Exploiting Artificial Neural Networks for Digital Twins in Sand Casting

Abstract:

Dedicated simulation software is used to create a dataset which for a particular sand-cast part and various casting parameter values provide casting quality metrics based on temperature and solidification evolution. Based on simulation data, ANNs are trained to predict the successful or failed filling of the mold (a classification problem), as well as the quality of the part through solidification time, maximum microporosity, maximum von Mises residual stress, maximum displacement of any point in the casting and total volumetric shrinkage (a regression problem). Such ANNs can provide augmented information much faster than the simulation model to the process planner. A third category of ANNs (of the regression type, too) determine the temperature evolution with time at an inaccessible point, where no thermocouple can be placed, from the measurement history at two other thermocouples close to it. This data comes from real-time monitoring during casting. Such ANNs can aid the process supervisor in a ‘digital shadow’ context. The issues associated with generalizing these predictors to become independent of specific part geometry are also discussed.

• Periodical: Materials Science Forum (Volume 1188)
• Pages: 85-94
• Online since: April 2026
• Authors: Evangelos Pischinas, Emmanouil Stathatos, George Christopher Vosniakos*
• Keywords: Digital Twin, Metal Casting, Neural Networks, Process Simulation

References

[1] Z. Chen, Y. Li, F. Zhao, S. Li, J. Zhang. Progress in numerical simulation of casting process. Measurement and Control, 55/5-6 (2022) 257-264.

DOI: 10.1177/00202940221102656

[2] I. Onaji, D. Tiwari, P. Soulatiantork, B. Song, A. Tiwari. Digital twin in manufacturing: Conceptual framework and case studies. International Journal of Computer Integrated Manufacturing, 35/8 (2022) 831–858.

DOI: 10.1080/0951192x.2022.2027014

[3] C. Brecher, M. Dalibor, B. Rumpe, K. Schilling, A. Wortmann, A. An ecosystem for digital shadows in manufacturing. Procedia CIRP, 104 (2021) 833–838.

DOI: 10.1016/j.procir.2021.11.140

[4] B. He, K.-J. Bai, J. Ren. Digital twin-based sustainable intelligent manufacturing: A review. Advances in Manufacturing, 9/1 (2021) 1–21.

[5] Á. Bárkányi, T. Chován, S.Németh, J. Abonyi. Modelling for digital twins—Potential role of surrogate models and uncertainty quantification. Processes, 9/3 (2021) 476.

DOI: 10.3390/pr9030476

[6] J. Kang, J. Wang, X. Han, Q. Zhao. Deep learning based heat transfer simulation of the casting process. Scientific Reports, 14 (2024) 29068.

DOI: 10.1038/s41598-024-80515-x

[7] Q. Zhao, B. Wang, J. Kang. A PIKAN-based model for the prediction of the temperature fields of castings. Scientific Reports (2025) in-press.

DOI: 10.1038/s41598-025-32973-0

[8] Z. Lu, N. Ren, X. Xu, J. Li, C. Panwisawas, M. Xia, H. Dong, E. Tsang, J. Li. Real-time prediction and adaptive adjustment of continuous casting based on deep learning. Communications Engineering, 2 (2023) 34.

DOI: 10.1038/s44172-023-00084-1

[9] D. Mery. Aluminum casting inspection using deep learning: A method based on convolutional neural networks. Journal of Nondestructive Evaluation, 39/1 (2020) 12.

DOI: 10.1007/s10921-020-0655-9

[10] İ. E. Parlak, E. Emel. Deep learning-based detection of aluminum casting defects and their types. Engineering Applications of Artificial Intelligence, 118 (2023) 105636.

DOI: 10.1016/j.engappai.2022.105636

[11] L. Jiang, Y. Wang, Z. Tang, Y. Miao, S. Chen. Casting defect detection in X-ray images using convolutional neural networks and attention-guided data augmentation. Measurement, 170 (2021) 108736.

DOI: 10.1016/j.measurement.2020.108736

[12] Y. Zhang, Z. Gao, J. Sun, L. Liu. Machine-Learning Algorithms for Process Condition Data-Based Inclusion Prediction in Continuous-Casting Process: A Case Study. Sensors, 23/15 (2023) 6719.

DOI: 10.3390/s23156719

[13] J. Nieves, B. Bravo, D.-C. Sierra. A Smart Digital Twin to Stabilize Return Sand Temperature without Using Coolers. Metals, 12/5 (2022) 730.

DOI: 10.3390/met12050730

[14] D. A. Howard, M. Værbak, Z. Ma, B. N. Jørgensen, Z. Ma. Data-driven digital twin for foundry production process: Facilitating best practice operations investigation and impact analysis. In: Energy Informatics: 4th Energy Informatics Academy Conference (EI.A 2024), Proceedings, Part I, Lecture Notes in Computer Science, 15271 (2025) 259–273.

DOI: 10.1007/978-3-031-74738-0_17

[15] T. Bauernhansl, S. Hartleif, T. Felix. The digital shadow of production—A concept for the effective and efficient information supply in dynamic industrial environments. Procedia CIRP, 72 (2018) 69–74.

DOI: 10.1016/j.procir.2018.03.188

[16] D. Liu, Y. Du, W. Chai, C. Lu, M. Cong. Digital twin and data-driven quality prediction of complex die-casting manufacturing. IEEE Transactions on Industrial Informatics, 18/11 (2022) 8119–8128.

DOI: 10.1109/tii.2022.3168309

[17] A. Shafyei, S. H. M. Anijdan, A. Bahrami. Prediction of porosity percent in Al–Si casting alloys using ANN. Materials Science and Engineering: A, 431/1–2 (2006) 206–210.

DOI: 10.1016/j.msea.2006.05.150

[18] S. Shahane, N. Aluru, P. Ferreira, S. G. Kapoor, S. P. Vanka. Optimization of solidification in die casting using numerical simulations and machine learning. Journal of Manufacturing Processes, 51 (2020) 130–141.

DOI: 10.1016/j.jmapro.2020.01.016

[19] Z. Jiang, C. Xu, J. Liu, W. Luo, Z. Chen, W. Gui. A dual closed-loop digital twin construction method for optimizing the copper disc casting process. IEEE/CAA Journal of Automatica Sinica, 11/3 (2024) 581–594.

DOI: 10.1109/jas.2023.123777

[20] A. Ktari, M. El Mansori. Digital twin of functional gating system in 3D printed molds for sand casting using a neural network. Journal of Intelligent Manufacturing, 33/3 (2022) 897–909.

DOI: 10.1007/s10845-020-01699-3

[21] https://www.bayrammetal.com.tr/uploads/docs/en-ab-and-ac-44000.pdf.

[22] J. Campbell. Complete Casting Handbook: Metal Casting Processes, Metallurgy, Techniques and Design, 2nd edition, Butterworth-Heinemann,2015.

[23] G. C. Vosniakos, A. Vassiliou, S. Tsekouras. Numerical simulation of sand casting of an aluminium part. In: B. Katalinic (ed), Annals of DAAAM for 2011 & Proceedings of the 22nd International DAAAM Symposium, Vol. 22, No. 1, November 2011, 445-446, Danube Adria Association for Automation and Manufacturing (DAAAM), Vienna, Austria.

DOI: 10.2507/22nd.daaam.proceedings.221

[24] A. N. Vasileiou, G.-C. Vosniakos, D.I. Pantelis. Determination of local heat transfer coefficients in precision castings by genetic optimisation aided by numerical simulation. Proc. Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science 229/4 (2015) 735-750.

DOI: 10.1177/0954406214539468

[25] N. Kovachki, Z. Li, B. Liu, K. Azizzadenesheli, K. Bhattacharya, A. Stuart, A. Anandkumar. Neural Operator: Learning Maps Between Function Spaces, Journal of Machine Learning Research, 24/89 (2023) 1–97.

DOI: 10.52202/068431-1220

[26] G. Baruffa, A. Pieressa, M. Sorgato, G. Lucchetta. Transfer learning-based artificial neural network for predicting weld line occurrence through process simulations and molding trials. Journal of Manufacturing and Materials Processing, 8/3 (2024) 98.

DOI: 10.3390/jmmp8030098