Influence of Stretching Boundary Conditions on Fracture in Phantom Star Polymer Networks: From Volume to Cross-sectional Area Conservation

This study systematically investigates the effect of stretching boundary conditions, ranging from conservation of cross-sectional area to conservation of volume, on the rupture behavior of phantom star polymer networks using energy-minimizing coarse-grained molecular simulations. By continuously varying the deformation parameter, the simulations reveal that true stress and rupture characteristics, such as strain and stress at break and work for rupture, systematically decrease as the boundary condition approaches cross-sectional area conservation. In contrast, nominal stress and the corresponding rupture characteristics exhibit near-independence from boundary conditions, indicating that bond tension remains largely unaffected for phantom networks under the examined conditions. These results clarify that volume expansion primarily drives deviations in true stress and highlight a critical distinction between true and nominal stress-strain definitions. The difference between true and nominal stress-strain relations also affected the scaling exponent for strand length dependence on stretch at break. The findings stress the importance of specifying both deformation boundary conditions and stress-strain definitions in polymer network simulations for accurate interpretation of mechanical properties.

💡 Research Summary

This paper presents a systematic investigation of how stretching boundary conditions influence the fracture behavior of phantom star polymer networks, using energy‑minimizing coarse‑grained molecular dynamics simulations. The authors introduce a continuous deformation parameter ! that interpolates between two extreme protocols: cross‑sectional area conservation ( ! = 0) and volume conservation ( ! = 0.5). Intermediate values represent mixed conditions, effectively varying the Poisson‑like lateral response during uniaxial tension.

The model consists of monodisperse star polymers (arm number ‘ = 3–8, equal arm length L) placed in a periodic cubic box at fixed bead density. Inter‑bead interactions are limited to bonded, non‑linear springs (equilibrium length ℓ₀, non‑linearity γ); excluded‑volume and attractive forces are omitted to maintain a phantom network where individual bond tension is not screened by entanglements or steric effects. After thermal equilibration, a Brownian‑dynamics driven end‑linking reaction creates a gel, suppressing intra‑molecular loops. The resulting network (≈35 000 beads) is then energy‑minimized (BFGS) and subjected to incremental affine elongation. At each step, any bond exceeding a critical length ℓ_c is removed, and the system is re‑minimized, allowing the identification of the rupture point.

Two stress–strain definitions are employed: (i) true (Hencky) stress σₜ and true strain εₜ, which reference the instantaneous volume, and (ii) nominal stress σₙ and nominal strain εₙ, which reference the initial volume and length. This dual approach enables a direct comparison of how volumetric changes affect measured mechanical quantities.

Key findings:

1. True stress–strain curves show a systematic decrease in both the stress at break (σ₀) and the strain at break (ε₀) as ! decreases (i.e., as the protocol moves toward cross‑sectional area conservation). The volume of the simulation box expands markedly under low ! conditions, increasing the current cross‑sectional area and thereby diluting the true stress for a given bond tension. Consequently, the work required for rupture (W₀) also diminishes.

2. Nominal stress–strain curves are essentially invariant with respect to !. The peak stress, break strain, and rupture work extracted from σₙ vs εₙ collapse onto a single master curve for all ! values. This indicates that, for the phantom network, the actual bond tension is insensitive to the macroscopic deformation protocol; only the geometric normalization (current vs. initial area) changes the apparent stress.

3. Scaling with network topology is examined using the cycle‑rank density * (number of independent loops per branch point). When σ₀ and W₀ are normalized by the branch‑point density Y₀, master curves emerge for both true and nominal definitions. For true quantities, the apparent power‑law exponent slightly increases as ! decreases, reflecting the influence of volume expansion on strand deformation. Nominal quantities display a constant exponent, suggesting that under large stretch the strands align predominantly along the loading direction regardless of lateral constraints.

4. Physical interpretation: The results clarify that in phantom networks the deformation protocol does not alter the microscopic bond forces; instead, volumetric expansion governs the observed differences between true and nominal stresses. True stress reflects work per unit current volume, which is reduced when the system swells, whereas nominal stress reflects work per unit reference volume, remaining constant across protocols.

The authors conclude that (a) specifying both the stretching boundary condition and the stress‑strain definition is essential for reproducible reporting of fracture data, (b) nominal stress–strain analysis may be more appropriate when comparing to experiments that measure force per initial cross‑section, and (c) the introduced ! parameter provides a flexible framework to bridge experimental and computational protocols. The study highlights a subtle yet crucial distinction that can lead to misinterpretation of mechanical properties if overlooked, and it sets the stage for future work on non‑phantom (entangled, excluded‑volume) networks where the coupling between bond tension and macroscopic deformation may be more complex.