Modelling for the flow behavior of a new metastable beta titanium alloy by GA-based Arrhenius equation

Abstract

The hot compression tests of a new metastable beta titanium alloy (Ti-5Al-5Mo-5V-3Cr-1Zr alloy) were conducted at the temperature range of 1033–1123 K with an interval of 30 K and the strain rates of 0.001–1 s−1 on a Gleeble-3500 isothermal simulator for the acquisition of true stress strain curves. The true stress strain curves at the experimental conditions show different curve-types which reveal different microstructure evolution mechanisms including dynamic recrystallization (DRX), dynamic globularization (DG) and dynamic recovery (DRV). The Arrhenius constitutive model of this alloy was developed based on the experimental results and used to predict the true stress. Based on the developed Arrhenius model, a genetic-algorithm-based (GA-based) Arrhenius model was constructed and applied in true stress prediction. To validate the superiority of the GA-based Arrhenius model in true stress prediction, statistical indicators including mean value (μ), standard deviation (ω), correlation coefficient (R) and average absolute relative error (AARE) were introduced to assess and compare the experimental values and the prediction values. As a result, the μ-value, ω-value, R-value and AARE-value of Arrhenius equation method are −3.8998%, 5.505 62%, 0.9960 and 5.1817% respectively and of the GA-based Arrhenius equation method are 0.0505%, 0.8083%, 0.9999 and 0.6099% respectively, which means the GA-based Arrhenius equation method is promoted in true stress prediction and high accuracy finite element simulation can be achieved by this method.