Data-driven approach to predict the fatigue properties of ferrous metal materials using the cGAN and machine-learning algorithms

Abstract

The stress-life curve (S–N) and low-cycle strain-life curve (E–N) are the two primary representations used to characterize the fatigue behavior of a material. These material fatigue curves are essential for structural fatigue analysis. However, conducting material fatigue tests is expensive and time-intensive. To address the challenge of data limitations on ferrous metal materials, we propose a novel method that utilizes the Random Forest Algorithm and transfer learning to predict the S–N and E–N curves of ferrous materials. In addition, a data-augmentation framework is introduced using a conditional generative adversarial network (cGAN) to overcome data deficiencies. By incorporating the cGAN-generated data, the accuracy (R ²) of the Random Forest Algorithm-trained model is improved by 0.3–0.6. It is proven that the cGAN can significantly enhance the prediction accuracy of the machine-learning model and balance the cost of obtaining fatigue data from the experiment.

Keywords

Fatigue life curve / Machine learning / Transfer learning / Conditional generative adversarial network (cGAN)

References

[1.] Long X, Lu C, Su Y, et al. Machine learning framework for predicting the low cycle fatigue life of lead-free solders. Eng Fail Anal, 2023, 148: 107228.

[2.] Hao S, Cui L, Jiang D, et al. A transforming metal nanocomposite with large elastic strain, low modulus, and high strength. Science, 2013, 339(6124): 1191-1194.

[3.] Sun X, Zhou K, Shi S, et al. A new cyclical generative adversarial network based data augmentation method for multiaxial fatigue life prediction. Int J Fatigue, 2022, 162: 106996.

[4.] Zhang M, Sun CN, Zhang X, et al. High cycle fatigue life prediction of laser additive manufactured stainless steel: a machine learning approach. Int J Fatigue, 2019, 128: 105194.

[5.] Yang WK, Hu BL, Luo YW, et al. Understanding geometrical size effect on fatigue life of A588 steel using a machine learning approach. Int J Fatigue, 2023, 172: 107671.

[6.] Wei X, Zhang C, Han S, et al. High cycle fatigue S-N curve prediction of steels based on transfer learning guided long short term memory network. Int J Fatigue, 2022, 163: 107050.

[7.] Zhan Z, Li H. Machine learning based fatigue life prediction with effects of additive manufacturing process parameters for printed SS 316L. Int J Fatigue, 2021, 142: 105941.

[8.] He L, Wang Z, Ogawa Y, et al. Machine-learning-based investigation into the effect of defect/inclusion on fatigue behavior in steels. Int J Fatigue, 2022, 155: 106597.

[9.] Li H, Zhang J, Hu L, et al. Notch fatigue life prediction of micro-shot peened 25CRMO4 alloy steel: a comparison between fracture mechanics and machine learning methods. Eng Fract Mech, 2023, 277.

[10.] Zhou T, Sun X, Chen X. A multiaxial low-cycle fatigue prediction method under irregular loading by ANN model with knowledge-based features. Int J Fatigue, 2023, 176: 107868.

[11.] Srinivasan V. Low cycle fatigue and creep–fatigue interaction behavior of 316L(N) stainless steel and life prediction by artificial neural network approach. Int J Fatigue, 2003, 25(12): 1327-1338.

[12.] Michalski RS, Carbonell JG, Mitchell TM (1983) Machine learning an artificial intelligence approach. Springer Berlin, Heidelberg

[13.] Jones BA, Li W, Nachtsheim CJ et al (2007) Model discrimination—another perspective on model-robust designs. J Stat Plan Infer 137:1577–1583

[14.] Roy A (n.d.) A novel conditional wasserstein deep convolutional generative adversarial network_supp1-3288851.PDF. https://doi.org/10.1109/tai.2023.3288851/mm1

[15.] Goodfellow IJ, Pouget-Abadie J, Mirza M et al. (2014) Generative adversarial networks. arXiv:1406.2661. https://arxiv.org/abs/1406.2661

[16.] Zhu JY, Park T, Isola P et al (2017) Unpaired image-to-image translation using cycle-consistent adversarial networks. In: IEEE international conference on computer vision (ICCV), Venice, Italy, pp 2242–2251

[17.] Liu MY, Tuzel O (2016) Coupled generative adversarial networks. arXiv:1606.07536. https://arxiv.org/abs/1606.07536

[18.] Zhang H, Goodfellow I, Metaxas D et al. (2019) Self-attention generative adversarial networks. arXiv:1805.08318. https://arxiv.org/abs/1805.08318

[19.] Brock A, Donahue J, Simonyan K (2019) Large scale gan training for high fidelity natural image synthesis. arXiv:1809.11096. https://arxiv.org/abs/1809.11096

[20.] Ma B, Wei X, Liu C, et al. Data augmentation in microscopic images for material data mining. NPJ Comput Mater, 2020, 6(1): 125.

[21.] Buehler EL, Buehler MJ. End-to-end prediction of multimaterial stress fields and fracture patterns using cycle-consistent adversarial and transformer neural networks. Biomed Eng Adv, 2022, 4: 100038.

[22.] Liu F, Zhou S, Xia C, et al. Optimization of fatigue life distribution model and establishment of probabilistic S-N curves for a 165 KSI grade super high strength drill pipe steel. J Petrol Sci Eng, 2016, 145: 527-532.

[23.] Li S, Xie X, Cheng C, et al. A modified coffin-manson model for ultra-low cycle fatigue fracture of structural steels considering the effect of stress triaxiality. Eng Fract Mech, 2020, 237: 107223.

[24.] Farhat H. Operation, maintenance, and repair of land-based gas turbines, 2021, Amsterdam: Elsevier

[25.] Cooper CV, Fine ME. Fatigue microcrack initiation in polycrystalline alpha-iron with polished and oxidized surfaces. Metall Trans A, 1985, 16(4): 641-649.

[26.] Baldi P, Sadowski P. The dropout learning algorithm. ArtifIntell, 2014, 210: 78-122.

[27.] Goh ATC. Back-propagation neural networks for modeling complex systems. Artif Intell Eng, 1995, 9(3): 143-151.

[28.] Yin L, Zhang B. Time series generative adversarial network controller for long-term smart generation control of microgrids. Appl Energ, 2021, 281.

[29.] Razmjoo A, Xanthopoulos P, Zheng QP. Online feature importance ranking based on sensitivity analysis. Expert Syst Appl, 2017, 85: 397-406.

[30.] Colin Cameron A, Windmeijer FAG. An R-squared measure of goodness of fit for some common nonlinear regression models. J Econom, 1997, 77(2): 329-342.

[31.] Miao C, Li R, Yu J. Effects of characteristic parameters of corrosion pits on the fatigue life of the steel wires. J Constr Steel Res, 2020, 168: 105879.

[32.] Parzinger M, Hanfstaengl L, Sigg F, et al. Comparison of different training data sets from simulation and experimental measurement with artificial users for occupancy detection—using machine learning methods random forest and lasso. Build Environ, 2022, 223: 109313.

[33.] Itabashi M, Kawata K. Carbon content effect on high-strain-rate tensile properties for carbon steels. Int J Impact Eng, 2000, 24(2): 117-131.

[34.] Rodrigues CAD, Bandeira RM, Duarte BB, et al. Effect of phosphorus content on the mechanical, microstructure and corrosion properties of supermartensitic stainless steel. Mat Sci Eng A, 2016, 650: 75-83.

[35.] Meiners T, Peng Z, Gault B, et al. Sulfur-induced embrittlement in high-purity, polycrystalline copper. Acta Mater, 2018, 156: 64-75.

[36.] Jain S, Jain P, Pandey K et al (2022) Artificial intelligence, machine learning, and mental health in pandemics. https://doi.org/10.1016/c2020-0-04085-5

[37.] He K, Zhang X, Ren S et al (2015) Deep residual learning for image recognition. arXiv:1512.03385. https://arxiv.org/abs/1512.03385

[38.] Yamashita R, NishioM DRK, et al. Convolutional neural networks: an overview and application in radiology. Insights Imaging, 2018, 9(4): 611-629.

[39.] Ana Cláudia OES, de Souza MB, da Silva FV (2022) Exploring the potential of fully convolutional neural networks for FDD of a chemical process. Comput Aided Chem Eng 49:1621–1626

[40.] Kurek A, Kurek M, Łagoda T. Stress-life curve for high and low cycle fatigue. J Theor App Mech, 2019, 57(3): 677-684.

Funding

Key Basic Research Project of Suzhou(#SJC2022029); Key Technologies Research and Development Program http://dx.doi.org/10.13039/501100012165(No. 2022YFB4601804); Innovative Research Group Project of the National Natural Science Foundation of China http://dx.doi.org/10.13039/100014718(No. 52205377); Major Basic Research Project of the Natural Science Foundation of the Jiangsu Higher Education Institutions http://dx.doi.org/10.13039/501100013280(#SJC2022031)