Determination of Maximum Inclusion Size in Pressure Vessel Steel: Methods, Standards, and Applicatio

Maximum inclusion size determination is a critical steel cleanliness assessment for pressure vessel components, quantifying the largest non-metallic inclusions that govern fatigue crack initiation, hydrogen damage resistance, and fracture safety in service. This technical guide explains how accredited third-party laboratories sample, prepare, measure, and statistically extrapolate inclusion sizes in pressure vessel steel in accordance with ASTM E45, ASTM E2283, ASTM E1245, ISO 4967, GB/T 10561, and JIS G 0555.

Technical Overview

Non-metallic inclusions are unavoidable constituents of pressure vessel steel, originating from deoxidation products (alumina, silicates), sulfide precipitation (MnS), nitrides and carbonitrides (TiN, Nb(C,N)), refractory erosion, slag entrapment, and reoxidation during casting. Although inclusions are present at very low volume fractions, their size, shape, distribution, and chemistry exert a decisive influence on the structural integrity of boilers, reactors, heat exchangers, and storage vessels operating under sustained pressure, cyclic loading, and aggressive media.

For pressure vessel integrity, the governing parameter is typically the maximum inclusion size rather than the average cleanliness level. The single largest hard oxide inclusion is the most probable site for fatigue crack initiation, particularly in the very-high-cycle fatigue regime, and elongated manganese sulfide stringers promote anisotropic toughness loss, lamellar tearing, and hydrogen-induced cracking. Because only an extremely small fraction of a component volume can be examined metallographically, the true maximum inclusion size is a statistical quantity that must be estimated through a combination of microscopic rating, quantitative image analysis, and extreme value statistics.

Established evaluation frameworks classify inclusions into five categories per ISO 4967, ASTM E45, and GB/T 10561: Type A (sulfides), Type B (alumina, fragmented stringers), Type C (silicates, elongated), Type D (dispersed globular oxides), and Type DS (single large globular inclusions). Quantitative maximum-size determination and extrapolation to larger steel volumes are performed under ASTM E2283 (extreme value analysis), ASTM E1245 (image analysis), and EN 10247 (automated inclusion rating). These methods support material specifications such as ASTM A20, ASTM A516, ASTM A387, EN 10028, GB 713, and the material rules of the ASME Boiler and Pressure Vessel Code.

Testing Samples

Boiler and Pressure Vessel Plate

Carbon and low-alloy steel plates supplied to ASTM A516, ASTM A387, EN 10028, or GB 713 are sampled at representative locations, typically at quarter-thickness or mid-thickness positions, with both longitudinal and transverse sections prepared so that stringer-shaped inclusions (Types A, B, C) are correctly oriented relative to the principal working direction.

Forgings

Nozzles, flanges, tubesheets, ring forgings, and shaft forgings are sectioned at locations defined by the applicable product specification, usually at mid-radius, with the polished plane aligned to capture the maximum inclusion elongation in the principal forging direction.

Pipe and Tube

Seamless and welded pressure piping samples are taken from body, weld, and heat-affected zone locations, with transverse and longitudinal metallographic sections prepared to compare inclusion populations in base metal and weld regions.

Weld Metal and Heat-Affected Zone

For welded pressure components, cross sections through the weld joint are prepared to characterize oxide inclusions in weld metal, which control weld metal toughness and fatigue performance, and to detect exogenous slag entrapment or lack-of-fusion features that may be inclusion-related.

Specimen Preparation Requirements

Specimens are cut by abrasive or precision sawing with abundant coolant to avoid thermal alteration and plastic deformation of inclusions. After mounting, samples are ground through successive silicon carbide grades and polished with diamond suspensions down to 1 µm or finer to produce a scratch-free, feature-preserving mirror surface. Examination is performed in the unetched condition, because etching obscures inclusion boundaries. For chart rating per ASTM E45, a minimum polished area of 160 mm² (approximately 320 fields of 0.5 mm² at 100× magnification) is examined per specimen, with special care taken to preserve edge retention and prevent smearing of soft sulfides.

Testing Items

• Inclusion morphology classification into Types A, B, C, D, and DS per ISO 4967 / ASTM E45 / GB/T 10561.
• Individual inclusion dimensions: length, width (thickness), equivalent diameter, projected area, and the square-root-of-area parameter widely used in fatigue modeling.
• Maximum observed inclusion size per specimen, per location, and per heat, reported by type.
• Worst-field severity ratings in the thin and thick series for each inclusion type.
• Statistically extrapolated maximum inclusion size for a defined reference area or volume, with confidence intervals, per ASTM E2283.
• Inclusion area fraction, number density, size distribution, and aspect ratio (anisotropy index).
• Chemical composition of the largest inclusions by scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDS), distinguishing deoxidation products, complex calcium aluminates, exogenous slag, and refractory particles.

Testing Methods

Optical Microscopy Chart Rating (ASTM E45 Method A, ISO 4967, GB/T 10561, JIS G 0555)

The classicalJK-chart method examines at least 160 mm² of polished surface at 100× magnification using a calibrated field of view of 0.5 mm². Each field is compared against standard rating diagrams, and the worst field for each inclusion type and thickness series is reported. The DS scale expresses the diameter of single large globular inclusions, with rating thresholds increasing from approximately 13 µm at DS 0.5 to more than 300 µm at DS 5. Chart rating provides a robust, widely specified acceptance index, and the DS rating gives a direct, standardized indication of maximum inclusion size in the examined plane.

Quantitative Automated Image Analysis (ASTM E1245, EN 10247)

Digitized micrographs acquired under calibrated optical conditions are segmented by gray-level thresholding to measure every inclusion object. The method outputs precise maximum length, width, equivalent diameter, and area statistics, together with number density and area fraction. Automated image analysis removes operator subjectivity, produces full size distributions rather than discrete ratings, and supplies the objective dimensional data required for extreme value analysis. Threshold settings must be validated carefully to avoid counting polishing pits, dust, or etch artifacts as inclusions.

Extreme Value Statistics (ASTM E2283)

Because the physically largest inclusion in a plate or forging is almost never present in the examined metallographic section, ASTM E2283 provides a rigorous statistical extrapolation procedure:

• Define a fixed inspection field area (commonly 0.5 mm² at 100×) and examine a large number of fields, typically one hundred to several hundred, distributed across the polished surface.
• Measure the characteristic size (length or square root of projected area) of the largest inclusion in each field.
• Rank the maxima and plot them against the reduced variate of the Gumbel (Type I) extreme value distribution.
• Fit the linearized distribution by least squares and verify goodness of fit.
• Extrapolate to a return period equal to the ratio of the reference area of interest (for example, the full plate surface or the surface stressed above a defined threshold) to the field area, yielding the predicted maximum inclusion size with quantified confidence bounds.

The method assumes a single, homogeneous inclusion population with an unbounded exponential tail. For ultra-clean steels in which most fields contain no inclusions, modified approaches based on maxima per specimen block or on peaks-over-threshold distributions are applied, and larger inspected areas are used to maintain confidence in the extrapolated maximum.

SEM-EDS Verification of the Largest Inclusions

The few largest inclusions identified optically are examined by SEM to confirm their identity, morphology, and chemistry. Typical findings include corundum alumina clusters, calcium aluminates from calcium treatment, manganese sulfide stringers, titanium nitride cubes, and complex exogenous particles whose composition indicates slag or refractory entrainment, providing direct feedback to steelmaking process control.

Supplementary Volumetric Techniques

For screening very large steel volumes that cannot be covered metallographically, accredited third-party laboratories apply high-frequency ultrasonic scanning, which detects compact inclusions above the tens-of-micrometers range; X-ray computed microtomography for three-dimensional size measurement; electrolytic extraction, which dissolves the steel matrix and recovers inclusions for direct three-dimensional sizing and mass-fraction determination; and electron beam button melting, which aggregates inclusions from a melt sample into a surface button for rapid screening of the largest particles in clean steels.

The measured or extrapolated maximum size feeds directly into fitness calculations, most notably fatigue assessments based on the square-root-of-area parameter model, in which the predicted fatigue limit decreases monotonically with increasing inclusion size, and into hydrogen damage susceptibility evaluations for wet sour service.

Applications

• Acceptance testing and quality release of carbon and alloy steel plate, forgings, and pipe for boilers, pressure vessels, heat exchangers, and reactors.
• Fatigue design verification and very-high-cycle fatigue life assessment of safety-critical pressure boundary components.
• Qualification of steels for wet sour hydrocarbon service against hydrogen-induced cracking, stepwise cracking, and sulfide stress cracking requirements referenced in ISO 15156.
• Failure analysis of leaking, cracked, or brittle-fractured vessels, identifying inclusion-related initiation sites and their metallurgical origin.
• Steelmaking process optimization, including deoxidation practice, ladle metallurgy, calcium treatment, tundish and shrouding design, and casting cleanliness control.
• Welding procedure qualification and weld consumable evaluation, where weld metal oxide inclusion content governs toughness.
• Remaining-life and fitness-for-service assessments of aging pressure equipment in refining, petrochemical, and power generation plants.
• Research into inclusion engineering and oxide metallurgy for grain-refined, high-toughness pressure vessel steels.

FAQ

Why is the maximum inclusion size, rather than the average cleanliness rating, decisive for pressure vessel safety?

Failure processes such as fatigue crack initiation and brittle fracture are controlled by extreme defects, not by averages. A single large oxide inclusion can act as an intrinsic crack starter whose severity scales with its size, so a heat of steel with an excellent average rating but one oversized inclusion may still be unsuitable for cyclic or sour service. Extreme value statistics were introduced precisely to quantify these rare, life-limiting defects.

What does the DS category mean in the rating standards?

DS designates single, large, approximately globular inclusions that occur individually rather than in aligned strings or dispersed groups. The DS rating scale converts the observed diameter into a standardized index, from about 13 µm at DS 0.5 up to more than 300 µm at DS 5, making it a direct contractual instrument for limiting maximum inclusion size in pressure vessel plate specifications.

How can a laboratory state an inclusion size larger than anything actually observed?

Under ASTM E2283, the largest inclusion measured in each of many equally sized fields follows the statistics of extremes. By fitting a Gumbel distribution to the field maxima and extrapolating to the return period corresponding to the total area of engineering interest, the laboratory predicts the most probable maximum inclusion size in that area, with confidence intervals reflecting the number of fields examined. The result is a statistical estimate, and its reliability improves as the inspected area increases.

How much surface area must be examined?

For chart rating, ASTM E45, ISO 4967, and GB/T 10561 require at least 160 mm² of polished surface per specimen at 100×. For extreme value analysis, several hundred fields totaling on the order of 100 to 300 mm² or more are typically inspected, with cleaner steels requiring larger inspected areas to keep the confidence interval of the extrapolated maximum acceptably narrow.

What acceptance criteria are commonly applied to pressure vessel steel?

Individual purchaser specifications vary, but typical plate requirements limit thin-series ratings to approximately 2.0 to 2.5 and thick-series ratings to 1.0 to 1.5 for Types A through D, with DS ratings restricted to roughly 1.5 to 2.5 depending on wall thickness, design fatigue life, and service medium. More severe cyclic or hydrogen-charging service justifies tighter DS limits and formal extreme value acceptance thresholds.

Can the standard JK rating charts be applied to weld metal?

The charts were developed for wrought steel microstructures and are not directly representative of weld metal inclusion populations, which are dominated by finely dispersed deoxidation oxides. Weld metal is therefore more reliably evaluated by quantitative image analysis, with acceptance criteria agreed between the purchaser, fabricator, and accredited third-party laboratory on the basis of measured size distributions rather than JK ratings.

How is the measured maximum inclusion size used in fatigue assessment?

The square-root-of-projected-area parameter of the largest surface or subsurface inclusion is the key input to established fatigue limit models, in which predicted fatigue strength decreases as this parameter increases. Combining ASTM E2283 extrapolation with such models allows designers to select cleanliness levels consistent with the required design life, including the giga-cycle regime where subsurface inclusion initiation dominates.