Back-propagation neural network-based approximate analysis of true stress-strain behaviors of high-strength metallic material

Jaehyeok Doh, Seung Uk Lee, Jongsoo Lee

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

In this study, a Back-propagation neural network (BPN) is employed to conduct an approximation of a true stress-strain curve using the load-displacement experimental data of DP590, a high-strength material used in automobile bodies and chassis. The optimized interconnection weights are obtained with hidden layers and output layers of the BPN through intelligent learning and training of the experimental data; by using these weights, a mathematical model of the material’s behavior is suggested through this feed-forward neural network. Generally, the material properties from the tensile test cannot be acquired until the fracture regions, since it is difficult to measure the cross-section area of a specimen after diffusion necking. For this reason, the plastic properties of the true stress-strain are extrapolated using the weighted-average method after diffusion necking. The accuracies of BPN-based meta-models for predicting material properties are validated in terms of the Root mean square error (RMSE). By applying the approximate material properties, the reliable finite element solution can be obtained to realize the different shapes of the finite element models. Furthermore, the sensitivity analysis of the approximate meta-model is performed using the first-order approximate derivatives of the BPN and is compared with the results of the finite difference method. In addition, we predict the tension velocity’s effect on the material property through a first-order sensitivity analysis.

Original languageEnglish
Pages (from-to)1233-1241
Number of pages9
JournalJournal of Mechanical Science and Technology
Volume30
Issue number3
DOIs
Publication statusPublished - 2016 Mar 1

Fingerprint

Backpropagation
Materials properties
Neural networks
Sensitivity analysis
Automobile bodies
Feedforward neural networks
Chassis
Stress-strain curves
Finite difference method
Mean square error
Mathematical models
Plastics
Derivatives

All Science Journal Classification (ASJC) codes

  • Mechanics of Materials
  • Mechanical Engineering

Cite this

@article{89a3eb3715e2486aa89b788e8c06954b,
title = "Back-propagation neural network-based approximate analysis of true stress-strain behaviors of high-strength metallic material",
abstract = "In this study, a Back-propagation neural network (BPN) is employed to conduct an approximation of a true stress-strain curve using the load-displacement experimental data of DP590, a high-strength material used in automobile bodies and chassis. The optimized interconnection weights are obtained with hidden layers and output layers of the BPN through intelligent learning and training of the experimental data; by using these weights, a mathematical model of the material’s behavior is suggested through this feed-forward neural network. Generally, the material properties from the tensile test cannot be acquired until the fracture regions, since it is difficult to measure the cross-section area of a specimen after diffusion necking. For this reason, the plastic properties of the true stress-strain are extrapolated using the weighted-average method after diffusion necking. The accuracies of BPN-based meta-models for predicting material properties are validated in terms of the Root mean square error (RMSE). By applying the approximate material properties, the reliable finite element solution can be obtained to realize the different shapes of the finite element models. Furthermore, the sensitivity analysis of the approximate meta-model is performed using the first-order approximate derivatives of the BPN and is compared with the results of the finite difference method. In addition, we predict the tension velocity’s effect on the material property through a first-order sensitivity analysis.",
author = "Jaehyeok Doh and Lee, {Seung Uk} and Jongsoo Lee",
year = "2016",
month = "3",
day = "1",
doi = "10.1007/s12206-016-0227-1",
language = "English",
volume = "30",
pages = "1233--1241",
journal = "Journal of Mechanical Science and Technology",
issn = "1738-494X",
publisher = "Korean Society of Mechanical Engineers",
number = "3",

}

Back-propagation neural network-based approximate analysis of true stress-strain behaviors of high-strength metallic material. / Doh, Jaehyeok; Lee, Seung Uk; Lee, Jongsoo.

In: Journal of Mechanical Science and Technology, Vol. 30, No. 3, 01.03.2016, p. 1233-1241.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Back-propagation neural network-based approximate analysis of true stress-strain behaviors of high-strength metallic material

AU - Doh, Jaehyeok

AU - Lee, Seung Uk

AU - Lee, Jongsoo

PY - 2016/3/1

Y1 - 2016/3/1

N2 - In this study, a Back-propagation neural network (BPN) is employed to conduct an approximation of a true stress-strain curve using the load-displacement experimental data of DP590, a high-strength material used in automobile bodies and chassis. The optimized interconnection weights are obtained with hidden layers and output layers of the BPN through intelligent learning and training of the experimental data; by using these weights, a mathematical model of the material’s behavior is suggested through this feed-forward neural network. Generally, the material properties from the tensile test cannot be acquired until the fracture regions, since it is difficult to measure the cross-section area of a specimen after diffusion necking. For this reason, the plastic properties of the true stress-strain are extrapolated using the weighted-average method after diffusion necking. The accuracies of BPN-based meta-models for predicting material properties are validated in terms of the Root mean square error (RMSE). By applying the approximate material properties, the reliable finite element solution can be obtained to realize the different shapes of the finite element models. Furthermore, the sensitivity analysis of the approximate meta-model is performed using the first-order approximate derivatives of the BPN and is compared with the results of the finite difference method. In addition, we predict the tension velocity’s effect on the material property through a first-order sensitivity analysis.

AB - In this study, a Back-propagation neural network (BPN) is employed to conduct an approximation of a true stress-strain curve using the load-displacement experimental data of DP590, a high-strength material used in automobile bodies and chassis. The optimized interconnection weights are obtained with hidden layers and output layers of the BPN through intelligent learning and training of the experimental data; by using these weights, a mathematical model of the material’s behavior is suggested through this feed-forward neural network. Generally, the material properties from the tensile test cannot be acquired until the fracture regions, since it is difficult to measure the cross-section area of a specimen after diffusion necking. For this reason, the plastic properties of the true stress-strain are extrapolated using the weighted-average method after diffusion necking. The accuracies of BPN-based meta-models for predicting material properties are validated in terms of the Root mean square error (RMSE). By applying the approximate material properties, the reliable finite element solution can be obtained to realize the different shapes of the finite element models. Furthermore, the sensitivity analysis of the approximate meta-model is performed using the first-order approximate derivatives of the BPN and is compared with the results of the finite difference method. In addition, we predict the tension velocity’s effect on the material property through a first-order sensitivity analysis.

UR - http://www.scopus.com/inward/record.url?scp=84961275882&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84961275882&partnerID=8YFLogxK

U2 - 10.1007/s12206-016-0227-1

DO - 10.1007/s12206-016-0227-1

M3 - Article

VL - 30

SP - 1233

EP - 1241

JO - Journal of Mechanical Science and Technology

JF - Journal of Mechanical Science and Technology

SN - 1738-494X

IS - 3

ER -