Mechanics of deformation of carbon nanotube-polymer nanocomposites

Abstract

The goal is to develop finite element techniques to evaluate the mechanical behavior of carbon nanotube enabled composites and gain a thorough understanding of the parameters that affect the properties of the composite, both micro- and macroscopically. Micromechanical models of representative volume elements (RVEs) of unit-cell and random multi-particle distributions are used to study such parameters and their performance and accuracy in doing is compared and discussed. The microstructural parameters of interest can be loosely categorized in two groups: those related to the geometry of the composite and those associated with the matrix-nanotube interactions as well as the load transfer mechanisms along the interface and inside the nanotubes. Among the geometry-related parameters, of particular interest are the nanotube aspect ratio, the number of walls, as well as the weight and volume fraction of nanotubes, their distribution and alignment in the matrix and their curvature. In terms of the matrix-nanotube interactions, emphasis is given on the bonds developed between the matrix and the nanotube and their effect on load transfer. The amount of load transferred internally in multi-wall nanotubes is also investigated. A number of models have been created and finite element methods have been employed to analyze the macroscopic mechanical behavior of nanotube-enabled composites, using the axial stiffness as the common metric in all cases. Fully functionalized matrix-nanotube interfaces have enabled the separate investigation of load transfer internally in multi-wall nanotubes. Unit-cell RVEs with appropriate periodic boundary conditions to emulate regular stacked or regular staggered arrays of nanotubes within a matrix,

(cont.) highlight the deficiency of using stacked array RVEs for assessing macroscopic properties. Unit-cell RVEs with staggered boundary conditions enable the detailed examination of issues, regarding modelling of the layered nature of nanotube walls. However, they do not fully capture the effects associated with the distribution of the nanotubes in the matrix. The focus shifted on accurately defining a RVE by analyzing nanotube dispersion in the matrix statistically, with emphasis on the proximity of neighboring particles. Simulated random distributions are studied in terms of the degree of filler clustering and its effect on composite stiffness and compared to nanotube distributions obtained from SEM images of actual composites. As a result, multi-particle finite element models are developed, based on these random distributions. They allow investigation of randomness in alignment, dispersion and curvature and are able to capture the characteristics and behavior of actual nanotube-enabled polymer composites more accurately than unit-cell models.