Global modeling for elevated temperature flow behavior of 6013 aluminum alloy during two-pass deformation

Two-pass hot plane strain compression tests of 6013 aluminum alloy were conducted at different temperatures, strain rates, and holding times. Using the experimental data, four popular metamodel types - Kriging, radial basis function, polynomial regression and artificial neural network - were investigated as potential methods for global modeling of the flow behavior during two-pass deformation. The global model developed from the Kriging method was superior in terms of the accuracy and stability of the prediction. Furthermore, this model was successfully used to predict not only the two-pass deformation behavior beyond the experimental conditions, but also the multipass deformation behavior. It is proved that the Kriging method is an effective and reliable approach to develop a global model for optimization of the hot forming process.