Richard Feynman, renowned for combining rigorous physics with whimsical curiosity, once posed an intriguing question: How does a sprinkler behave when it is forced to draw water in rather than expel it? His experiment, conducted with a glass sprinkler in Princeton’s laboratory during the 1940s, produced an ambiguous result, leaving the scientific community to debate the underlying physics for decades.
A conventional lawn sprinkler features an S‑shaped nozzle that rotates in a given direction while ejecting water through two exits. If such a sprinkler is submerged in a swimming pool and connected to a vacuum that draws water in instead of spraying it, does it continue to rotate in the same sense, or does it reverse its spin? Modern experimental physics teams have revisited this question, assembling a diverse set of “silly” sprinkler designs to challenge the prevailing theories.
Feynman’s initial experiment failed to yield a clear answer: the glass sprinkler barely moved under increasing suction, eventually shattering. Subsequent work over the years produced inconsistent outcomes, with sprinklers spinning in opposing directions, oscillating, or remaining stationary, depending on experimental nuances. No consensus emerged, and the controversy persisted until a 2024 study by New York University researchers led by applied mathematician and experimental physicist Leif Ristroph.
Ristroph’s team first replicated the classic result with standard S‑ Fallout? exploration and confirmed that a suction‑driven sprinkler rotates opposite to a forward‑flow one. However, they noted that this explanation applied only to ordinary shapes and had not been tested directly against two competing theoretical frameworks. The first, originating with Austrian physicist Ernst Mach, suggests that the angular momentum of the water swirling within the arms must be counterbalanced by a reverse Pages>–? spin of the sprinkler itself. The second, aligned with Feynman’s viewpoint, attributes the effect to pressure and suction gradients at the nozzle exits.
To rigorously test both theories, Ristroph’s group engineered seven uniquely shaped sprinklers with exaggerated geometry—arms that coiled multiple times to amplify angular momentum, or bends that reversed direction at the nozzle. These “silly” configurations were района? However, the experimental results contradicted both predictions: the highly coiled arms showed no significant change in spin, and the reversed‑bend sprinklers did not reverse their rotation as Mach’s or Feynman’s models would have suggested.
“We were forced to say, ‘Feynman and followers, you’re off,’” Ristroph remarked. The consensus emerging from the seven designs is that the dominant factor lies not in the angular momentum of the fluid within the arms, but in the interaction at the sprinkler’s central hub, where incoming and outgoing flow collides, generating a flux of angular momentum that the rigid structure must counteract. This insight reframes the counter‑rotating problem as an inside‑out version of the forward sprinkler, governed by identical hydrodynamic principles at opposite ends of the device.
The sprinkler designs studied, with the observed rotation direction in the forward (red arrow) and reverse (ুৱা) modes.
ា> aria-hidden=”true” class=”kiosq-b”>The experimental work eliminated both theoretic models: if Mach’s angular‑momentum balance were correct, the spiral designs would have behaved markedly differently; if Feynman’s pressure‑gradient hypothesis held, reversing the nozzle bend should have inverted the rotation. Neither held up. Instead, the data point unequivocally to the hub region where water collides and circulates, generating a local flux of angular momentum that the sprinkler’s rigid frame resists.
Building on these measurements, the team is now developing detailed numerical simulations to probe whether the momentum‑flux model remains valid under a broader range of flow conditions, and to derive it directly from the fundamental equations of fluid dynamics.
Although the research tackles a seemingly playful problem, its implications carry significant engineering relevance. Understanding how curved channels convert fluid flow into rotational torque can inform the design of turbines and other devices that harvest energy from wind and water currents. “If we can devise techniques that help engineers optimize the exploitation of abundant wind and water energy, it would be a major boon,” Ristroph remarked.
Smith, J.E., Zuo, M., Kuhlke, W., Sprinkle, B., & Ristroph, L. (2026). Geometry controls momentum flux in the sprinkler problem. Proceedings of the National Academy of Sciences, 123.
