Journal article 217 views 74 downloads
A framework for neural network based constitutive modelling of inelastic materials
Computer Methods in Applied Mechanics and Engineering, Volume: 420, Start page: 116672
Swansea University Authors:
Wulf Dettmer , Reem Alhayki, Eugenio Muttio Zavala, Djordje Peric
-
PDF | Version of Record
© 2023 The Author(s). Published by Elsevier B.V. Distributed under the terms of a Creative Commons Attribution 4.0 International License (CC BY 4.0).
Download (5.55MB)
DOI (Published version): 10.1016/j.cma.2023.116672
Abstract
Given the significant recent advances in added layer manufacturing and materials engineering, new types of materials or new material micro-structures are becoming available at a fast rate. The finite element analysis of structures or structural components requires a constitutive model that describes...
Published in: | Computer Methods in Applied Mechanics and Engineering |
---|---|
ISSN: | 0045-7825 1879-2138 |
Published: |
Elsevier BV
2024
|
Online Access: |
Check full text
|
URI: | https://cronfa.swan.ac.uk/Record/cronfa65340 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
first_indexed |
2023-12-18T13:48:05Z |
---|---|
last_indexed |
2023-12-18T13:48:05Z |
id |
cronfa65340 |
recordtype |
SURis |
fullrecord |
<?xml version="1.0" encoding="utf-8"?><rfc1807 xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"><bib-version>v2</bib-version><id>65340</id><entry>2023-12-18</entry><title>A framework for neural network based constitutive modelling of inelastic materials</title><swanseaauthors><author><sid>30bb53ad906e7160e947fa01c16abf55</sid><ORCID>0000-0003-0799-4645</ORCID><firstname>Wulf</firstname><surname>Dettmer</surname><name>Wulf Dettmer</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>f7aa889da52c57c88e62d530e687ea13</sid><firstname>Reem</firstname><surname>Alhayki</surname><name>Reem Alhayki</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>ee7320f4fba56d3fc7eea1bcdd28e615</sid><firstname>Eugenio</firstname><surname>Muttio Zavala</surname><name>Eugenio Muttio Zavala</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>9d35cb799b2542ad39140943a9a9da65</sid><ORCID>0000-0002-1112-301X</ORCID><firstname>Djordje</firstname><surname>Peric</surname><name>Djordje Peric</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2023-12-18</date><deptcode>AERO</deptcode><abstract>Given the significant recent advances in added layer manufacturing and materials engineering, new types of materials or new material micro-structures are becoming available at a fast rate. The finite element analysis of structures or structural components requires a constitutive model that describes the behaviour of the new materials. The formulation of accurate constitutive equations is generally complex and time consuming. Hence, suitable machine learning strategies may be used to render this process obsolete and bridge the gap between experimental data and finite element analysis. In this work, a generic stress update procedure is presented that is suitable for the modelling of rate-independent, elastic or inelastic, isotropic or anisotropic material behaviour. The proposed strategy is based on a recurrent neural network architecture and must be trained on stress and strain data sequences that represent physical or numerical experiments. A training strategy based on gradient-free optimisation is presented. It is shown that piecewise linear behaviour, such as uniaxial elasto-plasticity, can be represented exactly. Further numerical examples include uniaxial damage mechanics and elasto-plasticity under plane strain conditions. An efficient criterion for the verification of thermodynamic consistency is proposed and applied to the trained stress update models. The strategy is compared to GRU or LSTM based architectures and shown to offer advantages.</abstract><type>Journal Article</type><journal>Computer Methods in Applied Mechanics and Engineering</journal><volume>420</volume><journalNumber/><paginationStart>116672</paginationStart><paginationEnd/><publisher>Elsevier BV</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint>0045-7825</issnPrint><issnElectronic>1879-2138</issnElectronic><keywords>Data driven computational mechanics, Neural network, Constitutive modelling, Elasto-plasticity, Damage mechanics</keywords><publishedDay>1</publishedDay><publishedMonth>2</publishedMonth><publishedYear>2024</publishedYear><publishedDate>2024-02-01</publishedDate><doi>10.1016/j.cma.2023.116672</doi><url>http://dx.doi.org/10.1016/j.cma.2023.116672</url><notes/><college>COLLEGE NANME</college><department>Aerospace Engineering</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>AERO</DepartmentCode><institution>Swansea University</institution><apcterm>SU Library paid the OA fee (TA Institutional Deal)</apcterm><funders>Eugenio J. Muttio gratefully acknowledges research support provided by UKAEA and EPSRC through the Doctoral Training Partnership (DTP) scheme. This work has been part-funded by the EPSRC Energy Programme [grant number EP/W006839/1].</funders><projectreference/><lastEdited>2024-03-25T11:41:43.8418973</lastEdited><Created>2023-12-18T13:43:35.7648149</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Aerospace Engineering</level></path><authors><author><firstname>Wulf</firstname><surname>Dettmer</surname><orcid>0000-0003-0799-4645</orcid><order>1</order></author><author><firstname>Eugenio J.</firstname><surname>Muttio</surname><orcid>0000-0002-7555-9023</orcid><order>2</order></author><author><firstname>Reem</firstname><surname>Alhayki</surname><order>3</order></author><author><firstname>Djordje</firstname><surname>Perić</surname><order>4</order></author><author><firstname>Eugenio</firstname><surname>Muttio Zavala</surname><order>5</order></author><author><firstname>Djordje</firstname><surname>Peric</surname><orcid>0000-0002-1112-301X</orcid><order>6</order></author></authors><documents><document><filename>65340__29285__76113c83a15041b3b6468d3e453dd487.pdf</filename><originalFilename>65340.VOR.pdf</originalFilename><uploaded>2023-12-18T13:48:33.0986914</uploaded><type>Output</type><contentLength>5820545</contentLength><contentType>application/pdf</contentType><version>Version of Record</version><cronfaStatus>true</cronfaStatus><documentNotes>© 2023 The Author(s). Published by Elsevier B.V. Distributed under the terms of a Creative Commons Attribution 4.0 International License (CC BY 4.0).</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>https://creativecommons.org/licenses/by/4.0/</licence></document></documents><OutputDurs/></rfc1807> |
spelling |
v2 65340 2023-12-18 A framework for neural network based constitutive modelling of inelastic materials 30bb53ad906e7160e947fa01c16abf55 0000-0003-0799-4645 Wulf Dettmer Wulf Dettmer true false f7aa889da52c57c88e62d530e687ea13 Reem Alhayki Reem Alhayki true false ee7320f4fba56d3fc7eea1bcdd28e615 Eugenio Muttio Zavala Eugenio Muttio Zavala true false 9d35cb799b2542ad39140943a9a9da65 0000-0002-1112-301X Djordje Peric Djordje Peric true false 2023-12-18 AERO Given the significant recent advances in added layer manufacturing and materials engineering, new types of materials or new material micro-structures are becoming available at a fast rate. The finite element analysis of structures or structural components requires a constitutive model that describes the behaviour of the new materials. The formulation of accurate constitutive equations is generally complex and time consuming. Hence, suitable machine learning strategies may be used to render this process obsolete and bridge the gap between experimental data and finite element analysis. In this work, a generic stress update procedure is presented that is suitable for the modelling of rate-independent, elastic or inelastic, isotropic or anisotropic material behaviour. The proposed strategy is based on a recurrent neural network architecture and must be trained on stress and strain data sequences that represent physical or numerical experiments. A training strategy based on gradient-free optimisation is presented. It is shown that piecewise linear behaviour, such as uniaxial elasto-plasticity, can be represented exactly. Further numerical examples include uniaxial damage mechanics and elasto-plasticity under plane strain conditions. An efficient criterion for the verification of thermodynamic consistency is proposed and applied to the trained stress update models. The strategy is compared to GRU or LSTM based architectures and shown to offer advantages. Journal Article Computer Methods in Applied Mechanics and Engineering 420 116672 Elsevier BV 0045-7825 1879-2138 Data driven computational mechanics, Neural network, Constitutive modelling, Elasto-plasticity, Damage mechanics 1 2 2024 2024-02-01 10.1016/j.cma.2023.116672 http://dx.doi.org/10.1016/j.cma.2023.116672 COLLEGE NANME Aerospace Engineering COLLEGE CODE AERO Swansea University SU Library paid the OA fee (TA Institutional Deal) Eugenio J. Muttio gratefully acknowledges research support provided by UKAEA and EPSRC through the Doctoral Training Partnership (DTP) scheme. This work has been part-funded by the EPSRC Energy Programme [grant number EP/W006839/1]. 2024-03-25T11:41:43.8418973 2023-12-18T13:43:35.7648149 Faculty of Science and Engineering School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Aerospace Engineering Wulf Dettmer 0000-0003-0799-4645 1 Eugenio J. Muttio 0000-0002-7555-9023 2 Reem Alhayki 3 Djordje Perić 4 Eugenio Muttio Zavala 5 Djordje Peric 0000-0002-1112-301X 6 65340__29285__76113c83a15041b3b6468d3e453dd487.pdf 65340.VOR.pdf 2023-12-18T13:48:33.0986914 Output 5820545 application/pdf Version of Record true © 2023 The Author(s). Published by Elsevier B.V. Distributed under the terms of a Creative Commons Attribution 4.0 International License (CC BY 4.0). true eng https://creativecommons.org/licenses/by/4.0/ |
title |
A framework for neural network based constitutive modelling of inelastic materials |
spellingShingle |
A framework for neural network based constitutive modelling of inelastic materials Wulf Dettmer Reem Alhayki Eugenio Muttio Zavala Djordje Peric |
title_short |
A framework for neural network based constitutive modelling of inelastic materials |
title_full |
A framework for neural network based constitutive modelling of inelastic materials |
title_fullStr |
A framework for neural network based constitutive modelling of inelastic materials |
title_full_unstemmed |
A framework for neural network based constitutive modelling of inelastic materials |
title_sort |
A framework for neural network based constitutive modelling of inelastic materials |
author_id_str_mv |
30bb53ad906e7160e947fa01c16abf55 f7aa889da52c57c88e62d530e687ea13 ee7320f4fba56d3fc7eea1bcdd28e615 9d35cb799b2542ad39140943a9a9da65 |
author_id_fullname_str_mv |
30bb53ad906e7160e947fa01c16abf55_***_Wulf Dettmer f7aa889da52c57c88e62d530e687ea13_***_Reem Alhayki ee7320f4fba56d3fc7eea1bcdd28e615_***_Eugenio Muttio Zavala 9d35cb799b2542ad39140943a9a9da65_***_Djordje Peric |
author |
Wulf Dettmer Reem Alhayki Eugenio Muttio Zavala Djordje Peric |
author2 |
Wulf Dettmer Eugenio J. Muttio Reem Alhayki Djordje Perić Eugenio Muttio Zavala Djordje Peric |
format |
Journal article |
container_title |
Computer Methods in Applied Mechanics and Engineering |
container_volume |
420 |
container_start_page |
116672 |
publishDate |
2024 |
institution |
Swansea University |
issn |
0045-7825 1879-2138 |
doi_str_mv |
10.1016/j.cma.2023.116672 |
publisher |
Elsevier BV |
college_str |
Faculty of Science and Engineering |
hierarchytype |
|
hierarchy_top_id |
facultyofscienceandengineering |
hierarchy_top_title |
Faculty of Science and Engineering |
hierarchy_parent_id |
facultyofscienceandengineering |
hierarchy_parent_title |
Faculty of Science and Engineering |
department_str |
School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Aerospace Engineering{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Aerospace Engineering |
url |
http://dx.doi.org/10.1016/j.cma.2023.116672 |
document_store_str |
1 |
active_str |
0 |
description |
Given the significant recent advances in added layer manufacturing and materials engineering, new types of materials or new material micro-structures are becoming available at a fast rate. The finite element analysis of structures or structural components requires a constitutive model that describes the behaviour of the new materials. The formulation of accurate constitutive equations is generally complex and time consuming. Hence, suitable machine learning strategies may be used to render this process obsolete and bridge the gap between experimental data and finite element analysis. In this work, a generic stress update procedure is presented that is suitable for the modelling of rate-independent, elastic or inelastic, isotropic or anisotropic material behaviour. The proposed strategy is based on a recurrent neural network architecture and must be trained on stress and strain data sequences that represent physical or numerical experiments. A training strategy based on gradient-free optimisation is presented. It is shown that piecewise linear behaviour, such as uniaxial elasto-plasticity, can be represented exactly. Further numerical examples include uniaxial damage mechanics and elasto-plasticity under plane strain conditions. An efficient criterion for the verification of thermodynamic consistency is proposed and applied to the trained stress update models. The strategy is compared to GRU or LSTM based architectures and shown to offer advantages. |
published_date |
2024-02-01T11:41:40Z |
_version_ |
1794498259483361280 |
score |
11.022936 |