YAPAY ZEKA KULLANILARAK TREN TEKERLEKLERİNİN YORULMA ÖZELLİKLERİNİN VEKİL MODELLENMESİ

Yıl 2024, Cilt: 12 Sayı: 2, 277 - 284, 30.06.2024

Mehran Mahouti Mehmet Sinan Komek Suat Yılmaz

https://doi.org/10.21923/jesd.1434972

Öz

Sonlu elemanlar yöntemi (FEM), tren tekerlekleri gibi karmaşık yapıların analiz edilmesi ve tasarlanması için mühendislikte hayati bir araçtır. Tren tekerlekleri, işletme ömürleri boyunca karşılaştıkları aşırı ve değişken yükler nedeniyle yorulmaya maruz kalmaktadır ve bu durum, ömür süresi ve güvenlik üzerindeki etkileri nedeniyle tren tekerleği tasarımında kritik bir endişe kaynağıdır. Ancak, özellikle tren tekerlekleri gibi karmaşık geometrilere sahip büyük ölçekli yapıların modellenmesinde FEM'in geniş hesaplama ihtiyaçları önemli zorluklar sunmaktadır. Doğru yorgunluk analizi için gereken detaylı modelleme, genellikle büyük hesaplama yükleri ve uzun zaman dilimleri ile sonuçlanmakta ve bu durum, hızlı karar verilmesi gereken durumlarda daha az uygulanabilir bir seçenek haline gelmektedir. Bu sınırlamaları ele almak için, Yapay Zeka (AI), yenilikçi bir çözüm olarak ortaya çıkmıştır. FEM simülasyonlarından elde edilen veri setleri üzerinde eğitilen YZ modelleri, geleneksel hesaplama maliyeti ve zamanının bir kısmında yorgunluk ömrünü tahmin ederek etkin bir alternatif sunmaktadır. Bu vekil modeller, mühendislik tasarım optimizasyonu süreçleri için gerekli olan hızlı ve doğru tahmini sağlamaktadır. Bu çalışmada YZ tabanlı vekil modelleme yaklaşımı ile tren tekerlekleri optimizasyon problemini geleneksel FEM yaklaşımına kıyas ile nerdeyse %90 oranında hızlandırma başarısına erişilmiştir.

Anahtar Kelimeler

Yapay Zeka, Optimizasyon, Vekil modelleme, Sonlu elemanlar yöntemi

SURROGATE MODELLING OF TRAIN WHEELS FATIGUE CHARACTERISTICS USING ARTIFICIAL INTELLIGENCE

Öz

The finite element method (FEM) is a vital tool in engineering for analyzing and designing complex structures such as train wheels. Train wheels are subject to fatigue due to the extreme and variable loads they encounter throughout their operating life, and this is a critical concern in train wheel design due to its effects on lifespan and safety. However, the large computational needs of FEM present significant challenges, especially in modeling large-scale structures with complex geometries such as train wheels. The detailed modeling required for accurate fatigue analysis often results in large computational loads and long time periods, making it a less feasible option in situations where rapid decisions must be made. To address these limitations, Artificial Intelligence (AI) has emerged as an innovative solution. AI models trained on data sets obtained from FEM simulations offer an effective alternative by predicting fatigue life at a fraction of the traditional computational cost and time. These surrogate models provide the fast and accurate prediction required for engineering design optimization processes. In this study, the AI-based surrogate modeling approach succeeded in accelerating the train wheels optimization problem by almost 90% compared to the traditional FEM approach.

Anahtar Kelimeler

Artificial intelligence, Optimization, Surrogate modeling, Finite element method.

Kaynakça

• Adriano, V. S. R., Martínez, J. M. G., Ferreira, J. L. A., Araújo, J. A., & Da Silva, C. R. M. (2018). The influence of the fatigue process zone size on fatigue life estimations performed on aluminum wires containing geometric discontinuities using the Theory of Critical Distances. Theoretical and Applied Fracture Mechanics, 97, 265-278.
• Andújar, Rabindranath, Jaume Roset, and Vojko Kilar. "Beyond FEM: overview on physics simulation tools for structural engineers." Technics Technologies Education Management 6, no. 3 (2011): 555-571.
• Ansys, https://www.ansys.com/. Available on (24.11.2023)
• Bhat, S., and R. Patibandla. "Metal fatigue and basic theoretical models: a review." Alloy steel-properties and use 22 (2011).
• Bian, Jian, Yuantong Gu, and Martin Howard Murray. "A dynamic wheel–rail impact analysis of railway track under wheel flat by finite element analysis." Vehicle System Dynamics 51, no. 6 (2013): 784-797.
• Bracamonte, A. J., Mercado-Puche, V., Martínez-Arguelles, G., Pumarejo, L. F., Ortiz, A. R., & Herazo, L. C. S. (2023). Effect of Finite Element Method (FEM) Mesh Size on the Estimation of Concrete Stress–Strain Parameters. Applied Sciences, 13(4), 2352.
• Cwik, Tom, Daniel S. Katz, and Jean Patterson. "Scalable solutions to integral-equation and finite-element simulations." IEEE Transactions on Antennas and Propagation 45, no. 3 (1997): 544-555.
• de Gooijer, B. M., Havinga, J., Geijselaers, H. J., & van den Boogaard, A. H. (2021). Evaluation of POD based surrogate models of fields resulting from nonlinear FEM simulations. Advanced Modeling and Simulation in Engineering Sciences, 8(1), 1-33.
• Drucker, H., Burges, C. J., Kaufman, L., Smola, A., & Vapnik, V. (1996). Support vector regression machines. Advances in neural information processing systems, 9.
• Du, X., He, P., & Martins, J. R. (2021). Rapid airfoil design optimization via neural networks-based parameterization and surrogate modeling. Aerospace Science and Technology, 113, 106701.
• Efthimeros, G. A., D. I. Photeinos, Z. G. Diamantis, and D. T. Tsahalis. "Vibration/noise optimization of a FEM railway wheel model." Engineering Computations 19, no. 8 (2002): 922-931.
• Feather, W. G., Lim, H., & Knezevic, M. (2021). A numerical study into element type and mesh resolution for crystal plasticity finite element modeling of explicit grain structures. Computational Mechanics, 67, 33-55.
• Gardner, M. W., & Dorling, S. R. (1998). Artificial neural networks (the multilayer perceptron)—a review of applications in the atmospheric sciences. Atmospheric environment, 32(14-15), 2627-2636.
• Ghiasi, R., Ghasemi, M. R., & Noori, M. (2018). Comparative studies of metamodeling and AI-Based techniques in damage detection of structures. Advances in Engineering Software, 125, 101-112.
• Güneş, F., Demirel, S., & Mahouti, P. (2016). A simple and efficient honey bee mating optimization approach to performance characterization of a microwave transistor for the maximum power delivery and required noise. International Journal of Numerical Modelling: Electronic Networks, Devices and Fields, 29(1), 4-20.
• Koziel, S., Belen, M. A., Çalişkan, A., & Mahouti, P. (2023). Rapid Design of 3D Reflectarray Antennas by Inverse Surrogate Modeling and Regularization. IEEE Access, 11, 24175-24184.
• Kraus, M. A., Bischof, R., Kaufmann, W., & Thoma, K. (2022). Artificial intelligence-finite element method-hybrids for efficient nonlinear analysis of concrete structures. Acta Polytechnica CTU Proceedings, 36, 99-108.
• Kudela, J., & Matousek, R. (2022). Recent advances and applications of surrogate models for finite element method computations: A review. Soft Computing, 26(24), 13709-13733.
• Kukulski, J., Jacyna, M., & Gołębiowski, P. (2019). Finite element method in assessing strength properties of a railway surface and its elements. Symmetry, 11(8), 1014.
• Li, Yang, Mi Zhao, Cheng-shun Xu, Xiu-li Du, and Zheng Li. "Earthquake input for finite element analysis of soil-structure interaction on rigid bedrock." Tunnelling and Underground Space Technology 79 (2018): 250-262.
• Liu, Kai, and Lin Jing. "A finite element analysis-based study on the dynamic wheel–rail contact behaviour caused by wheel polygonization." Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit 234, no. 10 (2020): 1285-1298.
• Liu, Yongming, Liming Liu, and Sankaran Mahadevan. "Analysis of subsurface crack propagation under rolling contact loading in railroad wheels using FEM." Engineering fracture mechanics 74, no. 17 (2007): 2659-2674.
• Loh, W. L. (1996). On Latin hypercube sampling. The annals of statistics, 24(5), 2058-2080.
• Long, Y. Q., Cen, S., & Long, Z. F. (2009). Advanced finite element method in structural engineering (pp. 495-586). Beijing: Tsinghua University Press.
• Mendes-Moreira, J., Soares, C., Jorge, A. M., & Sousa, J. F. D. (2012). Ensemble approaches for regression: A survey. Acm computing surveys (csur), 45(1), 1-40.
• Naboulsi, S., & Mall, S. (2003). Fretting fatigue crack initiation behavior using process volume approach and finite element analysis. Tribology international, 36(2), 121-131.
• Okereke, M., Keates, S., Okereke, M., & Keates, S. (2018). Finite element mesh generation. Finite Element Applications: A Practical Guide to the FEM Process, 165-186.
• Patel, Sunil, Veerendra Kumar, and Raji Nareliya. "Fatigue analysis of rail joint using finite element method." International Journal of Research in Engineering and Technology 2, no. 1 (2013): 80-84.
• Plevris, V., & Tsiatas, G. C. (2018). Computational structural engineering: Past achievements and future challenges. Frontiers in Built Environment, 4, 21.
• Saneie, Hamid, Ramin Alipour-Sarabi, Zahra Nasiri-Gheidari, and Farid Tootoonchian. "Challenges of finite element analysis of resolvers." IEEE Transactions on Energy Conversion 34, no. 2 (2018): 973-983.
• Schulz, E., Speekenbrink, M., & Krause, A. (2018). A tutorial on Gaussian process regression: Modelling, exploring, and exploiting functions. Journal of Mathematical Psychology, 85, 1-16.
• Sun, G., & Wang, S. (2019). A review of the artificial neural network surrogate modeling in aerodynamic design. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 233(16), 5863-5872.
• Swiler, L. P., Gulian, M., Frankel, A. L., Safta, C., & Jakeman, J. D. (2020). A survey of constrained Gaussian process regression: Approaches and implementation challenges. Journal of Machine Learning for Modeling and Computing, 1(2).
• Szabó, B., & Babuška, I. (2021). Finite Element Analysis: Method, Verification and Validation.
• Wang, J., Jiang, H., Chen, G., Wang, H., Lu, L., Liu, J., & Xing, L. (2023). Integration of multi-physics and machine learning-based surrogate modelling approaches for multi-objective optimization of deformed GDL of PEM fuel cells. Energy and AI, 14, 100261.
• Wu, S. W., Wan, D. T., Jiang, C., Liu, X., Liu, K., & Liu, G. R. (2023). A finite strain model for multi-material, multi-component biomechanical analysis with total Lagrangian smoothed finite element method. International Journal of Mechanical Sciences, 243, 108017.
• Yang, X. S., Koziel, S., & Leifsson, L. (2014). Computational optimization, modelling and simulation: Past, present and future. Procedia Computer Science, 29, 754-758.
• Yao, J., Ye, Z., & Wang, Y. (2014). An efficient SRAM yield analysis and optimization method with adaptive online surrogate modeling. IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 23(7), 1245-1253.
• Zahavi, E. (2019). Fatigue design: life expectancy of machine parts. CRC press.
• Zhang, Guanzhen, and Ruiming Ren. "Study on typical failure forms and causes of high-speed railway wheels." Engineering Failure Analysis 105 (2019): 1287-1295.
• Zhou, M., Mei, G., & Xu, N. (2023). Enhancing Computational Accuracy in Surrogate Modeling for Elastic–Plastic Problems by Coupling S-FEM and Physics-Informed Deep Learning. Mathematics, 11(9), 2016.
• Zhu, Yi, Wenjian Wang, Roger Lewis, Wenyi Yan, Stephen R. Lewis, and Haohao Ding. "A review on wear between railway wheels and rails under environmental conditions." Journal of Tribology 141, no. 12 (2019): 120801.