Press Release, 8/27/99
Work at a new research institute using supercomputers in Germany and the USA has provided the first glimpse of what might happen when a small black hole tries to slide past a bigger one but gets swallowed instead.
At the Albert Einstein Institute (Max Planck Institute for Gravitational Physics) in Golm near Potsdam, Germany, which opened in 1995, an international team of physicists has assembled a powerful computer laboratory to study black holes. Among the computers in Golm is a powerful Origin 2000 supercomputer built by SGI Computers in Mountain View, California. Using 32 separate computer processors working in parallel, it can perform 3 billion computations per second.
In June, the team virtually owned a much larger computer, a 256-processor SGI Origin 2000 reserved for special "capability computing" simulations, for more than two weeks. Operated at the National Center for Supercomputing Applications (NCSA) at the University of Illinois, it is one of the largest civilian supercomputers in the USA. Capability computing refers to superjobs that require dedicated use of a high-performance computing system for days or weeks. By the time the team, which also includes researchers at NCSA, Washington University in St. Louis, USA, and the Konrad-Zuse-Zentrum in Berlin, had finished their simulations, they had output nearly a trillion bytes (terabyte) of data and logged an astonishing 140,000 CPU-hours on the Origin2000. That's equivalent to more than 16 years on a single-processor machine. Never before has any research team exceeded 100,000 CPU-hours in one month on an NCSA Origin system.
One of the team, Dr B Brügmann, had demonstrated the feasibility of studying black-hole collisions with the Origin 2000 at the Albert-Einstein-Institut by simulating the merging of two holes in a grazing collision. Using a new method that he and his colleague Dr S Brandt developed for calculating the starting conditions of a black-hole collision, he studied what happens when one black hole moves very close to another twice its size. Like two liquid drops, they quickly merge together into a single black hole, sending off ripples of pure gravity that physicists call gravitational waves. Many research groups, particularly in the USA, are working to perform simulations like this, but Brügmann's simulations were the first to show a collision which is not head on, making it the first fully three-dimensional simulation of a black hole collision. The huge runs performed at NCSA in June now bring the general black hole collision problem into the reach of quantitative science.
Besides their intrinsic fascination with black holes, the group is motivated by the need to understand the observations of black holes that may soon be made by detectors designed to register gravitational waves. As the group's leader, Prof E Seidel, says: "The enormous investment in huge gravitational wave detectors in several countries, including the 1.2 km-long GEO detector near Hannover, may only pay off if we can provide the observers with reliable ways of recognizing the waves produced by black holes." Seidel and several collaborators moved to Germany three years ago from the USA, where they had performed the most accurate simulations of head-on collisions to date. Grazing encounters of the type calculated now in Potsdam are thought to be much more common in astronomy than directly head-on collisions.
In order to make sense of the vast, trillion bytes of ouput produced by the simulations, the group worked with visualization software developed at the Konrad-Zuse-Zentrum in Berlin, customized for the black hole work by Werner Benger. This software allows one to create visual images of the data, which can then be assembled as movies to enable the scientists to see the black hole collisions they have simulated. Working for three weeks on a special graphics supercomputer provided by NCSA and SGI, Benger was able to create stunning visualizations of the process of black hole collisions, showing how they merge, and in the process, create bursts of gravitational waves that may soon be detected by gravitational wave detectors like GEO in Hannover.
With the advancement of such supercomputer simulations, scientists now stand on the threshold of a new kind of experimental physics. Prof B Schutz, a director of the AEI, says: "Astronomers now tell us that they know the locations of many thousands of black holes, but we can't do any experiments with them on Earth. The only way we will learn the details is to build numerical substitutes for them inside our computers and watch what they do. I believe that studying black holes will be a key theme of astronomy in the first decade of the next century."
A vizualization of two colliding black holes. Shown are the so-called
apparent horizons: whatever falls in, even light, cannot escape again.
Snapshots of two colliding black holes, now also shown with
gravitational
waves generated by the merger process, displayed in shades
from red to
white, that would propagate towards an observer on earth.
