arXiv subjects
Publications and source records attributed to H. J. Hilhorst.
We consider the average number B_m(t) of bonds traversed exactly m times by a t step simple random walk. We determine B_m(t) explicitly in the scaling limit t -> oo with m/sqrt(t) fixed in dimension d=1 and m/log(t) fixed in dimension d=2. The scaling function is an erfc in d=1 and an exponential in d=2.
We show that the contact process in an ensemble with conserved total particle number, as simulated recently by Tome and de Oliveira [Phys. Rev. Lett. 86 (2001) 5463], is equivalent to the ordinary contact process, in agreement with what the authors assumed and believed. Similar conserved ensembles and equivalence proofs are easily constructed for other models.
Our recent Monte Carlo results [Phys. Rev. E 61 (2000) 6330] for the one-dimensional reaction-diffusion process A+B->2B and B->A lead to the correlation length exponent estimate nu=2.21 +/- 0.05. In a comment on our work [cond-mat/0007366] the exact value is claimed to be nu=2. We reply that the arguments advanced in the comment fail to prove the claim.
The Riemann walk is the lattice version of the Levy flight. For the one-dimensional Riemann walk of Levy exponent 0<α<2 we study the statistics of the support, i.e. the set of visited sites, after t steps. We consider a wide class of support related observables M(t), including the number S(t) of visited sites and the number I(t) of sequences of visited sites. For t->\infty we obtain the asymptotic power laws for the averages, variances, and correlations of these observables. Logarithmic correction factors appear for α=2/3 and α=1. Bulk and surface observables have different power laws for 1\leqα<2. Fluctuations are shown to be universal for 2/3\leqα<2. This means that in the limit t->\infty the deviations from average \DeltaM(t) are fully described (i) either by a single M independent stochastic process (when 2/3\leqα\leq 1) (ii) or by two such processes, one for the bulk and one for the surface observables (when 1<α<2).
Motivated by certain problems of statistical physics we consider a stationary stochastic process in which deterministic evolution is interrupted at random times by upward jumps of a fixed size. If the evolution consists of linear decay, the sample functions are of the "random sawtooth" type and the level dependent persistence exponent θcan be calculated exactly. We then develop an expansion method valid for small curvature of the deterministic curve. The curvature parameter g plays the role of the coupling constant of an interacting particle system. The leading order curvature correction to θis proportional to g^{2/3}. The expansion applies in particular to exponential decay in the limit of large level, where the curvature correction considerably improves the linear approximation. The Langevin equation, with Gaussian white noise, is recovered as a singular limiting case.
We consider the d-dimensional diffusion equation for a field phi(x,t) with random initial condition, and observe that, when appropriately scaled, phi(0,t) is Gaussian and Markovian in the limit d->0. This leads via the Majumdar-Sire perturbation theory to a small-d expansion for the persistence exponent theta(d). We find theta(d) = d/4 - 0.12065...d^{3/2} + ...
We present the first example of a phase transition in a nonequilibrium steady-state that can be argued analytically to be first order. The system of interest is a two-species reaction-diffusion problem whose control parameter is the total density ρ. Mean-field theory predicts a second-order transition between two stationary states at a critical density ρ=ρ_c. We develop a phenomenological picture that, instead, below the upper critical dimension d_c=4, predicts a first-order transition. This picture is confirmed by hysteresis found in numerical simulations, and by the study of a renormalization-group improved equation of state. The latter approach is inspired by the Weinberg-Coleman mechanism in QED.
We consider two stochastic processes, the Gribov process and the general epidemic process, that describe the spreading of an infectious disease. In contrast to the usually assumed case of short-range infections that lead, at the critical point, to directed and isotropic percolation respectively, we consider long-range infections with a probability distribution decaying in d dimensions with the distance as 1/R^{d+σ}. By means of Wilson's momentum shell renormalization-group recursion relations, the critical exponents characterizing the growing fractal clusters are calculated to first order in an ε-expansion. It is shown that the long-range critical behavior changes continuously to its short-range counterpart for a decay exponent of the infection σ=σ_c>2.
Healthy and sick individuals (A and B particles) diffuse independently with diffusion constants D_A and D_B. Sick individuals upon encounter infect healthy ones (at rate k), but may also spontaneously recover (at rate 1/τ). The propagation of the epidemic therefore couples to the fluctuations in the total population density. Global extinction occurs below a critical value ρ_{c} of the spatially averaged total density. The epidemic evolves as the diffusion--reaction--decay process A + B --> 2B, B --> A , for which we write down the field theory. The stationary state properties of this theory when D_A=D_B were obtained by Kree et al. The critical behavior for D_A D_B remains unsolved.
A random walk starts from the origin of a d-dimensional lattice. The occupation number n(x,t) equals unity if after t steps site x has been visited by the walk, and zero otherwise. We study translationally invariant sums M(t) of observables defined locally on the field of occupation numbers. Examples are the number S(t) of visited sites; the area E(t) of the (appropriately defined) surface of the set of visited sites; and, in dimension d=3, the Euler index of this surface. In d > 3, the averages (t) all increase linearly with t as t-->infinity. We show that in d=3, to leading order in an asymptotic expansion in t, the deviations from average Delta M(t)= M(t)- (t) are, up to a normalization, all identical to a single "universal" random variable. This result resembles an earlier one in dimension d=2; we show that this universality breaks down for d>3.
We study the support (i.e. the set of visited sites) of a t step random walk on a two-dimensional square lattice in the large t limit. A broad class of global properties M(t) of the support is considered, including, e.g., the number S(t) of its sites; the length of its boundary; the number of islands of unvisited sites that it encloses; the number of such islands of given shape, size, and orientation; and the number of occurrences in space of specific local patterns of visited and unvisited sites. On a finite lattice we determine the scaling functions that describe the averages on appropriate lattice size dependent time scales. On an infinite lattice we first observe that the all increase with t as t/\log^k t, where k is an M dependent positive integer. We then consider the class of random processes constituted by the fluctuations around average Delta M(t). We show that to leading order as t gets large these fluctuations are all proportional to a single universal random process eta(t), normalized to =1$. For t--> infinity the probability law of eta(t) tends to that of Varadhan's renormalized local time of self-intersections. An implication is that in the long time limit all Delta M(t) are proportional to Delta S(t).
We analyse the critical region of finite-($d$)-dimensional Ising spin glass, in particular the limit of $d$ closely above the lower critical dimension $d_\ell$. At criticality the thermally active degrees of freedom are surfaces (of width zero) enclosing clusters of spins that may reverse with respect to their environment. The surfaces are organised in finite interacting structures. These may be called {\em protodroplets}\/, since in the off-critical limit they reduce to the Fisher and Huse droplets. This picture provides an explanation for the phenomenon of critical chaos discovered earlier. It also implies that the spin-spin and energy-energy correlation functions are multifractal and we present scaling laws that describe them. Several of our results should be verifiable in Monte Carlo studies at finite temperature in $d=3$.
We analyse the chiral symmetry in the random $\pm J$ $XY$ model on a $N\times 2$ square lattice with periodic boundary conditions in the transverse direction. This ``tube" lattice may be seen as a two-dimensional lattice of which one dimension has been compactified. In the Villain formulation the discrete-valued {\em chiralities}\/ or {\em charges}\/ associated with the plaquettes of the lattice decouple from the continuous degrees of freedom. The difficulty of the problem lies in the fact that the chiralities interact through the long range ``strong" one-dimensional Coulomb potential - which increases linearly with distance - as well as through an exponentially decaying ``weak" interaction. By comparing the ground state energies for periodic, antiperiodic, and reflecting boundary conditions in the longitudinal direction, we show that the chiralities and the $XY$ spins have the {\em same}\/ zero-$T$ correlation length exponent, whose exact value $ν_c = 0.5564\ldots$ we determine. The equality of these correlation lengths even in the presence of long range chirality-chirality interactions lends support to the view that chiral glass order cannot be sustained without simultaneous spin glass order.
This is an analytic study of the two-dimensional XY spin glass with $\pm J$ disorder. The Hamiltonian has a continuous spin symmetry and a discrete chiral symmetry, and therefore possesses, potentially, two different order parameters and correlation lengths. The cost of breaking the symmetries is probed by comparing the ground state energy under periodic (P) boundary conditions with the one under antiperiodic (AP) and under reflecting (R) boundary conditions. Two energy differences (``domain wall energies'') appear, $ΔE^{AP}$ and $ΔE^R$, whose scaling behavior with system size is nontrivially related to the correlation length exponents. For a specific distribution of the $\pm J$ disorder we show that the chiral and spin correlation lengths diverge with the same exponent as $T\downarrow 0$. The common exponent has a common cause, viz. the low reversal energy of domains of chiral variables. For general disorder we give a heuristic argumentation in terms of droplet excitations that leads to spin ordering on a longer, or equal, scale than chiral ordering. These results are in contrast with interpretations of Monte Carlo simulations.