Methods of jump-like modeling of the discontinuous yield of AMg6 aluminum alloy

Abstract

Various approaches for studying the jump-like deformation of AMg6 aluminum alloy are being compared. AMg6 alloy is characterized by instantaneous deformation increases during uniaxial stretching in the area of plasticity. It was assumed that the process of jump-like tensile deformation is caused by the cracking of dispersoids in the volume of the material. Based on that assumption, the methods that predict the initiation and magnitude of jump-like deformation depending on the proportion of destroyed inclusions were proposed. In particular, the ANSYS software complex was used to predict jump-like deformation, in which the groups of finite element models were developed to determine the main patterns of influence of structural heterogeneity parameters of the simulated environment on the stress-strain state. In addition, given the large amount of experimental data, it is important to learn how to solve such problems using machine learning (ML), particularly neural networks. It has been established that the prediction accuracy by one of the most common ML methods, that was neural networks, comprised more than 90%.