Development of subgrid-scale model for LES of Burgers turbulence with large filter size

A neural network (NN) with one hidden layer is implemented to establish a relationship between the resolved-scale flow field and the subgrid-scale (SGS) stress for large eddy simulation (LES) of the Burgers equation. Five sets of input are considered for the neural network by combining the velocity...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Physics of fluids (1994) 2022-04, Vol.34 (4)
1. Verfasser: Tabe Jamaat, A. Golsa
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue 4
container_start_page
container_title Physics of fluids (1994)
container_volume 34
creator Tabe Jamaat, A. Golsa
description A neural network (NN) with one hidden layer is implemented to establish a relationship between the resolved-scale flow field and the subgrid-scale (SGS) stress for large eddy simulation (LES) of the Burgers equation. Five sets of input are considered for the neural network by combining the velocity gradient and the filter size. The training datasets are obtained by filtering the direct numerical simulation (DNS) results of the Burgers equation with random forcing function. The number of modes is sufficiently large (N = 65 536) to resolve extremely small scales of motion. In the a priori test, a correlation coefficient over 0.93 is achieved for the SGS stress between the NN models and the filtered DNS data. The results of the a posteriori test reveal that the obtained solutions are stable for all NN models without applying any stabilization techniques. However, not all NN models have a reasonable performance when embedded in the LES code. The applicability of the NN models to the Burgers equation with higher and lower viscosity is also investigated, and it is indicated that the most reliable NN models obtained in this paper can be applied to a set of parameters which are different from those used in training. The results of the SGS models constructed using the neural network are also compared with the existing models, and it is shown that the best obtained NN models outperform the Smagorinsky model and the gradient model, and are comparable to the dynamic Smagorinsky model. However, the NN models have an advantage over the dynamic Smagorinsky model in numerical cost.
doi_str_mv 10.1063/5.0087761
format Article
fullrecord <record><control><sourceid>proquest_scita</sourceid><recordid>TN_cdi_scitation_primary_10_1063_5_0087761</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2652998510</sourcerecordid><originalsourceid>FETCH-LOGICAL-c257t-becf0857206a7cece764f4ba37bed83cdf7aff03d1ea8cc8c9a0125ec6173fa83</originalsourceid><addsrcrecordid>eNqd0E1LAzEQBuAgCtbqwX8Q8KSwNdk0H3vUWj-g4EHFY8hmJ3XLdrMm2Yr-ere04N3TDDMPM_AidE7JhBLBrvmEECWloAdoRIkqMimEONz2kmRCMHqMTmJcEUJYkYsRer-DDTS-W0ObsHc49uUy1FUWrWkAr30FDXY-4MX8Zbu-7cMSQsSpD2XfQGsBf9XpAzdmmGNXNwkCjvUPnKIjZ5oIZ_s6Rm_389fZY7Z4fnia3Swym3OZshKsI4rLnAgjLViQYuqmpWGyhEoxWzlpnCOsomCUtcoWhtCcgxVUMmcUG6OL3d0u-M8eYtIr34d2eKlzwfOiUJySQV3ulA0-xgBOd6Fem_CtKdHb3DTX-9wGe7Wz0dbJpNq3_8MbH_6g7irHfgETMHy-</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2652998510</pqid></control><display><type>article</type><title>Development of subgrid-scale model for LES of Burgers turbulence with large filter size</title><source>AIP Journals Complete</source><source>Alma/SFX Local Collection</source><creator>Tabe Jamaat, A. Golsa</creator><creatorcontrib>Tabe Jamaat, A. Golsa</creatorcontrib><description>A neural network (NN) with one hidden layer is implemented to establish a relationship between the resolved-scale flow field and the subgrid-scale (SGS) stress for large eddy simulation (LES) of the Burgers equation. Five sets of input are considered for the neural network by combining the velocity gradient and the filter size. The training datasets are obtained by filtering the direct numerical simulation (DNS) results of the Burgers equation with random forcing function. The number of modes is sufficiently large (N = 65 536) to resolve extremely small scales of motion. In the a priori test, a correlation coefficient over 0.93 is achieved for the SGS stress between the NN models and the filtered DNS data. The results of the a posteriori test reveal that the obtained solutions are stable for all NN models without applying any stabilization techniques. However, not all NN models have a reasonable performance when embedded in the LES code. The applicability of the NN models to the Burgers equation with higher and lower viscosity is also investigated, and it is indicated that the most reliable NN models obtained in this paper can be applied to a set of parameters which are different from those used in training. The results of the SGS models constructed using the neural network are also compared with the existing models, and it is shown that the best obtained NN models outperform the Smagorinsky model and the gradient model, and are comparable to the dynamic Smagorinsky model. However, the NN models have an advantage over the dynamic Smagorinsky model in numerical cost.</description><identifier>ISSN: 1070-6631</identifier><identifier>EISSN: 1089-7666</identifier><identifier>DOI: 10.1063/5.0087761</identifier><identifier>CODEN: PHFLE6</identifier><language>eng</language><publisher>Melville: American Institute of Physics</publisher><subject>Burgers equation ; Computer simulation ; Correlation coefficients ; Direct numerical simulation ; Fluid dynamics ; Large eddy simulation ; Mathematical models ; Neural networks ; Physics ; Subgrid scale models ; Training ; Velocity gradient</subject><ispartof>Physics of fluids (1994), 2022-04, Vol.34 (4)</ispartof><rights>Author(s)</rights><rights>2022 Author(s). Published under an exclusive license by AIP Publishing.</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c257t-becf0857206a7cece764f4ba37bed83cdf7aff03d1ea8cc8c9a0125ec6173fa83</citedby><cites>FETCH-LOGICAL-c257t-becf0857206a7cece764f4ba37bed83cdf7aff03d1ea8cc8c9a0125ec6173fa83</cites><orcidid>0000-0002-9220-1904 ; 0000-0002-1601-6416</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>314,776,780,790,4498,27901,27902</link.rule.ids></links><search><creatorcontrib>Tabe Jamaat, A. Golsa</creatorcontrib><title>Development of subgrid-scale model for LES of Burgers turbulence with large filter size</title><title>Physics of fluids (1994)</title><description>A neural network (NN) with one hidden layer is implemented to establish a relationship between the resolved-scale flow field and the subgrid-scale (SGS) stress for large eddy simulation (LES) of the Burgers equation. Five sets of input are considered for the neural network by combining the velocity gradient and the filter size. The training datasets are obtained by filtering the direct numerical simulation (DNS) results of the Burgers equation with random forcing function. The number of modes is sufficiently large (N = 65 536) to resolve extremely small scales of motion. In the a priori test, a correlation coefficient over 0.93 is achieved for the SGS stress between the NN models and the filtered DNS data. The results of the a posteriori test reveal that the obtained solutions are stable for all NN models without applying any stabilization techniques. However, not all NN models have a reasonable performance when embedded in the LES code. The applicability of the NN models to the Burgers equation with higher and lower viscosity is also investigated, and it is indicated that the most reliable NN models obtained in this paper can be applied to a set of parameters which are different from those used in training. The results of the SGS models constructed using the neural network are also compared with the existing models, and it is shown that the best obtained NN models outperform the Smagorinsky model and the gradient model, and are comparable to the dynamic Smagorinsky model. However, the NN models have an advantage over the dynamic Smagorinsky model in numerical cost.</description><subject>Burgers equation</subject><subject>Computer simulation</subject><subject>Correlation coefficients</subject><subject>Direct numerical simulation</subject><subject>Fluid dynamics</subject><subject>Large eddy simulation</subject><subject>Mathematical models</subject><subject>Neural networks</subject><subject>Physics</subject><subject>Subgrid scale models</subject><subject>Training</subject><subject>Velocity gradient</subject><issn>1070-6631</issn><issn>1089-7666</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><recordid>eNqd0E1LAzEQBuAgCtbqwX8Q8KSwNdk0H3vUWj-g4EHFY8hmJ3XLdrMm2Yr-ere04N3TDDMPM_AidE7JhBLBrvmEECWloAdoRIkqMimEONz2kmRCMHqMTmJcEUJYkYsRer-DDTS-W0ObsHc49uUy1FUWrWkAr30FDXY-4MX8Zbu-7cMSQsSpD2XfQGsBf9XpAzdmmGNXNwkCjvUPnKIjZ5oIZ_s6Rm_389fZY7Z4fnia3Swym3OZshKsI4rLnAgjLViQYuqmpWGyhEoxWzlpnCOsomCUtcoWhtCcgxVUMmcUG6OL3d0u-M8eYtIr34d2eKlzwfOiUJySQV3ulA0-xgBOd6Fem_CtKdHb3DTX-9wGe7Wz0dbJpNq3_8MbH_6g7irHfgETMHy-</recordid><startdate>202204</startdate><enddate>202204</enddate><creator>Tabe Jamaat, A. Golsa</creator><general>American Institute of Physics</general><scope>AAYXX</scope><scope>CITATION</scope><scope>8FD</scope><scope>H8D</scope><scope>L7M</scope><orcidid>https://orcid.org/0000-0002-9220-1904</orcidid><orcidid>https://orcid.org/0000-0002-1601-6416</orcidid></search><sort><creationdate>202204</creationdate><title>Development of subgrid-scale model for LES of Burgers turbulence with large filter size</title><author>Tabe Jamaat, A. Golsa</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c257t-becf0857206a7cece764f4ba37bed83cdf7aff03d1ea8cc8c9a0125ec6173fa83</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Burgers equation</topic><topic>Computer simulation</topic><topic>Correlation coefficients</topic><topic>Direct numerical simulation</topic><topic>Fluid dynamics</topic><topic>Large eddy simulation</topic><topic>Mathematical models</topic><topic>Neural networks</topic><topic>Physics</topic><topic>Subgrid scale models</topic><topic>Training</topic><topic>Velocity gradient</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Tabe Jamaat, A. Golsa</creatorcontrib><collection>CrossRef</collection><collection>Technology Research Database</collection><collection>Aerospace Database</collection><collection>Advanced Technologies Database with Aerospace</collection><jtitle>Physics of fluids (1994)</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Tabe Jamaat, A. Golsa</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Development of subgrid-scale model for LES of Burgers turbulence with large filter size</atitle><jtitle>Physics of fluids (1994)</jtitle><date>2022-04</date><risdate>2022</risdate><volume>34</volume><issue>4</issue><issn>1070-6631</issn><eissn>1089-7666</eissn><coden>PHFLE6</coden><abstract>A neural network (NN) with one hidden layer is implemented to establish a relationship between the resolved-scale flow field and the subgrid-scale (SGS) stress for large eddy simulation (LES) of the Burgers equation. Five sets of input are considered for the neural network by combining the velocity gradient and the filter size. The training datasets are obtained by filtering the direct numerical simulation (DNS) results of the Burgers equation with random forcing function. The number of modes is sufficiently large (N = 65 536) to resolve extremely small scales of motion. In the a priori test, a correlation coefficient over 0.93 is achieved for the SGS stress between the NN models and the filtered DNS data. The results of the a posteriori test reveal that the obtained solutions are stable for all NN models without applying any stabilization techniques. However, not all NN models have a reasonable performance when embedded in the LES code. The applicability of the NN models to the Burgers equation with higher and lower viscosity is also investigated, and it is indicated that the most reliable NN models obtained in this paper can be applied to a set of parameters which are different from those used in training. The results of the SGS models constructed using the neural network are also compared with the existing models, and it is shown that the best obtained NN models outperform the Smagorinsky model and the gradient model, and are comparable to the dynamic Smagorinsky model. However, the NN models have an advantage over the dynamic Smagorinsky model in numerical cost.</abstract><cop>Melville</cop><pub>American Institute of Physics</pub><doi>10.1063/5.0087761</doi><tpages>19</tpages><orcidid>https://orcid.org/0000-0002-9220-1904</orcidid><orcidid>https://orcid.org/0000-0002-1601-6416</orcidid></addata></record>
fulltext fulltext
identifier ISSN: 1070-6631
ispartof Physics of fluids (1994), 2022-04, Vol.34 (4)
issn 1070-6631
1089-7666
language eng
recordid cdi_scitation_primary_10_1063_5_0087761
source AIP Journals Complete; Alma/SFX Local Collection
subjects Burgers equation
Computer simulation
Correlation coefficients
Direct numerical simulation
Fluid dynamics
Large eddy simulation
Mathematical models
Neural networks
Physics
Subgrid scale models
Training
Velocity gradient
title Development of subgrid-scale model for LES of Burgers turbulence with large filter size
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-02T20%3A29%3A44IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_scita&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Development%20of%20subgrid-scale%20model%20for%20LES%20of%20Burgers%20turbulence%20with%20large%20filter%20size&rft.jtitle=Physics%20of%20fluids%20(1994)&rft.au=Tabe%20Jamaat,%20A.%20Golsa&rft.date=2022-04&rft.volume=34&rft.issue=4&rft.issn=1070-6631&rft.eissn=1089-7666&rft.coden=PHFLE6&rft_id=info:doi/10.1063/5.0087761&rft_dat=%3Cproquest_scita%3E2652998510%3C/proquest_scita%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2652998510&rft_id=info:pmid/&rfr_iscdi=true