Hot Deformation Characterization of Mg-Sm-Zn-Zr Alloy Using Artificial Neural Network and 3D Processing Map
The true stress-strain data from isothermal hot compression tests on Gleeble-1500D thermo mechanical simulator, in a wide range of temperatures (350-450°C) and strain rates (0.001-1s-1), was employed to establish the PSO-BP network prediction model and 3D processing map of Mg-Sm-Zn-Zr alloy. It was found that the PSO-BP model could be efficient and accurate in predicting flow stress, most of relative errors were in the range of -4% to 6%, and the average relative error was found to be 1.52%. Then considering the effect of strain, the 3D processing map was established to characterize the hot workability of the alloy. The 3D processing map exhibited the maximum efficiency domain and the instability domain, which could be used to determine the optimal deformation conditions. The optimum processing parameters of Mg-Sm-Zn-Zr alloy were deformation temperatures of 400-450°C and strain rates of 0.003-0.1s-1.