Physics-informed MeshGraphNets (PI-MGNs): Neural finite element solvers for non-stationary and nonlinear simulations on arbitrary meshes
Engineering components must meet increasing technological demands in ever shorter development cycles. To face these challenges, a holistic approach is essential that allows for the concurrent development of part design, material system and manufacturing process. Current approaches employ numerical s...
Gespeichert in:
Veröffentlicht in: | arXiv.org 2024-02 |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | |
---|---|
container_issue | |
container_start_page | |
container_title | arXiv.org |
container_volume | |
creator | Würth, Tobias Freymuth, Niklas Zimmerling, Clemens Neumann, Gerhard Kärger, Luise |
description | Engineering components must meet increasing technological demands in ever shorter development cycles. To face these challenges, a holistic approach is essential that allows for the concurrent development of part design, material system and manufacturing process. Current approaches employ numerical simulations, which however quickly becomes computation-intensive, especially for iterative optimization. Data-driven machine learning methods can be used to replace time- and resource-intensive numerical simulations. In particular, MeshGraphNets (MGNs) have shown promising results. They enable fast and accurate predictions on unseen mesh geometries while being fully differentiable for optimization. However, these models rely on large amounts of expensive training data, such as numerical simulations. Physics-informed neural networks (PINNs) offer an opportunity to train neural networks with partial differential equations instead of labeled data, but have not been extended yet to handle time-dependent simulations of arbitrary meshes. This work introduces PI-MGNs, a hybrid approach that combines PINNs and MGNs to quickly and accurately solve non-stationary and nonlinear partial differential equations (PDEs) on arbitrary meshes. The method is exemplified for thermal process simulations of unseen parts with inhomogeneous material distribution. Further results show that the model scales well to large and complex meshes, although it is trained on small generic meshes only. |
doi_str_mv | 10.48550/arxiv.2402.10681 |
format | Article |
fullrecord | <record><control><sourceid>proquest_arxiv</sourceid><recordid>TN_cdi_arxiv_primary_2402_10681</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2928439595</sourcerecordid><originalsourceid>FETCH-LOGICAL-a951-466bdc690c34f26220347908c77a814f05f5bdf5094918e0b42821c7a0f19e8c3</originalsourceid><addsrcrecordid>eNotkM9Kw0AQhxdBsNQ-gCcXvOghdf8mu96kaBXa2kPvYZPs0i3Jpu4kxb6Bj23aehqY-Wbmx4fQHSVToaQkzyb--MOUCcKmlKSKXqER45wmSjB2gyYAO0IISzMmJR-h3_X2CL6ExAfXxsZWeGlhO49mv13ZDvDj-jNZzlfw9IJXto-mxs4H31lsa9vY0GFo64ONgIdtHNqQQGc63wYTj9iE6tSqfbAmYvBNX59ngNuATSx8F09YMzy0cIuunanBTv7rGG3e3zazj2TxNf-cvS4SoyVNRJoWVZlqUnLhWMoY4SLTRJVZZhQVjkgni8pJooWmypJCMMVomRniqLaq5GN0fzl71pTvo2-GDPlJV37WNRAPF2If2-_eQpfv2j6GIVPONFOCa6kl_wOehG66</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2928439595</pqid></control><display><type>article</type><title>Physics-informed MeshGraphNets (PI-MGNs): Neural finite element solvers for non-stationary and nonlinear simulations on arbitrary meshes</title><source>arXiv.org</source><source>Free E- Journals</source><creator>Würth, Tobias ; Freymuth, Niklas ; Zimmerling, Clemens ; Neumann, Gerhard ; Kärger, Luise</creator><creatorcontrib>Würth, Tobias ; Freymuth, Niklas ; Zimmerling, Clemens ; Neumann, Gerhard ; Kärger, Luise</creatorcontrib><description>Engineering components must meet increasing technological demands in ever shorter development cycles. To face these challenges, a holistic approach is essential that allows for the concurrent development of part design, material system and manufacturing process. Current approaches employ numerical simulations, which however quickly becomes computation-intensive, especially for iterative optimization. Data-driven machine learning methods can be used to replace time- and resource-intensive numerical simulations. In particular, MeshGraphNets (MGNs) have shown promising results. They enable fast and accurate predictions on unseen mesh geometries while being fully differentiable for optimization. However, these models rely on large amounts of expensive training data, such as numerical simulations. Physics-informed neural networks (PINNs) offer an opportunity to train neural networks with partial differential equations instead of labeled data, but have not been extended yet to handle time-dependent simulations of arbitrary meshes. This work introduces PI-MGNs, a hybrid approach that combines PINNs and MGNs to quickly and accurately solve non-stationary and nonlinear partial differential equations (PDEs) on arbitrary meshes. The method is exemplified for thermal process simulations of unseen parts with inhomogeneous material distribution. Further results show that the model scales well to large and complex meshes, although it is trained on small generic meshes only.</description><identifier>EISSN: 2331-8422</identifier><identifier>DOI: 10.48550/arxiv.2402.10681</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Computer Science - Artificial Intelligence ; Computer Science - Computational Engineering, Finance, and Science ; Computer Science - Learning ; Computer simulation ; Finite element method ; Iterative methods ; Machine learning ; Mathematical models ; Neural networks ; Nonlinear differential equations ; Optimization ; Partial differential equations ; Simulation ; Thermal simulation ; Time dependence ; Time dependent analysis</subject><ispartof>arXiv.org, 2024-02</ispartof><rights>2024. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><rights>http://creativecommons.org/licenses/by/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,784,885,27925</link.rule.ids><backlink>$$Uhttps://doi.org/10.48550/arXiv.2402.10681$$DView paper in arXiv$$Hfree_for_read</backlink><backlink>$$Uhttps://doi.org/10.1016/j.cma.2024.117102$$DView published paper (Access to full text may be restricted)$$Hfree_for_read</backlink></links><search><creatorcontrib>Würth, Tobias</creatorcontrib><creatorcontrib>Freymuth, Niklas</creatorcontrib><creatorcontrib>Zimmerling, Clemens</creatorcontrib><creatorcontrib>Neumann, Gerhard</creatorcontrib><creatorcontrib>Kärger, Luise</creatorcontrib><title>Physics-informed MeshGraphNets (PI-MGNs): Neural finite element solvers for non-stationary and nonlinear simulations on arbitrary meshes</title><title>arXiv.org</title><description>Engineering components must meet increasing technological demands in ever shorter development cycles. To face these challenges, a holistic approach is essential that allows for the concurrent development of part design, material system and manufacturing process. Current approaches employ numerical simulations, which however quickly becomes computation-intensive, especially for iterative optimization. Data-driven machine learning methods can be used to replace time- and resource-intensive numerical simulations. In particular, MeshGraphNets (MGNs) have shown promising results. They enable fast and accurate predictions on unseen mesh geometries while being fully differentiable for optimization. However, these models rely on large amounts of expensive training data, such as numerical simulations. Physics-informed neural networks (PINNs) offer an opportunity to train neural networks with partial differential equations instead of labeled data, but have not been extended yet to handle time-dependent simulations of arbitrary meshes. This work introduces PI-MGNs, a hybrid approach that combines PINNs and MGNs to quickly and accurately solve non-stationary and nonlinear partial differential equations (PDEs) on arbitrary meshes. The method is exemplified for thermal process simulations of unseen parts with inhomogeneous material distribution. Further results show that the model scales well to large and complex meshes, although it is trained on small generic meshes only.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Computational Engineering, Finance, and Science</subject><subject>Computer Science - Learning</subject><subject>Computer simulation</subject><subject>Finite element method</subject><subject>Iterative methods</subject><subject>Machine learning</subject><subject>Mathematical models</subject><subject>Neural networks</subject><subject>Nonlinear differential equations</subject><subject>Optimization</subject><subject>Partial differential equations</subject><subject>Simulation</subject><subject>Thermal simulation</subject><subject>Time dependence</subject><subject>Time dependent analysis</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><sourceid>GOX</sourceid><recordid>eNotkM9Kw0AQhxdBsNQ-gCcXvOghdf8mu96kaBXa2kPvYZPs0i3Jpu4kxb6Bj23aehqY-Wbmx4fQHSVToaQkzyb--MOUCcKmlKSKXqER45wmSjB2gyYAO0IISzMmJR-h3_X2CL6ExAfXxsZWeGlhO49mv13ZDvDj-jNZzlfw9IJXto-mxs4H31lsa9vY0GFo64ONgIdtHNqQQGc63wYTj9iE6tSqfbAmYvBNX59ngNuATSx8F09YMzy0cIuunanBTv7rGG3e3zazj2TxNf-cvS4SoyVNRJoWVZlqUnLhWMoY4SLTRJVZZhQVjkgni8pJooWmypJCMMVomRniqLaq5GN0fzl71pTvo2-GDPlJV37WNRAPF2If2-_eQpfv2j6GIVPONFOCa6kl_wOehG66</recordid><startdate>20240216</startdate><enddate>20240216</enddate><creator>Würth, Tobias</creator><creator>Freymuth, Niklas</creator><creator>Zimmerling, Clemens</creator><creator>Neumann, Gerhard</creator><creator>Kärger, Luise</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20240216</creationdate><title>Physics-informed MeshGraphNets (PI-MGNs): Neural finite element solvers for non-stationary and nonlinear simulations on arbitrary meshes</title><author>Würth, Tobias ; Freymuth, Niklas ; Zimmerling, Clemens ; Neumann, Gerhard ; Kärger, Luise</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a951-466bdc690c34f26220347908c77a814f05f5bdf5094918e0b42821c7a0f19e8c3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Computational Engineering, Finance, and Science</topic><topic>Computer Science - Learning</topic><topic>Computer simulation</topic><topic>Finite element method</topic><topic>Iterative methods</topic><topic>Machine learning</topic><topic>Mathematical models</topic><topic>Neural networks</topic><topic>Nonlinear differential equations</topic><topic>Optimization</topic><topic>Partial differential equations</topic><topic>Simulation</topic><topic>Thermal simulation</topic><topic>Time dependence</topic><topic>Time dependent analysis</topic><toplevel>online_resources</toplevel><creatorcontrib>Würth, Tobias</creatorcontrib><creatorcontrib>Freymuth, Niklas</creatorcontrib><creatorcontrib>Zimmerling, Clemens</creatorcontrib><creatorcontrib>Neumann, Gerhard</creatorcontrib><creatorcontrib>Kärger, Luise</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science & Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection><collection>arXiv Computer Science</collection><collection>arXiv.org</collection><jtitle>arXiv.org</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Würth, Tobias</au><au>Freymuth, Niklas</au><au>Zimmerling, Clemens</au><au>Neumann, Gerhard</au><au>Kärger, Luise</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Physics-informed MeshGraphNets (PI-MGNs): Neural finite element solvers for non-stationary and nonlinear simulations on arbitrary meshes</atitle><jtitle>arXiv.org</jtitle><date>2024-02-16</date><risdate>2024</risdate><eissn>2331-8422</eissn><abstract>Engineering components must meet increasing technological demands in ever shorter development cycles. To face these challenges, a holistic approach is essential that allows for the concurrent development of part design, material system and manufacturing process. Current approaches employ numerical simulations, which however quickly becomes computation-intensive, especially for iterative optimization. Data-driven machine learning methods can be used to replace time- and resource-intensive numerical simulations. In particular, MeshGraphNets (MGNs) have shown promising results. They enable fast and accurate predictions on unseen mesh geometries while being fully differentiable for optimization. However, these models rely on large amounts of expensive training data, such as numerical simulations. Physics-informed neural networks (PINNs) offer an opportunity to train neural networks with partial differential equations instead of labeled data, but have not been extended yet to handle time-dependent simulations of arbitrary meshes. This work introduces PI-MGNs, a hybrid approach that combines PINNs and MGNs to quickly and accurately solve non-stationary and nonlinear partial differential equations (PDEs) on arbitrary meshes. The method is exemplified for thermal process simulations of unseen parts with inhomogeneous material distribution. Further results show that the model scales well to large and complex meshes, although it is trained on small generic meshes only.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><doi>10.48550/arxiv.2402.10681</doi><oa>free_for_read</oa></addata></record> |
fulltext | fulltext |
identifier | EISSN: 2331-8422 |
ispartof | arXiv.org, 2024-02 |
issn | 2331-8422 |
language | eng |
recordid | cdi_arxiv_primary_2402_10681 |
source | arXiv.org; Free E- Journals |
subjects | Computer Science - Artificial Intelligence Computer Science - Computational Engineering, Finance, and Science Computer Science - Learning Computer simulation Finite element method Iterative methods Machine learning Mathematical models Neural networks Nonlinear differential equations Optimization Partial differential equations Simulation Thermal simulation Time dependence Time dependent analysis |
title | Physics-informed MeshGraphNets (PI-MGNs): Neural finite element solvers for non-stationary and nonlinear simulations on arbitrary meshes |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-06T23%3A51%3A49IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_arxiv&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Physics-informed%20MeshGraphNets%20(PI-MGNs):%20Neural%20finite%20element%20solvers%20for%20non-stationary%20and%20nonlinear%20simulations%20on%20arbitrary%20meshes&rft.jtitle=arXiv.org&rft.au=W%C3%BCrth,%20Tobias&rft.date=2024-02-16&rft.eissn=2331-8422&rft_id=info:doi/10.48550/arxiv.2402.10681&rft_dat=%3Cproquest_arxiv%3E2928439595%3C/proquest_arxiv%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2928439595&rft_id=info:pmid/&rfr_iscdi=true |