Searcharxiv⌕ Search

arXiv subjects

D. A. Head

Publications and source records attributed to D. A. Head.

32 records · Page 2Linked to original sources

Universal persistence exponents in an extremally driven system

The local persistence R(t), defined as the proportion of the system still in its initial state at time t, is measured for the Bak--Sneppen model. For 1 and 2 dimensions, it is found that the decay of R(t) depends on one of two classes of initial configuration. For a subcritical initial state, R(t)\sim t^{-θ}, where the persistence exponent θcan be expressed in terms of a known universal exponent. Hence θis universal. Conversely, starting from a supercritical state, R(t) decays by the anomalous form 1-R(t)\sim t^{τ_{\rm ALL}} until a finite time t_{0}, where τ_{\rm ALL} is also a known exponent. Finally, for the high dimensional model R(t) decays exponentially with a non--universal decay constant.

cond-mat.stat-mech↗

Jamming, hysteresis and oscillation in scalar models for shear thickening

We investigate shear thickening and jamming within the framework of a family of spatially homogeneous, scalar rheological models. These are based on the `soft glassy rheology' model of Sollich et al. [Phys. Rev. Lett. 78, 2020 (1997)], but with an effective temperature x that is a decreasing function of either the global stress σor the local strain l. For appropiate x=x(σ), it is shown that the flow curves include a region of negative slope, around which the stress exhibits hysteresis under a cyclically varying imposed strain rate \gd. A subclass of these x(σ) have flow curves that touch the \gd=0 axis for a finite range of stresses; imposing a stress from this range {\em jams} the system, in the sense that the strain γcreeps only logarithmically with time t, γ(t)\sim\ln t. These same systems may produce a finite asymptotic yield stress under an imposed strain, in a manner that depends on the entire stress history of the sample, a phenomenon we refer to as history--dependent jamming. In contrast, when x=x(l) the flow curves are always monotonic, but we show that some x(l) generate an oscillatory strain response for a range of steady imposed stresses. Similar spontaneous oscillations are observed in a simplified model with fewer degrees of freedom. We discuss this result in relation to the temporal instabilities observed in rheological experiments and stick--slip behaviour found in other contexts, and comment on the possible relationship with `delay differential equations' that are known to produce oscillations and chaos.

cond-mat.soft↗

Rheological instability in a simple shear thickening model

We study the strain response to steady imposed stress in a spatially homogeneous, scalar model for shear thickening, in which the local rate of yielding Γ(l) of mesoscopic `elastic elements' is not monotonic in the local strain l. Despite this, the macroscopic, steady-state flow curve (stress vs. strain rate) is monotonic. However, for a broad class of Γ(l), the response to steady stress is not in fact steady flow, but spontaneous oscillation. We discuss this finding in relation to other theoretical and experimental flow instabilities. Within the parameter ranges we studied, the model does not exhibit rheo-chaos.

cond-mat.soft↗

Robust propagation direction of stresses in a minimal granular packing

By employing the adaptive network simulation method, we demonstrate that the ensemble-averaged stress caused by a local force for packings of frictionless rigid beads is concentrated along rays whose slope is consistent with unity: forces propagate along lines at 45 degrees to the horizontal or vertical. This slope is shown to be independent of polydispersity or the degree to which the system is sheared. Further confirmation of this result comes from fitting the components of the stress tensor to the null stress constitutive equation. The magnitude of the response is also shown to fall off with the -1/2 power of distance. We argue that our findings are a natural consequence of a system that preserves its volume under small perturbations.

cond-mat.stat-mech↗

Extremal driving as a mechanism for generating long-term memory

It is argued that systems whose elements are renewed according to an extremal criterion can generally be expected to exhibit long-term memory. This is verified for the minimal extremally driven model, which is first defined and then solved for all system sizes N\geq2 and times t\geq0, yielding exact expressions for the persistence R(t)=[1+t/(N-1)]^{-1} and the two-time correlation function C(t_{\rm w}+t,t_{\rm w})=(1-1/N)(N+t_{\rm w})/(N+t_{\rm w}+t-1). The existence of long-term memory is inferred from the scaling of C(t_{\rm w}+t,t_{\rm w})\sim f(t/t_{\rm w}), denoting {\em aging}. Finally, we suggest ways of investigating the robustness of this mechanism when competing processes are present.

cond-mat.stat-mech↗

Temperature scaling, glassiness and stationarity in the Bak-Sneppen model

We show that the emergence of criticality in the locally-defined Bak-Sneppen model corresponds to separation over a hierarchy of timescales. Near to the critical point the model obeys scaling relations, with exponents which we derive numerically for a one-dimensional system. We further describe how the model can be related to the glass model of Bouchaud [{\em J. Phys. I France {\bf 2}, 1705 (1992)}], and we use this insight to comment on the usual assumption of stationarity in the Bak-Sneppen model. Finally, we propose a general definition of self-organised criticality which is in partial agreement with other recent definitions.

cond-mat.stat-mech↗

Phenomenological glass model for vibratory granular compaction

A model for weakly excited granular media is derived by combining the free volume argument of Nowak et al. [Phys. Rev. E 57, 1971 (1998)] and the phenomenological model for supercooled liquids of Adam and Gibbs [J. Chem. Phys. 43, 139 (1965)]. This is made possible by relating the granular excitation parameter Γ, defined as the peak acceleration of the driving pulse scaled by gravity, to a temperature-like parameter η(Γ). The resulting master equation is formally identical to that of Bouchaud's trap model for glasses [J. Phys. I 2, 1705 (1992)]. Analytic and simulation results are shown to compare favourably with a range of known experimental behaviour. This includes the logarithmic densification and power spectrum of fluctuations under constant η, the annealing curve when ηis varied cyclically in time, and memory effects observed for a discontinuous shift in η. Finally, we discuss the physical interpretation of the model parameters and suggest further experiments for this class of systems.

cond-mat.stat-mech↗

Stretched exponentials and power laws in granular avalanching

We introduce a model for granular avalanching which exhibits both stretched exponential and power law avalanching over its parameter range. Two modes of transport are incorporated, a rolling layer consisting of individual particles and the overdamped, sliding motion of particle clusters. The crossover in behaviour observed in experiments on piles of rice is attributed to a change in the dominant mode of transport. We predict that power law avalanching will be observed whenever surface flow is dominated by clustered motion.

cond-mat.stat-mech↗

A model of macroevolution with a natural system size

We describe a simple model of evolution which incorporates the branching and extinction of species lines, and also includes abiotic influences. A first principles approach is taken in which the probability for speciation and extinction are defined purely in terms of the fitness landscapes of each species. Numerical simulations show that the total diversity fluctuates around a natural system size $N_{\rm nat}$ which only weakly depends upon the number of connections per species. This is in agreement with known data for real multispecies communities. The numerical results are confirmed by approximate mean field analysis.

adap-org↗

The Anisotropic Bak-Sneppen Model

The Bak-Sneppen model is shown to fall into a different universality class with the introduction of a preferred direction, mirroring the situation in spin systems. This is first demonstrated by numerical simulations and subsequently confirmed by analysis of the multi-trait version of the model, which admits exact solutions in the extremes of zero and maximal anisotropy. For intermediate anisotropies, we show that the spatiotemporal evolution of the avalanche has a power law ``tail'' which passes through the system for any non-zero anisotropy but remains fixed for the isotropic case, thus explaining the crossover in behaviour. Finally, we identify the maximally anisotropic model which is more tractable and yet more generally applicable than the isotropic system.

cond-mat.stat-mech↗

A coarse grained model of granular compaction and relaxation

We introduce a theoretical model for the compaction of granular materials by discrete vibrations which is expected to hold when the intensity of vibration is low. The dynamical unit is taken to be clusters of granules that belong to the same collective structure. We rigourously construct the model from first principles and show that numerical solutions compare favourably with a range of experimental results. This includes the logarithmic relaxation towards a statistical steady state, the effect of varying the intensity of vibration resulting in a so-called ``annealing'' curve, and the power spectrum of density fluctuations in the steady state itself. A mean field version of the model is introduced which shares many features with the exact model and is open to quantitative analysis.

cond-mat.stat-mech↗

Slowly Driven Sandpile Formation with Granular Mixtures

We introduce a one-dimensional sandpile model with $N$ different particle types and an infinitesimal driving rate. The parameters for the model are the N^2 critical slopes for one type of particle on top of another. The model is trivial when N=1, but for N=2 we observe four broad classes of sandpile structure in different regions of the parameter space. We describe and explain the behaviour of each of these classes, giving quantitative analysis wherever possible. The behaviour of sandpiles with N>2 essentially consists of combinations of these four classes. We investigate the model's robustness and highlight the key areas that any experiment designed to reproduce these results should focus on.

cond-mat.stat-mech↗

Crossover to Self-Organized Criticality in an Inertial Sandpile Model

We introduce a one-dimensional sandpile model which incorporates particle inertia. The inertial dynamics are governed by a new parameter which, as it passes through a threshold value, alters the toppling dynamics in such a way that the system no longer evolves to a self-organized critical state. A range of mean-field theories based on a kinetic equation approach is presented which confirm the numerical findings. We conclude by considering the physical applications of this model, particularly with reference to recent experimental results.

cond-mat↗

Kinetics of Catalysis with Surface Disorder

We study the effects of generalised surface disorder on the monomer-monomer model of heterogeneous catalysis, where disorder is implemented by allowing different adsorption rates for each lattice site. By mapping the system in the reaction-controlled limit onto a kinetic Ising model, we derive the rate equations for the one and two-spin correlation functions. There is good agreement between these equations and numerical simulations. We then study the inclusion of desorption of monomers from the substrate, first by both species and then by just one, and find exact time-dependent solutions for the one-spin correlation functions.

cond-mat↗