Microstructural evolution and constitutive relationship of Al–Zn–Mg alloy containing small amount of Sc and Zr during hot deformation based on Arrhenius-type and artificial neural network models

Abstract

Abstract Hot compression tests of Al–Zn–Mg alloy containing small amount of Sc and Zr were carried out on a Gleeble-1500 thermo-mechanical simulator in deformation temperature range from 340 to 500 °C and at strain rate range from 0.001 to 10 s−1. The microstructural evolution and flow behavior of the alloy under various conditions were investigated. It was observed that the flow stress increased with increasing strain rate and decreasing deformation temperature, the main dynamic soften mechanism was caused by dynamic recovery and dynamic recrystallization. Based on the experimental results, Arrhenius-type constitutive equations and artificial neural network (ANN) model with a back-propagation learning algorithm were established to predict hot deformation behavior of the alloy. A comparative predictability of constitutive equations and ANN model were further evaluated using a wide variety of statistical indices. The average absolute relative error was 0.58% and 3.57% corresponding to ANN model and constitutive equations, respectively. The relative errors between experimental and predicted flow stress computed from constitutive equations and ANN model were respectively in the range of 15.58% to −15.32% and 3.16% to −3.25%. The results showed that the predicted flow stresses by ANN model were in a better agreement with experimental values, indicating that ANN model is more accurate and efficient than Arrhenius-type constitutive equations.