Constitutive modelling of the flow behaviour of a β titanium alloy at high strain rates and elevated temperatures using the Johnson–Cook and modified Zerilli–Armstrong models

Abstract

The objectives of this work are to characterize the flow behaviour of the Ti–6Cr–5Mo–5V–4Al (Ti6554) alloy at high strain rates and elevated temperatures using the Johnson–Cook (JC) model and a modified Zerilli–Armstrong (ZA) model, and to make a comparative study on the predictability of these two models. The stress–strain data from Split Hopkinson Pressure Bar (SHPB) tests over a wide range of temperatures (293–1173K) and strain rates (103–104s−1) were employed to fit parameters for the JC and the modified ZA models. It is observed that both the JC and the modified ZA models have good capacities of describing the flow behaviour of the Ti6554 alloy at high strain rates and elevated temperatures in terms of the average absolute error. The modified ZA model is able to capture the strain-hardening behaviour of the Ti6554 alloy better as it incorporates the coupling effects of strain and temperature. However, dynamic recovery or dynamic recrystallization that may happen at elevated temperatures should be taken into consideration when selecting data set for parameters fitting for the modified ZA model. Also the modified ZA model requires more stress–strain data for the parameters fitting than the JC model.