Efficient reconstruction of prior austenite grains in steel from etched light optical micrographs using deep learning and annotations from correlative microscopy
The high-temperature austenite phase is the initial state of practically all technologically relevant hot forming and heat treatment operations in steel processing. The phenomena occurring in austenite, such as recrystallization or grain growth, can have a decisive influence on the subsequent proper...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-10-01
|
| Series: | Frontiers in Materials |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmats.2022.1033505/full |
| _version_ | 1817984469712240640 |
|---|---|
| author | Björn-Ivo Bachmann Björn-Ivo Bachmann Martin Müller Martin Müller Dominik Britz Dominik Britz Ali Riza Durmaz Marc Ackermann Oleg Shchyglo Thorsten Staudt Frank Mücklich Frank Mücklich |
| author_facet | Björn-Ivo Bachmann Björn-Ivo Bachmann Martin Müller Martin Müller Dominik Britz Dominik Britz Ali Riza Durmaz Marc Ackermann Oleg Shchyglo Thorsten Staudt Frank Mücklich Frank Mücklich |
| author_sort | Björn-Ivo Bachmann |
| collection | DOAJ |
| description | The high-temperature austenite phase is the initial state of practically all technologically relevant hot forming and heat treatment operations in steel processing. The phenomena occurring in austenite, such as recrystallization or grain growth, can have a decisive influence on the subsequent properties of the material. After the hot forming or heat treatment process, however, the austenite transforms into other microstructural constituents and information on the prior austenite morphology are no longer directly accessible. There are established methods available for reconstructing former austenite grain boundaries via metallographic etching or electron backscatter diffraction (EBSD) which both exhibit shortcomings. While etching is often difficult to reproduce and strongly depend on the investigated steel’s alloying concept, EBSD acquisition and reconstruction is rather time-consuming. But in fact, though, light optical micrographs of steels contrasted with conventional Nital etchant also contain information about the former austenite grains. However, relevant features are not directly apparent or accessible with conventional segmentation approaches. This work presents a deep learning (DL) segmentation of prior austenite grains (PAG) from Nital etched light optical micrographs. The basis for successful segmentation is a correlative characterization from EBSD, light and scanning electron microscopy to specify the ground truth required for supervised learning. The DL model shows good and robust segmentation results. While the intersection over union of 70% does not fully reflect the model performance due to the inherent uncertainty in PAG estimation, a mean error of 6.1% in mean grain size derived from the segmentation clearly shows the high quality of the result. |
| first_indexed | 2024-04-13T23:45:54Z |
| format | Article |
| id | doaj.art-0d528dbce9a84a30a9c196d7c90d36b5 |
| institution | Directory Open Access Journal |
| issn | 2296-8016 |
| language | English |
| last_indexed | 2024-04-13T23:45:54Z |
| publishDate | 2022-10-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Materials |
| spelling | doaj.art-0d528dbce9a84a30a9c196d7c90d36b52022-12-22T02:24:20ZengFrontiers Media S.A.Frontiers in Materials2296-80162022-10-01910.3389/fmats.2022.10335051033505Efficient reconstruction of prior austenite grains in steel from etched light optical micrographs using deep learning and annotations from correlative microscopyBjörn-Ivo Bachmann0Björn-Ivo Bachmann1Martin Müller2Martin Müller3Dominik Britz4Dominik Britz5Ali Riza Durmaz6Marc Ackermann7Oleg Shchyglo8Thorsten Staudt9Frank Mücklich10Frank Mücklich11Department of Materials Science, Saarland University, Saarbruecken, GermanyMaterial Engineering Center Saarland, Saarbruecken, GermanyDepartment of Materials Science, Saarland University, Saarbruecken, GermanyMaterial Engineering Center Saarland, Saarbruecken, GermanyDepartment of Materials Science, Saarland University, Saarbruecken, GermanyMaterial Engineering Center Saarland, Saarbruecken, GermanyFraunhofer Institute for Mechanics of Materials IWM, Freiburg, GermanySteel Institute, RWTH Aachen University, Aachen, GermanyInterdisciplinary Centre for Advanced Materials Simulation (ICAMS), Ruhr-Universität Bochum, Bochum, GermanyAktien-Gesellschaft der Dillinger Hüttenwerke, Dillingen, GermanyDepartment of Materials Science, Saarland University, Saarbruecken, GermanyMaterial Engineering Center Saarland, Saarbruecken, GermanyThe high-temperature austenite phase is the initial state of practically all technologically relevant hot forming and heat treatment operations in steel processing. The phenomena occurring in austenite, such as recrystallization or grain growth, can have a decisive influence on the subsequent properties of the material. After the hot forming or heat treatment process, however, the austenite transforms into other microstructural constituents and information on the prior austenite morphology are no longer directly accessible. There are established methods available for reconstructing former austenite grain boundaries via metallographic etching or electron backscatter diffraction (EBSD) which both exhibit shortcomings. While etching is often difficult to reproduce and strongly depend on the investigated steel’s alloying concept, EBSD acquisition and reconstruction is rather time-consuming. But in fact, though, light optical micrographs of steels contrasted with conventional Nital etchant also contain information about the former austenite grains. However, relevant features are not directly apparent or accessible with conventional segmentation approaches. This work presents a deep learning (DL) segmentation of prior austenite grains (PAG) from Nital etched light optical micrographs. The basis for successful segmentation is a correlative characterization from EBSD, light and scanning electron microscopy to specify the ground truth required for supervised learning. The DL model shows good and robust segmentation results. While the intersection over union of 70% does not fully reflect the model performance due to the inherent uncertainty in PAG estimation, a mean error of 6.1% in mean grain size derived from the segmentation clearly shows the high quality of the result.https://www.frontiersin.org/articles/10.3389/fmats.2022.1033505/fullsteelprior austenite grainssegmentationmachine learning/deep learningquantification |
| spellingShingle | Björn-Ivo Bachmann Björn-Ivo Bachmann Martin Müller Martin Müller Dominik Britz Dominik Britz Ali Riza Durmaz Marc Ackermann Oleg Shchyglo Thorsten Staudt Frank Mücklich Frank Mücklich Efficient reconstruction of prior austenite grains in steel from etched light optical micrographs using deep learning and annotations from correlative microscopy Frontiers in Materials steel prior austenite grains segmentation machine learning/deep learning quantification |
| title | Efficient reconstruction of prior austenite grains in steel from etched light optical micrographs using deep learning and annotations from correlative microscopy |
| title_full | Efficient reconstruction of prior austenite grains in steel from etched light optical micrographs using deep learning and annotations from correlative microscopy |
| title_fullStr | Efficient reconstruction of prior austenite grains in steel from etched light optical micrographs using deep learning and annotations from correlative microscopy |
| title_full_unstemmed | Efficient reconstruction of prior austenite grains in steel from etched light optical micrographs using deep learning and annotations from correlative microscopy |
| title_short | Efficient reconstruction of prior austenite grains in steel from etched light optical micrographs using deep learning and annotations from correlative microscopy |
| title_sort | efficient reconstruction of prior austenite grains in steel from etched light optical micrographs using deep learning and annotations from correlative microscopy |
| topic | steel prior austenite grains segmentation machine learning/deep learning quantification |
| url | https://www.frontiersin.org/articles/10.3389/fmats.2022.1033505/full |
| work_keys_str_mv | AT bjornivobachmann efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT bjornivobachmann efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT martinmuller efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT martinmuller efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT dominikbritz efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT dominikbritz efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT alirizadurmaz efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT marcackermann efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT olegshchyglo efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT thorstenstaudt efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT frankmucklich efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy AT frankmucklich efficientreconstructionofprioraustenitegrainsinsteelfrometchedlightopticalmicrographsusingdeeplearningandannotationsfromcorrelativemicroscopy |