"The Bounded Catastrophe"
The oldest strategy for proving that fluid equations behave well is to show that energy stays finite. If the total energy of the flow is bounded, the flow cannot develop infinite velocities — or so the intuition goes. This intuition is wrong.
Shi constructs smooth solutions to a system derived from the 3D axisymmetric Euler equations that explode in finite time. The velocity field develops a singularity. But a natural weighted energy — the quantity you would monitor to detect trouble — remains uniformly bounded throughout. The catastrophe happens without the energy budget noticing.
The mechanism is geometric. The blow-up concentrates along specific "ridge ray" angles in the domain. Along these rays, the dynamics reduce to a one-dimensional Riccati equation — the simplest kind of ODE that can blow up. The energy, being a spatial integral, averages over all angles. The catastrophic concentration at a set of measure zero is invisible to any integral quantity.
This doesn't solve the millennium problem of Navier-Stokes regularity — the system studied is a reduction, not the full equations. But it eliminates one of the main strategies people have tried. Energy boundedness, by itself, cannot rule out singularity formation. Whatever proof eventually works will need something more than energy.
The lesson extends beyond fluid mechanics. In any system where a conserved quantity is a spatial average, singularities can hide at points of concentration that the average cannot see. The budget is balanced. The catastrophe is local.
