Testing Relativity with Neutron Star Mergers This article is the third in a series on the nine NASA HPCC Earth and Space Science (ESS) Project Science Team II Grand Challenge Investigations. Above: Paul Saylor (left) and Doug Swesty discuss the simulated merger of two 1.4 solar mass neutron stars. by Jarrett Cohen	Issue 3, September 1997 Welcome to the on-line version of NASA's Insights Newsletter. Insights was published by the High Performance Computing and Communications (HPCC) Program Office. Program Manager: Dr. William Feiereisen Editor, Photography, Design: Ms. Judy Conlon WWW Design and Conversion: Mr. Lawrence Picha
In the three snapshots above, the lighter color signifies higher matter density and darker color, lower density. Embodying some gravitational effects, simulations show that gravitational radiation losses influence instabilities during the merger. Only full relativistic calculations, now underway, can predict the gravitational wave signal and probe the possibility of black holes resulting.
"The Einstein field equations are an elegant creation like the score of a Mozart concerto waiting to be played," said Paul Saylor, professor of computer science at the University of Illinois at Urbana-Champaign (UIUC). With terms numbering in the thousands, these equations make up the general theory of relativity, which states that space and time are dynamic. Saylor's orchestra of astrophysicists, relativity experts and computer scientists are testing the theory's limits by simulating neutron star mergers. Paul Saylor, professor of computer science, University of Illinois at Urbana-Champagn. Neutron stars result from supernova explosions: the core of an old, bloated star collapses and violently spits out the envelope. Left behind is a city-sized atomic nucleus with the mass of our sun. "The atoms are squeezed so closely together that their boundaries disappear; even the nuclei merge," said Jim Lattimer, professor of physics and astronomy at the State University of New York at Stony Brook. He related that a liter of neutron star matter would weigh as much as Halley's Comet. Neutron star coalescence occurs in binary systems where orbiting stars have undergone supernovas. Affirmed by observations, "Einstein's theory of gravity predicts that two objects in orbit around each other will emit gravitational radiation," said Doug Swesty, research scientist at UIUC's National Center for Supercomputing Applications (NCSA). The radiation causes the orbit to decay. Jim Lattimer, professor of physics and astronomy, State Univeristy of New York. After about one billion years, an ever-accelerating inspiral draws the stars together, "like two gigantic sumo wrestlers circling each other," Saylor said. Only this match occurs at near-light speeds. In the final round, the orbital period is about one millisecond, and "these massive objects . . . are orbiting around each other at 60,000 revolutions-per-minute!" Swesty exclaimed. The merger itself takes only one to two milliseconds. The velocities involved make neutron star mergers relativistic, and only general relativity can describe gravitational radiation and its effects. This radiation is essentially gravitational waves, or ripples of space-time, that approximate cosmic Jell-O, quipped Wai-Mo Suen. "If you punch Jell-O, it will vibrate," said the associate professor of physics at Washington University in St. Louis. "At the end, the neutron stars collide in a big punch and send out a burst of gravitational waves." "If you punch Jell-O, it will vibrate. At the end, the neutron stars collide in a big punch and send out a burst of gravitational waves." --Wai-Mo Suen Washington University Updating Newton Most computational approaches have at best made some corrections to Newtonian gravity, now over 300 years old. Saylor's team is one of the few doing a "full relativistic calculation," Swesty said.	Their "Cactus" general relativity code builds on four years of simulating the more massive black holes by investigators at UIUC, Washington University and the Max Planck Institute for Gravitational Physics in Potsdam, Germany. Max Planck research scientists Joan Massó and Paul Walker were Cactus's main developers. Since black holes exist in a vacuum, Cactus is coupled to a new hydrodynamics code depicting changes in matter. Obtaining fast performance on microprocessor-based supercomputers such as the CRAY T3E requires high utilization of each processor's memory cache. Keeping the operands in cache as much as possible before doing a load from main memory will result in a higher percentage of peak speed, team members said. As of this writing, additional efficiency improvements brought the coupled code to approximately 40 billion floating-point operations per second on the High Performance Computing and Communications' 512-processor CRAY T3E at Goddard Space Flight Center. "These are all violent events that can't be studied fully unless relativity is taken into account." -Ed Seidel, Max Planck and NCSA Users must be able to take advantage of computing cycles whenever they are available. Accordingly, the Saylor group is using message passing for communication and synchronization among processors so that the code can run on "every major scalable architecture produced in North America," Swesty said. Distributing a highly optimized code will allow others to make their own enhancements to model the formation of black holes, supernovas, white dwarf stars and other phenomena, said Ed Seidel, senior research scientist at Max Planck and NCSA. "These are all violent events that can't be studied fully unless relativity is taken into account." Gazing through gravitational waves For neutron star coalescence, such simulations intertwined with contemporary observations will begin answering questions about the dense stars' properties and merger outcomes. Neutron stars are primarily detected at radio or X-ray frequencies, with three binary systems known to exist in our galaxy, but the most telling and direct signature is gravitational waves. The central mystery gravitational waves can unravel is neutron star matter's equation of state, which relates pressure to density and temperature, Lattimer said. It is challenging to understand the equation of state because scientists cannot yet replicate the density conditions on Earth. Each equation of state may produce a unique gravitational wave signal. In the models "we vary the properties of matter at high densities to see if the emitted signal could constrain the matter's behavior," Swesty said. A strong function of the equation of state is the maximum mass a neutron star can have; the estimated range is 1.4 to 3.2 solar masses. Swesty explained that mergers have three likely outcomes: a larger neutron star, a black hole or an object that will undergo a delayed collapse to a black hole. A relativistic simulation generating anything but the first option would point to an upper mass limit. The most ambitious effort to measure the predicted gravitational waves is the Laser Interferometric Gravitational Observatory (LIGO), coming on-line in 2000. [See the LIGO web site] It may locate as many as 300 neutron star mergers per year, but "a single observation would be phenomenal," Swesty said. "LIGO, long prior to coalescense, will detect masses of the individual objects. That, combined with what we learn from the coalescense signal in the models, could really help us to pin down the structure of neutron stars." With the additional potential of using gravitational wave strength to mark the universe's expansion rate, "this is an auspicious era for astronomical research," Saylor said. "It is like when Galileo looked through the universe with his telescope. We will be looking at the universe through the lens of gravitational waves." Other co-investigators include Steven Ashby, Lawrence Livermore National Laboratory; Ian Foster, Argonne National Laboratory; Michael Norman, NCSA/UIUC; and Clifford Will, Washington University. For more information about the project, visit their web site.

[Return to Top]