Feynman's sprinkler puzzle solved: momentum flux theory confirmed

The reverse sprinkler problem represents one of physics' most enduring counterintuitive puzzles. Ernst Mach posed it in 1883 as a thought experiment: if you reverse the flow of a spinning lawn sprinkler so it sucks water instead of spraying it, will it still rotate? The question languished until the 1940s, when physicists at Princeton—including a young graduate student named Richard Feynman—began debating the answer earnestly. Feynman even tested the idea experimentally in a cyclotron laboratory, but his result (no sustained rotation) left the question unresolved, as other researchers later obtained different outcomes depending on experimental conditions.

The core difficulty is that intuition misleads: one might expect a reverse sprinkler to behave like a regular sprinkler played backward, but fluid dynamics does not work that way. Mach argued the forces would cancel; Feynman observed a transient tremor but no sustained rotation; and still others reported steady reverse rotation under different conditions. Since then, experimental results have been inconsistent, reflecting variations in friction, flow rate, and sprinkler geometry. The 2024 momentum flux theory proposed by Leif Ristroph and colleagues offered a unified explanation: in reverse mode, water jets shoot into the central chamber where the arms meet and collide off-center, generating rotational torque. The latest work validates this theory by extending it to multiple silly sprinkler designs tested in both forward and reverse modes, with observations that strongly support momentum flux theory while contradicting both Mach's and Feynman's original hypotheses.

Beyond the historical puzzle, the findings carry practical significance. Understanding how fluid flows generate torque and rotation has direct applications to engineering systems—turbines, pumps, and other devices that convert or harness fluid motion. By identifying specific design principles (particularly how arm shape controls jet flow), the research provides a foundation for future engineers to predict and control how fluid dynamics will drive rotation in real-world systems.