Efficient Deep Learning for Gradient-Enhanced Stress Dependent Damage Model

This manuscript introduces a computational approach to micro-damage problems using deep learning for the prediction of loading deflection curves. The location of applied forces, dimensions of the specimen and material parameters are used as inputs of the process. The micro-damage is modelled with a...

Full description

Bibliographic Details
Main Authors: Xiaoying Zhuang, L. C. Nguyen, Hung Nguyen-Xuan, Naif Alajlan, Timon Rabczuk
Format: Article
Language:English
Published: MDPI AG 2020-04-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/10/7/2556
id doaj-66a042c978ab41a49203464edaa07697
record_format Article
spelling doaj-66a042c978ab41a49203464edaa076972020-11-25T02:37:27ZengMDPI AGApplied Sciences2076-34172020-04-01102556255610.3390/app10072556Efficient Deep Learning for Gradient-Enhanced Stress Dependent Damage ModelXiaoying Zhuang0L. C. Nguyen1Hung Nguyen-Xuan2Naif Alajlan3Timon Rabczuk4Division of Computational Mechanics, Ton Duc Thang University, Ho Chi Minh City, Viet NamInstitute for Continuum Mechanics, Leibniz Universität Hannover, Appelstr. 11, 30167 Hannover, GermanyCIRTech Institute, Ho Chi Minh City University of Technology (HUTECH), Ho Chi Minh City, VietnamDepartment of Computer Engineering College of Computer and Information Sciences King Saud University, Riyadh 11543, Saudi ArabiaDepartment of Computer Engineering College of Computer and Information Sciences King Saud University, Riyadh 11543, Saudi ArabiaThis manuscript introduces a computational approach to micro-damage problems using deep learning for the prediction of loading deflection curves. The location of applied forces, dimensions of the specimen and material parameters are used as inputs of the process. The micro-damage is modelled with a gradient-enhanced damage model which ensures the well-posedness of the boundary value and yields mesh-independent results in computational methods such as FEM. We employ the Adam optimizer and Rectified linear unit activation function for training processes and research into the deep neural network architecture. The performance of our approach is demonstrated through some numerical examples including the three-point bending specimen, shear bending on L-shaped specimen and different failure mechanisms.https://www.mdpi.com/2076-3417/10/7/2556deep neural networkdeep learninggradient enhanced damagestress-level dependent damage model
collection DOAJ
language English
format Article
sources DOAJ
author Xiaoying Zhuang
L. C. Nguyen
Hung Nguyen-Xuan
Naif Alajlan
Timon Rabczuk
spellingShingle Xiaoying Zhuang
L. C. Nguyen
Hung Nguyen-Xuan
Naif Alajlan
Timon Rabczuk
Efficient Deep Learning for Gradient-Enhanced Stress Dependent Damage Model
Applied Sciences
deep neural network
deep learning
gradient enhanced damage
stress-level dependent damage model
author_facet Xiaoying Zhuang
L. C. Nguyen
Hung Nguyen-Xuan
Naif Alajlan
Timon Rabczuk
author_sort Xiaoying Zhuang
title Efficient Deep Learning for Gradient-Enhanced Stress Dependent Damage Model
title_short Efficient Deep Learning for Gradient-Enhanced Stress Dependent Damage Model
title_full Efficient Deep Learning for Gradient-Enhanced Stress Dependent Damage Model
title_fullStr Efficient Deep Learning for Gradient-Enhanced Stress Dependent Damage Model
title_full_unstemmed Efficient Deep Learning for Gradient-Enhanced Stress Dependent Damage Model
title_sort efficient deep learning for gradient-enhanced stress dependent damage model
publisher MDPI AG
series Applied Sciences
issn 2076-3417
publishDate 2020-04-01
description This manuscript introduces a computational approach to micro-damage problems using deep learning for the prediction of loading deflection curves. The location of applied forces, dimensions of the specimen and material parameters are used as inputs of the process. The micro-damage is modelled with a gradient-enhanced damage model which ensures the well-posedness of the boundary value and yields mesh-independent results in computational methods such as FEM. We employ the Adam optimizer and Rectified linear unit activation function for training processes and research into the deep neural network architecture. The performance of our approach is demonstrated through some numerical examples including the three-point bending specimen, shear bending on L-shaped specimen and different failure mechanisms.
topic deep neural network
deep learning
gradient enhanced damage
stress-level dependent damage model
url https://www.mdpi.com/2076-3417/10/7/2556
work_keys_str_mv AT xiaoyingzhuang efficientdeeplearningforgradientenhancedstressdependentdamagemodel
AT lcnguyen efficientdeeplearningforgradientenhancedstressdependentdamagemodel
AT hungnguyenxuan efficientdeeplearningforgradientenhancedstressdependentdamagemodel
AT naifalajlan efficientdeeplearningforgradientenhancedstressdependentdamagemodel
AT timonrabczuk efficientdeeplearningforgradientenhancedstressdependentdamagemodel
_version_ 1724795589101617152