Inverse analysis for transient moisture diffusion through fiber-reinforced composites

A new approach has been developed, based on an inverse analysis technique, to determine critical moisture diffusion parameters for a fiber-reinforced composite. This technique incorporates two distinct features: direct experimental observations of the weight gained by a composite material exposed to a humid environment, and highly detailed computational analyses that capture the actual heterogeneous microstructure of the composite. The latter feature was carried out by modeling more than 1000 individual carbon fibers that are randomly distributed within an epoxy matrix. The verification and efficacy of this technique was established by conducting an experiment on a high-grade IM7/997 carbon fiber-reinforced epoxy to determine the maximum moisture content at saturation and the diffusivity of epoxy. With the inverse analysis, the time duration required to estimate these moisture diffusion parameters could be drastically reduced as compared to conventional procedures. Subsequently, the established models were employed to characterize transient moisture absorption process within the composite. Here, it was demonstrated that modeling the heterogeneous microstructure of the composite is critical for obtaining accurate diffusion parameters, and an analytical model with effective properties does not produce correct transient moisture absorption behavior. Furthermore, the evolution of stress fields due to moisture induced volumetric expansion was quantified. It was observed that high stress concentrations develop in regions of fiber concentration. These regions then act as potential failure initiation sites that can lead to lower damage tolerance.