2023, Volume 23, Issue 4
Engineering >> 2023, Volume 23, Issue 4 doi: 10.1016/j.eng.2022.09.015
Predictions of Additive Manufacturing Process Parameters and Molten Pool Dimensions with a Physics-Informed Deep Learning Model
School of Mechanical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China
Next Previous
Abstract
Molten pool characteristics have a significant effect on printing quality in laser powder bed fusion (PBF), and quantitative predictions of printing parameters and molten pool dimensions are critical to the intelligent control of the complex processes in PBF. Thus far, bidirectional predictions of printing parameters and molten pool dimensions have been challenging due to the highly nonlinear correlations involved. To
address this issue, we integrate an experiment on molten pool characteristics, a mechanistic model, and deep learning to achieve both forward and inverse predictions of key parameters and molten pool characteristics during laser PBF. The experiment provides fundamental data, the mechanistic model significantly augments the dataset, and the multilayer perceptron (MLP) deep learning model predicts the molten pool dimensions and process parameters based on the dataset built from the experiment and the mechanistic model. The results show that bidirectional predictions of the molten pool dimensions and process parameters can be realized, with the highest prediction accuracies approaching 99.9% and mean prediction accuracies of over 90.0%. Moreover, the prediction accuracy of the MLP model is closely related to the characteristics of the dataset—that is, the learnability of the dataset has a crucial impact on the prediction accuracy. The highest prediction accuracy is 97.3% with enhancement of the dataset via the mechanistic model, while the highest prediction accuracy is 68.3% when using only the experimental dataset. The prediction accuracy of the MLP model largely depends on the quality of the dataset as well. The research results demonstrate that bidirectional predictions of complex correlations using MLP are feasible for laser PBF, and offer a novel and useful framework for the determination of process conditions and outcomes for intelligent additive manufacturing.
Keywords
Additive manufacturing ; Molten pool ; Model ; Deep learning ; Learnability
Figures
Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16
Fig. 17
References
[ 1 ] DebRoy T, Wei HL, Zuback JS, Mukherjee T, Elmer JW, Milewski JO, et al. Additive manufacturing of metallic components—process, structure and properties. Prog Mater Sci 2018;92:112‒224. link1
[ 2 ] Wei HL, Mukherjee T, Zhang W, Zuback JS, Knapp GL, De A, et al. Mechanistic models for additive manufacturing of metallic components. Prog Mater Sci 2021;116:100703. link1
[ 3 ] Shi R, Khairallah SA, Roehling TT, Heo TW, McKeown JT, Matthews MJ. Microstructural control in metal laser powder bed fusion additive manufacturing using laser beam shaping strategy. Acta Mater 2020;184:284‒305. link1
[ 4 ] Wei HL, Knapp GL, Mukherjee T, DebRoy T. Three-dimensional grain growth during multi-layer printing of a nickel-based alloy inconel 718. Addit Manuf 2019;25:448‒59. link1
[ 5 ] Wei HL, Elmer JW, DebRoy T. Three-dimensional modeling of grain structure evolution during welding of an aluminum alloy. Acta Mater 2017;126:413‒25. link1
[ 6 ] DebRoy T, Mukherjee T, Wei HL, Elmer JW, Milewski JO. Metallurgy, mechanistic models and machine learning in metal printing. Nat Rev Mater 2021;6(1):48‒68. link1
[ 7 ] Cao Y, Wei HL, Yang T, Liu TT, Liao WH. Printability assessment with porosity and solidification cracking susceptibilities for a high strength aluminum alloy during laser powder bed fusion. Addit Manuf 2021;46:102103. link1
[ 8 ] Wei HL, Cao Y, Liao WH, Liu TT. Mechanisms on inter-track void formation and phase transformation during laser Powder Bed Fusion of Ti‒6Al‒4V. Addit Manuf 2020;34:101221. link1
[ 9 ] Mukherjee T, Wei HL, De A, DebRoy T. Heat and fluid flow in additive manufacturing—Part I: modeling of powder bed fusion. Comput Mater Sci 2018;150:304‒13. link1
[10] Mukherjee T, Wei HL, De A, DebRoy T. Heat and fluid flow in additive manufacturing—Part II: powder bed fusion of stainless steel, and titanium, nickel and aluminum base alloys. Comput Mater Sci 2018;150:369‒80. link1
[11] McCann R, Obeidi MA, Hughes C, McCarthy É, Egan DS, Vijayaraghavan RK, et al. In-situ sensing, process monitoring and machine control in laser powder bed fusion: a review. Addit Manuf 2021;45:102058. link1
[12] Yavari R, Riensche A, Tekerek E, Jacquemetton L, Halliday H, Vandever M, et al. Digitally twinned additive manufacturing: detecting flaws in laser powder bed fusion by combining thermal simulations with in-situ melt pool sensor data. Mater Des 2021;211:110167. link1
[13] Karniadakis GE, Kevrekidis IG, Lu L, Perdikaris P, Wang S, Yang L. Physicsinformed machine learning. Nat Rev Phys 2021;3(6):422‒40. link1
[14] Liu Q, Wu H, Paul MJ, He P, Peng Z, Gludovatz B, et al. Machine-learning assisted laser powder bed fusion process optimization for AlSi10Mg: new microstructure description indices and fracture mechanisms. Acta Mater 2020;201:316‒28. link1
[15] Xiong J, Zhang G, Hu J, Wu L. Bead geometry prediction for robotic GMAW-based rapid manufacturing through a neural network and a second-order regression analysis. J Intell Manuf 2014;25:157‒63. link1
[16] Nagesh DS, Datta GL. Prediction of weld bead geometry and penetration in shielded metal-arc welding using artificial neural networks. J Mater Process Technol 2002;123(2):303‒12. link1
[17] Le-Hong T, Lin PC, Chen JZ, Pham TDQ, Van Tran X. Data-driven models for predictions of geometric characteristics of bead fabricated by selective laser melting. J Intell Manuf 2021;9:1‒17.
[18] Caiazzo F, Caggiano A. Laser direct metal deposition of 2024 AI alloy: trace geometry prediction via machine learning. Materials 2018;11(3):444. link1
[19] Jeon I, Yang L, Ryu K, Sohn H. Online melt pool depth estimation during directed energy deposition using coaxial infrared camera, laser line scanner, and artificial neural network. Addit Manuf 2021;47:102295. link1
[20] Qi X, Chen G, Li Y, Cheng X, Li C. Applying neural-network-based machine learning to additive manufacturing: current applications, challenges, and future perspectives. Engineering 2019;5(4):721‒9. link1
[21] Sing SL, Kuo CN, Shih CT, Ho CC, Chua CK. Perspectives of using machine learning in laser powder bed fusion for metal additive manufacturing. Virtual Phys Prototyp 2021;16(3):372‒86. link1
[22] Tian C, Li T, Bustillos J, Bhattacharya S, Turnham T, Yeo J, et al. Data-driven approaches toward smarter additive manufacturing. Adv Intell Syst 2021;3(12):2100014. link1
[23] Meng L, McWilliams B, Jarosinski W, Park HY, Jung YG, Lee J, et al. Machine learning in additive manufacturing: a review. JOM 2020;72(6):2363‒77. link1
[24] Gan Z, Li H, Wolff SJ, Bennett JL, Hyatt G, Wagner GJ, et al. Data-driven microstructure and microhardness design in additive manufacturing using a self-organizing map. Engineering 2019;5(4):730‒5. link1
[25] Baturynska I, Martinsen K. Prediction of geometry deviations in additive manufactured parts: comparison of linear regression with machine learning algorithms. J Intell Manuf 2021;32:179‒200. link1
[26] Maleki E, Bagherifard S, Guagliano M. Application of artificial intelligence to optimize the process parameters effects on tensile properties of Ti‒6Al‒4V fabricated by laser powder-bed fusion. Int J Mech Mater Des 2022;18:199‒222. link1
[27] Li J, Zhou Q, Huang X, Li M, Cao L. In situ quality inspection with layer-wise visual images based on deep transfer learning during selective laser melting. J Intell Manuf 2023;34:853‒67. link1
[28] Scime L, Beuth J. A multi-scale convolutional neural network for autonomous anomaly detection and classification in a laser powder bed fusion additive manufacturing process. Addit Manuf 2018;24:273‒86. link1
[29] Caggiano A, Zhang J, Alfieri V, Caiazzo F, Gao R, Teti R. Machine learning-based image processing for on-line defect recognition in additive manufacturing. CIRP Ann 2019;68(1):451‒4. link1
[30] Mohammadi MG, Mahmoud D, Elbestawi M. On the application of machine learning for defect detection in L-PBF additive manufacturing. Opt Laser Technol 2021;143:107338. link1
[31] Schmid S, Krabusch J, Schromm T, Jieqing S, Ziegelmeier S, Grosse CU, et al. A new approach for automated measuring of the melt pool geometry in laserpowder bed fusion. Prog Addit Manuf 2021;6(2):269‒79. link1
[32] Scime L, Beuth J. Using machine learning to identify in-situ melt pool signatures indicative of flaw formation in a laser powder bed fusion additive manufacturing process. Addit Manuf 2019;25:151‒65. link1
[33] Bag S, De A, DebRoy T. A genetic algorithm-assisted inverse convective heat transfer model for tailoring weld geometry. Mater Manuf Process 2009;24(3):384‒97. link1
[34] Das D, Pratihar DK, Roy GG, Pal AR. Phenomenological model-based study on electron beam welding process, and input‒output modeling using neural networks trained by back-propagation algorithm, genetic algorithms, particle swarm optimization algorithm and bat algorithm. Appl Intell 2018;48(9):2698‒718. link1
[35] Johnson NS, Vulimiri PS, To AC, Zhang X, Brice CA, Kappes BB, et al. Invited review: machine learning for materials developments in metals additive manufacturing. Addit Manuf 2020;36:101641. link1
[36] Wang B, Tao F, Fang X, Liu C, Liu Y, Freiheit T. Smart manufacturing and intelligent manufacturing: a comparative review. Engineering 2021;7(6):738‒57. link1
[37] Gunasegaram DR, Murphy AB, Barnard A, DebRoy T, Matthews MJ, Ladani L, et al. Towards developing multiscale‒multiphysics models and their surrogates for digital twins of metal additive manufacturing. Addit Manuf 2021;46:102089. link1
[38] Wei HL, Liu FQ, Liao WH, Liu TT. Prediction of spatiotemporal variations of deposit profiles and inter-track voids during laser directed energy deposition. Addit Manuf 2020;34:101219. link1
[39] Liu FQ, Wei L, Shi SQ, Wei HL. On the varieties of build features during multilayer laser directed energy deposition. Addit Manuf 2020;36:101491. link1
[40] Wei HL, Liu FQ, Wei L, Liu TT, Liao WH. Multiscale and Multiphysics explorations of the transient deposition processes and additive characteristics during laser 3D printing. J Mater Sci Technol 2021;77:196‒208. link1
[41] Ren K, Chew Y, Liu N, Zhang YF, Fuh JYH, Bi GJ. Integrated numerical modelling and deep learning for multi-layer cube deposition planning in laser aided additive manufacturing. Virtual Phys Prototyp 2021;16(3):318‒32. link1
[42] Gu D, Ma C, Xia M, Dai D, Shi Q. A multiscale understanding of the thermodynamic and kinetic mechanisms of laser additive manufacturing. Engineering 2017;3(5):675‒84. link1
[43] Du Y, Mukherjee T, DebRoy T. Physics-informed machine learning and mechanistic modeling of additive manufacturing to reduce defects. Appl Mater Today 2021;24:101123. link1
[44] Yang T, Liu T, Liao W, MacDonald E, Wei H, Chen X, et al. The influence of process parameters on vertical surface roughness of the AlSi10Mg parts fabricated by selective laser melting. J Mater Process Technol 2019;266:26‒36. link1
[45] Yale K. Preparing the right data diet for training neural networks. IEEE Spectr 1997;34(3):64‒6. link1
[46] Wang J, Li S, An Z, Jiang X, Qian W, Ji S. Batch-normalized deep neural networks for achieving fast intelligent fault diagnosis of machines. Neurocomputing 2019;329:53‒65. link1
[47] Abadi M, Agarwal A, Barham P, Brevdo E, Chen Z, Citro C, et al. TensorFlow: large-scale machine learning on heterogeneous distributed systems. 2016. arXiv:1603.04467v2.
[48] Bircanoǧlu C, Arıca N. A comparison of activation functions in artificial neural networks. In: Proceedings of the 26th Signal Processing and Communications Applications Conference (SIU); 2018 May 2‒5; Izmir, Turkey. IEEE; 2018. p. 1‒4. link1
[49] He S, Guo F, Zou Q, Ding H. MRMD2.0: a python tool for machine learning with feature ranking and reduction. Curr Bioinform 2020;15(10):1213‒21. link1
[50] Ruder S. An overview of gradient descent optimization algorithms. 2016. arXiv:1609.04747.
[51] Wu X, Zheng W, Chen X, Zhao Y, Yu T, Mu D. Improving high-impact bug report prediction with combination of interactive machine learning and active learning. Inf Softw Technol 2021;133:106530. link1
[52] Klawonn M, Heim E, Hendler J. Exploiting class learnability in noisy data. Proc AAAI Conf Artif Intell 2019;33(1):4082‒9. link1
[53] Zheng W, Liu X, Yin L. Research on image classification method based on improved multi-scale relational network. PeerJ Comput Sci 2021;7:e613. link1
[54] Khairallah SA, Sun T, Simonds BJ. Onset of periodic oscillations as a precursor of a transition to pore-generating turbulence in laser melting. Addit Manuf Lett 2021;1:100002. link1
[55] Gan Z, Kafka OL, Parab N, Zhao C, Fang L, Heinonen O, et al. Universal scaling laws of keyhole stability and porosity in 3D printing of metals. Nat Commun 2021;12:2379. link1
[56] Khairallah SA, Martin AA, Lee JRI, Guss G, Calta NP, Hammons JA, et al. Controlling interdependent meso-nanosecond dynamics and defect generation in metal 3D printing. Science 2020;368(6491):660‒5. link1
[57] Matthews MJ, Guss G, Khairallah SA, Rubenchik AM, Depond PJ, King WE. Denudation of metal powder layers in laser powder bed fusion processes. Acta Mater 2016;114:33‒42. link1
[58] Lee S, Peng J, Shin D, Choi YS. Data analytics approach for melt-pool geometries in metal additive manufacturing. Sci Technol Adv Mater 2019;20(1):972‒8. link1
[59] Mukherjee T, Debroy T. A digital twin for rapid qualification of 3D printed metallic components. Appl Mater Today 2019;14:59‒65. link1
[60] Knapp GL, Mukherjee T, Zuback JS, Wei HL, Palmer TA, De A, et al. Building blocks for a digital twin of additive manufacturing. Acta Mater 2017;135:390‒9. link1
[61] Tao F, Qi Q, Wang L, Nee AYC. Digital twins and cyber‒physical systems toward smart manufacturing and Industry 4.0: correlation and comparison. Engineering 2019;5(4):653‒61. link1
京公网安备 11010502051620号
