Prediction and control of equiaxed α in near-β forging of TA15 Ti-alloy based on BP neural network: For purpose of tri-modal microstructure

For TA15 Ti-alloy in near-β forging and subsequent heat treatment, the evolution of equiaxed α is complex and difficult to control, but a tri-modal microstructure has strict requirements of the volume fraction, size and so on for equiaxed α. In this paper, a prediction model based on improved BP neural network was adopted to investigate quantitative evolution laws of the volume fraction, average grain size, and average aspect ratio of equiaxed α under different deformation temperatures, degrees and strain rates in near-β forging and subsequent high and low temperature double heat treatments (HLT, 950 °C/100 min/WQ+800 °C/8 h/AC). Then, taking the tri-modal microstructure as target, the control of equiaxed α was realized and a reasonable processing parameters match of near-β forging under HLT treatment was determined. Finally the reliability of prediction model and results were verified through experiments, and the tri-modal microstructure with excellent mechanical properties was obtained. The results provide a guide for obtaining a tri-modal microstructure of Ti-alloy through the near-β forging technology.