Berlin 2024 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
MM: Fachverband Metall- und Materialphysik
MM 20: Data Driven Material Science: Big Data and Workflows III
MM 20.9: Vortrag
Dienstag, 19. März 2024, 12:30–12:45, C 243
Influence of the neighboring phases of MnS inclusions on damage accumulation in case-hardening steel — •Clara Reinhart1, Tom Reclik1, Maximilian A. Wollenweber1, Ulrich Kerzel2, Talal Al-Samman1, and Sandra Korte-Kerzel1 — 1Institute for Physical Metallurgy and Materials Physics, RWTH Aachen University, Aachen, Germany — 2Data Science and Artificial Intelligence in Materials and Geoscience, Faculty of Georesources and Materials Engineering, RWTH Aachen University, Aachen, Germany
Microstructural damage sites that are created during forming processes are usually observable in the form of voids and known to impede the mechanical properties of materials, especially during cyclic and rapid loading. In the case of 16MnCrS5 case-hardening steel, MnS inclusions lead to the creation of damage sites by cracking or delamination due to a pronounced mechanical contrast in the microstructure. This mechanical contrast depends on whether the inclusion is surrounded by ferrite, pearlite or both phases simultaneously. In this work we set out to characterize damage sites based on the neighboring phase of the MnS inclusion by training a machine learning network to automatically segment the etched microstructure and characterize the interfaces. In a second step damages sites are automatically detected and correlated to the determined neighboring phase(s). With this approach we show that a large difference of damage accumulation emerges for different neighboring phases, distinguishing not only inclusions with one-phase and two-phase interfaces, but also inclusions surrounded exclusively by either ferrite or pearlite.
Keywords: Steel; Phase Segmentation; Microstructural damage; Machine learning; Scanning electron microscopy