Turbulence in pipe flow at Reynolds numbers above 10,000 is treated as a fait accompli. The flow is turbulent, the pressure losses are high, and any attempt to suppress the turbulence requires active intervention — oscillating walls, polymer injection, blowing and suction through the pipe surface. The energy cost of fighting turbulence is considered part of the price of moving fluid.
Bagheri, Becker, and Schlatter found that a passive geometric modification — increasing the streamwise curvature of a 180-degree pipe bend and changing the cross-section from circular to oval — collapses the turbulence entirely. At Re = 10,000 and Re = 20,000, the modified bend relaminarizes the flow. Pressure losses drop by 53% compared to a standard bend and 36% compared to an equal-length straight pipe. No energy input. No sensors. No feedback loop. The shape of the pipe does the work.
The mechanism is specific: the combined curvature and oval cross-section suppress streamwise Reynolds stresses and weaken the secondary flow patterns that feed the near-wall regeneration cycle. Turbulence in a pipe is sustained by a self-reinforcing loop near the wall — streaks generate vortices, vortices regenerate streaks. Disrupt that loop and turbulence has no mechanism for self-maintenance. The modified geometry disrupts the loop by altering the velocity field in a way that starves the cycle of the conditions it requires.
The distinction from active control matters. Active methods suppress turbulence by adding energy to counteract the turbulent fluctuations. The passive geometry does something different: it removes the conditions under which turbulence can sustain itself. The turbulence doesn't fight back because there's nothing to fight against. The intervention isn't opposing the turbulence — it's removing the substrate on which turbulence depends.
The broader implication: some phenomena assumed to require active suppression may be sustainably eliminated by modifying the context in which they arise. The turbulence was never an inherent property of the flow. It was a property of the flow in that geometry. Change the geometry, and the same fluid at the same velocity does something different. The container was always a control parameter — it was just being held constant.