Energy dispersive spectroscopy (EDS) analysis of non-metallic inclusions is a key metallographic test for pressure vessel steel, revealing the type, morphology, size, and chemical composition of inclusions that govern toughness, fatigue life, and hydrogen-cracking resistance. By combining scanning electron microscopy (SEM) with EDS microanalysis, accredited third-party laboratories help steel producers, fabricators, and plant operators verify steel cleanliness, diagnose defects, and satisfy the requirements of ASTM, ISO, EN, and GB standards. This article explains the sample types, testing items, analytical methods, applications, and frequently asked questions related to SEM-EDS inclusion analysis for pressure vessel steels.
Technical Overview
Pressure vessel steels—including the carbon and low-alloy plate grades covered by GB 713 (for example Q245R, Q345R, 15CrMoR, 14Cr1MoR, 13MnNiMoNbR), ASTM A516/A516M, ASTM A387/A387M, and EN 10028—are used in boilers, reactors, heat exchangers, spherical tanks, and other equipment operating under sustained pressure, elevated temperature, and aggressive media. Because these components must contain stored energy safely over decades of service, the cleanliness of the steel is a first-order quality attribute.
Non-metallic inclusions are an inevitable consequence of deoxidation, desulfurization, refining, casting, and welding. Their influence on performance depends not only on size and quantity but also on composition and shape. Elongated manganese sulfide stringers reduce through-thickness ductility and provide initiation sites for hydrogen-induced cracking in wet hydrogen sulfide service; hard alumina and spinel particles promote fatigue initiation and surface defects; large exogenous inclusions can act as internal discontinuities that lower fracture toughness. Conversely, a controlled dispersion of fine, stable oxides can be deliberately engineered to refine the microstructure.
Energy dispersive spectroscopy, usually abbreviated EDS or EDX and performed in a scanning electron microscope, measures the characteristic X-rays generated when the electron beam strikes an inclusion. This provides an elemental fingerprint of each individual particle, typically at the micrometer scale. While optical methods such as ASTM E45 and ISO 4967 (or the equivalent GB/T 10561) rate overall inclusion severity, SEM-EDS identifies exactly what each inclusion is made of, which allows its metallurgical origin to be traced and its service significance to be judged.
The principal inclusion families encountered in pressure vessel steel include:
- Sulfides: MnS (the dominant type), CaS, and FeS, often elongated during hot rolling.
- Oxides: alumina (Al2O3), calcium aluminates (CaO–Al2O3 system), magnesia-alumina spinel (MgAl2O3), silica (SiO2), and complex deoxidation products.
- Silicates: deformable manganese–iron silicates that form stringers during thermomechanical processing.
- Nitrides and carbonitrides: TiN, Nb(C,N), and AlN, typically angular and stable at high temperature.
- Duplex inclusions: an oxide core surrounded by a sulfide rim, common in calcium-treated steels.
- Exogenous inclusions: entrained ladle or mold slag and eroded refractory fragments, usually large and irregular.
Testing Samples
SEM-EDS inclusion analysis can be applied to virtually any solid steel coupon with a properly polished surface. Typical sample types submitted to accredited third-party laboratories include the following.
Plate and Strip Samples
Full-thickness coupons cut from boiler and pressure vessel plates. For statistical cleanliness evaluation, the polished observation surface is prepared parallel to the rolling plane at approximately mid-thickness, with a polished area of about 300 mm² as prescribed by ISO 4967 and GB/T 10561. For targeted SEM-EDS characterization, smaller coupons of roughly 15 mm × 15 mm taken at mid-thickness, quarter-thickness, or sub-surface positions are sufficient.
Forging, Pipe and Component Samples
Coupons from forged heads, flanges, nozzles, tube sheets, and seamless pipe. Because forgings develop directional flow lines, specimens are usually prepared in longitudinal, transverse, and radial orientations so that inclusion alignment relative to the principal working stress can be assessed.
Weld Metal, Heat-Affected Zone and Overlay Samples
Transverse cross sections of welded joints, including the weld metal, fusion line, and heat-affected zone. These samples are examined for weld-metal deoxidation products, flux or slag entrapment, and coarse inclusions that may degrade joint toughness or correlate with lack-of-fusion indications.
Failure and Fracture Samples
Sections from cracked components, delamination surfaces, and fracture faces where inclusions are suspected of acting as crack initiation sites. Where possible, the fracture surface itself is examined first without cleaning, after which a polished section through the initiation region is prepared to associate the crack origin with specific particles.
Testing Items
A complete SEM-EDS inclusion characterization for pressure vessel steel typically covers the following items:
- Inclusion morphology: shape (globular, elongated, angular, dendritic), two-dimensional size, aspect ratio, and internal structure such as core-and-rim texture.
- Elemental composition: qualitative and semi-quantitative analysis of Mg, Al, Si, S, Ca, Ti, Mn, Cr, Nb, V, Fe, O, and other relevant elements.
- Inclusion type identification: classification of each particle as sulfide, oxide, silicate, nitride, carbonitride, or duplex inclusion.
- Exogenous inclusion identification: recognition of entrained slag or refractory particles by their characteristic compositions, for example CaO–Al2O3–SiO2 with MgO, or alumina-rich and zirconia-rich refractory fragments.
- Element distribution analysis: line scans and X-ray mapping that reveal zoning and multi-phase structure within individual inclusions.
- Population statistics: number density, size distribution, maximum particle size, area fraction, and clustering tendency over a defined observation area.
- Crack–inclusion association: documentation of whether inclusions are attached to, or nucleated from, cracks, voids, or delamination surfaces.
- Correlation with cleanliness ratings: side-by-side comparison with the severity indices obtained by ASTM E45 or ISO 4967 / GB/T 10561.
The deliverable test report normally contains backscattered and secondary electron micrographs, point-analysis spectra, quantitative element tables expressed in weight percent and atomic percent, representative element maps, a classified inventory of the analyzed inclusions, and an interpretive discussion of their probable steelmaking or welding origin.
Testing Methods
The analysis follows a well-established metallographic workflow designed to preserve the inclusions in their true condition and to generate X-ray data of adequate quality for classification.
Sample Preparation
Coupons are sectioned at low stress to avoid smearing, pulling out, or cracking the particles, then mounted and ground through successively finer silicon carbide papers. Polishing proceeds with diamond suspensions down to 1 µm, frequently finished with colloidal silica for a deformation-free surface as guided by ASTM E3 and ASTM E768. The surface is evaluated in the unetched condition so that inclusions remain visible and chemically intact. After ultrasonic cleaning and drying, most mounted steel samples are conductive enough to be examined directly; a thin conductive coating is applied only when charging occurs and its spectral interference is accounted for.
SEM-EDS Instrumentation and Parameters
A field-emission or thermionic SEM fitted with a silicon drift detector is used. Typical operating conditions are an accelerating voltage of 15–20 kV, sufficient to excite the characteristic lines of the elements of interest, and a probe current adjusted to give a detector dead time of roughly 20–40 percent. Backscattered electron imaging is the primary survey mode because inclusions contrast strongly with the iron matrix by atomic number. Magnification is calibrated so that inclusion dimensions can be measured directly from the images.
Analysis Procedure
- Survey imaging: the polished surface is scanned in backscattered electron mode to locate inclusions and document their distribution.
- Point analysis: the beam is positioned at the center of each selected inclusion and a spectrum is acquired for 30–100 seconds of live time to ensure adequate counting statistics.
- Line scans and mapping: for particles showing internal contrast or suspected duplex structure, element line scans across the particle and full X-ray maps are collected.
- Classification: each inclusion is assigned to a type according to its composition and morphology, following the classification framework of ASTM E2142.
- Statistics: the analyzed population is summarized by type, size class, and frequency, and the largest and most damaging particles are highlighted.
- Optional automated feature analysis: when large inclusion populations must be screened, automated SEM-EDS feature detection is used to analyze thousands of particles and produce statistically robust size and chemistry distributions.
Data Quantification and Applicable Standards
- ASTM E2142: standard test methods for inclusion analysis of steel by scanning electron microscopy and EDS-based classification.
- ASTM E45: determining the inclusion content of steel by the chart-rating method.
- ISO 4967 and GB/T 10561: micrographic determination of the non-metallic inclusion content using standard rating diagrams.
- ASTM E3: preparation of metallographic specimens.
- ASTM E768: preparing and evaluating specimens for automatic inclusion assessment of steel.
- ASTM E1508 and ISO 22309: guides for quantitative analysis by energy dispersive spectroscopy.
Because oxygen K-line analysis is intrinsically semi-quantitative, oxide stoichiometry is often used to convert measured cation ratios into probable phases, such as MgO·Al2O3 spinel or 12CaO·7Al2O3 calcium aluminate, which strengthens the metallurgical interpretation.
Applications
SEM-EDS inclusion analysis supports every stage of the pressure vessel life cycle, from steelmaking to post-service failure diagnosis:
- Incoming material qualification: verifying that plate and forging deliveries meet the cleanliness expectations implied by the purchase specification before fabrication begins.
- Steelmaking process optimization: distinguishing over- or under-calcium treatment, evaluating deoxidation practice, ladle slag control, and refractory performance so that inclusion populations can be pushed toward smaller, more dispersed, and less harmful particles.
- Hydrogen damage and sour service assessment: identifying elongated MnS and calcium-aluminate populations that govern susceptibility to hydrogen-induced cracking and blistering in wet hydrogen sulfide environments.
- Failure analysis: confirming whether fatigue cracks, lamellar tearing, brittle fractures, or weld defects initiated at specific inclusions, and attributing the origin to the steelmaking or welding process.
- Through-thickness property studies: correlating inclusion stringers with reduction of area in the thickness direction and with the performance of attachment welds on shell components.
- Weld procedure qualification and defect investigation: characterizing oxide and slag particles in weld metal that affect notch toughness and ultrasonic indications.
- Dispute resolution and certification: providing objective, instrument-based evidence of inclusion type and origin when supplier claims or batch quality are contested.
- Research and development: supporting inclusion engineering and oxide metallurgy programs aimed at refining grain size and improving heat-affected-zone toughness.
FAQ
What is the difference between inclusion rating and EDS inclusion analysis?
Inclusion rating by ASTM E45 or ISO 4967 / GB/T 10561 is an optical microscopy method that assigns severity indices to inclusion categories on a polished surface, giving a fast statistical measure of steel cleanliness. EDS analysis, usually performed in an SEM, does not replace the rating but adds compositional identification of individual particles, so each inclusion can be named, its origin inferred, and its individual significance to the intended service assessed.
How small an inclusion can be analyzed by EDS?
The practical lower limit is about 0.5 to 1 µm. Below this size, the electron interaction volume exceeds the particle diameter and the spectrum becomes contaminated by the surrounding steel matrix. Fine sub-micrometer dispersions are better characterized statistically or by complementary high-resolution techniques.
Can EDS accurately quantify oxygen in oxide inclusions?
Only semi-quantitatively. Oxygen has a low-energy characteristic X-ray that is strongly absorbed in the detector window and in the specimen itself, so reported oxygen values should be treated as indicative. Analysts therefore rely primarily on the cation ratios and convert them to oxide phases by stoichiometry, which gives a reliable inclusion identification.
How many inclusions are typically analyzed?
For a targeted investigation, 10 to 30 representative inclusions of each observed population are usually sufficient for classification. For statistical cleanliness assessment or litigation support, automated feature analysis of several hundred to several thousand particles across a defined area, such as the 300 mm² surface of ISO 4967, is recommended.
Is the test destructive?
A coupon must be cut from the material and destroyed during metallographic preparation, so the test is destructive on a local scale. On finished components, sampling locations are selected in non-critical regions or on excess material such as test prolongations, which is standard practice in pressure vessel quality plans.
How long does SEM-EDS inclusion analysis take?
For routine samples, preparation, microscope time, data reduction, and reporting typically require three to seven working days. Complex failure analyses involving fracture-surface examination, multiple sections, or automated large-population screening may require additional time, and expedited scheduling is generally available from accredited third-party laboratories.
Which sample orientation should be examined?
For rating purposes, the surface is prepared parallel to the rolling plane near mid-thickness as specified in ISO 4967. For engineering interpretation, a longitudinal section is often preferred because it displays elongated sulfide and silicate stringers in their most damaging form and allows the inclusion shape to be related to the principal stress direction in the vessel.