In recent years, polymer nano-composites (PNCs) have increasingly gained more attention due to their improved mechanical, barrier, thermal, optical, electrical and biodegradable properties in comparison with the conventional micro-composites or pristine polymer. With a modest addition of nanoparticles (usually less than 5wt. %), PNCs offer a wide range of improvements in moduli, strength, heat resistance, biodegradability, as well as decrease in gas permeability and flammability. Although PNCs offer enormous opportunities to design novel material systems, development of an effective numerical modeling approach to predict their properties based on their complex multi-phase and multiscale structure is still at an early stage. Developing a computational framework to predict the mechanical properties of PNC is the focus of this dissertation. A computational framework has been developed to predict mechanical properties of polymer nano-composites. In chapter 1, a microstructure inspired material model has been developed based on statistical technique and this technique has been used to reconstruct the microstructure of Halloysite nanotube (HNT) polypropylene composite. This technique also has been used to reconstruct exfoliated Graphene nanoplatelet (xGnP) polymer composite. The model was able to successfully predict the material behavior obtained from experiment. Chapter 2 is the summary of the experimental work to support the numerical work. First, different processing techniques to make the polymer nanocomposites have been reviewed. Among them, melt extrusion followed by injection molding was used to manufacture high density polyethylene (HDPE) – xGnP nanocomposties. Scanning electron microscopy (SEM) also was performed to determine particle size and distribution and to examine fracture surfaces. Particle size was measured from these images and has been used for calculating the probability density function for GNPs in chapter 1. A series of nanoindentation tests have been conducted to reveal the spatial variation of the superstructure developed along and across the flow direction of injection-molded HDPE/GNP.The uniaxial tensile test and shear test have been conducted on HDPE and xGnP/HDPE specimens. The stress-strain curves for HDPE obtained from these experiments have been used in chapter 5 to calibrate the modified Gurson–Tvergaard–Needleman to capture the damage progression in HDPE. In chapter 3, the 3D microstructure model developed in chapter 1 was incorporated in a damage modeling problem in nanocomposite where damage initiation has been modeled using cohesive-zone model. There is a significant difference between the properties of inclusion and the host polymer in polymer nanocomposite, which leads to the damage evolution during deformation due to a huge stress concentration between nanofiller and polymer. The finite element model of progressive debonding in nano-reinforced composite has been proposed based on the cohesive-zone model of the interface. In order to model cohesive-zone, a cohesive zone traction displacement relation is needed. This curve may be obtained either through a fiber pullout experiment or by simulating the test using molecular dynamics. In the case of nano-fillers, conducting fiber pullout test is very difficult and result is often not reproducible. In chapter 4, molecular dynamics simulation of polymer nanocomposite has been performed. One of the goals was to extract the load-displacement curves of graphene/HDPE pullout test and obtain cohesive zone parameters in chapter 3. Finally, in chapter 5, a damage model of HDPE/GNP nanocomposite has been developed based on matrix cracking and fiber debonding. This 3D microstructure model was incorporated in a damage modeling problem in nanocomposite where damage initiation and damage progression have been modeled using cohesive-zone and modified Gurson-Tvergaard-Needleman (GTN) material models.
Read
