Baysesian Neural Network Approach to Detecting Temper Embrittlement of 30Cr2MoV Rotor Steels

Abstract:

Baysesian Neural Network approach for predicting the temper embrittlement of steam turbine rotor in service was proposed. The FATT50 (the fracture appearance transition temperature) of the rotors was predicted as a function of ratio of the two peak current densities (Ipr / Ip ) tested by electrochemical potentiodynamic reaction method, temperature of electrolyte, J-factor and grain size ( N ). A database was obtained from the test of electrochemical potentiodynamic reaction and Charpy impact. The Bayesian neural network technique was used for modeling of temper embrittlement. The neural network shows a more precise prediction of temper embrittlement of rotor steels than the prediction using multiple linear regression. The training error and verifying error is with the scatter of ±20°C. The results show that, for the temper embrittlement of rotor steels prediction, the prediction model based on Bayesian neural network is feasible and effective.

• Periodical: Applied Mechanics and Materials (Volumes 34-35)
• Pages: 1046-1051
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.34-35.1046
• Online since: October 2010
• Authors: Yong Zhe Fan, Rui Na Ma, An Du
• Keywords: Bayesian Neural Network, Electrochemical Potential, Steam Turbine Rotor, Temper Embrittlement

References

[1] J. Kameda, R. Ranjan. Nondestructive evaluation of steels using acoustic and magnetic Barkhausen signals－II. Effect of intergranular impurity segregation. Acta Metall, Vol. 35（1987), p.1527.

DOI: 10.1016/0001-6160(87)90097-6

[2] Fujita, Akitsugu, Shinohara, Masatomo, Yokota Hiroshi, Kaku Katsuo; Soeda Kihei; Kuroda Yasufumi. Effect of chemical composition and manufacturing procedure on toughness of CrMoV HP rotor forgings. Journal of the Iron and Steel Institute of Japan, Vol. 84（1998), p.236.

DOI: 10.2355/tetsutohagane1955.84.3_236

[3] S. M. Bruemmer. Evaluation of chemical and electrochemical etching techniques to determine phosphorus segregation in NiCrMoV rotor steels. Corrosion, Vol. 42（1986), p.180.

DOI: 10.5006/1.3584899

[4] R. Viswanathan, S. M. Bruemmer, R. H. Richmann. Etching technique for assessing toughness degradation of in-service components. Journal of engineering materials and technology, Vol. 110 （1988), p.313.

DOI: 10.1115/1.3226055

[5] Kwon II-Hyun, Baek Seung-Se, Yu Hyo-Sun. The use of electrochemical and mechanical property correlations to monitor the degradation of Cr-Mo-V turbine casing steel. International Journal of Pressure Vessels and Piping, Vol. 80（2003), p.157.

DOI: 10.1016/s0308-0161(02)00140-0

[6] X. Mao and W. Zhao. Electrochemical Polarization Method to Detect Aging Embrittlement of 321 Stainless Steel. Corrosion, Vol. 49（1993), p.335.

DOI: 10.5006/1.3316058

[7] Shin-ichi Komazaki, Tetsuo Shoji, Iwao Abe, Ikuo Okada. Nondestructive evaluation of creep damage and life prediction of Ni-base super alloy used in advanced gas turbine blades by electrochemical technique. Ninth International Symposium on Nondestructive Characterization of Materials Shield. Sydney, 1999. 113-119.

DOI: 10.1063/1.1301992

[8] Yoshikuni Kadoya, Toru Goto, Mao Takei. Nondestructive Evaluation of Temper Embrittlement in Cr-Mo-V Rotor Steels. JSME International Journal, Vol. 35（1992), p.226.

DOI: 10.1299/jsmea1988.35.2_226

[9] Danaher S. , Datta S.; Waddle I. , Hackney P. Erosion modeling using Bayesian regulated artificial neural networks. Wear, Vol. 256（2004, p.879.

DOI: 10.1016/j.wear.2003.08.006

[10] Yoo Y.S. , Jo C.Y. , Jones C.N. Compositional prediction of creep rupture life of single crystal Ni base super alloy by Bayesian neural network. Materials Science and Engineering A, Vol. 336 （2002), p.22.

DOI: 10.1016/s0921-5093(01)01965-7

[11] Arpad Kelemen, Yulan Liang. Bayesian Regularized Neural Network for Multiple Gene Expression Pattern Classification, Proceedings of the International Joint Conference on Neural Networks, Portland, 2003. 654-659.

DOI: 10.1109/ijcnn.2003.1223441

[12] Darrin C. Edwards, Charles E. Metz, Robert M. Nishikawa, Estimation of three-class ideal observer decision functions with a Bayesian artificial neural network, Proceedings of SPIE - The International Society for Optical Engineering, San Diego, 2002. 1-12.

DOI: 10.1117/12.462662

[13] Chua C.G. , Goh, A.T.C. , A hybrid Bayesian back-propagation neural network approach to multivariate modeling. International Journal for Numerical and Analytical Methods in Geomechanics, Vol. 27（2003), p.651.

DOI: 10.1002/nag.291