Network Mechanics
Decoding the Mechanics of Molecular Networks
The elasticity of a rubber band has nothing to do with the stiffness of its chemical bonds. Rather, it emerges from a restoring force born from thermal fluctuations and the statistics of long, disordered chains. This phenomenon is known as entropic elasticity. For most of the nineteenth century, rubber was treated as an anomaly. However, the realization that molecular disorder and random fluctuation could be the source of rich and diverse mechanical behavior, such as elasticity, viscoelasticity and fracture remains one of the more elegant results in soft condensed matter physics.
The properties of soft matter depend not on the specific chemistry of individual bonds, but on complex mesoscale interactions between larger structural units under constant thermal noise.
Recent advances in synthetic chemistry now allow the synthesis of polymer networks with increasingly precise architectures, yet the ability to predict how that architecture governs macroscopic behavior has lagged behind. Closing this gap requires the full machinery of statistical mechanics, and new modeling frameworks that sit squarely between the molecular and the continuum scale. Our research is therefore articulated around two main objectives:
- The introduction of a network-level (or mesoscale) computational approach to study the collective mechanics of complex molecular networks.
- The development of statistical mechanics approaches (namely the transient network theory) to relate network level mechanisms to physics-based constitutive relations (elasticity, viscoelasticity, damage, fracture) at the continuum scale.
Mesoscale Modeling: Molecular Networks as Dynamic Graphs
A polymer network is, at its core, a graph made of a collection of nodes connected by edges. This is not merely a metaphor: it is the natural mathematical structure for describing soft matter at the mesoscale, sitting between the atomic detail of molecular dynamics and the averaged descriptions of continuum mechanics.
In this framework, each edge represents a single chain segment, carrying the full entropic elasticity of a polymer strand under thermal agitation. The nodes represent junctions and junctions come in many flavors: a covalent crosslink is permanent, a topological entanglement constrains but can slide, while a dynamic bond can break under force and reform elsewhere. By assigning different mechanical and kinetic rules to different node types, the same framework describes a diversity of physics, from tight elastomers to self-healing hydrogels. Thermal fluctuations enter naturally as stochastic events: bonds break and reform with rates governed by thermal activation, and the network is never truly at rest.
This scale unlocks phenomena invisible to molecular dynamics, which is limited to nanometer volumes and microsecond timescales, while retaining molecular resolution. A single simulation volume can represent tens of microns of material, giving access to slow processes such as reptation, bond-driven network reorganization, void nucleation, and crack propagation. Because every chain and junction is individually tracked, one can extract full statistical distributions of local forces and deformations. This matters because phenomena like fracture are rare-event effects: it is the tail of the force distribution, not its mean, that determines where a network fails.
Finally, the graph structure provides a principled bridge to continuum models. By averaging the statistical behavior of individual chains and junctions, one can derive constitutive laws that carry genuine molecular information, connecting molecular architecture to macroscopic properties from the bottom up.
Open Questions & Current Directions
Despite a century of progress in soft matter physics, a complete, predictive theory of polymer network mechanics remains out of reach. Below are the central questions driving our research.
Network Architecture
How does the spatial distribution of crosslinks and entanglements, set at synthesis, determine the full mechanical response, including toughness and fatigue life? Can we design heterogeneity to be beneficial?
Far-from-Equilibrium Dynamics
At large deformations, chains explore non-Gaussian regimes; the tube model breaks down. What replaces it? How do entanglements disengage under large, fast deformations, and what is the role of thermal fluctuations in this process?
Fracture Nucleation
Macroscopic fracture begins at the molecular scale, a single chain breaking, a cluster of entanglements failing. How do these rare, local events organize into a propagating crack? What statistical mechanics governs the nucleation?