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

Condensed Matter > Disordered Systems and Neural Networks

Title: Localization of thermal packets and metastable states in Sinai model

Authors: Cecile Monthus (SPhT Saclay France), Pierre Le Doussal (ENS Paris France)
Abstract: We consider the Sinai model describing a particle diffusing in a 1D random force field. As shown by Golosov, this model exhibits a strong localization phenomenon for the thermal packet: the disorder average of the thermal distribution of the relative distance y=x-m(t), with respect to the (disorder-dependent) most probable position m(t), converges in the limit of infinite time towards a distribution P(y). In this paper, we revisit this question of the localization of the thermal packet. We first generalize the result of Golosov by computing explicitly the joint asymptotic distribution of relative position y=x(t)-m(t) and relative energy u=U(x(t))-U(m(t)) for the thermal packet. Next, we compute in the infinite-time limit the localization parameters Y_k, representing the disorder-averaged probabilities that k particles of the thermal packet are at the same place, and the correlation function C(l) representing the disorder-averaged probability that two particles of the thermal packet are at a distance l from each other. We moreover prove that our results for Y_k and C(l) exactly coincide with the thermodynamic limit of the analog quantities computed for independent particles at equilibrium in a finite sample of length L. Finally, we discuss the properties of the finite-time metastable states that are responsible for the localization phenomenon and compare with the general theory of metastable states in glassy systems, in particular as a test of the Edwards conjecture.
Comments: 17 pages
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn)
Journal reference: Phys. Rev. E 65 (2002) 66129
DOI: 10.1103/PhysRevE.65.066129
Cite as: arXiv:cond-mat/0202295 [cond-mat.dis-nn]
  (or arXiv:cond-mat/0202295v1 [cond-mat.dis-nn] for this version)

Submission history

From: Cecile Monthus [view email]
[v1] Mon, 18 Feb 2002 13:22:40 GMT (24kb)

Link back to: arXiv, form interface, contact.