Waiting for Gravity

Many millennia ago, in a distant patch of space some 6500 lightâ''years from Earth, a hot blue giant star exploded in catastrophic but glorious stellar suicide. In one stupendous runaway thermo­nuclear reaction, the star blasted off 95 percent of its gaseous outer layers, and its core collapsed, blazing so fiercely that for a few magnificent days that single star rivaled the total brilliance of all the other million or so stars around it. Over the next few months, the star’s naked core cooled and faded away, leaving a dim, dense neutron star one-twentieth of its original mass surrounded by a rapidly expanding multi­colored cloud of gases. Eventually, the star’s outer layers attained immortality as the gorgeous, gaseous Crab Nebula [see photo, ” ”].

The story doesn’t end there, though. For when an exploding star, or supernova, suddenly redistributes its mass, something strange also happens to its gravitational field. Thanks to Einstein’s general theory of relativity, we have a good idea what that something might be.

According to Einstein, the explosion’s powerful acceleration of star matter should generate distortions in the normal curvature of space. These hypothesized distortions, known as gravitational waves, would ripple outward into the universe at the speed of light, stretching and compressing the space around any objects they happen to pass through. Other acts of astronomical ­violence—galaxies colliding or black holes cannibalizing other black holes—should also emit gravitational waves.

Astronomers believe that if we could detect these waves, they would illuminate much about the universe that is now obscured. Detecting gravitational waves would also give physicists a defini­tive new test of general relativity. Newtonian physics doesn’t come close to explaining gravitation from black holes or other regions of strong gravity, but Einstein’s theory does; the observation of gravitational waves from such sites would serve to confirm, or possibly enhance, the theory. Gravitational waves may also give astronomers the means to look back to the earliest moments of cosmic evolution, when the universe was still small and dense. And they could provide a completely new way to survey the contents of the universe, perhaps even revealing phenomena that may not have electromagnetic signatures.

None of the current astronomical observatories—whether detecting visible light, radio waves, X-rays, or any other type of electromagnetic radiation—can peer too far below the surfaces of stars and other objects. Photons that originate deep in a star’s interior get absorbed, reemitted, or otherwise altered on their way out to the surface—it can take a million years for a photon to work its way from the core of our sun to the surface, for example. And once photons leave the surface, they may be further altered or blocked by gas and dust in space before ever arriving at a detector on Earth.

In contrast, scientists believe gravitational waves pass unaffected through all intervening matter, carrying with them intimate secrets about the universe’s most violent events that can’t be learned in any other way. It’s analogous to the way sounds detected by a stethoscope can reveal essential information about a person’s heart or lungs, details unobtainable by simply looking at the surface of the skin. Indeed, astronomers hope that gravitational waves may let us effectively listen to the very pulse of the cosmos’s most brutal and exotic events.

The hitch is that because gravitational waves travel right through matter unaltered, they are extraordinarily difficult to detect—some would even say impossible.

Undeterred, pioneering astronomers and physicists around the world have teamed with engineers to build technologically ingenious detectors to seek evidence of gravitational waves. This past November, the world’s two largest gravitational-wave detectors began their first full-scale run of observations. Like a pair of ears listening simultaneously for the same sounds, they are the twin L-shaped instruments of the Laser Interferometer Gravitational Wave Observatory, or LIGO (pronounced LYE-go). One of the observatory’s two sites, LIGO-Livingston, is located in a dense forest in Louisiana, 42 kilometers northeast of Baton Rouge in the southeastern United States [see photo, ” ”], and the other, LIGO-Hanford, is 3000 km away in the sagebrush desert of eastern Washington state. If gravity waves are to be detected anytime soon, these are probably the machines that will do it.

Talk about the middle of nowhere. After landing at the tiny five-gate airport in Pasco, Wash., a visitor has to travel another 20 km by car past brown-gray desert scrub to get to LIGO’s northern outpost. This also happens to be part of the now-desolate Hanford Reservation, famous for its pivotal role in the Manhattan Project, which built the first atomic bomb [see ”The Atomic Fortress That Time Forgot,” IEEE Spectrum, April 2005]. It’s a challenge following roads whose intersections are unmarked (a remnant of World War II secrecy), especially in a winter fog so thick that only three dashed lines are visible ahead on the wet asphalt. Finally, at a lone access road, a modest steel sign proclaims in black letters, ”LIGO: Hanford Observatory.”

LIGO-Hanford, like its Lâ''shaped twin in Louisiana, is based on a century-old and fundamentally simple instrument for precisely measuring length. Day and night, 16 000 times a second, LIGO repeatedly compares its two 4-km-long perpendicular arms by measuring the time it takes laser light to travel their lengths. If a gravitational wave passed through Earth, it would momentarily distort space, and scientists are betting it would also distort the lengths of LIGO’s arms in a characteristic pattern.

Even to hope to capture such a subtle phenomenon, though, the scientists and engineers who built LIGO had to tackle two key problems: how to measure that minute distortion and how to reject any other noise coming from terrestrial sources that might mimic or mask the distortion.

The idea of space ”distorting,” or having any kind of structure at all, may be foreign to nonphysicists, but physicists talk about the fabric of space and time—and they mean ”fabric” almost in the sense of cloth. Just as a smooth piece of cloth is actually made of vertical and horizontal threads woven in a grid, the three-dimensional vacuum of outer space can be envisioned as a grid that extends in all three dimensions. Any points in space, such as the two ends and the apex of the L-shaped LIGO instrument, can be thought of as fixed to three specific locations on that grid.

So what happens when a giant star explodes? Any mass curves the fabric of space and time, stretching the grid—the more massive the object, the greater the curvature. If a mass accelerates, such as when the supernova blasts away most of the star’s matter, the abrupt change in the gravitational field distorts that fabric, and the distortion radiates outward through the universe as a gravitational wave. It’s comparable to the wave you see quickly rippling outward from your hands when you snap a bedsheet.

As the wave passes through it, any pattern on that sheet—say, the capital letter L—is momentarily distorted as well. The two ends and the apex of the L haven’t moved their positions with respect to the threads of the fabric. But the fabric itself has been momentarily warped. In an analogous way, a passing gravitational wave stretches and compresses the fabric of space—stretching and compressing the grid of space itself. In effect, a gravitational wave should momentarily change the lengths of LIGO’s two arms without altering the positions of the arms on Earth.

A gravitational wave coming from directly above LIGO would have a characteristic pattern: first lengthening the X arm while compressing the Y arm, then compressing the X arm while stretching the Y arm [see diagram, ”Seeing the Light”]. LIGO is looking for length displacements that oscillate back and forth at frequencies between about 50 and 2000 hertz, within the audio range of a piano. ”If your ears were sensitive to gravitational waves, you could literally hear them,” notes LIGO beam tube designer Rainer Weiss, professor emeritus of physics at the Massachusetts Institute of Technology, in Cambridge. The wave’s duration—from seconds to minutes to months—would depend on the type of astronomical event causing it.

Physicists and astronomers hope LIGO is sensitive enough not only to detect the presence of a gravitational wave but also to measure its intensity, duration, and frequency, along with any changes in those characteristics over time. Because as surely as DNA can identify a culprit, gravitational waves emitted by a supernova are expected to look very different from those given off by a black hole swallowing a nearby star, by two galaxies ripping each other apart, or by other acts of astronomical carnage.

Having two LIGO sites instead of just one allows scientists to corroborate observations and reject any spurious measurements that may just mimic gravitational waves. Moreover, if a real gravitational wave should be detected at both sites, the minuscule difference in when the observations occurred and in the patterns of oscillation could give a rough idea of where the waves came from, much as two ears can home in on a cricket’s chirping in the night.

Throughout LIGO-Hanford and LIGO-Livingston, sensors monitor thousands of signals indicating the position, temperature, motion, noise, and other characteristics of every optical, mechanical, and electronic component in the observatory. All that information is funneled into the control room at each site, where it is displayed as brightly colored blinking images on dozens of computer monitors. Encompassing the Hanford control room with a sweep of his arm, LIGO-Hanford director Frederick J. Raab summarizes: ”Each LIGO interferometer has only one channel you care about”—namely, the photodetector, where a glimmer of light might signal a gravitational wave—”and several thousand other channels to determine whether you should believe that single channel.”

”LIGO needs to be as isolated as possible from man-made vibrations,” says Raab, as he climbs into a pickup truck for a tour of the site. ”Even so, the instrument is so sensitive to ground movements that we can tell when the pile drivers at a construction site 20 miles away stop pounding when the workers break for lunch.”

The LIGO sites don’t look like astronomical observatories—there’s no gleaming dome sheltering a huge optical telescope, no rotating parabolic dish listening for radio signals. From the outside, LIGO looks like an enormous L-shaped concrete bunker, each arm of the L running low to the ground toward the horizon.

Step inside, and the daylight streaming in through the open door illuminates a foil-wrapped tube extending into darkness. That’s the beam tube, one of the biggest ultrahigh-vacuum chambers on Earth: fully 1.2 meters in diameter and 4 km long. Each LIGO site has a pair of these mammoth tubes, one for each perpendicular arm. The tubes are pumped down to 10-9 or 10-10 pascals—equivalent to the atmospheric pressure at an altitude of 150 km above Earth.

With a vacuum almost as rarefied as what exists on the daytime side of the moon, ”LIGO is the biggest ’hole’ in the atmosphere ever built,” says Michael E. Zucker, codesigner of the vacuum system and head of LIGO’s Louisiana site.

At the apex where the two beam tubes meet sits a single 10-watt diode-pumped, solid-state neodymium-doped yttrium-aluminum-garnet laser, emitting a continuous beam at the near infrared wavelength of 1.064 micrometers. A beam splitter in front of the laser divides the beam in half. Four kilometers away, at the far end of each arm, a 10-kilogram precision mirror hangs from an elaborate suspension system that isolates it from ground vibrations. Each half-beam shoots down one of the instrument’s arms, speeds through 4 km of vacuum, reflects off the massive end mirror, and returns to the apex. There, it is reflected by another suspended mirror and sent down its beam tube again. After 50 such round-trips that bring the power of the light in each beam tube to 10 to 15 kilowatts, the two half-beams are recombined.

It is at that moment of recombination that astronomers anxiously seek evidence of gravitational waves.

”LIGO doesn’t form an image , even though it’s using optics,” explains Barry C. Barish, Linde Professor of Physics emeritus at the California Institute of Technology, in Pasadena. Barish had been director of the LIGO project for most of its construction and first science runs—from 1994 until late March of this year. ”Instead,” he says, ”each LIGO is continually comparing the lengths of its two arms to extraordinary precision.”

A thousandth of the diameter of a neutron—or 10^-16 centimeter—over a distance of 4 km: that’s the minute subatomic difference in the lengths of LIGO’s arms that the astronomers and physicists hope to see. Not many technologies can measure something so small. LIGO does it by precisely timing how long light takes to travel down each arm and back.

”LIGO’s basic principle for comparing the lengths of its arms is actually very simple,” declares LIGO-Hanford’s Raab. ”Basically, LIGO is the world’s biggest Michelson interferometer,” he says, referring to an instrument invented by 19th-century physicist A.A. Michelson, who for that invention and other achievements became the first American to win a Nobel Prize. ”Stripped down to basics, LIGO—or any Michelson interferometer—consists of just these five elements,” Raab says, pointing to a small table he uses to demonstrate to school groups. On it sit a pen-size laser pointer, a pair of prisms glued together to form a beam splitter, two mirrors at the ends of perpendicular arms, and a projection lens that serves as a photodetector through which laser light is projected onto a screen, ”This is a one-ten-thousandth-scale miniature LIGO that any high school student can build for under $150,” he says. ”Makes a great science fair project.”

”Can’t use a flashlight,” Raab adds. ”You need a laser because its light is coherent. That means all its wave fronts are in phase, so all the light waves’ crests and troughs are in step.” He attaches a clothespin to the laser pointer to keep its on-off button in the on position. The laser light enters the beam splitter, and each half-beam bounces off its respective end mirror and then returns to the beam splitter, where the beams are recombined.

”But when the half-beams recombine, they interfere with each other,” he continues. ”Watch this.” Gently, Raab pulls outward on a string attached to one of the two mirrors, effectively increasing the length of the beam’s path. ”If the two arms are the same length—or differ in length by a whole number of half-wavelengths—the half-beams interfere destructively: the wave fronts recombine out of phase, and nothing shows on the detector.”

Sure enough, the projection screen behind Raab’s model remains dark. ”But if the two arms differ in length by a quarter-wavelength—or any odd whole number of quarter-wavelengths—the two half-beams interfere constructively: the wave fronts recombine in phase, and a bright spot shines on the detector.” Raab pulls the string slightly more, until a bright spot appears on the projection screen.

In the full-size LIGO, the end mirrors are adjusted so that they’re in a state the scientists call optical lock—the ideal condition when the instrument’s closed-loop feedback control systems are holding all the mirrors properly aligned so that the laser light resonates in the beam tube. Absent any passing gravitational wave, the wave fronts from the half-beams recombine exactly out of phase, and essentially no light reaches the photodetector, which sits near the apex of the two beam tubes.

But when, or if, a gravitational wave passes through, altering the relative lengths of the light paths in the beam tubes, a small amount of light proportional to the strength of the gravitational wave will appear at the photodetector. If the length change is a quarter of a wavelength, or 266 nanometers, the change will be detected as a sudden bright spot on the photodetector.

In the central control room at each LIGO, one of the many overhead television and computer monitors allows the astronomers and operators to keep an eye on the photodetector. Says MIT’s Weiss, ”We’re just waiting for a glimmer of light through the dark port”—light that would signal the first detection of a gravitational wave.

The Michelson interferometer technique is so sensitive that even Raab’s $150 model can detect changes in the relative lengths of its 12.7-cm-long arms as small as 10-10 cm. The full-size LIGO is a million times as sensitive as that, the result of the laser’s higher power, the beams’ repeated round-trips through the tubes, and electronics and optics that ensure that the beam is as round, coherent, and aligned as human engineering can make it.

Being able to detect a gravitational wave is just part of the story. ”Isolating LIGO from things that aren’t gravitational waves, that’s the other whole trick and art of doing this experiment,” remarks Zucker, who in addition to directing LIGO-Livingston is also a physicist at MIT. Isolating LIGO means identifying and then rejecting all other sources of vibration that could distort the lengths of LIGO’s arms, thereby mimicking gravitational waves.

No place on Earth is free from vibrations. Every day, the earth’s elastic crust swells and contracts as the coming and going of sunlight heats it up and cools it down. It also rises and falls about 30 cm because of tides in the solid earth caused by the sun and the moon. Meanwhile, ocean waves pounding against the shore hundreds of kilometers away—LIGO-Hanford is most troubled by the Pacific Ocean and LIGO-Livingston by the Gulf of Mexico and the Atlantic Ocean—cause a microseism, a rolling motion of the crust every 6 or 7 seconds, day and night. Even winds higher than 35 km per hour can be problematic.

Moreover, a couple of dozen times a week, an earthquake of 5 or greater on the Richter scale sets the earth’s crust quivering. ”An earthquake of magnitude 5.5 or 6.0 anywhere in the world is enough to make us drop lock”—cause the suspended mirrors to sway enough to throw off the instrument—”for a good 10 minutes,” notes Joseph Giaime, associate professor of physics and astronomy at Louisiana State University, in Baton Rouge, and chief scientist of LIGO-Livingston.

Indeed, the magnitude 9.0 earthquake that gave rise to the December 2004 tsunami in southeast Asia, half a world away, set LIGO’s mirrors swinging for several days. Washington state even has its own occasional earthquake. A 6.8 earthquake near Olympia in February 2001 damaged some equipment at LIGOâ''Hanford and set the project back three or four months. And then there are the sporadic vibrations from lightning strikes, earth-moving machinery, and passing railroad trains—even from garbage trucks emptying dumpsters on the LIGO grounds.

All those vibrations could change the length of LIGO’s arms in a way that more or less mimics a passing gravitational wave. For that reason, LIGO’s physicists and engineers devoted much head-scratching to figuring out how to isolate the instrument’s mirrors from the ground, almost as if the apparatus were floating in outer space instead of attached to Earth. They came up with two isolation systems: one passive and one active.

The passive isolation system is used at both LIGO sites. Inside the vacuum at the far end of each beam tube, each 10â''kg mirror hangs in a delicate balance, resting on a thin steel piano wire that loops around its circumference. The two ends of the piano wire are strung from a steel cage bolted to an aluminum optical table, forming a pendulum that helps isolate the mirror from vibrations.

The table, whose surface is riddled with a grid of holes for mounting lasers and other optical equipment, rests on stacks of stiff springs alternating with massive circular steel plates that also help damp vibrations up to about 40 Hz. In turn, the springs and masses sit on heavy steel tubes that pass through the walls of the vacuum chamber and are anchored in the meter-thick concrete floor.

To illustrate the effectiveness of simple springs, masses, and pendulums as a low-pass filter, Dennis Coyne, LIGO chief engineer at Caltech, hangs a small weight from a Slinky toy. ”If you move the Slinky very slowly either from side to side or up and down, the weight will follow your hand,” Coyne explains. In the same way, LIGO’s optical tables will follow low-frequency movements of the earth (such as tides) without disturbing the positions of the mirrors.

”But if you wiggle your hand fast either vertically or horizontally, the weight will just stay in one place,” he continues, demonstrating with the Slinky. In the same way, the pendulums, springs, and masses in LIGO isolate the optical tables from higher-frequency motions.

For the most part, this passive system works just fine at Hanford. Indeed, passive isolation was all that was originally planned for the first decade or so at both sites. But LIGOâ''Livingston is built on a tree farm where pines and other fast-growing trees are grown for paper. Throughout the year, sometimes just tens of meters from the interferometer’s delicate optics, ”the loggers, bless ’em, chop trees, drop them onto the ground with a thud, buzz off the branches, then claw the trunks up in a crane and drop them rumbling into the back of a truck that roars away,” recounts LIGO-Livingston’s Zucker. Until mid-2004, even with the passive isolation system, LIGO-Livingston ”could only run at night. Every morning we’d lose lock as the logging machines started up,” he says.

To counter these large-motion disturbances, which occur mainly at the low frequencies of 1 to 10 Hz, Giaime’s team—which included scientists and engineers from Stanford, MIT, and Caltech, as well as from LIGO-Livingston—accelerated the development of an active isolation system they call HEPI, for Hydraulic External Pre-Isolator [see photo, ” ”].

Each of the nine optical tables at Livingston is fitted with a HEPI system. The system consists of four oddly shaped assemblies that sit between the ground and the passive spring-and-weight stacks holding up the tables. Inside each assembly are two stainless steel shoebox-size actuators—one horizontal and one vertical—attached to a vertical beam. The beams from the four assemblies support the optical table.

As the earth’s crust moves, the HEPI system measures the motion of LIGO’s concrete slab floor with a set of seismometers. Then it applies vertical and horizontal forces to the assemblies that are equal in magnitude but in the opposite direction, to counteract any shaking of the earth—thereby keeping the optical tables and their suspended mirrors more stationary than the earth itself.

The fastest a HEPI can move is 80 wm per second, or about a millimeter in 10 seconds. That speed is barely perceptible to a patient observer, but it’s more than enough to quiet not only vibrations from nearby logging, but also the periodic tremors of the microseism, which is about three times worse at Livingston than at Hanford because of its proximity to the Gulf of Mexico. Proof of the isolators’ effectiveness is the fact that they have enabled LIGO-Livingston to run night and day, despite the logging, since they were installed two years ago.

Now, astronomers watch and wait. But what if LIGO detects no gravitational waves? ”In a completely new field, you build successively more sensitive instruments as technology allows and see what you find at each stage,” Zucker says. Even as LIGO works through its first full-scale science run, which will end in the summer of 2007, its scientists and engineers are building upgrades for what they call ”advanced” LIGO—including larger mirrors, more powerful lasers, and a HEPI system at Hanford as well as at Livingston, all of which will make both instruments an order of magnitude more sensitive. If approved, installation of the new components is to take place from 2011 to 2014.

”If we find events and uncover a new phenomenon, that would be very exciting,” Zucker adds. ”But if we see nothing—especially after advanced LIGO—that would be very curious, and would carry serious implications for our understanding of general relativity and what’s out there in the universe.”