Artificial intelligence design algorithm for nanocomposites optimized for shear crack resistance

Chi Hua Yu, Zhao Qin, Markus J. Buehler

Research output: Contribution to journalArticle

Abstract

Here we report a design approach for optimizing the toughness of nanocomposite materials using artificial intelligence (AI), implemented in a novel ‘AutoComp Designer’ algorithm. The algorithm consists of a machine learning predictor combined with an AI improved genetic algorithm, which is capable of discovering de novo materials designs in a vast space of possible solutions. Facilitated by a deep convolutional neural network that is trained with a dataset of hundreds of thousands of combinations of soft and brittle materials originating from a finite element analysis, we predict the. Through the algorithm, we extend the capability of physical simulations beyond property predictions to optimize the fracture toughness by altering the material distribution. The solutions are generated by our AI model at a dramatically lower computational cost compared to brute-force searching methods. Wefurther investigate the physical mechanism for improving material performance behind the AI approach, and demonstrate the ability of AI to search for optimal designs with very limited sampling. Brute-force molecular dynamics simulations of the nanocomposite designs confirm that our AI design improves the performance by effectively decreasing the stress concentration at the crack tip. This AI approach can be easily applied to other nanocomposites, biomaterials, and other material classes, and provides a transferable and reliable rapid design approach expanding current capabilities.

Original languageEnglish (US)
Article number035001
JournalNano Futures
Volume3
Issue number3
DOIs
StatePublished - Sep 2019

Fingerprint

artificial intelligence
Artificial intelligence
Nanocomposites
nanocomposites
cracks
shear
Cracks
brittle materials
machine learning
stress concentration
crack tips
Biocompatible Materials
toughness
Brittleness
predictions
fracture strength
Biomaterials
Crack tips
genetic algorithms
Toughness

Keywords

  • Artificial intelligence
  • Genetic algorithm
  • Nanocomposite
  • Neural network
  • Optimization
  • Strength
  • Toughness

ASJC Scopus subject areas

  • Bioengineering
  • Chemistry(all)
  • Atomic and Molecular Physics, and Optics
  • Biomedical Engineering
  • Materials Science(all)
  • Electrical and Electronic Engineering

Cite this

Artificial intelligence design algorithm for nanocomposites optimized for shear crack resistance. / Yu, Chi Hua; Qin, Zhao; Buehler, Markus J.

In: Nano Futures, Vol. 3, No. 3, 035001, 09.2019.

Research output: Contribution to journalArticle

@article{3a3dafc4f77f4481b773e5751d0126ae,
title = "Artificial intelligence design algorithm for nanocomposites optimized for shear crack resistance",
abstract = "Here we report a design approach for optimizing the toughness of nanocomposite materials using artificial intelligence (AI), implemented in a novel ‘AutoComp Designer’ algorithm. The algorithm consists of a machine learning predictor combined with an AI improved genetic algorithm, which is capable of discovering de novo materials designs in a vast space of possible solutions. Facilitated by a deep convolutional neural network that is trained with a dataset of hundreds of thousands of combinations of soft and brittle materials originating from a finite element analysis, we predict the. Through the algorithm, we extend the capability of physical simulations beyond property predictions to optimize the fracture toughness by altering the material distribution. The solutions are generated by our AI model at a dramatically lower computational cost compared to brute-force searching methods. Wefurther investigate the physical mechanism for improving material performance behind the AI approach, and demonstrate the ability of AI to search for optimal designs with very limited sampling. Brute-force molecular dynamics simulations of the nanocomposite designs confirm that our AI design improves the performance by effectively decreasing the stress concentration at the crack tip. This AI approach can be easily applied to other nanocomposites, biomaterials, and other material classes, and provides a transferable and reliable rapid design approach expanding current capabilities.",
keywords = "Artificial intelligence, Genetic algorithm, Nanocomposite, Neural network, Optimization, Strength, Toughness",
author = "Yu, {Chi Hua} and Zhao Qin and Buehler, {Markus J.}",
year = "2019",
month = "9",
doi = "10.1088/2399-1984/ab36f0",
language = "English (US)",
volume = "3",
journal = "Nano Futures",
issn = "2399-1984",
publisher = "Institute of Physics Publishing",
number = "3",

}

TY - JOUR

T1 - Artificial intelligence design algorithm for nanocomposites optimized for shear crack resistance

AU - Yu, Chi Hua

AU - Qin, Zhao

AU - Buehler, Markus J.

PY - 2019/9

Y1 - 2019/9

N2 - Here we report a design approach for optimizing the toughness of nanocomposite materials using artificial intelligence (AI), implemented in a novel ‘AutoComp Designer’ algorithm. The algorithm consists of a machine learning predictor combined with an AI improved genetic algorithm, which is capable of discovering de novo materials designs in a vast space of possible solutions. Facilitated by a deep convolutional neural network that is trained with a dataset of hundreds of thousands of combinations of soft and brittle materials originating from a finite element analysis, we predict the. Through the algorithm, we extend the capability of physical simulations beyond property predictions to optimize the fracture toughness by altering the material distribution. The solutions are generated by our AI model at a dramatically lower computational cost compared to brute-force searching methods. Wefurther investigate the physical mechanism for improving material performance behind the AI approach, and demonstrate the ability of AI to search for optimal designs with very limited sampling. Brute-force molecular dynamics simulations of the nanocomposite designs confirm that our AI design improves the performance by effectively decreasing the stress concentration at the crack tip. This AI approach can be easily applied to other nanocomposites, biomaterials, and other material classes, and provides a transferable and reliable rapid design approach expanding current capabilities.

AB - Here we report a design approach for optimizing the toughness of nanocomposite materials using artificial intelligence (AI), implemented in a novel ‘AutoComp Designer’ algorithm. The algorithm consists of a machine learning predictor combined with an AI improved genetic algorithm, which is capable of discovering de novo materials designs in a vast space of possible solutions. Facilitated by a deep convolutional neural network that is trained with a dataset of hundreds of thousands of combinations of soft and brittle materials originating from a finite element analysis, we predict the. Through the algorithm, we extend the capability of physical simulations beyond property predictions to optimize the fracture toughness by altering the material distribution. The solutions are generated by our AI model at a dramatically lower computational cost compared to brute-force searching methods. Wefurther investigate the physical mechanism for improving material performance behind the AI approach, and demonstrate the ability of AI to search for optimal designs with very limited sampling. Brute-force molecular dynamics simulations of the nanocomposite designs confirm that our AI design improves the performance by effectively decreasing the stress concentration at the crack tip. This AI approach can be easily applied to other nanocomposites, biomaterials, and other material classes, and provides a transferable and reliable rapid design approach expanding current capabilities.

KW - Artificial intelligence

KW - Genetic algorithm

KW - Nanocomposite

KW - Neural network

KW - Optimization

KW - Strength

KW - Toughness

UR - http://www.scopus.com/inward/record.url?scp=85073448389&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85073448389&partnerID=8YFLogxK

U2 - 10.1088/2399-1984/ab36f0

DO - 10.1088/2399-1984/ab36f0

M3 - Article

AN - SCOPUS:85073448389

VL - 3

JO - Nano Futures

JF - Nano Futures

SN - 2399-1984

IS - 3

M1 - 035001

ER -