On Wednesday, November 8th Ed Seidel gave a talk to Physics Society members on Relativity and black holes. He works in the NCSA Relativity Group, which brings together computer scientists, theoretical physcists (and everything in between) to solve problems like that of two colliding black holes with supercomputing.
Prof. Seidel began with an introduction to Einstein's Special Relativity. From two fairly simple assumptions (physics is the same in any inertial reference frame, and the speed of light is constant and the same in any inertial frame) one can deduce some fairly bizarre results. Objects travelling at high rates of speed will be seen by "stationary" observers to be shorter and moving slower in time.
General Relativity involves something called the Equivalence Principle. This states that accelerating reference frames and reference frames in a gravitational field are the same. This allows gravitation to be described with curved (non-Euclidean) geometry, via math Gauss had worked out but hadn't been applied to physical systems before Einstein. General Relativity states that light should follow a curved path in this geometry, a prediction that was exactly confirmed by experiment, making Einstein the hero he is today. :-) Light from a distant star was shown to be bent by the sun's gravitation, if only by a small amount.
Prof. Seidel also talked about gravitational waves, the gravitational analog to electromagnetic waves. General Relativity predicts that such waves should exist, though no such evidence exists yet. Laser interferometric studies may prove their existence by the turn of the century. They are considered to be very weak, and so major events like supernovae and binary collisions are needed to produce waves strong enough to be detected.
Seidel's work with the Relativity group involves studying what happens when two black holes collide. Black holes are stars so massive/dense that gravity overpowers internucleic repulsion and the stars collapse down to a point -- a singularity where there is severe curvature of spacetime and lots of "bad things" happen. Within the event horizon (a theoretical surface around the black hole) nothing can escape -- the potential energy is so significant that even light cannot escape. Black holes, and their event horizons, can be perturbed though -- by adding a mass or any number of similar actions. The black hole essentially acts like a damped harmonic oscillator (yes, that model really is used for something...) and emits gravitational waves until it settles down to a spherical event horizon again. Prof. Seidel showed several video clips illustrating how these waves propagate in space, and how the black hole responds to perturbations.
The computer graphics were all done in 2-D because of the incredible amount of computing power it requires to solve the equations from general relativity -- it is not an activity for those with pencil and paper. In order to solve the 3-D problem of two colliding black holes, an "old" Cray supercomputer would require about 100,000 hours to solve the problem. Advances in computing speed and the development of faster and cheaper (?) parallel-processing computers have begun to make such calculations more feasible.
Although we actually haven't "seen" a black hole, we do have significant evidence confirming their existence. It now seems possible that black holes can exist at the centers of galaxies. Data collected by the Hubble Space Telescope allowed the estimation of a galaxy's core size (dimensions) and mass, and given what we know (and theorize) about astrophysics it would seem that only a black hole could explain so great a mass in such a small space.