‘Liquid Gears’ That Rotate Without Contact Between Parts Have Been Created
Scientists have developed a technology of ‘liquid gears’ in which rotation is transmitted via magnetic fields without physical contact between parts. This opens up prospects for creating fundamentally new, nearly wear-free mechanical devices.
Liquid Gears: How Engineers Made Water Work Without a Single Tooth and Why Robotics Needs It
Introduction
Humanity has used gear transmissions for over five thousand years. The ancient Chinese employed them in mills and agricultural machines; the Greeks built them into the Antikythera mechanism to predict celestial motion. Over millennia, materials changed—wood gave way to metal and plastic—but the fundamental principle remained unshaken: motion is transmitted through physical contact of teeth. And where there is contact, there is friction, wear, and the need for lubrication. Researchers at New York University decided to rethink the very concept of a gear, creating a transmission without teeth and without any contact between parts. Their ‘liquid gears’ transmit rotation solely through fluid flows, and this breakthrough could transform entire fields in robotics, medical technology, and micromachinery.
Event Details and Timeline
The study was published on January 13, 2026, in the journal Physical Review Letters, and it attracted widespread attention from the technical press in late April with the release of detailed technology reviews. The project was led by NYU physics professor Jun Zhang and mathematics professor Leif Ristroph.
The experimental setup is surprisingly simple. Two plastic disks, 50 mm in diameter, 3D-printed from ordinary resin, are placed in a tank filled with a viscous fluid—a mixture of water and glycerin. The disks are separated by a gap of about 3 mm and do not touch at all. When the first disk starts rotating at 40 to 200 rpm, the fluid adhering to its surface is set in motion. Shear flows arise, transmitting momentum layer by layer to the second disk. Within milliseconds, the driven disk begins to rotate synchronously with the driving disk, lagging by only a few degrees—exactly like a classic gear pair.
The nature of motion transmission depends on the distance between the disks—this is the second key result of the experiments. When the cylinders are close together, the fluid flows form micro-vortices that cause the driven disk to spin in the opposite direction—the classic behavior of a gear transmission. When the distance increases and the driving disk speed is high enough, the flow forms a loop resembling an invisible belt: both disks rotate in the same direction. In other words, by changing the distance or fluid viscosity, one can instantly switch the direction of rotation and the gear ratio—something that in a mechanical gearbox requires a complex set of gears.
The researchers introduced tiny bubbles into the fluid to visualize the flows and confirm exactly how the fluid acts as both teeth and belt simultaneously.
Impact and Significance
The significance of the development becomes clear when considering the limitations it overcomes. Traditional gear transmissions require extremely high manufacturing precision: clearances are measured in microns, the slightest misalignment leads to accelerated wear, and a single grain of sand can jam the entire mechanism. A liquid transmission does not need precision machining—parts can be printed on a consumer 3D printer, and the 3 mm gap is three orders of magnitude larger than the tolerances required for metal gears.
The absence of contact means no wear, no noise, and no need for lubrication. Moreover, the system has built-in overload protection: if the driven disk suddenly jams, the fluid simply slips, damaging neither the parts nor the drive. No shear pins, friction clutches, or complex current-limiting algorithms—the physics of the fluid does this automatically.
However, the technology also has fundamental limitations. Viscous friction converts mechanical energy into heat, so efficiency drops as the gap widens or viscosity decreases. The current prototype transmits only a few milliwatts of power—enough for a miniature pump, but incomparable to the kilowatts passing through an e-bike gearbox. For applications requiring high torque—such as automotive transmissions—the liquid approach is categorically unsuitable. The rigidity of metal teeth in such cases is not a drawback but a necessary condition for power transmission.
The niche for this technology lies at the opposite end of the spectrum: soft robotics, medical devices, and microsystems. Polymer rotors can be autoclaved, sterilized with gamma radiation, or printed from biodegradable materials—this eliminates the problem of metal particles and toxic lubricants, critical for implantable devices. The absence of seams and gaps makes the design ideal for disposable drug infusion cassettes.
Reactions from Key Players
The first to respond to the development were medical device creators and soft robotics researchers. Medical equipment manufacturers were attracted by the possibility of autoclaving polymer rotors and the complete absence of metal particles, which complicate the operation of implantable pumps.
In the developer community, the project sparked explosive interest. Within weeks of the source data being published on GitHub, versions appeared with herringbone grooves and rotors on flexible hinges capable of changing geometry on the fly. One YouTube channel has already demonstrated a two-speed ‘gearbox’ controlled solely by redirecting water flows—without a single moving mechanical part.
Parallel research in related fields confirms that interest in contactless transmissions is growing across a broad front. Tohoku University and the University of Surrey have developed a magnetic transmission for reconfigurable 6G antennas, also eliminating physical contact. The Max Planck Institute and the University of Michigan have shown how swarms of magnetic microrobots create controlled fluid flows to rotate objects without touching them. The German Aerospace Center funded a student project, Ferrowheel, an orientation system using a ferrofluid bearing for the International Space Station. NYU’s liquid gears naturally fit into this growing landscape of contactless mechanical solutions.
Forecast and Conclusions
Liquid gears are at an early laboratory prototype stage, and their commercial future depends on solving the energy efficiency problem. Smart fluids—magnetorheological suspensions that increase shear strength tenfold under a magnetic field, and shear-thickening compositions that harden under load—could radically change the efficiency equation. Early tests at ETH Zurich and Osaka University show that torque densities above 10 N·m/L are quite achievable, which is already comparable to the characteristics of small planetary gearboxes used in surgical instruments.
The path to market will likely begin with niches where silence, sterility, and inherent safety are more important than energy efficiency: MRI-compatible surgical robots, soft exoskeleton joints, disposable micropumps for drugs. It is there that the quiet rotation of water will replace the metallic click of teeth—places where the traditional gear never felt quite at home.
The main lesson of this story goes beyond the engineering solution. For five millennia, the gear remained a symbol of mechanical inevitability: if you want to transmit rotation, make teeth. It turns out that water flowing in a three-millimeter gap between two plastic disks does the job just as well—and without wearing out, making noise, or forgiving assembly errors. This is not a replacement for industrial gearboxes, but the discovery of a new class of mechanical systems—soft, adaptive, and human-friendly. And in a world where robots increasingly work side by side with humans, such quality may prove more important than brute power.
— Editorial Team
No comments yet.