Multiscale Modeling of Epoxies and Epoxy-Based Composites
269
This also highlights the need to develop a multiscale modeling approach in which
information is upscaled from the appropriate length scale to the next one, either
hieratically or concurrently, to better predict the deformation and failure of epoxies
and composites. This would provide a promising avenue to greatly accelerate the
design of advanced epoxy-based materials.
While a number of comprehensive review articles on epoxies and epoxy-based
composites have been published in the last 10 years [36–38], there is still a
lack of a comprehensive review focused on multiscale epoxy modeling efforts.
Thus, in this chapter, recent research on multiscale simulations of epoxies and
epoxy-based composites are discussed. In Sect. 2, common simulation methods are
briefly discussed for the purpose of outlining the building blocks of a multiscale
framework. In Sect. 3, various models that predicte the molecular structure and
thermomechanical properties of epoxies are reviewed. The majority of studies
discussed in this paper are on two of the most common epoxies, diglycidyl ether
of bisphenol A (DGEBA) and diglycidyl ether of bisphenol F (DGEBF). The
properties of epoxy-based composites, coatings, and adhesives are then discussed
in Sect. 4. Finally, a summary and conclusions are given in Sect. 5.
2 Overview of Multiscale Simulation Methods for Epoxies
2.1 Molecular Dynamics Simulation
Molecular dynamics simulations are based on modeling many-body dynamics at
the atomic scale using Newton’s second law. Atomic structures and force fields are
two important inputs for an MD simulation, and both can be obtained from DFT
calculations.
The commonly used force fields in classical MD simulations include CVFF,
PCFF, COMPASS, Dreiding, and other modified and mixed potentials [16, 39, 40].
Many studies have been performed using classical (nonreactive force fields) MD
simulations to explore the molecular topology, degree of cross-linking, water
absorption, strain rate, temperature, and effects of force fields on the density,
modulus, fracture, glass transition temperature, and coefficient of thermal expansion
of various epoxy systems [16, 31, 40–49]. In these simulations, nonreactive force
fields are utilized, and bonds are created or broken based on criterion upscaled
from DFT simulations. In addition, reactive force fields (i.e., ReaxFF) are also
widely used when the chemical reactions are of interest in the epoxy simulations
[28, 29, 50]. Different types of force fields used in various epoxy studies are listed
in Table 1.
The MD simulations can typically model epoxy systems that are up to tens of
nanometers in edge lengths and reach time scales of up to a few nanoseconds. As
the examples in Table 1 show, the size of the MD simulation cells often consists of
few hundreds of combined prepolymer and curing agents, and a simulation system
that is larger than a few thousand molecules is rare due to the high computational
cost demands associated with such simulations.
269
This also highlights the need to develop a multiscale modeling approach in which
information is upscaled from the appropriate length scale to the next one, either
hieratically or concurrently, to better predict the deformation and failure of epoxies
and composites. This would provide a promising avenue to greatly accelerate the
design of advanced epoxy-based materials.
While a number of comprehensive review articles on epoxies and epoxy-based
composites have been published in the last 10 years [36–38], there is still a
lack of a comprehensive review focused on multiscale epoxy modeling efforts.
Thus, in this chapter, recent research on multiscale simulations of epoxies and
epoxy-based composites are discussed. In Sect. 2, common simulation methods are
briefly discussed for the purpose of outlining the building blocks of a multiscale
framework. In Sect. 3, various models that predicte the molecular structure and
thermomechanical properties of epoxies are reviewed. The majority of studies
discussed in this paper are on two of the most common epoxies, diglycidyl ether
of bisphenol A (DGEBA) and diglycidyl ether of bisphenol F (DGEBF). The
properties of epoxy-based composites, coatings, and adhesives are then discussed
in Sect. 4. Finally, a summary and conclusions are given in Sect. 5.
2 Overview of Multiscale Simulation Methods for Epoxies
2.1 Molecular Dynamics Simulation
Molecular dynamics simulations are based on modeling many-body dynamics at
the atomic scale using Newton’s second law. Atomic structures and force fields are
two important inputs for an MD simulation, and both can be obtained from DFT
calculations.
The commonly used force fields in classical MD simulations include CVFF,
PCFF, COMPASS, Dreiding, and other modified and mixed potentials [16, 39, 40].
Many studies have been performed using classical (nonreactive force fields) MD
simulations to explore the molecular topology, degree of cross-linking, water
absorption, strain rate, temperature, and effects of force fields on the density,
modulus, fracture, glass transition temperature, and coefficient of thermal expansion
of various epoxy systems [16, 31, 40–49]. In these simulations, nonreactive force
fields are utilized, and bonds are created or broken based on criterion upscaled
from DFT simulations. In addition, reactive force fields (i.e., ReaxFF) are also
widely used when the chemical reactions are of interest in the epoxy simulations
[28, 29, 50]. Different types of force fields used in various epoxy studies are listed
in Table 1.
The MD simulations can typically model epoxy systems that are up to tens of
nanometers in edge lengths and reach time scales of up to a few nanoseconds. As
the examples in Table 1 show, the size of the MD simulation cells often consists of
few hundreds of combined prepolymer and curing agents, and a simulation system
that is larger than a few thousand molecules is rare due to the high computational
cost demands associated with such simulations.