Evaluation and prediction of hot rheological properties of Ti-6Al-4V in dual-phase region using processing map and artificial neural network

Abstract

Isothermal compression tests of Ti-6Al-4V are conducted in the dual-phase region at different temperatures (1053-1203 K) and strain rates (0.01-10 s^-1). Processing maps based on the principles of the dynamic material model are then constructed using the experimental data. Stable and unstable regions on the maps are distinguished to evaluate the hot forming performance of the alloy. The regions suitable for hot forming are validated by the microstructure evolutions. The domain with a high power dissipation efficiency <i style="mso-bidi-font-style:normal">η</i> larger than 0.3 is found to be the optimal processing region on the map at a strain of 0.7, where the temperatures range from 1090 K to 1203 K and the strain rates from 0.01 s^-1 to 0.1 s^-1. Moreover, an artificial neural network model with a back-propagation algorithm was developed to predict the hot rheological properties of the alloy involving complex nonlinear intrinsic relationships between the processing parameters. Theoretical processing maps are drawn by using the predicted data and then compared with the experimental maps. The results indicate that the model can track the experimental data with sound precision, and the theoretical maps can definitely give guidance to the design of hot forming process of Ti-6Al-4V in the dual-phase region.<b style="mso-bidi-font-weight:normal"></b>