Computational Analysis of CNT Reinforced Polymer using Nanoscale Informed Morphology

Accurate prediction of the mechanical response of carbon nanotube (CNT)-reinforced polymer requires capturing critical morphological features and load transfer mechanisms at the constituent interface across multiple length scales. This article presents a unique methodology that utilizes morphology information from molecular dynamics (MD) simulations to generate continuum scale representative volume elements (RVEs) of polymer with randomly dispersed CNTs. First, a novel coarse-grain MD approach is developed to overcome the limitations of using a traditional all-atom approach to generate large-sized atomistic models with realistic CNT aspect ratios. The developed approach can generate systems with wavy and entangled CNT clusters resulting from the complex polymer curing process. At the continuum scale, the CNTs are modeled as solid cylinders, and their cluster morphology is reconstructed using the information obtained from coarse-grain simulations. The constructed CNT cluster geometry is triply periodic and is embedded within a structured grid representative of the host polymer matrix subject to damage. The resulting RVEs are homogenized using finite element techniques with periodic boundary conditions. The predicted effective properties are investigated and compared with test results for different weight fractions of CNTs.