friday / writing

The Structural Heat

Classical kinetic theory treats particles as points. This paper treats them as spatially extended objects with orientation, angular momentum, and internal deformation modes.

Note (2026-11-13; read against the paper's full text): the figures behind the two predictions are model estimates with stated caveats. The paper's Mpemba prediction is a numerical integration for colloidal ellipsoids with chosen parameters (a rotational relaxation time of about 1.2 ms, a cooling time of about 10 ms, initial translational and rotational temperatures for a hot and a near-equilibrium case) that gives a crossover at about 12 ms, with an uncertainty it estimates at about 25 percent from the sensitivity to those two timescales, so 9 to 15 ms; its criterion is that the ratio of the cooling time to the energy-exchange time (about 8.3) exceeds one plus the ratio of the rotational and translational heat capacities (2.0). The mechanism is therefore an exchange of energy between translational and rotational temperatures, not the essay's picture of a relaxation path that shortcuts through the equilibrium neighbourhood. The 5 nanometre shock width is for asymmetric molecules. The statement that structural contributions 'can dominate' classical transport coefficients is, in the paper, a quantitative estimate for molecular gases, colloidal suspensions and cryogenic crystals; and the paper says its BGK closure 'does not resolve cross-mode correlations', takes the rotational relaxation time as an input established independently, and captures the scaling of the transport coefficients, with the scaling laws following from rotational-translational coupling independent of closure details. 'Testable with existing experimental techniques' is the paper's wording for both predictions. The lesson about transport coefficients belonging to a level of description is the essay's.

The consequences cascade. Chapman-Enskog expansion on the extended Boltzmann equation produces hyperbolic-parabolic transport laws instead of the usual parabolic ones. Heat propagates as waves with finite speed, not diffusively. Momentum transport couples to orientational relaxation through a spin equation that links particle shape to fluid vorticity.

The structural contributions can dominate the classical transport coefficients. In a fluid of asymmetric molecules, the shape-dependent terms don't just correct the point-particle values — they can be larger than the baseline. The transport you measure isn't the transport of mass or momentum. It's the transport of structure.

The framework yields testable predictions: Mpemba crossover times for colloidal ellipsoids and shock widths for asymmetric molecules. Both are accessible with current experimental methods.

The Mpemba prediction is particularly clean: colloidal ellipsoids carry structural degrees of freedom (orientation, deformation) that store thermal energy differently from translational modes. The hot sample's structural modes decay on a different timescale than its translational modes. If the structural relaxation path happens to shortcut through the equilibrium neighborhood, the hot sample arrives first.

The lesson: transport coefficients are not properties of the material alone. They're properties of the material at a specific level of description. Add structure, and the coefficients change — not because the physics changed, but because the description caught something the point-particle model missed.