Influence of loading mode on the prediction accuracy of nonlinear viscoelastic models for soft adhesives

The precise modeling of nonlinear viscoelastic behavior in soft adhesives plays an essential role in their application in advanced technologies. Linear viscoelastic models, such as hyperelasticity with the Prony series, have exhibited limitations in accurate prediction of material behavior under large strains. This study calibrates a nonlinear viscoelastic three-network viscoplastic (TNV) model based on experimental data to effectively capture the complex mechanical behavior of soft adhesives under large strains. A series of dynamic oscillatory shear tests, monotonic (i.e., continuous uniaxial shear tests) and creep tests, and cyclic loading tests were employed to explore the full range of material responses. Based on the results, the TNV model outperformed linear ones in prediction of stress, energy dissipation, and material softening. Furthermore, a low-cycle loading–unloading test was identified as an effective and cost-efficient approach for calibration of the constitutive models. The validation of the extracted model demonstrated its high accuracy in predicting the behavior of soft materials under various loading scenarios, particularly cyclic loading conditions, thereby offering valuable insights into the mechanical performance of soft adhesives, especially in flexible display technologies. This research establishes the groundwork for more efficient material characterization with minimum testing, providing a reliable framework for predicting the long-term performance of visco-hyperelastic materials in large strain applications.