Formation and characterization of non-metallic inclusions in steel produced by melting secondary ferrous scrap in an electric arc furnace: A case study of AISI 1066 steel
DOI:
https://doi.org/10.36547/ams.32.2.2284Keywords:
non-metallic inclusions, scrap-based, inclusion size distribution, deep learning, oxide and sulfide, metallurgical cleanlinessAbstract
This study investigated the formation of non-metallic inclusions and their characterization during the melting of AISI1066 steel from secondary metal scrap in an electric arc furnace. The analyses were carried out using scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS). In addition, a deep learning-based segmentation method was applied to improve the identification of inclusions. The results confirmed a heterogeneous microstructure consisting of non-metallic inclusions dispersed within a Fe-based matrix. The size of the inclusions ranged from a few micrometers up to approximately 40 µm. The majority of inclusions were found within the range of 10–25 µm. SEM analysis showed that the inclusions predominantly exhibit irregular and globular morphologies. EDS results confirmed that they consist of manganese sulfides (MnS), oxides (Al₂O₃ and SiO₂), and complex multi-phase particles. The elevated sulfur (S) (0.12 wt.%) and copper (Cu) (0.81 wt.%) contents in the investigated samples indicate that deoxidation and modification processes were not sufficiently effective.
The deep learning-based model made it possible to accurately separate non-metallic inclusions in the alloy composition. According to the model results, inclusions larger than ≥30 µm mainly represent the fraction of oxide phases, while fine inclusions smaller than <20 µm represent the fraction of sulfide phases.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Nozimjon Kholmirzaev

This work is licensed under a Creative Commons Attribution 4.0 International License.