Artificial Neural Network Algorithms for 3D Printing
Abstract
:1. Introduction
2. Three-Dimensional (3D) Printing Processes
- Firstly, CAD software is used to develop a CAD model.
- This CAD model is converted into stereolithography format (.STL), which is the wedge-shaped drawing of a 3D CAD model.
- Then, the file is sliced into several thin cross-sectional layers, using slicing software known as a slicer.
- Finally, postprocessing steps, including surface treatments, sintering, or finishing, are usually performed [29].
2.1. Binder Jetting
2.2. Direct Energy Deposition
2.3. Material Extrusion
2.4. Material Jetting
2.5. Powder Bed Fusion
2.6. Sheet Lamination
2.7. Vat Polymerization
2.8. Color 3D Printing Technology
2.8.1. Color Measurement of Color 3D Printing
2.8.2. Color Specification of Color 3D Printing
2.8.3. Color Reproduction of Color 3D Printing
3. Introduction to the Artificial Neural Network (ANN) Algorithm
4. Applications of ANN in 3D Printing
4.1. Process Monitoring
4.2. Designing
4.3. Correlation between Process Parameters’ and Parts’ Final Characteristics
4.4. ANN for Metals’ and Polymers’ 3D Printing
5. Potential Challenges for the Implementation of ANN and Their Solutions
5.1. Datasets Optimization
5.2. Selection of Significant Input Parameters
5.3. Under- and Over-fitting in the ANN Model
5.4. Linking the Analytical Modeling and Numerical Simulations with ANN
5.5. Real-Time Monitoring of the 3D Printing Process
- (1)
- Collection of a larger number of part designs, based on different variety of geometries, topologies, and material types;
- (2)
- Increasing and/or altering the input variables/factors for evaluating the part designs based on the application requirement. However, these factors may be updated based on user requirements;
- (3)
- Developing software with a graphical user interface (GUI) that can integrate the different subsystems. It can assist both novice users in inputting their datasets as well as 3D printing specialists to modify the ANN algorithm parameters to tune their results; and
- (4)
- Collaboration with industry partners so that a joint system specific to each industry sector (such as automotive, aerospace, consumer appliances, etc.) be developed.
6. Future Outlook
7. Conclusions
- ANN involves supervised learning primarily composed of three layers: (a) an input layer, (b) a hidden layer, and (c) an output layer. Three classes, including multilayer perceptron, convolutional ANN, and recurrent ANN, have been found. The ANN structure contains four hyperparameters: (a) the number of the hidden layers, (b) neurons, (c) the activation function, and (d) the loss function. Two types of error functions have been identified in the case of the 3D printing process: (a) Tanh and (b) Sigmoid.
- ANN can be used for product designing, process monitoring, and to correlate the input parameters with the properties of the final produced part. In the 3D printing process, it is very tough to optimize operating parameters, as they are highly nonlinear in nature. This task, however, can efficiently be completed using an ANN model, owing to its nonlinear nature. In this context, convolutional ANN has proved capabilities to forecast with better precision compared to other classes of ANN. According to Table 4, 5–10 neurons in the hidden layer are suggested to determine the optimal solution in the 3D printing process.
- The performance of an ANN model depends on the quantity and type of data provided while training. Further, it is expensive and time-consuming to collect and organize the data for the training of an ANN model. Therefore, it is necessary to determine the significant set of parameters to save time and train an ANN model, effectively. It will also avoid the over- or underfitting of the ANN model. On the other, artificial datasets can be generated via analytical or numerical modeling to avoid the deficiency of datasets needed for ANN training and testing.
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Lu, S.; Zhou, L.N.; Wang, Z. Damage Evolution of Concrete by Electrical Resistivity Monitoring Methods. Appl. Mech. Mater. 2015, 796, 85–89. [Google Scholar] [CrossRef]
- Derby, B. Additive Manufacture of Ceramics Components by Inkjet Printing. Engineering 2015, 1, 113–123. [Google Scholar] [CrossRef] [Green Version]
- 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, 675–684. [Google Scholar] [CrossRef]
- Herzog, D.; Seyda, V.; Wycisk, E.; Emmelmann, C. Additive manufacturing of metals. Acta Mater. 2016, 117, 371–392. [Google Scholar] [CrossRef]
- Liu, L.; Ding, Q.; Zhong, Y.; Zou, J.; Wu, J.; Chiu, Y.L.; Li, J.; Zhang, Z.; Yu, Q.; Shen, Z. Dislocation network in additive manufactured steel breaks strength–ductility trade-off. Mater. Today 2018, 21, 354–361. [Google Scholar] [CrossRef] [Green Version]
- Mahmood, M.A.; Popescu, A.C.; Hapenciuc, C.L.; Ristoscu, C.; Visan, A.I.; Oane, M.; Mihailescu, I.N. Estimation of clad geometry and corresponding residual stress distribution in laser melting deposition: Analytical modeling and experimental correlations. Int. J. Adv. Manuf. Technol. 2020, 111, 77–91. [Google Scholar] [CrossRef]
- Acharya, R.; Sharon, J.A.; Staroselsky, A. Prediction of microstructure in laser powder bed fusion process. Acta Mater. 2017, 124, 360–371. [Google Scholar] [CrossRef]
- Fergani, O.; Berto, F.; Welo, T.; Liang, S.Y. Analytical modelling of residual stress in additive manufacturing. Fatigue Fract. Eng. Mater. Struct. 2017, 40, 971–978. [Google Scholar] [CrossRef]
- Chen, Q.; Guillemot, G.; Gandin, C.A.; Bellet, M. Three-dimensional finite element thermomechanical modeling of additive manufacturing by selective laser melting for ceramic materials. Addit. Manuf. 2017, 16, 124–137. [Google Scholar] [CrossRef] [Green Version]
- Wang, Y.; Li, X. An accurate finite element approach for programming 4D-printed self-morphing structures produced by fused deposition modeling. Mech. Mater. 2020, 151, 103628. [Google Scholar] [CrossRef]
- Wu, X.; Xu, C.; Zhang, Z.; Guo, C. Modeling and visualization of layered curing conversion profile in ceramic mask projection stereolithography process. Ceram. Int. 2020, 46, 25750–25757. [Google Scholar] [CrossRef]
- Zhang, Y.; Chou, Y.K. Three-dimensional finite element analysis simulations of the fused deposition modelling process. Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 2006, 220, 1663–1671. [Google Scholar] [CrossRef]
- Zhang, Y.; Chou, Y.K. A parametric study of part distortions in fdm using 3d fea. Proc. Inst. Mech. Eng. 2008, 222, 959–968. [Google Scholar] [CrossRef]
- Bellini, A.; Güçeri, S.; Bertoldi, M. Liquefier dynamics in fused deposition. J. Manuf. Sci. Eng. Trans. ASME 2004, 126, 237–246. [Google Scholar] [CrossRef]
- Venkataraman, N.; Rangarajan, S.; Matthewson, M.J.; Harper, B.; Safari, A.; Danforth, S.C.; Wu, G.; Langrana, N.; Guceri, S.; Yardimci, A. Feedstock material property—Process relationships in fused deposition of ceramics (FDC). Rapid Prototyp. J. 2000, 6, 244–252. [Google Scholar] [CrossRef]
- Ju, Y.; Xie, H.; Zheng, Z.; Lu, J.; Mao, L.; Gao, F.; Peng, R. Visualization of the complex structure and stress field inside rock by means of 3D printing technology. Chin. Sci. Bull. 2014, 59, 5354–5365. [Google Scholar] [CrossRef]
- Sachs, E.; Vezzetti, E. Numerical simulation of deposition process for a new 3DP printhead design. J. Mater. Process. Technol. 2005, 161, 509–515. [Google Scholar] [CrossRef]
- Curodeau, A. Three Dimensional Printing of Ceramic Molds with Accurate Surface Macro-Textures for Investment Casting of Orthopaedic Implants; Massachusetts Institute of Technology: Cambridge, MA, USA, 1995. [Google Scholar]
- Lecun, Y.; Bengio, Y.; Hinton, G. Deep learning. Nature 2015, 521, 436–444. [Google Scholar] [CrossRef]
- Krizhevsky, A.; Sutskever, I.; Hinton, G.E. ImageNet classification with deep convolutional neural networks. In Proceedings of the Advances in Neural Information Processing Systems, Lake Tahoe, NV, USA, 3–6 December 2012; Volume 2, pp. 1097–1105. [Google Scholar]
- Anusuya, M.; Katti, S. Speech recognition by machine, a review. Int. J. Comput. Sci. Inf. Secur. 2010, 6, 181–205. [Google Scholar]
- Devlin, J.; Chang, M.; Lee, K.; Toutanova, K. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the NAACL-HLT 2019, Minneapolis, MN, USA, 1 June–7 June 2019; pp. 4171–4186. [Google Scholar]
- Ondruska, P.; Posner, I. Deep tracking: Seeing beyond seeing using recurrent neural networks. In Proceedings of the AAAI-16 Conference, Phoenix, AZ, USA, 12–17 February 2016. [Google Scholar]
- ISO/ASTM 52900:2015(en), Additive Manufacturing—General Principles—Terminology. Available online: https://www.iso.org/obp/ui/#iso:std:iso-astm:52900:ed-1:v1:en (accessed on 9 December 2020).
- Mahmood, M.A.; Popescu, A.C.; Mihailescu, I.N. Metal Matrix Composites Synthesized by Laser-Melting Deposition: A Review. Materials 2020, 13, 2593. [Google Scholar] [CrossRef]
- Mahmood, M.A.; Popescu, A.C.; Oane, M.; Ristoscu, C.; Chioibasu, D.; Mihai, S.; Mihailescu, I.N. Three-Jet Powder Flow and Laser–Powder Interaction in Laser Melting Deposition: Modelling Versus Experimental Correlations. Metals 2020, 10, 1113. [Google Scholar] [CrossRef]
- Mahmood, M.A.; Han, C.F.; Chu, H.Y.; Sun, C.C.; Wu, W.H.; Lin, W.J.; Liu, L.C.; Lai, J.Y.; Mihailescu, I.N.; Lin, J.F. Effects of roll pattern and reduction ratio on optical characteristics of A1008 cold–rolled steel specimens: Analytical approach and experimental correlations. Int. J. Adv. Manuf. Technol. 2020, 111, 2001–2020. [Google Scholar] [CrossRef]
- Mahmood, M.A.; Tsai, T.Y.; Hwu, Y.J.; Lin, W.J.; Liu, L.C.; Lai, J.Y.; Pan, J.W.; Li, W.L.; Lin, J.F. Effect of fractal parameters on optical properties of cold rolled aluminum alloy strips with induced surface deflection: Simulations and experimental correlations. J. Mater. Process. Technol. 2020, 279, 116554. [Google Scholar] [CrossRef]
- Chioibasu, D.; Mihai, S.; Mahmood, M.A.; Lungu, M.; Porosnicu, I.; Sima, A.; Dobrea, C.; Tiseanu, I.; Popescu, A.C. Use of X-ray Computed Tomography for Assessing Defects in Ti Grade 5 Parts Produced by Laser Melting Deposition. Metals 2020, 10, 1408. [Google Scholar] [CrossRef]
- Ziaee, M.; Crane, N.B. Binder jetting: A review of process, materials, and methods. Addit. Manuf. 2019, 28, 781–801. [Google Scholar] [CrossRef]
- Sing, S.L.; Tey, C.F.; Tan, J.H.K.; Huang, S.; Yeong, W.Y. 3D printing of metals in rapid prototyping of biomaterials: Techniques in additive manufacturing. In Rapid Prototyping of Biomaterials: Techniques in Additive Manufacturing; Elsevier: Amsterdam, The Netherlands, 2019; pp. 17–40. ISBN 9780081026632. [Google Scholar]
- Sidambe, A. Biocompatibility of Advanced Manufactured Titanium Implants—A Review. Materials 2014, 7, 8168–8188. [Google Scholar] [CrossRef] [Green Version]
- Sireesha, M.; Lee, J.; Kranthi Kiran, A.S.; Babu, V.J.; Kee, B.B.T.; Ramakrishna, S. A review on additive manufacturing and its way into the oil and gas industry. RSC Adv. 2018, 8, 22460–22468. [Google Scholar] [CrossRef] [Green Version]
- Zhang, Y.; Jarosinski, W.; Jung, Y.G.; Zhang, J. Additive manufacturing processes and equipment. In Additive Manufacturing: Materials, Processes, Quantifications and Applications; Elsevier: Amsterdam, The Netherlands, 2018; pp. 39–51. ISBN 9780128121559. [Google Scholar]
- Safari, A.; Allahverdi, M. Electroceramics: Rapid Prototyping. In Encyclopedia of Materials: Science and Technology; Elsevier: Amsterdam, The Netherlands, 2001; pp. 2510–2518. [Google Scholar]
- MacAdam, D.L. Subtractive Color Mixture and Color Reproduction. J. Opt. Soc. Am. 1938, 28, 466. [Google Scholar] [CrossRef]
- Brunton, A.; Arikan, C.A.; Tanksale, T.M.; Urban, P. 3D printing spatially varying color and translucency. ACM Trans. Graph. 2018, 37, 1–13. [Google Scholar] [CrossRef] [Green Version]
- Shishkovsky, I. New Trends in 3D Printing; BoD—Books on Demand: Norderstedt, Germany, 2016; ISBN 9789535124795. [Google Scholar]
- Chen, C.; Chen, G.X.; Yu, Z.H.; Wang, Z.H. A new method for reproducing oil paintings based on 3D printing. Appl. Mech. Mater. 2014, 644–650, 2386–2389. [Google Scholar] [CrossRef]
- Goldberg, Y. Neural Network Methods for Natural Language Processing. Synth. Lect. Hum. Lang. Technol. 2017, 10, 1–311. [Google Scholar] [CrossRef]
- Rumelhart, D.E.; Hinton, G.E.; Williams, R.J. Learning representations by back-propagating errors. Nature 1986, 323, 533–536. [Google Scholar] [CrossRef]
- Gardner, M.W.; Dorling, S.R. Artificial neural networks (the multilayer perceptron)—A review of applications in the atmospheric sciences. Atmos. Environ. 1998, 32, 2627–2636. [Google Scholar] [CrossRef]
- Rawat, W.; Wang, Z. Deep convolutional neural networks for image classification: A comprehensive review. Neural Comput. 2017, 29, 2352–2449. [Google Scholar] [CrossRef]
- Mikolov, T.; Karafiát, M.; Burget, L.; Černocký, J.; Khudanpur, S. Recurrent Neural Network Based Language Model. In Proceedings of the 11th Annual Conference of the International Speech Communication Association, Chiba, Japan, 26–30 September 2010; ISCA Archive: Yokohama, Japan, 2010; pp. 1045–1048. [Google Scholar]
- Bre, F.; Gimenez, J.M.; Fachinotti, V.D. Prediction of wind pressure coefficients on building surfaces using artificial neural networks. Energy Build. 2018, 158, 1429–1441. [Google Scholar] [CrossRef]
- Sood, A.K.; Ohdar, R.K.; Mahapatra, S.S. Experimental investigation and empirical modelling of FDM process for compressive strength improvement. J. Adv. Res. 2012, 3, 81–90. [Google Scholar] [CrossRef] [Green Version]
- Xu, S.; Chen, L. A novel approach for determining the optimal number of hidden layer neurons for FNN’s and its application in data mining. In Proceedings of the 5th International Conference on Information Technology and Applications, Cairns, Australia, 23–26 June 2008. [Google Scholar]
- Staiano, A.; Tagliaferri, R.; Pedrycz, W. Improving RBF networks performance in regression tasks by means of a supervised fuzzy clustering. Neurocomputing 2006, 69, 1570–1581. [Google Scholar] [CrossRef]
- Acharya, U.R.; Oh, S.L.; Hagiwara, Y.; Tan, J.H.; Adam, M.; Gertych, A.; Tan, R.S. A deep convolutional neural network model to classify heartbeats. Comp. Biol. Med. 2017, 89, 389–396. [Google Scholar] [CrossRef]
- Cheng, H.D.; Xu, H. A novel fuzzy logic approach to contrast enhancement. Pattern Recogit. 2000, 33, 809–819. [Google Scholar] [CrossRef]
- Wu, S.; Chow, T.W.S. Self-organizing and self-evolving neurons: A new neural network for optimization. IEEE Trans. Neural Netw. 2007, 18, 385–396. [Google Scholar] [CrossRef]
- Joseph, J.; Pignatiello, J.R. An overview of the strategy and tactics of Taguchi. IIE Trans. 2007, 20, 247–254. [Google Scholar] [CrossRef]
- Murphey, Y.L.; Guo, H.; Feldkamp, L.A. Neural learning from unbalanced data. Appl. Integ. 2004, 21, 117–128. [Google Scholar] [CrossRef]
- Zhang, S.U. Degradation classification of 3D printing thermoplastics using Fourier transform infrared spectroscopy and artificial neural networks. Appl. Sci. 2018, 8, 1224. [Google Scholar] [CrossRef] [Green Version]
- Rong-Ji, W.; Xin-Hua, L.; Qing-Ding, W.; Lingling, W. Optimizing process parameters for selective laser sintering based on neural network and genetic algorithm. Int. J. Adv. Manuf. Technol. 2009, 42, 1035–1042. [Google Scholar] [CrossRef]
- Garg, A.; Tai, K.; Savalani, M.M. State-of-the-art in empirical modelling of rapid prototyping processes. Rapid Prototyp. J. 2014, 20, 164–178. [Google Scholar] [CrossRef]
- Abiodun, O.I.; Jantan, A.; Omolara, A.E.; Dada, K.V.; Mohamed, N.A.; Arshad, H. State-of-the-art in artificial neural network applications: A survey. Heliyon 2018, 4, e00938. [Google Scholar] [CrossRef] [Green Version]
- Ioffe, S.; Szegedy, C. Batch normalization: Accelerating deep network training by reducing internal covariate shift. In Proceedings of the 32nd International Conference on Machine Learning, ICML, Lille, France, 6–11 July 2015; International Machine Learning Society (IMLS): Washington, DC, USA, 2015; Volume 1, pp. 448–456. [Google Scholar]
- 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, 721–729. [Google Scholar] [CrossRef]
- Shen, X.; Yao, J.; Wang, Y.; Yang, J. Density prediction of selective laser sintering parts based on artificial neural network. Lect. Notes Comput. Sci. 2004, 3174, 832–840. [Google Scholar] [CrossRef]
- Li, X.; Dong, J.; Zhang, Y. Modeling and applying of RBF neural network based on fuzzy clustering and pseudo-inverse method. In Proceedings of the International Conference on Information Engineering and Computer Science, Wuhan, China, 19–20 December 2009. [Google Scholar]
- Munguía, J.; Ciurana, J.; Riba, C. Neural-network-based model for build-time estimation in selective laser sintering. Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 2009, 223, 995–1003. [Google Scholar] [CrossRef]
- Wang, R.; Gutierrez-Farewik, E.M. The effect of excessive subtalar inversion/eversion on the dynamic function of the soleus and gastrocnemius during the stance phase. In Proceedings of the ASME Summer Bioengineering Conference 2009, Lake Tahoe, CA, USA, 17–21 June 2009; American Society of Mechanical Engineers Digital Collection: New York, NY, USA, 2009; pp. 1035–1036. [Google Scholar]
- Wang, C.Y.; Jiang, N.; Chen, Z.L.; Chen, Y.; Dong, Q. Prediction of sintering strength for selective laser sintering of polystyrene using artificial neural network. J. Donghua Univ. 2015, 32, 825–830. [Google Scholar]
- Lee, S.H.; Park, W.S.; Cho, H.S.; Zhang, W.; Leu, M.C. A neural network approach to the modelling and analysis of stereolithography processes. Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 2001, 215, 1719–1733. [Google Scholar] [CrossRef]
- Caiazzo, F.; Caggiano, A. Laser Direct Metal Deposition of 2024 Al Alloy: Trace Geometry Prediction via Machine Learning. Materials 2018, 11, 444. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Sood, A.K.; Equbal, A.; Toppo, V.; Ohdar, R.K.; Mahapatra, S.S. An investigation on sliding wear of FDM built parts. CIRP J. Manuf. Sci. Technol. 2012, 5, 48–54. [Google Scholar] [CrossRef]
- Vosniakos, G.C.; Maroulis, T.; Pantelis, D. A method for optimizing process parameters in layer-based rapid prototyping. Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 2007, 221, 1329–1340. [Google Scholar] [CrossRef]
- Equbal, A.; Sood, A.K.; Ohdar, R.K.; Mahapatra, S.S. Prediction of dimensional accuracy in fused deposition modelling: A fuzzy logic approach. Int. J. Product. Qual. Manag. 2011, 7, 22–43. [Google Scholar] [CrossRef]
- Sood, A.K.; Ohdar, R.K.; Mahapatra, S.S. Parametric appraisal of fused deposition modelling process using the grey Taguchi method. Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 2010, 224, 135–145. [Google Scholar] [CrossRef]
- Chen, H.; Zhao, Y. Learning algorithm based modeling and process parameters recommendation system for binder jetting additive manufacturing process. In Proceedings of the International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, Boston, MA, USA, 2–5 August 2015; p. V01AT02A029. [Google Scholar]
- Everton, S.K.; Hirsch, M.; Stavroulakis, P.I.; Leach, R.K.; Clare, A.T. Review of in-situ process monitoring and in-situ metrology for metal additive manufacturing. Mater. Des. 2016, 95, 431–445. [Google Scholar] [CrossRef]
- Wasmer, K.; Le-Quang, T.; Meylan, B.; Shevchik, S.A. In Situ Quality Monitoring in AM Using Acoustic Emission: A Reinforcement Learning Approach. J. Mater. Eng. Perform. 2019, 28, 666–672. [Google Scholar] [CrossRef]
- Shevchik, S.A.; Kenel, C.; Leinenbach, C.; Wasmer, K. Acoustic emission for in situ quality monitoring in additive manufacturing using spectral convolutional neural networks. Addit. Manuf. 2018, 21, 598–604. [Google Scholar] [CrossRef]
- Zhang, Y.; Hong, G.S.; Ye, D.; Zhu, K.; Fuh, J.Y.H. Extraction and evaluation of melt pool, plume and spatter information for powder-bed fusion AM process monitoring. Mater. Des. 2018, 156, 458–469. [Google Scholar] [CrossRef]
- 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, 273–286. [Google Scholar] [CrossRef]
- Scime, L.; Beuth, J. Anomaly detection and classification in a laser powder bed additive manufacturing process using a trained computer vision algorithm. Addit. Manuf. 2018, 19, 114–126. [Google Scholar] [CrossRef]
- 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, 151–165. [Google Scholar] [CrossRef]
- Jafari-Marandi, R.; Khanzadeh, M.; Tian, W.; Smith, B.; Bian, L. From in-situ monitoring toward high-throughput process control: Cost-driven decision-making framework for laser-based additive manufacturing. J. Manuf. Syst. 2019, 51, 29–41. [Google Scholar] [CrossRef]
- Ye, D.; Hsi Fuh, J.Y.; Zhang, Y.; Hong, G.S.; Zhu, K. In situ monitoring of selective laser melting using plume and spatter signatures by deep belief networks. ISA Trans. 2018, 81, 96–104. [Google Scholar] [CrossRef]
- Bin Maidin, S.; Campbell, I.; Pei, E. Development of a design feature database to support design for additive manufacturing. Assem. Autom. 2012, 32, 235–244. [Google Scholar] [CrossRef] [Green Version]
- Yao, X.; Moon, S.K.; Bi, G. A hybrid machine learning approach for additive manufacturing design feature recommendation. Rapid Prototyp. J. 2017, 23, 983–997. [Google Scholar] [CrossRef]
- Shi, Y.; Zhang, Y.; Baek, S.; De Backer, W.; Harik, R. Manufacturability analysis for additive manufacturing using a novel feature recognition technique. Comput. Aided Des. Appl. 2018, 15, 941–952. [Google Scholar] [CrossRef]
- Zhang, X.; Le, X.; Panotopoulou, A.; Whiting, E.; Wang, C.C.L. Perceptual models of preference in 3D printing direction. ACM Trans. Graph. 2015, 34, 1–12. [Google Scholar] [CrossRef]
- Williams, G.; Meisel, N.A.; Simpson, T.W.; McComb, C. Design repository effectiveness for 3D convolutional neural networks: Application to additive manufacturing. J. Mech. Des. Trans. ASME 2019, 141. [Google Scholar] [CrossRef]
- Khadilkar, A.; Wang, J.; Rai, R. Deep learning–based stress prediction for bottom-up SLA 3D printing process. Int. J. Adv. Manuf. Technol. 2019, 102, 2555–2569. [Google Scholar] [CrossRef]
- Koeppe, A.; Hernandez Padilla, C.A.; Voshage, M.; Schleifenbaum, J.H.; Markert, B. Efficient numerical modeling of 3D-printed lattice-cell structures using neural networks. Manuf. Lett. 2018, 15, 147–150. [Google Scholar] [CrossRef]
- Chowdhury, S.; Anand, S. Artificial Neural Network Based Geometric Compensation for Thermal Deformation in Additive Manufacturing Processes; ASME International: New York, NY, USA, 2016. [Google Scholar]
- Meng, L.; Zhao, J.; Lan, X.; Yang, H.; Wang, Z. Multi-objective optimisation of bio-inspired lightweight sandwich structures based on selective laser melting. Virtual Phys. Prototyp. 2020, 15, 106–119. [Google Scholar] [CrossRef]
- Gu, G.X.; Chen, C.T.; Richmond, D.J.; Buehler, M.J. Bioinspired hierarchical composite design using machine learning: Simulation, additive manufacturing, and experiment. Mater. Horiz. 2018, 5, 939–945. [Google Scholar] [CrossRef] [Green Version]
- Gu, G.X.; Chen, C.T.; Buehler, M.J. De novo composite design based on machine learning algorithm. Extrem. Mech. Lett. 2018, 18, 19–28. [Google Scholar] [CrossRef]
- Zohdi, T.I. Dynamic thermomechanical modeling and simulation of the design of rapid free-form 3D printing processes with evolutionary machine learning. Comput. Methods Appl. Mech. Eng. 2018, 331, 343–362. [Google Scholar] [CrossRef]
- Gan, Z.; Li, H.; Wolff, S.J.; Bennett, J.L.; Hyatt, G.; Wagner, G.J.; Cao, J.; Liu, W.K. Data-Driven Microstructure and Microhardness Design in Additive Manufacturing Using a Self-Organizing Map. Engineering 2019, 5, 730–735. [Google Scholar] [CrossRef]
- Petrov, R.A.; Pernot, J.-P.; Giannini, F.; Falcidieno, B.; Véron, P. Mapping Aesthetic Properties to 3D Free Form Shapes through the Use of a Machine Learning Based Framework; IMATI Report Series; IMATI: Milan, Italy, 2016. [Google Scholar]
- Sarlo, R.; Tarazaga, P.A. A neural network approach to 3D printed surrogate systems. In Topics in Modal Analysis & Testing; Proceedings of the Society for Experimental Mechanics Series; Springer LLC: New York, NY, USA, 2016; Volume 10, pp. 215–222. [Google Scholar]
- Wang, R.J.; Li, J.; Wang, F.; Li, X.; Wu, Q. ANN model for the prediction of density in Selective Laser Sintering. Int. J. Manuf. Res. 2009, 4, 362–373. [Google Scholar] [CrossRef]
- Pasquet, I.; Baco-Carles, V.; Chamelot, P.; Gibilaro, M.; Massot, L.; Tailhades, P. A multimaterial based on metallic copper and spinel oxide made by powder bed laser fusion: A new nanostructured material for inert anode dedicated to aluminum electrolysis. J. Mater. Process. Technol. 2020, 278, 116452. [Google Scholar] [CrossRef]
- Yu, W.; Sing, S.L.; Chua, C.K.; Tian, X. Influence of re-melting on surface roughness and porosity of AlSi10Mg parts fabricated by selective laser melting. J. Alloys Compd. 2019, 792, 574–581. [Google Scholar] [CrossRef]
- Kuo, C.N.; Chua, C.K.; Peng, P.C.; Chen, Y.W.; Sing, S.L.; Huang, S.; Su, Y.L. Microstructure evolution and mechanical property response via 3D printing parameter development of Al–Sc alloy. Virtual Phys. Prototyp. 2020, 15, 120–129. [Google Scholar] [CrossRef]
- Ding, D.; Pan, Z.; Cuiuri, D.; Li, H.; Van Duin, S.; Larkin, N. Bead modelling and implementation of adaptive MAT path in wire and arc additive manufacturing. Robot. Comput. Integr. Manuf. 2016, 39, 32–42. [Google Scholar] [CrossRef]
- Ding, D.; Shen, C.; Pan, Z.; Cuiuri, D.; Li, H.; Larkin, N.; Van Duin, S. Towards an automated robotic arc-welding-based additive manufacturing system from CAD to finished part. CAD Comput. Aided Des. 2016, 73, 66–75. [Google Scholar] [CrossRef] [Green Version]
- Khaw, J.F.C.; Lim, B.S.; Lim, L.E.N. Optimal design of neural networks using the Taguchi method. Neurocomputing 1995, 7, 225–245. [Google Scholar] [CrossRef]
- 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, 157–163. [Google Scholar] [CrossRef]
- Mohamed, O.A.; Masood, S.H.; Bhowmik, J.L. Investigation of dynamic elastic deformation of parts processed by fused deposition modeling additive manufacturing. Apem J. 2016, 13, 227–238. [Google Scholar] [CrossRef] [Green Version]
- Bayraktar, Ö.; Uzun, G.; Çakiroğlu, R.; Guldas, A. Experimental study on the 3D-printed plastic parts and predicting the mechanical properties using artificial neural networks. Polym. Adv. Technol. 2017, 28, 1044–1051. [Google Scholar] [CrossRef]
- Asadi-Eydivand, M.; Solati-Hashjin, M.; Fathi, A.; Padashi, M.; Abu Osman, N.A. Optimal design of a 3D-printed scaffold using intelligent evolutionary algorithms. Appl. Soft Comput. J. 2016, 39, 36–47. [Google Scholar] [CrossRef]
- Zhu, Y.; Wu, Z.; Hartley, W.D.; Sietins, J.M.; Williams, C.B.; Yu, H.Z. Unraveling pore evolution in post-processing of binder jetting materials: X-ray computed tomography, computer vision, and machine learning. Add. Manuf. 2020, 34, 101183. [Google Scholar] [CrossRef]
- Fang, Z.; Wang, R.; Wang, M.; Zhong, S.; Ding, L.; Chen, S. Effect of reconsideration algorithm on the identification of 3D printing polymers on hyperspectral CT technology combined with artificial neural network. Materials 2020, 13, 1963. [Google Scholar] [CrossRef]
- Boillat, E.; Kolossov, S.; Glardon, R.; Loher, M.; Saladin, D.; Levy, G. Finite element and neural network models for process optimization in selective laser sintering. Proc. Inst. Mech. Eng. Part B J. Eng. Manf. 2004, 218, 607–614. [Google Scholar] [CrossRef]
- Ahmadi, A.; Mirzaeifar, R.; Moghaddam, N.S.; Turabi, A.S.; Karaca, H.E.; Elahinia, M. Effect of manufacturing parameters on mechanical properties of 316L stainless steel parts fabricated by selective laser melting: A computational framework. Mater. Des. 2016, 112, 328–338. [Google Scholar] [CrossRef]
- Snell, R.; Tammas-Williams, S.; Chechik, L.; Lyle, A.; Hernández-Nava, E.; Boig, C.; Panoutsos, G.; Todd, I. Methods for Rapid Pore Classification in Metal Additive Manufacturing. JOM 2020, 72, 101–109. [Google Scholar] [CrossRef] [Green Version]
- Cho, H.S.; Park, W.S.; Choi, B.W.; Leu, M.C. Determining optimal parameters for stereolithography processes via genetic algorithm. J. Manuf. Syst. 2000, 19, 18–27. [Google Scholar] [CrossRef]
- He, H.; Yang, Y.; Pan, Y. Machine learning for continuous liquid interface production: Printing speed modelling. J. Manuf. Syst. 2019, 50, 236–246. [Google Scholar] [CrossRef]
- Shamsaei, N.; Yadollahi, A.; Bian, L.; Thompson, S.M. An overview of direct laser deposition for additive manufacturing; part II: Mechanical behavior, process parameter optimization and control. Addit. Manu. 2015, 8, 12–35. [Google Scholar] [CrossRef]
- Saqib, S.; Urbanic, R.J.; Aggarwal, K. Analysis of laser cladding bead morphology for developing additive manufacturing travel paths. In Proceedings of the 47th CIRP Conference on Manufacturing Systems, Windsor, UK, 28–30 April 2014; pp. 824–829. [Google Scholar]
- Zhang, W.; Mehta, A.; Desai, P.S.; Fred Higgs, C., III. Machine Learning Enabled Powder Spreading Process Map for Metal Additive Manufacturing (AM). In Proceedings of the 28th Annual International Solid Freeform Fabrication Symposium, Austin, TX, USA, 7–9 August 2017. [Google Scholar]
- Tak, J.; Kantemur, A.; Sharma, Y.; Xin, H. A 3-D-printed W-band slotted waveguide array antenna optimized using machine learning. IEEE Antennas Wirel. Propag. Lett. 2018, 17, 2008–2012. [Google Scholar] [CrossRef]
- Wang, T.; Kwok, T.H.; Zhou, C.; Vader, S. In-situ droplet inspection and closed-loop control system using machine learning for liquid metal jet printing. J. Manuf. Syst. 2018, 47, 83–92. [Google Scholar] [CrossRef] [Green Version]
- Kwon, O.; Kim, H.G.; Ham, M.J.; Kim, W.; Kim, G.H.; Cho, J.H.; Kim, N., II; Kim, K. A deep neural network for classification of melt-pool images in metal additive manufacturing. J. Intell. Manuf. 2020, 31, 375–386. [Google Scholar] [CrossRef]
- Pant, M.; Singari, R.M.; Arora, P.K.; Moona, G.; Kumar, H. Wear assessment of 3-D printed parts of PLA (polylactic acid) using Taguchi design and Artificial Neural Network (ANN) technique. Mater. Res. Express 2020, 7, 115307. [Google Scholar] [CrossRef]
- Ahmadi Dastjerdi, A.; Movahhedy, M.R.; Akbari, J. Optimization of process parameters for reducing warpage in selected laser sintering of polymer parts. Addit. Manuf. 2017, 18, 285–294. [Google Scholar] [CrossRef]
- Fountas, N.A.; Kostazos, P.; Pavlidis, H.; Antoniou, V.; Manolakos, D.E.; Vaxevanidis, N.M. Experimental investigation and statistical modelling for assessing the tensile properties of FDM fabricated parts. In Procedia Structural Integrity; Elsevier B.V.: Amsterdam, The Netherlands, 2020; Volume 26, pp. 139–146. [Google Scholar]
- Fountas, N.A.; Kechagias, J.D.; Manolakos, D.E.; Vaxevanidis, N.M. Single and multi-objective optimization of FDM-based additive manufacturing using metaheuristic algorithms. Procedia Manuf. 2020, 51, 740–747. [Google Scholar] [CrossRef]
- Géron, A. Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems; O’Reilly Media, Inc.: Sebastopol, CA, USA, 2019; ISBN 9781492032618. [Google Scholar]
- Kingma, D.P.; Welling, M. Auto-encoding variational bayes. In Proceedings of the 2nd International Conference on Learning Representations, ICLR 2014—Conference Track Proceedings, Banff, AB, Canada, 14–16 April 2014. [Google Scholar]
- Goodfellow, I.J.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; Bengio, Y. Generative Adversarial Nets. arXiv 2014, arXiv:1406.2661v1. [Google Scholar]
- Makhzani, A.; Shlens, J.; Jaitly, N.; Goodfellow, I.; Frey, B. Adversarial Autoencoders. arXiv 2015, arXiv:1511.05644v2. [Google Scholar]
- Yusuf, S.M.; Gao, N. Influence of energy density on metallurgy and properties in metal additive manufacturing. Mater. Sci. Technol. 2017, 33, 1269–1289. [Google Scholar] [CrossRef]
- Srivastava, N.; Hinton, G.; Krizhevsky, A.; Sutskever, I.; Salakhutdinov, R. Dropout: A simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 2014, 15, 1929–1958. [Google Scholar]
- Ng, A.Y. Feature selection, L1 vs. L2 regularization, and rotational invariance. In Proceedings of the Twenty-First International Conference on Machine Learning, ICML 2004, Banff, AB, Canada, 4–8 July 2004; ACM Press: New York, NY, USA, 2004; pp. 615–622. [Google Scholar]
- Online Materials Information Resource—MatWeb. Available online: http://www.matweb.com/ (accessed on 17 November 2020).
Sr. No. | 3D Printing Processes | Process Illustration |
---|---|---|
1 | Binder jetting | A liquid bonding agent is selectively poured to join powder materials |
2 | Direct energy deposition | Focused thermal energy (laser, electron, or plasma arc) is used to fuse materials by melting as they are being deposited |
3 | Material extrusion | Material is selectively dispensed through a nozzle or orifice |
4 | Material jetting | Droplets of building material are selectively deposited |
5 | Powder bed fusion | Thermal energy selectively fuses regions of a powder bed |
6 | Sheet lamination | Sheets of material are bonded to form a part |
7 | Vat photopolymerization | Liquid photopolymer in a vat is selectively cured by light-activated polymerization |
Multilayer Perceptron | The most typical ANN, common in linear and nonlinear summations, e.g., sigmoid functions. It is frequently used for the data in the tabular [42]. |
Convolutional ANN (CANN) | Deliberates the relationship between image pixels and is therefore used in image processing [43]. |
Recurrent ANN | Has a significant role in temporal dynamics, as it can build a connection between the number of nodes in a given layer. Therefore, it is used in long short-term memory that can regenerate the simulation results accurately [44]. |
No. of hidden layers | A combination of the input-hidden-output layer with a respective number of neurons is used to define the ANN model [46]. For instance, 5-8-1 means that the input layer has 5 neurons, 8 neurons in the hidden layer, while the output layer has only 1 neuron. | |
No. of neurons in a given layer | The required number of neurons in the input and output layer depends on the particular problem. However, the number of neurons must be chosen carefully. If the number of neurons is not selected optimally, it will lead to under- or overfitting in the given dataset [47]. Various studies have been carried out that suggest choosing a number of neurons between 5 and 10 [48,49,50,51,52,53,54,55,56,57]. | |
Activation function | A nonlinear transformation on a given input signal. In other words, it decides whether to activate and deactivate a particular neuron. Based on its performance, the activation function is considered a vital part of an ANN model. It is important to mention that a network deprived of an activation function behaves like a linear regression model, which cannot deal with the complicated tasks [58]. A few activation functions are: | |
(1) | ||
(2) | ||
(3) | ||
Loss function | The loss function is usually determined using real-world problems and carries the interpretation of real-time data. The root means square (RMS) and the absolute mean error (AME) are two commonly used methods to estimate the difference between the predicted vector and the target value [59]. Their mathematical expressions are given as: | |
(4) | ||
(5) | ||
where i is a sample index, xi is a predicted value, and xt is the target value. |
3D Printing Process | Input-Hidden-Output Layers | Activation Function | Error Between Absolute Output and Anticipated Output (%) | References |
---|---|---|---|---|
Selective laser sintering | 4-9-1 | Sigmoid | 6.99 | [60] |
4-6-1 | 1.06 | [61] | ||
3-7-1 | 15.01 | [62] | ||
7-7-1 | 4.36 | [63] | ||
5-27-1 | 0.91 | [64] | ||
7-8-1 | 0.90 | [64] | ||
3-9-1 | Tanh | 0.50 | [56] | |
Stereolithography | 6-20-5 | Sigmoid | 5.98 | [65] |
Laser-melting deposition | 3-9-3 | Sigmoid | 3.0 | [66] |
Fused deposition modeling | 5-8-1 | Tanh | 1.99 | [46] |
5-8-1 | 1.02 | [67] | ||
4-15-12-1 | Sigmoid | 5.2 | [68] | |
5-6-4 | 4.08 | [69] | ||
5-7-3 | 0.11 | [70] | ||
Binder jetting | 4-20-1 | Sigmoid | 0.40 | [71] |
Features | ANN Technique | Remarks | References |
---|---|---|---|
Composite design | Linear model and CANN |
| [90,91] |
Process planning | Genetic algorithm (GA) and classical gradient-based schemes |
| [92] |
Design feature recommendation | Hierarchical clustering and SVM |
| [82] |
Tuning microstructure and microhardness | Self-organizing map |
| [93] |
Optimize build orientation concerning build time and part mass | 10-layer CANN and linear Regression model |
| [85] |
Flatness perception | Classification tree (C4.5) |
| [94] |
Geometric compensation | Feedforward ANN |
| [88] |
Part orientation | DLEANN |
| [84] |
Designing surrogate systems | ANN |
| [95] |
Composite design | CANN |
| [90,91] |
Stress prediction | 2-Stream CANN |
| [86] |
3D Printing Process | ANN Input Parameters | ANN Output Parameters | References |
---|---|---|---|
Selective laser sintering | Laser power, scanning speed, hatch spacing, powder layer thickness | Density | [60] |
Laser power, scanning speed, hatch spacing, powder layer thickness | Geometrical dimensions | [61] | |
Vertical height, deposited volume, bounding box | Manufacturing time | [62] | |
Laser power, scanning speed, hatch distance, powder layer thickness, scanning mode, temperature distribution, the processing time | Shrinkage percentage | [63] | |
Powder layer thickness, laser power, scanning speed | Part porosity | [64] | |
Laser power, scanning speed, hatch distance, powder layer thickness, temperature distribution | Tensile strength | [64] | |
Laser power, scanning speed, hatch distance, powder layer thickness, scanning mode, temperature distribution, the processing time | Density | [96] | |
Stereolithography | Powder layer thickness, curation time, hatch distance, filling cure depth, filling spacing depth | Geometrical dimensions (precision) | [65] |
Laser-melting deposition | Laser power, scanning speed, powder feed rate | Geometrical dimensions (precision) | [66] |
Fused deposition modeling | Layer thickness, positioning, raster angle and width, air gap | Compressive strength | [46] |
Layer thickness, positioning, raster angle and width, air gap | Wear | [67] | |
Positioning, slice width | Deposition error in volume | [68] | |
Layer thickness, positioning, raster angle and width, air gap | Dimensional precision | [69,70] | |
Binder jetting | Layer thickness, printing saturation, heater power ration, drying time | Surface roughness, shrinkage | [71] |
Process | Purpose | ANN Method | Input Parameters | References |
---|---|---|---|---|
Binder Jetting | Predicting surface roughness, shrinkage rate in y- and z-directions | 3-layer BP-ANN | Layer thickness, printing saturation, heater power ratio, drying time | [53] |
Compressive strength, open porosity | Aggregated ANN | Orientation, layer thickness, delay time | [106] | |
To characterize defects evolution | Gaussian mixture model | To reduce pore decomposition, shrinkage and smoothing during post-processing | [107] | |
Selective laser sintering | Dimension | Radial basic function ANN, fuzzy C-means, and pseudoinverse method, k-means | Laser power, scan speed, scan spacing, layer thickness | [48] |
Material analysis | ANN | Structural characterization | [108] | |
Shrinkage ratio | ANN | Laser power, scan speed, hatch spacing, layer thickness, scan mode, temperature, interval time | [55] | |
Tensile strength | ANN | Laser power, scan speed, hatch spacing, layer thickness, powder temperature | [109] | |
Density | ANN | Laser power, scan speed, hatch spacing, layer thickness, scan mode, temperature, interval time | [110] | |
Selective laser melting | Keyhole porosity | K-means clustering | Energy density | [111] |
Stereolithography | Dimensional accuracy | ANN | Layer thickness, border overcure, hatch overcure, fill cure depth, fill spacing, and hatch spacing | [112] |
Printability | Ensemble method, Siamese network | Printability | [113] | |
Laser-melting deposition | Geometrical accuracy | ANN | Laser power, scanning speed, powder feeding rate | [114] |
Melt-pool width | ANN | Laser power, powder feed rate, laser speed, focal length, contact tip to workpiece distance | [115] | |
Electron beam melting | Volume, roughness | ANN | Spreader translation speed, rotation speed | [116] |
Feature Selection | Illustration |
This technique assists in determining the most influencing parameters from a given list using statistical tools. | To determine the parameter significantly affecting the printing process, a Pearson’s coefficient can be determined to figure out the dependency between the given parameters on output. If Pearson’s value (max = 1) is higher for one parameter compared to the other parameter, it will affect the desired output significantly. |
Feature Combination | Illustration |
This technique helps to carry out dimensionality lessening for input attributes and thereby concentrate on the newly generated features. Once the translation regulation is identified, manual manipulations are usually preferred. Mathematical tools such as principal components analysis can be utilized for the same purpose based on the attribute. | Energy density (ED) influences the solidification, metallurgical, microstructure, and mechanical properties of a 3D-printed part. Laser power, scanning speed, hatch distance, and layer thickness combine and generate a new ED feature. |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Mahmood, M.A.; Visan, A.I.; Ristoscu, C.; Mihailescu, I.N. Artificial Neural Network Algorithms for 3D Printing. Materials 2021, 14, 163. https://doi.org/10.3390/ma14010163
Mahmood MA, Visan AI, Ristoscu C, Mihailescu IN. Artificial Neural Network Algorithms for 3D Printing. Materials. 2021; 14(1):163. https://doi.org/10.3390/ma14010163
Chicago/Turabian StyleMahmood, Muhammad Arif, Anita Ioana Visan, Carmen Ristoscu, and Ion N. Mihailescu. 2021. "Artificial Neural Network Algorithms for 3D Printing" Materials 14, no. 1: 163. https://doi.org/10.3390/ma14010163
APA StyleMahmood, M. A., Visan, A. I., Ristoscu, C., & Mihailescu, I. N. (2021). Artificial Neural Network Algorithms for 3D Printing. Materials, 14(1), 163. https://doi.org/10.3390/ma14010163