Graphene nanoplatelets (GNP), dispersed in epoxy resins, have been reported to show remarkable improvements in fracture toughness. This work seeks to investigate the nanoscale mechanisms near the crack tip in the presence of graphene nanoplatelets that promote the improvement in fracture toughness. The size of the fracture process zone formed in epoxy near a crack, typically on the order of tens of microns, is beyond the range of pure molecular dynamics (MD) modeling. Hence, MD simulations, concurrently coupled with finite element method (FEM) to bridge length and time scales, are used to simulate the crack initiation behavior under fracture loading. A novel concurrent multiscale coupling methodology is used, that partitions the full simulation domain into MD and FEM subdomains, and iteratively computes solutions for a given load step. The nonlinear behavior of the polymer in the MD domain is approximated in the FEM domain using a Ramberg–Osgood constitutive model, parameterized based on MD representative volume elements (RVE). K-dominant displacement fields, derived from linear elastic fracture mechanics (LEFM) are used as boundary conditions, away from the crack tip and surrounding the fracture process zone (FPZ). Stresses and strains in the MD domain (near-crack-tip region) are compared at a given level of far-field loading for baseline specimens and GNP-embedded specimens. The overarching goal is to be able to predict fracture toughness at crack initiation at room temperature in a brittle epoxy polymer. The effect of GNP size and position on crack tip shielding and hence on enhanced fracture properties will be investigated and the quantified using MD simulations.

Thin laminated composite structures under mechanical loading exhibit coupled interlaminar delamination and intralaminar transverse cracking, which govern progressive failure and residual capacity. Capturing the interaction between these damage modes typically requires separate numerical frameworks and fine through-thickness discretization, resulting in high computational cost. This work presents a computationally efficient, thermodynamically consistent layerwise decohesion model for simulating coupled delamination and transverse cracking in laminated composite beams. Fracture is treated as an intra-element phenomenon, with multiple potential crack directions at each material point. A rate-independent decohesion model governs crack-plane degradation through internal damage variables derived from the Helmholtz free energy, ensuring irreversible evolution and non-negative dissipation. Tensile-shear mixed-mode fracture is modeled using the Benzeggagh-Kenane criterion, and crack opening/closure is captured via smooth kinematic ramp functions. The model is validated using three-point bend experiments on carbon/epoxy laminates with varying pre-embedded delamination lengths. A single calibrated parameter set accurately predicts delamination-dominated, transverse-crack-dominated, and coupled responses across all configurations. Comparisons with Abaqus simulations using cohesive-zone and Hashin damage models show comparable accuracy while reducing computational cost by over two orders of magnitude. The framework enables rapid, robust analysis of laminated composites, providing a practical tool for structural assessment and design optimization.

Understanding and improving delamination resistance in carbon fiber-reinforced polymer composites is crucial for advancing their applications in structural and multifunctional systems. While the influence of graphene nanoplatelet (GNP) weight fraction on composite performance has been widely reported, the role of GNP particle size remains largely unexplored, which poses a challenge to predict delamination initiation behaviour across different length scales accurately. This study investigates the effects of GNP size and weight fraction on the Mode I delamination fracture energy of carbon/epoxy composites at room temperature. Building on previous work, an analytical approach based on experimental data is developed to quantify the influence of nanoparticle size on delamination initiation and propagation. Double cantilever beam (DCB) specimens are used to evaluate fracture energy as a function of dispersed GNP size as well as weight fraction. In addition to matrix reinforcement, the study introduces a coating of the carbon-fiber fabric with GNPs prior to resin infusion to assess synergistic effects on toughness. A novel predictive model for delamination fracture energy based on crack-tip shielding is proposed and validated using experimental data. Subsequently, it is applied to elucidate why composite delamination toughness is generally lower than the fracture toughness in epoxy for the same GNP loading. Scanning electron microscopy (SEM) is used to characterize toughening mechanisms such as crack-tip shielding and microcrack development. The findings reveal that both nanoparticle size and its weight fraction significantly influence delamination resistance, offering practical insights for the tailoring of tougher, more reliable composite structures.

Comprehending the influence of reactive atomic species on the nanoscale mechanical response of Carbon-Carbon (C-C) nanocomposites remains a fundamental challenge in predicting coupled mechanical-chemical response in materials exposed to extreme environments. Specifically, leading-edge/nose-cone of hypersonic vehicles are exposed to extreme environments where shock-induced loading, particulate impact, and high-temperature oxidation occur simultaneously. Therefore, a detailed study of coupled synergistic thermo-mechanical and chemical processes occurring in thermal protection materials (TPM) is of great importance and poses significant challenges due to its complexity. In this work, reactive molecular dynamics (RMD) simulations are employed to study nano dust nanoparticle (~10 nm) impact at hypersonic velocity on a C-C composite surface, after traversing through a leading-edge shock layer. Unlike rain droplets, solid dust nanoparticles maintain their structural integrity while traversing through the shock layer, leading to concentrated momentum transfer and significantly deeper craters. The crater depth increases non-linearly with impact velocity ranging from Mach 6 to Mach 12, in accordance with hypervelocity impact theory. In addition, atomic oxygen (AO) interactions with the C-C composites’ surface are also examined, revealing oxidation-induced surface morphological reconstruction primarily characterized by CO production, resulting in C-C composite surfaces’ structural reordering. Interestingly, subsequent impacts from dust nanoparticle on the AO reacted C-C surface result in decreased crater depth due to AO induced morphological changes. A non-dimensional scaling law is established to correlate RMD predictions with experimental data, exhibiting consistent power-law behavior across different length scales.

It is now well documented in the literature that the inclusion of nanoparticles, such as graphene nanoplatelets (GNP), in matrix materials, such as epoxy, has resulted in significantly improved fracture toughness in mode I and mixed-mode. One of the mechanisms postulated to increase the effective crack initiation fracture toughness is the crack tip shielding effect due to nanoparticles in the fracture process zone. This effect is deemed to arise due to debonding of nanoparticles from the matrix material in the process zone, which in turn reduces the stress state at the tip of the primary crack via shielding. Thus, nanoparticles act to redistribute stress in the crack tip region, thereby lowering the near tip stress intensity factor, depending on their orientation relative to the crack. Therefore, higher far-field loads can be achieved before the critical stress intensity is reached at the crack tip. In this paper the K-test approach is used in conjunction with molecular dynamics (MD) to model fracture in an amorphous carbon matrix material, with embedded GNPs. Amorphous carbon matrix is deliberately selected to facilitate the computational efficiency of the solution process, because the fracture process zone size for amorphous carbon is relatively small from a MD simulation viewpoint. The effect of GNPs on the shielding of the crack tip, with varying orientation and location relative to the crack is investigated using detailed virial stress plots, the atomistic J-integral, and compared with linear elastic fracture mechanics (LEFM) results.