Austenite grain size determination is a core metallographic test used to evaluate the microstructural quality of alloy steels before and after heat treatment. Because prior-austenite grain size directly governs hardenability, strength, and toughness, measurement in accordance with ASTM E112, ISO 643, and GB/T 6394 is widely specified for material qualification, heat-treatment validation, and failure analysis. This article summarizes the sampling, grain-boundary revelation, measurement, and reporting requirements that accredited third-party laboratories apply when determining the austenite grain size of alloy steel.
Technical Overview
When alloy steel is austenitized, that is, held above its Ac3 transformation temperature, the microstructure converts to face-centered-cubic austenite. Upon quenching, martensite or bainite forms inside each original austenite grain, and the outline of that original grain persists as the prior-austenite grain boundary (PAGB). Determination of austenite grain size therefore consists of two linked tasks: first revealing these boundaries metallographically, and second quantifying the grain population according to a standardized procedure.
The prior-austenite grain size is one of the most influential microstructural parameters in quenched-and-tempered alloy steels. It controls several engineering-critical properties:
- Strength, through the Hall–Petch relationship: finer grains raise yield strength.
- Toughness: grain refinement lowers the ductile-to-brittle transition temperature and is essentially the only mechanism that improves strength and impact toughness simultaneously.
- Hardenability: coarser austenite grains shift the continuous-cooling transformation to longer times, increasing hardenability but degrading toughness.
- Fabrication risk: coarse grains increase susceptibility to quench cracking, intergranular fracture, and grain-boundary embrittlement.
Grain size is normally reported as the ASTM grain size number G, defined by the relationship N = 2^(G−1), where N is the number of grains per square inch observed at 100× magnification. Typical correspondences are G5 ≈ 65 µm mean diameter, G6 ≈ 46 µm, G7 ≈ 32 µm, and G8 ≈ 23 µm. Most quenched-and-tempered specifications require G5 or finer, while fine-grain plate standards frequently require G6 or finer.
In microalloyed and aluminum-killed alloy steels, fine pinning precipitates such as AlN, TiN, Nb(C,N), and VC restrain grain growth during austenitizing. When these particles dissolve or coarsen at elevated temperature, rapid or abnormal grain coarsening may occur, producing a mixed or duplex grain structure. Grain size testing after actual or simulated austenitizing is therefore a key tool for establishing the grain-coarsening behavior of a heat and for verifying fine-grain practice.
Testing Samples
Specimens are prepared from locations representative of the product and its processing history. Sectioning must be performed without introducing heat, burning, or heavy plastic deformation at the examination surface, since cold work and thermal artifacts can obscure prior-austenite boundaries. Specimens are typically mounted, with an examination face of roughly 15 to 25 mm providing a statistically adequate measurement area.
Bar, Rod, and Billet Samples
Transverse sections taken at locations specified by the product standard are used for routine release testing. Longitudinal sections are added when elongated grains from rolling or drawing must be characterized, using oriented intercept measurements.
Plate and Strip Samples
Samples are taken at quarter-thickness or as specified in the applicable plate specification. Through-thickness grain size surveys may be performed to detect surface decarburization-affected zones or centerline coarsening.
Forging Samples
Specimens represent the critical stressed locations of die forgings, shafts, crankshafts, and rings, including regions of minimum hot reduction where grain refinement may be incomplete or where critical-strain recrystallization has produced coarse grains.
Pipe and Tube Samples
Ring segments or longitudinal strips are sectioned from pipe and tube products, with attention to the inner and outer surfaces where abnormal grain growth can occur during through-heating.
Finished and Heat-Treated Components
Gears, shafts, spindles, bearings, fasteners, and other finished parts are sampled at functionally critical sections. For case-hardened parts, measurements target the effective hardened case after carburizing or carbonitriding plus quenching.
Weldments and Heat-Affected Zones
Cross sections through weld joints are examined to quantify grain coarsening in the coarse-grained heat-affected zone, which often governs the toughness and hydrogen-cracking susceptibility of welded alloy steel fabrications.
Failure Analysis Specimens
Sections adjacent to fracture surfaces are prepared to determine whether abnormally coarse prior-austenite grains, mixed grains, or grain-boundary phenomena contributed to the failure.
Testing Items
Typical test items requested from accredited third-party laboratories include:
- Average prior-austenite grain size, reported as the ASTM grain size number G, mean lineal intercept length, and/or grains per square millimeter.
- Grain size distribution and uniformity assessment across the section.
- ALA (as-largest-alloy? no—largest observed) grain size determination per ASTM E930, quantifying the single largest grain in the section.
- Duplex or mixed grain size characterization per ASTM E1181, including the area fraction and size of coarse and fine grain populations.
- McQuaid-Ehn effective grain size after simulated carburizing, for carburizing-grade gear steels.
- Verification of fine-grain practice and determination of grain-coarsening temperature through a series of isothermal holds at increasing temperatures.
- Comparative grain size before and after austenitizing, normalizing, or carburizing heat treatment.
Testing Methods
Specimen Preparation
Specimens are ground and polished to a scratch-free mirror finish, typically to 1 µm or finer diamond suspension. Because prior-austenite boundaries are subtle, careful polishing with minimal smearing is essential; over-etching with ordinary nital is generally ineffective on quenched martensitic structures.
Revelation of Prior-Austenite Grain Boundaries
Several standardized revelation techniques are used, selected according to the steel composition and heat-treatment condition:
- Etching method: for hardened martensitic or bainitic alloy steels, a saturated aqueous picric acid solution containing a wetting agent (for example, sodium tridecylbenzene sulfonate) selectively attacks prior-austenite boundaries; the etchant is often used warm with repeated etch-and-repolish cycles. Vilella's reagent and electrolytic etching are alternatives for specific grades.
- Ferrite network method: for slowly cooled or isothermally treated hypoeutectoid steels, proeutectoid ferrite precipitates along prior-austenite grain boundaries and outlines the grains directly; ordinary nital is sufficient.
- Oxidation method: a pre-polished specimen is heated briefly at the test temperature in air so that oxide preferentially forms at grain boundaries, then quenched and lightly repolished to reveal the oxide network, as described in ISO 643.
- Carburizing method (McQuaid-Ehn): specimens are carburized at approximately 925 °C for 8 hours and slowly cooled, precipitating a cementite network along the prior-austenite boundaries; this is the reference method for assessing grain growth of carburizing steels.
- Partial transformation methods: a short hold in the ferrite or bainite region decorates boundaries with transformation products before the final quench.
- Thermal etching: grooves formed at boundaries during vacuum heating outline the austenite grains at temperature.
Comparison Chart Method
The microstructure at 100× (or at an approved alternate magnification with appropriate correction) is matched against the standard comparison charts of ASTM E112, ISO 643, or GB/T 6394. The method is fast and suitable for routine acceptance of equiaxed grains, with an accuracy of roughly ±0.5 to 1 grain size number. It is not valid for elongated or duplex structures.
Planimetric Method
The Jeffries planimetric procedure counts all grains completely inside a test circle of known area plus one half of the grains intersected by the circle, giving grains per unit area. This directly yields the grain count metrics from which G is calculated, and it is well suited to coarse or non-uniform structures.
Intercept Method
The lineal intercept procedure, commonly implemented with the three-concentric-circle test grid of the Abrams procedure, counts intersections of test lines with grain boundaries. The mean lineal intercept is converted to the ASTM grain size number using the formulas in ASTM E112. The intercept method is the preferred technique because it is statistically robust, applicable to elongated grains, and readily yields a 95 percent confidence interval.
Automatic Image Analysis
Semiautomatic and fully automatic image analysis per ASTM E1382 digitizes the revealed boundary network, measures intercepts or grain areas across multiple calibrated fields, and computes the grain size distribution. Digital microscopy with stage-micrometer calibration is mandatory, and sufficient fields are measured so that the reported confidence interval meets the specification limits.
Data Processing and Reporting
The test report states the material and heat-treatment condition, sampling location and orientation, revelation technique and etchant, magnification, measurement method, number of fields and grains measured, the mean grain size number with its 95 percent confidence interval (or mean intercept length), any duplex or ALA findings, and a statement of conformance to the specified acceptance limit.
Applications
Austenite grain size determination supports a broad range of industrial and research activities:
- Heat-treatment validation: verifying that quench-and-temper, normalize, and austenitizing cycles produce the specified grain refinement without coarsening.
- Gear and bearing steel quality control: confirming fine grain after carburizing via the McQuaid-Ehn procedure and controlling grain growth during long carburizing cycles.
- Pressure vessel and boiler fabrication: demonstrating fine-grain practice for normalized or normalized-and-tempered alloy steel plate used in boilers, reactors, and piping.
- Fastener manufacturing: verifying grain size for high-strength bolts and studs where coarse grains degrade toughness and delayed-fracture resistance.
- Aerospace and energy forgings: acceptance testing of turbine shafts, discs, and structural forgings where grain size limits are mandatory.
- Weld procedure qualification: quantifying heat-affected-zone coarsening to optimize heat input and preheat practice.
- Failure analysis: linking intergranular fracture, quench cracks, or brittle service failures to abnormally coarse or mixed prior-austenite grains.
- Materials research: measuring grain-growth kinetics and grain-coarsening temperatures, and developing microalloyed steels with stable fine grains.
FAQ
Question: Why is the measured grain size in hardened alloy steel different from the martensite structure visible under the microscope?
Answer: In the quenched condition, martensite packets, blocks, and laths are much finer than the parent austenite envelope. The property-governing dimension is the prior-austenite grain size, so specialized etching is used to reveal those original boundaries before measurement.
Question: Which standards govern the test?
Answer: Average grain size is measured to ASTM E112, ISO 643, or GB/T 6394; the largest observed grain to ASTM E930; duplex grain structures to ASTM E1181; and image-analysis-based measurement to ASTM E1382. Japanese specifications commonly reference JIS G 0551. The applicable standard is selected by the purchaser's specification.
Question: What acceptance criteria are typical?
Answer: Requirements vary by specification, but quenched-and-tempered alloy steel components commonly require grain size G5 or finer, fine-grain pressure vessel plate typically requires G6 or finer, and carburizing gear steels are commonly required to remain within G5 to G8 after the McQuaid-Ehn cycle. The governing product or purchaser specification always prevails.
Question: Why does nital fail to reveal prior-austenite grain boundaries?
Answer: Nital etches the martensite matrix uniformly and shows packet and lath structure rather than prior-austenite boundaries. A saturated aqueous picric acid etch with a wetting agent, the oxidation method, or boundary precipitation methods must be used instead. Note that picric acid reagents must be kept wet and handled under controlled laboratory safety procedures.
Question: What is the McQuaid-Ehn test?
Answer: It is a standardized simulation in which specimens are carburized at about 925 °C for 8 hours and slowly cooled so that a cementite network precipitates along the prior-austenite boundaries. It evaluates the grain-growth tendency of carburizing steels under realistic case-hardening conditions.
Question: How is the ASTM grain size number converted to a diameter?
Answer: G is defined by N = 2^(G−1), where N is the grains per square inch at 100×. Approximately, G5 corresponds to 65 µm, G6 to 46 µm, G7 to 32 µm, and G8 to 23 µm mean diameter; ASTM E112 provides exact conversion tables and formulas.
Question: What causes duplex or mixed grain size, and how is it handled?
Answer: Duplex structures arise from abnormal grain growth after pinning precipitates dissolve, from critical-range hot-working reduction, or from non-uniform austenitizing. They are characterized per ASTM E1181 by reporting the sizes and area fractions of the coarse and fine populations, since a single average number would be misleading.
Question: How many specimens are needed and what is the typical turnaround?
Answer: Routine acceptance testing usually requires one to three specimens per lot as defined by the product specification. Standard etching-based measurement is typically completed within a few working days, while McQuaid-Ehn testing requires the additional 8-hour carburizing cycle plus slow cooling and sectioning.