Research Objectives
Theory predicts that at temperatures of order 150 MeV there will be a phase transition in which protons and neutrons effectively "melt" into their constituents. Temperatures this high probably occurred in the very early universe and will be created in collisions of heavy ions at the Relativistic Heavy Ion Collider (RHIC) under construction at Brookhaven National Laboratory. This transition is expected to be a second-order transition in the same universality class as the O4 spin model. However, recent studies by the Japanese Lattice QCD (JLQCD), Bieleld, and MIMD Lattice Computation (MILC) collaborations have failed to find the expected O4 scaling behavior. This leaves open the question of the order of the phase transition -- perhaps it is really a first-order transition at small-enough quark masses. Our research objective is to check this possibility to a higher accuracy than previous tests.
Computational Approach
We use the standard "refreshed molecular dynamics" method to generate sample configurations for the QCD gluon fields at high temperature. Thermodynamic quantities such as the order parameter and free energy are monitored as a function of a fictitious "simulation time". Although this simulation time is not the same as real time, behaviour of the system as a function of simulation time can be used as an indicator of the type of transition. The code is the MILC collaboration's QCD code, which runs on a variety of MIMD machines, in this case on the NERSC T3E.
Accomplishments
We have run Monte Carlo simulations of the system on a larger spatial volume (L=24) than previous tests, using starting configurations in the hot and cold phase. This is a standard technique for testing for first-order transitions. If the transition is first order, and you have adjusted the temperature near enough to the phase transition, large systems will be metastable in the phase where they start. In contrast, for higher-order transitions, observables in the two time histories will evolve together.
For two values of the quark mass, we found hot and cold starts evolving to the same final values, so this simulation favors a second-order transition. Previous studies have found that results are strongly dependent on the spatial size of the system being simulated. These simulations use a spatial size 1.5 times as large as previous tests, as well as a quark mass which is slightly closer to the real-world value. These results will be analyzed together with a set of runs spanning a range of temperatures to see if the expected scaling forms are seen with these quark masses.
Publications
Bernard, C., T. Blum, T. A. DeGrand, C. DeTar, S. Gottlieb, U. M. Heller, J. Hetrick, L. Karkkainnen, K. Rummukainen, R. Sugar, D. Toussaint, and M. Wingate. 1997. MILC studies of high-temperature QCD -- A progress report. QCD on Parallel Machines Workshop, Tsukuba, Japan.
Bernard, C., T. Blum, C. E. DeTar, S. Gottlieb, U. M. Heller, J. E. Hetrick, B. Jegerlehner, K. Rummukainen, R. L. Sugar, D. Toussaint, and M. Wingate. 1997. Critical behavior at the chiral phase transition. Lattice-97 Conference, Edinburgh.
URL
http://www.physics.arizona.edu/~doug
Time histories of the order parameter for hot and cold starts with a quark mass of 0.05*T
and 0.032*T, with two separate starts for the larger quark mass. The temperature in these
runs has been adjusted to be right at the peaks in the susceptibilities.