EUROPEAN CONGRESS AND EXHIBITION ON ADVANCED MATERIALS AND PROCESSES - EUROMAT 2021

Computational techniques have been important complementary materials design tools because they are able to predict mechanical properties of new materials formulations as well as provide microscopic details of deformation and failure mechanisms. Continuum mechanics based techniques have been extensively used in the design of conventional composites. However, when the material contains nanoscopic material heterogeneities, a particle-based physical description is required which is able to account for the different deformation and failure mechanisms characteristic of the molecular level. Although desirable, molecular simulations of systems longer than a few nm would require a prohibitively large number of particles at a fully atomic resolution. As is recognized by the simulation community, an effective coarse-graining is necessary in order to extend the viable system sizes while keeping the adequate physical description.

In this study, we adapt a complex coarse-graining procedure in order to generate coarse-grained (CG) models for crosslinked epoxies. In the first step, we perform cross-linking simulations in stoichiometric mixtures of an epoxy resin and a hardener at the all-atom level. We finetune the simulation parameters of the reactive steps in order to obtain a volume evolution consistent with experiments. In the next step, we compute characteristic distribution functions (of distances and angles) which are used to derive the CG interaction parameters through a combination of Boltzmann inversion and requiring mechanical equivalence between the CG and atomic models. In order to enable the system to undergo failure, in the last step, we introduce a quantum chemically derived bond potential able to break upon large separations. Besides demonstrating the feasibility of such a bottom-up coarse-graining strategy, we will present fracture properties of a model epoxy system as a function of the cross-linking degree. This work is intended to lay the groundwork for future coarse-grained computer studies dedicated to the failure properties of nano-reinforced epoxies.