piece of freshly quenched high-carbon martensitic steel can shatter like glass if you try to bend it.
This difference traces back to those crystal structures. The symmetrical face-centered cubic lattice of austenite has many slip systems, meaning there are numerous planes along which atoms can slide past each other under stress. More slip systems means more ways to deform plastically before breaking. The distorted tetragonal lattice of martensite, clogged with trapped carbon, has far fewer easy slip paths. Dislocations (the line defects that carry plastic deformation) have a much harder time moving through it.
Carbon content amplifies the effect. In iron-chromium stainless steels, the hardness of as-quenched martensite climbs as carbon increases, but only up to a point. At higher carbon levels, more austenite is retained alongside the martensite even after quenching, and the mix of phases can actually reduce overall hardness.6PubMed Central. Influence of Carbon on the Microstructure Evolution and Hardness of Fe-13Cr-xC (x = 0-0.7 wt.%) Stainless Steel Hardness in a real quenched steel depends not just on how tetragonal the martensite is, but on how much soft austenite survived the quench.
Mixing Both Phases on Purpose
Modern steels increasingly use austenite and martensite together rather than choosing one or the other. The idea is to combine martensite’s strength with austenite’s ability to absorb energy and stretch before breaking. A dual-phase steel containing both can achieve tensile strengths above 1.4 GPa while retaining around 30% ductility, a combination that neither phase manages alone.7International Journal of Plasticity. Mechanical properties and deformation mechanisms of a novel austenite-martensite dual phase steel In these materials, the hard martensite forms a load-bearing skeleton while the soft austenite fills the spaces in between, deforming and absorbing energy.
The ratio between the two phases matters. In layered composite steels with alternating martensite and austenite, researchers found that a martensite-to-austenite thickness ratio of about 2:1 gave the best balance, increasing uniform ductility by roughly 190% compared to purely martensitic steel with only a modest strength gain.8Materials Science and Engineering: A. Mechanism of increased strength and ductility of martensite-austenite composite steel with different thickness ratios Too much martensite and the material becomes brittle again. Too much austenite and you lose the strength advantage. The engineering challenge is hitting the sweet spot.
Nanotwinning offers another path. When austenite is processed to contain extremely fine internal twin boundaries, it can reach a yield strength of about 900 MPa while still stretching to around 21% uniform elongation. Martensite-austenite mixes of comparable strength managed only about 12% elongation.9Materials Science and Engineering: A. Comparison of strength–ductility combinations between nanotwinned austenite and martensite–austenite stainless steels This suggests that how you engineer the microstructure within a single phase can sometimes outperform blending two phases, though dual-phase approaches remain dominant in commercial production.
Transformation Under Stress
One of the more useful tricks in metallurgy is letting austenite transform into martensite during deformation rather than during cooling. When a steel part is loaded in service and a pocket of retained austenite feels enough stress, it can snap into martensite right at the spot where it is most needed. This is called transformation-induced plasticity, or the TRIP effect, and it improves strain-hardening behavior: the material gets stronger in exactly the zone where it is being stretched the hardest.10Advanced Engineering Materials. The Role of Transformation‐Induced Plasticity in the Development of Advanced High Strength Steels
The TRIP effect is central to several families of advanced high-strength steels used in automotive crash structures. During a collision, the steel absorbs more energy than either pure martensite or pure austenite could on its own, because the phase change itself consumes energy. The deformation-induced formation of both epsilon and alpha-prime martensite contributes to this transformation plasticity.11steel research international. Deformation Induced Martensite Formation and its Effect on Transformation Induced Plasticity (TRIP)
How quickly you deform the steel affects how much martensite forms. In metastable austenitic stainless steel, the volume fraction of martensite formed during tensile loading dropped from about 0.86 at a slow strain rate to only about 0.26 at a fast strain rate.12International Journal of Plasticity. Mechanical properties and deformation mechanisms of a novel austenite-martensite dual phase steel Faster deformation generates more heat, which raises the local temperature above the range where martensite is stable, suppressing the transformation. Engineers designing TRIP steels have to account for this: a component tested slowly in the lab may behave differently from one loaded rapidly in a crash.
Tempering and Retained Austenite
Freshly quenched martensite is generally too brittle for direct use. The standard fix is tempering: reheating the quenched steel to a moderate temperature and holding it there. During tempering, carbon gradually diffuses out of the martensite lattice to form tiny carbide particles, relieving internal strain. The result is a material that retains most of the strength of martensite but with significantly improved toughness.
The carbide sequence during tempering is surprisingly complex. At low temperatures (around 200°C), fine epsilon-carbides appear inside the martensite laths. Above 300°C, needle-shaped iron-rich carbides form along lath boundaries. As the temperature climbs further to around 500°C, these give way to different carbide types enriched in chromium and molybdenum. Above 600°C, yet another set of carbide transformations takes place.13PubMed Central. Sequential Carbide Precipitation During Tempering and Its Influence on Strength–Toughness Balance in 31CrMoNiNbV Secondary Hardening Martensitic Steel Each step in this cascade affects the final balance of strength and toughness, which is why tempering temperature is one of the most tightly controlled parameters in steel heat treatment.
Not all austenite transforms to martensite during quenching. Some fraction, called retained austenite, survives. In a martensitic stainless steel with about 13% chromium, the retained austenite fraction after conventional quenching was measured at roughly 9.5%. Cryogenic treatment (submerging the steel in liquid nitrogen for extended periods) pushed more of it to transform: two hours brought it down to about 8.4%, and twenty hours reduced it to about 6.5%.14PubMed Central. The Effect of Cryogenic Treatment and Tempering Duration on the Microstructure and Mechanical Properties of Martensitic Stainless Steel 13Cr-2Ni-2Mo Whether you want to eliminate retained austenite depends on the application. For a bearing or a gauge block where dimensional stability matters, retained austenite is a liability because it can transform unpredictably later. For a tool that needs some toughness, a small fraction of retained austenite acts as a shock absorber.
Interestingly, not all retained austenite is equally transformable. In nanostructured bainitic steels, thin films of austenite between bainite plates were found to resist martensitic transformation even when cooled to nearly absolute zero, down to -271°C.15PubMed Central. Cryogenic study of the magnetic and thermal stability of retained austenite in nanostructured bainite The mechanical constraint from the surrounding bainite, combined with chemical stabilization from carbon enrichment, can make thin austenite films extraordinarily stubborn. Blocky austenite pools in the same steel may transform readily, but films between the plates simply will not budge. This is a practical headache for anyone trying to fully eliminate retained austenite through cryogenic processing.
Shape Memory and Superelasticity
The martensite-austenite transformation is not limited to making things hard. In shape-memory alloys, it is the mechanism behind one of the more remarkable material behaviors in engineering. Nitinol (a roughly 50-50 nickel-titanium alloy) can be bent, twisted, or compressed into a new shape, then heated to recover its original form. This works because the deformation rearranges martensite variants without breaking atomic bonds, and heating transforms the martensite back to austenite, which remembers only one shape.16Materialia. Functional properties and shape memory effect of Nitinol manufactured via electron beam powder bed fusion
Superelasticity works on the same principle but at a constant temperature. If you deform Nitinol above its transformation temperature, stress forces the austenite to convert to martensite. The moment you release the stress, the martensite is thermodynamically unstable at that temperature and snaps back to austenite, recovering all the deformation. Orthodontic archwires exploit this: a Nitinol wire bent to fit crooked teeth exerts a nearly constant restoring force over a wide range of deflection, something no ordinary spring could do. The same principle enables self-expanding vascular stents that are compressed for insertion through a catheter and then spring open once deployed.
The crystallographic relationship between the austenite and martensite phases in these alloys is precise and reproducible, which is what makes the shape recovery so exact. The martensite that forms has multiple orientation variants, and mechanical loading selects among them, but all variants share the same geometric relationship with the parent austenite lattice. That shared relationship is what guarantees reversibility.
The Hydrogen Embrittlement Problem
Austenitic stainless steels are generally considered resistant to hydrogen embrittlement because the face-centered cubic lattice holds hydrogen differently than body-centered structures. But this assumption breaks down when the austenite is metastable. If deformation or welding induces some of the austenite to transform into martensite, the newly formed martensite becomes a trap for hydrogen and a site for embrittlement. In 316L stainless steel welds, strain-induced transformation of austenite to martensite was directly responsible for high hydrogen embrittlement susceptibility, and a greater amount of transformed martensite caused a greater loss in ductility in the presence of hydrogen.17International Journal of Plasticity. Mechanical properties and deformation mechanisms of a novel austenite-martensite dual phase steel
This is a real-world concern in petrochemical piping, hydrogen storage systems, and offshore structures where austenitic stainless steels are chosen precisely because they are supposed to resist hydrogen attack. Cold-worked zones near welds, bolt holes, or other stress concentrators can contain significant deformation-induced martensite, creating localized vulnerability exactly where you least want it. Selecting more stable austenitic grades (those with higher nickel content, for instance) reduces the risk because higher nickel suppresses the deformation-induced transformation. But any metastable austenitic steel in a hydrogen-rich environment deserves scrutiny.
Beyond Steel
The martensite-austenite framework extends well past iron-based alloys. Titanium alloys form martensitic phases during rapid cooling that can strengthen the material significantly, and the principles mirror those in steel: a shear-driven, diffusionless transformation that traps the lattice in a strained state.
Perhaps the most surprising application is in ceramics. Zirconia (zirconium dioxide) undergoes a martensitic-type transformation from a tetragonal to a monoclinic crystal structure, and engineers have harnessed this to toughen otherwise brittle ceramics. In partially stabilized zirconia, metastable tetragonal particles sit embedded in the ceramic matrix. When a crack begins to propagate and the stress field reaches these particles, they transform to the monoclinic phase. The transformation absorbs energy and the associated volume expansion squeezes the crack shut, dramatically increasing fracture toughness.18Journal of the American Ceramic Society. Transformation Toughening in Zirconia‐Containing Ceramics The mechanical energy absorbed near the crack tip through these stress-induced phase transformations is the key to the toughening effect.19Acta Metallurgica. Transformation-toughening in partially-stabilized zirconia (PSZ)
Zirconia dental crowns, hip-joint implants, and high-performance cutting tools all rely on this transformation toughening. The parallel with TRIP steels is striking: in both cases, a metastable phase transforms under stress, absorbing energy exactly where it is needed most. The atomic details differ completely, but the strategy is the same. It is a testament to how broadly useful the idea of a controlled martensitic transformation has become, far beyond the blacksmith’s quench tank where the phenomenon was first exploited centuries ago.
Crystallographic Orientation Relationships
When martensite forms inside an austenite grain, the two lattices are not randomly oriented with respect to each other. Specific planes and directions in the austenite align with specific planes and directions in the martensite, following what metallurgists call an orientation relationship. The most commonly cited ones in steel are the Kurdjumov-Sachs and Nishiyama-Wasserman relationships, though the actual orientation can vary considerably from one alloy system to another.20Materials Science and Technology. Austenite–martensite/bainite orientation relationship: Characterisation parameters and their application
These relationships dictate the shape of the martensite plates, the planes on which they form (their habit planes), and how they interact with neighboring grains. In martensitic steels, the well-known {225} habit plane has been studied for decades, and modeling work has shown that it arises naturally from the geometric constraints of the transformation distortion and the Kurdjumov-Sachs relationship.21Scientific Reports. {225}γ habit planes in martensitic steels: from the PTMC to a continuous model For most people working with steel, these crystallographic details live deep in the background. But for anyone designing heat treatments for precision components, or interpreting microstructures under the microscope, the orientation relationship is what tells you which variants of martensite formed and whether the transformation was well-behaved or chaotic.