After some early numerical experiments (particle-mesh code with 100,000
particles) on counter-rotating cores in elliptical galaxies, our brave bud
discovered warped galaxies. The attraction was obvious and instantaneous,
and we're not talking gravitation here. During this time Greg developed
a new type of astrophysical hydrodynamics code: a 2.5-dimensional
Lagrangian method, or, more precisely, a method for solving the Navier-Stokes
Research is the best place to be: you work your buns off, and if it works you're a hero; if it doesn't, well -- nobody else has done it yet either, so you're still a valiant nerd.
The code was good for a small amount of amusement however. At one point
its author made an error in the value of the gravitational constant (by a
factor of 1000); the galaxy sucked itself inside out in a single integration
step. Oops. There aren't many people who can honestly say they've
destroyed an entire galaxy single-handedly, but Greg Roelofs is one
of the few who can.
So the FLOPPY DISC code was scrapped, but Greg's warped fascination remained. On the basis of 400,000-particle numerical experiments by his advisor, Dr. Richard H. Miller, and Miller's NASA/Ames collaborator, Dr. Bruce F. Smith, Greg turned to the question of whether a planar galactic disk embedded in an oscillating background potential could "fake" a warp. Without diving into the gory details, suffice it to say that the velocity map (Doppler data from radio observations of neutral hydrogen) of a warped galaxy without radial oscillations looks a great deal like the corresponding map for an unwarped galaxy that is oscillating radially. Greg's dissertation, entitled Radial Motions in Spiral Galaxies (1.4 MB PDF), considers whether so-called warped spirals may actually be flat after all (answer: probably not) and whether one could detect this if it were true. He defended his dissertation on 27 July (successfully, by golly!) and graduated on 15 December 1995. That's Dr. Greg to you, pal.