Finite-size-scaling analysis of subsystem data in the dilute Ising model.

Abstract

Monte Carlo simulation results for the magnetization of subsystems of finite lattices are used to determine the critical temperature and a critical exponent of the simple-cubic Ising model with quenched site dilution, at a concentration of [ital p]=40%. Particular attention is paid to the effect of the finite size of the systems from which the subsystem results are obtained. This finiteness of the lattices involved is shown to be a source of large deviations of critical temperatures and exponents estimated from subsystem data from their values in the thermodynamic limit. By the use of different lattice sizes, the results [ital T][sub [ital c]](40%)=1.209[plus minus]0.002 and [nu](40%)=0.78[plus minus]0.01 could be extrapolated.