Physics-Guided AI Co-Design and Control to Minimize Shrinkage and Gas Porosity in Metal Castings
DOI:
https://doi.org/10.5281/zenodo.22974976Keywords:
Metal Casting, Shrinkage Porosity, Solidification Modeling, Adaptive Process ControlAbstract
orosity in metal castings—caused by shrinkage during solidification, entrained air, and dissolved gases—remains one of the most persistent quality challenges in foundry engineering. This paper presents a hybrid, physics-guided artificial intelligence (AI) framework that combines predictive modeling, inverse design, and real-time control to minimize shrinkage and gas-related porosity. The system integrates solidification physics and process data through a 3D neural architecture that distinguishes porosity mechanisms, predicts their spatial distribution, and recommends corrective design or process measures. Bayesian optimization and reinforcement learning are applied for process tuning and adaptive control. Experimental validation on Al-Si castings shows a 40% reduction in total porosity through inverse design and an additional 25% improvement with real-time control. The framework demonstrates the synergy between physics-based reasoning and AI learning for intelligent defect prevention in modern foundries.
References
Abu-Mahfouz, I., El Ariss, O., Esfakur Rahman, A., & Banerjee, A. (2017). Surface roughness prediction as a classification problem using support vector machine. The International Journal of Advanced Manufacturing Technology, 92(1), 803-815. https://doi.org/10.1007/s00170-017-0165-9
Alajmi, M. S., & Almeshal, A. M. (2020). Predicting the tool wear of a drilling process using novel machine learning XGBoost-SDA. Materials, 13(21), 4952. https://doi.org/10.3390/ma13214952
Bai, J., Rabczuk, T., Gupta, A., Alzubaidi, L., & Gu, Y. (2023). A physics-informed neural network technique based on a modified loss function for computational 2D and 3D solid mechanics. Computational Mechanics, 71(3), 543-562. https://doi.org/10.1007/s00466-022-02252-0
Banerjee, S., Poria, S., Sutradhar, G., & Sahoo, P. (2019). Wear performance of Mg-WC metal matrix nanocomposites using taguchi methodology. Materials Today: Proceedings, 19, 177-181. https://doi.org/10.1016/j.matpr.2019.06.642
Bhagwat, V. B., Kamble, D. A., & Kore, S. S. (2024). An overview of machine learning applications in metal casting industries. Archives of Metallurgy and Materials, 69(4), 1577-1584. https://doi.org/10.24425/amm.2024.151428
Bosse, S., Lehmhus, D., & Kumar, S. (2024). Automated porosity characterization for aluminum die casting materials using X-ray radiography, synthetic X-ray data augmentation by simulation, and machine learning. Sensors, 24(9), 2933. https://doi.org/10.3390/s24092933
Deng, J., Xu, Y., Zuo, Z., Hou, Z., & Chen, S. (2019). Bead geometry prediction for multi-layer and multi-bead wire and arc additive manufacturing based on XGBoost. In Transactions on Intelligent Welding Manufacturing: Volume II No. 4 2018 (pp. 125-135). Springer. https://doi.org/10.1007/978-981-13-8668-8_7
Gao, K., Chen, H., Zhang, X., Ren, X., Chen, J., & Chen, X. (2019). A novel material removal prediction method based on acoustic sensing and ensemble XGBoost learning algorithm for robotic belt grinding of Inconel 718. The International Journal of Advanced Manufacturing Technology, 105(1), 217-232. https://doi.org/10.1007/s00170-019-04170-7
Gaspar-Cunha, A., & Covas, J. A. (2008). Robustness in multi-objective optimization using evolutionary algorithms. Computational optimization and applications, 39(1), 75-96. https://doi.org/10.1007/s10589-007-9053-9
Habibnejad-Korayem, M., Mahmudi, R., Ghasemi, H., & Poole, W. (2010). Tribological behavior of pure Mg and AZ31 magnesium alloy strengthened by Al2O3 nano-particles. Wear, 268(3-4), 405-412. https://doi.org/10.1016/j.wear.2009.08.031
Ivanina, E., Monastyrskiy, V., & Ershov, M. Y. (2022). Quantitative estimation of formation of shrinkage porosity by the niyama criterion. Inorganic Materials: Applied Research, 13(1), 100-105. https://doi.org/10.1134/S2075113322010130
Janardhan, M. (2015). An integrated evaluation approach for modelling and optimization of surface grinding process parameters. Materials Today: Proceedings, 2(4-5), 1622-1633. https://doi.org/10.1016/j.matpr.2015.07.089
Kathane, S., & Karagadde, S. (2024). A physics informed neural network (PINN) methodology for coupled moving boundary PDEs. arXiv preprint arXiv:2409.10910. https://doi.org/10.48550/arXiv.2409.10910
Kennedy, J., & Eberhart, R. (1995). Particle swarm optimization. In Proceedings of ICNN'95-international conference on neural networks (Vol. 4, pp. 1942-1948). IEEE. https://doi.org/10.1109/ICNN.1995.488968
Kiangala, S. K., & Wang, Z. (2021). An effective adaptive customization framework for small manufacturing plants using extreme gradient boosting-XGBoost and random forest ensemble learning algorithms in an Industry 4.0 environment. Machine Learning with Applications, 4, 100024. https://doi.org/10.1016/j.mlwa.2021.100024
Li, H., Miao, Y., Khodaei, Z. S., & Aliabadi, M. (2025). Finite-PINN: A physics-informed neural network with finite geometric encoding for solid mechanics. Journal of the Mechanics and Physics of Solids, 203, 106222. https://doi.org/10.1016/j.jmps.2025.106222
Makadia, A. J., & Nanavati, J. (2013). Optimisation of machining parameters for turning operations based on response surface methodology. Measurement, 46(4), 1521-1529. https://doi.org/10.1016/j.measurement.2012.11.026
Mamalis, A. (2005). Advanced manufacturing engineering. Journal of Materials Processing Technology, 161(1-2), 1-9. https://doi.org/10.1016/j.jmatprotec.2004.07.055
Nguyen, Q., Sim, Y., Gupta, M., & Lim, C. (2015). Tribology characteristics of magnesium alloy AZ31B and its composites. Tribology International, 82, 464-471. https://doi.org/10.1016/j.triboint.2014.02.024
Nieves, J., Garcia, D., Angulo-Pines, J., Santos, F., & Rodriguez, P. P. (2025). An artificial intelligence-based digital twin approach for rejection rate and mechanical property improvement in an investment casting plant. Applied Sciences, 15(4), 2013. https://doi.org/10.3390/app15042013
Pal, A., Poria, S., Sutradhar, G., & Sahoo, P. (2018). Tribological behavior of Al-WC nano-composites fabricated by ultrasonic cavitation assisted stir-cast method. Materials Research Express, 5(3), 036521. https://doi.org/10.1088/2053-1591/aab577
Ruppert, D. (2004). The elements of statistical learning: data mining, inference, and prediction. In. Taylor & Francis. https://doi.org/10.1198/jasa.2004.s339
Song, K., Yan, F., Ding, T., Gao, L., & Lu, S. (2020). A steel property optimization model based on the XGBoost algorithm and improved PSO. Computational Materials Science, 174, 109472. https://doi.org/10.1016/j.commatsci.2019.109472
Stefanescu, D. M. (2015). Science and engineering of casting solidification. Springer. https://doi.org/10.1007/978-3-319-15693-4
Uyan, T. C., Otto, K., Silva, M. S., Vilaça, P., & Armakan, E. (2023). Industry 4.0 foundry data management and supervised machine learning in low-pressure die casting quality improvement. International journal of metalcasting, 17(1), 414-429. https://doi.org/10.1007/s40962-022-00783-z
Vijayaram, T. R., Sulaiman, S., Hamouda, A., & Ahmad, M. (2006). Numerical simulation of casting solidification in permanent metallic molds. Journal of Materials Processing Technology, 178(1-3), 29-33. https://doi.org/10.1016/j.jmatprotec.2005.09.025
Wang, S., Teng, Y., & Perdikaris, P. (2020). Understanding and mitigating gradient pathologies in physics-informed neural networks. arXiv preprint arXiv:2001.04536. https://doi.org/10.48550/arXiv.2001.04536
Wickham, H., Averick, M., Bryan, J., Chang, W., McGowan, L. D. A., François, R., Grolemund, G., Hayes, A., Henry, L., & Hester, J. (2019). Welcome to the Tidyverse. Journal of open source software, 4(43), 1686. https://doi.org/10.21105/joss.01686
Zafari, A., Ghasemi, H., & Mahmudi, R. (2012). Tribological behavior of AZ91D magnesium alloy at elevated temperatures. Wear, 292, 33-40. https://doi.org/10.1016/j.wear.2012.06.002
Zitzler, E., & Thiele, L. (1998). Multiobjective optimization using evolutionary algorithms—a comparative case study. In International conference on parallel problem solving from nature (pp. 292-301). Springer. https://doi.org/10.1007/bfb0056872
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Reports in Mechanical Engineering
This work is licensed under a Creative Commons Attribution 4.0 International License.