We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cond-mat

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Condensed Matter > Statistical Mechanics

Title: How a spin-glass remembers. Memory and rejuvenation from intermittency data: an analysis of temperature shifts

Abstract: The memory and rejuvenation aspects of intermittent heat transport are explored theoretically and by numerical simulation for Ising spin glasses with short-ranged interactions. The theoretical part develops a picture of non-equilibrium glassy dynamics recently introduced by the authors. Invoking the concept of marginal stability, this theory links irreversible `intermittent' events, or `quakes' to thermal fluctuations of record magnitude. The pivotal idea is that the largest energy barrier $b(t_w,T)$ surmounted prior to $t_w$ by thermal fluctuations at temperature $T$ determines the rate $r_q \propto 1/t_w$ of the intermittent events occurring near $t_w$. The idea leads to a rate of intermittent events after a negative temperature shift given by $r_q \propto 1/t_w^{eff}$, where the `effective age' $t_w^{eff} \geq t_w$ has an algebraic dependence on $t_w$, whose exponent contains the temperatures before and after the shift. The analytical expression is verified by numerical simulations. Marginal stability suggests that a positive temperature shift $T \to T'$ could erase the memory of the barrier $b(t_w,T)$. The simulations show that the barrier $b(t_w,T') \geq b(t_w,T)$ controls the intermittent dynamics, whose rate is hence $r_q \propto 1/t_w$.
Additional `rejuvenation' effects are also identified in the intermittency data for shifts of both signs.
Comments: Revised introduction and discussion. Final version to appear in Journal of Statistical Mechanics: Theory and Experiment
Subjects: Statistical Mechanics (cond-mat.stat-mech); Disordered Systems and Neural Networks (cond-mat.dis-nn); Soft Condensed Matter (cond-mat.soft)
Journal reference: JSTAT(2004)P10013
DOI: 10.1088/1742-5468/2004/10/P10013
Cite as: arXiv:cond-mat/0406614 [cond-mat.stat-mech]
  (or arXiv:cond-mat/0406614v2 [cond-mat.stat-mech] for this version)

Submission history

From: Paolo Sibani [view email]
[v1] Thu, 24 Jun 2004 18:14:12 GMT (42kb)
[v2] Tue, 26 Oct 2004 09:42:50 GMT (46kb)

Link back to: arXiv, form interface, contact.