PSI - Issue 68

ScienceDirect Structural Integrity Procedia 00 (2025) 000–000 Structural Integrity Procedia 00 (2025) 000–000 Available online at www.sciencedirect.com Available online at www.sciencedirect.com ScienceDirect Available online at www.sciencedirect.com ScienceDirect

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

Procedia Structural Integrity 68 (2025) 84–90

European Conference on Fracture 2024 Performance evaluation of artificial neural networks developed for estimation of fatigue behavior of low-alloy steels Tea Marohnić a, *, Robert Basan a a University of Rijeka, Faculty of Engineering, Vukovarska 58, HR-51000 Rijeka, Croatia Abstract In the literature, numerous empirical and machine learning-based techniques for estimation of strain-life fatigue parameters from monotonic properties have been proposed. Existing machine learning-based methods are evaluated in different manners, making their comparison difficult. Most authors use metrics such as root mean square error RMSE and neglect fatigue life estimations criteria, or use only conventional error criterion E f ( s ). In this study, ANNs developed for estimation of fatigue parameters have been evaluated regarding their accuracy and applicability for estimation of low- and high-cycle fatigue lives of low-alloy steels, further divided into low- and high-strength subgroups. © 2025 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 ECF24 organizers Keywords: artificial neural networks; evaluation; fatigue parameters; fatigue life; low-alloy steels 1. Introduction Numerous methods for estimation of strain-life fatigue parameters from monotonic properties have been proposed in the literature, and new ones are being developed shifting the approach from empirical (analytical) methods like Muralidharan and Manson (1988), Bäumel and Seeger (1990), Roessle and Fatemi (2000), Wachter and Esderts (2018) and others to addressing the same topic using machine learning, such as Genel (2004), Troschenko et al. (2011), Marohnić (2017) and Soyer (2022). European Conference on Fracture 2024 Performance evaluation of artificial neural networks developed for estimation of fatigue behavior of low-alloy steels Tea Marohnić a, *, Robert Basan a a University of Rijeka, Faculty of Engineering, Vukovarska 58, HR-51000 Rijeka, Croatia Abstract In the literature, numerous empirical and machine learning-based techniques for estimation of strain-life fatigue parameters from monotonic properties have been proposed. Existing machine learning-based methods are evaluated in different manners, making their comparison difficult. Most authors use metrics such as root mean square error RMSE and neglect fatigue life estimations criteria, or use only conventional error criterion E f ( s ). In this study, ANNs developed for estimation of fatigue parameters have been evaluated regarding their accuracy and applicability for estimation of low- and high-cycle fatigue lives of low-alloy steels, further divided into low- and high-strength subgroups. © 2025 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 ECF24 organizers Keywords: artificial neural networks; evaluation; fatigue parameters; fatigue life; low-alloy steels 1. Introduction Numerous methods for estimation of strain-life fatigue parameters from monotonic properties have been proposed in the literature, and new ones are being developed shifting the approach from empirical (analytical) methods like Muralidharan and Manson (1988), Bäumel and Seeger (1990), Roessle and Fatemi (2000), Wachter and Esderts (2018) and others to addressing the same topic using machine learning, such as Genel (2004), Troschenko et al. (2011), Marohnić (2017) and Soyer (2022). © 2025 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 ECF24 organizers

* Corresponding author. Tel.: +385-51-651-531; fax: +385-51-651-416. E-mail address: tmarohnic@riteh.uniri.hr * Corresponding author. Tel.: +385-51-651-531; fax: +385-51-651-416. E-mail address: tmarohnic@riteh.uniri.hr

2452-3216 © 2025 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 ECF24 organizers 2452-3216 © 2025 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 ECF24 organizers

2452-3216 © 2025 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 ECF24 organizers 10.1016/j.prostr.2025.06.026