Results (
Thai) 2:
[Copy]Copied!
Graphene sheets (GS) have Young’s modulus and thermal
conductivity rivalling that of graphite (1.06 TPa and
3000W m−1 K−1 respectively) [1, 2]. They may exist as single
layered or multi-layer structures. It is possible to harness
the multifunctional properties of graphene sheets and design
novel class of advanced composites with superior mechanical
and electric performance [1–3], as well as innovative strain
sensors [5]. An approach to produce graphene–polymer
composites by complete exfoliation of graphite and molecularlevel
dispersion of GS in a polymer host has been described
in [4]. The latter work, from Stankovich et al, has fuelled
4 Address for correspondence: Department of Aerospace Engineering,
University of Bristol, Queens Building, University Walk, Bristol BS8 1TR,
UK.
a growing interest into the mechanical determination and
characterization of single layer graphene sheets (SLGS),
although from the experimental point of view advances have
been made in measuring magneto-transport properties [9],
while experimental mechanical data are still confined to
graphene layers only. The enhanced flexibility of GS,
despite their high Young’s modulus, has been attributed to the
change in curvature given by reversible elongation of sp2 C–C
bonds [6, 8, 49]. Vibrational properties of SLGS [10] or multilayer
graphene assemblies [7] have also been evaluated using
analytical and finite element simulation methods.
Molecular mechanistic modelling of single layer graphene
sheets has been pursued by several authors. Simple lattice
models with force constants derived from an assumed
potential have been developed by Bacon and Nicholson [11]
and Gillis [12]. Ab initio methods have been used by Kudin et al [13], who predict a Young’s modulus of 1.02 TPa
and Poisson’s ratio of 0.149, and Van Lier et al [14],
reporting a Young’s modulus for graphene equal to 1.11 TPa.
Several authors have made use of Tersoff–Brenner potentials
to describe the mechanical properties of single graphene
sheets, with Young’s modulus predictions between 0.694 and
0.714 TPa, and Poisson’s ratios from 0.397 to 0.417 [15, 16].
Brenner’s potential and Cauchy–Born rule have also been used
by Reddy et al [23] to describe the mechanical properties of
finite size graphene sheets, highlighting the difference between
minimized and unminimized strain energy configurations for
the SLGS under different types of loading. Rajendran and
Reddy have also indicated the maximum numerical precision
attainable using the above methods to calculate the stiffness
of SLGS and carbon nanotubes [22]. Second generation
Brenner potentials have also been used by Huang et al
[17] to calculate the in-plane Young’s moduli, Poisson’s
ratios and thickness of GS and single wall carbon nanotubes
(SWCNTs), with stiffness values for the GS ranging from
2.99 to 4.23 TPa, and Poisson’s ratio of around 0.397. A
rigorous homogenization technique has been also developed
by Caillerie et al [25] to calculate first Piola–Kirchhoff and
Cauchy stress tensors considering stretching and bond angle
variation. Other analytical models of SLGS incorporating
energy contributions from bond stretching and changes in
bond angle have been proposed as subset of analogous models
for single wall carbon nanotubes, [50]. The mechanical
properties of different graphene configurations, with chiral
index n towards infinity, have been extracted from analytical
nanotubes models. Hemmasizadeh et al [24] have also used a
mixed MD–continuum mechanics model based on thin shell
theory to obtain the properties of SLGS from a numerical
nanoindentation simulation.
Another approach widely used recently for nanostructures
modelling is the equivalent atomistic continuum–structural
mechanics approach, pioneered by Odegard et al [18] and Li
and Chou [19]. In this approach, typical elements of structural
mechanics, such as rods, beams and shells are used with
appropriate mechanical properties to simulate the static and
dynamic behaviour of graphene layers and carbon nanotubes.
The mechanical properties for the structural elements are
derived from equilibrium between harmonic steric potentials
of the C–C bonds and mechanical strain energies associated to
tension, torsion and bending related to the mechanical elements
simulating the bonds themselves. A truss model was proposed
in [18], wherein rods of different stiffnesses represent the
stretching and in-plane bending capabilities of the C–C bonds.
Reddy et al [21] extended the model from [18] to account for
the orthotropy generated in finite size graphene sheets. Meo
and Rossi [20] developed a finite element model comprising
uniaxial links and nonlinear rotational spring to represent
the modified Morse potential when simulating graphene and
carbon nanotube structures. Tserpes and Papanikos [33]
identified thickness and stiffness properties (Young’s and shear
modulus) for an equivalent material associated to the C–C
bonds represented by finite element beams in single walled
carbon nanotubes (SWCNTs). The approach of [33] has been
used by Sakhaee-Pour et al [10] to compute natural frequencies
and modes of single graphene sheets, and to characterize the
in-plane properties of SLGS with different chirality [34].
The wide dispersion of the mechanical properties of
graphene sheets can be attributed principally to the uncertainty
associated to the thickness of these nanostructures. For
the majority of models used, the assumed thickness of the
graphene layer is 3.4 A° , equal to the one of a graphite layer.
The 3.4 °A value provides in-plane Young’s modulus of the
order of 1 TPa. However, several models related to graphene
and single wall nanotubes have indicated thickness values
ranging from 0.57 A° [17] to 6.9 A° [18]. In SWCNTs, the
dispersion of mechanical properties associated with thickness
and stiffness has been known as the ‘Yakobson’s paradox’ [29].
From the modelling point of view, thickness becomes
also important when considering the equivalent structural
mechanics approach. Sun et al [30] determined a thickness for
the C–C bond in SWCNTs of 1.2 °A for an equilibrium length
of 1.42 A° , coupling chemical potentials with Kirchhoff–Love
thin shell theory. However, the use of an isotropic thin shell
can be considered valid for nanotubes with radius/thickness
ratio higher than 10 [31], and for the first order of error of
the ratio between atomic spacing and SWCNT radius [32].
In [33], the thickness of the Euler–Bernoulli (EB) beam
element representing the C–C bond for a carbon nanotube is
1.47 A° , corresponding to an equilibrium length of 1.42 A° . On
the other hand, EB theory can be applied only to slender beams
with aspect ratios higher than 10 [38]. An improved model
has been proposed by Scarpa and Adhikari [33] considering
deep beam theory, where shear correction factors depending
on the cross-section and Poisson’s ratio of the equivalent C–C
bond material are taken into account. Using the AMBER force
model [41], the thickness value for the C–C bond for 1.42 °A
of length is 0.84 A° , with a Poisson’s ratio of 0.0032. The C–C
bond has therefore a negligible mechanical lateral deformation
when stretched or compressed, behaving like cork [26].
Reddy et al [23] have also highlighted the variation
in bond lengths present in finite graphene sheets, as well
as the in-plane orthotropy of single GS. Equilibrium bond
lengths in finite graphene sheets up to 120 atoms have been
observed varying between 1.39 and 1.47 °A under potential
energy minimization for different in-plane loading behaviour.
Special orthotropy [37] has also been recorded in two and
four straight edges SGS configurations, with anisotropy degree
between 0.92 and 0.99 [23]. The special orthotropy behaviour
is particularly interesting, because it is also observed in
common structural honeycomb configurations, even when
regular hexagonal topologies are considered [27]. The other
linear elastic models available for SGS predict however
an isotropic in-plane mechanical behaviour, with Young’s
modulus Y constant along the principal directions, with inplane
shear modulus G obeying the relation G = Y/2/(1+ν).
Analytical models of structural honeycombs and cellular solids
are able to simulate in-plane mechanical properties [36, 37].
Structural honeycombs have ribs made of elements behaving
like structural beams, with stretching and bending capabilities,
and hinging important for high relative densities and damage
at the base of the ribs [36].
In this work we develop closed form solutions for
the in-plane elastic properties of SLGS using mechanical mechanics approaches. A full analytical truss-lattice model
of the SGS based on the geometry proposed in [18] is
developed, and the rigidity matrix coefficients computed. A
cellular solid micromechanics model is also developed for the
hexagonal honeycomb configuration of the graphene, based on
the theoretical framework proposed by Masters and Evans [36].
The evaluation of the force constants to be used in this
model is based on the equivalent mechanical properties of
the C–C bond. Using a deep beam theory model with
Timoshenko shear correction factor, it is possible to determine
the thickness, Poisson’s ratio and equivalent Young’s modulus
of the C–C bond material, and use this values to provide the
stretching, flexural and hinging capabilities of the equivalent
beam elements constituting the SGS lattice. Equilibrium C–
C bond lengths are evaluated from models of the bracedtruss
and honeycomb lattice using finite element models with
2096 atoms under different in-plane loading (uniaxial and pure
shear). The models are used to identify bond lengths and
thickness distributions related to minimum potential energy
configurations for imposed strains up to 0.01%. The finite
element results provide a benchmark of the analytical models
developed, together with a critical assessment from results in
open literature.
Being translated, please wait..
