1: %% LyX 1.3 created this file. For more info, see http://www.lyx.org/.
2: %% Do not edit unless you really know what you are doing.
3: \documentclass[english,showpacs,superscriptaddress]{revtex4}
4: \usepackage[T1]{fontenc}
5: \usepackage[dvips]{graphicx}
6:
7: \makeatletter
8:
9: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% LyX specific LaTeX commands.
10: %% Bold symbol macro for standard LaTeX users
11: \newcommand{\boldsymbol}[1]{\mbox{\boldmath $#1$}}
12:
13:
14: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% User specified LaTeX commands.
15: % Uncomment to print out only slides and overlays
16: %
17: %\onlyslides{\slides}
18:
19: % Uncomment to print out only notes
20: %
21: %\onlynotes{\notes}
22:
23: \usepackage{babel}
24: \makeatother
25: \begin{document}
26:
27: \title{Incompressible Turbulence as Nonlocal Field Theory}
28:
29:
30: \author{Mahendra K. Verma }
31:
32:
33: \email{mkv@iitk.ac.in}
34:
35:
36: \homepage{http://home.iitk.ac.in/~mkv}
37:
38:
39: \affiliation{Department of Physics, Indian Institute of Technology, Kanpur 208016,
40: INDIA}
41:
42: \begin{abstract}
43: It is well known that incompressible turbulence is nonlocal in real
44: space because sound speed is infinite in incompressible fluids. The
45: equation in Fourier space indicates that it is nonlocal in Fourier
46: space as well. Contrast this with Burgers equation which is local
47: in real space. Note that the sound speed in Burgers equation is zero.
48: In our presentation we will contrast these two equations using nonlocal
49: field theory. Energy spectrum and renormalized parameters will be
50: discussed.
51: \end{abstract}
52:
53: \pacs{47.27.Ak, 47.27Gs}
54:
55: \maketitle
56:
57: \section{Introduction}
58:
59: Generic \textbf{}Equations in Physics, like Diffusion Equation, Schrodinger
60: Equation are local in real space. Take Schrodinger's equation for
61: example:\[
62: -i\hbar\frac{\partial\psi}{\partial t}=-\frac{\hbar^{2}}{2m}\nabla\psi+V(\mathbf{x})\psi,\]
63: where $V(\mathbf{x})$ is the potential, and $\psi(\mathbf{x},t)$
64: is the wavefunction. Clearly, to compute $\psi(\mathbf{x},t+dt)$
65: given $\psi(\mathbf{x},t)$, we need the local function, and finite
66: number of its derivatives. In this talk we investigate whether the
67: equations for fluid flows is local or not.
68:
69: Fluid flows are described by Navier-Stokes (NS) equation, continuity
70: equation, and the equation of state given below:\begin{eqnarray}
71: \frac{\partial\mathbf{u}}{\partial t}+\left(\mathbf{u}\cdot\nabla\right)\mathbf{u} & = & -\frac{1}{\rho}\nabla p+\nu\nabla^{2}\mathbf{u},\label{eq:NS1}\\
72: \frac{\partial\rho}{\partial t}+\nabla\cdot\left(\rho\mathbf{u}\right) & = & 0,\label{eq:rho}\\
73: p & = & p(\rho),\label{eq:p}\end{eqnarray}
74: where $\mathbf{u},p,\rho$ are the velocity, pressure, and density
75: field respectively. $\nu$is the kinematic viscosity of the fluid.
76: We nondimensionalize the above equations by scaling the quantities
77: appropriately, e. g., divide $\mathbf{u}$ by velocity scale $U$.
78: Navier-Stokes equation gets modified to
79:
80: \begin{eqnarray}
81: \frac{\partial\mathbf{u}}{\partial t}+\left(\mathbf{u}\cdot\nabla\right)\mathbf{u} & = & -\frac{C_{s}^{2}}{U^{2}}\nabla\rho+\frac{\nu}{UL}\nabla^{2}\mathbf{u}\label{eq:NS2}\end{eqnarray}
82: where $L$ is the length scale, and $C_{s}=\sqrt{{\frac{dp_{0}}{d\rho_{0}}}}$
83: is the Sound Speed. If the sound speed $C_{s}$ and $\frac{\nu}{UL}$
84: are finite, then we can find $\mathbf{u}(\mathbf{x},t+dt)$ and $\rho(\mathbf{x},t+dt)$
85: given $\mathbf{u}(\mathbf{x})$ and $\rho(\mathbf{x},t)$ ( assuming
86: that $\mathbf{u}(\mathbf{x},t)$,$\rho(\mathbf{x},t)$, and their
87: first and second derivatives are finite). For typical flows $\frac{\nu}{UL}$
88: is finite, so for finite $C_{s}$, Navier-Stokes equation is local
89: in real space.
90:
91: Note that the disturbances propagate with the sound speed. The larger
92: the sound speed, larger the range of influence per unit time. Still
93: the influence moves locally as long as the sound speed is finite.
94: When the sound speed is infinite, then disturbances can propagate
95: instantaneously, and all parts of the system starts interacting; hence
96: the system becomes nonlocal. Hence, the equations for fluid flows
97: become nonlocal when the sound speed is infinite, which is the case
98: for incompressible fluids $(\delta\rho=0)$. The speed of propagation
99: is infinite in Newton's law of gravitation as well as in Coulomb's
100: interactions between the charged particles. These are some other examples
101: of nonlocal interactions in physics.
102:
103: We can abstract the above physics using a 2D mesh of spring-mass system.
104: For finite spring constant, the disturbance propagates with a finite
105: speed, and the physics is local. When the spring constant is very
106: large, the physics is still local, but the range of propagation per
107: unit time becomes quite large. Here mass is pulled-pushed by local
108: spring only (4 of them). When the spring constant become infinite,
109: then whole system behaves like a solid and the speed of propagation
110: becomes infinite. This system has nonlocal interactions; the movement
111: of the mass at a point is affected by masses and springs at large
112: distances. In fact, in this nonlocal system, we can think of a mass
113: connected by all other masses, like in Coulomb's interactions or in
114: Calegaro-Sutherland model.
115:
116: Any real fluid has finite sound speed however large it may be. In
117: practice, the fluid is considered incompressible if $C_{s}/U\gg1$.
118: The properties of this fluid is expected to match with the ideal incompressible
119: fluid. This is based on an assumption that the properties of fluid
120: with $C_{s}\rightarrow\infty$ matches with $C_{s}=\infty$, or $(C_{s}\rightarrow\infty)=(C_{s}=\infty)$.
121: This assumption seems reasonable, but we are not aware of any strict
122: mathematical result showing this. Note that $(\nu\rightarrow0)\ne(\nu=0)$.
123:
124: In the next section we discuss incompressible Navier-Stokes equation.
125:
126:
127: \section{Incompressible Navier-Stokes}
128:
129: Before we proceed further, we remark that for incompressible fluids
130: the normalized $-\nabla p=-\frac{C_{s}^{2}}{U^{2}}\nabla\rho$ is
131: finite even though $C_{s}$ is infinite. Also note that the normalized
132: term $\frac{\nu}{UL}\nabla^{2}\mathbf{u}$ is finite.
133:
134: The continuity equation yields a constraint equation\[
135: \nabla\cdot\mathbf{u}=0,\]
136: the substitution of which in NS equation gives Poission's equation
137: for $p$,\[
138: \nabla^{2}p=-\nabla\cdot\left\{ \mathbf{u}\cdot\nabla\mathbf{u}\right\} \]
139: Therefore,\[
140: p(\mathbf{x},t)=\int\frac{-\nabla\cdot\left\{ \mathbf{u}\cdot\nabla\mathbf{u}\right\} }{\left|\mathbf{x}-\mathbf{x}'\right|},\]
141: which is the Coulomb's operator (nonlocal). Clearly the computation
142: of $p(\mathbf{x},t)$ and consequently that of $\mathbf{u}(\mathbf{x},t+dt)$
143: requires knowledge of $\mathbf{u}(\mathbf{x}')$ at all positions.
144: This is another way to infer that incompressible NS is nonlocal in
145: real space. Landau \cite{LandFlui:book}, Frisch \cite{Fris:book},
146: and others reached the above conclusion.
147:
148: It is customary to study NS in Fourier space. Let us investigate whether
149: NS is local or nonlocal in Fourier space. NS equation in Fourier space
150: is
151:
152: \[
153: \frac{\partial u_{i}(\mathbf{k},t)}{\partial t}=-ik_{j}\int u_{j}(\mathbf{q})u_{i}(\mathbf{p})+ik_{i}\frac{k_{j}k_{m}}{k^{2}}\int u_{j}(\mathbf{q})u_{m}(\mathbf{p})\]
154: with $\mathbf{k=p+q}$. Note that the second term arises due to pressure.
155:
156: Since $u_{i}(\mathbf{k},t)$ requires knowledge of $u_{j}(\mathbf{q})u_{i}(\mathbf{p})$
157: where $\mathbf{p}$ and \textbf{$\mathbf{q}$} could be very different
158: from $\mathbf{k}$, hence NS is nonlocal interactions in Fourier Space.
159: If we interpret NS in terms of energy transfer, we find that the energy
160: is exchanged between the all the three modes of the triad. Kraichnan
161: \cite{Krai:59} and Dar et al. \cite{Dar:flux} devised formulas to
162: measure the strength of interactions in fluid turbulence. In this
163: paper we will use Dar et al.'s \emph{mode-to-mode formalism} \cite{Dar:flux}
164: in which the energy transfer rate from Fourier mode $\mathbf{p}$
165: to Fourier mode $\mathbf{k}$ with Fourier mode $\mathbf{q}$ acting
166: as a mediator is given by\begin{equation}
167: S(\mathbf{k|p|q)}=Im[\mathbf{k}\cdot\mathbf{u}(\mathbf{q})\mathbf{u}(\mathbf{p})\cdot\mathbf{u}(\mathbf{k})].\label{eq:S}\end{equation}
168: The above quantity can be computed using numerical simulation or using
169: analytic tools. Earlier, Domaradzki and Rogallo \cite{Doma:Local2}
170: and Waleffe \cite{Wale} calculated the above using EDQNM approximation.
171: Recently Verma et al. \cite{Ayye} calculated the above using field-theoretic
172: technique. In this paper we will report analytical result obtained
173: using first-order field-theoretic calculation. In this scheme, under
174: the assumption of homogeneity and isotropy we obtain\begin{equation}
175: \left\langle S(k'|p|q))\right\rangle =\frac{T_{1}(k,p,q)C(p)C(q)+T_{2}(k,p,q)C(k)C(q)+T_{3}(k,p,q)C(k)C(p)}{\nu(k)k^{2}+\nu(p)p^{2}+\nu(q)q^{2}},\label{eq:Savg}\end{equation}
176: where $T_{i}$'s are functions of $k,p$ and $q$. To save space,
177: we have skipped all the details for which the reader is referred to
178: Verma et al. \cite{Ayye}.
179:
180: We focus our attention on the inertial range where the interactions
181: are self-similar. Therefore, it is sufficient to analyze $S(k'|p|q)$
182: for triangles $(1,p/k,q/k)=(1,v,w)$. Since, $|k-p|\le q\le k+p$,
183: $|1-v|\le w\le1+v$, hence any interacting triad $(1,v,w)$ is represented
184: by a point $(v,w)$ in the hatched region of Fig. \ref{Fig:vw} \cite{Lesl:book}.
185:
186: %
187: \begin{figure}
188: \includegraphics[%
189: scale=0.5]{vw.eps}
190:
191:
192: \caption{\label{Fig:vw} The interacting triad $(\mathbf{k,p,q})/k=(1,v,w)$
193: under the condition $\mathbf{k}=\mathbf{p+q}$ is represented by a
194: point $(v,w)$ in the hatched region. The axis $(v',w')$ are inclined
195: to axis $(v,w)$ by 45 degrees. Note that the local wavenumbers are
196: $v\approx1,w\approx1$ or $v'\approx w'\approx1/\sqrt{{2}}$. }
197: \end{figure}
198: For convenience, $\left\langle S(v',w')\right\rangle $ are represented
199: in terms of new variables $(v',w')$ measured from the rotated axis
200: shown in the figure \ref{Fig:vw}. It is easy to show that $v=1+(v'-w')/\sqrt{{2}},w=(v'+w')/\sqrt{{2}}$.
201:
202: The local wavenumbers are $v\approx1,w\approx1$, while the rest are
203: called nonlocal wavenumbers. We substitute $C(k)$ and $\nu(k)$ in
204: Eq. (\ref{eq:Savg}) and rewrite $S(k|p|q)$ in terms of $v',w'$.
205: For details refer to Verma et al. \cite{Ayye}. Fig. \ref{Fig:S}
206: illustrates the density plots of $\left\langle S(v',w')\right\rangle $.
207: Fig. (a) shows the plot for 3D, while Fig. (b) shows the one for 2D.
208: %
209: \begin{figure}
210: \includegraphics[%
211: scale=0.5]{S.ps}
212:
213:
214: \caption{\label{Fig:S} Density plot of $\left\langle S(v',w')\right\rangle $
215: of Eq. (\ref{eq:Savg}) without the bracketed factor for (a) 3D (b)
216: 2D. }
217: \end{figure}
218: We can draw the following conclusions from the plots.
219:
220: \begin{enumerate}
221: \item When $v'\rightarrow1$ or $v\rightarrow0$, we find that $S(k|p|q)$
222: is large positive for 3D and large negative for 3D. This shows that
223: the nonlocal interactions are strong.
224: \item The value of $S$ at $(v,w)=(1,1)$, or $(v',w')=(1/\sqrt{{2}},1/\sqrt{{2}})$
225: is zero in both 2D and 3D. When $v\approx w\approx1$, $S$ is small
226: indicating that local interactions are small.
227: \item When $v\rightarrow0$, $S>0$ for 3D and $S<0$ for 2D. This is reminiscent
228: of forward cascade in 3D, and backward cascade in 2D.
229: \end{enumerate}
230: Hence, we find that the interactions in the incompressible fluid turbulence
231: is nonlocal. This result appears to contradict Kolmogorov's phenomenology
232: which predicts local energy transfer in Fourier space. We will show
233: below that the shell-to-shell energy transfer rates are local even
234: though the interactions are nonlocal.
235:
236:
237: \section{Shell-to-shell energy transfers in turbulence.}
238:
239: The wavenumber space is divided into shells $(k_{0}s^{n},k_{0}s^{n+1})$,
240: where $s>1$, and $n$ can take both positive and negative values.
241: The energy transfer rate from $m$th shell $(k_{0}s^{m},k_{0}s^{m+1})$
242: to $n$th shell $(k_{0}s^{n},k_{0}s^{n+1})$ is given by \cite{Dar:flux}\begin{equation}
243: T_{nm}=\sum_{k_{0}s^{n}\leq k\leq k_{0}s^{n+1}}\sum_{k_{0}s^{m}\leq p\leq k_{0}s^{m+1}}\left\langle S(k|p|q)\right\rangle .\label{eq:shell}\end{equation}
244: If the shell-to-Shell energy transfer rate is maximum for the nearest
245: neighbours, and decreases monotonically with the increase of $|n-m|$,
246: then the shell-to-shell energy transfer is said to be local.
247:
248: $T_{nm}$ can be computed using numerical simulations or using analytical
249: tools. Zhou \cite{Zhou:Local} calculated similar quantity. In the
250: following we plot $T_{nm}$ obtained using numerical simulation \cite{Ayye}.
251: Clearly, \emph{shell-to-shell energy transfer is local as envisaged
252: by Kolmogorov.}
253:
254: \begin{center}%
255: \begin{figure}
256: \includegraphics[%
257: scale=0.3]{Tnmsim.eps}
258:
259:
260: \caption{\label{Fig:T-sim} Plots of normalized shell-to-shell energy transfer
261: $T_{nm}/\Pi$ vs. $n-m$ for $m=4..9$. The plots collapse on each
262: other indicating self-similarity.}
263: \end{figure}
264: \end{center}
265:
266: We \cite{Ayye} have also computed the shell-to-shell energy transfer
267: rates using field-theoretic method. The reader is referred to the
268: original paper for the details. The plots of $T_{nm}$ for both 3D
269: and 2D fluid turbulence given below.
270:
271: %
272: \begin{figure}
273: \includegraphics[%
274: scale=0.5]{shell3D.ps}
275:
276:
277: \caption{\label{Fig:shell3D} Plot of normalized shell-to-shell energy transfer
278: $T_{nm}/\Pi$ vs $n-m$ for $d=3$. The $n$th shell is $(k_{0}s^{n},k_{0}s^{n+1})$
279: with $s=2^{1/4}$. The energy transfer is maximum for $n=m\pm1$,
280: hence the energy transfer is local. The energy transfer is also forward.}
281: \end{figure}
282: From the above plot we can clearly deduce that energy transfer in
283: 3D fluid turbulence is local. In fact, the values obtained from analytical
284: calculations match very well with the numerical values shown in Fig.
285: \ref{Fig:T-sim}.
286:
287: We have done similar analysis for 2D fluid turbulence. The result
288: is shown below:
289:
290: %
291: \begin{figure}
292: \includegraphics[%
293: scale=0.5]{shell2D.ps}
294:
295:
296: \caption{\label{Fig:shell2D} Plot of normalized shell-to-shell energy transfer
297: $T_{nm}/\left|\Pi\right|$ vs $n-m$ for $d=2$ in the inertial range.
298: The energy transfer rate from the shell $m$ to the shells $m+1,$
299: $m+2$, $m+3$ is forward, but $m+4$ onward it is negative. The net
300: effect of all these transfer is the inverse energy flux $\Pi$. }
301: \end{figure}
302: The shell-to-shell energy transfer rates to the nearby shells are
303: forward, whereas the transfer rates to the far off shells are backward.
304: The net effect is a negative energy flux. This theoretical result
305: is consistent with Dar et al.'s numerical finding \cite{Dar:flux}.
306: The inverse cascade of energy is consistent with the backward nonlocal
307: energy transfer in mode-to-mode picture {[}$S(k|p|q)${]} (see Fig.
308: \ref{Fig:S}). Verma et al. \cite{Ayye} have shown that the transition
309: from backward energy transfer to forward transfer takes place at $d_{c}\approx2.25$.
310:
311: To summarize, the triad interactions in incompressible fluids is nonlocal
312: both in real and Fourier space. However, the shell-to-shell energy
313: transfer is local in Fourier space. Verma et al. \cite{Ayye} attribute
314: this behaviour to the fact that the nonlocal triads occupy much less
315: Fourier space volume than the local ones.
316:
317:
318: \section{Fully compressible limit: Burgers equation}
319:
320: Let us go back to Eq. (\ref{eq:NS2}) and take the limit $C_{s}=0$.
321: This is the fully compressible limit, and the resulting equation was
322: first studied by Burgers. This equation, given below, is known as
323: Burgers equation (strictly speaking in 1D).\begin{eqnarray*}
324: \frac{\partial\mathbf{u}}{\partial t}+\left(\mathbf{u}\cdot\nabla\right)\mathbf{u} & = & \nu\nabla^{2}\mathbf{u}.\end{eqnarray*}
325: Clearly this equation is local in real space. What about in Fourier
326: space?
327:
328: In Fourier space, the above equation is given by\[
329: \frac{\partial u_{i}(\mathbf{k},t)}{\partial t}=-i\int u_{j}(\mathbf{q})p_{j}u_{i}(\mathbf{k-q})-\nu k^{2}u_{i}(\mathbf{k}),\]
330: which implies that the interactions are nonlocal in Fourier space.
331: Note that the pressure term is absent in the above equation.
332:
333: The field-theoretic treatment of the above equation is rather complex
334: for arbitrary dimension. Here we attempt the self-consistent field-theoretic
335: treatment of one-dimensional Burgers equation for 1D Burgers equation.
336: In 1D, the energy spectrum of Burgers equation is given by \begin{equation}
337: C(k)=A\frac{\mu^{2}}{L}k^{-2},\label{eq:Burg:Ck}\end{equation}
338: where $L$ is the length of the system, $\mu$ is the shock strength,
339: and $A$ is a constant. Using dimensional arguments, we write the
340: renormalized viscosity of the following form \cite{McCo:book,MKV:MHD_RG}:
341:
342: \begin{eqnarray}
343: \nu(k_{n}k') & = & \nu_{*}(k')\mu\sqrt{{\frac{A}{L}}}k_{n}^{-3/2}.\label{eq:Burg:nu}\end{eqnarray}
344: Unfortunately straight-forward application of self-consistent
345: Renormalization Group (RG) procedure
346: of McComb \cite{McCo:book,MKV:MHD_RG} does not work. The contributions
347: of $<u^{>}(p)u^{>}(q)>$ is negligible; one needs to come up with
348: a cleverer renormalization scheme to obtain the renormalized viscosity.
349:
350: To make a connection with Kolmogorov's theory of fluid turbulence,
351: we rewrite Eq. (\ref{eq:Burg:Ck}) as
352:
353: \[
354: C(k)=A[\Pi(k)]^{2/3}k^{-5/3}\]
355: with the flux function $\Pi(k)$ as\begin{equation}
356: \Pi(k)=\frac{\mu^{3}}{L^{3/2}}k^{-1/2}.\label{eq:Burg:Pi}\end{equation}
357: Note that the flux $\Pi$ has become $k$-dependent. Verma \cite{MKV:KPZ}
358: and Frisch \cite{Fris:book} have shown that the flux function follows
359: a multifractal distribution.
360:
361: Question is whether we can compute the flux using field-theoretic
362: method. Since Burgers equation is compressible, the formula $S(K|p|q)$
363: is not applicable \cite{MKV:PR}. However we can still write the flux
364: using Kraichnan's combined energy transfer formula \cite{Krai:59}.
365: The energy flux crossing a wavenumber $k_{0}$ is given by\begin{eqnarray*}
366: \Pi(k_{0}) & = & \int_{k>k_{0}}dk\int_{p<k_{0}}\frac{1}{2}\Im\left[-ku(p)u(q)u(k)\right],\end{eqnarray*}
367: with $k+p+q=0$. We apply first-order perturbative method assuming
368: $u(k)$ to be quasi-normal as in fluid turbulence. We also make a
369: change of variable to \[
370: k=\frac{k_{0}}{u};\,\,\, p=\frac{k_{0}v}{u};\,\,\, q=\frac{k_{0}w}{u}.\]
371: To first order,\begin{eqnarray*}
372: \Pi(k_{0}) & = & \frac{k_{0}^{2}}{2}\int_{-1}^{1}du\int_{-u}^{u}(k)\frac{kC(p)C(q)+pC(k)C(q)+qC(k)C(p)}{\nu(p)p^{2}+\nu(q)q^{2}+\nu(k)k^{2}}\\
373: & = & \frac{A^{3/2}}{\nu_{*}}\Pi(k_{0})\int_{-1}^{1}dv\,2(1-v^{1/2})\frac{w^{-2}+v^{-2}-(vw)^{-2}}{1+v^{1/2}+w^{1/2}},\end{eqnarray*}
374: with $w=1-v$. We find that the above integral converges and is equal
375: to 2.45. Hence,\[
376: 1=\frac{A^{3/2}}{\nu_{*}}2.45.\]
377: Thus we show that $C(k),\nu(k),\Pi(k)$ given by Eqs. (\ref{eq:Burg:Ck},
378: \ref{eq:Burg:nu}, \ref{eq:Burg:Pi}) are consistent solution of 1D
379: Burgers equation. Note however that the renormalization group analysis
380: of Burgers equation is somewhat uncertain.
381:
382: The spectral index of Burgers equation ($-2$) is very different from
383: the the spectral index of incompressible fluid turbulence ($-5/3$).
384: The difference arises due to the neglect of $-\nabla p$ term in Burgers
385: equation. The compressible effects are different in these two equations.
386: Burgers equation is local real space, while incompressible NS is nonlocal
387: in real space.
388:
389: It is interesting to compare the above results with Noncommutative
390: field theory, where the nonlocal interactions are included using parameter
391: $\theta$. Burgers equation is local, while incompressible NS is nonlocal
392: due to $-\nabla p$ term. Note that the $\nabla p$ term is nonlocal
393: in Coulomb's operator sense $(V\sim1/r)$. We are not aware of field-theoretic
394: ideas applied to Coulomb;s operator, which is one of the most important
395: operator in physics. We hope this investigation and its connection
396: with fluid turbulence will be carried out in future.
397:
398: \begin{acknowledgments}
399: The above work is a result of collaborative work and discussions with
400: Arvind Ayyer, Shishir Kumar, Amar V. Chandra, V. Eswaran, and G. Dar.
401: \end{acknowledgments}
402: %\bibliographystyle{apsrev}
403: %\bibliography{/home/mkv/research/mypapers/bib/mkv,/home/mkv/research/mypapers/bib/fluid/general,/home/mkv/research/mypapers/bib/mhd/Bisk,/home/mkv/research/mypapers/bib/mhd/Bran,/home/mkv/research/mypapers/bib/mhd/analytic,/home/mkv/research/mypapers/bib/mhd/anisotropy,/home/mkv/research/mypapers/bib/mhd/compressible,/home/mkv/research/mypapers/bib/mhd/dnsles,/home/mkv/research/mypapers/bib/mhd/dynamo,/home/mkv/research/mypapers/bib/mhd/field-th,/home/mkv/research/mypapers/bib/mhd/general,/home/mkv/research/mypapers/bib/mhd/intermittency,/home/mkv/research/mypapers/bib/mhd/soc,/home/mkv/research/mypapers/bib/mhd/solar,/home/mkv/research/mypapers/bib/mhd/turbulence,/home/mkv/research/mypapers/bib/fluid-turb/dnsles,/home/mkv/research/mypapers/bib/fluid-turb/field-th,/home/mkv/research/mypapers/bib/fluid-turb/general,/home/mkv/research/mypapers/bib/fluid-turb/intermittency,/home/mkv/research/mypapers/bib/fluid-turb/local,/home/mkv/research/mypapers/bib/fluid-turb/2d,/home/mkv/research/mypapers/bib/field-th/general,/home/mkv/research/mypapers/bib/fluid-turb/anisotropy,/home/mkv/research/mypapers/bib/mhd/misc,/home/mkv/research/mypapers/bib/fluid-turb/analytic,/home/mkv/research/mypapers/bib/fluid-turb/Krai,/home/mkv/research/mypapers/bib/misc}
404:
405: \begin{thebibliography}{13}
406: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
407: \expandafter\ifx\csname bibnamefont\endcsname\relax
408: \def\bibnamefont#1{#1}\fi
409: \expandafter\ifx\csname bibfnamefont\endcsname\relax
410: \def\bibfnamefont#1{#1}\fi
411: \expandafter\ifx\csname citenamefont\endcsname\relax
412: \def\citenamefont#1{#1}\fi
413: \expandafter\ifx\csname url\endcsname\relax
414: \def\url#1{\texttt{#1}}\fi
415: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
416: \providecommand{\bibinfo}[2]{#2}
417: \providecommand{\eprint}[2][]{\url{#2}}
418:
419: \bibitem[{\citenamefont{Landau and Lifsitz}(1987)}]{LandFlui:book}
420: \bibinfo{author}{\bibfnamefont{L.~D.} \bibnamefont{Landau}} \bibnamefont{and}
421: \bibinfo{author}{\bibfnamefont{E.~M.} \bibnamefont{Lifsitz}},
422: \emph{\bibinfo{title}{Fluid Mechanics}} (\bibinfo{publisher}{Pergamon Press},
423: \bibinfo{address}{Oxford}, \bibinfo{year}{1987}).
424:
425: \bibitem[{\citenamefont{Frisch}(1995)}]{Fris:book}
426: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Frisch}},
427: \emph{\bibinfo{title}{Turbulence}} (\bibinfo{publisher}{Cambridge University
428: Press}, \bibinfo{address}{Cambridge}, \bibinfo{year}{1995}).
429:
430: \bibitem[{\citenamefont{Kraichnan}(1959)}]{Krai:59}
431: \bibinfo{author}{\bibfnamefont{R.~H.} \bibnamefont{Kraichnan}},
432: \bibinfo{journal}{J. Fluid Mech.} \textbf{\bibinfo{volume}{5}},
433: \bibinfo{pages}{497} (\bibinfo{year}{1959}).
434:
435: \bibitem[{\citenamefont{Dar et~al.}(2001)\citenamefont{Dar, Verma, and
436: Eswaran}}]{Dar:flux}
437: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Dar}},
438: \bibinfo{author}{\bibfnamefont{M.~K.} \bibnamefont{Verma}}, \bibnamefont{and}
439: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Eswaran}},
440: \bibinfo{journal}{Physica D} \textbf{\bibinfo{volume}{157}},
441: \bibinfo{pages}{207} (\bibinfo{year}{2001}).
442:
443: \bibitem[{\citenamefont{Domaradzki and Rogallo}(1990)}]{Doma:Local2}
444: \bibinfo{author}{\bibfnamefont{J.~A.} \bibnamefont{Domaradzki}}
445: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.~S.}
446: \bibnamefont{Rogallo}}, \bibinfo{journal}{Phys. Fluids A}
447: \textbf{\bibinfo{volume}{2}}, \bibinfo{pages}{413} (\bibinfo{year}{1990}).
448:
449: \bibitem[{\citenamefont{Waleffe}(1992)}]{Wale}
450: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Waleffe}},
451: \bibinfo{journal}{Phys. Fluids A} \textbf{\bibinfo{volume}{4}},
452: \bibinfo{pages}{350} (\bibinfo{year}{1992}).
453:
454: \bibitem[{\citenamefont{Verma et~al.}(2004)\citenamefont{Verma, Ayyer, ,
455: Debliquy, Kumar, and Chandra}}]{Ayye}
456: \bibinfo{author}{\bibfnamefont{M.~K.} \bibnamefont{Verma}},
457: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Ayyer}}, ,
458: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Debliquy}},
459: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Kumar}}, \bibnamefont{and}
460: \bibinfo{author}{\bibfnamefont{A.~V.} \bibnamefont{Chandra}},
461: \bibinfo{journal}{Pramana}
462: \textbf{\bibinfo{volume}{65}}, \bibinfo{pages}{297}
463: (\bibinfo{year}{2004}).
464:
465: \bibitem[{\citenamefont{Leslie}(1973)}]{Lesl:book}
466: \bibinfo{author}{\bibfnamefont{D.~C.} \bibnamefont{Leslie}},
467: \emph{\bibinfo{title}{Development in the Theory of Turbulence}}
468: (\bibinfo{publisher}{Oxford University Press}, \bibinfo{address}{Claredon},
469: \bibinfo{year}{1973}).
470:
471: \bibitem[{\citenamefont{Zhou}(1993)}]{Zhou:Local}
472: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Zhou}}, \bibinfo{journal}{Phys.
473: Fluids} \textbf{\bibinfo{volume}{5}}, \bibinfo{pages}{1092}
474: (\bibinfo{year}{1993}).
475:
476: \bibitem[{\citenamefont{McComb}(1990)}]{McCo:book}
477: \bibinfo{author}{\bibfnamefont{W.~D.} \bibnamefont{McComb}},
478: \emph{\bibinfo{title}{The Physics of Fluid Turbulence}}
479: (\bibinfo{publisher}{Oxford University Press}, \bibinfo{address}{Claredon},
480: \bibinfo{year}{1990}).
481:
482: \bibitem[{\citenamefont{Verma}(2001)}]{MKV:MHD_RG}
483: \bibinfo{author}{\bibfnamefont{M.~K.} \bibnamefont{Verma}},
484: \bibinfo{journal}{Phys. Plasmas} \textbf{\bibinfo{volume}{8}},
485: \bibinfo{pages}{3945} (\bibinfo{year}{2001}).
486:
487: \bibitem[{\citenamefont{Verma}(2000)}]{MKV:KPZ}
488: \bibinfo{author}{\bibfnamefont{M.~K.} \bibnamefont{Verma}},
489: \bibinfo{journal}{Physica A} \textbf{\bibinfo{volume}{277}},
490: \bibinfo{pages}{359} (\bibinfo{year}{2000}).
491:
492: \bibitem[{\citenamefont{Verma}(2004)}]{MKV:PR}
493: \bibinfo{author}{\bibfnamefont{M.~K.} \bibnamefont{Verma}},
494: \bibinfo{journal}{Phys. Rep.}
495: \textbf{\bibinfo{volume}{401}},
496: \bibinfo{pages}{229}, (\bibinfo{year}{2004}).
497:
498: \end{thebibliography}
499:
500:
501: \end{document}
502: