Back to search

National Technical University of Athens (NTUA)

Multiscale analysis of CNT and GnP-reinforced composites using FEM/XFEM

Abstract

dc:description

The subject of the current thesis is the multiscale analysis of nanocomposites reinforced by carbon nanotubes (CNTs) and graphene nanoplatelets (GnPs), using the conventional and extended finite element method (FEM/XFEM). Specifically, various modeling techniques are hierarchically applied, through different length scales from atomistic to nano, then to micro and finally to macro-scale, in order to study the mechanical and damping properties of the nano-reinforced polymer composites. Simulations of representative volume elements (RVEs) of nanocomposites are performed, where various stochastic parameters have been considered in order to account for the real microstructure geometry of the heterogeneous media.In the context of carbon nanotube-reinforced composites (CNT-RCs), the effect of interfacial shear strength (ISS) on the mechanical and damping properties of the material is investigated. The atomic lattice of CNTs is modeled using the molecular structural mechanics (MSM) approach and is reduced to an equivalent beam element (EBE). This beam is used as the basic building block for the construction of full length CNTs, which are then embedded in the polymer matrix. Elastic, as well as plastic, material properties are assigned to the EBEs for modeling their linear or nonlinear behavior, while the Maxwell-Wiechert material model is used for modeling viscoelasticity of the polymer. The interfacial load transfer mechanism between the lateral surface of the CNT and the surrounding matrix is taken into account with a nonlinear bond-slip friction-type model. Finite element (FE) models of RVEs are constructed comprised of two independent meshes: a structured with solid elements for the matrix and a series of embedded EBEs for the full length CNTs. Straight, as well as wavy CNTs, are considered. In the case of wavy CNTs, random CNT geometries are generated using the spectral representation method with evolutionary power spectra (EPS), which are derived from processing scanning electron microscope (SEM) images. Stochastic average properties are derived through Monte Carlo (MC) simulation. The mechanical and damping properties of CNT-RCs are assessed on the basis of sensitivity analyses with respect to various weight fractions (wf) and ISS values. Numerical results are presented, showing the significant effect of the ISS, as well as the influence of CNT waviness, on the damping behavior of CNT-RCs. Then, the multiscale modeling proceeds to macro-scale through the implementation of a nonlinear homogenization method. In the context of sequential homogenization, a novel viscoplastic constitutive model is introduced, which accounts for anisotropic stiffness and energy dissipation of the composite due to CNT reinforcement and slip. Sensitivity analysis is again performed with respect to various wf and ISS values where the mechanical and damping properties of the homogeneous models are assessed and compared with direct calculations on detailed fine scale heterogeneous models.The other class of materials studied in the current thesis is the graphene nanoplatelet-reinforced composites (GnP-RCs). Effective elastic properties are calculated from GnP-RC RVEs through a computational homogenization method, which accounts for arbitrarily shaped platelet inclusions. The homogenization combines the extended finite element method (XFEM) for the microstructural analysis with Monte Carlo simulation (MCS). The implementation of XFEM is particularly suitable for this type of problems since there is no need to generate a new finite element mesh at each MCS. The inclusions are randomly distributed and oriented within the medium while their shape is implicitly modeled by the iso-zero of an analytically defined random level set function, which also serves as the enrichment function in the framework of XFEM. Hill's energy condition is satisfied by the proposed homogenization method, which involves the generation of a large number of random RVE realizations. The microstructure geometries of these RVEs include specific volume fraction (vf) of inclusions with various stochastic parameters (e.g. number, shape, spatial distribution and orientation). The influence of the inclusion shape on the effective properties of the random media is highlighted. It is shown that the statistical characteristics of the effective properties can be significantly affected by the shape of the inclusions, especially in the case of large volume fraction and stiffness ratio.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2014

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Savvas, Dimitrios
  • Σάββας, Δημήτριος

Subjects

dc:subject × 16

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/36531

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Savvas, Dimitrios; Σάββας, Δημήτριος. Multiscale analysis of CNT and GnP-reinforced composites using FEM/XFEM. National Technical University of Athens (NTUA), 2014. http://hdl.handle.net/10442/hedi/36531