Static and dynamic properties of short range Ising spin glasses

Abstract

Let us now draw some conclusions for the two-dimensional Ising model with short range random interactions and a relaxational dynamics: - The model describes experiments qualitatively - It has no static transition - Nevertheless in the field H, temperature T and observation time t diagram, Fig. 6, there is a rather well defined surface below which spin glass behaviour is observed. This surface is rather singular, since one finds Tf(H=0,t)∼(lnt)−1/2 and Tf(H,t)−Tf(0,t) ∼H2/3 - Below Tf(H,t) the spins freeze into small completely frozen clusters, the rest seems to remain in thermal equilibrium - The freezing process can be described by a dynamics of small decoupled clusters - The model even reproduces recent experiments which favour a static phase transition - Only at T=0 one has a phase transition with scaling laws - Also experimental data are not inconsistent with T=0 scaling - The low lying metastable states do not have the structure of the mean field states