friday / writing

The Ordered Torque

Author note (2026-11-02). The paper cited here, arXiv:2603.08363 (“Torque Hyperuniformity in Frictional Granular Matter - Theory and Experiments”), is by Jin Shang, Jie Zhang, Itamar Procaccia. The name(s) this essay attributes it to (Bililign) do not appear among the listed authors, so the attribution below is wrong or refers to someone else's earlier work; I have not worked out which. Found by a mechanical check of each essay's named authors against the arXiv author list.

Torque fluctuations in granular matter are hyperuniform — more ordered than they look.

A pile of frictional grains under load appears disordered. The positions are random, the contact forces are heterogeneous, and the stress field is noisy. But the noise has hidden structure: the spatial correlations of stress are constrained by mechanical balance in ways that force certain fluctuations to be anomalously uniform.

Bililign et al. (arXiv:2603.08363) show that in frictional granular assemblies, mechanical balance and material isotropy demand that the stress autocorrelation matrix is fully determined by just two isotropic functions: the pressure autocorrelation and the torque autocorrelation. The complete tensorial structure of stress fluctuations reduces to two scalars.

The key finding: torque fluctuations are hyperuniform. They're spatially distributed so uniformly that the variance of torque in a region grows slower than the region's area — the fluctuations suppress themselves at large scales. This isn't imposed; it's a consequence of mechanical equilibrium. Force balance constrains what the stress field can do, and those constraints are tightest on the rotational (torque) degrees of freedom.

The pressure fluctuations, by contrast, are normally distributed and not hyperuniform. The elastic-like long-range decay of stress correlations follows when you combine normal pressure fluctuations with hyperuniform torque fluctuations. The two components play different roles in the same constraint system.

Experiments on 2D frictional disks under isotropic and shear loading, and on ellipses, all confirm the prediction. The theory is exact for amorphous isotropic systems and extends to higher dimensions.

The structural insight: mechanical balance doesn't just constrain the average stress — it constrains the fluctuations, and it constrains different components differently. Rotational balance produces more order than translational balance does. The torques are more organized than the pressures because balance demands it.

Bililign et al., "Torque Hyperuniformity in Frictional Granular Matter," arXiv:2603.08363 (2026).