Predicting length of fatigue cracks by means of machine learning algorithms in the small-data regime

• Autorzy: Badora Maciej , Sepe Marzia , Bielecki Marcin , Graziano Antonino , Szolc Tomasz

Identyfikatory

DOI

10.17531/ein.2021.3.19

• Języki publikacji: EN

Abstrakty

EN

In this paper several statistical learning algorithms are used to predict the maximal length of fatigue cracks based on a sample composed of 31 observations. The small-data regime is still a problem for many professionals, especially in the areas where failures occur rarely. The analyzed object is a high-pressure Nozzle of a heavy-duty gas turbine. Operating parameters of the engines are used for the regression analysis. The following algorithms are used in this work: multiple linear and polynomial regression, random forest, kernel-based methods, AdaBoost and extreme gradient boosting and artificial neural networks. A substantial part of the paper provides advice on the effective selection of features. The paper explains how to process the dataset in order to reduce uncertainty; thus, simplifying the analysis of the results. The proposed loss and cost functions are custom and promote solutions accurately predicting the longest cracks. The obtained results confirm that some of the algorithms can accurately predict maximal lengths of the fatigue cracks, even if the sample is small.

Słowa kluczowe

EN

empirical models; fatigue cracks; predictive maintenance; regression analysis; small data; statistical learning; turbomachinery

• Wydawca: Polskie Naukowo-Techniczne Towarzystwo Eksploatacyjne
• Czasopismo: Eksploatacja i Niezawodność
• Rocznik: 2021
• Tom: Vol. 23, no. 3
• Strony: 575--585
• Opis fizyczny: Bibliogr. 45 poz., rys., tab.

Twórcy

autor

Badora Maciej

• Institute of Fundamental Technological Research, Polish Academy of Sciences, ul. Pawińskiego 5b, 02-106 Warsaw, Poland
• Baker Hughes, BH Poland sp. z o.o., Aleja Krakowska 110/114, 02-256 Warsaw, Poland

autor

Sepe Marzia

• Baker Hughes, Via Felice Matteucci 2, 50127 Florence, Italy

autor

Bielecki Marcin

• Baker Hughes, BH Poland sp. z o.o., Aleja Krakowska 110/114, 02-256 Warsaw, Poland

autor

Graziano Antonino

• Baker Hughes, Via Felice Matteucci 2, 50127 Florence, Italy

autor

Szolc Tomasz

• Institute of Fundamental Technological Research, Polish Academy of Sciences, ul. Pawińskiego 5b, 02-106 Warsaw, Poland

Bibliografia

• 1. Abernethy RB. An Overview Of Weibull Analysis. The New Weibull Handbook: Reliability & Statistical Analysis for Predicting Life, Safety, Risk, Support Costs, Failures, and Forecasting Warranty Claims, Substantiation and Accelerated Testing, Usin Weibull, Log Normal, Crow-AMSAA, Probit, and Kaplan-Meier Models, 5th edition. North Palm Beach, Florida, Robert B. Abernethy: 2004: 13-24.
• 2. Allegorico C, Mantini V. A Data-Driven Approach for on-line Gas Turbine Combustion Monitoring using Classification Models. PHM Society European Conference 2014, https://doi.org/10.36001/phme.2014.v2i1.1461.
• 3. Awad M, Khanna R. Support Vector Regression. In Awad M, Khanna R (eds): Efficient Learning Machines: Theories, Concepts, and Applications for Engineers and System Designers. Berkeley, CA, Apress: 2015: 67-80, https://doi.org/10.1007/978-1-4302-5990-9_4.
• 4. Beden S, Abdullah S, Ariffin AK. Review of Fatigue Crack Propagation Models for Metallic Components. European Journal of Scientific Research 2009.
• 5. Breiman L. Random Forests. Machine Learning 2001; 45(1): 5-32, https://doi.org/10.1023/A:1010933404324.
• 6. Carlevaro F, Cioncolini S, Sepe M et al. Use of Operating Parameters, Digital Replicas and Models for Condition Monitoring and Improved Equipment Health. Proceedings of the ASME Turbo Expo 2018: Turbomachinery Technical Conference and Exposition, Oslo, Norway, American Society of Mechanical Engineers Digital Collection: 2018, https://doi.org/10.1115/GT2018-76849.
• 7. Chen T, Guestrin C. XGBoost: A Scalable Tree Boosting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, New York, NY, USA, ACM: 2016: 785-794, https://doi.org/10.1145/2939672.2939785.
• 8. Keras. [https://keras.io/].
• 9. Dourado AD, Viana F. Physics-Informed Neural Networks for Bias Compensation in Corrosion-Fatigue. AIAA Scitech 2020 Forum, Orlando, FL, American Institute of Aeronautics and Astronautics: 2020. doi:10.2514/6.2020-1149, https://doi.org/10.2514/6.2020-1149.
• 10. Dresia K, Waxenegger-Wilfing G, Riccius J et al. Numerically Efficient Fatigue Life Prediction of Rocket Combustion Chambers using Artificial Neural Networks. 2019.
• 11. Drucker H. Improving Regressors Using Boosting Techniques. Proceedings of the 14th International Conference on Machine Learning 1997.
• 12. Durodola JF, Ramachandra S, Gerguri S, Fellows NA. Artificial neural network for random fatigue loading analysis including the effect of mean stress. International Journal of Fatigue 2018; 111: 321-332, https://doi.org/10.1016/j.ijfatigue.2018.02.007.
• 13. Escobedo E, Arguello L, Sepe M et al. Enhanced Early Warning Diagnostic Rules for Gas Turbines Leveraging on Bayesian Networks. Proceedings of the ASME Turbo Expo 2020: Turbomachinery Technical Conference and Exposition, American Society of Mechanical Engineers Digital Collection: 2021. doi:10.1115/GT2020-16082, https://doi.org/10.1115/GT2020-16082.
• 14. Faraway JJ, Augustin NH. When small data beats big data. Statistics & Probability Letters 2018; 136: 142-145, https://doi.org/10.1016/j.spl.2018.02.031.
• 15. Gan L, Zhao X, Wu H, Zhong Z. Estimation of remaining fatigue life under two-step loading based on kernel-extreme learning machine. International Journal of Fatigue 2021; 148: 106190, https://doi.org/10.1016/j.ijfatigue.2021.106190.
• 16. Harris CR, Millman KJ, van der Walt SJ et al. Array programming with NumPy. Nature 2020; 585(7825): 357-362, https://doi.org/10.1038/s41586-020-2649-2.
• 17. Iannitelli M, Allegorico C, Garau F, Capanni M. A Hybrid Model for on-line Detection of Gas Turbine Lean Blowout Events. PHM Society European Conference 2018, https://doi.org/10.36001/phme.2018.v4i1.405.
• 18. James G, Witten D, Hastie T, Tibshirani R. Linear Model Selection and Regularization. In James G, Witten D, Hastie T, Tibshirani R (eds): An Introduction to Statistical Learning: with Applications in R, New York, NY, Springer: 2013: 203-264, https://doi.org/10.1007/978-1-4614-7138-7_6.
• 19. James G, Witten D, Hastie T, Tibshirani R. Statistical Learning. In James G, Witten D, Hastie T, Tibshirani R (eds): An Introduction to Statistical Learning: with Applications in R, New York, NY, Springer: 2013: 15-57, https://doi.org/10.1007/978-1-4614-7138-7_2.
• 20. Jimenez-Martinez M, Alfaro-Ponce M. Fatigue damage effect approach by artificial neural network. International Journal of Fatigue 2019; 124: 42-47, https://doi.org/10.1016/j.ijfatigue.2019.02.043.
• 21. Kalayci CB, Karagoz S, Karakas Ö. Soft computing methods for fatigue life estimation: A review of the current state and future trends. Fatigue & Fracture of Engineering Materials & Structures 2020; 43(12): 2763-2785, https://doi.org/10.1111/ffe.13343.
• 22. Kalombo RB, Pestana MS, Freire Júnior RCS et al. Fatigue life estimation of an all aluminium alloy 1055 MCM conductor for different mean stresses using an artificial neural network. International Journal of Fatigue 2020; 140: 105814, https://doi.org/10.1016/j.ijfatigue.2020.105814.
• 23. Kamath C, Fan Y J. Regression with small data sets: a case study using code surrogates in additive manufacturing. Knowledge and Information Systems 2018; 57(2): 475-493, https://doi.org/10.1007/s10115-018-1174-1.
• 24. Karagiannopoulos M, Anyfantis D, Kotsiantis S B, Pintelas P E. Feature Selection for Regression Problems. Proceedings of HERCMA 2007, Athens, 2007.
• 25. Kingma DP, Ba J. Adam: A Method for Stochastic Optimization. 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, San Diego, CA, USA, 2015.
• 26. Kitchin R, Lauriault T. Small data in the era of big data. GeoJournal 2015; 80: 463-475, https://doi.org/10.1007/s10708-014-9601-7.
• 27. Liu X, Athanasiou CE, Padture NP et al. A machine learning approach to fracture mechanics problems. Acta Materialia 2020; 190: 105-112, https://doi.org/10.1016/j.actamat.2020.03.016.
• 28. Martens HA, Dardenne P. Validation and verification of regression in small data sets. Chemometrics and Intelligent Laboratory Systems 1998; 44(1): 99-121, https://doi.org/10.1016/S0169-7439(98)00167-1.
• 29. McKinney W. Data Structures for Statistical Computing in Python. Proceedings of the 9th Python in Science Conference, Austin, Texas, 2010; 445: 56-61, https://doi.org/10.25080/Majora-92bf1922-00a.
• 30. Michelassi V, Allegorico C, Cioncolini S et al. Machine Learning in Gas Turbines. Mechanical Engineering 2018; 140(09): S54-S55, https://doi.org/10.1115/1.2018-SEP5.
• 31. Nowell D, Nowell PW. A machine learning approach to the prediction of fretting fatigue life. Tribology International 2020; 141: 105913,https://doi.org/10.1016/j.triboint.2019.105913.
• 32. Paris P, Erdogan F. A Critical Analysis of Crack Propagation Laws. Journal of Basic Engineering 1963; 85(4): 528-533, https://doi.org/10.1115/1.3656900.
• 33. Pawełczyk M, Fulara S, Sepe M et al. Industrial gas turbine operating parameters monitoring and data-driven prediction. Eksploatacja i Niezawodnosc - Maintenance and Reliability 2020; 22: 391-399, https://doi.org/10.17531/ein.2020.3.2.
• 34. Pedregosa F, Varoquaux G, Gramfort A et al. Scikit-learn: Machine Learning in Python. Journal of Machine Learning Research 2011; 12 2825-2860.
• 35. Raissi M, Perdikaris P, Karniadakis GE. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics 2018; 378: 686-707, https://doi.org/10.1016/j.jcp.2018.10.045.
• 36. Rege K, Lemu H. A review of fatigue crack propagation modelling techniques using FEM and XFEM. IOP Conference Series Materials Science and Engineering 2017; 276: 012027, https://doi.org/10.1088/1757-899X/276/1/012027.
• 37. Rovinelli A, Sangid MD, Proudhon H, Ludwig W. Using machine learning and a data-driven approach to identify the small fatigue crack driving force in polycrystalline materials. npj Computational Materials 2018; 4(1): 1-10, https://doi.org/10.1038/s41524-018-0094-7.
• 38. Shanmugam M. Baker Hughes Company LLC: 2015.
• 39. Storn R, Price K. Differential Evolution - A Simple and Efficient Heuristic for Global Optimization over Continuous Spaces. Journal of Global Optimization 1997; 11: 341-359, https://doi.org/10.1023/A:1008202821328.
• 40. Virtanen P, Gommers R, Oliphant TE et al. SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python. Nature Methods 2020; 17(3): 261-272, https://doi.org/10.1038/s41592-019-0686-2.
• 41. Wang B, Xie L, Song J et al. Curved fatigue crack growth prediction under variable amplitude loading by artificial neural network. International Journal of Fatigue 2021; 142: 105886, https://doi.org/10.1016/j.ijfatigue.2020.105886.
• 42. Xu Y, Goodacre R. On Splitting Training and Validation Set: A Comparative Study of Cross-Validation, Bootstrap and Systematic Sampling for Estimating the Generalization Performance of Supervised Learning. Journal of Analysis and Testing 2018; 2(3): 249-262, https://doi.org/10.1007/s41664-018-0068-2.
• 43. Zhan Z, Li H. Machine learning based fatigue life prediction with effects of additive manufacturing process parameters for printed SS 316L. International Journal of Fatigue 2021; 142: 105941, https://doi.org/10.1016/j.ijfatigue.2020.105941.
• 44. Travel Weather Averages (Weatherbase). [https://www.weatherbase.com/].
• 45. XGBoost Documentation - xgboost 1.4.0-SNAPSHOT documentation. [https://xgboost.readthedocs.io/].

Uwagi

Opracowanie rekordu ze środków MNiSW, umowa Nr 461252 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2021).