Molecular Dynamics and Finite Element Investigation of Polymer Interphase Effects on Effective Stiffness of Wavy Aligned Carbon Nanotube Composites

Abstract

Interphase effects have been studied for their effect on composite properties for many decades, and it is well documented that an interphase can exist in polymer composites comprised of nanofibers as well. We present a first study of interphase effects on the basic elastic response of wavy aligned carbon nanotube (A-CNT) polymer nanocomposties (PNCs). Waviness is characterized by ex situ pre-fabrication imaging of A-CNT forests and used as an input to finite element analyses of the PNCs containing an interphase region in the thermoset polymer defined using molecular dynamics (MD) simulations. The interphase thickness of ~1nm is found to be independent of crosslink density and contain regions of both higher and lower mass density than the bulk polymer. Finite element analyses of wavy single and double-wall A-CNT PNCs incorporating this interphase, allow the effective stiffness based on a representative volume element to be calculated. Waviness of the A-CNTs dominates the effective axial stiffness of the PNCs, with the interphase having a negligible effect. The interphase changes the stress and strain distribution local to the CNT 'fiber' and this is expected to play an important role in failure, such as CNT pullout, of the CNTpolymer system. These findings, in addition to the relatively high volume fraction (V f ) of the interphase in PNCs with high CNT V f , suggests that the interphase may play a more important role in PNCs than in micron-scale (typical) collimated fiber composites. Future work in this area includes inelastic polymer response during nanofiber pullout. Nomenclature A = waviness amplitude of CNT A-PNC = aligned-CNT polymer nanocomposite Case-A = DWCNT outer wall contribution to stiffness Case-B = DWCNT all walls contribution to stiffness CNT = carbon nanotube CFRP = carbon fiber reinforced plastic DWCNT = double walled carbon nanotube d i = inner diameter of CNT d o = outer diameter of CNT 2 E RVE = axial modulus of representative volume element E w = wall modulus FEA = finite element analysis FRP = filament fiber reinforced plastic HR-SEM = high resolution scanning electron microscope L = length of wavy CNT MD = molecular dynamics PNC = polymer nanocomposite RVE = representative volume element SWCNT = single walled carbon nanotube t w = wall thickness w = waviness ratio of CNT

Mechanical properties of graphene nanoplatelet/carbon fiber/epoxy hybrid composites: Multiscale modeling and experiments

“…It is important to note that the properties in Table 1 However, they can be estimated using methods described elsewhere [11,18,36].…”

Influence of Waviness on the Elastic Properties of Aligned Carbon Nanotube Polymer Matrix Nanocomposites

“…67 showed that when the interphase size is very small and the CNTpolymer interfacial strength is 150 MPa, as reported by recent studies, [68][69][70] the assumption of perfect bonding will have a very small impact on the predicted effective modulus of the A-PNCs. A recent study showed that for an interphase thickness of ∼ 1 nm, the reinforcement modulus of the CNTs in the A-PNCs would be diminished by 3% for 0.1 ≤ w ≤ 0.5, 71 which also confirms that interphase effects will have a very weak influence on E pnc (V f ) (from Eq. 2c).…”

High-Volume-Fraction Textured Carbon Nanotube–Bis(maleimide) and −Epoxy Matrix Polymer Nanocomposites: Implications for High-Performance Structural Composites

“…It is well-known that interfacial interactions can dominate the mechanical and thermal behavior of composites even at low CNT V f (∼1 vol % as-grown A-CNTs with ∼70 nm Γ values) and particularly as the CNT V f increases, such as for densified 30 vol % A-CNT arrays with ∼6 nm Γ values. ,,, In these cases, the interfacial surface area-to-volume ratio is large (i.e., ∼700× larger for a typical 10-nm-diameter CNT compared to a 7-μm-diameter carbon fiber, both at 60 vol % in a composite), the interphase region can extend 1–100 nm from the CNT surface depending on the polymer, , and Γ is small and on the order of the radius of gyration for polymers (3–30 nm), which therefore has the potential to alter polymer behavior such as polymer chain mobility, crystallinity, glass transition temperature ( T g ), and curing . For example, recent molecular dynamics simulations have predicted a ∼1-nm-thick interphase layer for epoxies surrounding CNTs, as corroborated by experiments, highlighting the utility of both computational and experimental studies to inform composite design via interfacial engineering . It is possible and even likely that the entire polymer matrix can be affected by CNT confinement if only a few nanometers of space exist for the polymer chains to reside between each CNT in a high- V f CNT PNC .…”

“…8,11,22,23 In these cases, the interfacial surface area-to-volume ratio is large (i.e., ∼700× larger for a typical 10-nm-diameter CNT compared to a 7-μm-diameter carbon fiber, both at 60 vol % in a composite), the interphase region can extend 1−100 nm from the CNT surface depending on the polymer, 18,24 and Γ is small and on the order of the radius of gyration for polymers (3−30 nm), which therefore has the potential to alter polymer behavior such as polymer chain mobility, crystallinity, glass transition temperature (T g ), and curing. 25 For example, recent molecular dynamics simulations have predicted a ∼1-nm-thick interphase layer for epoxies surrounding CNTs, 24 as corroborated by experiments, 26 highlighting the utility of both computational and experimental studies to inform composite design via interfacial engineering. 18 It is possible and even likely that the entire polymer matrix can be affected by CNT confinement if only a few nanometers of space exist for the polymer chains to reside between each CNT in a high-V f CNT PNC.…”