Digital Twin-Driven Sheet Metal Forming: Modeling and Application for Stamping Considering Mold Wear

Abstract

Existing various constructed models of stamping provide great support to develop the forming quality improvement and energy-saving strategies. However, the immutable model cannot reflect the actual states of the process as the wear of the mold goes, and the inaccuracy model will lead to the failure of the strategies. To solve this problem, a Digital Twin-driven modeling method considering mold wear for stamping was proposed in this paper. The model of punch force and forming quality considering the coefficients that will vary with the states of mold wear was first built in the virtual space. The real-time punch force was acquired and inputted to the virtual space, and it was then compared with the punch force obtained by the Digital Twin model for monitoring the mold wear. If the difference of punch force is greater than the threshold, the friction coefficients update starts via the Particle Swarm Optimization with Differential Evolution (PSO-DE) algorithm. To validate the effectiveness, the method was applied in the process to form a clutch shell, and the results show that the maximum deviation of the punch force between the updated Digital Twin model and the measured value does not exceed 5%. Optimization results in the application show a 14.35% reduction in the maximum thinning ratio of the stamping part and an 8.9% reduction in the process energy. The Digital Twin-driven modeling assists in quality improvement and energy consumption reduction in sheet metal forming.