Physics-agnostic and physics-infused machine learning for thin films flows: modelling, and predictions from small data

Numerical simulations of multiphase flows are crucial in numerous engineering applications, but are often limited by the computationally demanding solution of the Navier–Stokes (NS) equations. The development of surrogate models relies on involved algebra and several assumptions. Here, we present a...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of fluid mechanics 2023-11, Vol.975, Article A41
Hauptverfasser: Martin-Linares, Cristina P., Psarellis, Yorgos M., Karapetsas, George, Koronaki, Eleni D., Kevrekidis, Ioannis G.
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 Journal of fluid mechanics
container_volume 975
creator Martin-Linares, Cristina P.
Psarellis, Yorgos M.
Karapetsas, George
Koronaki, Eleni D.
Kevrekidis, Ioannis G.
description Numerical simulations of multiphase flows are crucial in numerous engineering applications, but are often limited by the computationally demanding solution of the Navier–Stokes (NS) equations. The development of surrogate models relies on involved algebra and several assumptions. Here, we present a data-driven workflow where a handful of detailed NS simulation data are leveraged into a reduced-order model for a prototypical vertically falling liquid film. We develop a physics-agnostic model for the film thickness, achieving a far better agreement with the NS solutions than the asymptotic Kuramoto–Sivashinsky (KS) equation. We also develop two variants of physics-infused models providing a form of calibration of a low-fidelity model (i.e. the KS) against a few high-fidelity NS data. Finally, predictive models for missing data are developed, for either the amplitude, or the full-field velocity and even the flow parameter from partial information. This is achieved with the so-called ‘gappy diffusion maps’, which we compare favourably to its linear counterpart, gappy POD.
doi_str_mv 10.1017/jfm.2023.868
format Article
fullrecord <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_journals_2896734975</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><cupid>10_1017_jfm_2023_868</cupid><sourcerecordid>2896734975</sourcerecordid><originalsourceid>FETCH-LOGICAL-c340t-f42cbc2a0ddd4f1dbdd4e7116f22627598fd428d502520a770752e187388f18f3</originalsourceid><addsrcrecordid>eNptkMlKQzEUhoMoWIedDxBw23tNcofkupPiBAVd6DqkGdqUDDW5Vfr2prTgxtWB_3znP_ABcINRjRGmd2vja4JIU7OenYAJbvuhon3bnYIJQoRUGBN0Di5yXiOEGzTQCfh-X-2ylbkSyxDzaCUUQcHNMbTBbLNW0Au5skFDp0UKNiyhiQmOJYLGOp-hcfEn30MflXau7KeHlqSVlaONoRApepi9cA4qMYorcGaEy_r6OC_B59Pjx-ylmr89v84e5pVsWjRWpiVyIYlASqnWYLUoQ1OMe0NIT2g3MKNawlSHSEeQoBTRjmjMaMOYwcw0l-D20LtJ8Wur88jXcZtCeckJG3ratAPtCjU9UDLFnJM2fJOsF2nHMeJ7s7yY5XuzvJgteH3EhV8kq5b6r_Xfg1_csHxv</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2896734975</pqid></control><display><type>article</type><title>Physics-agnostic and physics-infused machine learning for thin films flows: modelling, and predictions from small data</title><source>Cambridge University Press Journals Complete</source><creator>Martin-Linares, Cristina P. ; Psarellis, Yorgos M. ; Karapetsas, George ; Koronaki, Eleni D. ; Kevrekidis, Ioannis G.</creator><creatorcontrib>Martin-Linares, Cristina P. ; Psarellis, Yorgos M. ; Karapetsas, George ; Koronaki, Eleni D. ; Kevrekidis, Ioannis G.</creatorcontrib><description>Numerical simulations of multiphase flows are crucial in numerous engineering applications, but are often limited by the computationally demanding solution of the Navier–Stokes (NS) equations. The development of surrogate models relies on involved algebra and several assumptions. Here, we present a data-driven workflow where a handful of detailed NS simulation data are leveraged into a reduced-order model for a prototypical vertically falling liquid film. We develop a physics-agnostic model for the film thickness, achieving a far better agreement with the NS solutions than the asymptotic Kuramoto–Sivashinsky (KS) equation. We also develop two variants of physics-infused models providing a form of calibration of a low-fidelity model (i.e. the KS) against a few high-fidelity NS data. Finally, predictive models for missing data are developed, for either the amplitude, or the full-field velocity and even the flow parameter from partial information. This is achieved with the so-called ‘gappy diffusion maps’, which we compare favourably to its linear counterpart, gappy POD.</description><identifier>ISSN: 0022-1120</identifier><identifier>EISSN: 1469-7645</identifier><identifier>DOI: 10.1017/jfm.2023.868</identifier><language>eng</language><publisher>Cambridge, UK: Cambridge University Press</publisher><subject>Accuracy ; Falling liquid films ; Film thickness ; JFM Papers ; Machine learning ; Mathematical models ; Missing data ; Multiphase flow ; Neural networks ; Partial differential equations ; Physics ; Prediction models ; Reduced order models ; Reynolds number ; Simulation ; Thin films ; Velocity ; Workflow</subject><ispartof>Journal of fluid mechanics, 2023-11, Vol.975, Article A41</ispartof><rights>The Author(s), 2023. Published by Cambridge University Press</rights><rights>The Author(s), 2023. Published by Cambridge University Press. This work is licensed under the Creative Commons Attribution License This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited. (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c340t-f42cbc2a0ddd4f1dbdd4e7116f22627598fd428d502520a770752e187388f18f3</citedby><cites>FETCH-LOGICAL-c340t-f42cbc2a0ddd4f1dbdd4e7116f22627598fd428d502520a770752e187388f18f3</cites><orcidid>0000-0002-3486-7484 ; 0000-0002-5229-4157</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://www.cambridge.org/core/product/identifier/S0022112023008686/type/journal_article$$EHTML$$P50$$Gcambridge$$Hfree_for_read</linktohtml><link.rule.ids>164,314,780,784,27924,27925,55628</link.rule.ids></links><search><creatorcontrib>Martin-Linares, Cristina P.</creatorcontrib><creatorcontrib>Psarellis, Yorgos M.</creatorcontrib><creatorcontrib>Karapetsas, George</creatorcontrib><creatorcontrib>Koronaki, Eleni D.</creatorcontrib><creatorcontrib>Kevrekidis, Ioannis G.</creatorcontrib><title>Physics-agnostic and physics-infused machine learning for thin films flows: modelling, and predictions from small data</title><title>Journal of fluid mechanics</title><addtitle>J. Fluid Mech</addtitle><description>Numerical simulations of multiphase flows are crucial in numerous engineering applications, but are often limited by the computationally demanding solution of the Navier–Stokes (NS) equations. The development of surrogate models relies on involved algebra and several assumptions. Here, we present a data-driven workflow where a handful of detailed NS simulation data are leveraged into a reduced-order model for a prototypical vertically falling liquid film. We develop a physics-agnostic model for the film thickness, achieving a far better agreement with the NS solutions than the asymptotic Kuramoto–Sivashinsky (KS) equation. We also develop two variants of physics-infused models providing a form of calibration of a low-fidelity model (i.e. the KS) against a few high-fidelity NS data. Finally, predictive models for missing data are developed, for either the amplitude, or the full-field velocity and even the flow parameter from partial information. This is achieved with the so-called ‘gappy diffusion maps’, which we compare favourably to its linear counterpart, gappy POD.</description><subject>Accuracy</subject><subject>Falling liquid films</subject><subject>Film thickness</subject><subject>JFM Papers</subject><subject>Machine learning</subject><subject>Mathematical models</subject><subject>Missing data</subject><subject>Multiphase flow</subject><subject>Neural networks</subject><subject>Partial differential equations</subject><subject>Physics</subject><subject>Prediction models</subject><subject>Reduced order models</subject><subject>Reynolds number</subject><subject>Simulation</subject><subject>Thin films</subject><subject>Velocity</subject><subject>Workflow</subject><issn>0022-1120</issn><issn>1469-7645</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>IKXGN</sourceid><sourceid>8G5</sourceid><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><sourceid>GNUQQ</sourceid><sourceid>GUQSH</sourceid><sourceid>M2O</sourceid><recordid>eNptkMlKQzEUhoMoWIedDxBw23tNcofkupPiBAVd6DqkGdqUDDW5Vfr2prTgxtWB_3znP_ABcINRjRGmd2vja4JIU7OenYAJbvuhon3bnYIJQoRUGBN0Di5yXiOEGzTQCfh-X-2ylbkSyxDzaCUUQcHNMbTBbLNW0Au5skFDp0UKNiyhiQmOJYLGOp-hcfEn30MflXau7KeHlqSVlaONoRApepi9cA4qMYorcGaEy_r6OC_B59Pjx-ylmr89v84e5pVsWjRWpiVyIYlASqnWYLUoQ1OMe0NIT2g3MKNawlSHSEeQoBTRjmjMaMOYwcw0l-D20LtJ8Wur88jXcZtCeckJG3ratAPtCjU9UDLFnJM2fJOsF2nHMeJ7s7yY5XuzvJgteH3EhV8kq5b6r_Xfg1_csHxv</recordid><startdate>20231127</startdate><enddate>20231127</enddate><creator>Martin-Linares, Cristina P.</creator><creator>Psarellis, Yorgos M.</creator><creator>Karapetsas, George</creator><creator>Koronaki, Eleni D.</creator><creator>Kevrekidis, Ioannis G.</creator><general>Cambridge University Press</general><scope>IKXGN</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>3V.</scope><scope>7TB</scope><scope>7U5</scope><scope>7UA</scope><scope>7XB</scope><scope>88I</scope><scope>8FD</scope><scope>8FE</scope><scope>8FG</scope><scope>8FK</scope><scope>8G5</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>ARAPS</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>BHPHI</scope><scope>BKSAR</scope><scope>C1K</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>F1W</scope><scope>FR3</scope><scope>GNUQQ</scope><scope>GUQSH</scope><scope>H8D</scope><scope>H96</scope><scope>HCIFZ</scope><scope>KR7</scope><scope>L.G</scope><scope>L6V</scope><scope>L7M</scope><scope>M2O</scope><scope>M2P</scope><scope>M7S</scope><scope>MBDVC</scope><scope>P5Z</scope><scope>P62</scope><scope>PCBAR</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PTHSS</scope><scope>Q9U</scope><scope>S0W</scope><orcidid>https://orcid.org/0000-0002-3486-7484</orcidid><orcidid>https://orcid.org/0000-0002-5229-4157</orcidid></search><sort><creationdate>20231127</creationdate><title>Physics-agnostic and physics-infused machine learning for thin films flows: modelling, and predictions from small data</title><author>Martin-Linares, Cristina P. ; Psarellis, Yorgos M. ; Karapetsas, George ; Koronaki, Eleni D. ; Kevrekidis, Ioannis G.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c340t-f42cbc2a0ddd4f1dbdd4e7116f22627598fd428d502520a770752e187388f18f3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Accuracy</topic><topic>Falling liquid films</topic><topic>Film thickness</topic><topic>JFM Papers</topic><topic>Machine learning</topic><topic>Mathematical models</topic><topic>Missing data</topic><topic>Multiphase flow</topic><topic>Neural networks</topic><topic>Partial differential equations</topic><topic>Physics</topic><topic>Prediction models</topic><topic>Reduced order models</topic><topic>Reynolds number</topic><topic>Simulation</topic><topic>Thin films</topic><topic>Velocity</topic><topic>Workflow</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Martin-Linares, Cristina P.</creatorcontrib><creatorcontrib>Psarellis, Yorgos M.</creatorcontrib><creatorcontrib>Karapetsas, George</creatorcontrib><creatorcontrib>Koronaki, Eleni D.</creatorcontrib><creatorcontrib>Kevrekidis, Ioannis G.</creatorcontrib><collection>Cambridge University Press Wholly Gold Open Access Journals</collection><collection>CrossRef</collection><collection>ProQuest Central (Corporate)</collection><collection>Mechanical &amp; Transportation Engineering Abstracts</collection><collection>Solid State and Superconductivity Abstracts</collection><collection>Water Resources Abstracts</collection><collection>ProQuest Central (purchase pre-March 2016)</collection><collection>Science Database (Alumni Edition)</collection><collection>Technology Research Database</collection><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>ProQuest Central (Alumni) (purchase pre-March 2016)</collection><collection>Research Library (Alumni Edition)</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>Advanced Technologies &amp; Aerospace Collection</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>Natural Science Collection</collection><collection>Earth, Atmospheric &amp; Aquatic Science Collection</collection><collection>Environmental Sciences and Pollution Management</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>ASFA: Aquatic Sciences and Fisheries Abstracts</collection><collection>Engineering Research Database</collection><collection>ProQuest Central Student</collection><collection>Research Library Prep</collection><collection>Aerospace Database</collection><collection>Aquatic Science &amp; Fisheries Abstracts (ASFA) 2: Ocean Technology, Policy &amp; Non-Living Resources</collection><collection>SciTech Premium Collection</collection><collection>Civil Engineering Abstracts</collection><collection>Aquatic Science &amp; Fisheries Abstracts (ASFA) Professional</collection><collection>ProQuest Engineering Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Research Library</collection><collection>Science Database</collection><collection>Engineering Database</collection><collection>Research Library (Corporate)</collection><collection>Advanced Technologies &amp; Aerospace Database</collection><collection>ProQuest Advanced Technologies &amp; Aerospace Collection</collection><collection>Earth, Atmospheric &amp; Aquatic Science 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>Engineering Collection</collection><collection>ProQuest Central Basic</collection><collection>DELNET Engineering &amp; Technology Collection</collection><jtitle>Journal of fluid mechanics</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Martin-Linares, Cristina P.</au><au>Psarellis, Yorgos M.</au><au>Karapetsas, George</au><au>Koronaki, Eleni D.</au><au>Kevrekidis, Ioannis G.</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Physics-agnostic and physics-infused machine learning for thin films flows: modelling, and predictions from small data</atitle><jtitle>Journal of fluid mechanics</jtitle><addtitle>J. Fluid Mech</addtitle><date>2023-11-27</date><risdate>2023</risdate><volume>975</volume><artnum>A41</artnum><issn>0022-1120</issn><eissn>1469-7645</eissn><abstract>Numerical simulations of multiphase flows are crucial in numerous engineering applications, but are often limited by the computationally demanding solution of the Navier–Stokes (NS) equations. The development of surrogate models relies on involved algebra and several assumptions. Here, we present a data-driven workflow where a handful of detailed NS simulation data are leveraged into a reduced-order model for a prototypical vertically falling liquid film. We develop a physics-agnostic model for the film thickness, achieving a far better agreement with the NS solutions than the asymptotic Kuramoto–Sivashinsky (KS) equation. We also develop two variants of physics-infused models providing a form of calibration of a low-fidelity model (i.e. the KS) against a few high-fidelity NS data. Finally, predictive models for missing data are developed, for either the amplitude, or the full-field velocity and even the flow parameter from partial information. This is achieved with the so-called ‘gappy diffusion maps’, which we compare favourably to its linear counterpart, gappy POD.</abstract><cop>Cambridge, UK</cop><pub>Cambridge University Press</pub><doi>10.1017/jfm.2023.868</doi><tpages>22</tpages><orcidid>https://orcid.org/0000-0002-3486-7484</orcidid><orcidid>https://orcid.org/0000-0002-5229-4157</orcidid><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 0022-1120
ispartof Journal of fluid mechanics, 2023-11, Vol.975, Article A41
issn 0022-1120
1469-7645
language eng
recordid cdi_proquest_journals_2896734975
source Cambridge University Press Journals Complete
subjects Accuracy
Falling liquid films
Film thickness
JFM Papers
Machine learning
Mathematical models
Missing data
Multiphase flow
Neural networks
Partial differential equations
Physics
Prediction models
Reduced order models
Reynolds number
Simulation
Thin films
Velocity
Workflow
title Physics-agnostic and physics-infused machine learning for thin films flows: modelling, and predictions from small data
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-02T10%3A55%3A44IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Physics-agnostic%20and%20physics-infused%20machine%20learning%20for%20thin%20films%20flows:%20modelling,%20and%20predictions%20from%20small%20data&rft.jtitle=Journal%20of%20fluid%20mechanics&rft.au=Martin-Linares,%20Cristina%20P.&rft.date=2023-11-27&rft.volume=975&rft.artnum=A41&rft.issn=0022-1120&rft.eissn=1469-7645&rft_id=info:doi/10.1017/jfm.2023.868&rft_dat=%3Cproquest_cross%3E2896734975%3C/proquest_cross%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2896734975&rft_id=info:pmid/&rft_cupid=10_1017_jfm_2023_868&rfr_iscdi=true