How pipe flow transitions from laminar to turbulent has been an open problem for 200 years. Classical fluid mechanics treats it as a stability problem — perturbations grow or decay depending on conditions. But the transition resists clean mathematical treatment. It is noisy, history-dependent, and sometimes discontinuous.
The resolution came from a different field entirely. In sufficiently curved pipes, the laminar-to-turbulent transition becomes discontinuous — the turbulent fraction jumps suddenly beyond a critical flow velocity, analogous to water freezing below a critical temperature. The framework is “tricritical directed percolation” from statistical mechanics.
The process is mathematically equivalent to percolation — the same mathematics that describes water filtering through coffee grounds or fluid finding connected paths through a porous medium. Turbulent puffs nucleate, grow, split, and merge through a process governed by directed percolation statistics, not by deterministic instability theory.
The 200-year-old open problem in fluid dynamics was a problem in statistical physics. The right abstraction was hiding in an adjacent discipline's formalism. Classical fluid mechanics looked for deterministic equations governing the transition; the answer was stochastic, requiring the tools of phase transitions and critical phenomena.
When a field's own tools cannot solve its oldest problem, the answer may not be deeper investigation within the field but lateral movement to a framework that treats the same phenomenon as a different kind of object. Turbulence isn't an instability. It's a phase transition.