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: Functional Renormalization for Disordered Systems, Basic Recipes and Gourmet Dishes

Abstract: We give a pedagogical introduction into the functional renormalization group treatment of disordered systems. After a review of its phenomenology, we show why in the context of disordered systems a functional renormalization group treatment is necessary, contrary to pure systems, where renormalization of a single coupling constant is sufficient. This leads to a disorder distribution, which after a finite renormalization becomes non-analytic, thus overcoming the predictions of the seemingly exact dimensional reduction. We discuss, how the non-analyticity can be measured in a simulation or experiment. We then construct a renormalizable field theory beyond leading order. We discuss an elastic manifold embedded in N dimensions, and give the exact solution for N to infinity. This is compared to predictions of the Gaussian replica variational ansatz, using replica symmetry breaking. We further consider random field magnets, and supersymmetry. We finally discuss depinning, both isotropic and anisotropic, and universal scaling function.
Comments: 29 pages
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn)
Journal reference: Markov Processes Relat. Fields 13 (2007) 777-818
Report number: NSF-KITP-06-108
Cite as: arXiv:cond-mat/0611346 [cond-mat.dis-nn]
  (or arXiv:cond-mat/0611346v1 [cond-mat.dis-nn] for this version)

Submission history

From: Kay Jörg Wiese [view email]
[v1] Mon, 13 Nov 2006 23:55:44 GMT (872kb)

Link back to: arXiv, form interface, contact.