[Zurück]

A.Y. Matveeva:
"Structure-Property Relationships in Polymer Nanocomposites";
Betreuer/in(nen), Begutachter/in(nen): F.W.J. van Hattum, L. Gorbatikh, H. J. Böhm; Universidade do Minho, 2015; Rigorosum: 08.05.2015.

Carbon nanotubes/ carbon nanofibres (CNTs/CNFs) are considered to be among the most promising reinforcements for improving the mechanical properties of polymers while at the same time offering enhanced electrical and thermal conductivity. Because of their exceptionally high aspect ratio and high surface area in combination with a low density, already small volume fractions can potentially transfer their superior properties to a polymer matrix. However when used as reinforcements in polymeric composites, phenomena at the nano- and microscales, such as agglomerations, waviness/curliness, surface defects or imperfect bonding between the reinforcement and the polymer, can dramatically decrease the composite properties. The main objective of this work is to understand and to analyse how microstructural effects influence the overall composite behaviour on the macroscale in order to enable the full potential of these materials and to establish guidelines for the design of materials that meet specific requirements for the mechanical performance.

Dependent on the polymer system, different processing techniques based on high shear mixing were applied to enhance the dispersion of nanoparticles in the polymer system. It was shown that the elastic moduli of polymers reinforced with CNTs/CNFs do not increase linearly for high volume fractions. The reasons for this behaviour were investigated by means of modelling approaches together with data analysis using different microscopy techniques. Information about the dispersion and distribution of nanotubes was obtained using optical light microscopy (OLM), whereas transmission electron microscopy (TEM) was used for direct observations and information about their spatial geometry and orientation. Probability distribution functions for agglomeration, length, curliness and orientation were determined for CNT reinforced nanocomposites and used as input parameters for the developed models.

Preliminary studies were conducted to compare analytical models with Finite Element (FE) simulations, for modelling polymers reinforced with nanofillers of simplified geometries. They were aimed at investigating the influence of structural characteristics, such as curvature, orientation and dispersion of reinforcements, as well as of different FEs and boundary conditions on the effective elastic stiffness tensors.

Based on the obtained findings a two-step homogenisation method was developed. In the first step, curved nanotubes were uniformly distributed in cube-shaped volume elements using Monte-Carlo simulations according to the prior experimentally obtained probability distribution functions of orientation and curvature parameters. The mechanical properties of these volume elements with three-dimensional (3D) spatial oriented nanostructures were evaluated by an Eshelby based analytical model and periodic homogenisation using the FE code ABAQUS. The nanoreinforcements were treated as curved hollow cylindrical structures perfectly bonded with the polymer matrix. In the FE method inclusions were incorporated into the polymer matrix using an embedded element technique. In the second homogenisation step the composite was considered as a two-phase material consisting of agglomerates in an isotropic medium made of the polymer matrix with perfectly dispersed curved nanostructures. Information about the number and sizes of agglomerates was obtained from OLM analyses. A Mori-Tanaka model was used to calculate the effective elastic properties of the resulting composite.

The obtained Young´s moduli of polymeric nanocomposites were validated with experimental results, showing that the proposed technique may provide an attractive combination of accuracy, computational costs and flexibility for modelling arbitrary nanocomposites.