High-performance, multi-component epoxy resin simulation for predicting thermo-mechanical property evolution during curing

Abstract

High-performance epoxy systems are extensively used in structural polymer‒matrix composites for aerospace vehicles. The evolution of the thermomechanical properties of these epoxies significantly impacts the evolution of process-induced residual stresses. The corresponding process parameters need to be optimized via multiscale process modeling to minimize the residual stresses and maximize the composite strength and durability. In this study, the thermomechanical properties of a multicomponent epoxy system are predicted via molecular dynamics (MD) simulation as a function of the degree of cure to provide critical property evolution data for process modeling. In addition, the experimentally validated results of this study provide critical insight into MD modeling protocols. Among these insights, harmonic- and Morse-bond-based force fields predict similar mechanical properties. However, simulations with the Morse-bond potential fail at intermediate strain values because of cross-term energy dominance. Additionally, crosslinking simulations should be conducted at the corresponding processing temperature, because the simulation temperature impacts shrinkage evolution significantly. Multiple analysis methods are utilized to process MD heating/cooling data for glass transition temperature prediction, and the results indicate that neither method has a significant advantage. These results are important for effective and comprehensive process modeling within the ICME (Integrated Computational Materials Engineering) and Materials Genome Initiative frameworks. Molecular dynamics simulations were utilized to model and predict thermomechanical properties of a multi-component high-performance epoxy. Virtual curing between the molecules is simulated using REACTER protocol in LAMMPS and LUNAR tool is utilized to assign force field parameters and post-processing of simulated data. Insights into the predicted properties using harmonic- and morse-bond-based force fields, strain-rate sensitivity, glass transition temperature predicted from heating and cooling simulations are provided. The comprehensive set of properties are predicted as a function of processing temperature and crosslinking density required for multi-scale process modeling high-performance epoxy.