Spark Mapping Analysis for Segregation Partitioning in Large-Scale Super-Critical-Power Steel

The material properties of P91 steel, a critical high-temperature heat-resistant steel, are critically dependent on the uniformity of its macro-composition distribution. This paper presents the first application of Spark Mapping Analysis for Large Samples (SMALS) for the non-destructive, full-field characterization of macro-composition distribution in P91 steel ingots and finished tubes. To address the analytical challenges posed by large-sized specimens, an innovative partition-based statistical analysis model was developed, enabling the effective demarcation of large-scale macro-segregation areas. Utilizing Sample A as the paradigm, a systematic methodology and workflow for the partition analysis were established, successfully identifying and quantifying the widths of its positive and negative segregation bands (namely 6 mm, 20 mm, and 8 mm). This approach was subsequently applied to samples from different smelting batches (B1, B2, C1, C2), effectively delineating macro-segregation areas within each sample and performing quantitative evaluations based on the statistical upper segregation limit. The findings provide essential experimental insights into the full-field compositional heterogeneity of P91 steel and deliver critical methodological guidelines for optimizing steel smelting processes to control and mitigate macro-segregation.