We report the observation of intermittency in gravity-capillary wave turbulence on the surface of mercury. We measure the temporal fluctuations of surface wave amplitude at a given location. We show that the shape of the probability density function of the local slope increments of the surface waves strongly changes across the time scales. The related structure functions and the flatness are found to be power laws of the time scale on more than one decade. The exponents of these power laws increase nonlinearly with the order of the structure function. All these observations show the intermittent nature of the increments of the local slope in wave turbulence. We discuss the possible origin of this intermittency.PACS numbers: 47.35.-i, 47.52.+j, 05.45.-a One of the most striking feature of turbulence is the occurrence of bursts of intense motion within more quiescent fluid flow. This generates an intermittent behavior [1,2]. One of the quantitative characterization of intermittency is given by the probability density function (PDF) of the velocity increments between two points separated by a distance r. Starting from a roughly Gaussian PDF at integral scale, the PDFs undergo a continuous deformation when r is decreased within the inertial range and develop more and more stretched exponential tails [3]. Deviation from the Gaussian shape can be quantified by the flatness of the PDF. The origin of nonGaussian statistics in three dimensional hydrodynamic turbulence has been ascribed to the formation of strong vortices since the early work of Batchelor and Townsend [1]. However, the physical mechanism of intermittency is still an open question that motivates a lot of studies in three dimensional turbulence [4]. Intermittency has been also observed in a lot of problems involving transport by a turbulent flow for which the analytical description of the anomalous scaling laws can be obtained [5].It has been known since the work of Zakharov and collaborators that weakly interacting nonlinear waves can also display Kolmogorov type spectra related to an energy flux cascading from large to small scales [6,7]. These spectra have been analytically computed using perturbation techniques, but can also be obtained by dimensional analysis using Kolmogorov-type arguments [8]. More recently, it has been proposed that intermittency corrections should be also taken into account in wave turbulence [9] and may be connected to singularities or coherent structures [8,10] such as wave breaking [11] or whitecaps [8] in the case of surface waves. However, intermittency in wave turbulence is often related to non Gaussian statistics of low wave number Fourier amplitudes [10], thus it is not obviously related to small scale intermittency of hydrodynamic turbulence. Surprisingly, there exist only a small number of experimental studies on wave turbulence [12,13,14,15,16] compared to hydrodynamic turbulence, and to the best of our knowledge, no experimental observation of intermittency has been reported in wave turbulence.In this letter, we repo...

No abstract

PACS. 46.10.+z Mechanics of discrete systems - 83.70.Fn Granular solids,

The relationship between the conductivity and the polarization noise is measured in a gel as a function of frequency in the range 1Hz − 40Hz. It is found that at the beginning of the transition from a fluid like sol to a solid like gel the fluctuation dissipation theorem is strongly violated. The amplitude and the persistence time of this violation are decreasing functions of frequency. At the lowest frequencies of the measuring range it persists for times which are about 5% of the time needed to form the gel. This phenomenology is quite close to the recent theoretical predictions done for the violation of the fluctuation dissipation theorem in glassy systems.Many physical systems in nature are not in thermodynamic equilibrium because they present very slow relaxation processes. A typical example of this phenomenon is the aging of glassy materials: when they are quenched from above their glass transition temperature T g to a temperature T < T g , any response function of these systems depends on the aging time t a spent at T . For example, the dielectric and elastic constants of polymers continue to evolve several years after the quench [1]. Because of these slow relaxation processes, the glass is out of equilibrium, and usual thermodynamics does not apply. However, as this time evolution is slow, some concepts of the classical approach may be useful for understanding the glass aging properties. A widely studied question, is how the temperature of these systems can be defined. One possible answer comes from the study of the deviation to the Fluctuation Dissipation Theorem (FDT) in an out of equilibrium system (for a review see ref. [2,3,4]). In this letter we show that this approach is relevant for the study of a sol-gel transition, where a strong violation of FDT is measured. Implications of this observation go beyond the physics interest. Indeed FDT is used as a tool to extract, from 1

We report an experimental study of subharmonic instabilities observed in a horizontal layer of glass spheres under vertical vibrations. Above a critical value of the acceleration, the layer undergoes a period-doubling instability. When the acceleration is increased further, the spatially homogeneous temporal oscillation at half the excitation frequency becomes unstable and a regular pattern of defects is generated. Each defect separates two regions oscillating out of phase, and is thus connected with the discrete broken symmetry at the period-doubling bifurcation.

The influence of large-scale flow on heat transport in turbulent thermal convection is experimentally investigated. Large-scale flow couples the upper and lower thermal boundary layers. This coupling produces a slow coherent oscillation of the temperature field and strongly influences the spatial distribution of temperature fluctuations. Moreover, when the large-scale flow is either suppressed or strongly modified no significant variation of the heat transport across the cell is observed. ͓S1063-651X͑96͒51012-3͔ PACS number͑s͒: 47.27.Te Turbulent thermal convection in a fluid layer heated from below, that is, Rayleigh-Benard convection, presents several open problems which are not yet very well understood ͓1͔. One of these is the role played by the mean circulation flow on the heat transport properties. This mean circulation flow consists of a large-scale convective roll which involves all the cell containing the convective fluid, thus it behaves like a large coherent structure of the turbulent flow. Its presence in turbulent Rayleigh-Benard convection was first noticed by Howart and Krishnamurty ͓2͔ but its importance on the heat flow has been pointed out only more recently ͓3-5͔. There is now a general agreement that the large-scale flow strongly perturbs the thermal boundary layers ͓5͔. However, it is not yet clear how this perturbation actually occurs. For example, one can ask whether the mean flow influences more either the mean temperature profile or the temperature fluctuations close to the boundaries. Furthermore, it is unclear whether the mean flow transports a relevant amount of heat and whether it couples the upper and lower thermal boundary layers, as recently proposed by Villermaux ͓6͔. Several of these questions have been analyzed in a very smart experiment ͓7͔ in which one of the two boundary layers was perturbed by a large-scale horizontal flow. In another recent experiment ͓8͔ the boundary layers were instead perturbed by increasing the roughness of the horizontal plates. However, in those experiments, the large-scale vertical flow was always present; thus, a clear answer to the above mentioned questions could not be given. In this paper we report several results of an experiment in which the large-scale flow is directly perturbed in order to study its influence on the thermal boundary layer and on the heat transport properties. We find that the spatial distribution of the temperature fluctuations at the boundaries is strongly modified when the large-scale flow is suppressed near the boundaries and the two boundary layers are decoupled. However, no significant difference in the heat flow is observed. These features have been studied as a function of the Rayleigh number Raϭ␣gd 3 ⌬T/(), where ␣ is the thermal-expansion coefficient, g the gravitational acceleration, ⌬T the temperature drop across the fluid layer, d the height of the layer, the kinematic viscosity, and the thermal diffusivity.The experimental apparatus has been already described in Ref. ͓9͔, thus we mention here only the main charact...

We report the observation of the crossover between gravity and capillary wave turbulence on the surface of mercury. The probability density functions of the turbulent wave height are found to be asymmetric and thus non-Gaussian. The surface wave height displays power-law spectra in both regimes. In the capillary region, the exponent is in fair agreement with weak turbulence theory. In the gravity region, it depends on the forcing parameters. This can be related to the finite size of the container. In addition, the scaling of those spectra with the mean energy flux is found in disagreement with weak turbulence theory for both regimes.

The key governing parameter of wave turbulence is the energy flux that drives the waves and cascades to small scales through nonlinear interactions. In the inertial range, the energy flux is conserved across the scales, and is assumed to be constant in most theoretical approaches. It is only recently that measurements of the injected power into wave turbulence have been performed at the scale of the wave maker (integral scale). In this review, we focus on the statistical properties of the injected power fluctuations in gravity-capillary wave turbulence in a stationary regime. Fluctuations of the injected power have been found much larger than their mean value. In addition, events related to a negative injected power, i.e. an instantaneous reversed energy flux, occur with a fairly large probability. Both features are well described using a Langevin type equation. Finally, we consider the experimental dependence of the scaling law of the wave spectrum with the mean injected power and discuss possible reasons for the discrepancy with weak turbulence theory.