PSI - Issue 40

Available online at www.sciencedirect.com Available online at www.sciencedirect.com ScienceDirect Structural Integrity Procedia 00 (2022) 000 – 000 Available online at www.sciencedirect.com ScienceDirect Structural Integrity Procedia 00 (2022) 000 – 000

www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia

ScienceDirect

Procedia Structural Integrity 40 (2022) 194–200

© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the15th International Conference on Mechanics, Resources and Diagnostics of Materials and Structures. Abstract A neural network is constructed to describe flow stress curves for the AlMg6/10% SiC metal matrix composite (MMC) at temperatures ranging from 300 to 500  С and strain rates ranging from 0.1 to 5 s − 1 . The metal matrix composite was produced by powder metallurgy technologies from the AlMg6 alloy (the 1560 aluminum alloy according to GOST 4784-97) with the addition of 10% SiC powder having a fraction of F1500. The obtained flow stress curves of the AlMg6/10% SiC MMC in the studied temperature – rate range have several sections (stages). At the first stage, the material hardens, then it softens, and a peak of deformation stress appears on the flow stress curve. After the softening stage, a steady section begins, at which the hardening and softening rates almost coincide. This is expressed in the constancy of the flow stress value with an increase in strain. The obtained flow stress curves are described by a neural network. The study shows that the constructed neural network can predict with acceptable engineering accuracy the behavior of flow stress curves for temperatures and strain rates that are not used in its training. © 2022 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the15th International Conference on Mechanics, Resource and Diagnostics of Materials and Structures. Keywords: metal matrix composite; neural network; strain; high temperature; aluminum; magnesium. Neural network modeling of the flow stress of the AlMg6/10% SiC metal matrix composite under deformation at high temperatures V.S. Kanakin*, A.S. Smirnov, and A.V. Konovalov Institute of Engineering Science, UB RAS, 34 Komsomolskaya St., 620049 Ekaterinburg, Russia Abstract A neural network is constructed to describe flow stress curves for the AlMg6/10% SiC metal matrix composite (MMC) at temper tures ranging fr m 300 to 500  С and strain rat ranging fr m 0.1 to 5 s − 1 . The metal matrix composite was produced by powder metallur y technologies from the AlMg6 alloy (the 1560 alu inum alloy according to GOST 4784-97) ith the addition of 10% SiC powder having a fraction of F1500. The obtained flow stress curves of the AlMg6/10% SiC MMC in the stud ed temperature – rate range have several sections (stages). At the irst stage, the material hardens, then it softens, a d a peak of deform ion str ss pp ars on the flow stress curve. After he so tening stage, a steady section begins, at which the hardening and so tening rates almost coincide. This is xpress d in the cons a cy of th flow stress value with n in reas in stra . The obtai ed flow stress curves are described by a neural network. The study shows tha the constructed neural network can predict wi h acceptable engineering accura y the eh vior of flo stress curves for temper tur s a d strain rat s that are not used in ts training. © 2022 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review u der responsibility of scientific committe of the15th International C ference o Mechanics, Resource and Diagnostics of Materials and S ructur s. Keywords: metal m rix composite; neural network; strain; high temperature; aluminum; magnesium. 15th International Conference on Mechanics, Resource and Diagnostics of Materials and Structures Neural network modeling of the flow stress of the AlMg6/10% SiC metal matrix composite under deformation at high temperatures V.S. Kanakin*, A.S. Smirnov, and A.V. Konovalov Institute of Engineering Science, UB RAS, 34 Komsomolskaya St., 620049 Ekaterinburg, Russia 15th International Conference on Mechanics, Resource and Diagnostics of Materials and Stru tures

* Corresponding author. Tel.: +7 (343) 375-35-89; fax: +7 (343) 374-53-30. E-mail address: kanakin.v.s@gmail.com * Corresponding author. Tel.: +7 (343) 375-35-89; fax: +7 (343) 374-53-30. E-mail ad ress: kanakin.v s@gmail.com

2452-3216 © 2022 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the15th International Conference on Mechanics, Resource and Diagnostics of Materials and Structures. 2452-3216 © 2022 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review u der re ponsibility of scientific committe of the15th Int rnational C ference o Mechanics, Resource and Diagnostics of Mate ials and Structures.

2452-3216 © 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the15th International Conference on Mechanics, Resources and Diagnostics

of Materials and Structures. 10.1016/j.prostr.2022.04.026