Feature extraction of fields of fluid dynamics data using sparse convolutional autoencoder

A neural network technique that extracts underlying flow features from the original flow field data is newly proposed. The technique here is based on the convolutional and sparse autoencoder learning algorithms and is called sparse convolutional autoencoder. Unlike the typical convolutional neural n...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:AIP advances Jg. 11; H. 10; S. 105211 - 105211-6
Hauptverfasser: Obayashi, Wataru, Aono, Hikaru, Tatsukawa, Tomoaki, Fujii, Kozo
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Melville American Institute of Physics 01.10.2021
AIP Publishing LLC
Schlagworte:
ISSN:2158-3226, 2158-3226
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract A neural network technique that extracts underlying flow features from the original flow field data is newly proposed. The technique here is based on the convolutional and sparse autoencoder learning algorithms and is called sparse convolutional autoencoder. Unlike the typical convolutional neural network (CNN) that changes the size of the data itself in the intermediate layers, flow field data size is not changed in the learning process of this method and only the numbers of channels are changed in each layer. Different but the same size of the data as the input are obtained by convolution with multiple spatially overlapping flow field data under the assumption of sparsity. When data restoration is realized in this autoencoder system, the channel numbers of data in the intermediate layers turn out to contain different flow characteristics of the original flow field. The proposed method is applied to the low Reynolds number flows over a circular cylinder. The high-fidelity unsteady flow data obtained by solving two-dimensional compressible Navier–Stokes equations with a high-resolution numerical scheme are used as a test case. In the proposed method, sparsity introduced in the middle-hidden layer is essential for the successful separation of the original data. The results presented in the example seem to correspond to positive and negative magnitudes of the original data, but future studies will reveal other features of the method. The present method shows flow features different from those of proper orthogonal decomposition in each mode, which is probably due to nonlinear decomposition in the CNN process.
AbstractList A neural network technique that extracts underlying flow features from the original flow field data is newly proposed. The technique here is based on the convolutional and sparse autoencoder learning algorithms and is called sparse convolutional autoencoder. Unlike the typical convolutional neural network (CNN) that changes the size of the data itself in the intermediate layers, flow field data size is not changed in the learning process of this method and only the numbers of channels are changed in each layer. Different but the same size of the data as the input are obtained by convolution with multiple spatially overlapping flow field data under the assumption of sparsity. When data restoration is realized in this autoencoder system, the channel numbers of data in the intermediate layers turn out to contain different flow characteristics of the original flow field. The proposed method is applied to the low Reynolds number flows over a circular cylinder. The high-fidelity unsteady flow data obtained by solving two-dimensional compressible Navier–Stokes equations with a high-resolution numerical scheme are used as a test case. In the proposed method, sparsity introduced in the middle-hidden layer is essential for the successful separation of the original data. The results presented in the example seem to correspond to positive and negative magnitudes of the original data, but future studies will reveal other features of the method. The present method shows flow features different from those of proper orthogonal decomposition in each mode, which is probably due to nonlinear decomposition in the CNN process.
Author Aono, Hikaru
Fujii, Kozo
Obayashi, Wataru
Tatsukawa, Tomoaki
Author_xml – sequence: 1
  givenname: Wataru
  surname: Obayashi
  fullname: Obayashi, Wataru
  organization: Department of Information and Computer Technology, Tokyo University of Science
– sequence: 2
  givenname: Hikaru
  surname: Aono
  fullname: Aono, Hikaru
  organization: Department of Mechanical Engineering and Robotics, Shinshu University
– sequence: 3
  givenname: Tomoaki
  surname: Tatsukawa
  fullname: Tatsukawa, Tomoaki
  organization: Department of Information and Computer Technology, Tokyo University of Science
– sequence: 4
  givenname: Kozo
  surname: Fujii
  fullname: Fujii, Kozo
  organization: Department of Information and Computer Technology, Tokyo University of Science
BookMark eNp9kUFrFTEQx4NUsNYe_AYBTwqvTbKb7OYoxWqh4EUvXsK8zKTksd08k2yx3755fW0tKs5lhvD7_ycz85odzGkmxt5KcSKF6U71iRBGm254wQ6V1OOqU8ocPKtfseNSNqJFb6UY-0P245ygLpk4_aoZfI1p5inwEGnCcl9NS0SOtzNcR184QgW-lDhf8bKFXIj7NN-kadkpYeKw1ESzT0j5DXsZYCp0_JCP2PfzT9_Ovqwuv36-OPt4ufK97evKIxB5oA5J4yCCkEIhKRyNIW_VOsjBWm_WGmzQ3mpUViGgF8OgrFmL7ohd7H0xwcZtc7yGfOsSRHf_kPKVg1yjn8gN2KPvJBlqw_cqrKEziK2V6gbraWxe7_Ze25x-LlSq26Qlt8GKU3psXxsHLRv1fk_5nErJFJ66SuF2l3DaPVyisad_sD5W2G2r7TtO_1R82CvKI_lkf5Pyb9BtMfwP_tv5Dt3iqW8
CODEN AAIDBI
CitedBy_id crossref_primary_10_1016_j_eswa_2023_121924
crossref_primary_10_1007_s40436_025_00545_0
crossref_primary_10_1063_5_0244416
crossref_primary_10_1063_5_0119688
Cites_doi 10.1126/science.1127647
10.1016/0021-9991(92)90324-r
10.1017/jfm.2019.238
10.1063/1.4913500
10.1007/s42979-021-00867-3
10.1145/3065386
10.1063/1.4816650
10.2514/1.j056060
10.1017/s0022112010001217
10.2514/3.45890
10.1063/1.4935357
10.2514/1.11672
10.1063/1.5067313
10.1016/j.jcp.2013.12.019
10.2514/1.j058462
10.1017/jfm.2019.822
ContentType Journal Article
Copyright Author(s)
2021 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Copyright_xml – notice: Author(s)
– notice: 2021 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
DBID AJDQP
AAYXX
CITATION
8FD
H8D
L7M
DOA
DOI 10.1063/5.0065637
DatabaseName AIP Open Access Journals
CrossRef
Technology Research Database
Aerospace Database
Advanced Technologies Database with Aerospace
DOAJ Directory of Open Access Journals
DatabaseTitle CrossRef
Technology Research Database
Aerospace Database
Advanced Technologies Database with Aerospace
DatabaseTitleList
CrossRef
Technology Research Database

Database_xml – sequence: 1
  dbid: DOA
  name: DOAJ Directory of Open Access Journals
  url: https://www.doaj.org/
  sourceTypes: Open Website
DeliveryMethod fulltext_linktorsrc
Discipline Physics
EISSN 2158-3226
EndPage 105211-6
ExternalDocumentID oai_doaj_org_article_7d4dc31e6e08442fba36ddd862379ce8
10_1063_5_0065637
adv
GroupedDBID 5VS
61.
AAFWJ
ABFTF
ACGFO
ADBBV
ADCTM
AEGXH
AENEX
AFPKN
AGKCL
AGLKD
AHSDT
AIAGR
AJDQP
ALMA_UNASSIGNED_HOLDINGS
BCNDV
EBS
FRP
GROUPED_DOAJ
HH5
KQ8
M~E
OK1
RIP
RNS
RQS
AAYXX
ABJGX
ADMLS
AKSGC
CITATION
8FD
H8D
L7M
ID FETCH-LOGICAL-c494t-cdaeecae3de5d70f0102de2d866ec92bf1799c6b5a9f5c95d292dadc077296b03
IEDL.DBID DOA
ISICitedReferencesCount 13
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000754519200018&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2158-3226
IngestDate Fri Oct 03 12:45:55 EDT 2025
Sun Nov 09 08:39:09 EST 2025
Sat Nov 29 07:51:01 EST 2025
Tue Nov 18 22:40:11 EST 2025
Fri Jun 21 00:30:11 EDT 2024
Thu Jun 23 13:36:40 EDT 2022
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 10
Language English
License All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c494t-cdaeecae3de5d70f0102de2d866ec92bf1799c6b5a9f5c95d292dadc077296b03
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ORCID 0000-0002-2070-6575
0000-0002-0697-7897
0000-0003-1425-8371
0000-0003-2427-0833
OpenAccessLink https://doaj.org/article/7d4dc31e6e08442fba36ddd862379ce8
PQID 2580108751
PQPubID 2050671
PageCount 6
ParticipantIDs doaj_primary_oai_doaj_org_article_7d4dc31e6e08442fba36ddd862379ce8
scitation_primary_10_1063_5_0065637
proquest_journals_2580108751
crossref_primary_10_1063_5_0065637
crossref_citationtrail_10_1063_5_0065637
PublicationCentury 2000
PublicationDate 20211001
2021-10-01
PublicationDateYYYYMMDD 2021-10-01
PublicationDate_xml – month: 10
  year: 2021
  text: 20211001
  day: 01
PublicationDecade 2020
PublicationPlace Melville
PublicationPlace_xml – name: Melville
PublicationTitle AIP advances
PublicationYear 2021
Publisher American Institute of Physics
AIP Publishing LLC
Publisher_xml – name: American Institute of Physics
– name: AIP Publishing LLC
References Fujii, Obayashi (c14) 1989; 26
Krizhevsky, Sutskever, Hinton (c13) 2012; 60
Sato, Nonomura, Okada, Asada, Aono, Yakeno, Abe, Fujii (c17) 2015; 27
Omata, Shirayama (c12) 2019; 9
Kawai, Fujii (c15) 2005; 43
Lee, Kawai, Nonomura, Anyoji, Aono, Oyama, Asai, Fujii (c18) 2015; 27
Lele (c19) 1992; 103
Taira, Brunton, Dawson, Rowley, Colonius, McKeon, Schmidt, Gordeyev, Theofilis, Ukeiley (c3) 2017; 55
Abe, Nonomura, Iizuka, Fujii (c20) 2014; 260
Taira, Hemati, Brunton, Sun, Duraisamy, Bagheri, Dawson, Yeh (c4) 2020; 58
Hinton, Salakhutdinov (c5) 2006; 313
Fukami, Hasegawa, Nakamura, Morimoto, Fukagata (c11) 2021; 2
Fukami, Fukagata, Taira (c9) 2019; 870
Schmid (c2) 2010; 656
Murata, Fukami, Fukagata (c10) 2020; 882
Li, Nonomura, Fujii (c16) 2013; 25
(2024031516523375500_c15) 2005; 43
(2024031516523375500_c10) 2020; 882
(2024031516523375500_c18) 2015; 27
(2024031516523375500_c20) 2014; 260
(2024031516523375500_c9) 2019; 870
(2024031516523375500_c16) 2013; 25
(2024031516523375500_c8) 1995
(2024031516523375500_c4) 2020; 58
(2024031516523375500_c14) 1989; 26
(2024031516523375500_c19) 1992; 103
(2024031516523375500_c13) 2012; 60
(2024031516523375500_c21) 1998
(2024031516523375500_c1) 1967
(2024031516523375500_c5) 2006; 313
2024031516523375500_c6
(2024031516523375500_c2) 2010; 656
(2024031516523375500_c11) 2021; 2
(2024031516523375500_c17) 2015; 27
(2024031516523375500_c7) 2011
(2024031516523375500_c12) 2019; 9
(2024031516523375500_c3) 2017; 55
References_xml – volume: 27
  start-page: 117101
  year: 2015
  ident: c17
  article-title: Mechanisms for laminar separated-flow control using dielectric-barrier-discharge plasma actuator at low Reynolds number
  publication-title: Phys. Fluids
– volume: 103
  start-page: 16
  year: 1992
  ident: c19
  article-title: Compact finite difference schemes with spectral-like resolution
  publication-title: J. Comput. Phys.
– volume: 58
  start-page: 998
  year: 2020
  ident: c4
  article-title: Modal analysis of fluid flows: Applications and outlook
  publication-title: AIAA J.
– volume: 25
  start-page: 086101
  year: 2013
  ident: c16
  article-title: Mechanism of controlling supersonic cavity oscillations using upstream mass injection
  publication-title: Phys. Fluids
– volume: 55
  start-page: 4013
  year: 2017
  ident: c3
  article-title: Modal analysis of fluid flows: An overview
  publication-title: AIAA J.
– volume: 43
  start-page: 953
  year: 2005
  ident: c15
  article-title: Analysis and prediction of thin-airfoil stall phenomena with hybrid turbulence methodology
  publication-title: AIAA J.
– volume: 870
  start-page: 106
  year: 2019
  ident: c9
  article-title: Super-resolution reconstruction of turbulent flows with machine learning
  publication-title: J. Fluid Mech.
– volume: 26
  start-page: 1123
  year: 1989
  ident: c14
  article-title: High-resolution upwind scheme for vortical-flow simulations
  publication-title: J. Aircr.
– volume: 27
  start-page: 023602
  year: 2015
  ident: c18
  article-title: Mechanisms of surface pressure distribution within a laminar separation bubble at different Reynolds
  publication-title: Phys. Fluids
– volume: 2
  start-page: 467
  year: 2021
  ident: c11
  article-title: Model order reduction with neural networks: Application to laminar and turbulent flows
  publication-title: SN Comput. Sci.
– volume: 882
  start-page: A13
  year: 2020
  ident: c10
  article-title: Nonlinear mode decomposition with convolutional neural networks for fluid dynamics
  publication-title: J. Fluid Mech.
– volume: 9
  start-page: 015006
  year: 2019
  ident: c12
  article-title: A novel method of low-dimensional representation for temporal behavior of flow fields using deep autoencoder
  publication-title: AIP Adv.
– volume: 656
  start-page: 5
  year: 2010
  ident: c2
  article-title: Dynamic mode decomposition of numerical and experimental data
  publication-title: J. Fluid Mech.
– volume: 60
  start-page: 84
  year: 2012
  ident: c13
  article-title: ImageNet classification with deep convolutional neural networks
  publication-title: Adv. Neural Inf. Process. Syst.
– volume: 260
  start-page: 163
  year: 2014
  ident: c20
  article-title: Geometric interpretations and spatial symmetry property of metrics in the conservative form for high-order finite-difference schemes on moving and deforming grids
  publication-title: J. Comput. Phys.
– volume: 313
  start-page: 504
  year: 2006
  ident: c5
  article-title: Reducing the dimensionality of data with neural networks
  publication-title: Science
– volume: 313
  start-page: 504
  year: 2006
  ident: 2024031516523375500_c5
  article-title: Reducing the dimensionality of data with neural networks
  publication-title: Science
  doi: 10.1126/science.1127647
– volume: 103
  start-page: 16
  year: 1992
  ident: 2024031516523375500_c19
  article-title: Compact finite difference schemes with spectral-like resolution
  publication-title: J. Comput. Phys.
  doi: 10.1016/0021-9991(92)90324-r
– start-page: 1
  year: 1998
  ident: 2024031516523375500_c21
  article-title: Simple ideas for the accuracy and efficiency improvement of compressible flow simulation methods
– volume: 870
  start-page: 106
  year: 2019
  ident: 2024031516523375500_c9
  article-title: Super-resolution reconstruction of turbulent flows with machine learning
  publication-title: J. Fluid Mech.
  doi: 10.1017/jfm.2019.238
– start-page: 1
  volume-title: CS294A Lecture notes
  year: 2011
  ident: 2024031516523375500_c7
  article-title: Sparse autoencoder
– volume: 27
  start-page: 023602
  year: 2015
  ident: 2024031516523375500_c18
  article-title: Mechanisms of surface pressure distribution within a laminar separation bubble at different Reynolds
  publication-title: Phys. Fluids
  doi: 10.1063/1.4913500
– volume: 2
  start-page: 467
  year: 2021
  ident: 2024031516523375500_c11
  article-title: Model order reduction with neural networks: Application to laminar and turbulent flows
  publication-title: SN Comput. Sci.
  doi: 10.1007/s42979-021-00867-3
– volume: 60
  start-page: 84
  year: 2012
  ident: 2024031516523375500_c13
  article-title: ImageNet classification with deep convolutional neural networks
  publication-title: Adv. Neural Inf. Process. Syst.
  doi: 10.1145/3065386
– volume: 25
  start-page: 086101
  year: 2013
  ident: 2024031516523375500_c16
  article-title: Mechanism of controlling supersonic cavity oscillations using upstream mass injection
  publication-title: Phys. Fluids
  doi: 10.1063/1.4816650
– volume: 55
  start-page: 4013
  year: 2017
  ident: 2024031516523375500_c3
  article-title: Modal analysis of fluid flows: An overview
  publication-title: AIAA J.
  doi: 10.2514/1.j056060
– ident: 2024031516523375500_c6
– start-page: 166
  volume-title: Atmospheric Turbulence and Radio Wave Propagation
  year: 1967
  ident: 2024031516523375500_c1
  article-title: The structure of inhomogeneous turbulent flows
– volume: 656
  start-page: 5
  year: 2010
  ident: 2024031516523375500_c2
  article-title: Dynamic mode decomposition of numerical and experimental data
  publication-title: J. Fluid Mech.
  doi: 10.1017/s0022112010001217
– volume: 26
  start-page: 1123
  year: 1989
  ident: 2024031516523375500_c14
  article-title: High-resolution upwind scheme for vortical-flow simulations
  publication-title: J. Aircr.
  doi: 10.2514/3.45890
– volume: 27
  start-page: 117101
  year: 2015
  ident: 2024031516523375500_c17
  article-title: Mechanisms for laminar separated-flow control using dielectric-barrier-discharge plasma actuator at low Reynolds number
  publication-title: Phys. Fluids
  doi: 10.1063/1.4935357
– volume: 43
  start-page: 953
  year: 2005
  ident: 2024031516523375500_c15
  article-title: Analysis and prediction of thin-airfoil stall phenomena with hybrid turbulence methodology
  publication-title: AIAA J.
  doi: 10.2514/1.11672
– start-page: 1
  volume-title: The Handbook of Brain Theory and Neural Networks
  year: 1995
  ident: 2024031516523375500_c8
  article-title: Convolutional networks for images, speech, and time-series
– volume: 9
  start-page: 015006
  year: 2019
  ident: 2024031516523375500_c12
  article-title: A novel method of low-dimensional representation for temporal behavior of flow fields using deep autoencoder
  publication-title: AIP Adv.
  doi: 10.1063/1.5067313
– volume: 260
  start-page: 163
  year: 2014
  ident: 2024031516523375500_c20
  article-title: Geometric interpretations and spatial symmetry property of metrics in the conservative form for high-order finite-difference schemes on moving and deforming grids
  publication-title: J. Comput. Phys.
  doi: 10.1016/j.jcp.2013.12.019
– volume: 58
  start-page: 998
  year: 2020
  ident: 2024031516523375500_c4
  article-title: Modal analysis of fluid flows: Applications and outlook
  publication-title: AIAA J.
  doi: 10.2514/1.j058462
– volume: 882
  start-page: A13
  year: 2020
  ident: 2024031516523375500_c10
  article-title: Nonlinear mode decomposition with convolutional neural networks for fluid dynamics
  publication-title: J. Fluid Mech.
  doi: 10.1017/jfm.2019.822
SSID ssj0000491084
Score 2.3156192
Snippet A neural network technique that extracts underlying flow features from the original flow field data is newly proposed. The technique here is based on the...
SourceID doaj
proquest
crossref
scitation
SourceType Open Website
Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 105211
SubjectTerms Algorithms
Artificial neural networks
Circular cylinders
Compressibility
Feature extraction
Flow characteristics
Flow control
Fluid dynamics
Fluid flow
Low Reynolds number flow
Machine learning
Neural networks
Proper Orthogonal Decomposition
Reynolds number
Sparsity
Unsteady flow
Title Feature extraction of fields of fluid dynamics data using sparse convolutional autoencoder
URI http://dx.doi.org/10.1063/5.0065637
https://www.proquest.com/docview/2580108751
https://doaj.org/article/7d4dc31e6e08442fba36ddd862379ce8
Volume 11
WOSCitedRecordID wos000754519200018&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVAON
  databaseName: DOAJ Directory of Open Access Journals
  customDbUrl:
  eissn: 2158-3226
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000491084
  issn: 2158-3226
  databaseCode: DOA
  dateStart: 20110101
  isFulltext: true
  titleUrlDefault: https://www.doaj.org/
  providerName: Directory of Open Access Journals
– providerCode: PRVHPJ
  databaseName: ROAD: Directory of Open Access Scholarly Resources
  customDbUrl:
  eissn: 2158-3226
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0000491084
  issn: 2158-3226
  databaseCode: M~E
  dateStart: 20110101
  isFulltext: true
  titleUrlDefault: https://road.issn.org
  providerName: ISSN International Centre
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV3BatwwEBUltCSXkrQJ2SYNIu2hFxOtJMvWMS0JvXTJoYWlF2FpRiEh7AZ7N8d8ezSyd7uFkF56McYMSIzHmmf76T3GPmtT-5AaX6ED-EJbqAtfqVAoGcltO2pfNdlsoppM6unUXm1YfREnrJcH7hN3VoGGoMZoUNRay-gbZQAgAXFV2YB5m6-o7MbL1G2Pe8ci2w2nllankaVZyQoZdUafURKOIe_zjWaUNfv_AprbqQv1P8Q3es7lLns7gEV-3k9yj73C2Tv2JpM2Q_ee_Sb8tmyRpwW27Tco8HnkmZTW5bO75Q1w6E3nO05sUE5E92ue1pG2Q06c86H20jjNcjEnXUvAdp_9urz4-e17MXglFEFbvSgCNIihQQVYQiUiScUBypQng8FKH0n5LRhfNjaWwZYgrYQGgiB0bbxQB2xrNp_hIeO1MBATiiJ2qQYBvgZh69JGpI1BiCP2ZZU0t8oP-VncufxD2yhXuiG_I3a6Dr3v1TOeC_pKmV8HkOB1vpDKwA1l4P5VBiN2vLpvbngKOyfL1H9Jsn88Yp_W9_KlmTwT9TBv_0S4e4gf_sd8j9iOJHZMpgUes61Fu8SP7HV4WNx07Uku5nT88XjxBEsi-gQ
linkProvider Directory of Open Access Journals
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Feature+extraction+of+fields+of+fluid+dynamics+data+using+sparse+convolutional+autoencoder&rft.jtitle=AIP+advances&rft.au=Obayashi%2C+Wataru&rft.au=Aono%2C+Hikaru&rft.au=Tatsukawa%2C+Tomoaki&rft.au=Fujii%2C+Kozo&rft.date=2021-10-01&rft.issn=2158-3226&rft.eissn=2158-3226&rft.volume=11&rft.issue=10&rft_id=info:doi/10.1063%2F5.0065637&rft.externalDBID=n%2Fa&rft.externalDocID=10_1063_5_0065637
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2158-3226&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2158-3226&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2158-3226&client=summon