Searcharxiv⌕ Search

arXiv subjects

Andreas Engel

Publications and source records attributed to Andreas Engel.

49 records · Page 3Linked to original sources

On large deviation properties of Erdos-Renyi random graphs

We show that large deviation properties of Erdös-Rényi random graphs can be derived from the free energy of the $q$-state Potts model of statistical mechanics. More precisely the Legendre transform of the Potts free energy with respect to $\ln q$ is related to the component generating function of the graph ensemble. This generalizes the well-known mapping between typical properties of random graphs and the $q\to 1$ limit of the Potts free energy. For exponentially rare graphs we explicitly calculate the number of components, the size of the giant component, the degree distributions inside and outside the giant component, and the distribution of small component sizes. We also perform numerical simulations which are in very good agreement with our analytical work. Finally we demonstrate how the same results can be derived by studying the evolution of random graphs under the insertion of new vertices and edges, without recourse to the thermodynamics of the Potts model.

cond-mat.stat-mech↗

Ferrofluids as thermal ratchets

Colloidal suspensions of ferromagnetic nano-particles, so-called ferrofluids, are shown to be suitable systems to demonstrate and investigate thermal ratchet behavior: By rectifying thermal fluctuations, angular momentum is transferred to a resting ferrofluid from an oscillating magnetic field without net rotating component. Via viscous coupling the noise driven rotation of the microscopic ferromagnetic grains is transmitted to the carrier liquid to yield a macroscopic torque. For a simple setup we analyze the rotation of the ferrofluid theoretically and show that the results are compatible with the outcome of a simple demonstration experiment.

cond-mat.mtrl-sci↗

Local amplitude equation from non-local dynamics

We derive a closed equation for the shape of the free surface of a magnetic fluid subject to an external magnetic field. The equation is strongly non-local due to the long range character of the magnetic interaction. We develop a systematic multiple scale perturbation expansion in which the non-locality is reduced to the occurrence of the Hilbert transform of the surface profile. The resulting third order amplitude equation describing the slow modulation of the basic pattern is shown to be purely local.

nlin.PS↗

Phase Transition in Multiprocessor Scheduling

The problem of distributing the workload on a parallel computer to minimize the overall runtime is known as Multiprocessor Scheduling Problem. It is NP-hard, but like many other NP-hard problems, the average hardness of random instances displays an ``easy-hard'' phase transition. The transition in Multiprocessor Scheduling can be analyzed using elementary notions from crystallography (Bravais lattices) and statistical mechanics (Potts vectors). The analysis reveals the control parameter of the transition and its critical value including finite size corrections. The transition is identified in the performance of practical scheduling algorithms.

cond-mat.dis-nn↗

Vortex dynamics and states of artificially layered superconducting films with correlated defects

Linear resistances and $IV$-characteristics have been measured over a wide range in the parameter space of the mixed phase of multilayered a-TaGe/Ge films. Three films with varying interlayer coupling and correlated defects oriented at an angle $\approx 25$ from the film normal were investigated. Experimental data were analyzed within vortex glass models and a second order phase transition from a resistive vortex liquid to a pinned glass phase. Various vortex phases including changes from three to two dimensional behavior depending on anisotropy have been identified. Careful analysis of $IV$-characteristics in the glass phases revealed a distinctive $T$ and $H$-dependence of the glass exponent $μ$. The vortex dynamics in the Bose-glass phase does not follow the predicted behavior for excitations of vortex kinks or loops.

cond-mat.supr-con↗

Jamming transitions and avalanches in the game of Dots-and-Boxes

We study the game of Dots-and-Boxes from a statistical point of view. The early game can be treated as a case of Random Sequential Adsorption, with a jamming transition that marks the beginning of the end-game. We derive set of differential equations to make predictions about the state of the lattice at the transition, and thus about the distribution of avalanches in the end-game.

cond-mat.dis-nn↗

Pattern and wavenumber selection in ferrofluids

The formation of patterns of peaks on the free surface of a ferrofluid subject to a magnetic field normal to the undisturbed interface is investigated theoretically. The relative stability of ridge, square, and hexagon planforms is studied using a perturbative energy minimization procedure. Extending previous studies the finite depth of the fluid layer is taken into account. Moreover, adding the wavenumber modulus $k$ to the set of variational parameters also the wavenumber selection problem is addressed. The results are compared with previous investigations and recent experimental findings.

nlin.PS↗

Multilayer neural networks with extensively many hidden units

The information processing abilities of a multilayer neural network with a number of hidden units scaling as the input dimension are studied using statistical mechanics methods. The mapping from the input layer to the hidden units is performed by general symmetric Boolean functions whereas the hidden layer is connected to the output by either discrete or continuous couplings. Introducing an overlap in the space of Boolean functions as order parameter the storage capacity if found to scale with the logarithm of the number of implementable Boolean functions. The generalization behaviour is smooth for continuous couplings and shows a discontinuous transition to perfect generalization for discrete ones.

cond-mat.stat-mech↗

Dynamics of a Single Peak of the Rosensweig Instability in a Magnetic Fluid

To describe the dynamics of a single peak of the Rosensweig instability a model is proposed which approximates the peak by a half-ellipsoid atop a layer of magnetic fluid. The resulting nonlinear equation for the height of the peak leads to the correct subcritical character of the bifurcation for static induction. For a time-dependent induction the effects of inertia and damping are incorporated. The results of the model show qualitative agreement with the experimental findings, as in the appearance of period doubling, trebling, and higher multiples of the driving period. Furthermore a quantitative agreement is also found for the parameter ranges of frequency and induction in which these phenomena occur.

patt-sol↗

Dissipation in ferrofluids: Mesoscopic versus hydrodynamic theory

Part of the field dependent dissipation in ferrofluids occurs due to the rotational motion of the ferromagnetic grains relative to the viscous flow of the carrier fluid. The classical theoretical description due to Shliomis uses a mesoscopic treatment of the particle motion to derive a relaxation equation for the non-equilibrium part of the magnetization. Complementary, the hydrodynamic approach of Liu involves only macroscopic quantities and results in dissipative Maxwell equations for the magnetic fields in the ferrofluid. Different stress tensors and constitutive equations lead to deviating theoretical predictions in those situations, where the magnetic relaxation processes cannot be considered instantaneous on the hydrodynamic time scale. We quantify these differences for two situations of experimental relevance namely a resting fluid in an oscillating oblique field and the damping of parametrically excited surface waves. The possibilities of an experimental differentiation between the two theoretical approaches is discussed.

patt-sol↗

Correlation of internal representations in feed-forward neural networks

Feed-forward multilayer neural networks implementing random input-output mappings develop characteristic correlations between the activity of their hidden nodes which are important for the understanding of the storage and generalization performance of the network. It is shown how these correlations can be calculated from the joint probability distribution of the aligning fields at the hidden units for arbitrary decoder function between hidden layer and output. Explicit results are given for the parity-, and-, and committee-machines with arbitrary number of hidden nodes near saturation.

cond-mat↗

Delta degrees of freedom in antisymmetrized molecular dynamics and (p,p') reactions in the delta region

Delta degrees of freedom are introduced into antisymmetrized molecular dynamics (AMD). This is done by increasing the number of basic states in the AMD wave function, introducing a Skyrme-type delta-nucleon potential, and including $NN\leftrightarrow NΔ$ reactions in the collision description. As a test of the delta dynamics, the extended AMD is applied to (p,p$'$) recations at $E_{\rm lab}=800$ MeV for a $^{12}$C target. It is found that the ratio and the absolute values for delta peak and quasielastic peak (QEP) in the $^{12}$C(p,p$'$) reaction are reproduced for angles $Θ_{\rm lab} \agt 40^\circ$, pointing to a correct treatment of the delta dynamics in the extended AMD. For forward angles the QEP is overestimated. The results of the AMD calculations are compared to one-step Monte Carlo (OSMC) calculations and a detailed analysis of multi-step and delta potential effects is given. As important side results we present a way to apply a Gallilei invariant theory for (N,N$'$) reactions up to $E_{\rm lab} \approx 800$ MeV which ensures approximate Lorentz invariance and we discuss how to fix the width parameter $ν$ of the single particle momentum distribution for outgoing nucleons in the AMD calculation.

nucl-th↗

Proton inelastic scattering to continuum studied with antisymmetrized molecular dynamics

Intermediate energy (p,p$'$x) reaction is studied with antisymmetrized molecular dynamics (AMD) in the cases of $^{58}$Ni target with $E_p = 120$ MeV and $^{12}$C target with $E_p = $ 200 and 90 MeV. Angular distributions for various $E_{p'}$ energies are shown to be reproduced well without any adjustable parameter, which shows the reliability and usefulness of AMD in describing light-ion reactions. Detailed analyses of the calculations are made in the case of $^{58}$Ni target and following results are obtained: Two-step contributions are found to be dominant in some large angle region and to be indispensable for the reproduction of data. Furthermore the reproduction of data in the large angle region $θ\agt 120^\circ$ for $E_{p'}$ = 100 MeV is shown to be due to three-step contributions. Angular distributions for $E_{p'} \agt$ 40 MeV are found to be insensitive to the choice of different in-medium nucleon-nucleon cross sections $σ_{NN}$ and the reason of this insensitivity is discussed in detail. On the other hand, the total reaction cross section and the cross section of evaporated protons are found to be sensitive to $σ_{NN}$. In the course of the analyses of the calculations, comparison is made with the distorted wave approach.

nucl-th↗