From the archives: Inside the tantalizing quest to sense gravity waves

To mark our 150th year, we’re revisiting the Popular Science stories (both hits and misses) that helped define scientific progress, understanding, and innovation—with an added hint of modern context. Explore the entire From the Archives series and check out all our anniversary coverage here.

When two blackholes collided 1.3 billion years ago, the impact released three suns’ worth of energy into the fabric of spacetime. On a Monday in 2015, at a remote facility in Hanford, Wash., researchers detected that ancient cosmic impact as its effects swung past Earth. They mapped its gravity wave, whose length was almost incomprehensibly small (1/10,000^th the diameter of a proton), to audio and heard a whoop. That tiny soundtrack was more than 100 years in the making.

Physicists had been seeking ways to detect gravity waves—ripples in spacetime caused by massive events—ever since Einstein predicted their existence in 1916. In an April 1981 article, Popular Science’s editor, Arthur Fisher, described the hunt for gravity waves, calling it “one of the most exciting in the whole history of science.” The Laser Interferometer Gravitational-Wave Observatory, or LIGO, was responsible for sensing the 2015 wave, but, as Fisher explains, in 1981 it was just one among many competing initiatives, each pursuing a different measurement technique. 

Rainer Weiss, MIT physics professor (now emeritus) and Kip Thorne (Caltech) were among the many scientists Fisher met and interviewed. Weiss devised the laser interferometer’s design in the 1970s and later teamed up with Thorne and Barry Barish to build LIGO (all three earned the 2017 Nobel Prize in Physics for their efforts). Ever since that first cosmic whoop in 2015, LIGO has detected 90 different gravitational wave events. 

In his story, Fisher describes the far-out wilds of space responsible for shaking space-time, including starquakes, gamma-ray bursts, and ticking neutron stars (pulsars). But it was Weiss, shortly after his device detected its first gravity wave in 2015, who captured space’s turbulence best: “monstrous things like stars, moving at the velocity of light, smashing into each other and making the geometry of space-time turn into some sort of washing machine.” 

“The tantalizing quest for gravity waves” (Arthur Fisher, April 1981) 

When scientists finally detect a form of energy they have never seen, they will open a new era in astronomy.

In the vast reaches of the cosmos, cataclysms are a commonplace: Something momentous is always happening. Perhaps the blazing death of an exhausted sun, or the collision of two black holes, or a warble deep inside a neutron star. Such an event spews out a torrent of radiation bearing huge amounts of energy. The energy rushes through space, blankets our solar system, sweeps through the Earth… and no one notices.

But there is a small band of experimenters, perhaps 20 groups worldwide, scattered from California to Canton, determined that some day they will notice. Pushed to the edge of contemporary technology and beyond, battling the apparent limits of natural law itself, they are developing what will be the most sensitive antennas ever built. And eventually, they are sure, they will detect these maddeningly intangible phenomena—gravity waves.

Even though gravity waves (more formally called gravitational radiation) have never been directly detected, virtually the entire scientific community is convinced they exist. This assurance stems, in part, from the bedrock on which gravity-wave notions are founded: Albert Einstein’s theory of general relativity, which, though still being tested, remains untoppled [PS, Dec. ‘79]. Says Caltech astrophysicist Kip Thorne, “I don’t know of any respectable expert in gravitational theory who has any doubt that gravity waves exist. The only way we could be mistaken would be if Einstein’s general relativity theory were wrong and if all the competing theories were also wrong, because they also predict gravity waves.”

In 1916, Einstein predicted that when matter accelerated in a suitable way, the moving mass would launch ripples in the invisible mesh of space-time, tugging momentarily at each point in the universal sea as they passed by. The ripples—gravity waves—would carry energy and travel at the speed of light. 

In many ways, this prediction was analogous to one made by James Clerk Maxwell, the brilliant British physicist who died in the year of Einstein’s birth—1879. Maxwell stated that the acceleration of an electric charge would produce electromagnetic radiation—a whole gamut of waves, including light, that would all travel at the same constant velocity. His ideas were ridiculed by many of his contemporaries. But a mere decade after his death, he was vindicated when Heinrich Hertz both generated and detected radio waves in the laboratory.

Why, then, more than 60 years after Einstein’s bold forecast, has no one seen a gravity wave? Why, despite incredible obstacles, are physicists still seeking them in a kind of modern quest for the Holy Grail, one of the most exciting in the whole history of science?

To find out, l visited experimenters who are building gravity-wave detectors and theoreticians whose esoteric calculations guide them. In the process, I learned about the problems, and how the attempts to solve them are already producing useful spinoffs. And I learned about the ultimate payoff if the quest is successful: a new and potent tool for penetrating, for the first time, what one physicist has called “the most overwhelming events in the universe.”

A kiss blown across the Pacific

The fundamental problem in gravity-wave detection is that gravity as a force is feeble in the extreme, some 40 orders of magnitude weaker than the electromagnetic force. (That’s 10⁴⁰, or a 1 followed by 40 zeros.)

Partly for this reason, and partly because of other properties of gravity waves, they interact with matter very weakly, making their passage almost imperceptible. And unlike the dipole radiation of electromagnetism, gravitational radiation is quadrupole.

If a gravity wave generated, for example, by a supernova in our galaxy passed through the page you are now reading, the quadrupole effect would first make the length expand and the width contract (or vice versa), and then the reverse. But the amount of energy deposited in the page would be so infinitesimal that the change in dimension would be less than the diameter of a proton. Trying to detect a gravity wave, then, is like standing in the surf at Big Sur and listening for a kiss blown across the Pacific. As for generating detectable waves on Earth, a la Hertz, theoreticians long ago dismissed the possibility. “Sure, you make gravity waves every time you wave your fist,” says Rainer Weiss, a professor of physics at MIT. “But anything you will ever be able to detect must be made by massive bodies moving very fast. That means events in space.”

Astrophysicists have worked up whole catalogs of such events, each associated with gravity waves of different energy, different characteristic frequencies, and different probabilities of occurrence. They include the supposed continuous background gravitational radiation of the “big bang” that began the universe [PS, Dec. ‘80], and periodic events like the regular pulses of radiation emitted by pulsars and binary systems consisting of superdense objects. And then there are the singular events: the births of black holes in globular clusters, galactic nuclei, and quasars; neutron-star quakes; and supernovas.

Probably the prime candidate for detection is what William Fairbank, professor of physics at Stanford University, calls “the most dramatic event in the history of the universe”—a supernova. As a star such as our sun ages, it converts parts of its mass into nuclear energy, perhaps one percent in five billion years. “The only reason a large star like the sun doesn’t collapse,” explains Fairbank, “is because the very high temperature in its core generates enough pressure to withstand gravitational forces. But as it cools from burning its fuel, the gravitational forces begin to overcome the electrical forces that keep its particles apart. It collapses faster and faster, and if it’s a supernova, the star’s outer shell blasts off. In the last thousandth of a second, it collapses to a neutron star, and if the original star exceeded three solar masses, maybe to a black hole.” One way of characterizing the energy of a gravity wave is the strain it induces in any matter it impinges on. If the mass has a dimension of a given length, then the strain equals the change in that length (produced by the gravity wave) divided by the length. Gravity waves have very, very tiny strains. A supernova occurring in our galaxy might produce a strain on Earth that would shrink or elongate a 100-cm-long detector only one one-hundredth the diameter of an atomic nucleus. (That is 10^-15 cm, and physicists would label the strain as 10^-17.) To the credit of tireless experimenters, there are detectors capable of sensing that iota of a minim of a scruple.

But there is a catch: Based on observations of other galaxies, a supernova can be expected to occur in the dense center of any given galaxy roughly about once in 30 years. That is a depressingly long interval. Over and over again, the scientists I spoke to despaired of doing meaningful work if it had to depend on such a rara avis. Professor David Douglass of the University of Rochester told me: “To build an experiment to detect an event once every 30 years—maybe—is not a very satisfying occupation. It’s hardly a very good Ph.D. project for a graduate assistant; it’s not even a good career project—you might be unlucky.”

Gravity waves: powerful astronomical tools?

What if we don’t confine ourselves to events in our own galaxy, but look farther afield? Instead of the “hopelessly rare” (in the words of one researcher) supernova in our galaxy, what if we looked for them in a really large arena—the Virgo cluster, which has some 2,500 galaxies, where supernovas ought to be popping from once every few days to once a month or so? That’s Catch-22². The Virgo cluster is about 1,000 times farther away than the center of our own galaxy. So a supernova event from the cluster would dispatch gravity waves whose effect on Earth would be some million times weaker (1,000 times 1,000, according to the inverse-square law governing all radiative energy). And that means building a detector a million times more sensitive. “There is no field of science,” says Ronald Drever of Caltech and the University of Glasgow, Scotland, “where such enormous increases in sensitivity are needed as they are here, in gravity-wave detection.” Trying to detect a supernova in a distant galaxy means having to measure a displacement one-millionth the size of an atomic nucleus.

Paradoxically, it is this very quality that gives gravity waves the ability to be, as Kip Thorne says, “a very powerful tool for astronomy. True, they go through a gravity-wave detector with impunity. But that means the gravity waves generated during the birth of a black hole can also get away through all the surrounding matter with impunity.” And neither light, nor gamma rays, nor radio waves can. During a supernova we can see the exploding shell via showers of electromagnetic radiation, but only hours or days after the initial massive implosion—the gravitational collapse. During the collapse, while a neutron star or black hole is being formed, nothing but gravity waves (and, theoretically, neutrinos) can escape.

“We’ve opened, at least partially, all the electromagnetic windows onto the universe,” says Thorne. “With gravity wave astronomy, we will open a unique new window onto fascinating, explosive events that cannot be well studied any other way—births and collisions of black holes, star quakes, collapses to neutron stars. This is the real bread and butter of modern high-energy astrophysics.”

But first, as the cookbooks say, you must catch your gravity wave. Until the 1950’s, no one presumed that the task was even feasible. Then Joseph Weber, a physicist at the University of Maryland, began to ponder the problem of building a gravity-wave detector, and proceeded to do so. It is no exaggeration to say that he fathered the entire field. By 1967, he and his assistants had built the first operating gravity-wave detector—a massive aluminum bar, isolated as well as possible from external vibrations and girdled by piezoelectric crystal sensors, which translated changes in the bar’s dimensions into electrical signals. Weber reported a number of events recorded on this and a twin detector at Argonne that he concluded were gravity waves [PS, May ‘72]. His report stimulated a host of other experimenters to build their own detectors. Designed by such investigators as J. A. Tyson at Bell Labs and David Douglass at Rochester, the detectors followed the same principles as Weber’s pioneering bar detector, but with greater sensitivity. These and subsequent experimenters were unable to confirm Weber’s findings; in fact, at the level Weber’s bar was capable of, theoreticians believe it was impossible to have detected gravity waves. “Either Joe Weber was wrong,” one told me, “or the whole universe is cockeyed.”

Today, three basic kinds of gravity-wave detectors are being developed. One is basically a Weber resonant-bar antenna, much refined; the second is the laser interferometer; and the third is a space-based system called Doppler tracking. Each has its advantages, and each its own devilish engineering problems.

Farthest along is the resonant bar, mostly because it has been in the works longest. The more massive such a bar is, the better (because it will respond to a gravity wave better). And its worth depends on the quality of resonating, or “ringing,” for a time after it has been struck by the wave. The longer it rings, the better an experimenter is able to pick out the effect of the wave. That quality is measured by the value called “Q”-the higher the Q, the better. For a while David Douglass and others, including Soviet scientists, have been seeking to make detectors out of such very-high-Q materials as sapphire-crystal balls. But Douglass, for one, has returned to aluminum. The reasons: New alloys of aluminum have been found with very high Q’s; sapphire can’t be fabricated in massive chunks (one of his detectors has a six-ton aluminum bar); and expense: “A 60-pound pure sapphire crystal,” he told me, “would cost about $50,000.”

Like virtually everyone else developing bar antennas, Douglass has abandoned room-temperature detectors and turned to cryogenic detectors, cooled down as close to absolute zero as possible. That includes groups at Perth, Australia, Tokyo, Moscow, Louisiana State University, Rome, Weber himself at the University of Maryland, and William Fairbank and colleagues at Stanford University.

Fairbank told me why the low-temperature route was essential: “At room temperature, the random thermal motion of the atoms in a bar is 300 times as big as the displacement we’re trying to detect. The only way to approach the sensitivities we’re after is to get rid of that thermal noise by cooling the bar.”

When I visited the Stanford campus, the detector’s five-ton aluminum bar was sealed inside its cryostat, a kind of oversized Thermos bottle. The whole assembly looked like something you could use if you wanted to freeze Frankenstein’s monster for a few centuries. And the environment was suitable, too: a vast, drafty, concrete building that could have been an abandoned zeppelin hangar.

This antenna, and others like it, is designed to respond to gravity waves with a frequency of about 1,000 Hz, characteristic of supernova radiation. Obviously the antenna must be isolated as far as possible from any external vibration at or around that frequency. This the Stanford group does by suspending the cylinder with special springs, consisting of alternating iron and rubber bars in what is called an isolation stack. “Otherwise, with our sensitivity,” Fairbank says, “this detector would make a dandy seismograph—just what we don’t want in California.” The Stanford suspension system attenuates outside noise by a factor of 10 °, enough so that you could drop a safe in its vicinity without disturbing the detector.

At LSU, William Hamilton, who is building an antenna very similar to Stanford’s (eventually it will become part of a Rome-Perth-Baton Rouge-Stanford axis looking for gravity-wave coincidences), takes another route toward seismic isolation. The very low temperature of the device allows him to levitate the bar magnetically; it is coated with a thin film of niobium-tin alloy, a material that becomes superconducting near absolute zero. If electromagnets are placed under the bar, the persistent currents running through its coating will interact with the magnetic field so that the bar literally floats in air.

Superconductivity is also the key to one of the most perplexing of all engineering problems: designing a transducer capable of sensing the tiny displacements of these antennas and converting them to a useful voltage that can be amplified and measured. “You can’t buy such things,” says David Douglass, “you have to make them, and go beyond the state of the art.” Both Douglass and Fairbank use superconducting devices whose elegant design makes them exquisitely sensitive—orders of magnitude more than the piezoelectric crystals originally used—although their approaches differ in details.

Superconducting devices may also one day—a day far in the future-allow gravity-wave astronomers to perform a feat of legerdemain called “quantum non-demolition.” To oversimplify, this means evading a fundamental limit for all resonant detectors, one that is imposed by the laws of quantum mechanics as the displacements become ever smaller. That problem will have to be faced if bar antennas are ever to be sensitive enough to detect gravity waves from supernovas in the Virgo cluster.

An alternative: laser interferometers

“One of the reasons we’re turning to laser detectors,” says Ronald Drever, “is to avoid the quantum-limit problem. Because we can make measurements over a much larger region of space, we effectively see a much larger signal. We don’t have to look for such minute changes as in a bar antenna.”

Laser interferometers bounce an argon-ion laser beam back and forth many times between two mirrors. (A generalized approach to the scheme appears in the drawing on page 92.) As a gravity wave ripples between the mirrors, the length of the light path changes, resulting in a change in the interference patterns that appear in photodetectors. Numbers of such detectors are in the planning and building stages, including ones at MIT, designed by Rainer Weiss, a pioneer in the field; at the Max Planck Institute of Astrophysics in Germany; at the University of Glasgow; and at Caltech.

“The one in Glasgow has 10-meter arms,” Drever told me, “and is working now. The one we’re working on at Caltech also has 10-meter arms, but will be stretched to 40 meters as soon as a building for it is ready. This will serve as a prototype for a much larger version—a kilometer to several kilometers long.”

Of course, laser interferometers have engineering problems, too, problems that become exacerbated as they grow larger. The laser beams must travel through vacuum pipes, and isolating pipes a kilometer long will not be simple. But Drever is convinced it can be done. “Maybe we’ll put it in a mine, or in the desert,” he says. This device may be ready by 1986, and has, Drever thinks, a chance of eventually detecting supernovas in the Virgo cluster.

One additional advantage of such laser detectors is that they are not restricted to a narrow frequency range, as are the resonant antennas, but would be sensitive to a broad band of frequencies from a few hertz to a few thousand hertz. They could therefore detect some massive black-hole events, which have lower frequencies than gravity waves from supernovas. To detect gravity waves with much lower frequencies, such as those from binary systems, you need very long baselines. “ln about 15 years,” says Rainer Weiss, “we will want big, space-based laser systems, using, say, a 10-kilometer frame in space. That way we could avoid all seismic noise.” 

The third kind of gravity-wave detector already exists in space, after a fashion. It has been used for spacecraft navigation for 20 years. It is called Doppler tracking, and is very simple—in theory. Here’s how ifs described by Richard Davies, program leader for space physics and astrophysics at Jet Propulsion Laboratory in Pasadena, Calif.: “You send a radio signal from Earth to a spacecraft, and a transponder aboard the craft sends the signal back to you. If a gravity wave passes through the solar system, it alters the distance between the two, and when you compare the frequency of the signal you sent out to the one you get back, you see that they are different—-the Doppler shift. However, the contribution of the gravity wave to this shift is minute compared to that of the spacecraft’s own velocity.

“We want to detect gravity waves with very low frequencies, maybe a thousandth of a hertz, using interplanetary spacecraft and the Deep Space Net that is used to track them. Such waves could be emitted from a collapsing system with a mass of a million to ten million suns, or from double stars that orbit each other in hours.”

A gravity-wave experiment had been planned for the International Solar Polar Mission. But, according to MIT’s Irwin Shapiro, who chaired the Committee on Gravitational Physics of the National Academy of Science’s Space Science Board, the experiment was dropped by NASA because of budget cuts.

Which of these methods will yield the first direct evidence of gravity waves? And when will that first contact come? No one really knows, and the gravity-wave seekers themselves are extremely diffident about making claims and predictions. But some time within the decade seems at least plausible.

ln the meantime, gravity-wave research is paying unexpected dividends. “It has opened up,” says Kip Thorne, “a modest new chapter in quantum electronics. Because it is pushing so hard against the bounds of modem technology, it is inventing new techniques that will have fallout elsewhere; for example, a new way to make laser frequencies more stable than ever. This will be useful in both physics and chemistry research.”

In the long run, however, the search for gravity waves is propelled by the basic drive of all scientists, and all mankind: to see a little farther, to understand a little more than we have ever done before.

Two indirect proofs for the existence of gravity waves 

The first evidence of any kind for the existence of gravity waves comes not from sensing them directly but from observing their effect on the behavior of a bizarre astronomical object called a binary pulsar. A pulsar, believed to be a rapidly spinning neutron star, emits strong radio signals in periodic beeps. But pulsar PSR 1913+16, discovered by a team of University of Massachusetts astronomers in 1974 with the world’s largest radio telescope (at Arecibo, P.R.), is unique. Its beeps decelerate and accelerate in a regular sequence lasting about eight hours. From this, the astronomers, led by Joseph Taylor, deduced that the pulsar was rapidly orbiting around another very massive object—perhaps another neutron star.

Einstein’s theory of general relativity predicts that this binary system should produce a considerable quantity of gravity waves, and that the energy radiated should be slowly extracted from the orbit of the system, gradually decreasing its period as the superdense stars spiral closer to one another. Einstein’s equations predict a decrease of one ten-thousandth of a second per year for a pulsar like PSR 1913+ 16. And after four years of observations Taylor’s team announced, in late 1978, that ultraprecise measurements of the radio signals gave a value almost exactly that amount. The closeness of the match not only provides good—even though indirect—evidence of the existence of gravity waves, but also further bolsters Einstein’s theory of gravity against some competing theories.

As Taylor said of what he called “an accidental discovery originally,” the astronomers had an ideal situation for testing the relativity theory—a moving clock (the pulsar) with a very precise rate of ticking and a high velocity—some 300 kilometers per second. “lt’s almost as if we had designed the system ourselves and put it out there just to do this measurement.” 

Another indirect indication that gravity waves do indeed exist came more recently, and more dramatically. It stemmed from an event that still has astronomers reeling. At exactly 15 hours, 52 minutes, five seconds, Greenwich time on March 5, 1979, a gamma-ray burst of unparalleled Intensity flashed through our solar system from somewhere in space. It triggered monstrous blips on detectors aboard a motley collection of nine different spacecraft throughout the solar system, which form, in effect, an international network maintained by the U.S., France, West Germany, and the Soviet Union.

Once-in-a-lifetime event

“This March 5 gamma-ray event was extraordinary,” says Thomas Cline of NASA Goddard Space Flight Center, who, with his colleague Reuven Ramaty and other U.S., French, and Russian astrophysicists, has been analyzing it ever since. “It was not like the gamma-ray bursts that have been seen a hundred times in the last decade. It’s a first and only, like something that’s seen once in a scientific lifetime.”

Because the surge of gamma rays was detected by so many satellites separated in space, astronomers were able to triangulate the position of its source and identify it with a visible object—the first time for such a feat. The object was a supernova remnant dubbed N49 in the Large Magellanic Cloud (LMC), a neighboring galaxy roughly 150,000 light-years away.

Ramaty, Cline, and colleagues posit that the genesis of the gamma-ray burst was a quivering neutron star—the ultradense, ultracompact object that many theorists believe is left over from a supernova explosion. “We believe,” Cline told me, “that a neutron star can undergo a transformation analogous to an avalanche. Snow falls on a mountain until there’s a slide. 

Similarly, dust and other material collect on a neutron star until it can’t stand being as heavy as it is. Then there’s a star quake, either in the crust or in the core. and the star shakes itself at a frequency of about 3,000 Hz, a note you could hear if you were listening to it in an atmosphere. The surface of the star-only five to 10 miles in diameter—is heaving up and down several feet, thousands of times a second. Its magnetosphere is shaken, and that’s what produces, indirectly, the gamma rays. But that’s secondary, in our model, to the gravitational waves caused by the oscillation of the neutron star.

“Could we detect these? The answer is no. After all, this is only a kind of after-gurgle, thousands of years after the star’s original collapse—the supernova. It’s like a tremor after a major earthquake, maybe only one percent as big.”

Nevertheless, Cline called all the U.S. gravity-wave experimenters who could have been “on-line” during the gamma ray burst to learn whether they had seen anything. Of them all, only Joseph Weber had an antenna working that March day, and he had observed nothing.

The gamma-ray detectors aboard the satellites were not capable of sensing the 3,000-Hz frequency predicted by the starquake model. If they had, says Cline, it would have been “a very direct link” to the existence of gravitational radiation.

But the star-quake model makes another prediction: The gravity waves generated should carry off an enormous amount of energy, far more than that in the gamma rays, and thus snuff out the star’s vibration very quickly. “The nice thing,” says Goddard’s Reuven Ramaty, “is that the damping time predicted for gravity waves in this event exactly corresponds to what we observed: The main part of the burst lasted just ‘I5 hundredths of a second, and that’s what we calculate from our model. So we now have for the second time indirect evidence of the existence of gravity waves. But both have problems, as do all indirect checks. They won’t replace direct evidence.”

Some text has been edited to match contemporary standards and style.