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

Download:

Current browse context:

cond-mat.soft

Change to browse by:

References & Citations

Bookmark

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

Condensed Matter > Disordered Systems and Neural Networks

Title: Mean-field density of states of a small-world model and a jammed soft spheres model

Authors: Mario Pernici (Istituto Nazionale di Fisica Nucleare, Sezione di Milano)
Abstract: We consider a class of random block matrix models in $d$ dimensions, $d \ge 1$, motivated by the study of the vibrational density of states (DOS) of soft spheres near the isostatic point. The contact networks of average degree $Z = z_0 + \zeta$ are represented by random $z_0$-regular graphs (only the circle graph in $d=1$ with $z_0=2$) to which Erd\"os-Renyi graphs having a small average degree $\zeta$ are superimposed.
In the case $d=1$, for $\zeta$ small the shifted Kesten-McKay DOS with parameter $Z$ is a mean-field solution for the DOS. Numerical simulations in the $z_0=2$ model, which is the $k=1$ Newman-Watts small-world model, and in the $z_0=3$ model lead us to conjecture that for $\zeta \to 0$ the cumulative function of the DOS converges uniformly to that of the shifted Kesten-McKay DOS, in an interval $[0, \omega_0]$, with $\omega_0 < \sqrt{z_0-1} + 1$.
For $2 \le d \le 4$, we introduce a cutoff parameter $K_d \le 0.5$ modeling sphere repulsion. The case $K_d=0$ is the random elastic network case, with the DOS close to the Marchenko-Pastur DOS with parameter $t=\frac{Z}{d}$. For $K_d$ large the DOS is close for small $\omega$ to the shifted Kesten-McKay DOS with parameter $t=\frac{Z}{d}$; in the isostatic case the DOS has around $\omega=0$ the expected plateau. The boson peak frequency in $d=3$ with $K_3$ large is close to the one found in molecular dynamics simulations for $Z=7$ and $8$.
Comments: 14 pages, 10 figures
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn); Soft Condensed Matter (cond-mat.soft); Statistical Mechanics (cond-mat.stat-mech)
Cite as: arXiv:2001.02622 [cond-mat.dis-nn]
  (or arXiv:2001.02622v1 [cond-mat.dis-nn] for this version)

Submission history

From: Mario Pernici [view email]
[v1] Wed, 8 Jan 2020 16:59:55 GMT (737kb)

Link back to: arXiv, form interface, contact.