Liquids may soon be able to be pumped through computers without the help of mechanical parts.

Yale electrical engineering professor Hur Koser and a small research team have developed a pump that allows liquids such as mineral oil, when mixed with magnetic particles, to move through pipe systems using only magnetic fields. The project, published Thursday in the physics journal Physical Review B, offers exciting opportunities in the development of cooling devices for electronics like computers, researchers said.

“When I studied more about magnetic fluids, I thought, ‘Why use bulky mechanical pumps when we can use magnetic fields to pump liquids?’” Koser said.

The researchers wrapped electrical wires around plastic pipes to propel magnetic fluids called ferrofluids through the pipes. This “ferrohydrodynamic pump” method uses the wires to generate a magnetic field, rotating the magnetic particles in the fluid so that the liquid is propelled forward.

Koser said this pump, with high flow rates and without the large mechanical components of conventional pumps, has great potential in applications such as portable electronics cooling, where one of the primary barriers to increasing processing speed and decreasing size is the cooling fan or pump.

Koser thought of the design after working as a graduate at MIT to power a turbo engine using a magnetic motor of similar design, he said. He realized at that time the potential that magnetic fluids could offer fluid pumping mechanisms if they were powered by similar technology. Conventional methods of ferrofluid pumping include pulling the liquid along the pipe with a moving magnet, which is over 10 times slower than the method Koser has developed, he said.

The project began around 2003 as a collaboration between Koser and Leidong Mao, who worked in Koser’s lab until 2007. After developing the initial theory, Mao said they consulted with MIT professor and former peer of Koser, Markus Zahn, for his expertise in ferrofluids and to use his equipment to track the motion of the fluid in the pipes.

“The data collection was all finished before I left Yale,” Mao said. “But then the question became how to make sense of the data. What we observed was not what our initial theory predicted.”

For the past three to four years, Koser and Mao have been developing a new theory to explain why the fluid was able to flow as fast as it did, said Mao. The team ran into difficulties during this process in finding computer software advanced enough to run simulations, as well as dealing with the complexity of measuring flow pressure and flow rate, Koser said.

Because of this, the study has also helped uncover new factors in describing fluid motion, Koser said.

The pump promises to be an efficient and inexpensive technology with many applications, Koser said. The materials used to construct the pump, including the ferrofluid, are widely commercially available.

“What I’m really envisioning is a circulation system in robots, for example, based entirely on magnetic fluid pumps,” Koser said.

The next steps, Mao said, are to optimize the design, especially for smaller scales, so that it can begin to be integrated into electronic devices.

The study was funded in part by the National Science Foundation.