Morphological Observation and Analysis of Non-Metallic Inclusions in Pressure Vessel Steel

Non-metallic inclusions are among the most influential microstructural features governing the safety, toughness, and service reliability of pressure vessel steel. This article provides a systematic overview of inclusion morphology observation and analysis, combining optical microscopy rating per ASTM E45, ISO 4967, JIS G 0555, and GB/T 10561 with SEM-EDS compositional identification and automated image analysis. Read on to understand how inclusion type, size, shape, and distribution are characterized, rated, and controlled in pressure-vessel-grade plates, forgings, and weldments.

Technical Overview

Pressure vessel steels, including carbon steel plates to ASTM A516/A516M and GB/T 713 (Q345R), chromium-molybdenum alloy plates to ASTM A387, and normalized fine-grain steels to EN 10028-2, operate under sustained pressure, thermal cycling, and often aggressive media. Because non-metallic inclusions act as local stress concentrators, hydrogen trap sites, and fatigue crack initiation points, their content and morphology are critical acceptance characteristics. Morphological observation and analysis therefore forms an essential part of quality control, failure analysis, and clean-steel development for pressure vessel applications.

Inclusions originate from deoxidation reactions, alloying additions, sulfide precipitation during solidification, reoxidation during casting, and entrainment of slag or refractory particles. Their morphology is determined largely by steelmaking practice: aluminum killing generates alumina, calcium treatment modifies alumina into globular calcium aluminates, and manganese sulfides deform into elongated stringers during rolling. Because morphology, more than simple quantity, controls the effect on toughness, anisotropy, and hydrogen-induced cracking (HIC) resistance, a systematic observation program must capture shape, size, distribution, and chemistry together.

• Type A, sulfide inclusions (MnS): ductile, elongated stringers or bands aligned with the rolling direction, typically light-gray and smooth-edged under bright-field optical microscopy.
• Type B, alumina inclusions (Al2O3): brittle, angular or irregular particles that fracture and align into broken chains or clusters during hot rolling.
• Type C, silicate inclusions: deformable, dark-gray elongated inclusions of variable length, often with visible internal elongation.
• Type D, globular oxide inclusions: rounded, undeformable oxides such as calcium aluminates, randomly distributed.
• Type DS, single large globular oxides: coarse isolated particles typically above approximately 13 micrometers, addressed specifically in ISO 4967 and GB/T 10561.
• Titanium carbonitrides, Ti(C,N): angular, cubic, or blocky particles with a characteristic yellow-orange optical appearance.
• Exogenous inclusions: large, irregular, multi-phase aggregates derived from slag or refractory entrainment, frequently flagged as ultrasonic indications.

The evaluation framework is defined by recognized standards: ASTM E45 (chart-based inclusion rating, including worst-field Method A), ISO 4967 and GB/T 10561 (equivalent diagram methods incorporating the DS category), JIS G 0555 (microscopic method with cleanliness indices), ASTM E1245 (inclusion content by automatic image analysis), and ASTM E7 (metallographic terminology). Accredited third-party laboratories apply these methods within the general requirements of product standards such as ASTM A20/A20M.

Testing Samples

Plate and Strip Samples

Samples are cut from pressure vessel plate at locations and orientations specified by the applicable product standard, typically quarter-width and quarter-thickness positions. Both longitudinal and transverse polished sections are prepared because sulfide stringers elongate in the rolling direction and oxide chains fracture across it, so orientation strongly affects the observed severity. ASTM E45 requires a minimum polished area of 160 mm², while ISO 4967 and GB/T 10561 specify an examination area of 200 mm² sampled in systematically spaced fields.

Forging, Pipe, and Cast Component Samples

For forgings, pipe hollows, and castings, specimens are taken with defined relationships to the principal metal-flow direction, since inclusions align with working direction and segregate toward mid-thickness and last-solidified regions. Mapping sampling position to solidification history allows inclusions to be traced back to ladle, tundish, or mold events.

Weld, HAZ, and Clad Samples

Cross-sections of weld metal, heat-affected zone, and cladding interfaces are examined for weld-metal deoxidation products, slag islands, and lack-of-fusion indications. Through-thickness sections of T-joints and nozzle attachments are especially relevant for lamellar tearing assessment, where banded sulfides and oxide stringers dominate susceptibility.

Service-Exposed and Failure Samples

Specimens containing fracture surfaces, fatigue origins, or stepwise HIC cracks are sectioned so that the polished plane intersects the crack path. Correlating inclusion morphology and position with crack initiation sites is one of the most powerful diagnostic outputs of this analysis.

Testing Items

Depending on the engineering objective, the scope of testing typically includes the following items, which may be selected individually or combined into a full characterization package:

• Morphological characterization: shape (globular, angular, elongated, stringer, dendritic, clustered), optical color and contrast, and distribution pattern (random, banded, chain-like, or segregation-associated).
• Classification and severity rating: assignment of Types A, B, C, D, and DS with thin-series and thick-series severity numbers from 0.5 to 3.0, plus worst-field ratings.
• Dimensional statistics: length, width, thickness, aspect ratio, equivalent circle diameter, and maximum particle dimensions.
• Quantitative cleanliness indices: area fraction, number of inclusions per mm², total inclusion length per mm², and statistical distribution across the examined area.
• Compositional identification: point analysis, line scans, and element mapping of individual particles to establish chemistry, for example Al2O3, MnS, calcium aluminates, magnesium aluminate spinels, and duplex oxide-core/sulfide-rim structures.
• Origin diagnosis: differentiation of endogenous deoxidation products from exogenous slag or refractory entrainment, and assessment of calcium-treatment effectiveness.

Testing Methods

Specimen Preparation

Careful metallographic preparation is decisive for valid inclusion rating. Samples are sectioned with abrasive cutting to avoid burning or smearing, mounted, ground through successive silicon-carbide grits, and polished with diamond suspensions down to 1 micrometer (optionally finished with colloidal silica) to obtain a scratch-free, relief-free surface. Rating is performed on the as-polished, unetched condition because etching attacks the matrix and alters the apparent size, color, and contrast of inclusions. Pull-out, comet-tailing, and abrasive embedding must be prevented, as they mimic or obscure real inclusions.

Optical Microscopy and Chart Rating

Bright-field examination at 100× magnification is performed with a calibrated field of view. Inclusion morphology is compared against standard diagrams, and severity ratings are assigned to thin and thick series of each inclusion type. ASTM E45 Method A records the worst field for each type and series, while ISO 4967 and GB/T 10561 apply an equivalent diagram method over a defined grid of fields and add the DS category for single large globular oxides. JIS G 0555 offers an alternative microscopic cleanliness index. Dark-field or polarized-light observation may be used to reveal internal structure and duplex character of complex inclusions.

Scanning Electron Microscopy with EDS

Scanning electron microscopy provides high-resolution secondary-electron and backscattered-electron imaging, in which atomic-number contrast separates oxide, sulfide, and nitride phases from the steel matrix. Energy-dispersive X-ray spectroscopy delivers qualitative and semi-quantitative element analysis of individual particles and element maps that expose core-rim structures, for example a calcium-aluminate core surrounded by a manganese sulfide rim. Electrolytic extraction of inclusions from the matrix can additionally reveal their true three-dimensional morphology, which two-dimensional sections only approximate. Because oxygen quantification by EDS is semi-quantitative, oxide stoichiometry is typically inferred from the measured cation ratios, with wavelength-dispersive techniques applied where higher accuracy for light elements is required.

Automated Image Analysis

Per ASTM E1245, automated stage-scanning systems acquire large numbers of fields and apply gray-level threshold segmentation to measure inclusion length, width, aspect ratio, and area fraction statistically. This approach reduces operator subjectivity, increases the examined area by an order of magnitude, and generates particle-size distributions suitable for cleanliness benchmarking and statistical process control. Results should be periodically cross-verified against manual chart ratings, since gray-level overlap between inclusion types and matrix features can bias automatic classification.

Reporting and Data Interpretation

The final report presents type and series ratings, worst-field results, quantitative cleanliness statistics, inclusion distribution maps, representative micrographs, and SEM-EDS spectra with elemental tables. Results are interpreted against the acceptance criteria in the purchase specification, correlated with Charpy toughness, fatigue, and HIC test data where available, and translated into steelmaking guidance such as deoxidation control, calcium-treatment optimization, or improved slag and refractory management.

Applications

Inclusion morphology analysis supports the entire life cycle of pressure vessel materials, from steelmaking control through fabrication qualification to failure investigation:

• Release and incoming inspection of pressure vessel plate for boiler drums, reactors, heat exchangers, spherical storage tanks, and low-temperature or LNG service.
• Supplier qualification and cleanliness benchmarking across heats, ladles, and casting sequences.
• Sour-service qualification programs, where elongated MnS stringers and clustered oxides are the primary hydrogen-induced cracking nucleation sites, supported by HIC test methods such as EN 10229.
• Failure analysis of fatigue cracks initiating at subsurface oxide clusters, lamellar tearing at welded attachments, and brittle fracture associated with banded inclusion arrays.
• Through-thickness property verification for Z-quality requirements, for example per EN 10164, for restrained welded connections.
• Research and development of clean steel: optimization of deoxidation practice, secondary refining, calcium treatment, and casting parameters, and evaluation of inclusion modification in aluminum-killed and calcium-treated grades.
• Fitness-for-service and remaining-life assessments of operating equipment, where inclusion populations inform fracture-toughness and crack-growth evaluations.

FAQ

Why must inclusion rating be performed on a polished, unetched surface?

Etching attacks the matrix surrounding inclusions, changing their apparent size, color, and contrast and dissolving sensitive phases such as certain sulfides and silicates. All chart-based standards therefore require evaluation in the as-polished condition to ensure ratings are comparable between laboratories.

What is the difference between ASTM E45 and ISO 4967 / GB/T 10561?

All three use equivalent inclusion diagrams derived from the same classification (A sulfides, B alumina, C silicates, D globular oxides) at 100× magnification. ISO 4967 and GB/T 10561 specify a 200 mm² examination area over a defined field grid and add the DS category for single large globular oxides, whereas ASTM E45 Method A reports worst-field ratings on a minimum polished area of 160 mm². In practice, results are closely comparable when sampling is consistent.

Which inclusions are most harmful in pressure vessel steel?

It depends on the degradation mechanism. Elongated Type A sulfides dominate anisotropy, through-thickness ductility, and HIC susceptibility; hard Type B alumina clusters and Type DS particles are the leading fatigue-initiation sites; and large exogenous inclusions can generate ultrasonic indications and initiate lamellar tearing. Size, shape, and distribution matter as much as chemical type, which is why full morphological analysis is preferred over simple counting.

How are inclusions in weld metal evaluated?

Polished transverse cross-sections of weld metal and heat-affected zone are rated using the same chart methods. Weld metal typically contains fine, uniformly dispersed deoxidation products; isolated coarse slag particles indicate flux entrapment or insufficient inter-pass cleaning and are reported separately with their composition.

What acceptance limits are typically applied?

Acceptance limits are defined by the product specification or purchase agreement rather than by the test method itself. For pressure vessel plate, purchasers commonly specify maximum thin-series and thick-series severities in the 1.0 to 2.5 range for each inclusion type, with stricter limits for critical or sour-service applications. Accredited third-party laboratories report measured values against the specified limits without imposing universal thresholds.

Can inclusion morphology indicate the steelmaking origin of a defect?

Yes. Chemistry and morphology together are strongly diagnostic: clustered angular alumina indicates deoxidation or reoxidation products; spherical calcium aluminates indicate effective calcium treatment; magnesium- or titanium-bearing multi-phase aggregates point to refractory interaction; and large irregular slag-type inclusions indicate ladle or tundish entrainment, guiding targeted corrective action in the melt shop.