Now Choptuik and Frans Pretorius of Princeton University have simulated such collisions, including all the extremely complex mathematical details from general relativity. For simplicity and to make the simulations generic, they modeled the two particles as hypothetical objects known as boson stars, which are similar to models that describe stars as spheres of fluid. Using hundreds of computers, Choptuik and Pretorius calculated the gravitational interactions between the colliding particles and found that a black hole does form if the two particles collide with a total energy of about one-third of the Planck energy, slightly lower than the energy predicted by hoop conjecture, as they report in a paper in press at Physical Review Letters.
Does that mean the LHC will make black holes? Not necessarily, Choptuik says. The Planck energy is a quintillion times higher than the LHC's maximum. So the only way the LHC might make black holes is if, instead of being three dimensional, space actually has more dimensions that are curled into little loops too small to be detected except in a high-energy particle collision. Predicted by certain theories, those extra dimensions might effectively lower the Planck energy by a huge factor. "I would be extremely surprised if there were a positive detection of black-hole formation at the accelerator," Choptuik says. Physicists say that such black hole would harmlessly decay into ordinary particles.
I'm guessing that this is the preprint of the paper in press.