Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A Kolmogorov spectrum for strongly vibrating plates

A Kolmogorov spectrum for strongly vibrating plates Is turbulence universal ? A Kolmogorov spectrum for strongly vibrating plates. Gustavo During  Facultad de F sica, PUC, Chile. Christophe Josserand LadHyX, CNRS & Ecole Polytechnique, UMR 7646, 91128, Palaiseau, France. Giorgio Krstulovic Universit e C^ ote d'Azur, Observatoire de la C^ ote d'Azur, CNRS, Laboratoire Lagrange, Bd de l'Observatoire, CS 34229, 06304 Nice cedex 4, France. Sergio Rica Facultad de Ingenier a y Ciencias, Universidad Adolfo Ib anez, ~ Santiago, Chile. In uid turbulence, energy is transferred from a scale to another by an energy cascade that depends only on the energy dissipation rate. It leads by dimensional arguments to the Kolmogorov 1941 (K41) spectrum. Remarkably the normal modes of vibrations in elastic plates manifests an energy cascade with the same K41 spectrum in the fully non-linear regime. Moreover, the elastic deformations present large \eddies" together with a myriad of small \crumpling eddies", such that folds, developable cones, and more complex stretching structures, in close analogy with spots, swirls, vortices and other structures in hydrodynamic turbulence. We characterize the energy cascade, the validity of the constant energy dissipation rate over the scales and the role of intermittency via the correlation functions. Turbulence has remained a central problem in uid dy- sions. Similarly, uid turbulence is a hard problem be- namics since the early experiments of Osborne Reynolds cause it is in the same footing that the Euler equation [1]. Perhaps the most salient feature is the setting of for perfect uids which is ab-initio a nonlinear partial an energy cascade that redistributes the energy among di erential equation without any small parameter that di erent Fourier modes of the velocity uctuations that may justify an asymptotic scheme nor a rational closure are independent of the dissipation caused by viscosity at for the statistical correlations [7]. Since uid turbulence small scales. This energy cascade depends only on the is studied in the framework of the NS equations, it is energy dissipation rate per unit mass, P and the wave important to question whether such cascade dynamics number k leading to the Kolmogorov K41 [2] spectrum: transporting the energy from the large injection scales 2=3 5=3 E  P k . Despite a century of e ort, turbulence to the small dissipative ones can be observed in di erent still remains nowadays a major challenge from the exper- contexts and whether the turbulence features described imental, theoretical and numerical points of view, and in uids are universal or dependent on the dynamical the very essence of the phenomena has not yet been re- equations. In particular, while the quadratic nonlinear vealed satisfactorily [3{5]. For instance, if the K41 law term of the NS equation is responsible of the mixing be- can be deduced using dimensional arguments, the statis- tween scales at the heart of the cascade process, noth- tical description of the turbulence fails to describe cor- ing in the theoretical description of turbulence takes into rectly the general correlation functions of the velocity account the details of the nonlinearity itself. In fact, elds. In fact, up to date, there is only a single analyt- turbulence has been observed in di erent contexts such ical result that maybe derived from the original Navier- as Magneto-Hydrodynamics where the magnetic eld is Stokes (NS) equations for incompressible uids, the so- coupled with the ow or viscoelastic turbulence in poly- called von K arm an-Howarth relation [6], that links the mer solutions at low Reynolds numbers for instance. In second and the third order correlations function of some these cases even though the nonlinearity di ers from the component of the uid velocity with the energy dissipa- one of NS equations and di erent spectra can be mea- tion rate. Higher moments of these quantities di er from sured, they have similar properties than the original Kol- mean- eld based predictions, and experimental as well as mogorov phenomenology. numerical data reveal extreme events, exhibiting heavy On the other hand, a di erent class of turbulence ex- tailed distributions, that are usually associated with in- ists when a linear term is present in the dynamics, the termittency. Moreover, one can argue that the diculty so-called wave turbulence. Wave propagation dominates of describing turbulence may come from the fundamen- at linear order and the waves interact through the non- tal issues raised by the uid mechanics equations, wit- linear terms. The dynamics is described by the weak nessed in particular by the Clay millennium problem on or wave turbulence theory (WTT), developed originally the regularity of the NS equation in three space dimen- in the sixties [8{11]. WTT is deduced in the limit of arXiv:1808.02025v1 [nlin.CD] 6 Aug 2018 2 weak wave amplitudes, so that the nonlinear terms can ber of studies in vibrating plates (for a review see [19]). be treated perturbatively, by contrast with uid turbu- Experiments performed soon after the theoretical pre- lence where the nonlinearity is present at rst order in dictions have shown slightly di erent spectrum power the dynamics [10, 12, 13]. Based on the long-time statis- laws [20, 21], that have been later explained by the spe- tics of randomly uctuating interacting waves, the WTT ci c features of the dissipation [22]. Furthermore, wave deduces a kinetic equation for the distribution of spectral turbulence of plates has shown to be a perfect system densities. Beside thermodynamical equilibrium solution, for investigating di erent concepts such as inverse cas- the key-feature of this equation is that non-equilibrium cade [23] and transitory dynamics [24]. More recently, turbulent stationary solutions also arise, called in general thin plates with a high forcing [25, 26], the breakdown Kolmogorov-Zakharov spectra (KZ). Indeed, as shown of the WTT and the onset of intermittency [27] have by Zakharov [10], they describe a constant ux transfer been extensively studied. In these high forcing regimes, (or cascade) of conserved quantities (e.g. energy) be- a wave turbulence spectrum is still observed at small tween large and small length scales. Examples of wave scales, while strong nonlinear regimes appear at large turbulence go much beyond the cases of surface gravity scales, whose nature remain unclear, suggesting the exis- or capillary waves and concerns systems as diverse as tence of a strong turbulent dynamics. plasma waves [10], nonlinear optics [14], vibrating elas- In the following, to investigate this strongly nonlinear tic plates [15] and gravitational waves in general rela- regime and the analogy with hydrodynamic turbulence, tivity [16], among others. Although the analogy with we will consider a thin elastic plate in the fully nonlinear turbulent ows is natural since it describes a ux of case which corresponds to the formal limit with no bend- the energy between scales, the links within the physi- ing and where linear waves are absent. With no surprise, cal mechanisms is less straightforward: in particular, the the WTT spectrum vanishes in this limit and a di erent WTT is based on the resonant interactions between the turbulent spectrum is expected to appear. The goal of waves prescribed by the nonlinearity so that the power the paper is to characterize such energy repartition and of the nonlinearity is crucial in the process. For instance highlight similarities with hydrodynamic turbulence. if the pertinent nonlinearity is quadratic, the dynamics is determined by three waves interactions such that both the wave-numbers and the frequencies of the three waves THEORETICAL MODEL form resonant triads. This mechanism implies a speci c dependence of the KZ spectrum with the energy ux P as The vibration of a bending-free elastic plate comes 1=N P where N is the power of the pertinent nonlinearity from the usual dynamical version of the F oppl{von of the dynamics (N = 2 for quadratic interactions, 3 for K arm an equations [28] for the vertical amplitude of the cubic, etc). This represents a crucial di erence between deformation  (x; y; t) and for the Airy stress function wave turbulence and hydrodynamic turbulence, because (x; y; t) that describes the three in{plane stresses: in the last one, the ux of energy appears in the second order correlation with the power 2=3, which cannot be obtained as a 1=N power. =   +   2  +F +D; (1) xx yy yy xx xy xy @t The turbulent asymptotic expansion of wave systems 2 2 in the strongly nonlinear limit is, by construction, sin-   =    : (2) xx yy xy gular, because the linear term becomes then subdomi- nant. However, a turbulent fully nonlinear dynamics is Here  = @ + @ is the usual Laplacian and  and E xx yy still expected to exist. The dependence of the statis- are, respectively, the mass density and the Young modu- tical properties with the potentially di erent nonlinear lus E of the plate. F and D are the forcing and dissipa- terms and the connexions with the NS uid turbulence tion respectively. is then worth to be investigated. In this paper, we pre- When F = D = 0, the equations (1) and (2) derive cisely address such singular limit in the case of vibrating from a Hamiltonian principle [15, 19]. More important, elastic plates. This system has indeed recently revealed it is a well posed system, in the sense that the energy a strong analogy with hydrodynamic turbulence through is compound by two positive (hence bound from below) the derivation of an exact law for a two-point correlation quantities, namely the kinetic energy and the stretching that is the equivalent of the Kolmogorov's 4/5-law [17]. energy per unit mass: The weakly nonlinear turbulence regime in vibrating Z Z elastic plates has been established ten years ago by three 1 E 2 1 2 E =  dr; E =     dr; kin stret xx yy xy of us [15]. Based on the dynamical F oppl{von K arm an 2S 8S equations [18] (described below), we have shown using (3) the WTT the existence of a spectrum of direct energy here S is the area plate. For dimensional purposes, we cascade. The rather simple comparison between the- use everywhere the energies per unit mass, E , E , and kin ory, numerics and experiments has led to a large num- E , with dimensions of the square of a speed. stret 3 RESULTS where wrinkles play the role of whirls and instead of vor- tex laments and sheets, the localized singularities come in the form of ridges and D-cones. To illustrate this idea Turbulent behavior further, we integrate the system (1) and (2) starting form a at state, with no forcing (F = 0), but with an initial We performed numerical simulations of equations (1) velocity at large-scales and let the system decay. Snap- and (2) by using a standard pseudo-spectral method in a shots of the velocity eld, shown on Fig. 2, exhibit the periodic domain with an additive random forcing at large dynamics of this decaying turbulence con guration. Ini- scales and a viscous damping acting at small scales. The tially, all the velocity uctuations are contained at large explicit form of forcing and dissipation are discussed in scales Fig. 2-a. Then, due to the non-linear mode inter- Methods. Turbulent states should not depend explicitly actions, instabilities appear creating smaller and smaller on their details, provided that these mechanisms are well structures Fig. 2-b. After this transient, a turbulent separated in the wavenumber space. state is observed where uctuations at all scales coexist Figures 1 a-b show the surface plate deformation _ Fig. 2-c. Finally, at large times, the dissipative terms (x; y) and the plate vertical speed  (x; y) respectively. acting at small scales kill the turbulent state, smoothing The displacement eld displays a coarse scale, super- out the uctuations (see Fig. 2-d). imposed with a ne scale uctuations dominated by a large number of wrinkles of various size. Such multi-scale states are the most prominent characteristic of turbulent Kolmogorov spectrum systems. The local speed of the plate in Fig.1-b demon- strates that the elastic plate is not at equilibrium as it By forcing at large scales and dissipating at small displays a myriad of excited modes. When bending is ab- scales, a turbulent out-of-equilibrium steady state is ob- sent, it is natural to expect the appearance of highly non- tained, such as the one observed in Fig. 1. A quantitative linear geometrical structures. Such structures are visible measurement of these steady turbulent states is obtained in Fig. 1-c that shows a close-up of the plate de ection. through the spectral densities of the plate velocity. As The small scale wrinkles consist in a random assembly in uid turbulence, here we compute the average kinetic of moving ridges and conical points. Ridges and conical energy spectrum and the kinetic energy ux for di er- points are in fact the fundamental equilibrium con gu- ent wave numbers. The kinetic energy spectrum per unit rations of elastic plates in the bending-free limit. These mass, E (k), is de ned through the kinetic energy per kin structures have been vastly studied since the pioneering unit mass by works performed in the 90s [29{31]. In the bending-free limit, the elastic deformations favor the bending modes E = E (k) dk; kin kin because they cost no energy. Then, the deformations of the plates are controlled by the stretching that modi es D E the locally plane metric of the sheet. Because of geomet- 2 _ _ where E (k) = 2k j j , the velocity eld  (t) = kin k k rical constrains, it is not always possible to have a fully ikr (r; t)e dr and isotropy is assumed. In addition, the developable surface everywhere. The system thus creates kinetic energy ux P (k) is de ned through the transfer kin singular structures: linear ridges [29] and punctual de- equation: velopable cones or more commonly named D-cones [30], which concentrate the plate curvature and the stretching @ @ E (k) = P (k): (5) kin kin energy. Therefore, the dynamics of a bending-free elas- @t @k tic plate corresponds to a myriad of singularities moving 3 2 Since E (k) has dimensions of L =T the dimensions (apparently) randomly over the sheet. In Supporting In- kin 2 3 of P (k) is L =T . Notice that the stretching energy formation one can see a movie of these defect induced kin (3) is a quadratic quantity, therefore analogous de ni- dynamics. tions can be given for its spectrum and ux (See Meth- To catch precisely these singular structures of dimen- sion 1 (riges) and 2 (D-cones) we have plotted in Fig. 1-d ods for explicit de nitions). We have measured directly the energy uxes for various turbulent runs with di erent the rst order correction to the instantaneous Gaussian forcing amplitudes. Figure 3-a shows the time-averaged curvature: kinetic (solid lines) energy uxes, normalized by their G(x; y; t) =    (4) xx yy xy mean value P in the well de ned transparency win- kin which, after (2), is the source of the in-plane stresses dow where they are at. This transparency window is via the Airy function (x; y; t). The Gaussian curvature usually called the inertial-range in the context of hydro- shows the complex network of ridges and D-cones. dynamic turbulence. At small scales, dissipation takes An immediate analogy can be established with hydro- place and produces a bottleneck that invades the iner- dynamic turbulence. Fully nonlinear elastic plate turbu- tial range as the mean energy ux increases, although an lence seems to be characterized by a \crumpling cascade" inertial range is still clearly present. This bottleneck is 4 a) c) b) d) FIG. 1. Turbulent states of a bending-free plate. a) Surface plot of the plate deformation  (x; y). b) Surface plot of the plate vertical velocity  (x; y). c) Detail of the wrinkles in a zoom of the surface plot of the plate deformation  (x; y). d) Snapshot of the gaussian curvature (4) of the respective area. The density plot displays the values of log jGj. Simulations were made at resolution of 4096 with an additive random forcing, see Methods for more details. corresponding energy spectrum. This is indeed the case for the kinetic energy spectrum, as shown in Fig. 3-b 5=3 where the k Kolmogorov law is apparent. The kinetic turbulent spectrum can be discussed on di- mensional grounds in the footprint of Kolmogorov. How- ever, here in addition to the constant energy ux per unit mass P and the wavenumber k (with dimension inverse of a length), the spectrum should depend a priori on the a) b) additional material parameter E=. Since a characteris- 3=2 tic length can be de ned, namely  = (E=) =P , one gets generically 2=3 3=2 P (E=) E (k) =  k ; (6) kin 5=3 where () is an arbitrary function of the dimensionless c) d) argument k. Contrary to uid turbulence, the existence of this extra length , does not allow to uniquely de- FIG. 2. Simulations for decaying turbulence. Snapshots of  . termine a xed exponent for the power law spectrum. Time goes as: a) t = 0, b) t = 0:5, c) t = 65; and d) t = 100 Remarkably, numerics suggest that the kinetic and the (in time units of (1) and (2)). stretching energy spectra (13) can be treated indepen- dently. Considering that for the kinetic energy the elastic property of the material does not intervene, the kinetic energy spectrum should follow related to dissipative e ects, that have been shown to be 2=3 5=3 non-trivial [22]. Figure 3-a also displays the stretching E (k) = C P k ; (7) kin kin energy uxes (dashed lines), which are notably smaller than the kinetic energy uxes. Therefore, numerical sim- with C = (0) 6= 0 a constant. The numerical simula- kin ulations indicate that the pertinent energy ux for the tions show indeed a good agreement with this predicted cascade is the kinetic energy one, because it shows a con- scalings (7) for both k and P . stant energy ux along scales in an inertial range. A Because the energy (3) combines two di erent contri- constant energy ux is usually associated with a turbu- butions we address the role of the stretching energy. We lent regime that leads to a power law behavior for the show (See methods for the proof ), via a simple argument 5 4 with C a constant proportional to the mean energy, stret that plays the role of an e ective temperature. In the 2.5 inset of Fig. 3 the stretching spectra for di erent forcing amplitude are shown to be consistent with the equipar- tition law 8, for the scales where the dissipation is negli- 1.5 0 1 2 3 gible. 10 10 10 10 Intermittency and beyond Kolmogorov 0.5 phenomenology The Kolmogorov phenomenology discussed in the pre- 0 1 2 3 10 10 10 10 vious section is based on dimensional analysis and mean- eld assumptions that neglect the existence of extreme uctuations. Nevertheless, it is well known that devia- tions exist to such predictions and they become impor- -2 10 tant when looking at higher order statistics [7]. These de- viations are associated with the intermittent statistics of the elds and are somehow inherent to uid turbulence. -4 -2 A complete understanding of such uctuations is still -3 missing. Unlike hydrodynamic turbulence, the theory of -4 -6 weak wave-turbulence predicts Gaussian statistics for the distribution of wave amplitudes, therefore no intermit- 0 1 2 3 10 10 10 10 tency can be observed within the range of validity of the -8 theory. However, when the non-linearities become of the 0 1 2 3 10 10 10 10 same order than the linear dispersive terms, the wave tur- bulence theory breaks down and an intermittent statistics can manifest [32]. Such intermittencies have been experi- FIG. 3. a) Time averaged kinetic energy ux (solid lines) mentally observed in gravity-capillary waves [33] and sug- and the stretching energy ux (dashed lines). Di erent mark- gested for thin elastic plates [34]. Moreover, numerical ers (and colors) are for di erent runs with increasing forcing simulations of elastic plates at strong forcing have shown amplitude or uxes (see arrow in gure b). The uxes have important deviations from the WTT predicted scalings been normalized by their total mean ux value in the inertial for the plate deformation [26, 27], leading to intermit- range (see methods for values). The inset displays the same tency signatures [27, 34]. plot but in log-log. b) Time averaged kinetic energy spectra 2=3 Intermittency is usually addressed looking at the mo- normalized by P . The dashed line displays the Kolmogorov ments of the so-called structure functions of the elds, scaling. Inset: Time averaged stretching energy. The dashed line displays the thermalization energy scaling. The arrow that provide information of the variation of the elds at indicates the di erent runs with increasing values of the ux a given scale [7]. For elastic plates, because of the fast P . The value of P varies in the range (0:03; 16:5) for all kin decay of the deformation spectrum, a second order di er- the runs (see Methods). ence is needed to observe an intermittent scaling [27]. In our system, where bending waves are absent and the dy- namics is thus fully non-linear, the intermittent behavior that the stretching energy ux related to the stretch- is expected to occur at any forcing. Following [27], we in- ing energy must vanish in a statistically steady state un- troduce the (second variation) increments   of the plate der an additive forcing and a viscous-like dissipation, as deformation and their corresponding structure functions the one considered in this work. This result follows di- S (`), namely rectly from the observation that the stretching energy 2 p 2 ux appears to be the time derivative of the stress corre- S (`) = hj j i with   =  (x + `) 2 (x) +  (x (9) `): ` ` lation function which must be zero in steady state turbu- At very small scales, the regularity of  (x) implies lent regime. Therefore, one would expect the stretching 2p the scaling S (`)  ` . However, in the inertial range modes to eventually thermalize. Since the stretching en- p 1 2 a non-trivial scaling can appear. We de ne then the ergy is quadratic in =     (see (3)), xx yy xy anomalous exponents (in the spirit of hydrodynamic tur- the equilibrium distribution corresponds to equipartition bulence)  , from the moments following S (`)  ` : of the Fourier modes of , leading to the equipartition p p From our numerics we measure that   2:2. In Kol- spectrum of the form: 2 mogorov theory as well as for WWT, a linear relation E (k) = C k; (8) between the di erent exponent is expected  = p, stret stret p 2 6 early with the distance `. This di erence can be ex- 2 plained by the existence of the extra-length , suggest- -0.5 ing that the exponent can not be uniquely determined by dimensional analysis. This fact is also related to the -1 complexity of the non-linear term, that is precisely taken -3 -2 -1 10 10 10 ℓ/L into account in (10). p = 0.1 p = 0.5 p = 1 p = 2 2 4 p = 3 DISCUSSION p = 4 p = 5 p = 6 -3 -2 -1 10 10 10 Substantial evidence shows robustness among turbu- ℓ/L lent behavior in ordinary uids and in the case of crum- 0 1 2 3 4 5 6 p pling vibrations of an elastic plate in the zero limit thick- ness. As presented, both manifest an energy cascade with 2=3 5=3 the well known K41 spectrum E  P k . More FIG. 4. Intermittency exponents  as a function of p showing k clearly a departure from a linear law, indicating the existence important, a myriad of randomly interacting highly non- of anomalous exponents and intermittency. The right down linear crumpling structures (folds, ridges and d-cones) at inset shows the structure functions local slopes as a function all relevant scales dominate the dynamics. They induce of `. The inertial range is delimited by the two vertical green intermittency that we have quantitatively investigated lines. The left-up panel shows evidence of the 1-law (10). by studying high-order correlators that con rm the ap- The numerics was done under the same conditions as previous pearance of an intermittent behavior. Nevertheless, the gures. underlying plate dynamics di ers notably from the one of ordinary uids: though in incompressible uids the total energy consists purely of kinetic energy, in elastic plates witnessing the self-similar nature of the dynamics. For the energy is compound of two separately (positive) ener- vibrating elastic plates however, a deviation from this gies: the kinetic and the stretching energy. Therefore, it linear law has been observed, indicating thus clearly the is expected that two di erent cascades may exist in the presence of intermittency that breaks the self-similarity elastic situation, a kinetic energy one and a stretching of crumpling dynamics [27]. In the bottom inset of Fig. energy one. Because the forcing is additive to the defor- 4, we present the local slope of the structure functions d log S (`) mation acceleration (1), only a constant kinetic energy de ned as  (`) = for di erent orders p. A rel- d log ` ux (see Fig. 3-a) exists resulting in a K41 cascade. On atively at behavior is observed in the inertial range and the other hand, the stretching energy ux vanishes (see the anomalous exponents  are measured by averaging Fig. 3-a) and consequently Gibbs equipartition is ob- (`) in this window. Fig. 4 shows  as a function of p p served for the stretching spectrum, as shown in the inset p, exhibiting a clear departure from the linear behavior of Fig. 3-b. Naturally it is expected that other forcing (dashed line) and thus intermittent statistics. Moreover, may display a stretching energy cascade without any ki- we have veri ed the exact result derived by two of us for netic energy cascade. Or perhaps two distinct cascades thin elastic plates that is valid for both, weak and strong a kinetic energy K41 cascade and a stretching cascade wave turbulence [17]. This result is the equivalent to the simultaneously. only exact result of hydrodynamic turbulence known as the 4=5-Kolmogorov law of turbulence. It relates incre- ments of the eld with the energy ux and the scales. APPENDIX In the case of a thin elastic plate, a similar relation was found for a di erent structure function, called Z (`), that depends on the rst order variation of the elds ,  and In the following we describe the mathematical and nu- (see Methods). In analogy with the 4=5-Kolmogorov merical methods used in the present paper. law, this relation was called the 1-law and reads Z (`) = P`: (10) Numerics The structure function Z (`), normalized by P` is dis- We solve the F oppl{von K arm an equations (1)-(2) with played in (top) inset of Fig. 4. An excellent agreement a standard pseudo-spectral code in a square domain of with the prediction (10) is observed over a decade with no size 2 with periodic boundary conditions. The forcing adjustable parameter. Finally, we have also studied the F is white-noise in time of variance f and its Fourier _ _ vertical speed structure function  (r + `)  (r) . modes are non-zero only for wave-vectors 1:5  jkj  3. Unlike, hydrodynamic turbulence, it does not scales lin- De-aliasing is made by using the standard 2=3-rule, that Z(ℓ)/(P ℓ) ξ (ℓ) p 7 is applied after computing each quadratic term. The dis- Steady state with zero stretching energy ux sipation is acting at small and large scales in the following way Following [17], we introduce the spatial correlation function n n _ _ D = ()  ()  (11) E (`) = h (r) 0(r )i; (16) stret r r 2E The run presented in Fig. 2 was performed with 1024 6 0 collocation points, F = 0, n = 2,  = 4 10 , = 0 with ` = r r, which on the limit of ` ! 0 converges and a random initial condition at large scale only on  . to the mean value of the stretching energy. Consider- All the other runs presented in this work were performed ing an homogeneous and isotropic system, then the time 2 12 with 4096 collocation points, n = 3,  = 2  10 , derivative of the correlation function (16) can be shown n = 2, = 100 and f = 1; 8; 27 and 100. For the di er- to be ent forcing amplitudes, the measured value of the uxes _ _ _ E (`) = h(r )f (r);  (r)gi = (r + `)f (r);  (r)(17) gdr: were 0:0284; 0:0354; 5:3358 and 16:5325 respectively. stret The last equality considers that for an homogeneous sys- tem the statistical average can be taken as a spatial av- The energy spectra erage. From the convolution theorem the Fourier trans- _ _ form E (p) =  f; g . In a steady state the time stret p p The energies in (3) maybe seen as quadratic contri- derivative of the correlation function must be zero hence 1 2 bution of  and (r) =     .The nal xx yy _ xy f; g = 0 and the stretching energy ux (15) must p p quadratic contributions read be zero. Z Z 1 E 2 2 E = j j dk and E = j j dk:(12) kin k stret k 2 2 1-law for thin elastic plate turbulence We de ne the kinetic energy and the stretching spectra In Ref. [17] an exact result for turbulence of thin elastic by averaging over the angular variables in the wavenum- plates was found. It provides a K arm an-Howarth-Monin ber space, hence type relationship for the energy ux: D E 2 2 _ 1 E (k) = k j j and E (k) =  k j j(13) ; kin k stret k _ r hJ i = P; (18) ` [;] where  stands for the rst di erence of the eld, e.g. where we have de ned h: : :i = : : : d' by the angu- =  (x + `)  (x). The vector J is de ned by lar average. J = f h f h ; f h f h : y yx x yy x xy y xx [f;h] The energy uxes Under the assumption of isotropy, (18) implies the 1-law (10) for the structure function Because the energy is quadratic in  and , the energy _ ^ Z (`)  hJ i ` = P `; (19) ux can be straightforward de ned as in hydrodynamic [;] turbulence. By making a scale-by-scale energy budget where ` is the unit vector along `. the energy uxes are: GD, GK and SR thank to FONDECYT grants N Z Z k k D E 1181382 and the Chilean-French scienti c exchange pro- @E (p) kin P (k) = dp =  f; g (14) dp; kin p p gram ECOS-Sud/CONICYT No. C14E04. @t 0 0 Z Z k k D E @E (p) stret P (k) = dp =  f; g (15) dp; stret p p @t 0 0 where the subscript 0 stands for the free of forcing [1] Reynolds O (1883) Phil. Trans. Roy. Soc. 174: 935. [2] Kolmogorov AN (1941) Dokl. Akad. Nauk SSSR 32: 15- and dissipation time variations of the elds through (1) 17 [English: Reprinted by Kolmogorov AN (1991) Proc. with F = D = 0. For simplicity, we de ne ff; gg R. Soc. London 434: 15-17]. f g + f g 2f g . We have also used the fact xx yy yy xx xy xy [3] Cardesa JI, Vela-Mart n A, Jim enez, J (2017) Science 1 1 that the angular averages satisfy ( f ) ( g) = p p 357: 782-784. ( f ) g . The total energy ux results from the p p [4] Castelvecchi D (2017) Nature 548: 382-383. addition of these previous expressions. [5] Ruelle DP (2012) PNAS 109: 20344-20346. 8 [6] von K arm an T, Howarth L (1938) Proc. R. Soc. London [18] F oppl A (1907) Vorlesungen ub  er technische Mechanik, 164: 192-216. Bd. 5, Leipzig p. 132; von K arm an T (1910) Ency. d. [7] Frisch U (1995) Turbulence: The legacy of A.N. Kol- math. Wiss., Bd. IV. 2, II, Leipzig, x8. mogorov. (Cambridge University Press, Cambridge). [19] During  G, Josserand C, Rica S (2017) Physica D 347: [8] Hasselmann K (1962) J. Fluid Mech. 12: 481-500; (1963) 42-73. J. Fluid Mech. 15: 273-281; (1963) J. Fluid Mech. 15: [20] Boudaoud A, Cadot O, Odille B, Touz e C (2008) Phys. 385-398. Rev. Lett. 100: 234504. [9] Benney DJ, Sa man PG (1966) Proc. Roy. Soc. London [21] Mordant N (2008) Phys. Rev. Lett. 100: 234505. A 289: 301-320. [22] Humbert T, Cadot O, During  G, Josserand C, Rica S, [10] Zakharov VE, Filonenko NN Dokl. Akad. Nauk SSSR Touz e C (2013) Euro. Phys. Lett. 102: 30002. 170 (1966): 1292 [English transl. in Sov. Math. Dokl.]; [23] Du ring G, Josserand C, Rica S (2015) Phys. Rev. E 91: Zakharov VE (1966) Zh. Eksper. Teoret. Fiz. 51: 686 052916. [English transl. in (1967) Sov. Phys. JETP 24: 455]; [24] Humbert T, Cadot O, Josserand C, Touz e C (2016) Phys- Zakharov VE, Filonenko NN (1967) Zh. Prikl. Mekh. I ica D 316: 34. Tekn. Fiz. 5: 62 [English transl. in J. Appl. Mech. Tech. [25] Miquel B, Alexakis A, Josserand C, Mordant N (2013) Phys.]; Zakharov VE, L'vov VS, G. Falkovich G (1992) Phys. Rev. Lett. 111: 054302. Kolmogorov Spectra of Turbulence I (Springer, Berlin). [26] Yokoyama N, Takaoka M. (2013) Phys. Rev. Lett. 110: [11] Benney DJ, Newell AC (1967) J. Math. Phys. 46: 363; 105501. Newell AC (1968) Rev. Geophys. 6: 1-31; Benney DJ, [27] Chibbaro S, Josserand C (2016) Phys. Rev. E 94: Newell AC. (1969) Stud. Appl. Math. 48, 29-53. 011101(R). [12] Newell AC, Rumpf B (2011) Annual Review of Fluid Me- [28] Landau LD, Lifshitz EM (1959) Theory of Elasticity chanics 43: 59-78. (Pergamon, New York). [13] Nazarenko S. (2011) Wave turbulence, Lecture Notes in [29] Witten TA, Li H (1993) Europhys. Lett. 23: 51-55. Physics Vol. 825, (Springer Berlin) . [30] Ben Amar M, Pomeau Y (1997) Proc. R. Soc. A 453: [14] Dyachenko S, Newell AC, Pushkarev A, Zakharov VE 729-755 (1992) Physica D 57: 96. [31] Cerda E, Mahadevan L (1998) Phys. Rev. Lett. 80: 2358- [15] Durin  g G, Josserand C, Rica S (2006) Phys. Rev. Lett. 2361. 97: 025503. [32] Newell AC, Nazarenko S, Biven L (2001) Physica D, 152- [16] Galtier S and Nazarenko S (2017) Phys. Rev. Lett. 119: 153: 520-550. 221101. [33] Falcon E, Fauve S, Laroche C (2007) Phys. Rev. Lett. 98: [17] Durin  g G, Krstulovic G (2018) Phys. Rev. E 97: 154501. 020201(R). [34] Mordant N, Miquel B. (2017) Phys. Rev. E 96, 042204. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nonlinear Sciences arXiv (Cornell University)

A Kolmogorov spectrum for strongly vibrating plates

Download PDF
8 pages

Loading next page...
 
/lp/arxiv-cornell-university/a-kolmogorov-spectrum-for-strongly-vibrating-plates-XSh62VSds4

References (33)

ISSN
2469-990X
eISSN
ARCH-3338
DOI
10.1103/PhysRevFluids.4.064804
Publisher site
See Article on Publisher Site

Abstract

Is turbulence universal ? A Kolmogorov spectrum for strongly vibrating plates. Gustavo During  Facultad de F sica, PUC, Chile. Christophe Josserand LadHyX, CNRS & Ecole Polytechnique, UMR 7646, 91128, Palaiseau, France. Giorgio Krstulovic Universit e C^ ote d'Azur, Observatoire de la C^ ote d'Azur, CNRS, Laboratoire Lagrange, Bd de l'Observatoire, CS 34229, 06304 Nice cedex 4, France. Sergio Rica Facultad de Ingenier a y Ciencias, Universidad Adolfo Ib anez, ~ Santiago, Chile. In uid turbulence, energy is transferred from a scale to another by an energy cascade that depends only on the energy dissipation rate. It leads by dimensional arguments to the Kolmogorov 1941 (K41) spectrum. Remarkably the normal modes of vibrations in elastic plates manifests an energy cascade with the same K41 spectrum in the fully non-linear regime. Moreover, the elastic deformations present large \eddies" together with a myriad of small \crumpling eddies", such that folds, developable cones, and more complex stretching structures, in close analogy with spots, swirls, vortices and other structures in hydrodynamic turbulence. We characterize the energy cascade, the validity of the constant energy dissipation rate over the scales and the role of intermittency via the correlation functions. Turbulence has remained a central problem in uid dy- sions. Similarly, uid turbulence is a hard problem be- namics since the early experiments of Osborne Reynolds cause it is in the same footing that the Euler equation [1]. Perhaps the most salient feature is the setting of for perfect uids which is ab-initio a nonlinear partial an energy cascade that redistributes the energy among di erential equation without any small parameter that di erent Fourier modes of the velocity uctuations that may justify an asymptotic scheme nor a rational closure are independent of the dissipation caused by viscosity at for the statistical correlations [7]. Since uid turbulence small scales. This energy cascade depends only on the is studied in the framework of the NS equations, it is energy dissipation rate per unit mass, P and the wave important to question whether such cascade dynamics number k leading to the Kolmogorov K41 [2] spectrum: transporting the energy from the large injection scales 2=3 5=3 E  P k . Despite a century of e ort, turbulence to the small dissipative ones can be observed in di erent still remains nowadays a major challenge from the exper- contexts and whether the turbulence features described imental, theoretical and numerical points of view, and in uids are universal or dependent on the dynamical the very essence of the phenomena has not yet been re- equations. In particular, while the quadratic nonlinear vealed satisfactorily [3{5]. For instance, if the K41 law term of the NS equation is responsible of the mixing be- can be deduced using dimensional arguments, the statis- tween scales at the heart of the cascade process, noth- tical description of the turbulence fails to describe cor- ing in the theoretical description of turbulence takes into rectly the general correlation functions of the velocity account the details of the nonlinearity itself. In fact, elds. In fact, up to date, there is only a single analyt- turbulence has been observed in di erent contexts such ical result that maybe derived from the original Navier- as Magneto-Hydrodynamics where the magnetic eld is Stokes (NS) equations for incompressible uids, the so- coupled with the ow or viscoelastic turbulence in poly- called von K arm an-Howarth relation [6], that links the mer solutions at low Reynolds numbers for instance. In second and the third order correlations function of some these cases even though the nonlinearity di ers from the component of the uid velocity with the energy dissipa- one of NS equations and di erent spectra can be mea- tion rate. Higher moments of these quantities di er from sured, they have similar properties than the original Kol- mean- eld based predictions, and experimental as well as mogorov phenomenology. numerical data reveal extreme events, exhibiting heavy On the other hand, a di erent class of turbulence ex- tailed distributions, that are usually associated with in- ists when a linear term is present in the dynamics, the termittency. Moreover, one can argue that the diculty so-called wave turbulence. Wave propagation dominates of describing turbulence may come from the fundamen- at linear order and the waves interact through the non- tal issues raised by the uid mechanics equations, wit- linear terms. The dynamics is described by the weak nessed in particular by the Clay millennium problem on or wave turbulence theory (WTT), developed originally the regularity of the NS equation in three space dimen- in the sixties [8{11]. WTT is deduced in the limit of arXiv:1808.02025v1 [nlin.CD] 6 Aug 2018 2 weak wave amplitudes, so that the nonlinear terms can ber of studies in vibrating plates (for a review see [19]). be treated perturbatively, by contrast with uid turbu- Experiments performed soon after the theoretical pre- lence where the nonlinearity is present at rst order in dictions have shown slightly di erent spectrum power the dynamics [10, 12, 13]. Based on the long-time statis- laws [20, 21], that have been later explained by the spe- tics of randomly uctuating interacting waves, the WTT ci c features of the dissipation [22]. Furthermore, wave deduces a kinetic equation for the distribution of spectral turbulence of plates has shown to be a perfect system densities. Beside thermodynamical equilibrium solution, for investigating di erent concepts such as inverse cas- the key-feature of this equation is that non-equilibrium cade [23] and transitory dynamics [24]. More recently, turbulent stationary solutions also arise, called in general thin plates with a high forcing [25, 26], the breakdown Kolmogorov-Zakharov spectra (KZ). Indeed, as shown of the WTT and the onset of intermittency [27] have by Zakharov [10], they describe a constant ux transfer been extensively studied. In these high forcing regimes, (or cascade) of conserved quantities (e.g. energy) be- a wave turbulence spectrum is still observed at small tween large and small length scales. Examples of wave scales, while strong nonlinear regimes appear at large turbulence go much beyond the cases of surface gravity scales, whose nature remain unclear, suggesting the exis- or capillary waves and concerns systems as diverse as tence of a strong turbulent dynamics. plasma waves [10], nonlinear optics [14], vibrating elas- In the following, to investigate this strongly nonlinear tic plates [15] and gravitational waves in general rela- regime and the analogy with hydrodynamic turbulence, tivity [16], among others. Although the analogy with we will consider a thin elastic plate in the fully nonlinear turbulent ows is natural since it describes a ux of case which corresponds to the formal limit with no bend- the energy between scales, the links within the physi- ing and where linear waves are absent. With no surprise, cal mechanisms is less straightforward: in particular, the the WTT spectrum vanishes in this limit and a di erent WTT is based on the resonant interactions between the turbulent spectrum is expected to appear. The goal of waves prescribed by the nonlinearity so that the power the paper is to characterize such energy repartition and of the nonlinearity is crucial in the process. For instance highlight similarities with hydrodynamic turbulence. if the pertinent nonlinearity is quadratic, the dynamics is determined by three waves interactions such that both the wave-numbers and the frequencies of the three waves THEORETICAL MODEL form resonant triads. This mechanism implies a speci c dependence of the KZ spectrum with the energy ux P as The vibration of a bending-free elastic plate comes 1=N P where N is the power of the pertinent nonlinearity from the usual dynamical version of the F oppl{von of the dynamics (N = 2 for quadratic interactions, 3 for K arm an equations [28] for the vertical amplitude of the cubic, etc). This represents a crucial di erence between deformation  (x; y; t) and for the Airy stress function wave turbulence and hydrodynamic turbulence, because (x; y; t) that describes the three in{plane stresses: in the last one, the ux of energy appears in the second order correlation with the power 2=3, which cannot be obtained as a 1=N power. =   +   2  +F +D; (1) xx yy yy xx xy xy @t The turbulent asymptotic expansion of wave systems 2 2 in the strongly nonlinear limit is, by construction, sin-   =    : (2) xx yy xy gular, because the linear term becomes then subdomi- nant. However, a turbulent fully nonlinear dynamics is Here  = @ + @ is the usual Laplacian and  and E xx yy still expected to exist. The dependence of the statis- are, respectively, the mass density and the Young modu- tical properties with the potentially di erent nonlinear lus E of the plate. F and D are the forcing and dissipa- terms and the connexions with the NS uid turbulence tion respectively. is then worth to be investigated. In this paper, we pre- When F = D = 0, the equations (1) and (2) derive cisely address such singular limit in the case of vibrating from a Hamiltonian principle [15, 19]. More important, elastic plates. This system has indeed recently revealed it is a well posed system, in the sense that the energy a strong analogy with hydrodynamic turbulence through is compound by two positive (hence bound from below) the derivation of an exact law for a two-point correlation quantities, namely the kinetic energy and the stretching that is the equivalent of the Kolmogorov's 4/5-law [17]. energy per unit mass: The weakly nonlinear turbulence regime in vibrating Z Z elastic plates has been established ten years ago by three 1 E 2 1 2 E =  dr; E =     dr; kin stret xx yy xy of us [15]. Based on the dynamical F oppl{von K arm an 2S 8S equations [18] (described below), we have shown using (3) the WTT the existence of a spectrum of direct energy here S is the area plate. For dimensional purposes, we cascade. The rather simple comparison between the- use everywhere the energies per unit mass, E , E , and kin ory, numerics and experiments has led to a large num- E , with dimensions of the square of a speed. stret 3 RESULTS where wrinkles play the role of whirls and instead of vor- tex laments and sheets, the localized singularities come in the form of ridges and D-cones. To illustrate this idea Turbulent behavior further, we integrate the system (1) and (2) starting form a at state, with no forcing (F = 0), but with an initial We performed numerical simulations of equations (1) velocity at large-scales and let the system decay. Snap- and (2) by using a standard pseudo-spectral method in a shots of the velocity eld, shown on Fig. 2, exhibit the periodic domain with an additive random forcing at large dynamics of this decaying turbulence con guration. Ini- scales and a viscous damping acting at small scales. The tially, all the velocity uctuations are contained at large explicit form of forcing and dissipation are discussed in scales Fig. 2-a. Then, due to the non-linear mode inter- Methods. Turbulent states should not depend explicitly actions, instabilities appear creating smaller and smaller on their details, provided that these mechanisms are well structures Fig. 2-b. After this transient, a turbulent separated in the wavenumber space. state is observed where uctuations at all scales coexist Figures 1 a-b show the surface plate deformation _ Fig. 2-c. Finally, at large times, the dissipative terms (x; y) and the plate vertical speed  (x; y) respectively. acting at small scales kill the turbulent state, smoothing The displacement eld displays a coarse scale, super- out the uctuations (see Fig. 2-d). imposed with a ne scale uctuations dominated by a large number of wrinkles of various size. Such multi-scale states are the most prominent characteristic of turbulent Kolmogorov spectrum systems. The local speed of the plate in Fig.1-b demon- strates that the elastic plate is not at equilibrium as it By forcing at large scales and dissipating at small displays a myriad of excited modes. When bending is ab- scales, a turbulent out-of-equilibrium steady state is ob- sent, it is natural to expect the appearance of highly non- tained, such as the one observed in Fig. 1. A quantitative linear geometrical structures. Such structures are visible measurement of these steady turbulent states is obtained in Fig. 1-c that shows a close-up of the plate de ection. through the spectral densities of the plate velocity. As The small scale wrinkles consist in a random assembly in uid turbulence, here we compute the average kinetic of moving ridges and conical points. Ridges and conical energy spectrum and the kinetic energy ux for di er- points are in fact the fundamental equilibrium con gu- ent wave numbers. The kinetic energy spectrum per unit rations of elastic plates in the bending-free limit. These mass, E (k), is de ned through the kinetic energy per kin structures have been vastly studied since the pioneering unit mass by works performed in the 90s [29{31]. In the bending-free limit, the elastic deformations favor the bending modes E = E (k) dk; kin kin because they cost no energy. Then, the deformations of the plates are controlled by the stretching that modi es D E the locally plane metric of the sheet. Because of geomet- 2 _ _ where E (k) = 2k j j , the velocity eld  (t) = kin k k rical constrains, it is not always possible to have a fully ikr (r; t)e dr and isotropy is assumed. In addition, the developable surface everywhere. The system thus creates kinetic energy ux P (k) is de ned through the transfer kin singular structures: linear ridges [29] and punctual de- equation: velopable cones or more commonly named D-cones [30], which concentrate the plate curvature and the stretching @ @ E (k) = P (k): (5) kin kin energy. Therefore, the dynamics of a bending-free elas- @t @k tic plate corresponds to a myriad of singularities moving 3 2 Since E (k) has dimensions of L =T the dimensions (apparently) randomly over the sheet. In Supporting In- kin 2 3 of P (k) is L =T . Notice that the stretching energy formation one can see a movie of these defect induced kin (3) is a quadratic quantity, therefore analogous de ni- dynamics. tions can be given for its spectrum and ux (See Meth- To catch precisely these singular structures of dimen- sion 1 (riges) and 2 (D-cones) we have plotted in Fig. 1-d ods for explicit de nitions). We have measured directly the energy uxes for various turbulent runs with di erent the rst order correction to the instantaneous Gaussian forcing amplitudes. Figure 3-a shows the time-averaged curvature: kinetic (solid lines) energy uxes, normalized by their G(x; y; t) =    (4) xx yy xy mean value P in the well de ned transparency win- kin which, after (2), is the source of the in-plane stresses dow where they are at. This transparency window is via the Airy function (x; y; t). The Gaussian curvature usually called the inertial-range in the context of hydro- shows the complex network of ridges and D-cones. dynamic turbulence. At small scales, dissipation takes An immediate analogy can be established with hydro- place and produces a bottleneck that invades the iner- dynamic turbulence. Fully nonlinear elastic plate turbu- tial range as the mean energy ux increases, although an lence seems to be characterized by a \crumpling cascade" inertial range is still clearly present. This bottleneck is 4 a) c) b) d) FIG. 1. Turbulent states of a bending-free plate. a) Surface plot of the plate deformation  (x; y). b) Surface plot of the plate vertical velocity  (x; y). c) Detail of the wrinkles in a zoom of the surface plot of the plate deformation  (x; y). d) Snapshot of the gaussian curvature (4) of the respective area. The density plot displays the values of log jGj. Simulations were made at resolution of 4096 with an additive random forcing, see Methods for more details. corresponding energy spectrum. This is indeed the case for the kinetic energy spectrum, as shown in Fig. 3-b 5=3 where the k Kolmogorov law is apparent. The kinetic turbulent spectrum can be discussed on di- mensional grounds in the footprint of Kolmogorov. How- ever, here in addition to the constant energy ux per unit mass P and the wavenumber k (with dimension inverse of a length), the spectrum should depend a priori on the a) b) additional material parameter E=. Since a characteris- 3=2 tic length can be de ned, namely  = (E=) =P , one gets generically 2=3 3=2 P (E=) E (k) =  k ; (6) kin 5=3 where () is an arbitrary function of the dimensionless c) d) argument k. Contrary to uid turbulence, the existence of this extra length , does not allow to uniquely de- FIG. 2. Simulations for decaying turbulence. Snapshots of  . termine a xed exponent for the power law spectrum. Time goes as: a) t = 0, b) t = 0:5, c) t = 65; and d) t = 100 Remarkably, numerics suggest that the kinetic and the (in time units of (1) and (2)). stretching energy spectra (13) can be treated indepen- dently. Considering that for the kinetic energy the elastic property of the material does not intervene, the kinetic energy spectrum should follow related to dissipative e ects, that have been shown to be 2=3 5=3 non-trivial [22]. Figure 3-a also displays the stretching E (k) = C P k ; (7) kin kin energy uxes (dashed lines), which are notably smaller than the kinetic energy uxes. Therefore, numerical sim- with C = (0) 6= 0 a constant. The numerical simula- kin ulations indicate that the pertinent energy ux for the tions show indeed a good agreement with this predicted cascade is the kinetic energy one, because it shows a con- scalings (7) for both k and P . stant energy ux along scales in an inertial range. A Because the energy (3) combines two di erent contri- constant energy ux is usually associated with a turbu- butions we address the role of the stretching energy. We lent regime that leads to a power law behavior for the show (See methods for the proof ), via a simple argument 5 4 with C a constant proportional to the mean energy, stret that plays the role of an e ective temperature. In the 2.5 inset of Fig. 3 the stretching spectra for di erent forcing amplitude are shown to be consistent with the equipar- tition law 8, for the scales where the dissipation is negli- 1.5 0 1 2 3 gible. 10 10 10 10 Intermittency and beyond Kolmogorov 0.5 phenomenology The Kolmogorov phenomenology discussed in the pre- 0 1 2 3 10 10 10 10 vious section is based on dimensional analysis and mean- eld assumptions that neglect the existence of extreme uctuations. Nevertheless, it is well known that devia- tions exist to such predictions and they become impor- -2 10 tant when looking at higher order statistics [7]. These de- viations are associated with the intermittent statistics of the elds and are somehow inherent to uid turbulence. -4 -2 A complete understanding of such uctuations is still -3 missing. Unlike hydrodynamic turbulence, the theory of -4 -6 weak wave-turbulence predicts Gaussian statistics for the distribution of wave amplitudes, therefore no intermit- 0 1 2 3 10 10 10 10 tency can be observed within the range of validity of the -8 theory. However, when the non-linearities become of the 0 1 2 3 10 10 10 10 same order than the linear dispersive terms, the wave tur- bulence theory breaks down and an intermittent statistics can manifest [32]. Such intermittencies have been experi- FIG. 3. a) Time averaged kinetic energy ux (solid lines) mentally observed in gravity-capillary waves [33] and sug- and the stretching energy ux (dashed lines). Di erent mark- gested for thin elastic plates [34]. Moreover, numerical ers (and colors) are for di erent runs with increasing forcing simulations of elastic plates at strong forcing have shown amplitude or uxes (see arrow in gure b). The uxes have important deviations from the WTT predicted scalings been normalized by their total mean ux value in the inertial for the plate deformation [26, 27], leading to intermit- range (see methods for values). The inset displays the same tency signatures [27, 34]. plot but in log-log. b) Time averaged kinetic energy spectra 2=3 Intermittency is usually addressed looking at the mo- normalized by P . The dashed line displays the Kolmogorov ments of the so-called structure functions of the elds, scaling. Inset: Time averaged stretching energy. The dashed line displays the thermalization energy scaling. The arrow that provide information of the variation of the elds at indicates the di erent runs with increasing values of the ux a given scale [7]. For elastic plates, because of the fast P . The value of P varies in the range (0:03; 16:5) for all kin decay of the deformation spectrum, a second order di er- the runs (see Methods). ence is needed to observe an intermittent scaling [27]. In our system, where bending waves are absent and the dy- namics is thus fully non-linear, the intermittent behavior that the stretching energy ux related to the stretch- is expected to occur at any forcing. Following [27], we in- ing energy must vanish in a statistically steady state un- troduce the (second variation) increments   of the plate der an additive forcing and a viscous-like dissipation, as deformation and their corresponding structure functions the one considered in this work. This result follows di- S (`), namely rectly from the observation that the stretching energy 2 p 2 ux appears to be the time derivative of the stress corre- S (`) = hj j i with   =  (x + `) 2 (x) +  (x (9) `): ` ` lation function which must be zero in steady state turbu- At very small scales, the regularity of  (x) implies lent regime. Therefore, one would expect the stretching 2p the scaling S (`)  ` . However, in the inertial range modes to eventually thermalize. Since the stretching en- p 1 2 a non-trivial scaling can appear. We de ne then the ergy is quadratic in =     (see (3)), xx yy xy anomalous exponents (in the spirit of hydrodynamic tur- the equilibrium distribution corresponds to equipartition bulence)  , from the moments following S (`)  ` : of the Fourier modes of , leading to the equipartition p p From our numerics we measure that   2:2. In Kol- spectrum of the form: 2 mogorov theory as well as for WWT, a linear relation E (k) = C k; (8) between the di erent exponent is expected  = p, stret stret p 2 6 early with the distance `. This di erence can be ex- 2 plained by the existence of the extra-length , suggest- -0.5 ing that the exponent can not be uniquely determined by dimensional analysis. This fact is also related to the -1 complexity of the non-linear term, that is precisely taken -3 -2 -1 10 10 10 ℓ/L into account in (10). p = 0.1 p = 0.5 p = 1 p = 2 2 4 p = 3 DISCUSSION p = 4 p = 5 p = 6 -3 -2 -1 10 10 10 Substantial evidence shows robustness among turbu- ℓ/L lent behavior in ordinary uids and in the case of crum- 0 1 2 3 4 5 6 p pling vibrations of an elastic plate in the zero limit thick- ness. As presented, both manifest an energy cascade with 2=3 5=3 the well known K41 spectrum E  P k . More FIG. 4. Intermittency exponents  as a function of p showing k clearly a departure from a linear law, indicating the existence important, a myriad of randomly interacting highly non- of anomalous exponents and intermittency. The right down linear crumpling structures (folds, ridges and d-cones) at inset shows the structure functions local slopes as a function all relevant scales dominate the dynamics. They induce of `. The inertial range is delimited by the two vertical green intermittency that we have quantitatively investigated lines. The left-up panel shows evidence of the 1-law (10). by studying high-order correlators that con rm the ap- The numerics was done under the same conditions as previous pearance of an intermittent behavior. Nevertheless, the gures. underlying plate dynamics di ers notably from the one of ordinary uids: though in incompressible uids the total energy consists purely of kinetic energy, in elastic plates witnessing the self-similar nature of the dynamics. For the energy is compound of two separately (positive) ener- vibrating elastic plates however, a deviation from this gies: the kinetic and the stretching energy. Therefore, it linear law has been observed, indicating thus clearly the is expected that two di erent cascades may exist in the presence of intermittency that breaks the self-similarity elastic situation, a kinetic energy one and a stretching of crumpling dynamics [27]. In the bottom inset of Fig. energy one. Because the forcing is additive to the defor- 4, we present the local slope of the structure functions d log S (`) mation acceleration (1), only a constant kinetic energy de ned as  (`) = for di erent orders p. A rel- d log ` ux (see Fig. 3-a) exists resulting in a K41 cascade. On atively at behavior is observed in the inertial range and the other hand, the stretching energy ux vanishes (see the anomalous exponents  are measured by averaging Fig. 3-a) and consequently Gibbs equipartition is ob- (`) in this window. Fig. 4 shows  as a function of p p served for the stretching spectrum, as shown in the inset p, exhibiting a clear departure from the linear behavior of Fig. 3-b. Naturally it is expected that other forcing (dashed line) and thus intermittent statistics. Moreover, may display a stretching energy cascade without any ki- we have veri ed the exact result derived by two of us for netic energy cascade. Or perhaps two distinct cascades thin elastic plates that is valid for both, weak and strong a kinetic energy K41 cascade and a stretching cascade wave turbulence [17]. This result is the equivalent to the simultaneously. only exact result of hydrodynamic turbulence known as the 4=5-Kolmogorov law of turbulence. It relates incre- ments of the eld with the energy ux and the scales. APPENDIX In the case of a thin elastic plate, a similar relation was found for a di erent structure function, called Z (`), that depends on the rst order variation of the elds ,  and In the following we describe the mathematical and nu- (see Methods). In analogy with the 4=5-Kolmogorov merical methods used in the present paper. law, this relation was called the 1-law and reads Z (`) = P`: (10) Numerics The structure function Z (`), normalized by P` is dis- We solve the F oppl{von K arm an equations (1)-(2) with played in (top) inset of Fig. 4. An excellent agreement a standard pseudo-spectral code in a square domain of with the prediction (10) is observed over a decade with no size 2 with periodic boundary conditions. The forcing adjustable parameter. Finally, we have also studied the F is white-noise in time of variance f and its Fourier _ _ vertical speed structure function  (r + `)  (r) . modes are non-zero only for wave-vectors 1:5  jkj  3. Unlike, hydrodynamic turbulence, it does not scales lin- De-aliasing is made by using the standard 2=3-rule, that Z(ℓ)/(P ℓ) ξ (ℓ) p 7 is applied after computing each quadratic term. The dis- Steady state with zero stretching energy ux sipation is acting at small and large scales in the following way Following [17], we introduce the spatial correlation function n n _ _ D = ()  ()  (11) E (`) = h (r) 0(r )i; (16) stret r r 2E The run presented in Fig. 2 was performed with 1024 6 0 collocation points, F = 0, n = 2,  = 4 10 , = 0 with ` = r r, which on the limit of ` ! 0 converges and a random initial condition at large scale only on  . to the mean value of the stretching energy. Consider- All the other runs presented in this work were performed ing an homogeneous and isotropic system, then the time 2 12 with 4096 collocation points, n = 3,  = 2  10 , derivative of the correlation function (16) can be shown n = 2, = 100 and f = 1; 8; 27 and 100. For the di er- to be ent forcing amplitudes, the measured value of the uxes _ _ _ E (`) = h(r )f (r);  (r)gi = (r + `)f (r);  (r)(17) gdr: were 0:0284; 0:0354; 5:3358 and 16:5325 respectively. stret The last equality considers that for an homogeneous sys- tem the statistical average can be taken as a spatial av- The energy spectra erage. From the convolution theorem the Fourier trans- _ _ form E (p) =  f; g . In a steady state the time stret p p The energies in (3) maybe seen as quadratic contri- derivative of the correlation function must be zero hence 1 2 bution of  and (r) =     .The nal xx yy _ xy f; g = 0 and the stretching energy ux (15) must p p quadratic contributions read be zero. Z Z 1 E 2 2 E = j j dk and E = j j dk:(12) kin k stret k 2 2 1-law for thin elastic plate turbulence We de ne the kinetic energy and the stretching spectra In Ref. [17] an exact result for turbulence of thin elastic by averaging over the angular variables in the wavenum- plates was found. It provides a K arm an-Howarth-Monin ber space, hence type relationship for the energy ux: D E 2 2 _ 1 E (k) = k j j and E (k) =  k j j(13) ; kin k stret k _ r hJ i = P; (18) ` [;] where  stands for the rst di erence of the eld, e.g. where we have de ned h: : :i = : : : d' by the angu- =  (x + `)  (x). The vector J is de ned by lar average. J = f h f h ; f h f h : y yx x yy x xy y xx [f;h] The energy uxes Under the assumption of isotropy, (18) implies the 1-law (10) for the structure function Because the energy is quadratic in  and , the energy _ ^ Z (`)  hJ i ` = P `; (19) ux can be straightforward de ned as in hydrodynamic [;] turbulence. By making a scale-by-scale energy budget where ` is the unit vector along `. the energy uxes are: GD, GK and SR thank to FONDECYT grants N Z Z k k D E 1181382 and the Chilean-French scienti c exchange pro- @E (p) kin P (k) = dp =  f; g (14) dp; kin p p gram ECOS-Sud/CONICYT No. C14E04. @t 0 0 Z Z k k D E @E (p) stret P (k) = dp =  f; g (15) dp; stret p p @t 0 0 where the subscript 0 stands for the free of forcing [1] Reynolds O (1883) Phil. Trans. Roy. Soc. 174: 935. [2] Kolmogorov AN (1941) Dokl. Akad. Nauk SSSR 32: 15- and dissipation time variations of the elds through (1) 17 [English: Reprinted by Kolmogorov AN (1991) Proc. with F = D = 0. For simplicity, we de ne ff; gg R. Soc. London 434: 15-17]. f g + f g 2f g . We have also used the fact xx yy yy xx xy xy [3] Cardesa JI, Vela-Mart n A, Jim enez, J (2017) Science 1 1 that the angular averages satisfy ( f ) ( g) = p p 357: 782-784. ( f ) g . The total energy ux results from the p p [4] Castelvecchi D (2017) Nature 548: 382-383. addition of these previous expressions. [5] Ruelle DP (2012) PNAS 109: 20344-20346. 8 [6] von K arm an T, Howarth L (1938) Proc. R. Soc. London [18] F oppl A (1907) Vorlesungen ub  er technische Mechanik, 164: 192-216. Bd. 5, Leipzig p. 132; von K arm an T (1910) Ency. d. [7] Frisch U (1995) Turbulence: The legacy of A.N. Kol- math. Wiss., Bd. IV. 2, II, Leipzig, x8. mogorov. (Cambridge University Press, Cambridge). [19] During  G, Josserand C, Rica S (2017) Physica D 347: [8] Hasselmann K (1962) J. Fluid Mech. 12: 481-500; (1963) 42-73. J. Fluid Mech. 15: 273-281; (1963) J. Fluid Mech. 15: [20] Boudaoud A, Cadot O, Odille B, Touz e C (2008) Phys. 385-398. Rev. Lett. 100: 234504. [9] Benney DJ, Sa man PG (1966) Proc. Roy. Soc. London [21] Mordant N (2008) Phys. Rev. Lett. 100: 234505. A 289: 301-320. [22] Humbert T, Cadot O, During  G, Josserand C, Rica S, [10] Zakharov VE, Filonenko NN Dokl. Akad. Nauk SSSR Touz e C (2013) Euro. Phys. Lett. 102: 30002. 170 (1966): 1292 [English transl. in Sov. Math. Dokl.]; [23] Du ring G, Josserand C, Rica S (2015) Phys. Rev. E 91: Zakharov VE (1966) Zh. Eksper. Teoret. Fiz. 51: 686 052916. [English transl. in (1967) Sov. Phys. JETP 24: 455]; [24] Humbert T, Cadot O, Josserand C, Touz e C (2016) Phys- Zakharov VE, Filonenko NN (1967) Zh. Prikl. Mekh. I ica D 316: 34. Tekn. Fiz. 5: 62 [English transl. in J. Appl. Mech. Tech. [25] Miquel B, Alexakis A, Josserand C, Mordant N (2013) Phys.]; Zakharov VE, L'vov VS, G. Falkovich G (1992) Phys. Rev. Lett. 111: 054302. Kolmogorov Spectra of Turbulence I (Springer, Berlin). [26] Yokoyama N, Takaoka M. (2013) Phys. Rev. Lett. 110: [11] Benney DJ, Newell AC (1967) J. Math. Phys. 46: 363; 105501. Newell AC (1968) Rev. Geophys. 6: 1-31; Benney DJ, [27] Chibbaro S, Josserand C (2016) Phys. Rev. E 94: Newell AC. (1969) Stud. Appl. Math. 48, 29-53. 011101(R). [12] Newell AC, Rumpf B (2011) Annual Review of Fluid Me- [28] Landau LD, Lifshitz EM (1959) Theory of Elasticity chanics 43: 59-78. (Pergamon, New York). [13] Nazarenko S. (2011) Wave turbulence, Lecture Notes in [29] Witten TA, Li H (1993) Europhys. Lett. 23: 51-55. Physics Vol. 825, (Springer Berlin) . [30] Ben Amar M, Pomeau Y (1997) Proc. R. Soc. A 453: [14] Dyachenko S, Newell AC, Pushkarev A, Zakharov VE 729-755 (1992) Physica D 57: 96. [31] Cerda E, Mahadevan L (1998) Phys. Rev. Lett. 80: 2358- [15] Durin  g G, Josserand C, Rica S (2006) Phys. Rev. Lett. 2361. 97: 025503. [32] Newell AC, Nazarenko S, Biven L (2001) Physica D, 152- [16] Galtier S and Nazarenko S (2017) Phys. Rev. Lett. 119: 153: 520-550. 221101. [33] Falcon E, Fauve S, Laroche C (2007) Phys. Rev. Lett. 98: [17] Durin  g G, Krstulovic G (2018) Phys. Rev. E 97: 154501. 020201(R). [34] Mordant N, Miquel B. (2017) Phys. Rev. E 96, 042204.

Journal

Nonlinear Sciences – arXiv (Cornell University)

Published: Aug 6, 2018

There are no references for this article.