Deep learning in metallography: adapting a grain size determination model to develop the new comparison charts for ISO 643
-
Z. Wu
Pursuing Doctor of Engineering in Central Iron and Steel Research Institute. Currently working at Central Iron and Steel Research Institute Co., Ltd and conducting research in microscopic analysis experiments.
M. E. Materials Engineering of University of Science and Technology Beijing. Currently working as Laboratory director of Central Iron and Steel Research Institute Co., Ltd, mainly engaged in the micro-analysis of steels – Vice Chairman of SAC/TC183/SC14 on Metallographic inspection – Expert of ISO/TC17/SC7/WG2 Steels – Micrographic determination of the apparent grain size.
,
J. Zhang
Abstract
The determination of average grain size is an important component in the microstructural characterization of metallic materials. The grain size is usually determined using the intercept and comparison methods, but there are problems with the relatively time-consuming intercept method and the low accuracy of the comparison method. In this paper, a new model is proposed to realize the task of automatic grain size grading using a combination of traditional image processing and deep learning. Deep learning semantic segmentation and image classification are used to achieve the grading of different quality images. The model is capable of extracting and complementing grain boundaries and calculating the pixel area of the grains. The accuracy of the model was tested on metallographic phases of pure iron and austenitic stainless steel. Using this model, grain size comparison charts were produced for ISO 643 based on real metallographic photographs.
Kurzfassung
Die Bestimmung der durchschnittlichen Korngröße ist ein wichtiges Element der Gefügecharakterisierung von metallischen Werkstoffen. In der Regel wird die Korngröße mithilfe des Linienschnitt- und Vergleichsverfahrens bestimmt. Dabei erweisen sich der relativ große Zeitaufwand beim Linienschnittverfahren und die geringe Genauigkeit des Vergleichsverfahrens jedoch als problematisch. In dieser Arbeit wird ein neues Modell vorgeschlagen, mit dem Korngrößen automatisch mithilfe einer Kombination aus herkömmlicher Bildverarbeitung und Deep Learning bestimmt werden können. Zur Bewertung von Aufnahmen unterschiedlicher Qualität wird mit auf Deep Learning basierender semantischer Segmentierung und Bildklassifizierung gearbeitet. Mit dem Modell lassen sich Korngrenzen extrahieren und vervollständigen sowie die Pixelfläche der Körner berechnen. Die Genauigkeit des Modells wurde an metallischen Phasen in reinem Eisen und austenitischem Edelstahl geprüft. Mithilfe des Modells wurden basierend auf realen metallographischen Aufnahmen ISO 643-Korngrößenrichtreihen erstellt.
About the authors
Pursuing Doctor of Engineering in Central Iron and Steel Research Institute. Currently working at Central Iron and Steel Research Institute Co., Ltd and conducting research in microscopic analysis experiments.
M. E. Materials Engineering of University of Science and Technology Beijing. Currently working as Laboratory director of Central Iron and Steel Research Institute Co., Ltd, mainly engaged in the micro-analysis of steels – Vice Chairman of SAC/TC183/SC14 on Metallographic inspection – Expert of ISO/TC17/SC7/WG2 Steels – Micrographic determination of the apparent grain size.
References / Literatur
[1] ISO 683-3:2022: Heat-treatable steels, alloy steels and free-cutting steels – Part 3: Casehardening steels.Suche in Google Scholar
[2] ISO 643:2024: Steels – Micrographic determination of the apparent grain size.Suche in Google Scholar
[3] ASTM E112-13: (Reapproved 2021) Standard Test Methods for Determining Average Grain Size.Suche in Google Scholar
[4] GB/T 6394-2017: Determination of estimating the average grain size of metal.Suche in Google Scholar
[5] Stücklin, S.; Pühringer, J.: Prakt. Met. Sonderband 53 (2019), pp. 107–112.Suche in Google Scholar
[6] GB/T 13298-2015: Inspection methods of micro-structure for metals.Suche in Google Scholar
© 2025 Walter de Gruyter GmbH, Berlin/Boston
Artikel in diesem Heft
- Inhalt
- Editorial
- Editorial
- Effect of tramp elements on the microstructural evolution of a ferritic-pearlitic steel
- Deep learning in metallography: adapting a grain size determination model to develop the new comparison charts for ISO 643
- Design and development of welding electrode for structural steel
- Failure Analysis
- Metallurgical failure analysis of small-bore tube compression fitting that fractured in a gas turbine burner test rig
- Picture of the Month
- Picture of the Month
- News
- News
- Meeting Diary
- Meeting Diary
Artikel in diesem Heft
- Inhalt
- Editorial
- Editorial
- Effect of tramp elements on the microstructural evolution of a ferritic-pearlitic steel
- Deep learning in metallography: adapting a grain size determination model to develop the new comparison charts for ISO 643
- Design and development of welding electrode for structural steel
- Failure Analysis
- Metallurgical failure analysis of small-bore tube compression fitting that fractured in a gas turbine burner test rig
- Picture of the Month
- Picture of the Month
- News
- News
- Meeting Diary
- Meeting Diary
