arXiv subjects

Itamar Procaccia

Publications and source records attributed to Itamar Procaccia.

At least 235 records · Page 13Linked to original sources

arXiv2002

Active and Passive Fields in Turbulent Transport: the Role of Statistically Preserved Structures

We have recently proposed that the statistics of active fields (which affect the velocity field itself) in well-developed turbulence are also dominated by the Statistically Preserved Structures of auxiliary passive fields which are advected by the same velocity field. The Statistically Preserved Structures are eigenmodes of eigenvalue 1 of an appropriate propagator of the decaying (unforced) passive field, or equivalently, the zero modes of a related operator. In this paper we investigate further this surprising finding via two examples, one akin to turbulent convection in which the temperature is the active scalar, and the other akin to magneto-hydrodynamics in which the magnetic field is the active vector. In the first example, all the even correlation functions of the active and passive fields exhibit identical scaling behavior. The second example appears at first sight to be a counter-example: the statistical objects of the active and passive fields have entirely different scaling exponents. We demonstrate nevertheless that the Statistically Preserved Structures of the passive vector dominate again the statistics of the active field, except that due to a dynamical conservation law the amplitude of the leading zero mode cancels exactly. The active vector is then dominated by the sub-leading zero mode of the passive vector. Our work thus suggests that the statistical properties of active fields in turbulence can be understood with the same generality as those of passive fields.

nlin.CD↗

arXiv2002

Stabilization of Hydrodynamic Flows by Small Viscosity Variations

Motivated by the large effect of turbulent drag reduction by minute concentrations of polymers we study the effects of a weakly space-dependent viscosity on the stability of hydrodynamic flows. In a recent Letter [Phys. Rev. Lett. {\bf 87}, 174501, (2001)] we exposed the crucial role played by a localized region where the energy of fluctuations is produced by interactions with the mean flow (the "critical layer"). We showed that a layer of weakly space-dependent viscosity placed near the critical layer can have a very large stabilizing effect on hydrodynamic fluctuations, retarding significantly the onset of turbulence. In this paper we extend these observation in two directions: first we show that the strong stabilization of the primary instability is also obtained when the viscosity profile is realistic (inferred from simulations of turbulent flows with a small concentration of polymers). Second, we analyze the secondary instability (around the time-dependent primary instability) and find similar strong stabilization. Since the secondary instability develops around a time-dependent solution and is three-dimensional, this brings us closer to the turbulent case. We reiterate that the large effect is {\em not} due to a modified dissipation (as is assumed in some theories of drag reduction), but due to reduced energy intake from the mean flow to the fluctuations. We propose that similar physics act in turbulent drag reduction.

nlin.CD↗

arXiv2002

Quasi-Static Brittle Fracture in Inhomogeneous Media and Iterated Conformal Maps: Modes I, II and III

The method of iterated conformal maps is developed for quasi-static fracture of brittle materials, for all modes of fracture. Previous theory, that was relevant for mode III only, is extended here to mode I and II. The latter require solution of the bi-Laplace rather than the Laplace equation. For all cases we can consider quenched randomness in the brittle material itself, as well as randomness in the succession of fracture events. While mode III calls for the advance (in time) of one analytic function, mode I and II call for the advance of two analytic functions. This fundamental difference creates different stress distribution around the cracks. As a result the geometric characteristics of the cracks differ, putting mode III in a different class compared to modes I and II.

cond-mat.stat-mech↗

arXiv2002

Transition in the Fractal Properties from Diffusion Limited Aggregation to Laplacian Growth via their Generalization

We study the fractal and multifractal properties (i.e. the generalized dimensions of the harmonic measure) of a 2-parameter family of growth patterns that result from a growth model that interpolates between Diffusion Limited Aggregation (DLA) and Laplacian Growth Patterns in 2-dimensions. The two parameters are βwhich determines the size of particles accreted to the interface, and C which measures the degree of coverage of the interface by each layer accreted to the growth pattern at every growth step. DLA and Laplacian Growth are obtained at β=0, C=0 and β=2, C=1, respectively. The main purpose of this paper is to show that there exists a line in the β-C phase diagram that separates fractal (D<2) from non-fractal (D=2) growth patterns. Moreover, Laplacian Growth is argued to lie in the non-fractal part of the phase diagram. Some of our arguments are not rigorous, but together with the numerics they indicate this result rather strongly. We first consider the family of models obtained for β=0, C>0, and derive for them a scaling relation D=2 * D_3. We then propose that this family has growth patterns for which D=2 for some C>C_{cr}, where C_{cr} may be zero. Next we consider the whole β-C phase diagram and define a line that separates 2-dimensional growth patterns from fractal patterns with D<2. We explain that Laplacian Growth lies in the region belonging to 2-dimensional growth patterns, motivating the main conjecture of this paper, i.e. that Laplacian Growth patterns are 2-dimensional. The meaning of this result is that the branches of Laplacian Growth patterns have finite (and growing) area on scales much larger than any ultra-violet cut-off length.

cond-mat.stat-mech↗

arXiv2002

Quasi-Gaussian Statistics of Hydrodynamic Turbulence in 3/4+εdimensions

The statistics of 2-dimensional turbulence exhibit a riddle: the scaling exponents in the regime of inverse energy cascade agree with the K41 theory of turbulence far from equilibrium, but the probability distribution functions are close to Gaussian like in equilibrium. The skewness $\C S \equiv S_3(R)/S^{3/2}_2(R)$ was measured as $\C S_{\text{exp}}\approx 0.03$. This contradiction is lifted by understanding that 2-dimensional turbulence is not far from a situation with equi-partition of enstrophy, which exist as true thermodynamic equilibrium with K41 exponents in space dimension of $d=4/3$. We evaluate theoretically the skewness $\C S(d)$ in dimensions ${4/3}\le d\le 2$, show that $\C S(d)=0$ at $d=4/3$, and that it remains as small as $\C S_{\text{exp}}$ in 2-dimensions.

nlin.CD↗

arXiv2002

Inverse Cascade Regime in Shell Models of 2-Dimensional Turbulence

We consider shell models that display an inverse energy cascade similar to 2-dimensional turbulence (together with a direct cascade of an enstrophy-like invariant). Previous attempts to construct such models ended negatively, stating that shell models give rise to a "quasi-equilibrium" situation with equipartition of the energy among the shells. We show analytically that the quasi-equilibrium state predicts its own disappearance upon changing the model parameters in favor of the establishment of an inverse cascade regime with K41 scaling. The latter regime is found where predicted, offering a useful model to study inverse cascades.

nlin.CD↗

arXiv2001

Statistically Preserved Structures and Anomalous Scaling in Turbulent Active Scalar Advection

The anomalous scaling of correlation functions in the turbulent statistics of active scalars (like temperature in turbulent convection) is understood in terms of an auxiliary passive scalar which is advected by the same turbulent velocity field. While the odd-order correlation functions of the active and passive fields differ, we propose that the even-order correlation functions are the same to leading order (up to a trivial multiplicative factor). The leading correlation functions are statistically preserved structures of the passive scalar decaying problem, and therefore universality of the scaling exponents of the even-order correlations of the active scalar is demonstrated.

nlin.CD↗

arXiv2001

Multifractal Structure of the Harmonic Measure of Diffusion Limited Aggregates

The method of iterated conformal maps allows to study the harmonic measure of Diffusion Limited Aggregates with unprecedented accuracy. We employ this method to explore the multifractal properties of the measure, including the scaling of the measure in the deepest fjords that were hitherto screened away from any numerical probing. We resolve probabilities as small as $10^{-35}$, and present an accurate determination of the generalized dimensions and the spectrum of singularities. We show that the generalized dimensions $D_q$ are infinite for $q<q^*$, where $q^*$ is of the order of -0.2. In the language of $f(α)$ this means that $α_{max}$ is finite. The $f(α)$ curve loses analyticity (the phenomenon of "phase transition") at $α_{max}$ and a finite value of $f(α_{max})$. We consider the geometric structure of the regions that support the lowest parts of the harmonic measure, and thus offer an explanation for the phase transition, rationalizing the value of $q^*$ and $f(α_{max})$. We thus offer a satisfactory physical picture of the scaling properties of this multifractal measure.

cond-mat.stat-mech↗

arXiv2001

Quasi-Static Fractures in Disordered Media and Iterated Conformal Maps

We study the geometrical characteristic of quasi-static fractures in disordered media, using iterated conformal maps to determine the evolution of the fracture pattern. This method allows an efficient and accurate solution of the Lamé equations without resorting to lattice models. Typical fracture patterns exhibit increased ramification due to the increase of the stress at the tips. We find the roughness exponent of the experimentally relevant backbone of the fracture pattern; it crosses over from about 0.5 for small scales to about 0.75 for large scales, in excellent agreement with experiments. We propose that this cross-over reflects the increased ramification of the fracture pattern.

cond-mat.stat-mech↗

arXiv2001

Thermodynamic Formalism of the Harmonic Measure of Diffusion Limited Aggregates: Phase Transition and Converged $f(α)$

We study the nature of the phase transition in the multifractal formalism of the harmonic measure of Diffusion Limited Aggregates (DLA). Contrary to previous work that relied on random walk simulations or ad-hoc models to estimate the low probability events of deep fjord penetration, we employ the method of iterated conformal maps to obtain an accurate computation of the probability of the rarest events. We resolve probabilities as small as $10^{-70}$. We show that the generalized dimensions $D_q$ are infinite for $q<q^*$, where $q^*= -0.17\pm 0.02$. In the language of $f(α)$ this means that $α_{max}$ is finite. We present a converged $f(α)$ curve.

cond-mat.stat-mech↗

arXiv2001

Statistically Preserved Structures in Shell Models of Passive Scalar Advection

It was conjectured recently that Statiscally Preserved Structures underlie the statistical physics of turbulent transport processes. We analyze here in detail the time-dependent (non compact) linear operator that governs the dynamics of correlation functions in the case of shell models of passive scalar advection. The problem is generic in the sense that the driving velocity field is neither Gaussian nor $δ$-correlated in time. We show how to naturally discuss the dynamics in terms of an effective compact operator that displays "zero modes" which determine the anomalous scaling of the correlation functions. Since shell models have neither Lagrangian structure nor "shape dynamics" this example differs significantly from standard passive scalar advection. Nevertheless with the necessary modifications the generality and efficacy of the concept of Statistically Preserved Structures are further exemplified. In passing we point out a bonus of the present approach, in providing analytic predictions for the time-dependent correlation functions in decaying turbulent transport.

nlin.CD↗

arXiv2001

Iterated Conformal Dynamics and Laplacian Growth

The method of iterated conformal maps for the study of Diffusion Limited Aggregates (DLA) is generalized to the study of Laplacian Growth Patterns and related processes. We emphasize the fundamental difference between these processes: DLA is grown serially with constant size particles, while Laplacian patterns are grown by advancing each boundary point in parallel, proportionally to the gradient of the Laplacian field. We introduce a 2-parameter family of growth patterns that interpolates between DLA and a discrete version of Laplacian growth. The ultraviolet putative finite-time singularities are regularized here by a minimal tip size, equivalently for all the models in this family. With this we stress that the difference between DLA and Laplacian growth is NOT in the manner of ultraviolet regularization, but rather in their deeply different growth rules. The fractal dimensions of the asymptotic patterns depend continuously on the two parameters of the family, giving rise to a "phase diagram" in which DLA and discretized Laplacian growth are at the extreme ends. In particular we show that the fractal dimension of Laplacian growth patterns is much higher than the fractal dimension of DLA, with the possibility of dimension 2 for the former not excluded.

cond-mat.stat-mech↗

arXiv2001

Retardation of the onset of yurbulence by minor viscosity contrasts

Motivated by the large effect of turbulent drag reduction by minute concentrations of polymers we study the effects of minor viscosity contrasts on the stability of hydrodynamic flows. The key player is a localized region where the energy of fluctuations is produced by interactions with the mean flow (the "critical layer"). We show that a layer of weakly space-dependent viscosity placed near the critical layer can have very large stabilizing effect on hydrodynamic fluctuations, retarding significantly the onset of turbulence. The effect is not due to a modified dissipation (as is assumed in theories of drag reduction), but due to reduced energy intake from the mean flow to the fluctuations. We propose that similar physics act in turbulent drag reduction.

nlin.CD↗

arXiv2001

Laplacian Growth and Diffusion Limited Aggregation: different universality classes

It had been conjectured that Diffusion Limited Aggregates and Laplacian Growth patterns (with small surface tension) are in the same universality class. Using iterated conformal maps we construct a 1-parameter family of fractal growth patterns with a continuously varying fractal dimension. This family can be used to bound the dimension of Laplacian Growth patterns from below. The bound value is higher than the dimension of Diffusion Limited Aggregates, showing that the two problems belong to two different universality classes.

cond-mat.stat-mech↗

arXiv2000

Outliers, Extreme Events and Multiscaling

Extreme events have an important role which is sometime catastrophic in a variety of natural phenomena including climate, earthquakes and turbulence, as well as in man-made environments like financial markets. Statistical analysis and predictions in such systems are complicated by the fact that on the one hand extreme events may appear as "outliers" whose statistical properties do not seem to conform with the bulk of the data, and on the other hands they dominate the (fat) tails of probability distributions and the scaling of high moments, leading to "abnormal" or "multi"-scaling. We employ a shell model of turbulence to show that it is very useful to examine in detail the dynamics of onset and demise of extreme events. Doing so may reveal dynamical scaling properties of the extreme events that are characteristic to them, and not shared by the bulk of the fluctuations. As the extreme events dominate the tails of the distribution functions, knowledge of their dynamical scaling properties can be turned into a prediction of the functional form of the tails. We show that from the analysis of relatively short time horizons (in which the extreme events appear as outliers) we can predict the tails of the probability distribution functions, in agreement with data collected in very much longer time horizons. The conclusion is that events that may appear unpredictable on relatively short time horizons are actually a consistent part of a multiscaling statistics on longer time horizons.

nlin.CD↗

arXiv2000

Convergent Calculation of the Asymptotic Dimension of Diffusion Limited Aggregates: Scaling and Renormalization of Small Clusters

Diffusion Limited Aggregation (DLA) is a model of fractal growth that had attained a paradigmatic status due to its simplicity and its underlying role for a variety of pattern forming processes. We present a convergent calculation of the fractal dimension D of DLA based on a renormalization scheme for the first Laurent coefficient of the conformal map from the unit circle to the expanding boundary of the fractal cluster. The theory is applicable from very small (2-3 particles) to asymptotically large (n \to \infty) clusters. The computed dimension is D=1.713\pm 0.003.

cond-mat.stat-mech↗

arXiv2000

Analytic Calculation of the Anomalous Exponents in Turbulence: Using the Fusion Rules to Flush Out a Small Parameter

The main difficulty of statistical theories of fluid turbulence is the lack of an obvious small parameter. In this paper we show that the formerly established fusion rules can be employed to develop a theory in which Kolmogorov's statistics of 1941 acts as the zero order, or background statistics, and the anomalous corrections to the K41 scaling exponents $ζ_n$ of the $n$th order structure functions can be computed analytically. The crux of the method consists of renormalizing a 4-point interaction amplitude on the basis of the fusion rules. The novelty is that this amplitude includes a small dimensionless parameter, which is shown to be of the order of the anomaly of $ζ_2$, $δ_2= ζ_2-2/3\approx 0.03$. Higher order interaction amplitudes are shown to be even smaller. The corrections to K41 to $O(δ_2)$ result from standard logarithmically divergent ladder-diagrams in which the 4-point interaction acts as a "rung". The theory allows a calculation of the anomalous exponents $ζ_n$ in powers of the small parameter $δ_2$. The $n$-dependence of the scaling exponents $ζ_n$ stems from pure combinatorics of the ladder diagrams. We calculate the exponents $ζ_n$ up to $O(δ_2^3)$. Previously derived bridge relations allow a calculation of the anomalous exponents of correlations of the dissipation field and of dynamical correlations in terms of the same parameter $δ_2$. The actual evaluation of the small parameter $δ_2$ from first principles requires additional developments that are outside the scope of this paper.

nlin.CD↗

arXiv2000

Analytic Theory of Fractal Growth Patterns in 2 Dimensions

Diffusion Limited Aggregation (DLA) is a model of fractal growth that was introduced in 1981 and had since attained a paradigmatic status due to its simplicity and its underlying role for a variety of pattern forming processes. Despite tremendous efforts there is no theory to compute the fractal dimension of DLA from first principles. In this Letter we offer such a theory for fractal growth patterns in two dimensions, including DLA as a particular case. In this theory the fractal dimension of the asymptotic cluster manifests iteself as a renormalization exponent observable already at very early growth stages. Using early stage dynamics we compute 1.6896<D<1.7135, and explain why traditional numerical estimates converged so slowly. We discuss similar computations for other fractal growth processes in 2-dimensions.

cond-mat.stat-mech↗