Electronic versions

Documents

  • Materials_today_communications_ZK60_Magnesium

    Accepted author manuscript, 2.51 MB, PDF document

    Embargo ends: 6/03/27

    Licence: CC BY-NC-ND Show licence

DOI

  • Hongwei Yan
    Chongqing University of Science and Technology
  • Qingshan Yang
    Chongqing University of Science and Technology
  • Dan Zhang
    Chongqing University of Science and Technology
  • Guanglin Liu
    Chongqing University of Science and Technology
  • Xianzheng Li
    Chongqing University of Science and Technology
  • Fei Ren
    Chongqing University of Science and Technology
  • Liyang Yue
  • Bin Jiang
    Chongqing University
This work presents a method of artificial neural network (ANN) to predict the stretch formability of ZK60 alloy associated with crystal deflection angle, tensile mechanical properties. To discuss the regulation of texture, the balance of strength and plasticity, an ANN model was constructed with datasets of texture and mechanical properties parameters that were prepared from published literature. The results of model training and prediction based on 7 predictors show that the model has good generalization ability and the prediction accuracy is over than 0.93. The quantitative evolution of microstructure was analyzed and calculated. Correlation analysis of the predicted results showed that the tensile mechanical properties and the texture distribution play a nonlinear role in regulating formability. Experimental results validate the reliability of the ANN model. Furthermore, high formability conditions were analyzed and discussed.

Keywords

  • Magnesium alloy, artificial neural network, Texture evolution, Stretch formability
Original languageEnglish
Pages (from-to)112150
Number of pages16
JournalMaterials Today Communications
Volume44
Early online date6 Mar 2025
DOIs
Publication statusE-pub ahead of print - 6 Mar 2025
View graph of relations