Statistical Physics in Random Media

Harvard University, Cambridge MA

Investigators

Abstract

0229243 Fisher This award supports theoretical research on the statistical physics of random media. Quenched randomness drastically alters the properties of many systems, especially their dynamics. A variety of systems in which many degrees of freedom interact in random media will be studied focusing on long time scale universal aspects of their behavior. One class of such system can undergo equilibrium phase transitions but they evolve extremely slowly towards equilibrium because of the necessity of overcoming large free energy barriers. For these, in particular spin glasses, both the nature of the non-equilibrium configurations that dominate observations and the equilibrium states towards which they strive are of interest. Many fundamental questions remain unanswered. A second class of random systems are those in which the dynamics and equilibrium properties are inextricably linked because of quantum mechanics. The destruction of superconductivity by quantum fluctuations will be investigated, focusing on the combined effects of randomness, quantum mechanics and dissipation. A third class, driven systems which are intrinsically far from equilibrium, can undergo novel types of non-equilibrium transitions, such as depinning from a stationary to a sliding "phase." Progress towards a better understanding of such transitions will be used to investigate the dynamics and statistics of earthquakes both on single faults and on interacting faults. Related issues in laboratory scale fracture will also be studied. %%% This theoretical research will study a variety of random systems ranging from magnetic glasses to earthquakes. The research addresses fundamental issues. The results of these studies may have broader applications to real materials. ***