2021 Nashville AISTech

Sessions Speakers Exhibitors Products Press Releases Show Specials Floor Plan Buzz

Analysis of Inclusion Clusters Using Machine-Learning Tools (Room 205 B)

Behavior of non-metallic inclusions is critical for steel processing. One of the main concerns is inclusion agglomeration, which can lead to large clusters in the solid product. Scanning electron microscopy (SEM) coupled with energy-dispersive spectroscopy (EDS) has been a prominent technique for inclusion characterization. This study utilizes the output generated from SEM/EDS analysis along with machine-learning tools to provide an automated method for inclusion cluster identification. The approach was to analyze multiple cross-sections from a steel sample, thereby examining inclusions from a detailed three-dimensional perspective. The aim was to relate the three-dimensional analysis to the two-dimensional cross-sections.