Alloy Steel Non-Metallic Inclusion Testing: Standards, Methods, and Evaluation

Non-metallic inclusion testing evaluates the cleanliness of alloy steel by identifying and quantifying the type, size, morphology, and severity of oxide, sulfide, silicate, and nitride inclusions embedded in the steel matrix. Accredited third-party laboratories perform this testing in accordance with internationally recognized standards such as ASTM E45, ISO 4967, EN 10247, JIS G 0555, and GB/T 10561 to verify compliance with product specifications. This article explains the classification of inclusions, sampling and preparation requirements, evaluation methods, and typical industrial applications of non-metallic inclusion detection for alloy steels.

Technical Overview

Non-metallic inclusions are non-metallic particles, such as oxides, sulfides, silicates, and nitrides, that are dispersed within the steel matrix. They originate primarily from deoxidation reactions during steelmaking, entrapment of slag or refractory particles, reoxidation during casting, and sulfide precipitation during solidification. Because inclusions interrupt the continuity of the metallic matrix, they act as stress concentrators and can significantly reduce fatigue life, transverse ductility, impact toughness, formability, and surface finish quality.

In modern metallurgy, the level of non-metallic inclusions is one of the most important indicators of steel cleanliness. Premium grades such as bearing steels, spring steels, and aerospace structural steels impose strict limits on inclusion severity, since hard and brittle oxide inclusions are well-documented initiation sites for fatigue cracks. Quantitative inclusion rating therefore serves as a fundamental quality-control step in both mill production and incoming material inspection.

Standard microscopic rating systems classify inclusions into well-defined categories based on their morphology and deformation behavior under hot working:

• Type A — Sulfides (for example, manganese sulfide): elongated, plastic inclusions that deform into stringers during rolling or forging.
• Type B — Alumina (aluminum oxides): discontinuous chains or clusters of hard, non-deformable particles.
• Type C — Silicates: elongated inclusions with variable plasticity, often appearing as tapered streaks.
• Type D — Globular (point) oxides: randomly dispersed, rounded, non-deformable particles such as calcium aluminates or spinels.
• Type DS — Single large globular oxide: isolated coarse particles, typically exceeding approximately 13 micrometers in diameter, rated separately in standards such as ISO 4967.

Rating is performed on polished, unetched specimens at 100× magnification by comparing observed fields against standard reference diagrams, or by automated image analysis. The results determine whether the material meets the cleanliness requirements of the applicable product specification.

Testing Samples

Typical Sample Forms

Non-metallic inclusion testing can be carried out on virtually any product form of alloy steel, including:

• Continuous casting blooms, billets, and slabs
• Hot-rolled and cold-rolled bars, rods, and wire rod
• Forged bars, forgings, and rolled rings
• Plates, strips, and sheets
• Seamless and welded tubes
• Finished components such as bearings, springs, gears, crankshafts, fasteners, and connecting rods
• Failed parts submitted for failure analysis

Sampling Location and Orientation

Sampling must be representative of the material being evaluated. For semi-finished products, samples are typically taken at the mid-radius or quarter-thickness position, where inclusion accumulation is often most severe. Standard rating methods are performed on a longitudinal section, that is, a plane parallel to the principal deformation direction, because inclusion morphology after rolling or forging is a key element of the classification. Sampling frequency, sample count, and exact locations are normally defined by the applicable product specification or by agreement between the purchaser and supplier.

Specimen Preparation

Specimen preparation quality directly determines the validity of the rating. General metallographic preparation follows the guidance of ASTM E3, while specimens intended for automatic inclusion assessment are prepared in accordance with ASTM E768. The essential requirements are:

• Sectioning performed with coolant and controlled feed to avoid burning or plastic deformation of the surface layer
• Mounting when necessary to protect edges and maintain flatness
• Progressive grinding through successively finer abrasives
• Final polishing with diamond suspension of 1 micrometer or finer to a mirror finish
• A polished surface free of scratches, smearing, comet tails, and inclusion pull-out, since preparation artifacts can be mistaken for or can mask real inclusions
• Evaluation in the unetched condition, as required by all major rating standards

For chart comparison methods, the polished area is normally at least approximately 160 square millimeters to ensure statistically meaningful coverage of the material.

Testing Items

A comprehensive non-metallic inclusion evaluation typically covers the following items, which may be reported individually or in combination according to the specified standard:

• Inclusion classification by type: A (sulfide), B (alumina), C (silicate), D (globular oxide), and DS (single globular oxide)
• Severity rating for each type in both the thin and heavy series, usually on a scale from 0.5 to 3.0 in half-step increments
• Worst-field rating, that is, the most severe field observed for each inclusion type and series
• Average rating and statistical distribution of field ratings across the examined area
• Inclusion dimensions: length, width, aspect ratio, and equivalent diameter
• Number density, area fraction, and size distribution of inclusions
• Chemical composition of individual inclusions, determined by energy-dispersive X-ray spectroscopy
• Assessment of nitride and carbonitride particles, such as titanium nitride, where relevant
• Extreme value analysis to estimate the maximum inclusion size expected in a larger steel volume
• Verification of conformity with cleanliness limits specified in product standards or purchase specifications

Testing Methods

Optical Microscopic Comparison Rating

The classical and most widely specified method uses an optical metallographic microscope at 100× magnification. The polished surface is systematically traversed, and each field of view is compared with standard rating diagrams. The principal standards are ASTM E45, ISO 4967, GB/T 10561, EN 10247, and JIS G 0555. Under ASTM E45, several procedures are defined, of which Method A (worst-field rating against the reference charts) and Method B (severity distribution based on field counts) are the most commonly applied. Results are reported as ratings for each inclusion type and series, for example, A2, B1.5, C1, D0.5, with higher values indicating more severe inclusion content.

Automatic Image Analysis

Automated systems use digital microscopes or dedicated image analyzers to detect and measure inclusions through contrast-based segmentation. Practices such as ASTM E1245 and test methods such as ASTM E2142 govern this approach, with specimen preparation guided by ASTM E768. Image analysis provides objective, repeatable measurements of inclusion length, width, area, number, and area fraction, and it can generate complete size distributions that chart comparison alone cannot provide. It is particularly valuable for research, process optimization, and high-volume quality control.

Scanning Electron Microscopy with EDS

Scanning electron microscopy combined with energy-dispersive X-ray spectroscopy reveals inclusion morphology at high magnification and determines the elemental composition of individual particles. This distinguishes, for example, aluminum oxide from calcium aluminate, manganese sulfide from complex oxy-sulfide, or titanium nitride from other cubic particles. SEM-EDS is indispensable for identifying inclusion origins, supporting failure investigations, and validating chart-based classifications in disputed cases.

Extreme Value Analysis

Because fatigue failure is usually governed by the largest inclusion rather than the average population, extreme value analysis per ASTM E2283 is applied to critical components such as bearing and spring steels. The maximum inclusion size is measured in each of numerous equal inspection areas, and statistical treatment of the maxima is used to estimate the largest inclusion likely to exist in the full production volume. This predicted maximum size can be linked directly to fatigue strength prediction models used in component design.

Complementary Techniques

Several additional techniques extend inclusion evaluation beyond two-dimensional metallography:

• Immersion ultrasonic testing and scanning acoustic microscopy for the detection of coarse internal inclusions over large volumes
• X-ray computed tomography for three-dimensional localization and sizing of large inclusions
• Electrolytic extraction, in which the steel matrix is dissolved to isolate inclusions for weighing, morphology study, and compositional analysis
• Spark-source optical emission spectroscopy with pulse discrimination analysis to provide statistical cleanliness indicators from inclusion-forming element signals

Applications

Non-metallic inclusion testing is applied throughout the alloy steel supply chain, from steelmaking process control to final component qualification. Key application areas include:

• Bearing steels: high-carbon chromium bearing grades and carburizing bearing grades are subject to the most stringent inclusion limits, since hard B-type and D-type oxides are leading causes of rolling contact fatigue. Product standards such as ASTM A295 and GB/T 18254 define explicit maximum ratings.
• Spring steels: inclusion severity controls fatigue life in valve springs, suspension springs, and other highly stressed elastic components.
• Gear and transmission steels: cleanliness ratings support pitting and bending fatigue performance requirements.
• Aerospace structural steels: inclusion evaluation is part of the qualification and acceptance testing required for flight-critical parts.
• Pipeline and pressure vessel steels: elongated sulfide inclusions are evaluated because they promote hydrogen-induced cracking and anisotropic toughness in sour service environments.
• Railway, automotive, and fastener steels: inclusion ratings verify consistency for safety-related forged and machined components.
• Steel plant process optimization: inclusion data guide improvements in deoxidation practice, ladle metallurgy, argon stirring, tundish design, and casting protection.
• Failure analysis: inclusion identification and localization help establish root causes of fatigue fracture, cracking, and surface defects.

FAQ

Why are specimens examined in the polished, unetched condition?

Etching attacks inclusion boundaries, may partially dissolve certain inclusions, and introduces microstructural contrast that interferes with comparison against the standard diagrams. All major rating standards therefore require evaluation of a carefully polished, unetched surface to ensure that inclusion size and contrast are reproduced accurately.

What is the difference between ASTM E45 Method A and Method B?

Method A records the single worst field observed for each inclusion type and series across the entire polished surface, and is the most common acceptance criterion. Method B records the number of fields at each severity level, producing a statistical distribution that describes the overall inclusion population rather than only the most severe field.

Are ISO 4967, GB/T 10561, EN 10247, and JIS G 0555 equivalent?

They are closely related micrographic methods based on the same inclusion classification principles and rating diagrams. GB/T 10561 is technically aligned with ISO 4967, while EN 10247 and JIS G 0555 apply the same concepts with differences in chart design, classes, and reporting details. Because results are not always numerically identical between standards, the governing specification must always be stated when reporting.

Why is the longitudinal section used for rating?

Inclusions deform differently during rolling or forging, and their post-deformation shape, whether elongated sulfide stringers, broken alumina chains, or undeformed globular oxides, is the basis of the standard classification. A longitudinal section parallel to the working direction captures these characteristic morphologies and yields ratings consistent with the reference charts.

What do the terms thin and heavy mean?

For each inclusion type, the reference charts provide two series distinguished by particle thickness or width. Thin series inclusions are finer and narrower, while heavy series inclusions are thicker and generally more harmful to fatigue and toughness. Each series is rated independently, so a complete report lists ratings such as A thin, A heavy, B thin, B heavy, and so on.

How are acceptance criteria established?

Acceptance limits are set by the applicable product standard or by agreement between the purchaser and the manufacturer. Premium bearing grades, for example, commonly specify that no field may exceed a low severity level for B-type and D-type inclusions, while structural grades generally permit somewhat higher ratings. The test report must therefore reference the specific standard, method, and limits against which conformity is judged.

Can image analysis replace manual chart comparison?

Automated image analysis offers superior repeatability, full size distributions, and higher throughput, and it is increasingly specified for modern cleanliness assessment. However, chart comparison methods remain deeply embedded in specifications worldwide, so many laboratories offer both and correlate results when transitioning between methods for a given product family.