Prediction of the Mechanical Properties of the Additive Friction Stir-Deposited Al2219 Using Machine Learning

Abstract

Additive friction stir deposition (AFSD) is an effective method for fabricating high-performance deposits, with process parameters directly influencing the mechanical properties of the resulting samples. In this study, three machine learning models, that is, multilayer perception (MLP), radial basis function (RBF), and back propagation (BP), are developed to predict the ultimate tensile strength (UTS) and elongation (EL) of AFSD Al2219 samples. The input variables include set parameters (rotation speed, traverse speed, layer thickness, and the presence or absence of a preheating system as well as data obtained from an in situ process monitoring kit (temperature, feedstock force, deposition interface force, and deposition interface torque)). Results show that the BP-trained neural network provides the best fit to the experimental data, achieving the highest coefficient of determination (R² = 0.821 for UTS and 0.817 for EL), the lowest mean absolute error (MAE = 8.692 for UTS and 1.003 for EL), and the lowest root mean square error (RMSE = 13.773 for UTS and 1.266 for EL). These findings demonstrate the effectiveness and advantages of BP-trained neural networks in predicting mechanical properties based on various input parameters. Recommendations are provided on how the prediction model can be applied in the field of additive manufacturing.

Keywords

additive friction stir deposition (AFSD) / in situ process monitoring kit / machine learning / mechanical properties / prediction

References

[1] D. Svetlizky, M. Das, B. Zheng, et al., “Directed Energy Deposition (DED) Additive Manufacturing: Physical Characteristics, Defects, Challenges and Applications,” Materials Today 49 (2021): 271–295, https://doi.org/10.1016/j.mattod.2021.03.020.

[2] V. Errico, A. Fusco, and S. L. Campanelli, “Effect of DED Coating and DED+ Laser Scanning on Surface Performance of L-PBF Stainless Steel Parts,” Surface and Coatings Technology 429 (2022): 127965, https://doi.org/10.1016/j.surfcoat.2021.127965.

[3] Z. Y. Hang, M. E. Jones, G. W. Brady, et al., “Non-Beam-Based Metal Additive Manufacturing Enabled by Additive Friction Stir Deposition,” Scripta Materialia 153 (2018): 122–130, https://doi.org/10.1016/j.scriptamat.2018.03.025.

[4] M. B. Williams, T. W. Robinson, C. J. Williamson, et al., “Elucidating the Effect of Additive Friction Stir Deposition on the Resulting Microstructure and Mechanical Properties of Magnesium Alloy WE43,” Metals 11, no. 11 (2021): 1739, https://doi.org/10.3390/met11111739.

[5] H. Ghadimi, H. Ding, S. Emanet, et al., “Hardness Distribution of Al2050 Parts Fabricated Using Additive Friction Stir Deposition,” Materials 16, no. 3 (2023): 1278, https://doi.org/10.3390/ma16031278.

[6] Q. Qiao, M. Zhou, X. Gong, et al., “In-Situ Monitoring of Additive Friction Stir Deposition of AA6061: Effect of Layer Thickness on the Microstructure and Mechanical Properties,” Additive Manufacturing 84 (2024): 104141, https://doi.org/10.1016/j.addma.2024.104141.

[7] N. Zhu, D. Z. Avery, Y. Chen, et al., “Residual Stress Distributions in AA6061 Material Produced by Additive Friction Stir Deposition,” Journal of Materials Engineering and Performance 32, no. 12 (2023): 5535–5544, https://doi.org/10.1007/s11665-022-07483-z.

[8] M. Mozaffar, A. Paul, R. Al-Bahrani, et al., “Data-Driven Prediction of the High-Dimensional Thermal History in Directed Energy Deposition Processes via Recurrent Neural Networks,” Manufacturing Letters 18 (2018): 35–39, https://doi.org/10.1016/j.mfglet.2018.10.002.

[9] O. A. Mohamed, S. H. Masood, and J. L. Bhowmik, “Investigation of Dynamic Elastic Deformation of Parts Processed by Fused Deposition Modeling Additive Manufacturing,” Advances in Production Engineering & Management 13, no. 3 (2016): 227–238, https://doi.org/10.14743/apem2016.3.223.

[10] Z. Zhan and H. Li, “Machine Learning Based Fatigue Life Prediction With Effects of Additive Manufacturing Process Parameters for Printed SS 316L,” International Journal of Fatigue 142 (2021): 105941, https://doi.org/10.1016/j.ijfatigue.2020.105941.

[11] A. Mishra, “Supervised Machine Learning and Physics-Based Machine Learning Approach for Prediction of Peak Temperature Distribution in Additive Friction Stir Deposition of Aluminium Alloy,” arXiv preprint arXiv:2309.06838 (2023), https://arxiv.org/abs/2309.06838.

[12] Q. Qiao, Q. Liu, J. Pu, et al., “A Comparative Study of Machine Learning in Predicting the Mechanical Properties of the Deposited AA6061 Alloys via Additive Friction Stir Deposition,” Materials Genome Engineering Advances 2, no. 1 (2024): e31, https://doi.org/10.1002/mgea.31.

[13] Q. Qiao, L. Wang, C. W. Tam, et al., “In-Situ Monitoring of Additive Friction Stir Deposition of AA6061: Effect of Rotation Speed on the Microstructure and Mechanical Properties,” Materials Science and Engineering: A 902 (2024): 146620, https://doi.org/10.1016/j.msea.2024.146620.

[14] G. Schuurmann, R. U. Ebert, J. Chen, B. Wang, and R. Kuhne, “External Validation and Prediction Employing the Predictive Squared Correlation Coefficient Test Set Activity Mean vs Training Set Activity Mean,” Journal of Chemical Information and Modeling 48, no. 11 (2008): 2140–2145, https://doi.org/10.1021/ci800253u.

[15] H. Li, L. Y. Hong, J. X. He, X. G. Xu, and J. Sun, “Small Sample-Oriented Case-based Kernel Predictive Modeling and Its Economic Forecasting Applications Under n-splits-k-times hold-out Assessment,” Economic Modelling 33 (2013): 747–761, https://doi.org/10.1016/j.econmod.2013.05.007.

[16] T. Chai and R. R. Draxler, “Root Mean Square Error (RMSE) or Mean Absolute Error (MAE)-Arguments Against Avoiding RMSE in the Literature,” Geoscientific Model Development 7, no. 3 (2014): 1247–1250, https://doi.org/10.5194/gmd-7-1247-2014.

[17] X. Qi, G. Chen, and Y. Li, “Applying Neural-Network-Based Machine Learning to Additive Manufacturing: Current Applications, Challenges, and Future Perspectives,” Engineer 5, no. 4 (2019): 721–729, https://doi.org/10.1016/j.eng.2019.04.012.

[18] C. J. Willmott and K. Matsuura, “Advantages of the Mean Absolute Error (MAE) Over the Root Mean Square Error (RMSE) in Assessing Average Model Performance,” Climate Research 30, no. 1 (2005): 79–82, https://doi.org/10.3354/cr030079.

[19] A. B. Khoshaim, E. B. Moustafa, O. T. Bafakeeh, and A. H. Elsheikh, “An Optimized Multilayer Perceptrons Model Using Grey Wolf Optimizer to Predict Mechanical and Microstructural Properties of Friction Stir Processed Aluminum Alloy Reinforced by Nanoparticles,” Coatings 11, no. 12 (2021): 1476, https://doi.org/10.3390/coatings11121476.

[20] K. Shan, Y. Zhang, Y. Lan, K. Jiang, G. Xiao, and B. Li, “Surface Roughness Prediction of Titanium Alloy During Abrasive Belt Grinding Based on an Improved Radial Basis Function (RBF) Neural Network,” Materials 16, no. 22 (2023): 7224, https://doi.org/10.3390/ma16227224.

[21] A. H. Alavi, A. H. Gandomi, M. Gandomi, and S. S. Sadat Hosseini, “Prediction of Maximum Dry Density and Optimum Moisture Content of Stabilised Soil Using RBF Neural Networks,” IES Journal Part A: Civil & Structural Engineering 2, no. 2 (2009): 98–106, https://doi.org/10.1080/19373260802659226.

[22] H. Wang, B. Li, and F. Z. Xuan, “Acoustic Emission for In Situ Process Monitoring of Selective Laser Melting Additive Manufacturing Based on Machine Learning and Improved Variational Modal Decomposition,” International Journal of Advanced Manufacturing Technology 122, no. 5–6 (2022): 2277–2292, https://doi.org/10.1007/s00170-022-10032-6.

[23] L. Liu, J. Chen, and L. Xu, “Realization and Application Research of BP Neural Network Based on MATLAB,” in 2008 International Seminar on Future BioMedical Information Engineering (IEEE, 2008), 130–133.

[24] A. Panyafong, N. Neamsorn, and C. Chaichana, “Heat Load Estimation Using Artificial Neural Network,” Energy Report 6 (2020): 742–747, https://doi.org/10.1016/j.egyr.2019.11.149.

[25] T. Jia-li, L. Yi-jun, and W. Fang-sheng, “Levenberg-Marquardt Neural Network for Gear Fault Diagnosis,” in 2010 International Conference on Networking and Digital Society, Vol. 1 (IEEE, 2010), 134–137.

[26] Z. Jiao, C. Ji, S. Yuan, Z. Zhang, and Q. Wang, “Development of Machine Learning Based Prediction Models for Hazardous Properties of Chemical Mixtures,” Journal of Loss Prevention in the Process Industries 67 (2020): 104226, https://doi.org/10.1016/j.jlp.2020.104226.

[27] I. Yilmaz and O. Kaynar, “Multiple Regression, ANN (RBF, MLP) and ANFIS Models for Prediction of Swell Potential of Clayey Soils,” Expert Systems with Applications 38, no. 5 (2011): 5958–5966, https://doi.org/10.1016/j.eswa.2010.11.027.

[28] M. Motahari-Nezhad and S. M. Jafari, “Comparison of MLP and RBF Neural Networks for Bearing Remaining Useful Life Prediction Based on Acoustic Emission,” Proceedings of the Institution of Mechanical Engineers - Part J: Journal of Engineering Tribology 237, no. 1 (2023): 129–148, https://doi.org/10.1177/13506501221106556.

[29] M. Feyzolahpour, M. Rajabi, and S. Roostaei, “Estimating Suspended Sediment Concentration Using Neural Differential Evolution (NDE), Multi Layer Perceptron (MLP) and Radial Basis Function (RBF) Models,” International Journal of the Physical Sciences 7, no. 29 (2012): 5106–5117, https://doi.org/10.5897/ijps12.269.

[30] Z. W. Dou, M. X. Ji, M. Wang, and Y. N. Shao, “Price Prediction of Pu’er Tea Based on ARIMA and BP Models,” Neural Computing & Applications 34 (2022): 3495–3511, https://doi.org/10.1007/s00521-021-05827-9.

[31] X. C. Zhong, R. K. Xie, S. H. Qin, and K. S. Zhang, “A Process-Data-Driven BP Neural Network Model for Predicting Interval-Valued Fatigue Life of Metals,” Engineering Fracture Mechanics 276 (2022): 108918, https://doi.org/10.1016/j.engfracmech.2022.108918.

RIGHTS & PERMISSIONS

2026 The Author(s). Materials Genome Engineering Advances published by Wiley-VCH GmbH on behalf of University of Science and Technology Beijing.