A New Multiscale Modeling Approach to Unravel the Influence of Interlayer sp 3 Bonds on the Nonlinear Large‐Deformation and Fracture Behaviors of 2D Carbon Nanostructures Under Tension
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | International Journal for Numerical Methods in Engineering |
| Published | 2026-10-25 |
| Authors | Xiangyang Wang, Biao Xu, Huibo Qi, Shichao Dai, Jiqiang Li |
| DOI | 10.1002/nme.70455 |
AA-stacked diamane reached a 25.8% zigzag fracture strain in the authors’ multiscale auxiliary nodes simulations, with an initial zigzag Young’s modulus of 1285.11 GPa. Uniformly distributed four rows of interlayer sp³ bonds produced a 22.9% fracture strain, above the 21.8% of bilayer graphene. The authors conclude that bond count alone does not predict tensile performance. Distribution must reduce boundary warping and necking to avoid lower fracture strain. These findings come from quasi-static numerical simulations, not experimental tests.
Model formulation
Section titled “Model formulation”The multiscale auxiliary nodes method, or MAN, represents bilayer graphene, diamane, and intermediate structures with two virtual continuum sheets. Nodes can be arranged freely on the sheets, while each carbon atom has an influence domain. Moving least squares interpolation maps atomic displacements to nodal displacements.
An sp³ hybridized unit cell serves as the common atomic template. The authors remove selected C-C bonds to represent sp² regions. Coefficients of 0 or 1 control included bonds and bond angles.
The second-generation reactive empirical bond order potential supplies atomic potential energy. Its bond-energy terms include bond length, bond angle, local coordination, conjugation, and torsional effects. MAN maps this energy into continuum strain energy through displacement interpolation.
Equilibrium follows by minimizing total potential energy with respect to nodal displacements. A Taylor expansion converts the nonlinear equation into a global stiffness equation. Newton-Raphson iterations solve that equation, and about a dozen iterations reached minimum energy in a load step. The authors classify the resulting computation as order N.
Numerical setup and verification
Section titled “Numerical setup and verification”The modeled sheets were 28.9 Å long along armchair and 27.5 Å wide along zigzag. The authors used AA-stacked diamane because cited work found AA and AB stacks nearly identical in stiffness and fracture response.
All simulations were quasi-static. Before loading, the authors relaxed each model to its undeformed equilibrium configuration. They then translated the rigid ends and minimized potential energy after every load step.
The comparison case used fully atomistic finite element simulations with the same potential and loading conditions. A two-layer 23 by 14 node arrangement matched the carbon-atom count and reproduced the atomistic response. The paper reports agreement when node count and distribution are comparable to or greater than the atom count.
The influence-domain scaling parameter ranged from 2.0 to 4.0 for bilayer graphene but only 2.0 to 3.0 for diamane. Larger domains slowed convergence and accumulated numerical error. The study therefore selected values between 2.0 and 3.0.
Reducing nodes smooths local deformation in graphene and falsely raises fracture strain. In diamane, fewer nodes distort the predicted configuration and lower fracture strain. These effects make node density a numerical choice, not an immaterial meshing detail.
Elastic response of diamane
Section titled “Elastic response of diamane”The elastic constants used a diamane thickness of 4.6 Å. The initial elastic-stage values were:
| Loading direction | Young’s modulus | Shear modulus | Poisson’s ratio |
|---|---|---|---|
| Zigzag | 1285.11 GPa | 608.48 GPa | 0.056 |
| Armchair | 1271.52 GPa | 598.08 GPa | 0.063 |
Young’s modulus and shear modulus decline as loading increases, showing nonlinear material response.
Predicted diamane fracture strain was approximately 25.8% along zigzag and 14.8% along armchair. The authors describe the directional difference as roughly 50%. The zigzag value agrees with Wu et al.
The calculated stiffness values agree with cited literature and remain close to reported diamond values. Earlier work also found nearly identical mechanics for AA and AB diamane stacks.
Interlayer bond placement under zigzag tension
Section titled “Interlayer bond placement under zigzag tension”The following results apply to zigzag uniaxial tension on the 28.9 Å by 27.5 Å sheets.
| Model | Interlayer sp³ arrangement | Fracture strain |
|---|---|---|
| Bilayer graphene | None | 21.8% |
| Mod-P 1 to 6 | Six patterned arrangements | 20.3%, 20.2%, 20.2%, 20.2%, 20.1%, 20.1% |
| Mod-L 1 to 3 | Linear bonds parallel to loading | 20.1%, 20.2%, 20.1% |
| Mod-L 4 to 6 | Linear bonds perpendicular to loading | 17%, 17.1%, 17% |
| Four end-arranged rows | Rows concentrated near each boundary | 20% |
| Four dispersed rows | Rows evenly distributed | 22.9% |
| AA-stacked diamane | Full interlayer bonding | 25.8% |
For Mod-P 1 to 6, cracks began at corner points and propagated through sp³ bonds. The authors attribute their lower fracture strain to edge warping caused by the inserted bonds.
Mod-L 1 to 3 placed bonds parallel to loading and gave nearly identical fracture strains. Mod-L 4 to 6 placed them perpendicular and lowered fracture strain to about 17%. The authors propose that boundary bonds and the zigzag edge formed notches that favored tear cracks.
Two full bond rows reduced strength below the bond-free sheet. As row count increased, edge warping weakened and strength approached AA-stacked diamane. The authors attribute the early reduction to combined edge warping and necking.
Four rows at the ends suppressed warping but reached only 20% fracture strain. Uniformly dispersed rows reached 22.9%. The authors suggest that adjacent sp³ bonds reinforced boundary sp² bonds and reduced nonuniform deformation.
Crack transfer and armchair tension
Section titled “Crack transfer and armchair tension”Armchair tension produced 14.1% fracture strain for bilayer graphene and 14.7% for diamane. The authors report little benefit from adding interlayer bonds in this direction.
Under zigzag loading, the crack starts in the lower layer at a corner and travels along zigzag. At an sp³ bond, it transfers to the upper layer, which ruptures from the middle. The two layers then fracture independently until failure.
Armchair loading produced a similar transfer sequence. The crack still followed the zigzag direction and began inside, near the boundary. The authors relate this origin to necking that moves and rotates selected boundary bonds.
Limits and engineering implications
Section titled “Limits and engineering implications”This study provides simulation evidence rather than experimental validation. Its conclusions apply to the modeled sheets, one interatomic potential, and the imposed quasi-static boundaries.
The model neglects interlayer van der Waals interactions. It also restricts the REBO cutoff to initial configurations because the 1.7 to 2.0 Å transition can create nonphysical bond forces.
Large deformation can make the stiffness matrix lose positive definiteness and prevent convergence. The authors address this with stiffness matrix modification.
Node reduction can bias fracture strain, especially for diamane. The text also reports 14.8% armchair fracture strain during validation and 14.7% in the bond-pattern comparison.
For a 2D carbon design needing stiffness and zigzag tensile strain, the simulations favor AA-stacked diamane. For patterned bilayer graphene, engineers should treat bond distribution and edge geometry as coupled design variables. The paper does not establish selection criteria for bulk diamond products.
View Original Abstract
ABSTRACT To delve deeply into the nonlinear large‐deformation and fracture behaviors of 2D carbon nanostructures (2D CNs), including bilayer graphene, diamane, and their transitional structures, this paper introduces a multiscale auxiliary nodes (MAN) method rooted in atomic structures and potentials. Different from traditional atomistic and multiscale solvers, the proposed framework builds a pair of decoupled, C 2 ‐continuous virtual continuum sheets separated from the physical atomic lattice as independent solution domains. Atomic geometric information, interatomic distances, and REBO bond‐order potentials are mapped onto the dual sheets via moving least squares (MLS) shape functions, reformulating the full atomic potential functional purely in terms of auxiliary nodal displacements rather than atomic coordinates. The derived equilibrium equations follow order‐N linear complexity with respect to auxiliary nodes, whose positions are unrestricted by lattice periodicity; flexible adjustment of node density enables smooth atom‐to‐continuum transitions without artificial interfacial forces. Through iterative solutions of nonlinear stiffness equations, the equilibrium configuration of the system under specified loading conditions can be obtained. Numerical simulations suggest that the MAN method can reasonably capture the nonlinear large‐deformation responses and fracture behaviors of 2D CNs. The Young’s modulus and shear modulus of diamane in both zigzag and armchair directions closely approach those of diamond and are notably higher than those of graphene. Furthermore, the quantity and distribution of interlayer sp 3 bonds significantly influence the fracture behavior of 2D CNs, with uniform dispersion of these bonds effectively mitigating strain localization and improving structural tensile strength.
Products
View Specs →
Advanced Thermal Simulation Service Diamond Heat Spreader Performance Analysis
View Specs →