SEM-EDS Analysis of Non-Metallic Inclusions in Pressure Vessel Steel

Scanning electron microscopy (SEM) combined with energy-dispersive X-ray spectroscopy (EDS) is the most definitive technique for characterizing non-metallic inclusions in pressure vessel steels such as Q345R, 15CrMoR, and SA516 Gr.70. This article explains how SEM-EDS identifies inclusion types, morphologies, sizes, and distributions, and why inclusion control is critical to pressure vessel integrity, hydrogen-induced cracking resistance, and compliance with ASTM E45, ASTM E2142, and ISO 4967. From routine cleanliness evaluation to failure analysis support, understanding SEM-based inclusion analysis enables better metallurgical and engineering decisions.

Technical Overview

Pressure vessel steels—including carbon and low-alloy grades such as Q345R, Q370R, 16MnDR, 15CrMoR, 14Cr1MoR, 12Cr2Mo1R, SA516 Gr.70, and SA387 Gr.22—are used in reactors, boilers, heat exchangers, and storage vessels operating under high pressure, elevated temperature, and often aggressive media. In this context, non-metallic inclusions are among the most influential microstructural defects governing safety and service life.

Inclusions originate primarily from deoxidation products, reoxidation during teeming, ladle slag and refractory entrainment, calcium treatment, and solidification segregation. Typical inclusion types include manganese sulfide (MnS), alumina (Al2O3), calcium aluminates, magnesium aluminate spinel, silicates, titanium nitride (TiN), and complex multiphase particles. Their effects on performance are well documented:

• Stress concentration at elongated sulfides and hard oxide clusters reduces transverse ductility and impact toughness.
• elongated MnS stringers and oxide bands act as initiation and propagation paths for hydrogen-induced cracking (HIC) in wet H2S (sour) service.
• Large globular oxides and silicate inclusions serve as fatigue crack initiation sites under cyclic loading.
• Inclusion clusters promote lamellar tearing and through-thickness anisotropy in heavy plate.

Conventional cleanliness rating by optical microscopy (ASTM E45, ISO 4967 / GB/T 10561) provides statistically standardized severity indexes but limited information on inclusion chemistry. Scanning electron microscopy with EDS overcomes this limitation by providing, simultaneously, high-resolution morphology, quantitative elemental composition, and—through automated feature analysis—large-population inclusion statistics with type classification per ASTM E2142.

Testing Samples

Plate and Forging Samples

Coupons are sectioned from plate or forging locations representative of the product, typically at the surface, quarter-thickness, and mid-thickness positions. Both longitudinal and transverse sections are prepared to reveal stringer-type inclusions elongated during rolling, while through-thickness (short-transverse) sections are essential when evaluating susceptibility to lamellar tearing and HIC. Samples are examined in the un-etched condition to preserve inclusion geometry and chemistry.

Weld Joint and HAZ Samples

Full-thickness cross-sections spanning weld metal, fusion line, and heat-affected zone (HAZ) are prepared to characterize weld-metal inclusions such as deoxidation oxides, silicates from flux reactions, and oxide islands at pass boundaries. This is particularly relevant for welding procedure qualification and investigation of weld-related cracking or toughness degradation.

Failure and Field-Returned Samples

Fracture surfaces, crack-path sections, and material adjacent to defects are extracted from failed vessels or components. The SEM examination correlates inclusion populations with crack initiation sites, while EDS identifies corrosive or foreign species associated with the failure. Even small irregular specimens, bolt sections, nozzle samples, and corrosion products can be accommodated.

Testing Items

SEM-EDS inclusion analysis for pressure vessel steels typically covers the following deliverables:

• Inclusion morphology characterization: two-dimensional shape, aspect ratio, clustering behavior, and internal structure of individual particles.
• Qualitative and semi-quantitative elemental composition of each inclusion by EDS point analysis, including Al, Si, Ca, Mg, Mn, Ti, S, O, and trace elements.
• Inclusion type classification into sulfide (Type A), alumina (Type B), silicate (Type C), globular oxide (Type D), and single large globular (DS) categories consistent with ASTM E45 terminology, extended to nitrides, spinels, and complex multiphase inclusions.
• Size distribution, number density, and area fraction statistics from automated feature analysis of scanned areas.
• Elemental X-ray mapping of inclusion-rich regions to visualize spatial association of S, Al, Ca, Ti, and O.
• Inclusion–crack association studies linking individual inclusions or clusters to microcrack initiation in failure analysis.
• Comparative cleanliness assessment between heats, suppliers, processing routes, or process improvements.

Testing Methods

Specimen Preparation

Samples are prepared in accordance with ASTM E3 metallographic practice. Sectioning is performed with adequate coolant to avoid thermal alteration, followed by mounting, grinding through successive SiC grit sizes, and polishing with diamond suspensions down to 1 µm, with a final oxide or colloidal silica polish recommended to eliminate relief and smearing. The surface must be free of embedded abrasive, pull-out, and staining. Because steel is electrically conductive, coating is generally unnecessary; if required for imaging stability on fractured or corroded surfaces, a thin removable conductive coating may be applied.

SEM Imaging Conditions

Secondary electron (SE) imaging is used for topographic detail on fracture surfaces, while backscattered electron (BSE) imaging provides atomic-number contrast that makes inclusions clearly distinguishable from the steel matrix—sulfides appear slightly brighter and silicates typically darker, enabling reliable segmentation. Routine imaging is performed at accelerating voltages of 15–20 kV to ensure adequate X-ray excitation for EDS, with lower voltages available for fine surface detail.

EDS Microanalysis

Each inclusion of interest is analyzed by EDS point spectra, with optional line scans across complex inclusions and area mapping of inclusion fields. Standardless quantitative processing yields semi-quantitative oxide or sulfide stoichiometries sufficient for type identification. It should be noted that light elements (O, N) are measured semi-quantitatively by EDS; phase identification therefore relies on the combination of characteristic metallic element ratios, morphology, and BSE contrast rather than oxygen content alone.

Automated Inclusion Feature Analysis

Automated SEM-EDS inclusion analysis is performed in accordance with ASTM E2142. The system scans programmed areas by BSE imaging, thresholds inclusion features against the matrix, relocates to each particle, acquires EDS spectra, and classifies each feature by rule-based chemistry logic (for example, Mn–S as sulfide, Al–O as alumina, Ca–Al–O as calcium aluminate, Ti–N as nitride, Mg–Al–O as spinel). The output includes size and aspect-ratio distributions, number densities per class, area fractions, and extreme-value statistics for the largest inclusions. This approach provides far greater chemical resolution than chart-based optical rating and is particularly valuable for clean-steel development and data-rich acceptance criteria.

Complementary Optical Rating

Where certification to standardized cleanliness indexes is required, ASTM E45 Method A or ISO 4967 / GB/T 10561 ratings are determined by optical microscopy on the same polished section. SEM-EDS then serves to confirm the chemical type of rated inclusions, resolve borderline classifications, and investigate disputed or unusually severe findings.

Applications

SEM-EDS inclusion analysis of pressure vessel steels supports a wide range of engineering and quality activities:

• Steel-mill quality control and clean-steel development, including optimization of deoxidation practice, calcium treatment, ladle refining, and tundish metallurgy.
• Acceptance evaluation of plates, forgings, and pipe for code-fabricated pressure vessels, boilers, and spherical tanks.
• Investigation of hydrogen-induced cracking and sulfide stress cracking in wet H2S service, including evaluation of inclusion morphology in HIC test panels per NACE TM0284.
• Failure analysis of fatigue cracks, brittle fractures, lamellar tearing, and weld defects where inclusions are suspected initiation sites.
• Welding procedure qualification and weld consumable evaluation through characterization of weld-metal inclusion content and its effect on toughness.
• Support for fitness-for-service assessments and remaining-life evaluations of vessels in continued service.
• Research and development of new pressure vessel grades with enhanced through-thickness properties and low-temperature toughness.

FAQ

What is the smallest inclusion size that can be detected?

With modern field-emission SEM systems, individual inclusions down to approximately 0.5 µm can be imaged and chemically analyzed; automated feature analysis is typically configured with a detection threshold of 0.5–1 µm depending on the required statistics and analysis time.

How does SEM-EDS analysis differ from ASTM E45 optical inclusion rating?

ASTM E45 (and the equivalent ISO 4967 / GB/T 10561) uses chart comparison on an optical microscope to assign severity indexes by morphological category, and is the standard basis for specification acceptance. SEM-EDS per ASTM E2142 adds chemical identification of each particle, larger measurable populations, and full size-distribution statistics. The two methods are complementary, and optical ratings are frequently confirmed by SEM when inclusion type or origin is in question.

Can the SEM determine whether an inclusion caused a crack?

Yes. By sectioning through the crack origin and combining SE imaging of the fracture surface with EDS of particles at the initiation site, it is possible to demonstrate spatial association between a specific inclusion and crack nucleation, which is central to many failure investigations.

Is conductive coating required for steel samples?

Polished steel is electrically conductive and is normally examined uncoated. Fracture surfaces with oxides or non-conductive corrosion products may be lightly carbon-coated to stabilize imaging; this coating does not affect inclusion analysis of the underlying polished section.

What sample dimensions are accepted?

Typical metallographic sections of approximately 15–25 mm square and a few millimeters thick are ideal, but the SEM chamber accommodates a wide range of sizes. Irregular fracture pieces and small components can usually be mounted and prepared without difficulty.

How should results be interpreted for sour service applications?

For HIC-resistant plate, attention is focused on the population of elongated MnS, oxide clusters, and Ca-aluminate stringers on the rolling plane, since these control hydrogen trap sites and crack-linking behavior. Results are interpreted alongside HIC exposure testing per NACE TM0284 and the applicable material specification limits.