PREDICTION OF HARDNESS DISTRIBUTION IN PLASMA ARC SURFACE HARDENING USING NEURAL NETWORK
Keywords:
Plasma arc, surface hardening, neural network, hardnessAbstract
In this paper, an attempt has been made to develop neural network models to predict the hardnessdistribution of hardened zone in plasma arc surface hardening process. The back propagation methodwith the Levenberg-Marquardt algorithm was used to train the neural network models. Hardnessdistributions were collected by the experimental setup in the laboratory and the associated data wereused to train the neural network models. Furthermore, the prediction of neural network models werecompared with those obtained from a statistical regression models. It is confirmed experimentallythat the hardness distribution can be accurately predicted by the trained neural network models. Theaccuracy of hardness distribution prediction using neural network is superior to that using otherstatistical regression models.
References
Ashby, M.F. and Easterling, K.E. (1984). Thetransformations hardening of steel surfaceby laser beams – I. hypoeutectoid steels.Acta Metallurgica, 32(11):1,935-1,948.
Bourithis, E., Tazedakis, A., and Papadimitriou,G. (2002). A study on the surface treatmentof Calmax tool steel by a plasma transferredarc (PTA) process. J. Mater. Process.Tech., 128:169-177.
Cheng, P.J. and Lin, S.C. (2000). Using neuralnetworks to predict bending angle of sheetmetal formed by laser. Int. J. Mach. ToolManufact., 40:1,185-1,197.
Demuth, H. and Beale, M. (2004). Neural NetwotkToolbox User’s Guide. Version 4.0. TheMathworks Inc.
Ezugwu, E.O., Fadare, D.A., Bonney, J., Da Silva,R.B., and Sales, W.F. (2005). Modelling thecorrelation between cutting and processparameters in high-speed machining ofInconel 718 alloy using an artificial neuralnetwork. Int. J. Mach. Tool Manufact.,45:1,375-1,385.
Ion, J.C. and Anisdahl, L.M. (1997). A PC basedsystem for produce development in lasertransformation hardening. J. Mater. Process.Tech., 65:261-267.
JIS. (1996). JIS G 0559 - Methods of measuringcase depth for steel hardened by flame or induction hardening process. StandardizationJournal, Japanese Standards Association.
Kim, I.S., Son, J.K., and Yarlagadda, P.K.D.V.(2003). A study on the quality improvementof robotic GMA welding process. Roboticsand Computer Integrated Manufacturing,19:67-572.
Kraposhin, V.S., Bobrov, A.V., and Gaponenko,O.S. (1989). Surface hardening of 9KhF steelby heating with a plasma gun. MetallovedTerm Obrab Met, 11:13-17.
Pan, W.X., Meng, X., Li, G., Fei, Q.X., and Wu,C.K. (2005). Feasibility of laminar plasmajet hardening of cast iron surface. Surfaceand Coatings Technology, 197:345-350.
Stich, T.J., Spoerre, J.K., and Velasco, T. (2000).The application of artificial neuralnetworks to monitoring and control of aninduction hardening process. J. Ind. Tech.,16(1):1-11.
Wang, J.Y., Kusumoto, K., and Nezu, K. (1999).Modelling and prediction of cut shape forplasma arc cutting based on artificialneural network. Science and Technologyof Welding and Joining, 4(4):195-200.
Woo, H.G. and Cho, H.S. (1998). Estimation ofhardened layer dimensions in laser surfacehardening processes with variations ofcoating thickness. Surface and CoatingsTechnology, 102:205-217.
Yan, M. (2003). Micro-beam plasma arc surfaceprocessing for ferrous and non-ferrousmetals. J. Mat. Sci., 38:3,219-3,222.
Yan, M. and Zhu, W.Z. (1997). Surface treatmentof 45 steel by plasma arc melting. Surfaceand Coatings Technology, 91:183-191.
Yan, M. and Zhu, W.Z. (1998). A new approachof surface treatment: micro-beam plasmaarc remelting. Materials Letters, 34:222-225.
Zhang, Y.F., Lu, L., and Yap, S.M. (1999). Predictionof the amount of PCA for mechanicalmilling. J. Mater. Process. Tech., 89-90:260-265.
