Computer simulation of curing and toughening of epoxy systems

Abstract

The crosslinking and toughening process of diglycidyl ether of bisphenol A (DGEBA) resin was simulated by various computational techniques, using a commercial polymer modeling software package. First, curing of DGEBA resin with three different curing agents was simulated using a Monte Carlo simulation technique. Results calculated for crosslinking conversion of the formed network showed that deviation from an ideal network due to loops and dangling chains increased with excess amounts of the curing agent and that the formed network is close to the ideal when stoichiometric concentrations are used. The maximal calculated modulus was an indication of the optimal curing agent concentration. The glassy modulus and Tg for the simulated systems were calculated using the group contribution method and semiempirical correlations. Second, the simulation results of multicomponent epoxy systems, comprising two curing agents, a reactive diluent and DGEBA resin, indicate that there is no difference in network quality compared with bicomponent epoxy systems, comprising DGEBA and a single curing agent. The simulation results exemplified the ability to choose optimal components concentration in a complicated multicomponent epoxy system. Third, the toughening process of amino-terminated butadiene acrylonitrile (ATBN) and carboxyl-terminated butadiene acrylonitrile (CTBN) in DGEBA resin was analyzed using the solubility parameter approach. This approach could explain the role of the rubber acrylonitrile group in epoxy/rubber blends. The interaction parameters of both systems (ATBN and CTBN) and their phase diagrams were estimated using modified Flory—Huggins theory. It was shown that this technique leads to good estimations of the optimal rubber concentration, leading to optimal mechanical properties.