TOuNN: Topology Optimization using Neural Networks
Neural networks, and more broadly, machine learning techniques, have been recently exploited to accelerate topology optimization through data-driven training and image processing. In this paper, we demonstrate that one can directly execute topology optimization (TO) using neural networks (NN). The p...
Saved in:
| Published in | Structural and multidisciplinary optimization Vol. 63; no. 3; pp. 1135 - 1149 |
|---|---|
| Main Authors | , |
| Format | Journal Article |
| Language | English |
| Published |
Berlin/Heidelberg
Springer Berlin Heidelberg
01.03.2021
Springer Nature B.V |
| Subjects | |
| Online Access | Get full text |
| ISSN | 1615-147X 1615-1488 |
| DOI | 10.1007/s00158-020-02748-4 |
Cover
Loading…
| Abstract | Neural networks, and more broadly, machine learning techniques, have been recently exploited to
accelerate
topology optimization through data-driven training and image processing. In this paper, we demonstrate that one can
directly execute
topology optimization (TO) using neural networks (NN). The primary concept is to use the NN’s activation functions to represent the popular Solid Isotropic Material with Penalization (SIMP) density field.
In other words, the density function is parameterized by the weights and bias associated with the NN, and spanned by NN’s activation functions; the density representation is thus independent of the finite element mesh.
Then, by relying on the NN’s built-in backpropogation, and a conventional finite element solver, the density field is optimized. Methods to impose design and manufacturing constraints within the proposed framework are described and illustrated. A byproduct of representing the density field via activation functions is that it leads to a crisp and differentiable boundary. The proposed framework is simple to implement and is illustrated through 2D and 3D examples. Some of the unresolved challenges with the proposed framework are also summarized. |
|---|---|
| AbstractList | Neural networks, and more broadly, machine learning techniques, have been recently exploited to accelerate topology optimization through data-driven training and image processing. In this paper, we demonstrate that one can directly execute topology optimization (TO) using neural networks (NN). The primary concept is to use the NN’s activation functions to represent the popular Solid Isotropic Material with Penalization (SIMP) density field. In other words, the density function is parameterized by the weights and bias associated with the NN, and spanned by NN’s activation functions; the density representation is thus independent of the finite element mesh. Then, by relying on the NN’s built-in backpropogation, and a conventional finite element solver, the density field is optimized. Methods to impose design and manufacturing constraints within the proposed framework are described and illustrated. A byproduct of representing the density field via activation functions is that it leads to a crisp and differentiable boundary. The proposed framework is simple to implement and is illustrated through 2D and 3D examples. Some of the unresolved challenges with the proposed framework are also summarized. Neural networks, and more broadly, machine learning techniques, have been recently exploited to accelerate topology optimization through data-driven training and image processing. In this paper, we demonstrate that one can directly execute topology optimization (TO) using neural networks (NN). The primary concept is to use the NN’s activation functions to represent the popular Solid Isotropic Material with Penalization (SIMP) density field. In other words, the density function is parameterized by the weights and bias associated with the NN, and spanned by NN’s activation functions; the density representation is thus independent of the finite element mesh. Then, by relying on the NN’s built-in backpropogation, and a conventional finite element solver, the density field is optimized. Methods to impose design and manufacturing constraints within the proposed framework are described and illustrated. A byproduct of representing the density field via activation functions is that it leads to a crisp and differentiable boundary. The proposed framework is simple to implement and is illustrated through 2D and 3D examples. Some of the unresolved challenges with the proposed framework are also summarized. |
| Author | Suresh, Krishnan Chandrasekhar, Aaditya |
| Author_xml | – sequence: 1 givenname: Aaditya surname: Chandrasekhar fullname: Chandrasekhar, Aaditya organization: Department of Mechanical Engineering, University of Wisconsin-Madison – sequence: 2 givenname: Krishnan surname: Suresh fullname: Suresh, Krishnan email: ksuresh@wisc.edu organization: Department of Mechanical Engineering, University of Wisconsin-Madison |
| BookMark | eNp9kEtPwzAQhC0EEm3hD3CKxDmwThw_uKGKl1S1lyJxs1zHqVzSONiJUPn1uA0CiUMPq9nDfLujGaPTxjUGoSsMNxiA3QYAXPAUMojDCE_JCRphiosUE85Pf3f2do7GIWwAgAMRI5QtF_18fpcsXetqt94li7azW_ulOuuapA-2WSdz03tVR-k-nX8PF-isUnUwlz86Qa-PD8vpczpbPL1M72epzrHoUpoZbAQVQECxsiq4YrQkmnFGMlHhQq8MEUSsCBWYaFJyBcoAIyovKdWM5hN0PdxtvfvoTejkxvW-iS9lVkD8wXK8d_HBpb0LwZtKatsd0nde2VpikPuG5NCQjA3JQ0OSRDT7h7bebpXfHYfyAQrR3KyN_0t1hPoGU3x48w |
| CitedBy_id | crossref_primary_10_1016_j_matdes_2022_110885 crossref_primary_10_1016_j_cma_2021_114197 crossref_primary_10_1007_s00158_022_03170_8 crossref_primary_10_1007_s10845_024_02490_4 crossref_primary_10_32604_cmes_2023_027603 crossref_primary_10_1016_j_energy_2022_124053 crossref_primary_10_1364_AO_470494 crossref_primary_10_1109_TITS_2023_3271313 crossref_primary_10_1016_j_finel_2022_103893 crossref_primary_10_1007_s00158_022_03347_1 crossref_primary_10_1142_S0219876221420135 crossref_primary_10_1155_2021_5058791 crossref_primary_10_1016_j_compbiomed_2022_106475 crossref_primary_10_1002_adts_202400589 crossref_primary_10_1177_09544062251324422 crossref_primary_10_1007_s00366_022_01760_0 crossref_primary_10_1007_s00158_024_03835_6 crossref_primary_10_1016_j_cma_2024_117698 crossref_primary_10_1007_s00158_022_03433_4 crossref_primary_10_1007_s00158_023_03686_7 crossref_primary_10_1093_pnasnexus_pgae186 crossref_primary_10_1007_s00158_021_03050_7 crossref_primary_10_1007_s10845_025_02568_7 crossref_primary_10_1016_j_ast_2023_108836 crossref_primary_10_1115_1_4062663 crossref_primary_10_1016_j_engappai_2024_109991 crossref_primary_10_1016_j_aei_2024_102763 crossref_primary_10_1016_j_compstruc_2023_107232 crossref_primary_10_1007_s00158_024_03888_7 crossref_primary_10_1007_s00366_024_02077_w crossref_primary_10_1007_s00366_022_01636_3 crossref_primary_10_1109_ACCESS_2021_3125014 crossref_primary_10_1007_s00466_023_02358_z crossref_primary_10_1016_j_engappai_2024_107916 crossref_primary_10_1016_j_probengmech_2022_103356 crossref_primary_10_1016_j_cma_2023_116278 crossref_primary_10_1016_j_ast_2023_108405 crossref_primary_10_3390_a15070241 crossref_primary_10_1007_s00366_023_01904_w crossref_primary_10_1080_15376494_2023_2253018 crossref_primary_10_1016_j_apm_2022_02_036 crossref_primary_10_3390_app11052400 crossref_primary_10_1016_j_jcp_2024_113506 crossref_primary_10_1007_s11433_024_2576_7 crossref_primary_10_3390_a16080380 crossref_primary_10_1007_s13160_023_00563_0 crossref_primary_10_1016_j_cma_2023_116704 crossref_primary_10_1115_1_4055377 crossref_primary_10_1016_j_aei_2024_102869 crossref_primary_10_1002_aisy_202200333 crossref_primary_10_1007_s11081_023_09806_y crossref_primary_10_1007_s00158_022_03369_9 crossref_primary_10_1007_s00707_022_03449_3 crossref_primary_10_1007_s00366_023_01827_6 crossref_primary_10_1115_1_4062969 crossref_primary_10_1007_s00366_023_01846_3 crossref_primary_10_1016_j_cad_2023_103639 crossref_primary_10_1007_s00466_024_02565_2 crossref_primary_10_1016_j_cad_2023_103665 crossref_primary_10_1007_s10999_023_09676_3 crossref_primary_10_1007_s00521_024_10132_2 crossref_primary_10_1016_j_cad_2024_103707 crossref_primary_10_1007_s12008_024_01905_z crossref_primary_10_1007_s00158_024_03908_6 crossref_primary_10_1080_0305215X_2024_2340062 crossref_primary_10_1109_TCI_2023_3347916 crossref_primary_10_1016_j_cma_2023_116052 crossref_primary_10_1038_s41598_024_57137_4 crossref_primary_10_1007_s00158_024_03818_7 crossref_primary_10_3389_fenrg_2023_1294531 crossref_primary_10_2139_ssrn_4010395 crossref_primary_10_1080_0305215X_2021_1902998 crossref_primary_10_1016_j_engstruct_2024_119194 crossref_primary_10_1038_s41598_022_26312_w crossref_primary_10_1093_jcde_qwad072 crossref_primary_10_3390_math12152350 crossref_primary_10_1007_s00158_024_03851_6 crossref_primary_10_1002_nme_7428 crossref_primary_10_1007_s00466_023_02434_4 crossref_primary_10_1016_j_euromechsol_2022_104869 crossref_primary_10_1007_s00366_022_01736_0 crossref_primary_10_1115_1_4055505 crossref_primary_10_1098_rsta_2022_0405 crossref_primary_10_1016_j_engappai_2023_107033 crossref_primary_10_1016_j_tws_2024_112595 crossref_primary_10_1016_j_matdes_2022_110672 crossref_primary_10_1017_S0890060423000100 crossref_primary_10_3390_app11199041 crossref_primary_10_1002_advs_202405835 crossref_primary_10_1115_1_4062352 crossref_primary_10_1007_s10409_024_24207_x crossref_primary_10_1115_1_4064131 crossref_primary_10_1115_1_4067528 crossref_primary_10_1016_j_cma_2024_117004 crossref_primary_10_3390_buildings14123825 crossref_primary_10_1007_s00158_022_03461_0 crossref_primary_10_1111_mice_12933 crossref_primary_10_1007_s11831_024_10100_y crossref_primary_10_1155_2021_7171816 crossref_primary_10_1016_j_finel_2022_103904 crossref_primary_10_1016_j_procir_2023_02_143 crossref_primary_10_1080_01430750_2023_2284263 crossref_primary_10_1016_j_neunet_2024_106957 crossref_primary_10_1007_s00158_024_03877_w crossref_primary_10_1016_j_cma_2023_116108 crossref_primary_10_1080_0951192X_2025_2478007 crossref_primary_10_1002_nme_7221 crossref_primary_10_3390_app15052753 crossref_primary_10_1007_s00158_022_03181_5 crossref_primary_10_1007_s00158_024_03856_1 crossref_primary_10_1016_j_cma_2023_116401 crossref_primary_10_1016_j_engstruct_2023_117103 crossref_primary_10_1016_j_matdes_2024_113086 crossref_primary_10_1007_s00366_024_02010_1 crossref_primary_10_1007_s41060_024_00688_6 crossref_primary_10_32604_cmes_2024_048118 crossref_primary_10_1016_j_advengsoft_2024_103717 crossref_primary_10_1016_j_advengsoft_2022_103359 crossref_primary_10_1016_j_compstruc_2023_107218 crossref_primary_10_1080_0305215X_2021_1974015 |
| Cites_doi | 10.1002/nme.5432 10.1007/s00158-013-0978-6 10.1115/1.4041319 10.1016/j.advengsoft.2016.07.002 10.1038/323533a0 10.1007/BF02551274 10.1016/0045-7949(93)90035-C 10.1115/1.2901581 10.1115/1.4028591 10.1016/S0045-7825(02)00559-5 10.1007/BF01214002 10.1007/s00158-019-02346-z 10.1007/s00158-014-1107-x 10.1007/s00158-014-1188-6 10.1007/s00158-010-0594-7 10.1515/rnam-2019-0018 10.1115/1.4031088 10.1007/s00158-009-0443-8 10.1007/s00158-018-2101-5 10.1016/j.addma.2017.11.007 10.4324/9780203451519 10.1007/s00158-015-1277-1 10.1007/978-3-662-05086-6_2 10.1007/978-3-662-03115-5 10.1115/1.4029110 10.1016/j.icheatmasstransfer.2018.07.001 10.1007/s001580050176 10.1002/nme.1620240207 10.1007/s00158-012-0807-3 10.1080/21681163.2015.1030775 10.1007/BF01743693 |
| ContentType | Journal Article |
| Copyright | Springer-Verlag GmbH Germany, part of Springer Nature 2020 Springer-Verlag GmbH Germany, part of Springer Nature 2020. |
| Copyright_xml | – notice: Springer-Verlag GmbH Germany, part of Springer Nature 2020 – notice: Springer-Verlag GmbH Germany, part of Springer Nature 2020. |
| DBID | AAYXX CITATION 8FE 8FG ABJCF AFKRA BENPR BGLVJ CCPQU DWQXO HCIFZ L6V M7S PHGZM PHGZT PKEHL PQEST PQGLB PQQKQ PQUKI PRINS PTHSS |
| DOI | 10.1007/s00158-020-02748-4 |
| DatabaseName | CrossRef ProQuest SciTech Collection ProQuest Technology Collection Materials Science & Engineering Collection ProQuest Central UK/Ireland ProQuest Central Technology Collection ProQuest One ProQuest Central Korea SciTech Premium Collection ProQuest Engineering Collection Engineering Database ProQuest Central Premium ProQuest One Academic (New) ProQuest One Academic Middle East (New) ProQuest One Academic Eastern Edition (DO NOT USE) ProQuest One Applied & Life Sciences ProQuest One Academic ProQuest One Academic UKI Edition ProQuest Central China Engineering collection |
| DatabaseTitle | CrossRef Engineering Database Technology Collection ProQuest One Academic Middle East (New) ProQuest One Academic Eastern Edition SciTech Premium Collection ProQuest One Community College ProQuest Technology Collection ProQuest SciTech Collection ProQuest Central China ProQuest Central ProQuest One Applied & Life Sciences ProQuest Engineering Collection ProQuest One Academic UKI Edition ProQuest Central Korea Materials Science & Engineering Collection ProQuest One Academic ProQuest Central (New) Engineering Collection ProQuest One Academic (New) |
| DatabaseTitleList | Engineering Database |
| Database_xml | – sequence: 1 dbid: 8FG name: ProQuest Technology Collection url: https://search.proquest.com/technologycollection1 sourceTypes: Aggregation Database |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Engineering |
| EISSN | 1615-1488 |
| EndPage | 1149 |
| ExternalDocumentID | 10_1007_s00158_020_02748_4 |
| GroupedDBID | -5B -5G -BR -EM -Y2 -~C .86 .VR 06D 0R~ 0VY 123 199 1N0 2.D 203 29Q 29~ 2J2 2JN 2JY 2KG 2KM 2LR 2P1 2VQ 2~H 30V 4.4 406 408 409 40D 40E 5QI 5VS 67Z 6NX 78A 8FE 8FG 8TC 8UJ 95- 95. 95~ 96X AAAVM AABHQ AACDK AAHNG AAIAL AAJBT AAJKR AANZL AARHV AARTL AASML AATNV AATVU AAUYE AAWCG AAYIU AAYQN AAYTO AAYZH ABAKF ABBBX ABBXA ABDZT ABECU ABFTD ABFTV ABHLI ABHQN ABJCF ABJNI ABJOX ABKCH ABKTR ABMNI ABMQK ABNWP ABQBU ABQSL ABSXP ABTEG ABTHY ABTKH ABTMW ABULA ABWNU ABXPI ACAOD ACBXY ACDTI ACGFS ACHSB ACHXU ACIWK ACKNC ACMDZ ACMLO ACOKC ACOMO ACPIV ACSNA ACZOJ ADHIR ADINQ ADKNI ADKPE ADPHR ADRFC ADTPH ADURQ ADYFF ADZKW AEBTG AEFIE AEFQL AEGAL AEGNC AEJHL AEJRE AEKMD AEMSY AENEX AEOHA AEPYU AESKC AETLH AEVLU AEXYK AFBBN AFEXP AFGCZ AFKRA AFLOW AFQWF AFWTZ AFZKB AGAYW AGDGC AGGDS AGJBK AGMZJ AGQEE AGQMX AGRTI AGWIL AGWZB AGYKE AHAVH AHBYD AHSBF AHYZX AIAKS AIGIU AIIXL AILAN AITGF AJBLW AJRNO AJZVZ ALMA_UNASSIGNED_HOLDINGS ALWAN AMKLP AMXSW AMYLF AOCGG ARCEE ARMRJ ASPBG AVWKF AXYYD AYJHY AZFZN B-. BA0 BDATZ BENPR BGLVJ BGNMA BSONS CAG CCPQU COF CS3 CSCUP DDRTE DL5 DNIVK DPUIP EBLON EBS EIOEI EJD ESBYG FEDTE FERAY FFXSO FIGPU FINBP FNLPD FRRFC FSGXE FWDCC GGCAI GGRSB GJIRD GNWQR GQ6 GQ7 GQ8 GXS H13 HCIFZ HF~ HG5 HG6 HMJXF HQYDN HRMNR HVGLF HZ~ I09 IHE IJ- IKXTQ IWAJR IXC IXD IXE IZIGR IZQ I~X I~Z J-C J0Z JBSCW JCJTX JZLTJ KDC KOV L6V LAS LLZTM M4Y M7S MA- N2Q N9A NB0 NDZJH NPVJJ NQJWS NU0 O9- O93 O9G O9I O9J OAM P19 P9P PF0 PT4 PT5 PTHSS QOK QOS R89 R9I RHV RIG RNI RNS ROL RPX RSV RZK S16 S1Z S26 S27 S28 S3B SAP SCLPG SDH SDM SEG SHX SISQX SJYHP SNE SNPRN SNX SOHCF SOJ SPISZ SRMVM SSLCW STPWE SZN T13 T16 TSG TSK TSV TUC U2A UG4 UOJIU UTJUX UZXMN VC2 VFIZW W23 W48 WK8 YLTOR Z45 Z5O Z7R Z7S Z7V Z7X Z7Y Z7Z Z81 Z83 Z85 Z86 Z88 Z8M Z8N Z8P Z8R Z8S Z8T Z8U Z8W Z8Z Z92 ZMTXR _50 ~02 AAPKM AAYXX ABBRH ABDBE ABFSG ACSTC ADHKG AEZWR AFDZB AFHIU AFOHR AGQPQ AHPBZ AHWEU AIXLP ATHPR AYFIA CITATION PHGZM PHGZT ABRTQ DWQXO PKEHL PQEST PQGLB PQQKQ PQUKI PRINS |
| ID | FETCH-LOGICAL-c319t-62e1e969040a7df58a76d4c787429f15cbe4949b46914c4d8a0ae074a3d66c763 |
| IEDL.DBID | U2A |
| ISSN | 1615-147X |
| IngestDate | Fri Jul 25 11:16:12 EDT 2025 Tue Jul 01 01:31:45 EDT 2025 Thu Apr 24 23:08:40 EDT 2025 Fri Feb 21 02:49:51 EST 2025 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Issue | 3 |
| Keywords | Topology optimization Neural networks Machine learning |
| Language | English |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-c319t-62e1e969040a7df58a76d4c787429f15cbe4949b46914c4d8a0ae074a3d66c763 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 |
| PQID | 2503197316 |
| PQPubID | 2043658 |
| PageCount | 15 |
| ParticipantIDs | proquest_journals_2503197316 crossref_citationtrail_10_1007_s00158_020_02748_4 crossref_primary_10_1007_s00158_020_02748_4 springer_journals_10_1007_s00158_020_02748_4 |
| ProviderPackageCode | CITATION AAYXX |
| PublicationCentury | 2000 |
| PublicationDate | 20210300 2021-03-00 20210301 |
| PublicationDateYYYYMMDD | 2021-03-01 |
| PublicationDate_xml | – month: 3 year: 2021 text: 20210300 |
| PublicationDecade | 2020 |
| PublicationPlace | Berlin/Heidelberg |
| PublicationPlace_xml | – name: Berlin/Heidelberg – name: Heidelberg |
| PublicationTitle | Structural and multidisciplinary optimization |
| PublicationTitleAbbrev | Struct Multidisc Optim |
| PublicationYear | 2021 |
| Publisher | Springer Berlin Heidelberg Springer Nature B.V |
| Publisher_xml | – name: Springer Berlin Heidelberg – name: Springer Nature B.V |
| References | AndreassenEEfficient topology optimization in MATLAB using 88 lines of codeStruct Multidiscipl Optim201143111610.1007/s00158-010-0594-7 Glorot X, Bengio Y (2010) Understanding the difficulty of training deep feedforward neural networks. In: Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, pp 249–256 Paszke A, et al (2019) Pytorch: An imperative style, high-performance deep learning library. In: Advances in neural information processing systems, pp 8026–8037 Bendsøe MP, Sigmund O (1995) Optimization of structural topology, shape, and material, volume 414 Springer GroenJPHigher-order multi-resolution topology optimization using the finite cell methodInt J Numer Methods Eng201711010903920367045410.1002/nme.5432 Lei X, et al (2019) Machine learning-driven real-time topology optimization under moving morphable component-based framework, journal of applied mechanics, 86, 1 UluEA data-driven investigation and estimation of optimal topologies under variable loading configurationsComput Methods Biomech Biomed Eng Imaging Vis2016426172349872910.1080/21681163.2015.1030775 VatanabeSLTopology optimization with manufacturing constraints: a unified projection-based approachAdv Eng Softw20161009711210.1016/j.advengsoft.2016.07.002 Goodfellow I, et al (2016) Deep learning MIT Press Bendsoe MP, Sigmund O (2003) Topology optimization: theory, methods, and applications, Springer Berlin Heidelberg 2 edition SigmundOA 99 line topology optimization code written in matlabStruct Multidiscipl Optim200121212012710.1007/s001580050176 SvanbergKThe method of moving asymptotes—a new method for structural optimizationInt J Numer Methods Eng198724235937387530710.1002/nme.1620240207 MirzendehdelAMSureshKA deflated assembly free approach to large-scale implicit structural dynamicsJ Comput Nonlinear Dyn201510606101510.1115/1.4029110 NguyenTHA computational paradigm for multiresolution topology optimization (MTOP)Struct Multidiscipl Optim2010414525539260147110.1007/s00158-009-0443-8 Pascanu R, et al (2012) Understanding the exploding gradient problem. arXiv:1211.5063 Sigmund O, Maute K (2013) Topology optimization approaches: a comparative review. Dec CybenkoGApproximation by superpositions of a sigmoidal functionMath Control Signals, Syst198924303314101567010.1007/BF02551274 BaydinAGAutomatic differentiation in machine learning: a surveyJ Mach Learn Res2017181559556373800512 Kingma DP , Ba JL (2015) Adam: a method for stochastic optimization. In: 3rd International Conference on Learning Representations, ICLR 2015 - Conference Track Proceedings. International Conference on Learning Representations, ICLR YuYDeep learning for determining a near-optimal topological design without any iterationStruct Multidiscipl Optim201959378779910.1007/s00158-018-2101-5 Smith LN (2018) A disciplined approach to neural network hyper-parameters: part 1–learning rate, batch size, momentum, and weight decay. arXiv:1803.09820 SosnovikIOseledetsINeural networks for topology optimizationRuss J Numer Anal Math Model2019344215223398503310.1515/rnam-2019-0018 Nocedal J, Wright S (2006) Numerical optimization Springer Science & Business Media Rojas-LabandaSStolpeMAutomatic penalty continuation in structural topology optimizationStruct Multidiscipl Optim201552612051221343332910.1007/s00158-015-1277-1 YadavPSureshKLarge scale finite element analysis via assembly-free deflated conjugate gradientJ Comput Inf Sci Eng201414404100810.1115/1.4028591 SigmundOPeterssonJNumerical instabilities in topology optimization: a survey on procedures dealing with checkerboards, mesh-dependencies and local minimaStruct Optim1998161687510.1007/BF01214002 Abadi M, et al (2016) Tensorflow: large-scale machine learning on heterogeneous distributed systems. arXiv:1603.04467 DíazASigmundOCheckerboard patterns in layout optimizationStruct Optim1995101404510.1007/BF01743693 BendsoeMPAn analytical model to predict optimal material properties in the context of optimal structural designJ Appl Mech2008614930132748310.1115/1.2901581 ChandrasekharABuild optimization of fiber-reinforced additively manufactured componentsStruct Multidiscipl Optim2020611779010.1007/s00158-019-02346-z Ioffe S, Szegedy C (2015) Batch normalization: Accelerating deep network training by reducing internal covariate shift. preprint arXiv:1502.03167 SontagEDVc dimension of neural networksNATO ASI Ser F Comput S Sci199816869960928.68091 DengSSureshKMulti-constrained topology optimization via the topological sensitivityStruct Multidiscipl Optim20155159871001335385610.1007/s00158-014-1188-6 MirzendehdelAMSureshKA pareto-optimal approach to multimaterial topology optimizationJ Mech Des20151371010170110.1115/1.4031088 Márquez-Neila P, et al (2017) Imposing hard constraints on deep networks: promises and limitations. arXiv:1706.02025 Arora R, et al (2016) Understanding deep neural networks with rectified linear units. arXiv:1611.01491 Mirzendehdel AM, et al (2018) Strength-based topology optimization for anisotropic parts, vol 19 Bishop CM (2006) Pattern recognition and machine learning Springer XieYMStevenGPA simple evolutionary procedure for structural optimizationComput Struct199349588589610.1016/0045-7949(93)90035-C LiuKTovarAAn efficient 3D topology optimization code written in MatlabStruct Multidiscipl Optim201450611751196331242510.1007/s00158-014-1107-x WangMYA level set method for structural topology optimizationComput Methods in Appl Mech Eng20031921-2227246195140810.1016/S0045-7825(02)00559-5 Ruder S (2016) An overview of gradient descent optimization algorithms. arXiv:1609.04747 Hoyer S, et al (2019) Neural reparameterization improves structural optimization. arXiv:1909.04240 Banga S, et al (2018) 3D topology optimization using convolutional neural networks. arXiv:1808.07440 Gurney K (1997) An introduction to neural networks CRC press RumelhartDELearning representations by back-propagating errorsNature1986323608853353610.1038/323533a0 Lu L, et al (2019) Dying relu and initialization:, Theory and numerical examples. arXiv:1903.06733 Zhang Y, et al (2019) A deep convolutional neural network for topology optimization with strong generalization ability. arXiv:1901.07761 SureshKEfficient generation of large-scale pareto-optimal topologiesStruct Multidiscipl Optim20134714961303892710.1007/s00158-012-0807-3 Bendsoe MP, Sigmund O (2013) Topology optimization: theory, methods, and applications Springer Science & Business Media Kervadec H, et al (2019) Constrained deep networks: Lagrangian optimization via log-barrier extensions. arXiv:1904.04205 Nie Z, et al (2020) Topology GAN: topology optimization using generative adversarial networks based on physical fields over the initial domain. arXiv:2003.04685 Lin Q, et al (2018) Investigation into the topology optimization for conductive heat transfer based on deep learning approach, vol 97 2748_CR3 2748_CR4 2748_CR1 E Ulu (2748_CR47) 2016; 4 2748_CR25 2748_CR26 2748_CR27 2748_CR21 2748_CR22 2748_CR23 2748_CR20 K Suresh (2748_CR45) 2013; 47 K Svanberg (2748_CR46) 1987; 24 2748_CR7 2748_CR8 2748_CR6 2748_CR18 2748_CR19 Y Yu (2748_CR52) 2019; 59 MY Wang (2748_CR49) 2003; 192 2748_CR15 2748_CR16 2748_CR10 YM Xie (2748_CR50) 1993; 49 2748_CR53 O Sigmund (2748_CR39) 2001; 21 MP Bendsoe (2748_CR9) 2008; 61 S Deng (2748_CR13) 2015; 51 K Liu (2748_CR24) 2014; 50 AG Baydin (2748_CR5) 2017; 18 A Chandrasekhar (2748_CR11) 2020; 61 DE Rumelhart (2748_CR38) 1986; 323 P Yadav (2748_CR51) 2014; 14 AM Mirzendehdel (2748_CR28) 2015; 137 S Rojas-Labanda (2748_CR36) 2015; 52 2748_CR41 2748_CR42 G Cybenko (2748_CR12) 1989; 2 TH Nguyen (2748_CR32) 2010; 41 ED Sontag (2748_CR43) 1998; 168 E Andreassen (2748_CR2) 2011; 43 2748_CR37 I Sosnovik (2748_CR44) 2019; 34 A Díaz (2748_CR14) 1995; 10 2748_CR33 2748_CR34 O Sigmund (2748_CR40) 1998; 16 2748_CR35 2748_CR30 SL Vatanabe (2748_CR48) 2016; 100 2748_CR31 JP Groen (2748_CR17) 2017; 110 AM Mirzendehdel (2748_CR29) 2015; 10 |
| References_xml | – reference: Glorot X, Bengio Y (2010) Understanding the difficulty of training deep feedforward neural networks. In: Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, pp 249–256 – reference: SureshKEfficient generation of large-scale pareto-optimal topologiesStruct Multidiscipl Optim20134714961303892710.1007/s00158-012-0807-3 – reference: Márquez-Neila P, et al (2017) Imposing hard constraints on deep networks: promises and limitations. arXiv:1706.02025 – reference: MirzendehdelAMSureshKA pareto-optimal approach to multimaterial topology optimizationJ Mech Des20151371010170110.1115/1.4031088 – reference: Gurney K (1997) An introduction to neural networks CRC press – reference: SigmundOPeterssonJNumerical instabilities in topology optimization: a survey on procedures dealing with checkerboards, mesh-dependencies and local minimaStruct Optim1998161687510.1007/BF01214002 – reference: Rojas-LabandaSStolpeMAutomatic penalty continuation in structural topology optimizationStruct Multidiscipl Optim201552612051221343332910.1007/s00158-015-1277-1 – reference: VatanabeSLTopology optimization with manufacturing constraints: a unified projection-based approachAdv Eng Softw20161009711210.1016/j.advengsoft.2016.07.002 – reference: Arora R, et al (2016) Understanding deep neural networks with rectified linear units. arXiv:1611.01491 – reference: Abadi M, et al (2016) Tensorflow: large-scale machine learning on heterogeneous distributed systems. arXiv:1603.04467 – reference: Goodfellow I, et al (2016) Deep learning MIT Press – reference: WangMYA level set method for structural topology optimizationComput Methods in Appl Mech Eng20031921-2227246195140810.1016/S0045-7825(02)00559-5 – reference: SvanbergKThe method of moving asymptotes—a new method for structural optimizationInt J Numer Methods Eng198724235937387530710.1002/nme.1620240207 – reference: Banga S, et al (2018) 3D topology optimization using convolutional neural networks. arXiv:1808.07440 – reference: Lin Q, et al (2018) Investigation into the topology optimization for conductive heat transfer based on deep learning approach, vol 97 – reference: Paszke A, et al (2019) Pytorch: An imperative style, high-performance deep learning library. In: Advances in neural information processing systems, pp 8026–8037 – reference: Bishop CM (2006) Pattern recognition and machine learning Springer – reference: DíazASigmundOCheckerboard patterns in layout optimizationStruct Optim1995101404510.1007/BF01743693 – reference: Bendsoe MP, Sigmund O (2003) Topology optimization: theory, methods, and applications, Springer Berlin Heidelberg 2 edition – reference: LiuKTovarAAn efficient 3D topology optimization code written in MatlabStruct Multidiscipl Optim201450611751196331242510.1007/s00158-014-1107-x – reference: Bendsøe MP, Sigmund O (1995) Optimization of structural topology, shape, and material, volume 414 Springer – reference: Nocedal J, Wright S (2006) Numerical optimization Springer Science & Business Media – reference: YadavPSureshKLarge scale finite element analysis via assembly-free deflated conjugate gradientJ Comput Inf Sci Eng201414404100810.1115/1.4028591 – reference: AndreassenEEfficient topology optimization in MATLAB using 88 lines of codeStruct Multidiscipl Optim201143111610.1007/s00158-010-0594-7 – reference: Pascanu R, et al (2012) Understanding the exploding gradient problem. arXiv:1211.5063 – reference: XieYMStevenGPA simple evolutionary procedure for structural optimizationComput Struct199349588589610.1016/0045-7949(93)90035-C – reference: SosnovikIOseledetsINeural networks for topology optimizationRuss J Numer Anal Math Model2019344215223398503310.1515/rnam-2019-0018 – reference: Mirzendehdel AM, et al (2018) Strength-based topology optimization for anisotropic parts, vol 19 – reference: YuYDeep learning for determining a near-optimal topological design without any iterationStruct Multidiscipl Optim201959378779910.1007/s00158-018-2101-5 – reference: Ioffe S, Szegedy C (2015) Batch normalization: Accelerating deep network training by reducing internal covariate shift. preprint arXiv:1502.03167 – reference: SigmundOA 99 line topology optimization code written in matlabStruct Multidiscipl Optim200121212012710.1007/s001580050176 – reference: Bendsoe MP, Sigmund O (2013) Topology optimization: theory, methods, and applications Springer Science & Business Media – reference: Ruder S (2016) An overview of gradient descent optimization algorithms. arXiv:1609.04747 – reference: BendsoeMPAn analytical model to predict optimal material properties in the context of optimal structural designJ Appl Mech2008614930132748310.1115/1.2901581 – reference: Zhang Y, et al (2019) A deep convolutional neural network for topology optimization with strong generalization ability. arXiv:1901.07761 – reference: Kingma DP , Ba JL (2015) Adam: a method for stochastic optimization. In: 3rd International Conference on Learning Representations, ICLR 2015 - Conference Track Proceedings. International Conference on Learning Representations, ICLR – reference: CybenkoGApproximation by superpositions of a sigmoidal functionMath Control Signals, Syst198924303314101567010.1007/BF02551274 – reference: NguyenTHA computational paradigm for multiresolution topology optimization (MTOP)Struct Multidiscipl Optim2010414525539260147110.1007/s00158-009-0443-8 – reference: SontagEDVc dimension of neural networksNATO ASI Ser F Comput S Sci199816869960928.68091 – reference: MirzendehdelAMSureshKA deflated assembly free approach to large-scale implicit structural dynamicsJ Comput Nonlinear Dyn201510606101510.1115/1.4029110 – reference: Hoyer S, et al (2019) Neural reparameterization improves structural optimization. arXiv:1909.04240 – reference: GroenJPHigher-order multi-resolution topology optimization using the finite cell methodInt J Numer Methods Eng201711010903920367045410.1002/nme.5432 – reference: Nie Z, et al (2020) Topology GAN: topology optimization using generative adversarial networks based on physical fields over the initial domain. arXiv:2003.04685 – reference: Lu L, et al (2019) Dying relu and initialization:, Theory and numerical examples. arXiv:1903.06733 – reference: Smith LN (2018) A disciplined approach to neural network hyper-parameters: part 1–learning rate, batch size, momentum, and weight decay. arXiv:1803.09820 – reference: RumelhartDELearning representations by back-propagating errorsNature1986323608853353610.1038/323533a0 – reference: BaydinAGAutomatic differentiation in machine learning: a surveyJ Mach Learn Res2017181559556373800512 – reference: ChandrasekharABuild optimization of fiber-reinforced additively manufactured componentsStruct Multidiscipl Optim2020611779010.1007/s00158-019-02346-z – reference: Sigmund O, Maute K (2013) Topology optimization approaches: a comparative review. Dec – reference: Lei X, et al (2019) Machine learning-driven real-time topology optimization under moving morphable component-based framework, journal of applied mechanics, 86, 1 – reference: UluEA data-driven investigation and estimation of optimal topologies under variable loading configurationsComput Methods Biomech Biomed Eng Imaging Vis2016426172349872910.1080/21681163.2015.1030775 – reference: Kervadec H, et al (2019) Constrained deep networks: Lagrangian optimization via log-barrier extensions. arXiv:1904.04205 – reference: DengSSureshKMulti-constrained topology optimization via the topological sensitivityStruct Multidiscipl Optim20155159871001335385610.1007/s00158-014-1188-6 – ident: 2748_CR1 – volume: 110 start-page: 903 issue: 10 year: 2017 ident: 2748_CR17 publication-title: Int J Numer Methods Eng doi: 10.1002/nme.5432 – ident: 2748_CR41 doi: 10.1007/s00158-013-0978-6 – ident: 2748_CR23 doi: 10.1115/1.4041319 – ident: 2748_CR4 – ident: 2748_CR8 – ident: 2748_CR42 – ident: 2748_CR27 – volume: 100 start-page: 97 year: 2016 ident: 2748_CR48 publication-title: Adv Eng Softw doi: 10.1016/j.advengsoft.2016.07.002 – volume: 323 start-page: 533 issue: 6088 year: 1986 ident: 2748_CR38 publication-title: Nature doi: 10.1038/323533a0 – ident: 2748_CR22 – volume: 2 start-page: 303 issue: 4 year: 1989 ident: 2748_CR12 publication-title: Math Control Signals, Syst doi: 10.1007/BF02551274 – volume: 49 start-page: 885 issue: 5 year: 1993 ident: 2748_CR50 publication-title: Comput Struct doi: 10.1016/0045-7949(93)90035-C – volume: 61 start-page: 930 issue: 4 year: 2008 ident: 2748_CR9 publication-title: J Appl Mech doi: 10.1115/1.2901581 – volume: 14 start-page: 041008 issue: 4 year: 2014 ident: 2748_CR51 publication-title: J Comput Inf Sci Eng doi: 10.1115/1.4028591 – volume: 192 start-page: 227 issue: 1-2 year: 2003 ident: 2748_CR49 publication-title: Comput Methods in Appl Mech Eng doi: 10.1016/S0045-7825(02)00559-5 – ident: 2748_CR16 – ident: 2748_CR53 – ident: 2748_CR37 – ident: 2748_CR33 – volume: 16 start-page: 68 issue: 1 year: 1998 ident: 2748_CR40 publication-title: Struct Optim doi: 10.1007/BF01214002 – volume: 61 start-page: 77 issue: 1 year: 2020 ident: 2748_CR11 publication-title: Struct Multidiscipl Optim doi: 10.1007/s00158-019-02346-z – volume: 50 start-page: 1175 issue: 6 year: 2014 ident: 2748_CR24 publication-title: Struct Multidiscipl Optim doi: 10.1007/s00158-014-1107-x – volume: 51 start-page: 987 issue: 5 year: 2015 ident: 2748_CR13 publication-title: Struct Multidiscipl Optim doi: 10.1007/s00158-014-1188-6 – ident: 2748_CR26 – ident: 2748_CR3 – volume: 43 start-page: 1 issue: 1 year: 2011 ident: 2748_CR2 publication-title: Struct Multidiscipl Optim doi: 10.1007/s00158-010-0594-7 – volume: 34 start-page: 215 issue: 4 year: 2019 ident: 2748_CR44 publication-title: Russ J Numer Anal Math Model doi: 10.1515/rnam-2019-0018 – ident: 2748_CR19 – ident: 2748_CR21 – ident: 2748_CR15 – ident: 2748_CR34 – volume: 137 start-page: 101701 issue: 10 year: 2015 ident: 2748_CR28 publication-title: J Mech Des doi: 10.1115/1.4031088 – volume: 41 start-page: 525 issue: 4 year: 2010 ident: 2748_CR32 publication-title: Struct Multidiscipl Optim doi: 10.1007/s00158-009-0443-8 – volume: 59 start-page: 787 issue: 3 year: 2019 ident: 2748_CR52 publication-title: Struct Multidiscipl Optim doi: 10.1007/s00158-018-2101-5 – ident: 2748_CR30 doi: 10.1016/j.addma.2017.11.007 – ident: 2748_CR18 doi: 10.4324/9780203451519 – ident: 2748_CR20 – volume: 52 start-page: 1205 issue: 6 year: 2015 ident: 2748_CR36 publication-title: Struct Multidiscipl Optim doi: 10.1007/s00158-015-1277-1 – ident: 2748_CR7 doi: 10.1007/978-3-662-05086-6_2 – volume: 18 start-page: 5595 issue: 1 year: 2017 ident: 2748_CR5 publication-title: J Mach Learn Res – volume: 168 start-page: 69 year: 1998 ident: 2748_CR43 publication-title: NATO ASI Ser F Comput S Sci – ident: 2748_CR35 – ident: 2748_CR6 doi: 10.1007/978-3-662-03115-5 – volume: 10 start-page: 061015 issue: 6 year: 2015 ident: 2748_CR29 publication-title: J Comput Nonlinear Dyn doi: 10.1115/1.4029110 – ident: 2748_CR31 – ident: 2748_CR25 doi: 10.1016/j.icheatmasstransfer.2018.07.001 – ident: 2748_CR10 – volume: 21 start-page: 120 issue: 2 year: 2001 ident: 2748_CR39 publication-title: Struct Multidiscipl Optim doi: 10.1007/s001580050176 – volume: 24 start-page: 359 issue: 2 year: 1987 ident: 2748_CR46 publication-title: Int J Numer Methods Eng doi: 10.1002/nme.1620240207 – volume: 47 start-page: 49 issue: 1 year: 2013 ident: 2748_CR45 publication-title: Struct Multidiscipl Optim doi: 10.1007/s00158-012-0807-3 – volume: 4 start-page: 61 issue: 2 year: 2016 ident: 2748_CR47 publication-title: Comput Methods Biomech Biomed Eng Imaging Vis doi: 10.1080/21681163.2015.1030775 – volume: 10 start-page: 40 issue: 1 year: 1995 ident: 2748_CR14 publication-title: Struct Optim doi: 10.1007/BF01743693 |
| SSID | ssj0008049 |
| Score | 2.6403666 |
| Snippet | Neural networks, and more broadly, machine learning techniques, have been recently exploited to
accelerate
topology optimization through data-driven training... Neural networks, and more broadly, machine learning techniques, have been recently exploited to accelerate topology optimization through data-driven training... |
| SourceID | proquest crossref springer |
| SourceType | Aggregation Database Enrichment Source Index Database Publisher |
| StartPage | 1135 |
| SubjectTerms | Computational Mathematics and Numerical Analysis Density Design optimization Engineering Engineering Design Image processing Isotropic material Machine learning Neural networks Optimization Research Paper Theoretical and Applied Mechanics Topology optimization |
| SummonAdditionalLinks | – databaseName: ProQuest Central dbid: BENPR link: http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwfV07T8MwED5Bu8CAeIpCQRnYwCIOtpOwIEBUFRIpQq3ULfIrLNAW2g78e86J0wISnZ2c5M_23Xf2PQDOuOJFyKOCMGMUYVIWRIrIkFQVQoSpEZK5fOenTHQH7HHIh_7CberDKmudWCpqM9bujvwSTTXuFtdn6WbyQVzXKPe66ltorEMTVXCCzlfz7iF7flno4qQiwI7WEMrioU-bKZPnHF1IiHOfnGuWEPbbNC355p8n0tLydLZhy1PG4LZa4x1Ys6Nd2PxRSHAPon5vnmXXQb9qefAV9FATvPsUy8DFtr8GrgwHismquO_pPgw6D_37LvHdEIjGic-IiCy1KTqzLJSxKXgiY2GYxgOHJqWgXCvrKs0o9Hcp08wkMpQWCYK8MkJoVCMH0BiNR_YQAl5Emmpq4pQqHLepsjLRVEZcIyPTpgW0BiLXvlS461jxli-KHJfg5QheXoKXsxacL_6ZVIUyVn7drvHN_aGZ5sslbsFFjfly-H9pR6ulHcNG5CJRysixNjRmn3N7glRipk79fvkGYZ7DLQ priority: 102 providerName: ProQuest |
| Title | TOuNN: Topology Optimization using Neural Networks |
| URI | https://link.springer.com/article/10.1007/s00158-020-02748-4 https://www.proquest.com/docview/2503197316 |
| Volume | 63 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV07T8MwED7RdoEB8RSFUmVgA0txsJ2ErUVNEYgUoVYqU-TYDgsURNuBf885rwICJCYPdizl7Lv7Tr77DuCEpzxzuZcRpnVKmJQZkcLTJEwzIdxQC8lsvfNtLK4m7HrKp2VR2LzKdq-eJHNLXRe7WfceEBvu2FAqIKwBLW5jd7zFE69X29-gAL0WyhDK_GlZKvPzHl_d0QpjfnsWzb1NtAWbJUx0esW5bsOame3AxifywF3wxqNlHF8446LNwbszQu1_LssqHZvP_uhY6g3cJi5yved7MIkG48srUnZAIApVY0GEZ6gJMYBlrvR1xgPpC80UKhm6kYxylRrLLpNijEuZYjqQrjQICuS5FkKh6diH5uxlZg7A4ZmnqKLaD2mK8yZMjQwUlR5XiMKUbgOtBJGokh7cdql4Smpi41x4CQovyYWXsDac1t-8FuQYf67uVPJNSkWZJ4jA8E9t-6w2nFUyX03_vtvh_5Yfwbpns1Hy7LEONBdvS3OMcGKRdqERRMMutHpRvx_bcfhwM8CxP4jv7rv53foA5LTERA |
| linkProvider | Springer Nature |
| linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV05T8MwFH6CMgAD4hTlzAATWNSu4yRICCGgFFrCEqRuwbEdFmgLbYX4U_xGnnO0gAQbs5M3fPneFb8DYM9N3LTmspRwrRPCpUyJFEyTIEmFqAVaSG77nW9D0bznNx23MwUfZS-MLassbWJmqHVP2X_kR-iqkS12z9Jp_4XYrVH2drVcoZHTomXe3zBlG5xcX-D33WescRmdN0mxVYAoFDAkghlqAkwKeU16OnV96QnNFRIXTXNKXZUYO7ElwbyRcsW1L2vSoKOVdS2EQnVEudMww-v1wGqU37gaW34_D7dtEEUo9zpFk07WqmeDE5_YZM0mgj7h3x3hJLr9cSGb-bnGIiwUAapzljNqCaZMdxnmv4wtXAEW3Y3C8NiJ8gUL784d2p3noqHTsZX0j44d-oFiwrzKfLAK9_-C0hpUur2uWQfHTZmiimovoAmemyAx0ldUMldh_Kd0FWgJRKyKweR2P8ZTPB6pnIEXI3hxBl7Mq3Awfqefj-X48-mtEt-4UNFBPCFUFQ5LzCfHv0vb-FvaLsw2o9t23L4OW5swx2wNTFaztgWV4evIbGMQM0x2MuY48PDfVP0EXN79iQ |
| linkToPdf | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV07T8MwELagSAgGxFMUCmRgA6u1sZ2ErQKq8koZWqmb5ScLhIqmA_8eO68WBEjMdiz5Yt99J9_3HQCnVFLbodhCorWERAgLBcMaxtIy1ok1E8TznR8T1h-RuzEdL7D482r36kmy4DR4laY0a0-0bdfENx_qI-hTH59WRZAsgxXnjpE_1yPcrX1xVABgD2sgIuG4pM38vMbX0DTHm9-eSPPI09sEGyVkDLrFP94CSybdBusLQoI7AA8HsyS5DIZFy4OPYOA8wWtJsQx8bftz4GU43DJJUfc93QWj3s3wqg_LbghQuX1lkGGDTOySWdIRobY0EiHTRLkL50KKRVRJ45VmpMt3EVFER6IjjAMI4kIzppwb2QON9C01-yCgFiukkA5jJN24iaURkUICU-UQmdJNgCpDcFVKhfuOFS-8FjnOjced8XhuPE6a4Kz-ZlIIZfw5u1XZl5eXZsodGnM79a20muC8svl8-PfVDv43_QSsPl33-MNtcn8I1rAvUsmLylqgkb3PzJFDGZk8zg_SJwg0xeM |
| 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=TOuNN%3A+Topology+Optimization+using+Neural+Networks&rft.jtitle=Structural+and+multidisciplinary+optimization&rft.au=Chandrasekhar%2C+Aaditya&rft.au=Suresh%2C+Krishnan&rft.date=2021-03-01&rft.pub=Springer+Berlin+Heidelberg&rft.issn=1615-147X&rft.eissn=1615-1488&rft.volume=63&rft.issue=3&rft.spage=1135&rft.epage=1149&rft_id=info:doi/10.1007%2Fs00158-020-02748-4&rft.externalDocID=10_1007_s00158_020_02748_4 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1615-147X&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1615-147X&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1615-147X&client=summon |