A Digital twin framework for Visco-Hyperelasticity calibration: Experiment and simulation

Modeling heavy gauge vacuum-assisted thermoforming is challenging due to material deformation, time-dependent behavior, and mould-sheet friction. This study develops an adaptive methodology that integrates Finite Element Model Updating (FEMU) to calibrate the visco-hyperelastic properties of ABS thermoplastic material within the thermo-vacuum forming range. Experimental data from a 3D Digital Image Correlation (DIC)-equipped biaxial bubble inflation test and step-strain relaxation tests were used to characterize the material behavior at 140 °C. A 2-term Ogden model combined with a Prony series captured material behavior. The objective function minimizes the Root Mean Square (RMS) error between experimental and simulated strain data, focusing on equibiaxial deformation at the bubble’s pole. The calibrated visco-hyperelastic model is further assessed under off-center biaxial deformation modes during bubble formation, leveraging the unique advantage of the bubble inflation test in capturing multiple deformation modes simultaneously. By integrating a real-time digital twin approach, an adapting method is proposed to dynamically optimize friction coefficient using strain evolution data from the contact zone during thermoforming. This ensures a more representative characterization of friction under actual forming conditions, capturing the interaction between the mould and the thermoplastic sheet more effectively. Subsequently, a vacuum-assisted thermoforming simulation on a positive semi-spherical mould was performed to validate the calibrated model. The thickness distribution along the centerline of the sheet showed strong agreement with experimental data, particularly in capturing the thinning effect in highly stretched regions near the corner. This research improves predictive modeling for industrial thermoforming processes.