PSI - Issue 68

Morteza Khomami Abadi et al. / Procedia Structural Integrity 68 (2025) 1312–1318 Author name / Structural Integrity Procedia 00 (2025) 000–000

1318

7

The results prove that this method is efficient and accurate by changing the crack location, crack length, and different loading conditions for determining stress intensity factors. The results show that increasing the length of the crack leads to an increase in the stress intensity factor, and the proximity of the crack location to the supports leads to a decrease in these coefficients. 5. Conclusion In this research, a new method based on statistical techniques such as experiment design and neural network was introduced to determine stress intensity coefficients. This method solves many limitations of previous methods. The stress intensity factors in beams and sheets were determined by implementing GMDH algorithm with appropriate errors of about 8%.Some of the most important advantages and limitation of this research are listed below: Advantages of method: • More flexibility over the change of boundary conditions and the crack depth and location. • Independence of the results from the range and values of input parameters. • The GMDH method will be more efficient than the finite element method from the viewpoint of the computational cost if the results for new samples are required. • Decreasing cost in the determination of results (no need for experimental tests, software results, and system training) Limitations of method: • The initial preparation of the GMDH neural network takes more time than the finite element method. • The error in this method is slightly more than finite elements, but it is acceptable. • Complexity in implementing relationships, setting the number of layers, normalizing data, etc. References Alijani, A., Abadi, M.Kh., Razzaghi, J., Jamali, A., 2019. Numerical Analysis of Natural Frequency and Stress Intensity Factor in Euler– Bernoulli Cracked Beam, Acta Mechanica 230, 4391–4415. Besarati, S.M., Myers, P.D., Covey, D.C., Jamali, A., 2015. Modelling Friction Factor in Pipeline Flow Using a GMDH-Type Neural Choi, M.J., Cho, S., 2015. Isogeometric Analysis of Stress Intensity Factors for Curved Crack Problems, Theoretical and Applied Fracture Mechanics 75, 89-103. He, M.Y., & Su, Y.X., 2000. Determination of Stress Intensity Factors Using Finite Element Method. Int. J. Fract 140(3), 235-249. Ivakhnenko, A.G., 1970. Heuristic Self-Organization in Problems of Engineering Cybernetics. J. Automat. 6, 207–219. Ivakhnenko, A.G., 1971. Polynomial Theory of Complex Systems. J. Trans. Syst. 1, 364–378. Kienzler, R., Herrmann, G., 1986. An Elementary Theory of Defective Beams. J. Acta Mech. 62, 37–46. Kim, K.B., Yoon, D.J., Jeong, J.C., Lee, S.S., 2004. Determining the Stress Intensity Factor of a Material with an Artificial Neural Network from Acoustic Emission Measurements. NDT & E International 37(6), 423-429. Kumar, P., 2009. Elements of Fracture Mechanics. McGraw-Hill, McGraw Hill Education (India) Private Limited, Bengaluru. Merrell, J., Emery, J., Kirby, R.M., Hochhalter, J., 2024, Stress Intensity Factor Models Using Mechanics-Guided Decomposition and Symbolic Regression, Engineering Fracture Mechanics 310, 110432. Network. J. Cogent Eng. 2, 1–14. Ricci, P., Viola, E., 2006. Stress Intensity Factors for Cracked T-Sections and Dynamic Behavior of T-Beams. J. Eng. Fract. Mech.73, 91– 111. Seenuan, P., Noraphaiphipaksa, N., Kanchanomai, Ch., 2023. Stress Intensity Factors for Pressurized Pipes with an Internal Crack: The Prediction Model Based on an Artificial Neural Network, Applied Sciences 13(20) 10.3390/app132011446. Tada, H., Paris, P.C., Irwin, G.R., 2000. The Stress Analysis of Crack Handbook, Third Edition, Paris Productions & (Del Research Corp.) Wu, Zh., Hu, Sh., Zhou F., 2014. Prediction of Stress Intensity Factors in Pavement Cracking With Neural Networks Based On Semi Analytical FEM, Expert Systems with Applications 41(4), 1021-1030. Yokoyama, T., Chen, M.C., 1998. Vibration Analysis of Edge-Cracked Beams Using a Line-Spring Model, J. Eng. Fract. Mech. 59, 403– 409.