The application of epoxy resin composites in the marine environment is hindered by long-term modulus changes and unpredictable structural failure due to water- induced internal damage. Water modifies the viscoelastic behavior of the epoxy resin matrix as well as the microstructure of the composite. One of the macroscopic observations of the effect of water is the water- induced w-transition in an isochrone of the dynamic modulus. Because this transition is linked to water sorption and internal damage within the composite, identification of its cause is important. This work hypothesizes, with supporting modeling, a possible mechanism for the w-transition. The hypothesis states thata heterogeneous distribution of water within the resin is the source of the transition. The validity of the hypothesis is tested by qualitatively comparing the transitions observed in the predicted dynamic moduli to the experimental moduli.
In order to predict the dynamic moduli of the heterogeneous material, the properties of the individual phases, including the viscoelastic moduli of the un-measurable water-rich epoxy resin phase, are necessary. A molecular-level model, developed to account for the interaction of time, temperature, and diluents on the viscoelastic properties of crosslinked polymers, predicts the properties of a model epoxy resin with various water contents. The volume-averaged dynamic moduli of the heterogeneous resins and composites with the hypothesized water-rich phase are calculated using micromechanical averaging techniques. The appearance of transitions in the predicted dynamic moduli similar to experimental w-transitions supports the proposed hypothesis. The molecular-level model not only provides a functional form to fit the experimental isochrone data of the resin and a predictive equation for the isothermal frequency response, but also aids in the study of time-temperature superposition and its validity in the presence of water.
