2005
DOI: 10.1103/physrevb.72.054417
|View full text |Cite
|
Sign up to set email alerts
|

Rejuvenation and memory in model spin glasses in three and four dimensions

Abstract: We numerically study aging for the Edwards-Anderson model in three and four dimensions using different temperature-change protocols. In D = 3, time scales a thousand times larger than in previous work are reached with the Spin Update Engine ͑SUE͒ machine. Deviations from cumulative aging are observed in the nonmonotonic time behavior of the coherence length. Memory and rejuvenation effects are found in a temperature-cycle protocol, revealed by vanishing effective waiting times. Similar effects are reported for… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

2
37
0

Year Published

2005
2005
2014
2014

Publication Types

Select...
7
1
1

Relationship

0
9

Authors

Journals

citations
Cited by 25 publications
(39 citation statements)
references
References 49 publications
2
37
0
Order By: Relevance
“…The apparent existence of a single length scale that grows algebraically was confirmed by a recent numerical work [6], where it was furthermore claimed that the response function is well described by local scale invariance [7]. In spite of this, the correlation function showed systematic deviations from a simple t/t w scaling [6] (although simple aging seems to work well in d = 3 [8]). In [6], the autocorrelation was then compared to the scaling form C(t, t w ) ∼ t −xc (t/t w ), which usually holds at a critical point with x > 0 [9].…”
mentioning
confidence: 60%
“…The apparent existence of a single length scale that grows algebraically was confirmed by a recent numerical work [6], where it was furthermore claimed that the response function is well described by local scale invariance [7]. In spite of this, the correlation function showed systematic deviations from a simple t/t w scaling [6] (although simple aging seems to work well in d = 3 [8]). In [6], the autocorrelation was then compared to the scaling form C(t, t w ) ∼ t −xc (t/t w ), which usually holds at a critical point with x > 0 [9].…”
mentioning
confidence: 60%
“…On the other hand, the dimensionality of the spin glass is determined by the dynamical correlation length. Values of the correlation length SG N SG n 1=3 imp (n imp the density of spins) extracted from field change experiments for various spin glasses [29] and extrapolation from recent numerical simulations [30] are of the order of N SG 30-50 spins after a waiting time t w 1000 s. For samples with transverse dimensions L y , L z larger than SG , the proposed conductance measurements will probe properties of an effective 3D spin glass. Reconsidering the experiments of [13], we obtain approximately 40 spins in the transverse dimensions L y ' 900 A, implying a 3D spin glass behavior.…”
mentioning
confidence: 88%
“…where a ≈ 0.4 for the three-dimensional case [59], whereas F (x) = exp[−x β ] with β ∼ 1.5 [60]. Using the method of integral estimators proposed in [30], we note that the dynamical correlation length is proportional to the following ratio of integrals:…”
Section: Dynamical Correlation Lengthmentioning
confidence: 99%