Constitutive behavior and failure prediction of crosslinked polymers exposed to concurrent fatigue and thermal aging: a reduced-order knowledge-driven machine-learned model

To assess the combined effects of thermal aging and cyclic fatigue on the constitutive and failure behavior of crosslinked polymers, a novel physics-informed data-driven constitutive model is proposed. While extensive research has been conducted on the individual degradation modes, the simultaneous impact of these two aging mechanisms has not been explored in depth. Our objective is to simulate the mechanical properties degradation resulting from simultaneous exposure to these two aging processes. To accurately capture the complex interactions between the two aging processes, we developed a physics-informed machine-learning approach that uses the respective aging mechanisms to model the effects of each aging condition separately. We then couple the kinetic equations representing the damage associated with each aging condition using the conditional neural network. By doing so, our model can account for the mechanical and environmental damage synergy on the constitutive response of the polymer macromolecular network. The model is built based on the assumption of complete independence of mechanical and environmental effects, building upon our recent models of thermal-oxidative aging and hydrolytic aging. The model is validated using a wide range of experimental data based on sequential aging. The suggested model showed a satisfactory correlation with the experimental results, demonstrating its potential for predicting the long-term durability and reliability of materials subjected to thermal aging and cyclic fatigue. Overall, our approach provides a valuable tool for designing and optimizing materials that are subjected to complex aging conditions.