Introduction This study proposes a novel prestress-impact synergistic plastic forming equipment for difficult-to-deform high-specific-strength materials, with a focus on TC4 titanium alloy. Methods A dynamic optimization framework integrating finite element analysis (FEA) and Gaussian Process Regression (GPR) surrogate modeling is developed to enhance the structural dynamic performance. Results Prestressed modal analysis identifies the sixth mode as the weak mode, and multi-objective optimization using GPR-assisted NSGA-II achieves a 6.1% reduction in structural mass, a 3.6% increase in the sixth natural frequency, and an 8.7% decrease in maximum von Mises stress compared to the initial design, outperforming traditional response surface methodology. Experimental validation on a constructed prototype demonstrates that the synergistic process, combining controllable prestress with low-frequency impact excitation, significantly improves formability: height reduction increases 48.6% (versus 35.2% in static pressing), forming force and energy consumption decrease by 32.5% and 30%, respectively, while average grain size is refined from 28.4 µm to 12.7 µm with equiaxed α-phase fraction rising to 78% Vickers hardness improves from 312 HV to 358 HV due to enhanced dynamic recrystallization and bimodal microstructure evolution. Discussions The integration of machine learning-based surrogate modeling provides an efficient and scalable approach for vibration-assisted forming systems, offering substantial energy savings and superior mechanical properties aligned with sustainable high-end manufacturing requirements.
Hu et al. (Wed,) studied this question.