Skip to main content

Retained Austenite in Quenched Steel: Effects, Detection, Reduction

Retained austenite is the fraction of austenite that fails to transform to martensite during a quench and remains as an untransformed phase in the hardened microstructure — commonly 5–25% in high-carbon and high-alloy grades such as 52100 bearing steel, D2 tool steel, M-series high-speed steels, 440C stainless, and carburized case layers. UTEC Industrial provides in-house induction hardening, through-hardening, and quench-and-temper heat treating services for industrial components in the Pacific Northwest, with integrated CNC machining and reverse-engineering capability. Because retained austenite is a metastable phase that can transform to fresh martensite under service stress, thermal cycling, or cold storage, its presence matters for dimensional stability, hardness uniformity, and long-term performance. This article surveys the sources of retained austenite, the detection methods (X-ray diffraction per ASTM E975, magnetic saturation, metallographic examination), service effects, and the two principal reduction routes — cryogenic treatment and multi-stage tempering — with guidance on when each is appropriate.

What is retained austenite, and why does it form during a quench?​

Retained austenite is austenite that remains in the microstructure after a quench that was intended to transform all austenite to martensite. The martensitic transformation is temperature-dependent rather than time-dependent — it begins at the martensite-start temperature (Ms) and continues as the steel cools toward the martensite-finish temperature (Mf). Both Ms and Mf depend strongly on carbon content and alloy content: as carbon content rises, both temperatures drop, so that at a carbon content of about 0.6% Mf falls below room temperature, and at 0.8–1.0% C (52100, M2 HSS, D2) Mf may be −100 °F to −200 °F (−75 °C to −130 °C) below room temperature. When a high-carbon or high-alloy steel is quenched from the austenite field to room temperature, the transformation is interrupted at the point where the steel cools through Ms but not below Mf — the fraction of austenite that has not yet transformed at room temperature remains as retained austenite. For medium-carbon alloy steels (4140, 4340, 8620 core) the Mf is typically above 0 °F and retained austenite is minimal (usually less than 5%). For high-carbon bearing steels (52100 at 1.0% C, 1.5% Cr), retained austenite of 5–15% is typical after a conventional oil quench. For air-hardening tool steels with high alloy content (D2, A2), 10–20% retained austenite is common. For carburized cases (8620 with a 0.9% C case layer) and high-speed steels (M2, M4), 15–25% retained austenite is typical after the initial quench. Retained austenite also forms preferentially in regions where cooling rate was slower (part interiors of thick sections) or where local alloy/carbon chemistry depresses Mf below the bulk value (ASM Handbook, Vol. 4A, ASM International, 2013; Totten, Steel Heat Treatment Handbook, 2nd ed., CRC Press, 2006; Heat Treater's Guide: Irons and Steels, 2nd ed., ASM International, 1995).

How is retained austenite detected and quantified?​

X-ray diffraction (XRD) per ASTM E975 is the direct and quantitative reference method. The test exposes a prepared surface to monochromatic X-rays (typically Cr-Kα or Mo-Kα radiation), and the diffracted intensity is measured at the angles corresponding to the principal crystallographic planes of martensite (body-centered tetragonal, appearing as BCC-like peaks) and austenite (face-centered cubic, distinct peaks at different angles). The ratio of integrated peak intensities — typically using the 200 and 211 martensite peaks against the 200, 220, and 311 austenite peaks — combined with a correction for the known X-ray scattering factors of each phase, yields the volume fraction of retained austenite to an accuracy of roughly ±1–2% for fractions above 5% and slightly poorer precision below that. ASTM E975 specifies the peak selection, correction factors, background subtraction, and sample preparation (typical requirements: surface finish better than 600-grit, no cold work from grinding, light etch to remove any work-hardened surface layer). A second method — magnetic saturation — exploits the fact that austenite is paramagnetic while martensite is ferromagnetic: the saturation magnetization of the hardened part is compared against a standard of known 100% martensite in the same composition, and the fractional shortfall is attributed to retained austenite content. Magnetic saturation is faster and does not require X-ray equipment, but it is indirect and assumes no other non-magnetic phases are present (carbides are largely paramagnetic but can introduce small errors). A third method is metallographic point counting on a selectively etched section — retained austenite etches differently from martensite under a two-stage etch (e.g., 4% Nital followed by sodium metabisulfite), and a point-count or image-analysis estimate of the austenite fraction can be performed on a representative field. Metallographic estimation is the least precise of the three but is useful for spatial distribution information that XRD and magnetic methods cannot provide (ASTM E975; ASM Handbook, Vol. 4A, ASM International, 2013; Totten, Steel Heat Treatment Handbook, 2nd ed., CRC Press, 2006).

Why does retained austenite matter — what does it do to part performance?​

Retained austenite has three significant effects on hardened-steel performance, and a good specification anticipates all three. First, dimensional growth: retained austenite is metastable at room temperature, and it transforms slowly to fresh martensite over weeks to years of service, during thermal cycling, or during cold exposure. This transformation is accompanied by a volume increase of roughly 4% (austenite is denser than martensite), producing a part that grows in its critical dimensions over time — bearing inner rings that tighten onto shafts, precision gauges that drift out of tolerance, gear teeth that interfere with mating teeth. For bearing and instrument applications, the transformation-driven dimensional change is often the most important reason to reduce retained austenite to low values (typically less than 5%). Second, reduced as-quenched hardness: austenite is substantially softer than martensite, so a hardened part containing 15% retained austenite reads lower on Rockwell C than the same composition with 5% retained austenite — the hardness reading may miss specification by 2–3 HRC points purely from the retained-austenite contribution, independent of the underlying martensite hardness. Third, in small amounts (under 5–10%), retained austenite can improve fracture toughness and fatigue resistance because the austenite phase is more ductile and its stress-induced transformation to martensite absorbs energy at crack tips — this is the basis of TRIP (transformation-induced plasticity) steels and is the reason some bearing-steel specifications tolerate a small fraction of retained austenite deliberately. For carburized gears, retained austenite of 10–15% in the case layer is often acceptable and sometimes beneficial; for precision bearings, less than 3% is the typical target; for hot-working tools and cold-work dies, retained austenite is typically reduced as low as possible to stabilize hardness under service thermal cycling (ASM Handbook, Vol. 4A, ASM International, 2013; Heat Treater's Guide: Irons and Steels, 2nd ed., ASM International, 1995; ASTM E975).

What grades and applications are most affected by retained austenite?​

The grades where retained austenite is a consistent specification concern are those with carbon content above about 0.6% or with sufficient alloy content to depress Mf below room temperature. 52100 bearing steel (1.0% C, 1.5% Cr) is the prototypical example — in the standard hardening cycle (1,550 °F austenitize, oil quench) it produces 6–15% retained austenite, and bearing manufacturers routinely use sub-zero treatment and double or triple tempering to reach the less-than-3% target for precision bearings. D2 cold-work tool steel (1.5% C, 12% Cr, plus Mo and V) produces 10–20% retained austenite after the standard oil quench and requires multi-stage tempering or cryogenic treatment for dimensional stability in die applications. The M-series high-speed steels (M2, M4, M42) can produce 20–30% retained austenite after the primary quench — these grades are typically subjected to triple tempering (three 2-hour cycles at 1,050 °F) that transforms retained austenite during the inter-temper cool to fresh martensite, which is then tempered by the next cycle. 440C martensitic stainless steel (1.0% C, 17% Cr) produces similarly high retained austenite fractions. Carburized cases on 8620, 9310, and similar grades typically contain 15–25% retained austenite in the high-carbon surface layer and require post-carburize tempering or sub-zero treatment when specified for precision gear applications. The medium-carbon alloy steels used for general industrial service — 4140 and 4340 with quench-and-temper — produce very low retained austenite (typically less than 5%) and do not normally require specific retained-austenite management beyond the standard temper cycle; programmable multi-cycle tempering (such as double temper at 1,050 °F with full cool between cycles) is the appropriate route for higher-carbon or higher-alloy grades where retained austenite is a documented concern and where sub-zero processing is not being specified. Parts specified in 52100, D2, M-series HSS, or 440C with tight dimensional stability or hardness uniformity requirements should be discussed with the heat treater at the specification stage — these grades often require specialty processing including cryogenic treatment that is not universally offered (ASM Handbook, Vol. 4A, ASM International, 2013; ASM Handbook, Vol. 1, ASM International, 1990; Heat Treater's Guide: Irons and Steels, 2nd ed., ASM International, 1995).

How is retained austenite reduced by cryogenic or sub-zero treatment?​

Sub-zero treatment cools the hardened part below its martensite-finish temperature (Mf), continuing the interrupted transformation from the original quench and converting retained austenite to fresh martensite. Two tiers are distinguished in practice. Cold treatment (or "sub-zero treatment") uses temperatures in the −100 °F to −150 °F range (−75 °C to −100 °C), achievable with dry ice / acetone baths, mechanical refrigeration, or low-temperature freezers; this range is below Mf for most high-carbon and alloy steels with Mf in the −50 °F to −125 °F range, driving most but not all of the retained austenite to transform. Deep cryogenic treatment uses liquid nitrogen at −320 °F (−196 °C), held for extended times (typically 12–36 hours), which drives the retained austenite transformation to near completion (typical residual less than 1–2%) and in addition appears to refine the fine-carbide precipitation in tool steels, improving wear resistance beyond what pure austenite-to-martensite conversion would predict. Both cold treatment and deep cryogenic treatment are applied after the initial quench but typically before the final temper — the newly formed martensite from the sub-zero cycle is then tempered along with the original martensite. Sub-zero and cryogenic treatment are specialty services; they require either liquid-nitrogen handling or low-temperature mechanical refrigeration not typically present in a general-purpose heat-treating facility. UTEC does not offer cryogenic or sub-zero treatment as part of its heat-treating service — parts specified for cryogenic treatment should be routed to specialty heat treaters that operate cryogenic equipment (typically aerospace and tool-and-die specialty heat-treat shops). For applications where cryogenic processing is specified, it is common to subcontract the cryogenic step to a specialist and return the part to the primary heat treater for final tempering — this multi-vendor coordination adds lead time and handling cost and should be planned at the specification and quotation stage (ASM Handbook, Vol. 4A, ASM International, 2013; Totten, Steel Heat Treatment Handbook, 2nd ed., CRC Press, 2006).

How is retained austenite reduced by multi-cycle tempering without going sub-zero?​

Multi-cycle tempering — specifically double or triple tempering at temperatures above 900 °F for high-alloy grades and at 400–500 °F for lower-carbon carburized cases — is the most common retained-austenite reduction method in production heat treating because it requires no specialty equipment beyond the tempering furnace. The mechanism works in three stages per cycle. During the first temper soak, carbides precipitate from the martensite and, at higher temper temperatures (above 900 °F), retained austenite begins to destabilize by losing carbon to the precipitating carbides — lowering the effective alloy content of the retained austenite and raising its Ms. During the cool from the first temper, the destabilized retained austenite (with its newly raised Ms) transforms to fresh martensite — but this fresh martensite is now untempered and brittle. The second temper cycle then tempers this newly formed martensite along with any remaining small fraction of destabilized austenite. For high-speed tool steels, a third cycle is typically added because the triple-alloyed composition produces multiple rounds of austenite destabilization. Typical cycles: 52100 double temper at 300–350 °F for 2 hours each with air cool between cycles targets low retained austenite with high hardness preservation; D2 triple temper at 1,000–1,050 °F for 2 hours each reduces retained austenite to less than 2% while producing secondary hardening from the alloy carbide precipitation; M2 HSS triple temper at 1,050 °F for 2 hours each reduces retained austenite to less than 3% and produces the secondary hardness peak characteristic of HSS. The multi-cycle approach is well within the capability of a programmable tempering furnace — each cycle is a ramp-soak-cool sequence with full part cool-down between cycles, typically adding 6–12 hours of total furnace time over a single-temper schedule. Where the specification calls for both sub-zero and multi-temper, the sequence is typically quench → sub-zero → temper #1 → temper #2 (→ temper #3 if specified), with the sub-zero cycle placed before any tempering to maximize the retained-austenite conversion before carbide precipitation begins (ASM Handbook, Vol. 4A, ASM International, 2013; Heat Treater's Guide: Irons and Steels, 2nd ed., ASM International, 1995).

What should a specification say about retained austenite, and what verification should it require?​

A thorough specification for a retained-austenite-sensitive part includes four elements. First, the maximum acceptable retained-austenite fraction — typically 3% for precision bearings, 5% for dimensional-stable gauges and instruments, 10–15% for general tool steel applications, or higher as application-specific tolerance permits. Second, the measurement method — almost always X-ray diffraction per ASTM E975 for precision work, with metallographic estimation acceptable for less critical applications. Third, the process steps required to meet the retained-austenite limit — for example, "austenitize at 1,550 °F, oil quench, sub-zero treat at −120 °F minimum for 1 hour, double temper at 325 °F for 2 hours each" — so that the heat treater can quote and schedule the full process, and so that the documentation package can record compliance step by step. Fourth, the location and frequency of retained-austenite measurement — a single test coupon accompanying the production batch, three locations on a sacrificial part, or 100% verification by non-destructive magnetic saturation are all possible and have very different cost and handling implications. For parts where the retained-austenite specification is not tight (many general industrial machined parts in 4140 or 4340), the specification may be silent on retained austenite and rely on the standard quench-and-temper cycle to produce an acceptable low level by default — this is the appropriate choice for medium-carbon alloy steel where retained austenite is inherently low. For high-carbon, high-alloy, or carburized-case parts where retained austenite is a documented concern, the specification must be explicit — omitting the retained-austenite requirement from a 52100 bearing drawing or a D2 die drawing is a common source of downstream dimensional-stability failures that manifest in service months after the part is delivered. Every retained-austenite specification should be paired with a documentation requirement: the XRD test result, the test location, the method certification, and the sample surface preparation should be included in the heat treatment record on file with the part (ASM Handbook, Vol. 4A, ASM International, 2013; ASTM E975; Heat Treater's Guide: Irons and Steels, 2nd ed., ASM International, 1995).

Related Articles

References​

  • ASM International. (2013). ASM Handbook, Volume 4A: Steel Heat Treating Fundamentals and Processes. ASM International.
  • ASM International. (1990). ASM Handbook, Volume 1: Properties and Selection — Irons, Steels, and High-Performance Alloys. ASM International.
  • ASM International. (1995). Heat Treater's Guide: Practices and Procedures for Irons and Steels (2nd ed.). ASM International.
  • Totten, G.E. (ed.). (2006). Steel Heat Treatment Handbook (2nd ed.). CRC Press / Taylor & Francis.
  • ASTM E975: Standard Test Method for X-Ray Determination of Retained Austenite in Steel with Near Random Crystallographic Orientation. ASTM International.

Need In-House Heat Treating for Heavy Industrial Parts?​

UTEC Industrial operates a 6' × 10' × 17' car-bottom furnace (1,800 °F, 50-ton capacity), in-house induction hardening with per-part hardness verification, and automated vibratory stress relief at our Spokane, WA facility. Weldment stress relief, annealing, quench and temper, and induction hardening — all under one roof, with full documentation on every job.

Request a Quote →

Questions? Call (509) 922-1832 or email sales@utec.co