Examination of 3D printing parameters using machine learning

Abstract

In this study, the mechanical properties of tensile samples produced in 3D printers with the fused deposition method (FDM) were investigated. Here, the parameters such as layer (filament) thickness, infill type and support angle in the FDM method were examined. The production was produced with Up-right and edge directions. As a result of the experiments, the best layer thickness in terms of tensile strength was 0.09 mm, and the infill type was full infill type, while different results were obtained in the support angle. According to the variance analysis (ANOVA) values, it was observed that the layer thickness and infill type were quite effective on the tensile strength, but the support angle was at a negligible level. In the second stage, the results were estimated with xgboost and catboost from the machine learning algorithms and linear regression models. The most effective algorithm on the examined mechanical properties was determined as the catboost algorithm.

Keywords

• Fused deposition model
• Mechanical properties
• Machine learning
• Regression
• Prediction

Makine öğrenmesi ile 3D yazdırma parametrelerinin incelenmesi

Öz

Bu çalışmada ergiyik yığma metodu (Fused Deposition Modelleme/FDM) yöntemi ile 3D yazıcılarda üretilen çekme numunelerinin mekanik özellikleri incelendi. Burada FDM yöntemindeki katman (filament) kalınlığı, dolgu tipi ve support açısı gibi parametreler incelendi. Üretim Up-right ve edge yönleri ve her bir yön için Taguchi L25 deney tasarımıyla üretildi. Deneyler neticesinde çekme mukavemeti açısından en iyi katman kalınlığı 0,09 mm, dolgu tipi olarak full dolgu tipi olurken support açısında farklı sonuçlar elde edildi. Varyans analizi (ANOVA) değerlerine göre parametrelerden katman kalınlığı ve dolgu tipinin çekme mukavemeti üzerinde oldukça etkili olduğu ancak support açısının göz ardı edilebilecek düzeyde olduğu gözlemlendi. İkinci aşamada sonuçlar makine öğrenmesi algoritmalarından xgboost ve catboost ile ve linear regression ile tahmin modelleri yapıldı. İncelenen mekanik özellikler üzerinde en etkin algoritma catboost algoritması olarak belirlendi.

Anahtar Kelimeler

• Ergiyik yığma metodu
• Mekanik özellikler
• Makine öğrenmesi
• Regresyon
• Tahmin

Kaynakça

1. [1] J. Cantrell vd., “Experimental Characterization of the Mechanical Properties of 3D-Printed ABS and Polycarbonate Parts Nomenclature 3D = Three-dimensional AM = Additive manufacturing ABS = Acrylonitrile butadiene styrene ASTM = American Society for Testing and Materials CA”, Proceedings of the 2016 Annual Conference on Experimental and Applied Mechanics, ss. 89–105, 2107.
2. [2] A. K. Sood, R. K. Ohdar, ve S. S. Mahapatra, “Parametric appraisal of mechanical property of fused deposition modelling processed parts”, Materials and Design, c. 31, sayı 1, ss. 287–295, 2010, doi: 10.1016/j.matdes.2009.06.016.
3. [3] H. L. Tekinalp vd., “High modulus biocomposites via additive manufacturing: Cellulose nanofibril networks as ‘microsponges’”, Composites Part B: Engineering, c. 173, s. 106817, Eyl. 2014, doi: 10.1016/j.compositesb.2019.05.028.
4. [4] F. Ning, W. Cong, Y. Hu, ve H. Wang, “Additive manufacturing of carbon fiber-reinforced plastic composites using fused deposition modeling: Effects of process parameters on tensile properties”, Journal of Composite Materials, c. 51, sayı 4, ss. 451–462, 2017, doi: 10.1177/0021998316646169.
5. [5] N. Sa’ude, S. H. Masood, M. Nikzad, ve M. Ibrahim, “Dynamic Mechanical Properties of Copper-ABS Composites for FDM Feedstock”, International Journal of Engineering Research and Applications (IJERA), c. 3, sayı 3, ss. 1257–1263, 2013, [Çevrimiçi]. Available at: http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:Dynamic+Mechanical+Properties+of+Copper-ABS+Composites+for+FDM+Feedstock#0
6. [6] S. Bacak, H. Varol Özkavak, ve M. Tatlı, “FDM Yöntemi ile Üretilen PLA Numunelerin Çekme Özelliklerine İşlem Parametrelerinin Etkisinin İncelenmesi”, Mühendislik Bilimleri ve Tasarım Dergisi, c. 9, sayı 1, ss. 209–216, Mar. 2021, doi: 10.21923/jesd.750264.
7. [7] H. Evlen, “Doluluk Oranının 3B Yazıcıda Üretilen TPU ve TPE Numunelerinin Mekanik Özellikleri Üzerine Etkilerinin İncelenmesi”, Deu Muhendislik Fakultesi Fen ve Muhendislik, c. 21, sayı 63, ss. 793–804, 2019, doi: 10.21205/deufmd.2019216310.
8. [8] M. Günay, S. Gündüz, H. Yılmaz, N. Yaşar, ve R. Kaçar, “PLA Esaslı Numunelerde Çekme Dayanımı İçin 3D Baskı İşlem Parametrelerinin Optimizasyonu”, Politeknik Dergisi, c. 23, sayı 1, ss. 73–79, Mar. 2020, doi: 10.2339/politeknik.422795.

9. [9] H. K. Sezer, O. Eren, H. R. Börklü, ve V. Özdemir, “karbon fiber takviyeli polimer kompozitlerin ergiyik biriktirme yöntemi ile eklemeli imalatı: fiber oranı ve yazdırma parametrelerinin mekanik özelliklere etkisi”, Gazi Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi, c. 2018, sayı 2018, ss. 663–674, Nis. 2018, doi: 10.17341/gazimmfd.416523.
10. [10] S. Demir, “The effect of printing parameters on hardness in the production of Poli lactic acid (PLA)-based samples with a 3D printer”, Pamukkale University Journal of Engineering Sciences, c. 30, sayı 2, ss. 136–144, 2024, doi: 10.5505/pajes.2023.49404.
11. [11] I. Malashin, V. Tynchenko, A. Gantimurov, V. Nelyub, ve A. Borodulin, “Boosting-Based Machine Learning Applications in Polymer Science: A Review”, Polymers, c. 17, sayı 4, ss. 1–42, 2025, doi: 10.3390/polym17040499.
12. [12] L. Zhang ve D. Jánošík, “Enhanced short-term load forecasting with hybrid machine learning models: CatBoost and XGBoost approaches”, Expert Systems with Applications, c. 241, sayı August 2023, 2024, doi: 10.1016/j.eswa.2023.122686.
13. [13] H. Mesghali, B. Akhlaghi, N. Gozalpour, J. Mohammadpour, F. Salehi, ve R. Abbassi, “Predicting maximum pitting corrosion depth in buried transmission pipelines: Insights from tree-based machine learning and identification of influential factors”, Process Safety and Environmental Protection, c. 187, sayı September 2023, ss. 1269–1285, 2024, doi: 10.1016/j.psep.2024.05.014.
14. [14] S. S. Dhaliwal, A. Al Nahid, ve R. Abbas, “Effective intrusion detection system using XGBoost”, Information (Switzerland), c. 9, sayı 7, 2018, doi: 10.3390/info9070149.
15. [15] S. Thongsuwan, S. Jaiyen, A. Padcharoen, ve P. Agarwal, “ConvXGB: A new deep learning model for classification problems based on CNN and XGBoost”, Nuclear Engineering and Technology, c. 53, sayı 2, ss. 522–531, 2021, doi: 10.1016/j.net.2020.04.008.
16. [16] R. Zabin, K. F. Haque, ve A. Abdelgawad, “PredXGBR: A Machine Learning Framework for Short-Term Electrical Load Prediction †”, Electronics (Switzerland), c. 13, sayı 22, ss. 1–26, 2024, doi: 10.3390/electronics13224521.
17. [17] J. Tian vd., “Synergetic Focal Loss for Imbalanced Classification in Federated XGBoost”, IEEE Transactions on Artificial Intelligence, c. 5, sayı 2, ss. 647–660, 2024, doi: 10.1109/TAI.2023.3254519.
18. [18] S. Alsulamy, “Predicting construction delay risks in Saudi Arabian projects: A comparative analysis of CatBoost, XGBoost, and LGBM”, Expert Systems with Applications, c. 268, sayı March 2024, s. 126268, 2025, doi: 10.1016/j.eswa.2024.126268.
19. [19] H. Zhuo, T. Li, W. Lu, Q. Zhang, L. Ji, ve J. Li, “Prediction model for spontaneous combustion temperature of coal based on PSO-XGBoost algorithm”, Scientific reports, c. 15, sayı 1, s. 2752, 2025, doi: 10.1038/s41598-025-87035-2.
20. [20] Y. Ueki, N. Seko, ve Y. Maekawa, “Machine learning approach for prediction of the grafting yield in radiation-induced graft polymerization”, Applied Materials Today, c. 25, s. 101158, 2021, doi: 10.1016/j.apmt.2021.101158.
21. [21] Z. A. Cevik, K. Ozsoy, A. Ercetin, ve G. Sariisik, “Machine Learning-Driven Optimization of Machining Parameters Optimization for Cutting Forces and Surface Roughness in Micro-Milling of AlSi10Mg Produced by Powder Bed Fusion Additive Manufacturing”, Applied Sciences (Switzerland), c. 15, sayı 12, ss. 1–23, 2025, doi: 10.3390/app15126553.
22. [22] X. Wang, L. Wen, ve R. Chai, “Optimization of Laser Additive Manufacturing Process Based on XGBoost Algorithm”, Journal of The Institution of Engineers (India): Series C, c. 105, sayı 6, ss. 1581–1590, 2024, doi: 10.1007/s40032-024-01119-y.
23. [23] L. Wang, W. Zhang, Q. Bao, ve Q. Wang, “XGBoost-based failure prediction method for metal additive manufacturing equipment”, Proceedings of the 34th Chinese Control and Decision Conference, CCDC 2022, ss. 3059–3064, 2022, doi: 10.1109/CCDC55256.2022.10034052.
24. [24] S. Rahmani Dabbagh, O. Ozcan, ve S. Tasoglu, “Machine learning-enabled optimization of extrusion-based 3D printing”, Methods, c. 206, sayı June 2022, ss. 27–40, 2022, doi: 10.1016/j.ymeth.2022.08.002.
25. [25] X. Zhu, F. Jiang, C. Guo, Z. Wang, T. Dong, ve H. Li, “Prediction of melt pool shape in additive manufacturing based on machine learning methods”, Optics and Laser Technology, c. 159, sayı June 2022, s. 108964, 2023, doi: 10.1016/j.optlastec.2022.108964.
26. [26] Z. Zhang, Z. Liu, ve D. Wu, “Prediction of melt pool temperature in directed energy deposition using machine learning”, Additive Manufacturing, c. 37, sayı October 2020, s. 101692, 2021, doi: 10.1016/j.addma.2020.101692.
27. [27] Y. Huang, X. Tian, Z. Zheng, D. Li, A. V. Malakhov, ve A. N. Polilov, “Multiscale concurrent design and 3D printing of continuous fiber reinforced thermoplastic composites with optimized fiber trajectory and topological structure”, Composite Structures, c. 285, sayı September 2021, s. 115241, 2022, doi: 10.1016/j.compstruct.2022.115241.
28. [28] J. Zhang, C. Yin, Y. Xu, ve S. L. Sing, “Machine learning applications for quality improvement in laser powder bed fusion: A state-of-the-art review”, International Journal of AI for Materials and Design, c. 1, sayı 1, s. 26, 2024, doi: 10.36922/ijamd.2301.
29. [29] D. Veeman, S. Sudharsan, G. J. Surendhar, R. Shanmugam, ve L. Guo, “Machine learning model for predicting the hardness of additively manufactured acrylonitrile butadiene styrene”, Materials Today Communications, c. 35, sayı March, s. 106147, 2023, doi: 10.1016/j.mtcomm.2023.106147.
30. [30] A. Garg ve K. Tai, “An ensemble approach of machine learning in evaluation of mechanical property of the rapid prototyping fabricated prototype”, Applied Mechanics and Materials, c. 575, ss. 493–496, 2014, doi: 10.4028/www.scientific.net/AMM.575.493.
31. [31] R. Agarwal, J. Singh, ve V. Gupta, “Predicting the compressive strength of additively manufactured PLA-based orthopedic bone screws: A machine learning framework”, Polymer Composites, c. 43, sayı 8, ss. 5663–5674, 2022, doi: 10.1002/pc.26881.
32. [32] J. Zhu vd., “Surface quality prediction and quantitative evaluation of process parameter effects for 3D printing with transfer learning-enhanced gradient-boosting decision trees”, Expert Systems with Applications, c. 237, sayı PB, s. 121478, 2024, doi: 10.1016/j.eswa.2023.121478.
33. [33] J. M. Barrios ve P. E. Romero, “Decision tree methods for predicting surface roughness in fused deposition modeling parts”, Materials, c. 12, sayı 16, 2019, doi: 10.3390/ma12162574.
34. [34] A. Boschetto, V. Giordano, ve F. Veniali, “Surface roughness prediction in fused deposition modelling by neural networks”, International Journal of Advanced Manufacturing Technology, c. 67, sayı 9–12, ss. 2727–2742, 2013, doi: 10.1007/s00170-012-4687-x.
35. [35] Z. Li, Z. Zhang, J. Shi, ve D. Wu, “Prediction of surface roughness in extrusion-based additive manufacturing with machine learning”, Robotics and Computer-Integrated Manufacturing, c. 57, sayı October 2018, ss. 488–495, Haz. 2019, doi: 10.1016/j.rcim.2019.01.004.
36. [36] N. Hooda, J. S. Chohan, R. Gupta, ve R. Kumar, “Deposition angle prediction of Fused Deposition Modeling process using ensemble machine learning”, ISA Transactions, c. 116, ss. 121–128, 2021, doi: 10.1016/j.isatra.2021.01.035.
37. [37] A. Douard, C. Grandvallet, F. Pourroy, ve F. Vignat, “An Example of Machine Learning Applied in Additive Manufacturing”, IEEE International Conference on Industrial Engineering and Engineering Management, c. 2019-December, ss. 1746–1750, 2018, doi: 10.1109/IEEM.2018.8607275.