Modeling Corrosion Property of High Chromium Cast Iron under H₃PO₄ Medium Condition Using Artificial Intelligence

Abstract:

The components in slurry pump suffer serious corrosion and abrasion in the phosphorus fertilizer manufacturing process because they undergo corrosion of H3PO4 medium and impact of particles at the same time. Presently, High chromium cast irons are often used to produce the components in slurry pump. In order to reveal the corrosive law, the corrosion properties of high chromium cast iron with 26wt.%Cr content (Cr26) were tested under different H3PO4 medium concentration conditions. Using back-propagation (BP) neural network, the non-linear relationship between the corrosion weight losses (W) and H3PO4 concentration, corrosion time (C, t) is established on the base of the dealing with experimental data. The results show that the well-trained BP neural network can predict the wear weight loss precisely according to H3PO4 concentration and corrosion time. The prediction results reveal that corrosion weight loss rises linearly with increasing corrosion time. The H3PO4 concentration has obvious effect on corrosion property. When H3PO4 concentration is lower than about 0.5mol/L, high chromium cast iron has well resistance to H3PO4 corrosion. However, the corrosion resistance of high chromium cast iron rapidly decreases when the H3PO4 concentration exceed about 0.8 mol/L. It is suggest the high chromium cast iron be used under the condition of H3PO4 concentration of lower 0.8 mol/L.

• Periodical: Advanced Materials Research (Volumes 150-151)
• Pages: 1054-1057
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.150-151.1054
• Online since: October 2010
• Authors: Song Min Zhang, Liu Jie Xu
• Keywords: Bayesian Regularization, BP Neural Network (BPNN), Corrosion, H₃PO₄, High Chromium Cast Iron

References

[1] X.H. Tang, R. Chung, D.Y. Li, et al.: Wear Vol. 267 (2009), p.116.

[2] B.T. Lu, J.L. Luo and S. Chiovelli: Metall. Mat. Trans. A Vol. 37 (2006), p.3029.

[3] S.H. Mousavi Anijdan, A. Bahrami, N. Varahram, et al.: Mater. Sci. Eng. A Vol. 455 (2007), p.623.

[4] X.H. Zhi, J.D. Xing, H.G. Fu, et al.: Materials Characterization Vol. 59(2008), p.1221.

[5] Z. Ma, Y.C. Zhao and R. Xu: Journal of Fuxin Mining Institute (Natural Science Edition) Vol. 16(1997), p.78.

[6] Z.M. Fu, J.H. Mo, C. Lin, ea al.: Materials & Design Vol. 31(2010), p.267.

[7] L.J. Xu, J. Paulo Davim and Rosária Cardoso: J. Mater. Process Technol. Vol. 189(2007), p.374.

[8] Trong Long Ho: Journal of Applied Geophysics Vol. 68(2009), p.489.

[9] Mackay and J.C. David: Neural Computation Vol. 4(1992), p.415.

[10] Ramazan Gençay and Min Qi: IEEE transactions on neural networks Vol. 12(2001), p.726.

[11] J. Guo, and Z. Yang: Analysis and Design of Neural Network Based on Matlab6. 5 (Publishing House of Electronics Industry, China 2003).