From its perch outside the International Space Station (ISS), the Alpha Magnetic Spectrometer (AMS) has spent the past 8.5 years collecting charged cosmic rays that move through space at nearly the speed of light. Unlike particle detectors on Earth, AMS is not attached to an accelerator but instead studies its quarries as they exist in the vacuum of space, unadulterated by interactions with our planet’s atmosphere. The experiment aims to tackle some of the universe’s biggest quandaries, such as what makes up the missing dark matter that seems to dominate the cosmos and why it contains more matter than antimatter.

So far AMS has studied more than 145 billion charged cosmic rays, resulting in 16 scientific papers and a bevy of intriguing questions. Yet a problem with the experiment’s cooling system that showed up in 2014 threatened to cut short the scientific work. NASA decided to replace the faltering hardware during a quartet of complicated, intricate spacewalks, or extravehicular activities (EVAs), that began on November 15. If successful, AMS should be able to continue taking data for years, potentially resolving some of the puzzling findings it has made so far.

Tantalizing Mysteries

One of the first mysteries AMS revealed had to do with the energies of the incoming positrons—the antimatter counterparts of electrons. It found that the spectrum of positrons picks up sharply at 25 billion electron volts (GeV), climbs to about 300 GeV, then drops precipitously and cuts out at around 1,000 GeV. “It’s totally irregular,” says Samuel Ting, a Nobel laureate physicist at the Massachusetts Institute of Technology, who leads the 600-member AMS science team.

The most logical explanation, Ting says, is that the particle detector, a project of 56 agencies in 16 countries, is seeing by-products of collisions of relatively high-mass dark matter particles. The peculiar pattern, however, could also stem from nearby pulsars or some other phenomenon.

“The drop-off of positrons is interesting, whether it is caused by dark matter, pulsars or something else,” says M.I.T. particle physicist Tracy Slatyer. “If it’s dark matter, it gives you a much more accurate handle on what its mass needs to be. If it’s coming from local pulsars, then it tells you something important about how the particles from those pulsars have to propagate through the galaxy. [By] doing better measurements of that high-energy cutoff, where the statistics are currently pretty limited, maybe you could exclude one of the explanations.”

Another intriguing finding is AMS’s detection of antihelium particles—helium’s antimatter counterparts, which contain two antiprotons and an antineutron—a result the AMS researchers have not yet published. “We want to make sure we can also see anticarbon and antioxygen,” Ting says. “We have less than 10 antihelium signals, and the rate is one antihelium to 100 million helium. That’s like looking for three people within the population of the United States.”

Repair Job

More definitive answers will come only with additional data from AMS, which was installed on the ISS during the final flight of the space shuttle Endeavour in May 2011. The $2-billion detector, which includes five science instruments, 300,000 electronic channels and 650 fast processors, is a wondrous machine. But it was not designed to be serviced, especially not by gloved and space-suited astronauts bumping around in microgravity. The AMS cooling system includes four small pumps, each outfitted with an impeller the size of a quarter that spins at 6,000 revolutions per minute to move six grams of carbon dioxide through the system. Only one pump is used at a time, and each has a three-year design life, so theoretically, the system should last 12 years. But the way the first pump failed raised concerns about the rest. “We came to realize that the pumps themselves were a bad design,” says Mark Sistilli, NASA’s AMS program manager.

It turned out the pumps did not have enough internal lubrication and were slowly grinding themselves up, which caused them to jam. NASA kicked off an effort to figure out if replacing the cooling system by space-walking astronauts was even possible. “We really didn’t want to see AMS fail for lack of this cooling,” Sistilli says. Drawing from techniques and tools developed for the five shuttle-based servicing missions for the Hubble Space Telescope, NASA spearheaded a four-year program to similarly repair the AMS.

In addition to a new pump system, engineers and technicians designed, prototyped and tested about 25 specialized tools for the job and arranged for all the items to be flown to the station onboard three different cargo ships. The agency also trained ISS crew members Luca Parmitano of the European Space Agency and Andrew Morgan of NASA for the grueling work ahead.

By the time the astronauts floated outside the station’s airlock last month for the first of four planned spacewalks, AMS was down to just one pump, working a 60 percent duty cycle every month or so and nearly out of CO2 coolant. “We knew it was only a matter of time until the fourth pump was going to go,” Sistilli says.

Delicate Maneuvers

During the first 6.5-hour outing, Parmitano and Morgan removed a protective debris shield and hoisted it into space so it would not pose a future impact threat to the station. The astronauts then removed insulation and a cover from an internal AMS support beam to expose 10 stainless steel tubes, each the width of a soda straw. Six would be cut and spliced into the new pump package on the next space walk along with another two tubes on the other side of the detector. A week later, the astronauts were back outside for another six hours of work, carefully cutting eight coolant line tubes, permanently disabling the original cooling system. During a third outing on December 2, they installed the new thermal plumbing system.

The timing of the fourth and final AMS-repair space walk has not yet been determined. During that outing the astronauts will fix any leaks that may appear in the spliced plumbing lines, which have not yet been pressurized. They will also install an insulation tent over the new pump box to help regulate temperature.

The cooling system is critical for operation of the heart of AMS—nine layers of silicon trackers that measure the trajectory and electrical charge of transiting particles. AMS’s powerful magnet curves the path of incoming particles, turning each one way or another, depending on whether they hold a positive or negative charge. Measurements of a particle’s curvature in a magnetic field can also be used to calculate momentum. Scientists are slowly assembling AMS’s energy and momentum measurements into a pointillistic—and possibly new—view of the universe.

NASA will not know until after the fourth space walk whether AMS is back in service. “We’ve essentially just done a heart transplant,” says Ken Bollweg, AMS project manager at NASA’s Johnson Space Center in Houston. “Our goal is to have it last for at least another 11 years, or as long as the space station is in orbit.”

Despite the complexities, NASA never hesitated over whether fixing AMS was worth the time and effort. “This is what station is for: to do science that can’t be done anywhere else on a scale that can’t be done anywhere else,” says ISS director Samuel Scimemi. “Whether or not Dr. Ting’s science does anything practical in the near term, I don’t know. But if this research turns out to be something that adds to our human existence in an area that no one has ever done any work on before, that will be value enough.”