PSI - Issue 68

Giovanni Chianese et al. / Procedia Structural Integrity 68 (2025) 1245–1251 Chianese et al. / Structural Integrity Procedia 00 (2025) 000–000

1251

7

In Fig. 5 (a), the true and the predicted values of the normalized crack length α are plotted. The four subplots show a good accuracy during validation for all of the four considered values of aspect ratio β , with a RMSE equal to 0.02. Fig. 5 (b) reports the performances of the CNN in predicting the crack front geometry by classifying β during the validation of the CNN. An average accuracy of about 91.7% was achieved. The confusion chart shows main misclassifications occur between classes with β equal to 1 and 1.5, which both represent scenarios with the more severe crack tunnelling. However, scenarios with straight crack front and less severe crack tunnelling were correctly predicted, indicating that the proposed methodology has potential to distinguish scenarios not requiring corrective factors from scenarios that would need this post-processing correction. 4. Conclusions and next steps for further developments In this work, a novel technique for in-situ real-time monitoring of the crack growth and geometry based on strain measurements was numerically demonstrated. In particular, two approaches have been considered: one involved the use of local strain measurements from the back and lateral face of the specimen, and the other one was based on processing full field strain measurements, e.g. carried out with DIC, and processed with a CNN. Simultaneous monitoring of the crack growth and geometry by using two local strain measurements resulted to be not feasible because the information was not representative of differences in scenarios with different crack front geometries. On the other hand, full field strain measurement on the lateral surface of the specimens proved to carry information useful for classification of the crack front geometry. Two separate CNNs were trained and validated for tracking the crack growth and predict the crack front geometry, with RMSE = 0.02 and classification accuracy above 90%, respectively, demonstrating that this approach has potential to be scaled in laboratory practice. Next steps for further development involve validation of the proposed methodology with experimental data collected during fatigue crack growth. References Chandawanich, N., Sharpe, W.N., 1979. An experimental study of fatigue crack initiation and growth from coldworked holes. Engineering Fracture Mechanics. 11(4), 609–620. Stephens, R.I., Fatemi, A., Stephens, R.R., Fuchs, H.O., 2000. Metal fatigue in engineering. II. New York: John Wiley & Sons, Ltd. Leonetti, D., Maljaars, J., & Snijder, H. B., 2021. Fracture mechanics based fatigue life prediction for a weld toe crack under constant and variable amplitude random block loading—Modeling and uncertainty estimation. Engineering Fracture Mechanics , 242 , 107487. De Iorio, A., Grasso, M., Penta, F., Pucillo, G.P., 2012. A three-parameter model for fatigue crack growth data analysis. Frattura ed Integrità Strutturale 21:21-29. Standard Test Method for Measurement of Fatigue Crack Growth Rates. ASTM International E647-24, June 24 2024. Standard Test Method for Linear-Elastic Plane-Strain Fracture Toughness of Metallic Materials. ASTM International E399-24, Oct 18 2024. Newman, J.C., Yamada, Y., James, M.A., 2011. Back-face strain compliance relation for compact specimens for wide range in crack lengths. Engineering Fracture Mechanics 78 (15) 2707-2711. Riddell, W., Piascik, R., 1999. A Back-Faced Strain Compliance Expression for the Compact Tension Specimen. ASTM International Journal of Testing and Evaluation,, 27 (2) 167-170. Carpinteri, A., 1993. Shape change of surface cracks in round bars under cyclic axial loading. International Journal of Fatigue 15(1),21–26. Jeong, D.Y., Tang, Y.H., Orringer, O., 1997. Damage tolerance analysis of detail fractures in rail. Theoretical and Applied Fracture Mechanics 28(2),109–115. Carpinteri, A., Ronchei, C., Vantadori, S., 2013. Stress intensity factors and fatigue growth of surface cracks in notched shells and round bars: two decades of research work. Fatigue & Fracture of Engineering Materials & Structures 36(11), 1164–1177. Pucillo, G.P., Esposito, L., Leonetti, D., 2019. On the effects of unilateral boundary conditions on the crack growth rate under cycling bending loads. International Journal of Fatigue 24, 245–252. Saeed, H., Chaudhuri, S., De Waele, W., 2023. Front face strain compliance for quantification of short crack growth in fatigue testing. The Journal of Strain Analysis for Engineering Design 58 (8), 649-660. Melching, D., Strohmann, T., Requena, G., Breitbarth, E., 2022. Explainable machine learning for precise fatigue crack tip detection. Scientific Reports 12, 9513. Pal, A., Meng, W., Nagarajaiah, S., 2023. Deep Learning-Based Subsurface Damage Localization Using Full-Field Surface Strains. Sensors 23 (17), 7445. Li, M., Jia, D., Wu, Z., Qiu, S., He, W., 2022. Structural damage identification using strain mode differences by the iFEM based on the convolutional neural network (CNN). Mechanical Systems and Signal Processing 165, 108289.