A vortex filament — a thin tube of concentrated rotation in a fluid — supports wave excitations discovered by Lord Kelvin in 1880. Kelvin waves are helical disturbances that propagate along the filament, rotating as they travel. Their dispersion relation — the relationship between frequency and wavelength — has been known for over a century. The waves are linear: small-amplitude perturbations of the filament's shape that obey superposition.
In 1972, Hasimoto showed that the vortex filament equation maps exactly onto the nonlinear Schrodinger equation through a gauge transformation that encodes the filament's curvature and torsion as the amplitude and phase of a complex field. The nonlinear Schrodinger equation is integrable — it supports solitons, localized disturbances that propagate without dispersing and pass through each other unchanged. This mathematical connection predicts that vortex filaments should support soliton-like localized structures: lumps of curvature that travel along the filament without spreading, surviving collisions.
Sterkers and Krstulovic (arXiv 2602.22439, February 2026) demonstrate this numerically in three-dimensional viscous Navier-Stokes simulations — not in the idealized filament model, but in the full equations of fluid mechanics.
The simulations create vortex filaments in a viscous incompressible fluid and excite localized disturbances on them. The Kelvin wave dispersion relation matches Kelvin's 1880 predictions — the linear theory works even in the full Navier-Stokes context. More remarkably, localized soliton-like structures propagate along the filaments with the properties predicted by the Hasimoto-NLS correspondence: they maintain their shape, travel at the expected velocity, and survive collisions with other solitons.
The survival is approximate, not exact. The Hasimoto map assumes an infinitely thin filament in an inviscid fluid — a one-dimensional object with no dissipation. A real vortex in a Navier-Stokes simulation has finite core size and viscous decay. The soliton loses energy to viscous dissipation and radiates small-amplitude Kelvin waves during collisions, effects absent in the integrable limit. But the qualitative behavior — shape preservation, nonlinear propagation, near-elastic collision — persists far beyond the formal validity of the mapping.
This is an integrable structure surviving into a non-integrable system. The nonlinear Schrodinger equation is special: its integrability depends on exact cancellations that a viscous three-dimensional fluid has no reason to preserve. Yet the soliton, which is a creature of that exact integrability, appears robustly in the full fluid simulation. The explanation is that the Hasimoto map captures the leading-order nonlinear dynamics of the filament, and the corrections from finite core size and viscosity are perturbatively small at the scales and Reynolds numbers studied. The soliton is not exact in the Navier-Stokes equations, but it is close enough to be recognizable and functional.
The authors propose an experimental protocol for generating vortex solitons in a laboratory — creating them on real fluid vortices, not just simulated ones. The mathematics predicted the structure in 1972. The simulation confirmed it in 2026. The experiment would close the loop: a soliton traveling along a vortex in a tank of water, maintaining its shape because the equations of fluid motion, in this narrow regime, remember the integrability they shouldn't have.