Microstructure Representation and Reconstruction of Heterogeneous Materials Via Deep Belief Network for Computational Material Design

Ruijin Cang, Yaopengxiao Xu, Shaohua Chen, Yongming Liu, Yang Jiao, Max Yi Ren

Research output: Contribution to journalArticle

24 Citations (Scopus)

Abstract

Integrated Computational Materials Engineering (ICME) aims to accelerate optimal design of complex material systems by integrating material science and design automation. For tractable ICME, it is required that (1) a structural feature space be identified to allow reconstruction of new designs, and (2) the reconstruction process be propertypreserving. The majority of existing structural presentation schemes relies on the designer's understanding of specific material systems to identify geometric and statistical features, which could be biased and insufficient for reconstructing physically meaningful microstructures of complex material systems. In this paper, we develop a feature learning mechanism based on convolutional deep belief network (CDBN) to automate a two-way conversion between microstructures and their lower-dimensional feature representations, and to achieve a 1000-fold dimension reduction from the microstructure space. The proposed model is applied to a wide spectrum of heterogeneous material systems with distinct microstructural features including Ti-6Al-4V alloy, Pb63-Sn37 alloy, Fontainebleau sandstone, and spherical colloids, to produce material reconstructions that are close to the original samples with respect to two-point correlation functions and mean critical fracture strength. This capability is not achieved by existing synthesis methods that rely on the Markovian assumption of material microstructures.

Original languageEnglish (US)
Article number071404
JournalJournal of Mechanical Design, Transactions Of the ASME
Volume139
Issue number7
DOIs
StatePublished - Jul 1 2017

Fingerprint

Bayesian networks
Microstructure
Materials science
Sandstone
Colloids
Fracture toughness
Automation

ASJC Scopus subject areas

  • Mechanics of Materials
  • Mechanical Engineering
  • Computer Science Applications
  • Computer Graphics and Computer-Aided Design

Cite this

Microstructure Representation and Reconstruction of Heterogeneous Materials Via Deep Belief Network for Computational Material Design. / Cang, Ruijin; Xu, Yaopengxiao; Chen, Shaohua; Liu, Yongming; Jiao, Yang; Ren, Max Yi.

In: Journal of Mechanical Design, Transactions Of the ASME, Vol. 139, No. 7, 071404, 01.07.2017.

Research output: Contribution to journalArticle

@article{0e00a0e3f7844a0aaccf6cd247689c4a,
title = "Microstructure Representation and Reconstruction of Heterogeneous Materials Via Deep Belief Network for Computational Material Design",
abstract = "Integrated Computational Materials Engineering (ICME) aims to accelerate optimal design of complex material systems by integrating material science and design automation. For tractable ICME, it is required that (1) a structural feature space be identified to allow reconstruction of new designs, and (2) the reconstruction process be propertypreserving. The majority of existing structural presentation schemes relies on the designer's understanding of specific material systems to identify geometric and statistical features, which could be biased and insufficient for reconstructing physically meaningful microstructures of complex material systems. In this paper, we develop a feature learning mechanism based on convolutional deep belief network (CDBN) to automate a two-way conversion between microstructures and their lower-dimensional feature representations, and to achieve a 1000-fold dimension reduction from the microstructure space. The proposed model is applied to a wide spectrum of heterogeneous material systems with distinct microstructural features including Ti-6Al-4V alloy, Pb63-Sn37 alloy, Fontainebleau sandstone, and spherical colloids, to produce material reconstructions that are close to the original samples with respect to two-point correlation functions and mean critical fracture strength. This capability is not achieved by existing synthesis methods that rely on the Markovian assumption of material microstructures.",
author = "Ruijin Cang and Yaopengxiao Xu and Shaohua Chen and Yongming Liu and Yang Jiao and Ren, {Max Yi}",
year = "2017",
month = "7",
day = "1",
doi = "10.1115/1.4036649",
language = "English (US)",
volume = "139",
journal = "Journal of Mechanical Design - Transactions of the ASME",
issn = "1050-0472",
publisher = "American Society of Mechanical Engineers(ASME)",
number = "7",

}

TY - JOUR

T1 - Microstructure Representation and Reconstruction of Heterogeneous Materials Via Deep Belief Network for Computational Material Design

AU - Cang, Ruijin

AU - Xu, Yaopengxiao

AU - Chen, Shaohua

AU - Liu, Yongming

AU - Jiao, Yang

AU - Ren, Max Yi

PY - 2017/7/1

Y1 - 2017/7/1

N2 - Integrated Computational Materials Engineering (ICME) aims to accelerate optimal design of complex material systems by integrating material science and design automation. For tractable ICME, it is required that (1) a structural feature space be identified to allow reconstruction of new designs, and (2) the reconstruction process be propertypreserving. The majority of existing structural presentation schemes relies on the designer's understanding of specific material systems to identify geometric and statistical features, which could be biased and insufficient for reconstructing physically meaningful microstructures of complex material systems. In this paper, we develop a feature learning mechanism based on convolutional deep belief network (CDBN) to automate a two-way conversion between microstructures and their lower-dimensional feature representations, and to achieve a 1000-fold dimension reduction from the microstructure space. The proposed model is applied to a wide spectrum of heterogeneous material systems with distinct microstructural features including Ti-6Al-4V alloy, Pb63-Sn37 alloy, Fontainebleau sandstone, and spherical colloids, to produce material reconstructions that are close to the original samples with respect to two-point correlation functions and mean critical fracture strength. This capability is not achieved by existing synthesis methods that rely on the Markovian assumption of material microstructures.

AB - Integrated Computational Materials Engineering (ICME) aims to accelerate optimal design of complex material systems by integrating material science and design automation. For tractable ICME, it is required that (1) a structural feature space be identified to allow reconstruction of new designs, and (2) the reconstruction process be propertypreserving. The majority of existing structural presentation schemes relies on the designer's understanding of specific material systems to identify geometric and statistical features, which could be biased and insufficient for reconstructing physically meaningful microstructures of complex material systems. In this paper, we develop a feature learning mechanism based on convolutional deep belief network (CDBN) to automate a two-way conversion between microstructures and their lower-dimensional feature representations, and to achieve a 1000-fold dimension reduction from the microstructure space. The proposed model is applied to a wide spectrum of heterogeneous material systems with distinct microstructural features including Ti-6Al-4V alloy, Pb63-Sn37 alloy, Fontainebleau sandstone, and spherical colloids, to produce material reconstructions that are close to the original samples with respect to two-point correlation functions and mean critical fracture strength. This capability is not achieved by existing synthesis methods that rely on the Markovian assumption of material microstructures.

UR - http://www.scopus.com/inward/record.url?scp=85019942478&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85019942478&partnerID=8YFLogxK

U2 - 10.1115/1.4036649

DO - 10.1115/1.4036649

M3 - Article

AN - SCOPUS:85019942478

VL - 139

JO - Journal of Mechanical Design - Transactions of the ASME

JF - Journal of Mechanical Design - Transactions of the ASME

SN - 1050-0472

IS - 7

M1 - 071404

ER -