Austenitic parent grain reconstruction in martensitic steel using deep learning

Patxi Fernandez-Zelaia; Andrés Márquez Rossy; Quinn Campbell; Andrzej Nycz; Christopher Ledford; Michael M. Kirka

doi:10.1016/j.matchar.2022.111759

Austenitic parent grain reconstruction in martensitic steel using deep learning

Patxi Fernandez-Zelaia, Andrés Márquez Rossy, Quinn Campbell, Andrzej Nycz, Christopher Ledford, Michael M. Kirka

Research output: Contribution to journal › Article › peer-review

6 Scopus citations

Abstract

In this work we develop a deep convolutional architecture to estimate the prior austenite structure from observed martensite electron backscatter diffraction micrographs. A novel data augmentation strategy randomizes the global reference coordinate system which makes it possible to train our model from only four micrographs. The model is much faster than algorithmic approaches and generalizes well when applied to micrographs of a different material. Empirical evidence suggests the efficacy of the model depends on the scale of the microstructure and receptive field of the vision model. This work demonstrates that modern computer vision approaches are well suited for capturing complex spatial-orientation patterns present in orientation imaging micrographs.

Original language	English
Article number	111759
Journal	Materials Characterization
Volume	185
DOIs	https://doi.org/10.1016/j.matchar.2022.111759
State	Published - Mar 2022

Funding

Research was sponsored by the US Department of Energy, Office of Energy Efficiency and Renewable Energy (EERE), Advanced Manufacturing Office, the Office of Fossil Energy, Crosscutting Research Program , under contract DE-AC05-00OR22725 with UT-Battelle LLC and performed in partiality at the Oak Ridge National Laboratory's Manufacturing Demonstration Facility , an Office of Energy Efficiency and Renewable Energy user facility.

Funders	Funder number
U.S. Department of Energy
Advanced Manufacturing Office
Office of Fossil Energy	DE-AC05-00OR22725
Office of Energy Efficiency and Renewable Energy
UT-Battelle

Keywords

Deep learning
Machine learning
Martensite
Phase transformations
Steel

Access to Document

10.1016/j.matchar.2022.111759

Cite this

@article{b9d89e435b9e4df38a57ee5587bff63d,

title = "Austenitic parent grain reconstruction in martensitic steel using deep learning",

abstract = "In this work we develop a deep convolutional architecture to estimate the prior austenite structure from observed martensite electron backscatter diffraction micrographs. A novel data augmentation strategy randomizes the global reference coordinate system which makes it possible to train our model from only four micrographs. The model is much faster than algorithmic approaches and generalizes well when applied to micrographs of a different material. Empirical evidence suggests the efficacy of the model depends on the scale of the microstructure and receptive field of the vision model. This work demonstrates that modern computer vision approaches are well suited for capturing complex spatial-orientation patterns present in orientation imaging micrographs.",

keywords = "Deep learning, Machine learning, Martensite, Phase transformations, Steel",

author = "Patxi Fernandez-Zelaia and Rossy, {Andr{\'e}s M{\'a}rquez} and Quinn Campbell and Andrzej Nycz and Christopher Ledford and Kirka, {Michael M.}",

note = "Publisher Copyright: {\textcopyright} 2022 Elsevier Inc.",

year = "2022",

month = mar,

doi = "10.1016/j.matchar.2022.111759",

language = "English",

volume = "185",

journal = "Materials Characterization",

issn = "1044-5803",

publisher = "Elsevier",

}

TY - JOUR

T1 - Austenitic parent grain reconstruction in martensitic steel using deep learning

AU - Fernandez-Zelaia, Patxi

AU - Rossy, Andrés Márquez

AU - Campbell, Quinn

AU - Nycz, Andrzej

AU - Ledford, Christopher

AU - Kirka, Michael M.

PY - 2022/3

Y1 - 2022/3

N2 - In this work we develop a deep convolutional architecture to estimate the prior austenite structure from observed martensite electron backscatter diffraction micrographs. A novel data augmentation strategy randomizes the global reference coordinate system which makes it possible to train our model from only four micrographs. The model is much faster than algorithmic approaches and generalizes well when applied to micrographs of a different material. Empirical evidence suggests the efficacy of the model depends on the scale of the microstructure and receptive field of the vision model. This work demonstrates that modern computer vision approaches are well suited for capturing complex spatial-orientation patterns present in orientation imaging micrographs.

AB - In this work we develop a deep convolutional architecture to estimate the prior austenite structure from observed martensite electron backscatter diffraction micrographs. A novel data augmentation strategy randomizes the global reference coordinate system which makes it possible to train our model from only four micrographs. The model is much faster than algorithmic approaches and generalizes well when applied to micrographs of a different material. Empirical evidence suggests the efficacy of the model depends on the scale of the microstructure and receptive field of the vision model. This work demonstrates that modern computer vision approaches are well suited for capturing complex spatial-orientation patterns present in orientation imaging micrographs.

KW - Deep learning

KW - Machine learning

KW - Martensite

KW - Phase transformations

KW - Steel

UR - http://www.scopus.com/inward/record.url?scp=85123900623&partnerID=8YFLogxK

U2 - 10.1016/j.matchar.2022.111759

DO - 10.1016/j.matchar.2022.111759

M3 - Article

AN - SCOPUS:85123900623

SN - 1044-5803

VL - 185

JO - Materials Characterization

JF - Materials Characterization

M1 - 111759

ER -

Austenitic parent grain reconstruction in martensitic steel using deep learning

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this