Neighborhood Identity Ordering and Quenched to Annealed Transition in Random Bond Models

Systems with many different interactions pose a challenge to traditionalmethods of statistical physics. In this paper, we develop the random bond model, which has a huge number of randomly chosen interaction parameters (quenched variables). Using heuristic arguments and Monte-Carlo simulations, we show that for any temperature there exists a sufficiently large system size above which one can forego the complicated quenched averaging familiar from spin glasses, and calculate statistical averages using standard methods of equilibrium statistical mechanics.