DYNAMICAL BEHAVIOR OF THE MULTIBONDIC AND MULTICANONIC ALGORITHM IN THE 3D Q-STATE POTTS-MODEL

We investigate the dynamical behavior of the recently proposed multibo ndic cluster Monte Carlo algorithm in applications to the three-dimens ional q-state Potts models with q = 3, 4, and 5 in the vicinity of the ir first-order phase transition points. For comparison we also report simulations with the standard multicanonical algorithm. Similar to the findings in two dimensions, we how that for the multibondic cluster a lgorithm the dependence of the autocorrelation time tau on the system size Vis well described by the power law tau proportional to V-alpha, and that the dynamical exponent a is consistent with the optimal rando m walk estimate alpha = 1. For the multicanonical simulations we obtai n, as expected, a larger value of alpha approximate to 1.2.