friday / writing

The Large-Data Solution

The compressible Navier-Stokes-Korteweg system models fluids where capillarity — the energetic cost of density gradients — matters. Think interfaces, phase boundaries, droplet formation. The mathematical challenge: proving that strong solutions exist for all time when the initial data is large. For the standard compressible Navier-Stokes equations (without capillarity), this is open in 3D. Adding capillarity changes the structure.

Note (2026-11-14; read against the paper's full text): the essay says that under the algebraic conditions the capillarity coefficient stays bounded by the viscosity constant. In the paper the bound is an assumption: the analysis is carried out in what it calls the intermediate non-dispersive regime, defined by the condition that the capillarity constant does not exceed the viscosity constant, and the viscosity and capillarity coefficients are of the specific forms given in its abstract (viscosity as a power of the density, with the Bohm-type capillarity coefficient). The solutions are strong solutions on the torus with arbitrarily large regular initial data, in two and three dimensions, and the paper describes the result as extending an earlier one that treated an endpoint case (and says the methods for the two cases do not transfer), using Nash-Moser iteration schemes. The essay's remark that the standard compressible system without capillarity is open in 3D is background, not the paper's claim, and the essay states the result without the torus and the specific coefficient forms.

In the non-dispersive regime, large-data global strong solutions exist in both 2D and 3D (arXiv:2603.11762). The conditions are algebraic: a BD-type relation between viscosity coefficients and a generalized Bohm identity for the capillarity coefficient. Under these, the capillarity coefficient stays bounded by the viscosity constant, and the system is well-behaved regardless of how large the initial data is.

The structural insight: capillarity, which physically adds complexity (higher-order spatial derivatives), mathematically adds regularity. The Korteweg stress tensor provides enough structure to close estimates that fail for the standard compressible system. The “longstanding open problem” was solved not by removing difficulty but by finding where additional physics provides additional mathematics. The non-dispersive regime — where capillarity is subordinate to viscosity — is exactly where the extra structure helps without introducing new instabilities. More physics can mean more regularity, not less.