Stochastic multi-scale modeling of randomly grown CNTs on carbon fiber

The main objective of this article is to develop a stochastic multi-scale approach for predicting mechanical properties of a carbon fiber coated with randomly oriented carbon nanotubes (CNT) known as a fuzzy fiber. The core fiber and surrounding CNT/polymer nanocomposites can be considered as two constitutive phases of a fuzzy fiber. Firstly, effective properties of carbon nanotubes reinforced polymer are estimated while non-bonded van der Waals (vdW) interaction is assumed between carbon nanotubes and polymer. Then, effective properties of the fuzzy fiber are predicted. The curvature, volume fraction and orientation of CNTs are considered as random variables in the developed modeling procedure. For validation purpose, a comparison is done between the obtained results from developed computational modeling and published experimental data in literature for a single fuzzy fiber containing randomly oriented CNTs. an excellent conformity is realized between developed modeling procedure and experimental observations demonstrating the importance of taking into account random parameters. Finally, the influence of CNT contents on improving the transverse Young's modulus of investigated fiber is studied.