From Connectivity to Rupture: A Coarse-Grained Stochastic Network Dynamics Approach to Polymer Network Mechanics

We introduce a coarse-grained stochastic network dynamics (CGSND) framework for modeling deformation and rupture in polymer networks. The method replaces explicit molecular dynamics (MD) or coarse-grained molecular dynamics (CGMD) with network-level evolution rules while retaining chain entropic elasticity and force-controlled bond failure. Under uniaxial loading, CGSND reproduces the characteristic nonlinear stress–stretch response of elastomeric networks, including a well-defined ultimate tensile strength and post-peak softening due to progressive bond rupture. Comparison with coarse-grained molecular dynamics (CGMD) simulations shows that CGSND captures the qualitative form of the stress response and the onset of catastrophic damage despite its rate-independent formulation. Analysis of rupture kinetics reveals a pronounced peak in the bond-breaking hazard rate near the ultimate tensile strength in both approaches. In addition, the distribution of broken segment lengths remains statistically indistinguishable from the initial network, indicating that rupture is not biased toward short or long chains. Finally, the evolution of the Gini coefficient of bond force magnitudes reveals strong force localization preceding failure. These results demonstrate that CGSND provides a computationally efficient and physically interpretable framework for connecting force localization and rupture kinetics to macroscopic failure in polymer networks.

💡 Research Summary

The paper introduces a coarse‑grained stochastic network dynamics (CGSND) framework that replaces explicit molecular dynamics (MD) or coarse‑grained MD (CGMD) with a graph‑based, network‑level description of polymer elastomers while preserving essential physics such as entropic elasticity and force‑controlled bond failure. The authors first map a cross‑linked polymer network onto an undirected weighted graph: nodes represent cross‑link junctions, edges represent polymer strands, and the integer weight of an edge equals the number of covalent bonds in that strand (weight = 1 for a direct cross‑link). During uniaxial loading, node positions are affinely stretched, and each edge’s end‑to‑end vector D and normalized extension p = |D|/L (with L = α w) are updated at every strain increment.

The mechanical response of each strand is modeled by an inverse‑Langevin constitutive law, approximated by a polynomial expression that reproduces linear Hookean behavior at small p and strong strain‑stiffening as p → 1. A nondimensional force vector f* = L⁻¹(p) D/|D| is computed for every intact edge. Bond rupture is deterministic: any edge whose force magnitude exceeds a prescribed cutoff fcut (≈ 1431.65, corresponding to ~10 GPa) is irreversibly removed, thereby degrading the network. Macroscopic stress is obtained from a bulk virial expression σ = (1/V*) ∑ D⊗f* and converted to physical units (MPa) using the Kuhn length b = 8.6 × 10⁻¹⁰ m.

Two additional observables are introduced to characterize damage evolution. First, the rupture hazard rate h(λ) = (1/N₀) dNb/dλ quantifies the instantaneous probability per unit stretch that a bond fails, where N₀ is the initial total bond count and Nb(λ) the cumulative number of broken bonds. Hazard rates are computed separately for cross‑links (w = 1) and backbone strands (w > 1). Second, the Gini coefficient G = (1/2Nb² \bar f) ∑|f_i − f_j| measures the heterogeneity of force magnitudes across the surviving network; G → 0 indicates uniform load sharing, while G → 1 signals extreme localization.

The authors apply CGSND to a representative elastomeric network and compare its predictions with full CGMD simulations of the same system. Both methods reproduce the characteristic nonlinear stress–stretch curve: an initial entropic regime, a strain‑stiffening region, a well‑defined peak stress (ultimate tensile strength), and post‑peak softening associated with progressive bond rupture. CGSND captures the overall shape and peak location but shows higher noise and slight differences in peak magnitude because it lacks explicit inertial, thermostat, and rate‑dependent effects present in CGMD.

Rupture kinetics differ markedly. In CGMD, the fraction of broken bonds grows approximately linearly with stretch for both backbone strands and cross‑links, reflecting thermally activated, stochastic breaking across the entire bond population. In contrast, CGSND exhibits a strongly nonlinear evolution: an early rapid rise dominated by cross‑link failure, followed by a crossover to backbone‑dominated rupture at higher stretches. This behavior stems from the deterministic force‑threshold rule, which preferentially removes the most highly stressed edges, leading to a rapid exhaustion of the “weakest” load‑bearing paths and a subsequent saturation as the remaining network bears load more diffusely.

Both approaches display a pronounced peak in the hazard rate near the ultimate tensile strength, confirming that failure onset is governed by force concentration rather than global strain. The Gini coefficient computed from CGSND rises sharply with deformation, reaching values close to one just before macroscopic failure, thereby quantitatively demonstrating that a small subset of bonds carries the majority of the load in the critical regime.

The study concludes that CGSND provides a computationally inexpensive yet physically interpretable framework for linking microscale force localization and rupture kinetics to macroscopic failure in polymer networks. Its main advantages are orders‑of‑magnitude speedup over CGMD and the ability to extract network‑level metrics (hazard rate, Gini coefficient) that are difficult to define in particle‑based simulations. Limitations include the omission of rate‑dependent, inertial, and thermal relaxation effects, and the use of a deterministic rupture criterion. The authors suggest future extensions such as incorporating stochastic failure thresholds, coupling to explicit relaxation steps, or hybrid multiscale schemes that combine CGMD accuracy with CGSND efficiency.