Issue 78

A. A. Frantsuzov et alii, Fracture and Structural Integrity, 78 (2026) 474-488; DOI: 10.3221/IGF-ESIS.78.27

Despite the well-established correlation between PAGS and SSC resistance [5,6], the reliable quantification of PAGS in modern high-strength OCTG steels remains a significant metallurgical challenge. Following quenching and tempering, the original austenite grain boundaries are obliterated by the martensitic transformation, making direct observation impossible and requiring usage of indirect characterization techniques that either reveal boundary traces through chemical contrast or reconstruct parent grains from crystallographic data. The most prevalent conventional approach involves chemical etching with picric acid-based solutions containing surfactants, which exploits the segregation of several solute elements, but most importantly phosphorus, to austenite grain boundaries during austenitization. Because the rapid cooling rates associated with quenching prevent the dissolution of these segregates, they remain as static markers that create localized electrochemical potential differences relative to the grain interior. Picric acid, functioning as a strong oxidizing organic agent, preferentially attacks these segregated regions, thereby delineating the prior austenite network. However, the efficacy of this method has diminished for contemporary high-grade steels characterized by ultra-fine PAGS (often ≤ 10 μ m), high chemical purity, and homogeneous martensitic distributions. In such materials, the degree of segregation is insufficient to produce continuous, high-contrast boundaries, resulting in faint, discontinuous, or controversial etch patterns that complicate accurate grain identification and introduce significant measurement uncertainty. Quantitative assessment of PAGS from etched micrographs traditionally relies on manual or semi-automated methods, including direct grain counting, linear intercept measurements, and comparison with standardized reference charts such as those specified in ASTM E112 [8]. While these approaches are entrenched in industrial quality control protocols due to their speed and simplicity, they suffer from inherent limitations regarding statistical robustness and operator subjectivity. The selection of representative fields of view, the interpretation of uncertain boundaries, and the exclusion of non-equiaxed or partially revealed grains can introduce substantial bias, particularly when dealing with the complex, non-uniform grain morphologies typical of thermomechanically processed steels. Advances in electron backscatter diffraction (EBSD) have enabled a methodology shift toward crystallography-based PAG reconstruction, leveraging the deterministic orientation relationships governing the martensitic transformation. Since each martensite variant inherits a specific crystallographic correspondence with its parent austenite grain, computational algorithms can reverse-engineer the prior austenite microstructure by grouping adjacent martensite laths or blocks that share a common parent orientation. Among the available tools, MTEX (an open-source texture analysis toolbox for MATLAB [9]) has emerged as a widely adopted platform for this purpose, offering sophisticated reconstruction routines that account for variant selection rules and misorientation tolerances. Unlike chemical etching, EBSD-based reconstruction provides an objective, quantitative grain identification procedure independent of etch quality, while simultaneously delivering rich morphological descriptors including individual grain area, aspect ratio, sphericity, and boundary character distribution. These parameters enable researchers to investigate not only mean grain size but also the influence of grain shape heterogeneity and topological connectivity on SSC susceptibility, potentially revealing microstructural features that conventional methods cannot resolve. Nevertheless, crystallographic reconstruction is not without its own constraints and sources of error. The accuracy of reconstructed PAG maps is critically dependent on EBSD data quality, including indexing success rate, angular resolution, and step size relative to the underlying microstructural scale. Poor pattern quality, surface preparation artifacts, or insufficient spatial resolution can lead to false variant assignments, artificial grain fragmentation, or failure to detect low angle sub-boundaries within parent grains. Additionally, reconstruction algorithms require user-defined parameters such as misorientation thresholds and minimum grain sizes, the selection of which can significantly affect outcomes and lacks universal standardization across laboratories. Most fundamentally, reconstructed PAGs are mathematical constructs rather than directly observed physical entities, raising important questions about their comparability with traditional etching-based measurements and their validity as actual prior austenite geometry. Discrepancies between the two methodologies may arise from differences in what each technique actually measures: etching reveals chemically decorated boundaries that may not correspond perfectly to crystallographic interfaces, while reconstruction assumes boundaries based on idealized transformation models that may not fully capture deviations caused by plastic deformation, incomplete transformation, or local compositional gradients. Therefore, validation studies correlating EBSD-reconstructed PAGS with both etched microstructures are essential to establish confidence in these advanced characterization approaches and will allow future appropriate application in correlation with failure analysis for sour resistance. While extensive comparative studies [7,10] have established the reliability of various prior austenite grain size characterization techniques for steels with moderate grain dimensions (typically 15 – 50 µm), where boundary resolution is straightforward and etching contrast is robust, significant knowledge gaps persist regarding the accuracy of these methods in the ultra-fine grain regime that is characteristic of modern high-strength OCTG steels. As PAGS approaches the lower threshold of approximately 10 µm or less, the limitations of conventional metallographic etching become pronounced due to insufficient segregation contrast, while EBSD-based reconstruction faces inherent challenges related to spatial resolution

476