This study looked into the thermo-mechanical properties and the recrystallization mechanisms of the Ti–5.5Cr–5Mo–5V-4Al–1Nb alloy (wt.%, Ti−65541). Hot compression experiments were conducted at two phase regions (740~950 °C) and strain rates from 0.001 to 1 s⁻¹. The compressive strain–stress curves were corrected by the adiabatic heating effect. The Arrhenius model was established and provided a reliable prediction of the value of stress with a 0.992 correlation coefficient. The constructed processing map demonstrated that when the temperature rose and the strain rate fell, the power dissipation efficiency (η) showed a rising trend. By utilizing electron backscattered diffraction (EBSD), the microstructural evolution and deformation process were analyzed. It was possible to witness both continuous dynamic recrystallization (cDRX) and discontinuous dynamic recrystallization (dDRX). The dynamic recovery (DRV) and dDRX eventually replaced dDRX while η decreased. Moreover, the deformation band (DB) impeded re-crystallization in the low η area. This study can supply a relatively reliable processing interval for the new Ti−65541 alloy.

The evolution of the microstructure and hardness changes in the Au-15Ag-12Cu-6Ni alloy during the processes of cold rolling and annealing were investigated and the heat treatment regimen for the alloy was optimized in this article. The hardness of the alloy continuously increases with the cold rolling reductions, leading to continuous deformation of the grains during the cold rolling process, ultimately resulting in smaller grain sizes. Subsequent annealing induces recovery and recrystallization, achieving complete recrystallization at 700 °C. An intriguing softening effect is observed after annealing at 700 °C, manifesting in a significant reduction in hardness to 238 (Hv_0.5). The cold deformation texture of the alloy aligns with the recrystallization texture type, exhibiting only a certain degree of angular deviation. This is primarily characterized by //RD texture and a texture deviating 60° from RD towards TD. The performance of the finished sheet improves with the precipitation of ordered phases AuCu after a 300 °C heat treatment for 0.5 h, resulting in a remarkable hardness of 380 (Hv_0.5).

In this paper, a CatBoost model for predicting superelastic strains of alloys was established by utilizing features construction and selection as well as model filtering and evaluation based on 125 existing data points of superelastic titanium alloys. The alloy compositions of a TiNbMoZrSnTa system were optimized and three nickel-free titanium alloys with potentially excellent superelastic properties were designed using the Bayesian optimization algorithm using a superelastic strain as the optimization target. The experimental results indicated that only Ti-12Nb-18Zr-2Sn and Ti-12Nb-16Zr-3Sn exhibited clear superelasticity due to the absence of relevant information about the alloys’ β stability in the machine learning model. Through experimental optimization of the heat treatment regimens, Ti-12Nb-18Zr-2Sn and Ti-12Nb-16Zr-3Sn ultimately achieved recovery strains of 4.65% after being heat treated at 853 K for 10 min and 3.01% after being heat treated at 1073 K for 30 min, respectively. The CatBoost model in this paper possessed a certain ability to design nickel-free superelastic titanium alloys but it was still necessary to combine it with existing knowledge of material theory for effective utilization.