Modeling of High Temperature Flow Stress of AZ80 Magnesium Alloy with Support Vector Machines and Artificial Neural Network

Abstract:

Support vector machines (SVM) and artificial neural network (ANN) were employed in modeling the flow stress of the AZ80 magnesium. The hot deformation behavior of extruded AZ80 magnesium was investigated by compression tests in the temperature 350-450 and strain rate range 0.01-50 s^-1. The maximum relative errors at different temperatures and different strain rates between experimental and predicted flow stresses by SVM and ANN were compared. The results show the SVM derives statistical models have better similar prediction ability to those of ANN, especially at high strain rate. This indicates that SVM can be used as an alternative modeling tool for high temperature rheological behavior studies.

• Periodical: Advanced Materials Research (Volume 486)
• Pages: 227-232
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.486.227
• Online since: March 2012
• Authors: Yan Lou
• Keywords: Artificial Neural Network (ANN), AZ80 Magnesium, Flow Stress, Modeling, Support Vector Machine (SVM)

References

[1] Zhou H.T., Li Q.B., Zhao Z.K. Hot workability characteristics of magnesium alloy AZ80-A study using processing map, Materials Science and Engineering A, 2010; 527: 2022-6.

DOI: 10.1016/j.msea.2009.12.009

[2] Saniee F.F., Badnava H., Najafabadi S.M.P. Application of T-shape friction test for AZ31 and AZ80 magnesium alloys at elevated temperatures, Materials and Design, 2011; 32: 3221-30.

DOI: 10.1016/j.matdes.2011.02.042

[3] Shahzad M., Wagner L. Influence of extrusion parameters on microstructure and texture developments, and their effects on mechanical properties of the magnesium alloy AZ80, Materials Science and Engineering A, 2009; 506: 141-7.

DOI: 10.1016/j.msea.2008.11.038

[4] Anaraki M.T., Sanjari M., Akbarzadeh A. Modeling of high temperature rheological behavior of AZ61 Mg-alloy using inverse method and ANN, Material and Design, 2008; 29: 1701-6.

DOI: 10.1016/j.matdes.2008.03.027

[5] Heesung Yoon, Seong-Chun Jun, Yunjung Hyun, et al. A comparative study of artificial neural networks and support vector machines for predictiong groundwater levels in a coastal aquifer, Journal of Hydrology, 2011; 396: 128-38.

DOI: 10.1016/j.jhydrol.2010.11.002

[6] Zhang L.Q., Li L.X., Ju H., Zhu B.W. Inverse identification of interfacial heat transfer coefficient between the casting and metal mold using neural network, Energy Conversion and Management, 2010; 51: 1898-904.

DOI: 10.1016/j.enconman.2010.02.020

[7] Sha W., Edwards K.L. The use of artificial neural networks in materials science based research, Material and Design, 2007; 28: 1747-52.

[8] Cortes C., Vapnik V. Support vector networks, Mach Learn, 1995; 20: 273–97.

DOI: 10.1007/bf00994018

[9] Yang Z., Gu X.S., Liang X.Y., Ling L.C. Genetic algorithm-least squares support vector regression based predicting and optimizing model on carbon fiber composite integrated conductivity. Material and Design, 2010; 31: 1042-9.

DOI: 10.1016/j.matdes.2009.09.057

[10] Huang Z., Chen H., Hsu C.J. Credit rating analysis with support vector machines and neural networks: a market comparative study, Decision Support Systems, 2004; 37: 534-58.

DOI: 10.1016/s0167-9236(03)00086-1

[11] Peng X.J. TSVR: An efficient Twin Support Vector Machine for regression, Neural Networks, 2010; 23: 365-72.

DOI: 10.1016/j.neunet.2009.07.002

[12] Li L., Zhou J., Duszczyk J. Determination of a constitutive relationship for AZ31B magnesium alloy and validation through comparison between simulated and real extrusion, Journal of Materials Process Technology. 2006; 172: 3372-80.

DOI: 10.1016/j.jmatprotec.2005.09.021

[13] Lou Y., Li L.X., Luan N. Flow stress correction of AZ80 magnesium alloy for deformation heating at high strain rates during hot compression, Advanced Materials Research, 2011; 263: 326-1330.

DOI: 10.4028/www.scientific.net/amr.129-131.1326

[14] Yang Z., Gu X.S., Liang X.Y., Ling L.C. Genetic algorithm-least squares support vector regression based predicting and optimizing model on carbon fiber composite integrated conductivity, Materials and Design, 2010; 31: 1042-9.

DOI: 10.1016/j.matdes.2009.09.057