Application of Artificial Neural Network to Forecast the Tensile Fatigue Life of Carbon Material

Abstract:

Artificial neural network (ANN) is widely applied to the modeling of complex systems, which has become a common modeling method in the study of materials science. As the ideal candidates for high temperature structural materials, carbon materials are no doubt involved in fatigue loads, so the study on forecasting fatigue life is meaningful. In this paper, the electrical resistance at various fatigue cycles and level of applied stress of the materials under tensile fatigue loading has been detected, and regarded the fracture or fatigue cycles equal to 106 as fatigue life of carbon materials. On the basis of the electrical resistance value, the fatigue life has been forecasted by applied the ANN. The results indicated that the ANN could forecast the fatigue life of carbon materials well; finally, the applications of ANN in the study of material, such as properties prediction, damage prediction and failure detection were reviewed.

• Periodical: Key Engineering Materials (Volumes 385-387)
• Pages: 533-536
• DOI: https://doi.org/10.4028/www.scientific.net/KEM.385-387.533
• Online since: July 2008
• Authors: Xiao Ling Liao, Wen Feng Xu, Zhi Qiang Gao
• Keywords: Artificial Neural Network (ANN), Carbon Material, Forecast, Tensile Fatigue Life

References

[1] Hejun Li, Xinghui Hou, Yixi Chen. Densification of unidirectional carbon/carbon composites by isothermal chemical vapor infiltration. Carbon, 2000, 38(3): 423-427.

DOI: 10.1016/s0008-6223(99)00122-0

[2] A Ozturk. The influence of cyclic fatigue damage on the fracture toughness of carbon/carbon composites. Compoties Part A, 1996, 27A: 641-646.

DOI: 10.1016/1359-835x(96)00035-8

[3] Ken Goto, Hiroshi Hatta, Daisuke Katsu, Terufumi Machida. Tensile fatigue of a laminated carbon-carbon composite at room temperature. Carbon, 2003, (41): 1249-1255.

DOI: 10.1016/s0008-6223(03)00040-x

[4] H Hatta, Y Kogo, T Tanimoto, T Morii. Static and fatigue fracture behavior of C/C composites. In: Proc of 4th Japan Int SAMPE Symposium, 1995: 368-373.

[5] C Tallaron, D Rouby, P Reynaud, G Fantozzi. Improvement of cyclic fatigue analysis by the use of a tensile master curve in carbon/carbon composites. Key Eng Mater, 1999, 164-165: 329-332.

DOI: 10.4028/www.scientific.net/kem.164-165.329

[6] Y Z Pappas, Markopoulos, V Kostopoulos. Failure mechanisms analysis of 2D carbon/carbon using acoustic emission monitoring, NDT & International, 2001, 31(3): 157-163.

DOI: 10.1016/s0963-8695(98)00002-4

[7] S Wang and D D L Cheng. Self-monitoring of strain and damage by a carbon-carbon composite. Carbon, 1997, 35(5): 621-630.

DOI: 10.1016/s0008-6223(97)00011-0

[8] Cai A H, Shun G X. The forecast of property of materials by using ANN. Foundry, 2003, 152 (4): 269-271.

[9] Liu Y L, Zhong Q P, Zhang Z. The forecast of fatigue damage degree of aluminum alloy by using ANN. J of Aviation, 2001, 22(2): 135-139.

[10] Valoor M T, Chandrashekhara K. A thick composite-beam model for delamination prediction by the use of neural networks. Composites Science and Technology, 2000, (60): 1773-1779.

DOI: 10.1016/s0266-3538(00)00063-4

[11] Qian Hancheng, Xia Bocai, Li Shangzheng, et al . Fuzzy neural network modeling of material properties. Journal of Materials Processing Technology, 2002, (122): 196-200 Cycle/n Applied stress/MPa �R/Ro/% Fig. 4 The network trainrecord of forecast Import Adjusting error Connect error of every nerve cell Target value Export Fig. 3 The study sketch of ANN Compare.

DOI: 10.1016/s0924-0136(02)00019-5