During hot working, material exhibits a complex nonlinear relationship between flow stress and thermo-mechanical processing parameters, due to the simultaneous occurrence of work hardening and softening. In order to express more accurately the nonlinear relationship between flow stress and thermo-mechanical processing parameters, a large number of constitutive models have been developed, which can be divided into three categories: a phenomenological constitutive model, a physically-based model and an artificial neural network model.
Phenomenological constitutive models are widely used to predict the hot deformation behavior of metallic materials including alloy steels and titanium alloys, due to their simple equations and small computational quantities. Physically-based constitutive models consider not only the thermo-mechanical processing parameters, but also the physical mechanism such as dislocation movement and thermal activation during hot working process. Compared with phenomenological constitutive models, physically-based models are more complex to establish. However, due to their high prediction accuracy, physically-based models are widely used in many commercial finite element analysis programs.
The Ti-6Al-2Sn-2Zr-3Mo-1.5Cr-2Nb (TC21) alloy is a 1 100MPa damage tolerance titanium alloy, with ultimate tensile strength of 1 130MPa, yield strength of 1 020MPa, elongation of 12%, toughness value of 80MPa·m1/2, and crack extension rate of 1.47×10-5 mm·cycle-1. Due to a good combination of strength and toughness, the application value of TC21 alloy is much higher than that of TC4 alloy.
So far, there have been many studies on the hot deformation behavior of TC21 alloy. However, few studies have been conducted or reported on other constitutive models for TC21 alloy including physically-based constitutive models.
Pak Hun, a student at the Faculty of Materials Science and Technology, has constructed phenomenological and physically-based models for predicting more accurately the hot deformation behavior of TC21 alloy, and compared their predictability. To this end, he carried out isothermal uniaxial compression tests at different strain rates in α + β two-phase and β phase fields. Based on the experimental data (flow stress curves), he developed a modified Johnson-Cook (m-JC) model, a strain-compensated Arrhenius type (sc-AT) model and a modified Zerilli-Armstrong (m-ZA) model for Ti-6Al-2Zr-2Sn-3Mo-1.5Cr-2Nb alloy. In addition, he compared the prediction accuracies of the three constitutive models using standard statistical parameters.
The results showed that the proposed m-ZA model is the preferred constitutive model to predict the hot deformation behavior of Ti-6Al-2Zr-2Sn-3Mo-1.5Cr-2Nb alloy.
You can find the details in his paper “Comparisons of phenomenological and physically-based constitutive models for Ti-6Al-2Zr-2Sn-3Mo-1.5Cr-2Nb alloy” in “Applied Physics” (SCI).
© 2021 KumChaek University of Technology