🔬 condensed matter

Investigating the Nature of Discontinuous Shear Thickening: Beyond a Mean-Field Description

This paper employs higher-order cluster analysis on force networks from dense suspension simulations to reveal significant changes in an effective interaction potential near the discontinuous shear thickening transition, thereby challenging mean-field descriptions and supporting the development of an emergent field theory for this phenomenon.

Original authors: Jetin E. Thomas, Abhay Goyal, Deshpreet Singh Bedi, Abhinendra Singh, Emanuela Del Gado, Bulbul Chakraborty

Published 2026-09-15

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Jetin E. Thomas, Abhay Goyal, Deshpreet Singh Bedi, Abhinendra Singh, Emanuela Del Gado, Bulbul Chakraborty

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a thick, gritty mixture of tiny solid grains suspended in a liquid, like cornstarch and water or wet sand. Under normal conditions, this mixture flows easily. But if you push it hard enough or stir it fast enough, something strange happens: it suddenly becomes as stiff as a solid, resisting your motion with a dramatic spike in thickness. Scientists call this phenomenon "discontinuous shear thickening." For years, the leading explanation has been a relatively simple idea: as the mixture is squeezed, the tiny grains are forced closer together. At a critical point, the smooth, slippery layer of liquid between them breaks down, and the grains begin to rub against each other with friction. This shift from slippery to sticky contact was thought to be the sole cause of the sudden stiffening, a view that could be described by averaging the behavior of all the grains together.

However, this simple picture leaves a gap. While the average behavior of the mixture is well understood, the microscopic details of what happens right at the moment of stiffening remain a mystery. Standard ways of looking at the mixture, such as tracking the exact positions of the grains, show almost no change when the thickening begins. The grains do not rearrange into a new pattern that a camera could easily spot. Yet, experiments suggest that something significant is happening deep within the network of forces holding the mixture together. The question for researchers has been: if the positions of the grains look the same, what is actually changing to make the mixture lock up so violently?

A team of physicists has now taken a fresh look at this problem by shifting their focus from where the grains are to how they push and pull on one another. Instead of mapping the locations of the particles, they mapped the invisible web of forces connecting them. They ran detailed computer simulations of a dense suspension of two thousand particles, subjecting them to controlled stress. By analyzing the data, they discovered that while the physical arrangement of the grains remained largely unchanged, the structure of the force network underwent a dramatic transformation. As the mixture approached the point of sudden stiffening, the forces between the grains began to organize into distinct, large-scale clusters. These clusters were not visible in the positions of the grains but were clearly evident in the pattern of the forces.

To make sense of these complex force patterns, the researchers developed a new way of simplifying the data. They treated the points where forces meet as if they were particles in their own right, interacting with each other. By filtering out the small, random jitters caused by the fluid and focusing on the larger, more significant contacts, they constructed an "effective potential." This is a mathematical description of how these force-points attract or repel one another. They found that the shape of this potential changes depending on whether the mixture is in a state of gradual thickening or sudden, discontinuous stiffening. In the state where the mixture stiffens abruptly, the forces between the points arrange themselves in a way that suggests the system is much more prone to large, unpredictable fluctuations.

The researchers then used this new description to run their own simulations, treating the force-points as a system of interacting particles. They found that the different shapes of the force potential led to qualitatively different behaviors. In the regime where the mixture thickens suddenly, the system showed signs of long-lived, unstable states that are very different from the steady flow seen in other conditions. This suggests that the sudden stiffening is not just a result of more grains rubbing together, but a fundamental shift in how the entire network of forces organizes itself. The study indicates that the transition involves a complex reorganization of stress that creates large, intermittent fluctuations, which standard theories that only look at averages cannot capture.

This work does not overturn the idea that friction between grains is important, but it reveals that the story is more complex than previously thought. The researchers suggest that to fully understand why these mixtures lock up, one must look beyond the simple average of frictional contacts and consider the collective, fluctuating nature of the force network. Their findings point toward a new way of thinking about these materials, one that treats the network of forces as a dynamic system capable of its own phase transitions. While the study relies on computer simulations and does not yet provide a complete, final theory, it offers a compelling new perspective. It suggests that the sudden stiffening of a dense suspension is driven by a hidden, large-scale reorganization of forces that creates a fragile, fluctuating state, waiting for the slightest push to snap into a solid.

Technical Summary: Investigating the Nature of Discontinuous Shear Thickening: Beyond a Mean-Field Description

Problem Statement
Dense suspensions of grains in a fluid exhibit Discontinuous Shear Thickening (DST), a phenomenon where viscosity increases abruptly at a critical shear stress. While a successful mean-field theory exists, attributing DST to an increase in the fraction of frictional particle contacts, this framework fails to capture the full microscopic physics of the transition. Conventional structural analysis, such as the pair correlation function of grain positions, shows no significant changes at the onset of DST. However, experimental evidence suggests the existence of higher-order microscopic correlations and large, intermittent stress fluctuations that a mean-field description cannot account for. The central problem addressed is the lack of a structural or statistical measure that captures the microscopic correlations and fluctuations driving the DST transition, particularly those that are elusive in real-space positional analysis.

Methodology
The authors employ a multi-stage approach combining numerical simulations, force-space representation, and statistical mechanics:

Microscopic Simulations: The study utilizes two-dimensional simulations of non-Brownian, frictional spherical particles in a Newtonian fluid under constant shear stress ( $\sigma_{xy}$ ). The model includes hydrodynamic lubrication, short-range repulsive forces, and frictional contact forces (Coulomb friction). The simulations track the time-dependent shear rate ( $\dot{\gamma}$ ) and particle positions.
Force-Space Representation: To analyze correlations beyond positional data, the authors map the suspension into "force space" using Maxwell-Cremona diagrams (force tilings). In this dual space, pairwise forces between grains correspond to edges, and the vertices represent height vectors ( $h$ ) derived from force balance constraints. The distance between vertices quantifies internal stresses.
Cluster Analysis: The authors apply the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm to both real-space grain positions and force-space vertices. This is used to identify characteristic clustering scales and distinguish between small-scale hydrodynamic correlations and larger-scale contact force correlations.
Coarse-Graining and Effective Potential: To overcome the strong clustering at small scales (dominated by hydrodynamic drag) which hinders analysis, a coarse-graining procedure is applied to the force-space vertices. This filters out small-scale noise to construct an effective interaction potential, $V^*(h) = -\ln[g^*(h)]$ , where $g^*(h)$ is the pair correlation function of the coarse-grained height vertices.
Molecular Dynamics (MD) Analogy: The constructed effective potentials are used as inputs for NVT molecular dynamics simulations of fictitious thermal systems. These simulations explore the accessible microstates and phase behavior (clustering vs. non-clustered fluid) to understand how changes in the potential shape affect the system's state.

Key Results

Fluctuations in DST: Analysis of strain rate distributions reveals that while Continuous Shear Thickening (CST) regimes exhibit narrow, Gaussian-like fluctuations, the DST regime (at high packing fractions, e.g., $\phi=0.80$ ) displays significant non-Gaussian, intermittent fluctuations. These fluctuations are not captured by mean-field theories.
Scale Separation in Force Space: DBSCAN analysis reveals a distinct scale separation in force-space vertices that is absent in real-space grain positions. In the DST regime, a "meta-clustering" emerges where small-scale clusters organize into larger structures. This meta-structure is sensitive to stress and packing fraction, unlike the grain positions which remain virtually unchanged across the transition.
Effective Potential Changes: The coarse-grained effective potential $V^*(h)$ $V^{*} (h)$ exhibits qualitative differences between CST and DST regimes.
- For CST ( $\phi=0.76$ ), the potential features a short-range attractive well and a repulsive shoulder around $h \approx 2$ .
- For DST ( $\phi=0.80$ ), the repulsive shoulder is suppressed, and the potential shape changes with increasing stress.
Phase Behavior Implications: MD simulations using these potentials show that the suppression of the repulsive shoulder in the DST regime alters the accessible microstates. The analysis of Fourier transforms of the potentials indicates a flattening of the potential at low wavevectors ( $q_h$ ) in the DST regime, suggesting the potential for long-ranged fluctuations and long-lived transients that differ from the steady state.

Significance and Claims
The paper claims to provide a pathway toward a statistical mechanics framework for shear thickening that goes beyond mean-field descriptions by explicitly incorporating microscopic fluctuations. The primary significance lies in demonstrating that:

Force Space is the Correct Order Parameter: While real-space positional correlations fail to detect the DST transition, correlations in force space (specifically the clustering of height vertices) capture the essential physics of the transition.
Potential Shape Drives Transition: The transition from CST to DST is associated with a qualitative change in the effective interaction potential in force space. This change in potential shape, rather than just a change in the fraction of frictional contacts, appears to drive the emergence of different flow states and fluctuation behaviors.
Beyond Mean-Field: By constructing an effective potential that captures these changes, the authors suggest that the DST transition can be viewed as a change in the underlying "state diagram" of the force space, analogous to phase transitions in equilibrium statistical mechanics. This offers a new perspective for developing field theories (e.g., Ginzburg-Landau functionals) for non-equilibrium shear thickening, where the density in height space serves as the order parameter.

The authors remain modest, noting that a direct mapping between the "temperature" in their MD simulations and the physical fluctuations (shear rate noise) in the suspension requires further investigation. However, the work establishes that the coarse-grained potential in force space contains critical information about the physics of different flow states and the nature of fluctuations near the DST transition.

More like this