本团队博士生郝一等在Journal of Alloys and Compounds发表研究论文。

摘要：The dynamic α→β phase transformation (DPT) is a prominent microstructural evolution characteristic of TC18 titanium alloy during deformation in the α+ β region, yet it has not been considered in the current constitutive modeling, resulting in the inability of the constitutive model to fully reflect the softening mechanisms. The flow behavior and microstructure evolution of TC18 titanium alloy were systematically investigated by hot compression tests. Microstructure analyses reveal that the flow softening primarily results from DPT and the dynamic recrystallization (DRX) of β phase. The DPT kinetics is accelerated by enhanced diffusion conditions and the DPT driving force at high temperatures and strain rates. The β phase DRX fraction increases with temperature at lower strain rates, but decreases with temperature at higher strain rates. Additionally, the impediment of α phase on dislocation motion promotes β phase DRX, and in turn promotes DPT. On this basis, a constitutive model fully considering the complex microstructural evolution was developed. The novelty of the proposed model lies in its integration of DPT kinetics modeling and machine learning-assisted DRX kinetics modeling through a phase decomposition approach, thereby filling the gap in previous studies that insufficiently considered the effects of DPT. The DPT kinetics modeling was based on the Johnson-Mehl-Avrami framework incorporating deformation parameters, and the DRX kinetics modeling accounted for the observed microstructure and the influence of DPT. The proposed model achieves average absolute relative errors of 2.3 % and 2.5 % in flow stress prediction for the modeling and the independent validation datasets, respectively.