No Cover Image

E-Thesis 319 views

A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids / EUGENIO ZAVALA

Swansea University Author: EUGENIO ZAVALA

  • E-Thesis under embargo until: 31st December 2026

DOI (Published version): 10.23889/SUThesis.69253

Abstract

A neural network-based surrogate model for inelastic solid materials simulations is presented. The network architecture incorporates the elastoplasticity equations and has been proven to be exact in one-dimensional elastoplasticity with hardening. This strategy provides an alternative to the complex...

Full description

Published: Swansea University, Wales, UK. 2024
Institution: Swansea University
Degree level: Doctoral
Degree name: Ph.D
Supervisor: Perić, D., and Dettmer, W. G.
URI: https://cronfa.swan.ac.uk/Record/cronfa69253
first_indexed 2025-04-10T11:02:16Z
last_indexed 2025-04-11T05:22:34Z
id cronfa69253
recordtype RisThesis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2025-04-10T12:07:47.7024225</datestamp><bib-version>v2</bib-version><id>69253</id><entry>2025-04-10</entry><title>A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids</title><swanseaauthors><author><sid>a43b6ee84ec8840b60ad9a640eaa344c</sid><firstname>EUGENIO</firstname><surname>ZAVALA</surname><name>EUGENIO ZAVALA</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2025-04-10</date><abstract>A neural network-based surrogate model for inelastic solid materials simulations is presented. The network architecture incorporates the elastoplasticity equations and has been proven to be exact in one-dimensional elastoplasticity with hardening. This strategy provides an alternative to the complex task of formulating constitutive equations for new materials due to the capability of learning directly from data. The strategy has a wide range of applications un- der the two main scenarios that generate the necessary data: (a) by performing physical experiments, and (b) by numerical homogenisation within the multi-scale analysis. The network is constructed based on the concept of internal states and utilises only observable variables (strains and stresses); hence, a recurrent structure is required to account for the deformation history. Due to the challenging task of training recurrent networks, a supervised parallel optimisation framework has been developed, which combines the exploration and exploitation capabilities of one (or several) meta-heuristic algorithms. The proposed optimisation strategy is a general-purpose optimisation framework. Therefore, this development offers a tool for a wide range of applications. Several analyses have been conducted, including hyperparameter analysis, algorithmic complexity and parallel message communications. The obtained performance of the proposed methodology surpasses current state-of-the-art algorithms. The trained network- based constitutive model is embedded as a surrogate model into a finite element code on a Gauss point level, replacing a traditional library of algorithmic constitutive models. The stress update is obtained from the network output, and the consistent tangent modulus is derived from the network architecture utilising the prescribed functional dependencies. Numerical experiments are presented using network surrogates trained with the Von Mises elastoplasticity model.</abstract><type>E-Thesis</type><journal/><volume/><journalNumber/><paginationStart/><paginationEnd/><publisher/><placeOfPublication>Swansea University, Wales, UK.</placeOfPublication><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic/><keywords>Data-driven computational mechanics, Neural Networks, Optimisation, Constitutive Modelling</keywords><publishedDay>2</publishedDay><publishedMonth>12</publishedMonth><publishedYear>2024</publishedYear><publishedDate>2024-12-02</publishedDate><doi>10.23889/SUThesis.69253</doi><url/><notes>A selection of content is redacted or is partially redacted from this thesis to protect sensitive and personal information.</notes><college>COLLEGE NANME</college><CollegeCode>COLLEGE CODE</CollegeCode><institution>Swansea University</institution><supervisor>Peri&#x107;, D., and Dettmer, W. G.</supervisor><degreelevel>Doctoral</degreelevel><degreename>Ph.D</degreename><degreesponsorsfunders>EPSRC / UKAEA</degreesponsorsfunders><apcterm/><funders>EPSRC / UKAEA</funders><projectreference/><lastEdited>2025-04-10T12:07:47.7024225</lastEdited><Created>2025-04-10T11:54:37.7876737</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Civil Engineering</level></path><authors><author><firstname>EUGENIO</firstname><surname>ZAVALA</surname><order>1</order></author></authors><documents><document><filename>Under embargo</filename><originalFilename>Under embargo</originalFilename><uploaded>2025-04-10T12:01:17.0742469</uploaded><type>Output</type><contentLength>76339742</contentLength><contentType>application/pdf</contentType><version>E-Thesis</version><cronfaStatus>true</cronfaStatus><embargoDate>2026-12-31T00:00:00.0000000</embargoDate><documentNotes>Copyright: The Author, Eugenio Jos&#xE9; Muttio Zavala, 2024 Distributed under the terms of a Creative Commons Attribution 4.0 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 2025-04-10T12:07:47.7024225 v2 69253 2025-04-10 A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids a43b6ee84ec8840b60ad9a640eaa344c EUGENIO ZAVALA EUGENIO ZAVALA true false 2025-04-10 A neural network-based surrogate model for inelastic solid materials simulations is presented. The network architecture incorporates the elastoplasticity equations and has been proven to be exact in one-dimensional elastoplasticity with hardening. This strategy provides an alternative to the complex task of formulating constitutive equations for new materials due to the capability of learning directly from data. The strategy has a wide range of applications un- der the two main scenarios that generate the necessary data: (a) by performing physical experiments, and (b) by numerical homogenisation within the multi-scale analysis. The network is constructed based on the concept of internal states and utilises only observable variables (strains and stresses); hence, a recurrent structure is required to account for the deformation history. Due to the challenging task of training recurrent networks, a supervised parallel optimisation framework has been developed, which combines the exploration and exploitation capabilities of one (or several) meta-heuristic algorithms. The proposed optimisation strategy is a general-purpose optimisation framework. Therefore, this development offers a tool for a wide range of applications. Several analyses have been conducted, including hyperparameter analysis, algorithmic complexity and parallel message communications. The obtained performance of the proposed methodology surpasses current state-of-the-art algorithms. The trained network- based constitutive model is embedded as a surrogate model into a finite element code on a Gauss point level, replacing a traditional library of algorithmic constitutive models. The stress update is obtained from the network output, and the consistent tangent modulus is derived from the network architecture utilising the prescribed functional dependencies. Numerical experiments are presented using network surrogates trained with the Von Mises elastoplasticity model. E-Thesis Swansea University, Wales, UK. Data-driven computational mechanics, Neural Networks, Optimisation, Constitutive Modelling 2 12 2024 2024-12-02 10.23889/SUThesis.69253 A selection of content is redacted or is partially redacted from this thesis to protect sensitive and personal information. COLLEGE NANME COLLEGE CODE Swansea University Perić, D., and Dettmer, W. G. Doctoral Ph.D EPSRC / UKAEA EPSRC / UKAEA 2025-04-10T12:07:47.7024225 2025-04-10T11:54:37.7876737 Faculty of Science and Engineering School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Civil Engineering EUGENIO ZAVALA 1 Under embargo Under embargo 2025-04-10T12:01:17.0742469 Output 76339742 application/pdf E-Thesis true 2026-12-31T00:00:00.0000000 Copyright: The Author, Eugenio José Muttio Zavala, 2024 Distributed under the terms of a Creative Commons Attribution 4.0 License (CC BY 4.0). true eng https://creativecommons.org/licenses/by/4.0/
title A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids
spellingShingle A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids
EUGENIO ZAVALA
title_short A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids
title_full A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids
title_fullStr A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids
title_full_unstemmed A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids
title_sort A Neural Network-Based Finite Element Framework for the Modelling of Inelastic Solids
author_id_str_mv a43b6ee84ec8840b60ad9a640eaa344c
author_id_fullname_str_mv a43b6ee84ec8840b60ad9a640eaa344c_***_EUGENIO ZAVALA
author EUGENIO ZAVALA
author2 EUGENIO ZAVALA
format E-Thesis
publishDate 2024
institution Swansea University
doi_str_mv 10.23889/SUThesis.69253
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 - Civil Engineering{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Civil Engineering
document_store_str 0
active_str 0
description A neural network-based surrogate model for inelastic solid materials simulations is presented. The network architecture incorporates the elastoplasticity equations and has been proven to be exact in one-dimensional elastoplasticity with hardening. This strategy provides an alternative to the complex task of formulating constitutive equations for new materials due to the capability of learning directly from data. The strategy has a wide range of applications un- der the two main scenarios that generate the necessary data: (a) by performing physical experiments, and (b) by numerical homogenisation within the multi-scale analysis. The network is constructed based on the concept of internal states and utilises only observable variables (strains and stresses); hence, a recurrent structure is required to account for the deformation history. Due to the challenging task of training recurrent networks, a supervised parallel optimisation framework has been developed, which combines the exploration and exploitation capabilities of one (or several) meta-heuristic algorithms. The proposed optimisation strategy is a general-purpose optimisation framework. Therefore, this development offers a tool for a wide range of applications. Several analyses have been conducted, including hyperparameter analysis, algorithmic complexity and parallel message communications. The obtained performance of the proposed methodology surpasses current state-of-the-art algorithms. The trained network- based constitutive model is embedded as a surrogate model into a finite element code on a Gauss point level, replacing a traditional library of algorithmic constitutive models. The stress update is obtained from the network output, and the consistent tangent modulus is derived from the network architecture utilising the prescribed functional dependencies. Numerical experiments are presented using network surrogates trained with the Von Mises elastoplasticity model.
published_date 2024-12-02T05:27:41Z
_version_ 1851097833885663232
score 11.089407

Similar Items