While at Swinburne University, I looked into the problem as to whether the two galaxies known as "Arp 271" would merge or not after colliding (because no one knows - for example Wikipedia, the ESO and even this site wonder about it: http://apod.nasa.gov/apod/ap130825.html). It takes a supercomputer to do the simulation. Here's a poster of the results, below which you can find my paper...