Monte Carlo simulations on the SK model have been done to investigate aging processes after a rapid quench from to the spin-glass phase. Taking care of time ranges of simulation, we examine time evolutions of energy of the system, Parisi's overlap distribution function, auto-correlation and clones-correlation functions, distribution functions of the two correlations, and magnetization induced by the field applied after a certain waiting time. The data simulated exhibit a rich variety of aging phenomena. Most of them can be interpreted in a unified way, though qualitatively, by the scenario of growth of quasi-equilibrium domains which we have recently introduced. The results are consistent qualitatively with asymptotic behaviours of some of the basic assumptions and their results in recent analytical theory on the same SK model, so long as the limiting procedures in finite systems are taken properly. Also they suggest that a basin of attraction of one dominant pure state spans almost an entire phase space of the system with a common time-reversal symmetry.
Aging phenomena of short-range Ising spin glass models have been investigated using Monte Carlo simulations. It is found that in the low-temperature spin-glass phase the mean domain size exhibits a crossover from a power-law growth associated with the critical fluctuation at the transition temperature to slower growth inherent in the low-temperature phase. The temperature dependence of the growth law of the domain size can be almost explained by this crossover. We also find that the spin-autocorrelation function in the quasi-equilibrium regime follows the expected scaling behavior from the droplet theory expressed in terms of the mean domain size and a characteristic length scale of droplet excitations. Spin glasses (SG) exhibit characteristic slow dynamics below the SG transition temperature. 1)-3) Recently such slow dynamics, in particular aging phenomena, has been much attractive both in experimental and theoretical studies. Several attempts have been made so far to explain the slow dynamics. There are mainly two distinct
We present a numerical study of the nonequilibrium dynamics of the Sherrington-Kirkpatrick model. We analyze the behavior of the system for different system sizes N and time scales t, in order to identify the correct out-of-equilibrium regime. We find evidence of aging behavior in the dynamics of the system, compatible with a weak-ergodicity breaking scenario. In addition, we obtain a pictorial characterization of the out-of-equilibrium phase space that presents some analogies with the hierarchy of states found in equilibrium.
By means of Monte Carlo simulations on the three-dimensional Ising spin-glass model, we have studied aging phenomena with various temperature(
T)-change protocols. Particularly, a
T-shift protocol, in which a system is first quenched to and aged for a period
tw1 at a temperature
T1, and subsequently aged at a new temperature
T2 is closely investigated. Most importantly, the mean size of domains, which is extracted from the replica-overlap function, was found to grow monotonically without any appreciable decrease by the
T-change. We also found the relaxation of energy density and spin auto-correlation function during the
T-shift process can be explained within the following picture: the dynamics finally crossovers to an isothermal aging of
T2 at around
t2≃tw1(eff) after the
T-change, where
tw1(eff) is an effective waiting time related with
tw1, but in the transient regime
t2<tw1(eff), adjustment of the population of thermally active droplets from that at
T1 to
T2 takes place very slowly. Implications of the results on aging phenomena in other
T-change protocols such as
T-cycling and continuous
T-change with an intermittent stop observed by simulations as well as experiments are discussed.
Using Sherrington-Kirkpatric （SK） model we investigated the nonequilibrum properties of spin glass by dynamical Monte Carlo simulation. Our results revealed that the susceptibility in the weak field shows a cusp near the glass transition, while the specific heat shows a more broad transition. The aging and zero-field colding memory effect is also reproduced in our simulation, and the phenomena can be explained by energy relaxation. Our simulation also reveals that homogeneous SK model can not reproduce the exchange bias effect observed in the experiment.