The effect of the interface between the carbon nanofibers (CNF) or carbon nanotubes (CNT) and the matrix on the properties of a nanocomposite is significant. Therefore, it is important to develop a model that accounts for the effect of the interface on the overall elastic modulus of a nanocomposite. A methodology for predicting the modulus of a nanocomposite is developed in this paper. This methodology uses a conceptual model called inclusion that consists of the nanofiber/nanotube inserted in a hollow cylinder representing the interface. The inclusion is embedded into the matrix. The flexibility of the hollow cylinder represents the effect of the interface. Inputs to this model are the moduli of the CNT/CNF, the matrix and the material representing the interface, as well as the volume fractions of these materials. These properties can be estimated from experiments or from molecular dynamics models.
There is significant uncertainty in the properties of the CNT/CNF, the matrix of nanocomposites and their interface. This is particularly true for the elastic modulus of the CNT/CNF. These uncertainties can have an important effect on the performance of nanocomposite materials. Therefore, it is important to develop an approach for quantifying the uncertainty in the properties of the CNT/CNF, the matrix and their interface and for estimating the resulting uncertainty in the elastic modulus of the nanocomposite. In this study, probabilistic models are used to model the CNT/CNF, matrix and interface properties. Uniform probability distributions are selected for the moduli of CNT/CNF, the matrix and the interface when only upper and lower bounds of these properties are known. A Monte-Carlo simulation approach is used together with a micromechanics model to estimate the probability distribution of the modulus of the composite. The method is demonstrated on an example.
Finally, a methodology for predicting the nonlinear static response and the reliability of plates made of nanocomposites is presented and demonstrated.