Πρόβλεψη συμπεριφοράς και αστοχιών νανοδομών με υπολογιστικές διαδικασίες

Abstract

The real progress in the last decade in nanotechnology, has been due to a series of advances in a variety of complementary areas, such as: the discoveries of atomically precise materials such as nanotubes and fullerenes; the ability of the scanning probe and the development of manipulation techniques to image and manipulate atomic and molecular configurations in real materials; the conceptualization and demonstration of individual electronic and logic devices with atomic or molecular level materials; the advances in the self-assembly of materials to be able to put together larger functional or integrated systems; and above all , the advances in computational nanotechnology ,i.e., physics- and chemistry- based modeling and simulation of possible nanomaterials, devices and applications. It turns out that at the nanoscale, devices and systems sizes have shrunk sufficiently small, so that, it is possible to describe their behavior fairly accurately. The simulation technologies have become also predictive in nature, and many novel concepts and designs have been first proposed based on modeling and simulations, and then were followed by their realization or verification through experiments. The development of computational procedures that are able to accurately simulate and predict the mechanical behaviour and failures of basic nanostructures is the aim of the present thesis. A spring based finite element method is developed and utilized to provide numerical results about the elastic, dynamic and nonlinear behaviour of the major carbon allotropes, such as graphene, graphite flakes and carbon nanotubes. The molecular mechanics theory provides the force fields that are used as the base for the spring elements formulation. The optimized atomistic geometry of nanostructures as graphene nanoribbons, graphite flakes, as well as single and multi walled carbon nanotubes as derived by the potential energy minimization is used to be defined their discrete geometry and the corresponding finite element models. Furthermore, multi-scale finite element models are formulated for the prediction of elastic and nonlinear mechanical behavior of carbon nanotube reinforced nanocomposites. Representative volume elements are implemented according to the micromechanical theory. Matrix materials such polymers, metals or rubber are considered as continuum mediums, whereas the reinforcement is modeled as a discrete structure. The interfacial zone between matrix and reinforcement is approached by appropriate elements and their stiffness properties are computed by using physical assumptions