EXPLAINABLE AI FRAMEWORK FOR ROOT CAUSE ANALYSIS OF MARINE WELDING DEFECTS: INTEGRATING PROCESS PARAMETERS AND MICROSTRUCTURAL FAILURE MODES
Indonesia 
Abstract
Marine welding defects threaten structural reliability because conventional inspection and black-box prediction models can identify discontinuities without adequately explaining their physical origins. This study aimed to develop an Explainable Artificial Intelligence framework for root-cause analysis by integrating welding process parameters with microstructural failure modes. An experimental-computational design was applied to 120 AH36 marine-steel specimens produced by Gas Metal Arc Welding under controlled variations in current, voltage, travel speed, wire-feed rate, shielding-gas flow, and heat input. Weld quality was evaluated using visual inspection, ultrasonic testing, metallography, scanning electron microscopy, and Vickers microhardness, while Random Forest, XGBoost, and neural-network models were compared and interpreted using SHAP. Parameter variation enabled AI explanations to be checked against metallurgical evidence and defect morphology. XGBoost achieved the best performance, with 90.0% accuracy, a macro F1-score of 0.886, and an AUROC of 0.956. Heat input, shielding-gas flow, travel speed, and HAZ microhardness emerged as the dominant explanatory variables. Porosity was mainly associated with inadequate shielding, incomplete fusion and penetration with insufficient effective heat input, and cracking with elevated HAZ hardness. The study concludes that physically validated XAI can extend defect classification toward transparent, mechanism-informed diagnosis and provide a stronger basis for corrective decision-making in safety-critical marine welding.
Full text article
References
Adeleke, O., & Jen, T.-C. (2025). Explainable AI and machine learning-based analysis of municipal solid waste generation rate: A South African case study. Waste Management, 206, 115036. https://doi.org/10.1016/j.wasman.2025.115036
Alkhamis, S. A., Alhalangy, A., Eltayeb, G. E. A., & Abouelkheir, E. (2026). Fed-CGIDS-UAV: Federated Causal Graph Learning for Cross-Domain Intrusion Detection in Cyber-Physical Drone Networks. Symmetry, 18(8), 1292. https://doi.org/10.3390/sym18081292
Ammar, S., Ben Fraj, B., Hentati, H., Saouab, A., Ben Amar, M., & Haddar, M. (2024). Mechanical performances of printed carbon fiber-reinforced PLA and PETG composites. Proceedings of the Institution of Mechanical Engineers, Part L: Journal of Materials: Design and Applications, 238(8), 1488–1499. https://doi.org/10.1177/14644207231225761
Awad, I. G. A. (2026). Failure Analysis and Simultaneous Optimization of Flexural Strength and Elastic Modulus in Friction Stir Welded PP / HDPE – TIO2 Nanocomposite Joints. Polymers for Advanced Technologies, 37(7), e70660. https://doi.org/10.1002/pat.70660
Das, S., & Ganesh Narayanan, R. (2025). Lap shear fracture prediction of friction stir spot welds between AA6063-T6 and CRCA/IS-513 sheets. Proceedings of the Institution of Mechanical Engineers, Part E: Journal of Process Mechanical Engineering, 09544089251335077. https://doi.org/10.1177/09544089251335077
Gangwal, A., & Lavecchia, A. (2026). IMPACT Framework: Establishing Global Standards for Artificial Intelligence Implementation, Methodology, and Translation in Drug Discovery. WIREs Computational Molecular Science, 16(2), e70072. https://doi.org/10.1002/wcms.70072
Ghazi, Y., Tabaa, M., Ennaji, M., & Zaz, G. (2025). An Explainable Markov Chain–Machine Learning Sequential-Aware Anomaly Detection Framework for Industrial IoT Systems Based on OPC UA. Sensors, 25(19), 6122. https://doi.org/10.3390/s25196122
Huang, Z., Jiang, M., Cao, S., & Yilmaz, E. (2026). Engineering strength through microstructure: Gradation and content effects in cemented high-sulfur tailings-crushed rock backfill. Results in Engineering, 32, 111841. https://doi.org/10.1016/j.rineng.2026.111841
Kewalramani, M., Mahmoodzadeh, A., Hassan, R., El Ouni, M. H., Alqahtani, A., Alanazi, A., Alsubai, S., & Palani, S. (2025). Predicting critical crack propagation length in sustainable additive-enhanced concrete using explainable machine learning. Scientific Reports, 16(1), 26. https://doi.org/10.1038/s41598-025-31900-7
Khadka, P., Varglund, S., Akinwamide, S., & Vilaça, P. (2024). Characterization of friction stir-based linear continuous joining of aluminium alloy to structural polymer. Welding in the World, 68(1), 71–90. https://doi.org/10.1007/s40194-023-01622-3
Khorshidi, M., Nejatipour, P., Teimortashlu, A., Oliveto, G., & Afaridegan, E. (2026). Explainable AI-based optimization framework for sustainable fly ash concrete design. Scientific Reports, 16(1), 22397. https://doi.org/10.1038/s41598-026-59294-0
Li, Y., Cheng, K., Wang, L., & Yang, W. (2026). Fault Analysis and Optimization of PT100 Temperature Sensor for Marine Exhaust Duct. In J. P. T. Mo (Ed.), Advances in Transdisciplinary Engineering. IOS Press. https://doi.org/10.3233/ATDE260562
Mallipudi, S. R., & Marampudi, R. (2025). Mechanical and Metallurgical Characterization of Friction Stir Welded Al–Mg–Mn Alloy with Minor Additions of Sc and Zr. Transactions of the Indian Institute of Metals, 78(6), 148. https://doi.org/10.1007/s12666-025-03624-9
Msomi, V., Mabuwa, S., Ngonda, T., Mehdi, H., & Saxena, K. K. (2025). Microstructure and mechanical properties of tungsten inert gas welded dissimilar aluminum alloys joint subjected to normal multi-pass friction stir process. Proceedings of the Institution of Mechanical Engineers, Part E: Journal of Process Mechanical Engineering, 239(4), 1925–1935. https://doi.org/10.1177/09544089231207424
Naik, R., & Majhi, B. (2025). Explainable AI reverse verification approach for monthly rainfall prediction in Chhattisgarh, India. Theoretical and Applied Climatology, 156(7), 412. https://doi.org/10.1007/s00704-025-05645-2
Onyelowe, K. C., Hanandeh, S., Kamchoom, V., Ebid, A. M., Imran, H., Duque Vaca, M. A., Herrera Morales, G. C., Ulloa, N., & Arunachalam, K. P. (2025). Data-driven framework for prediction of mechanical properties of waste glass aggregates concrete. Scientific Reports, 15(1), 20902. https://doi.org/10.1038/s41598-025-05229-0
Özüpak, Y. (2026). A hybrid physics-guided ensemble learning framework for predicting aviation co? emissions with explainable artificial intelligence. Theoretical and Applied Climatology, 157(7), 426. https://doi.org/10.1007/s00704-026-06356-y
Patel, R., Tiwari, S. K., Rakesh, D. K., Kailasanathan, N., & Selavarajan, S. (2025). An automated framework for traffic noise level analysis using explainable artificial intelligence techniques. Scientific Reports, 15(1), 36657. https://doi.org/10.1038/s41598-025-20377-z
Peng, J., Wang, S., Ding, W., Huang, X., Slamani, M., & Zhao, B. (2026). Machinability and surface formation mechanisms of SiCp/Al composites at different cutting speeds. Precision Engineering, 101, 310–322. https://doi.org/10.1016/j.precisioneng.2026.05.001
Piyar S, N., & S, D. B. (2026). An explainable spatio-temporal deep learning framework for crop yield prediction and recommendation. Scientific Reports, 16(1), 24662. https://doi.org/10.1038/s41598-026-54439-7
Rajagopal, V., Rajendran, D. K., & Muthukumaran, S. (2024). Multi-objective optimization and characterization of AA2219-T87 and AA6061-T6 dissimilar friction stir welded joints under the effect of copper backing using AUGMECON. S?dhan?, 49(1), 32. https://doi.org/10.1007/s12046-023-02360-4
Saroj, S., Sahoo, P., & Kore, S. D. (2026). Influence of induction preheating and post-weld heat treatment on the mechanical, microstructural and corrosion properties of friction stir welded marine-grade aluminium alloy. Welding International, 40(3), 211–228. https://doi.org/10.1080/09507116.2025.2525155
Shawon, S. M., Zaman, M., Maniam, S., Al Qwaid, M., Sarker, M. T., Kheng, T. Y., & Wong, H. Y. (2026). Explainable Quantile CNN-LSTM model for uncertainty-aware multi-layer soil moisture prediction in tropical cocoa plantations. Scientific Reports, 16(1), 16994. https://doi.org/10.1038/s41598-026-48517-z
Shen, L., Shojaeikhah, S., & Soliman, M. (2025). Experimental analysis of crack propagation in stiffened box girders under variable-amplitude loading. Journal of Constructional Steel Research, 235, 109812. https://doi.org/10.1016/j.jcsr.2025.109812
Shivaprasad, S., Chadaga, K., Dias, C. C., Sampathila, N., & Prabhu, S. (2024a). An interpretable schizophrenia diagnosis framework using machine learning and explainable artificial intelligence. Systems Science & Control Engineering, 12(1), 2364033. https://doi.org/10.1080/21642583.2024.2364033
Shivaprasad, S., Chadaga, K., Dias, C. C., Sampathila, N., & Prabhu, S. (2024b). An interpretable schizophrenia diagnosis framework using machine learning and explainable artificial intelligence. Systems Science & Control Engineering, 12(1), 2364033. https://doi.org/10.1080/21642583.2024.2364033
Singh, A., Sharma, S. K., & Batish, A. (2026). Achieving high-performance SS304 joints through dynamic recrystallization in friction stir welding. Proceedings of the Institution of Mechanical Engineers, Part L: Journal of Materials: Design and Applications, 240(9), 1567–1580. https://doi.org/10.1177/14644207251384748
Singh, A., Singh, S., Kamboj, R. K., & Bala, A. (2025). Explainable OptiCNN-SLSTM hybrid model for enhanced lithium-ion battery state of health prediction. Scientific Reports, 15(1), 44364. https://doi.org/10.1038/s41598-025-28091-6
Varma, S. O., Jadhav, P. P., Vishwakarma, A. L., Sonawane, M. R., & Chaudhari, A. (2026). Explainable neural network models for rapid prediction of dielectric constant and acid value in edible oils using FTIR spectroscopy. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 363, 128376. https://doi.org/10.1016/j.saa.2026.128376
Wang, B., Zhao, S., Ren, Q.-X., & Xie, R.-Q. (2025). Axial compressive behavior of circular hollow steel tube-reinforced UHTCC column. Structures, 71, 107943. https://doi.org/10.1016/j.istruc.2024.107943
Wei, F., & Frederick, A. (2025). Comprehensive Die Strengths Comparisions for Glass-Core Advanced Packaging Substrates using Different Singulation Methods—Dicing-Induced SeWaRe Failures Revisited Ten Years Later-. 2025 IEEE 75th Electronic Components and Technology Conference (ECTC), 146–153. https://doi.org/10.1109/ECTC51687.2025.00031
Wu, D., Wang, C., Liu, H., Zhang, Z., Liang, J., & Liu, Y. (2026). Correlation of breakage characteristics between single-particle and uniformly-graded volcanic soil: Insights from compression tests. Powder Technology, 468, 121611. https://doi.org/10.1016/j.powtec.2025.121611
Xu, Y., Saleem, M. F., Mingyu, G., Xie, W. L., Zhang, S.-H., Zhu, H. L., Khina, B. B., & Pokrovsky, A. I. (2026). Formability and failure mechanisms of ultra-thin TA2 titanium sheets in microchannel hydroforming: Effect of thickness-to-grain-size ratio. Thin-Walled Structures, 218, 114042. https://doi.org/10.1016/j.tws.2025.114042
Zhang, Y., He, C., Yu, Q., Li, X., Wang, X., Zhang, Y., Wang, J., Jiang, C., Jia, Y., Zhang, X.-C., Sun, B., Ritchie, R. O., & Tu, S.-T. (2024). Nacre-like surface nanolaminates enhance fatigue resistance of pure titanium. Nature Communications, 15(1), 6917. https://doi.org/10.1038/s41467-024-51423-5
Zhou, B., Li, Y., Wu, Y., Zhang, E., Chen, X., Tao, Z., & Xing, N. (2026). Investigation of the failure characteristics of goaf roof in a metal mine using the acoustic emission technique. Nondestructive Testing and Evaluation, 1–30. https://doi.org/10.1080/10589759.2026.2627620
Zhou, Q., Liu, H., Tu, W., Xue, T., & Xue, H. (2026). Multi-scale decoupling analysis of geometric imperfection effects on stress concentration in butt-welded joints of IMO type B tanks. Ocean Engineering, 355, 125078. https://doi.org/10.1016/j.oceaneng.2026.125078
Authors
Copyright (c) 2026 Mohammad Syaifi
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.