Abstract: Ultrasonic vibration-assisted microforming has become an important technique for fabricating micro-components owing to its ability to reduce forming load, enhance microcavity filling, and improve surface quality. To address insufficient filling and quality defects in the micro-extrusion of micro-shaft components, the effects of multi-mode ultrasonic vibration on the micro-extrusion behavior of T2 copper under different forming directions were systematically investigated. By quantitatively characterizing forming stress, filling capability, and microstructural evolution, the coupled mechanisms between vibration modes and extrusion directions in microplastic deformation were clarified. On this basis, a coupled theoretical model was developed to describe the attenuation behavior of forming stress. Furthermore, a genetic algorithm-optimized neural network model was established, achieving an average relative prediction error of 1.03%, to predict material filling performance under various ultrasonic parameters and process conditions, thereby supporting the synergistic optimization of shape and properties of micro-components. The results provide theoretical insights and practical guidance for micro- and nano-manufacturing.

Key words: multi-mode ultrasonic vibration, micro plastic forming, extrusion forming direction, forming mechanism, shape-property prediction