Hybrid RSM–ANN–GA Optimization of TIG Hot-Wire Welding for AISI 5160 Steel
Engineering, Technology & Applied Science Research · 06.06.2026 · 16 (3) · 34879-34886
DOI: 10.48084/etasr.17832 ↗Kurz gefasst
Fachartikel in der Zeitschrift Engineering, Technology & Applied Science Research, online erschienen am 06.06.2026. Verfasst von Phattharapong Keidlaphi und 3 weiteren Autor:innen. Die Arbeit nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Repair welding of AISI 5160 chromium spring steel is challenging due to its high sensitivity to thermal cycling, which may cause HAZ softening, carbide instability, and reduced toughness. This study proposes an integrated Response Surface Methodology–Artificial Neural Network–Genetic Algorithm (RSM–ANN–GA) framework for predicting and optimizing the multi-objective performance of TIG hot-wire repair welding, aiming to simultaneously enhance weld hardness and impact energy under limited experimental data. A compact Central Composite Design (CCD) with nine runs was employed, by varying Welding Current (WC) and Hot-Wire Current (HWC), while travel speed was kept constant. The measured responses ranged from 49.6–62.2 HRC and 19.8–35.5 J, indicating a strength–toughness interaction governed by heat-balance control. Quadratic RSM–CCD models assessed by ANOVA showed excellent adequacy for hardness (R² = 0.979; Adjusted R² = 0.944), dominated by WC and WC² effects, whereas impact energy was primarily influenced by nonlinear HWC terms with moderate predictability (R² = 0.882; Adjusted R² = 0.686). A feed-forward ANN (2–10–2) provided higher predictive accuracy (R² = 0.97 for hardness; R² = 0.99 for impact energy) compared to RSM. ANN was embedded into a desirability-based GA optimizer, yielding an optimal condition at WC = 200 A and HWC = 150 A, with predicted responses of 58.17 HRC and 35.55 J, with a high desirability (D) of 0.95. Microstructural and fractographic results supported enhanced weld uniformity and ductile fracture behavior.
Zitieren
APA
Keidlaphi, P., Charonerat, S., Prasomthong, S., & Chaiyason, T. (2026). Hybrid RSM–ANN–GA Optimization of TIG Hot-Wire Welding for AISI 5160 Steel. Engineering, Technology & Applied Science Research, 16(3), 34879-34886. https://doi.org/10.48084/etasr.17832
Vancouver
Keidlaphi P, Charonerat S, Prasomthong S, Chaiyason T. Hybrid RSM–ANN–GA Optimization of TIG Hot-Wire Welding for AISI 5160 Steel. Engineering, Technology & Applied Science Research. 2026;16(3):34879-34886. doi:10.48084/etasr.17832
BibTeX
@article{Keidlaphi2026,
author = {Keidlaphi, Phattharapong and Charonerat, Sittichai and Prasomthong, Suriya and Chaiyason, Teephet},
title = {Hybrid RSM–ANN–GA Optimization of TIG Hot-Wire Welding for AISI 5160 Steel},
journal = {Engineering, Technology & Applied Science Research},
year = {2026},
volume = {16},
number = {3},
pages = {34879--34886},
doi = {10.48084/etasr.17832}
}
EUVIMED