Pittsburgh Supercomputing Center

FOR IMMEDIATE RELEASE                         CONTACT:
March 18, 1994                                    Michael Schneider
                                                  Pittsburgh Supercomputing Center

                                                  Jacquelyn Savani
                                                  Princeton University

Supercomputing Aids Search for Pulsars

PITTSBURGH -- Using Cray supercomputers at the Pittsburgh Supercomputing Center, Princeton University physicist Joseph H. Taylor and his students have in the last three years found 20 previously unknown pulsars, including several with unusually fast rotation periods.

Taylor and his colleague Russell A. Hulse shared the 1993 Nobel Prize in Physics for their 1974 discovery of the first binary pulsar -- two neutron stars orbiting each other, one of them giving off the regular radio frequency "beeps" of a pulsar. This discovery has been important as a deep-space proving ground for Einstein's general theory of relativity.

Since 1974 Taylor has continued searching the heavens for pulsars, and he and his collaborators have found roughly half of more than 600 pulsars currently known. In recent years, Taylor added supercomputing to his pulsar-finding toolkit. This powerful technology makes it possible to analyze radio signals from space with far greater sensitivity and efficiency than was possible 20 years ago, and it is yielding a payoff in new pulsar discoveries.

To capture radio waves from space, Taylor and his Princeton research group use the largest radio telescope in the world, the 1,000-foot telescope at Arecibo, Puerto Rico. Taylor has used this same telescope since 1974, but now with supercomputing the group has improved ability to sift through the "noise" the telescope records. Supercomputing makes a difference especially, says Taylor, in searching for millisecond pulsars.

These superfast spinning stars turn a complete revolution -- called the rotation period -- in thousandths of a second, transmitting more "beeps" per second than slower pulsars.

"The first millisecond pulsar," says Taylor, "was the binary pulsar we found in 1974. Although that had a period of 59 milliseconds, it was the first of the genre. More recent ones found in the 80s have periods all the way down to 1.5 milliseconds. We would never have seen one of those with the computing equipment we had in 1974."

Only about 50 millisecond pulsars are known, 30 of them in distant globular star clusters, where gravity from other stars can interfere with precision measurements of the pulsar. Much more valuable for experiments in relativity are "field" millisecond pulsars -- outside of clusters. Only 20 of these are known, and five were discovered by Taylor's research group using supercomputing at Pittsburgh.

Probably several thousand millisecond pulsars lurk undetected in the galaxy, but finding them is much more difficult than finding slower pulsars. The noise from the telescope must be analyzed in smaller chunks of time, increasing the amount of data the computer has to process. "To find a pulsar with a period of a few milliseconds instead of half a second or a second," says Taylor, "is ten thousand to a million times more difficult."

When astronomers first discovered pulsars in 1967, they half-jokingly called them LGM for "Little Green Men," because the pulsing radio signal from space was so regular that it seemed it could be a sign of intelligent life. It is now clear that pulsars are neutron stars, incredibly dense spinning objects that beam radio waves in a cone shape through space, like a rotating lighthouse beacon.

Ticking like a cosmic metronome, pulsars are the most stable, accurate natural timekeepers known. "The neutron star is just sitting out there in interstellar space," says Taylor, "not touching anything, freely spinning on its axis. There's no friction of any consequence, and therefore it makes a very good clock." For this reason, pulsars are used to test accuracy of the most stable Earth clocks. They also have valuable cosmological applications in measuring gravity waves and in defining a precise astronomical frame of reference, important for space navigation.

Identifying pulsars as neutron stars brought astronomers face-to-face with an extreme state of matter. A neutron star is the cinder left when a giant star burns out and collapses on itself. Gravity overwhelms the electronic force that separates atoms from each other and crushes them into a mass ten trillion times denser than a lead brick. A single teaspoonful of neutron star material weighs as much as an ordinary mountain and is so dense that if it could be dropped over Earth it would fall like a stone through air, boring a hole in the planet as it fell.

The binary pulsar Taylor and Hulse discovered in 1974 gave physicists something they had been looking for since 1915 -- a way to test whether Einstein was right about gravity waves. Much like vibrating electrons give off electromagnetic radiation (radio waves for instance), Einstein's general theory of relativity predicts that an accelerating mass gives off energy as gravitational radiation, which can be thought of as ripples moving at the speed of light in the curvature of space-time. Gravity comparatively is a very weak force, however, and even Einstein questioned whether these waves could be detected.

Because the binary pulsar's two stars, each about 1.4 times the mass of the sun, are locked in close orbit with each other, their gravitational interaction is very strong and the resulting gravity waves, according to the theory, would measurably reduce the orbital energy. In other words, much as the orbit of an Earth satellite gradually decays toward Earth because it loses energy to the atmosphere, the pulsar and its companion should gradually spiral closer to each other and orbit faster and faster.

Detailed astronomical measurements by Taylor and his colleagues confirm that this is the case. The interval between radio pulses from the pulsar is getting about 75 millionths of a second shorter each year, almost precisely what general relativity predicts should happen as a result of the pulsar spiraling closer to its companion and shortening the period of its orbit. This is the strongest evidence to date that gravity waves exist.

The Pittsburgh Supercomputing Center, a joint project of Carnegie Mellon University and the University of Pittsburgh together with Westinghouse Electric Corporation, was established in 1986 by a grant from the National Science Foundation with support from the Commonwealth of Pennsylvania. Its purpose is to develop and make available state-of-the-art high-performance computing for scientific researchers nationwide.

# # #

Related article, with graphics, from Projects in Scientific Computing, PSC research report.

go back to contents page