PSI - Issue 42

Available online at www.sciencedirect.com Available online at www.sciencedirect.com ScienceDirect Structural Integrity Procedia 00 (2019) 000 – 000 il l li t . i ir t. i i Structural Integrity Procedia 00 (2019) 000 – 000

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

ScienceDirect

Procedia Structural Integrity 42 (2022) 1344–1349

© 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 the 23 European Conference on Fracture – ECF23 Abstract There was studied a deformation diagrams of AL-6061 aluminum alloy at various temperatures. Predicting the behavior of nonlinear systems is important. Experimental data have a certain variance that must be taken into account. This gives us a regression problem that can be solved by machine learning. Therefore, with sufficient experimental data, it is advisable to use machine learning methods, namely neural networks and boosted trees in science and technology, where stress-strain diagrams of structural materials is extremely important, in particular in metallurgy, aircraft, railways and more. The diagrams of AL-6061 aluminum alloy deformation was predicted by machine learning methods, in particular, neural networks and boosted trees at temperatures T = 343, 413  C. It is shown that the obtained results are in good agreement with the experimental ones. It was founded that the neural network method gives the least prediction error of 0.05% in the test sample. © 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Peer-review under responsibility of 23 European Conference on Fracture - ECF23 Keywords: AL-6061 aluminum alloy; deformation diagram; machine learning; neural networks, boosted trees; 1. Introduction Machine learning studies the method of constructing algorithms that are studied by examples by Mitchell (1997). It is used in many areas of science and technology, including the fracture mechanics by Pidaparti et al. (1995), str ct ere as st ie a ef r ati ia ra s f - al i all at ari s te erat res. re icti t e e a i r f li ear s ste s is i rta t. eri e tal ata a e a certai aria ce t at st e ta e i t acc t. is i es s a re ressi r le t at ca e s l e ac i e lear i . eref re, it s fficie t e eri e tal ata, it is a isa le t se ac i e lear i et s, a el e ral et r s a ste trees i scie ce a tec l , ere stress-strai ia ra s f str ct ral aterials is e tre el i rta t, i artic lar i etall r , aircraft, rail a s a re. e ia ra s f - aluminum alloy deformation was predicted by machine learning methods, in particular, neural networks and boosted trees at te erat res ,  . It is s t at t e tai e res lts are i a ree e t ith the e eri e tal es. It as f e t at t e e ral et r et i es t e least re icti err r f . i t e test sa le. 20 0 e t rs. lis e lse ier . . This is an ope access article er t e - - lice se ( tt ://creati ec s. r /lice ses/ - c- / . /) eer-re ie er res si ilit f r ea fere ce ract re - Keywords: AL-6061 aluminum alloy; deformation diagram; machine learning; neural networks, boosted trees; . I t ti i l r i st i s t t f str ti l rit s t t r st i l s it ll ( ). , i l i t fr t r i s i rti t l. ( ), 23 European Conference on Fracture - ECF23 Modelling of AL-6061 aluminum alloy deformation diagrams by machine learning methods Iryna Didych a, *, Oleh Yasniy a , Iaroslav Pasternak b , Lukash Sobashek c a Ternopil Ivan Puluj National Technical University, 56 Ruska str., Ternopil, 46001, Ukraine b Lesya Ukrainka Volyn National University,13 Voli Ave., Lutsk, 43025, Ukraine c Politechnika Lubelska, 38D Nadbystrzycka str., Lublin, 20-618, Poland r f r r t r - Modelling of AL-6061 aluminum alloy Iryna Didych a, *, Oleh Yasniy a , Iaroslav Pasterna b , c a ernopil Ivan uluj ational echnical niversity, 56 uska str., ernopil, 46001, kraine b esya krainka olyn ational niversity,13 oli ve., utsk, 43025, kraine c olitechnika ubelska, 38 adbystrzycka str., ublin, 20-618, oland It is s i r s f s i t l

* Corresponding author E-mail address: iryna.didych1101@gmail.com * orresponding author - ail address: iryna.didych1101 g ail.co

2452-3216 © 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Peer-review under responsibility of 23 European Conference on Fracture - ECF23 2452-3216 2020 he uthors. ublished by lsevier . . his is an open access article under the - - license (http://creativeco ons.org/licenses/by-nc-nd/4.0/) er res si ilit f r ea fere ce ract re - eer-re ie

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 the 23 European Conference on Fracture – ECF23 10.1016/j.prostr.2022.12.171