1:
2: \begin{filecontents*}{example.eps}
3:
4: gsave
5: newpath
6: 20 20 moveto
7: 20 220 lineto
8: 220 220 lineto
9: 220 20 lineto
10: closepath
11: 2 setlinewidth
12: gsave
13: .4 setgray fill
14: grestore
15: stroke
16: grestore
17: \end{filecontents*}
18: %
19: %\documentclass{svjour3} % onecolumn (standard format)
20: \documentclass[smallextended]{svjour3} % onecolumn (second format)
21: %\documentclass[twocolumn]{svjour3} % twocolumn
22: %
23: \smartqed % flush right qed marks, e.g. at end of proof
24: %
25: \usepackage{graphicx}
26: %
27: % \usepackage{mathptmx} % use Times fonts if available on your TeX system
28: %
29: % insert here the call for the packages your document requires
30: %\usepackage{latexsym}
31: % etc.
32: %
33: % please place your own definitions here and don't use \def but
34: % \newcommand{}{}
35: %
36: % Insert the name of "your journal" with
37: \journalname{Astrophysics and space science}
38: %
39: \begin{document}
40:
41: \title{Kelvin-Helmholtz instability in a weakly ionized layer}
42:
43:
44: \titlerunning{Kelvin-Helmholtz instability} % if too long for running head
45:
46: \author{Mohsen Shadmehri \and Turlough P. Downes }
47:
48: \authorrunning{Shadmehri \& Downes} % if too long for running head
49:
50: \institute{M. Shadmehri \at
51: School of Mathematical Sciences, Dublin City University, Glasnevin, Dublin 9, Ireland \\
52: Department of Physics, School of Science, Ferdowsi University, Mashhad, Iran\\
53: \email{mohsen.shadmehri@dcu.ie} % \\
54: % \emph{Present address:} of F. Author % if needed
55: \and
56: T. P. Downes \at
57: School of Mathematical Sciences, Dublin City University, Glasnevin, Dublin 9, Ireland\\
58: \email{turlough.downes@dcu.ie}
59: }
60:
61: \date{Received: date / Accepted: date}
62: % The correct dates will be entered by the editor
63:
64:
65: \maketitle
66:
67: \begin{abstract}
68: %
69: We study the linear theory of Kelvin-Helmholtz instability in a
70: layer of ions and neutrals with finite thickness. In the
71: short wavelength limit the thickness of the layer has a negligible
72: effect on the growing modes. However, perturbations with wavelength
73: comparable to layer's thickness are significantly affected by the
74: thickness of the layer. We show that the thickness of the layer has
75: a stabilizing effect on the two dominant growing modes. Transition
76: between the modes not only depends on the magnetic strength, but
77: also on the thickness of the layer.
78:
79:
80: \keywords{instabilities \and ISM: general \and ISM: jets and outflows}
81: % \PACS{PACS code1 \and PACS code2 \and more}
82: % \subclass{MSC code1 \and MSC code2 \and more}
83: \end{abstract}
84:
85: \section{Introduction}
86: \label{intro}
87:
88: %
89: When two fluids are in relative motion on either side of a common
90: boundary, the Kelvin-Helmholtz (KH) instability can occur. This
91: instability, the simplest example of shear flow instability, is a
92: well-known phenomenon in fluid mechanics and astrophysics (e.g.\
93: Chandrasekhar 1961). For example, the interactions between the solar
94: wind and planetary magnetospheres are studied based on the KH
95: instability (e.g. Nagano 1979). The KH instability, because of the
96: velocity shear in the mass flow between coronal plume structure and the
97: interplume region of the sun, could be a potential source for some of
98: the Alfv\'enic fluctuations observed in the solar wind (Andries, Tirry \&
99: Goossens 2000). The structure and the helical wave motion in the ionized
100: cometary tails as studied by Ershkovich \& Mendis (1986), Ershkovich,
101: Prialnik \& Eviatar (1986) and Ershkovich, Flammer \& Mendis (1986) using
102: the KH theory. The KH instability theory has been successfully used to
103: interpret the structure of the pc-scale jet in the radio source 3C273
104: (Lobanov \& Zensus 2001). Birk et al.\ (2000) studied the role of KH
105: instabilities in superwinds of primeval galaxies. The effect of radiative
106: losses on the evolution of the KH instability in jets and outflows has been
107: studied by many authors, either in linear regime (e.g., Massaglia et al. 1992; Hardee \& Stone 1997) or by direct numerical simulations (e.g., Rossi et al.
108: 1997; Downes \& Ray 1998). Recently, Michikoshi \& Inutsuka (2006) studied
109: the KH instability of the dust layer in protoplanetary discs to understand
110: the effect of relative motion between gas and dust.
111:
112:
113:
114:
115:
116: However, in many astrophysical flows the single fluid approximation is
117: inappropriate. For example, the fact that the winds and the intergalactic
118: medium may be only partially ionized should be considered in treatment
119: of role of the KH instability in superwinds of primeval galaxies. We
120: may have a similar situation at the interface between a jet and its
121: cocoon if the KH instability is to be viewed as a possible mechanism
122: for entrainment. The problem becomes more complicated if we note that
123: the development of the KH instability is strongly influenced by the
124: magnetic field (e.g.\ Malagoli et al. 1996; Jones et al. 1997).
125: Therefore it seems a two-fluid treatment of KH instability, taking
126: account of the different motions of ions and neutrals and of the
127: magnetic field, rather than a pure one fluid MHD description is a more
128: appropriate approach.
129:
130: By neglecting the thermal pressure forces of the neutrals or
131: perturbations in the neutral gas bulk velocities, the KH instability of
132: a system consisting of ions and neutrals with applications to partially
133: ionized cometary ionopauses, has been studied for both compressible
134: (Ershkovich \& Mendis 1986) and incompressible cases (Ershkovich,
135: Prialnik \& Eviatar 1986; Ershkovich, Flammer \& Mendis 1986). Chhajlani
136: \& Vyas (1991) incorporated the effects of finite resistivity and
137: rotation on the KH instability of two superposed fluids slipping past
138: each other. They showed that small rotation and the presence of neutral
139: particles destabilize the system. Birk et al.\ (2000) studied the KH
140: instability by taking into account the full incompressible dynamics of
141: both the neutral and the ionized gas components with applications to
142: multi-phase galactic outflow winds. In another similar study, Watson et
143: al.\ (2004) (hereafter WZHC) investigated the KH instability in the
144: linear, partially ionized regime to determine its possible effect on entrainment in massive bipolar outflows. They showed that for much of the relevant parameter space, neutral and ions are sufficiently decoupled that the neutrals are unstable while ions are held in place by the magnetic field. Birk \& Wiechen (2002) focused on unstable shear flows in partially ionized dense dusty plasma. They considered the dust and neutral gas components so that dust and neutral collisions is the dominant momentum transfer mechanism and dust component can interact with magnetic field lines, although dust charge fluctuations are negligible. They showed long wavelength modes can be stabilized by dust and neutral gas collisional momentum transfer. By doing multifluid numerical simulations, Wiechen (2006) studied KH modes in partially ionized, dusty plasmas for different masses and charges of dust.
145:
146: Analytical studies of KH instability in weakly ionized medium have so
147: far considered one interface between two flows. For example, WHZC
148: studied incompressible KH instability in a two-fluid system (charged
149: and neutral) in order to understand entrainment in outflows from massive
150: stars. They did the calculations in planar geometry with an interface at
151: $z=0$. However, it is not unreasonable to suppose that we may have a
152: flowing layer with finite thickness and two interfaces. In this paper,
153: we generalize the WHZC analysis by considering a finite thickness for a
154: weakly ionised flow. We will show that the growth time-scale of the
155: instability increases due to the thickness of the layer in particular
156: in the long wavelength limit. In the next section, we will present the basic equations and assumptions of the model. In the section 3, we will discuss growing modes and their dependence on the thickness of the layer.
157:
158:
159: \section{General Formulation}
160: \label{sec:2}
161: %
162: We follow the analysis of Chandrasekhar (1961) but take account of the
163: different bulk velocities and densities of the neutral and ions, both
164: inside and outside the layer. These two components of our model are
165: coupled by collisional interactions. We assume incompressibility for
166: the analytical calculations and also that the convective term in the
167: induction equation dominates the resistive one. For doing linear analysis, the unperturbed properties of the system are important. We suppose that the streaming takes place in the $x-$direction with velocity $U(z)$,
168: \[U(z) = \left\{
169: \begin{array}{l l}
170: +U & \quad \mbox{for $|z| \leq d$}\\
171: -U & \quad \mbox{for $|z| > d$}\\ \end{array} \right. \]
172: %
173: %
174: ($U$ is constant) and the neutral and ion components have the same
175: velocity in the unperturbed state. Note that $2d$ is the thickness
176: of the layer and $d\geq 0$. The magnetic field is assumed parallel
177: to the interface; that is, ${\bf B} = B_{x} {\bf e}_{x}$. Finally,
178: all unperturbed physical quantities are assumed constant in each
179: medium. The system under consideration consists of charged and
180: neutral fluid components coupled by collisional interactions. The
181: basic equations are
182:
183:
184: %
185: \begin{equation}
186: \nabla . {\bf v}_{n,i}=0,
187: \end{equation}
188: %
189: \begin{equation}
190: \rho_{n}(\frac{\partial{\bf v}_{n}}{\partial t}+ ({\bf v}_{n}.\nabla){\bf v}_{n})=-\nabla p_{n}- \gamma_{ni}\rho_{n}\rho_{i}({\bf v}_{n}-{\bf v}_{i})
191: \end{equation}
192: %
193: \begin{equation}
194: \rho_{i}(\frac{\partial{\bf v}_{i}}{\partial t}+ ({\bf v}_{i}.\nabla){\bf v}_{i})=-\nabla p_{i}+\frac{1}{4\pi}(\nabla\times{\bf B})\times {\bf B}- \gamma_{ni}\rho_{i}\rho_{n}({\bf v}_{i}-{\bf v}_{n})
195: \end{equation}
196: %
197: %
198: \begin{equation}
199: \frac{\partial {\bf B}}{\partial t}=\nabla\times ({\bf v}_{i}\times{\bf B}).
200: \end{equation}
201: %
202: Also note that $\nabla.{\bf B}=0$. In the above equations, $\gamma_{ni}$ is the collision rate coefficients per unit mass so that $\nu_{ni}=\gamma_{ni}\rho_{i}$ is the neutral-ion frequency. The collision frequency determine the scale of the coupling between the different components and the coupling between each component and the magnetic field.
203:
204: Now, we perturb the physical variables as $\chi (z,x,t) = \chi'(z) \exp[i(\omega t + k_{x}x)]$, and so
205: %
206: \begin{equation}
207: -k_{x}\rho_{ n} u_{n}' + i \rho_{n}\frac{dw_{n}'}{dz}=0,
208: \end{equation}
209: %
210: \begin{equation}
211: \phi\rho_{n}u_{n}'=-k_{x}p_{n}'+i\gamma_{ni}\rho_{n}\rho_{i}(u_{n}'-u_{i}'),
212: \end{equation}
213: %
214: \begin{equation}
215: \phi\rho_{n}w_{n}'=i\frac{dp_{n}'}{dz}+i\gamma_{ni}\rho_{n}\rho_{i}(w_{n}'-w_{i}'),
216: \end{equation}
217: %
218: where $\phi=\omega + k_{x}U$, and $u_{n}'$ and $w_{n}'$ are the $x$ and the $z$ components of the perturbed velocity of the neutrals. Also, the equations of the ions are
219: %
220: \begin{equation}
221: -k_{x}\rho_{ i} u_{i}' + i \rho_{i}\frac{dw_{i}'}{dz}=0,
222: \end{equation}
223: %
224: \begin{equation}
225: \phi\rho_{i}u_{i}'=-k_{x}p_{i}'+i\gamma_{ni}\rho_{n}\rho_{i}(u_{i}'-u_{n}'),
226: \end{equation}
227: %
228: \begin{equation}
229: \phi\rho_{i}w_{i}'=i\frac{dp_{i}'}{dz}-\frac{B_{x}^{2}}{4\pi\phi}\frac{d^{2}w_{i}'}{dz^{2}}+\frac{k_{x}^{2}B_{x}^{2}}{4\pi}\frac{w_{i}'}{\phi}
230: +i\gamma_{ni}\rho_{n}\rho_{i}(w_{i}'-w_{n}'),
231: \end{equation}
232: %
233: where $u_{i}'$ and $w_{i}'$ are the $x$ and the $z$ components of the perturbed velocity of the ions. We can reduce the above differential equations to a set of two
234: differential equations for $w_{i}'$ and $w_{n}'$ as follows (WZHC)
235: %
236: %
237: \begin{equation}\label{eq:main2}
238: (\rho_{i}\phi - \frac{B^{2}k_{x}^2}{4\pi\phi}){\mathcal D}w_{i}'=i\gamma_{ni}\rho_{n}\rho_{i}{\mathcal{D}}(w_{i}'-w_{n}'),
239: \end{equation}
240: %
241: %
242: \begin{equation}\label{eq:main1}
243: \rho_{n}\phi {\mathcal{D}}w_{n}'=i\gamma_{ni}\rho_{n}\rho_{i}{\mathcal{D}}(w_{n}'-w_{i}'),
244: \end{equation}
245: %
246: where $\mathcal{D}\equiv \frac{d^{2}}{dz^{2}}-k_{x}^{2}$. The behavior
247: of the flow both inside and outside the layer is determined by the
248: general solutions of the linear differential equations (\ref{eq:main1})
249: and (\ref{eq:main2}). One can easily show that the general solutions of these equations are linear combinations of $\exp(+k_{x}z)$ and $\exp(-k_{x}z)$. We impose the condition that the perturbed quantities do not diverge at $z=+\infty$ and $z=-\infty$. Thus, the general solutions are
250: %
251: \begin{equation}w_{i}'(z) = \left\{
252: \begin{array}{l l}
253: C_1 e^{+k_{x}z} + C_{2} e^{-k_x z}& \quad \mbox{for $0 \leq z \leq d$}\\
254: C_3 e^{-k_{x}z} & \quad \mbox{for $d \leq z$}\\ \end{array} \right.
255: \end{equation}
256:
257:
258: \begin{equation}w_{n}'(z) = \left\{
259: \begin{array}{l l}
260: C_4 e^{+k_{x}z} + C_{5} e^{-k_x z}& \quad \mbox{for $0 \leq z \leq d$}\\
261: C_6 e^{-k_{x}z} & \quad \mbox{for $d \leq z$}\\ \end{array} \right.
262: \end{equation}
263: %
264: where $C_1$, $C_2$, $C_3$, $C_4$, $C_5$ and $C_6$ are constants which
265: are determined by the boundary conditions. We note that, because of
266: symmetry of the unperturbed state, there are solutions where $w_{i}(z)
267: = - w_{i} (-z)$ and $w_{n}(z) = - w_{n} (-z)$ (odd solutions) and solutions where $w_{i}(z) = w_{i} (-z)$ and $w_{n}(z)
268: = w_{n} (z)$ (even solutions). Thus, we can write the boundary condition at $z=0$ as
269: %
270: \begin{equation}
271: C_{1} + S C_{2} = 0,\label{eq:bound1}
272: \end{equation}
273: %
274: \begin{equation}
275: C_{4} + S C_{5}=0,\label{eq:bound2}
276: \end{equation}
277: %
278: where $S = +1$ and $S= -1$ are corresponding to even and odd solutions, respectively. Thus, considering symmetry properties of the solutions (i.e. equations (\ref{eq:bound1}) and (\ref{eq:bound2})) our solutions are also valid for $z<0$, if we transform $z\rightarrow -z$. So, the solutions correctly vanish, as $z\rightarrow -\infty$.
279:
280: Another boundary condition is the continuity of the ion and neutral displacements across the interfaces. So, the boundary conditions at $z=d$ are
281: %
282: \begin{equation}
283: \frac{C_1 e^{k_x d} + C_2 e^{-k_x d}}{\omega + k_x U}=\frac{C_3 e^{-k_x d}}{\omega - k_x U},\label{eq:bound3}
284: \end{equation}
285: %
286: \begin{equation}
287: \frac{C_4 e^{k_x d} + C_5 e^{-k_x d}}{\omega + k_x U}=\frac{C_6 e^{-k_x d}}{\omega - k_x U}.\label{eq:bound4}
288: \end{equation}
289: %
290: Having equations (\ref{eq:bound1}), (\ref{eq:bound2}), (\ref{eq:bound3}) and (\ref{eq:bound4}), we can rewrite $w_i$ and $w_n$ as
291: %
292: %
293: \begin{equation}w_{i}'(z) = \left\{
294: \begin{array}{l l}
295: C_2 (e^{-k_x z} - S e^{k_x z})& \quad \mbox{for $0 \leq z \leq d$}\\
296: C_2 \frac{\omega - k_x U}{\omega + k_x U} (1-S e^{2k_{x}z}) & \quad \mbox{for $d \leq z$}\\ \end{array} \right.\label{eq:C2}
297: \end{equation}
298:
299:
300: \begin{equation}w_{n}'(z) = \left\{
301: \begin{array}{l l}
302: C_5 (e^{-k_x z} - S e^{k_x z})& \quad \mbox{for $0 \leq z \leq d$}\\
303: C_5 \frac{\omega - k_x U}{\omega + k_x U} (1-S e^{2k_{x}z})& \quad\mbox{for $d \leq z$}\\ \end{array} \right.\label{eq:C5}
304: \end{equation}
305: %
306:
307:
308: Clearly, the differential equations (\ref{eq:main1}) and (\ref{eq:main2}) are not valid at the interfaces $z=d$ (or $z=-d$), where velocities of the species are not analytic. However, we can integrate each equation over an infinitesimal region around the interface $z=d$ and use standard Gaussian pillbox arguments. Then we have
309: %
310: \begin{equation}
311: \triangle \{\rho_{n}\phi \frac{dw_{n}'}{dz}\}=i\triangle \{\gamma_{ni}\rho_{n}\rho_{i} \frac{d}{dz}(w_{n}'-w_{i}') \},
312: \end{equation}
313: %
314: \begin{equation}
315: \triangle \{\rho_{i}\phi \frac{dw_{i}'}{dz}\}=i\triangle \{\gamma_{ni}\rho_{n}\rho_{i} \frac{d}{dz}(w_{i}'-w_{n}') \}+\frac{B_{x}^{2}k_{x}^{2}}{4\pi}\triangle\{\frac{1}{\phi}\frac{dw_{i}'}{dz}\},
316: \end{equation}
317: %
318: By substituting equations (\ref{eq:C2}) and (\ref{eq:C5}) into the above equations and after lengthy (but straightforward) mathematical manipulations, we obtain
319: %
320: \begin{equation}
321: [\xi^2 + m\xi - (1-\lambda^2)h^2 + a^2] C_2 - m\xi C_5 =0,\label{eq:C2main}
322: \end{equation}
323: \begin{equation}
324: -\xi C_2 + [\xi^2 + \xi - (1-\lambda^2)h^2] C_5 =0,\label{eq:C5main}
325: \end{equation}
326: %
327: where
328: \begin{equation}
329: m=\frac{\rho_n}{\rho_{i}}, h = \frac{Uk_x}{\nu}, a = \frac{k v_{Ai}}{\nu}, \frac{\omega}{\nu}-\lambda h = -i\xi,
330: \end{equation}
331: %
332: and $\lambda = S^{-1}\exp (-2kd)$ and the ion Alfv\'en velocity $v_{Ai}$ is defined as $B/\sqrt{4\pi \rho_i}$. By the condition that the algebraic equations (\ref{eq:C2main}) and (\ref{eq:C5main}) have a nontrivial solution, we obtain the dispersion relation
333: %
334: \begin{displaymath}
335: \xi^4 + (m+1) \xi^3 - [2(1-\lambda^2)h^2-a^2]\xi^2
336: \end{displaymath}
337: \begin{equation}
338: -[(m+1)(1-\lambda^2)h^2-a^2]\xi + (1-\lambda^2)[(1-\lambda^2)h^2-a^2]h^2=0.\label{eq:mainm}
339: \end{equation}
340: %
341: This algebraic equation (\ref{eq:mainm}) is the basis of our stability analysis.
342: As we see, the effect of the thickness of the layer appears by term $\lambda^2$ which depends on the thickness of the layer. For a system with one interface at $z=0$, WZHC obtained this algebraic equation
343: %
344: \begin{displaymath}
345: \xi^4 + (m+1) \xi^3 - (2h^2-a^2)\xi^2-[(m+1)h^2-a^2]\xi + (h^2-a^2)h^2=0.
346: \end{displaymath}
347: %
348: We note that our algebraic equation (\ref{eq:mainm}) reduces to the above equation if we set $\lambda = 0$.
349:
350: \begin{figure}
351: \includegraphics[width=12cm]{figure1.eps}
352: \caption{Growth rate of the instability vs. wavelength of the
353: perturbations for two cases $B=1$ $\mu$G ({\it top}) and $B = 1$ mG
354: ({\it bottom}). In both cases, it is assumed $U=10$ Km s$^{-1}$. The
355: dashed curves correspond to the WZHC analysis and the other
356: curves are labeled by the thickness of layer. Black circles in the
357: bottom plot denote the wavelength at which the growing mode changes from
358: $(1-\lambda^2)^{1/2}h$ to $(1-\lambda^2)h^2$.} \label{fig:1}
359: \end{figure}
360:
361: \section{Analysis}
362: \label{sec:3}
363: %
364: Roots of equation (\ref{eq:mainm}) with positive real parts correspond to
365: growing unstable modes. This equation is applicable for various
366: weakly ionized systems. However, in this work the input parameters are chosen
367: to be appropriate for molecular outflows. This makes for easier comparison
368: with the analysis of WZHC. In our illustrative examples, the relative
369: velocity $2U$ is 20 Km s$^{-1}$, and the ratio, $m$, of neutral to
370: ionized density is assumed to be around $10^6$. The collision frequency is
371: $\nu=1.5\times 10^{-13}$ s$^{-1}$ (Draine et al.\ 1983). The magnetic
372: field strength is taken to be either $1$ $\mu$G (weak-field case) or
373: $1$ mG (strong-field case). With these parameters the ratio of the
374: Alfven velocity to the relative velocity is 4400 and 4.4 in the strong and weak magnetic field cases, respectively. Having these input parameters, we can solve the algebraic equation (\ref{eq:mainm}) numerically to find the
375: dispersion relation and the fastest growing modes.
376:
377: Figure \ref{fig:1} shows plots of the dispersion relation for the
378: above input parameters. When $\lambda \rightarrow 0$, the dispersion
379: relation tends to WZHC analysis (dashed curves in Figures \ref{fig:1}),
380: as expected. Each curve is also labeled by the thickness of
381: the layer, i.e. $2d$. It can be seen from Figure \ref{fig:1} that the effect
382: of the thickness of the layer is not significant for perturbations with
383: short wavelength. However, as the
384: wavelength of the perturbations increases, the thickness of the layer has a
385: stabilizing effect on the growing modes and the dispersion relation deviates
386: from the analysis of WZHC both in the weakly and strongly magnetised
387: cases. Considering the thickness of the layer implies longer growth
388: time-scale for the perturbations, in particular in the long wavelength
389: limit. Moreover, as one would expect, as the thickness of the layer
390: decreases the deviation from the WZHC analysis occurs at shorter
391: wavelength, and also the growth time-scale becomes longer.
392:
393:
394: Interestingly, the roots of equation (\ref{eq:mainm}) which correspond
395: to growing modes can be described by approximate analytical
396: solutions. There are three approximate positive roots for
397: this equation depending on its coefficients:
398:
399: \noindent (i) $(1-\lambda^2)^{1/2}h$, when $(1-\lambda^2)^{1/2}h >
400: 1$;
401:
402: \noindent (ii) $(1-\lambda^2)h^2$, when $(1-\lambda^2)^{1/2}h < 1$
403: and $(1-\lambda^2)^{1/2} < (a/h)m^{-1/2}$;
404:
405: \noindent (iii) $(1-\lambda^2)^{1/2}h
406: (1-(a/h)^2(1-\lambda^2)^{-1}m^{-1})^{1/2}$ and $(1-\lambda^2)^{1/2}
407: > (a/h)m^{-1/2}$.
408:
409: The agreement between these approximate roots and numerical roots
410: of the equation (\ref{eq:mainm}) is excellent. Thus, as long as
411: $(1-\lambda^2)^{1/2}h > 1$ the only growing mode is the first
412: growing mode, irrespective of the strength of the magnetic field
413: (see top plot of Figure \ref{fig:1}). Once this inequality is violated,
414: we may have the second growing mode, depending on whether
415: $(1-\lambda^2)^{1/2} < (a/h)m^{-1/2}$. For example, for $B=1$ mG we have
416: $(a/h)m^{-1/2}=4.5 > 1 \geq (1-\lambda^2)^{1/2}$ and so the transition from
417: the first to the second growing modes occurs when $(1-\lambda^2)^{1/2}h = 1$.
418: We can denote this transition wavelength by $\lambda_{t}$ and is
419: shown by black circles in Figure \ref{fig:1} (bottom plot). But in
420: the weakly magnetic case where $B=1$ $\mu$G, we have
421: $(a/h)m^{-1/2}=4.5\times 10^{-3} < (1-\lambda^2)^{1/2}$. So, for
422: wavelengths greater than $\lambda_{t}$, the third growing mode
423: appears, although the second term inside the parenthesis of this
424: root is much smaller than unity and the third mode actually tends
425: toward first mode (top plot). Considering our ranges of the input
426: parameters, the third mode is very close to the first mode.
427:
428: The first mode corresponds to the usual hydrodynamic instability, but the
429: second mode appears in our weakly ionized case as a new mode (WZHC). The transition between these two
430: modes depends on the thickness of the layer. Figure \ref{fig:2} shows the
431: transition wavelength $\lambda_{t}$ versus the the half thickness of the
432: layer. For thin layers, the transition wavelength is smaller than
433: the thick layers. In fact, as the thickness of the layer increases, this
434: transition wavelength increase and tends to an asymptotic value for the very
435: thick layers. WZHC analysis corresponds to this asymptotic regime.
436:
437:
438:
439:
440: \begin{figure}
441: \includegraphics[width=12cm]{figure2.eps}
442: \caption{The transition wavelength $\lambda_{t}$ vs. the half
443: thickness $d$ of the layer.} \label{fig:2}
444: \end{figure}
445:
446: \section{Conclusions}
447: \label{sec:4}
448: %
449: In this study we have extended the analysis of WZHC by considering a
450: layer of ions and neutrals with finite thickness interacting with an
451: ambient medium in the incompressible limit. However, this model can not
452: be applied directly for systems, such as outflows or jets because of
453: our simplifying assumptions. We can,
454: however, gain insight into the possible effects of shear between a moving
455: layer of ions and neutrals and the surrounding medium.
456:
457: To our knowledge, there are a few numerical simulations of KH
458: instability of a multifluid system (Birk et al. 2000; Birk \&
459: Wiechen 2002; Wiechen 2006). While Birk et al. (2000) showed that KH
460: modes can operate fast enough to amplify the magnetic field
461: strengths in superwinds of primeval galaxies within the timescale of
462: the outflow, Birk \& Wiechen (2002) and Wiechen (2006) studied KH
463: instability in a dusty magnetized medium. It is difficult to make a
464: direct comparison between our results and these numerical
465: simulations because of different basic assumptions. We considered an
466: incompressible two-fluid layer, but all these numerical simulations
467: are for compressible systems. Moreover, the effect of charged dust
468: particles are considered in Birk \& Wiechen (2002) and Wiechen
469: (2006), though we have neglected charged grains in our analysis.
470:
471:
472: We have shown that the growth time-scale of KH instability in a layer
473: of ions and neutrals increases when compared to a system with one
474: interface, this effect being more evident in the long wavelength limit.
475: There are two dominant growing modes of KH instability in such a
476: system, depending on the magnetic strength and the thickness of the
477: layer. As the thickness of the layer decreases, the transition between
478: these two unstable modes occurs at shorter wavelengths. Given these
479: results, it remains to future work to extend this analysis to compressible case and make
480: direct comparisons with jets or outflows with finite thickness.
481:
482:
483: \begin{acknowledgements}
484: The research of M. S. was funded under the Programme for Research in
485: Third Level Institutions (PRTLI) administered by the Irish Higher
486: Education Authority under the National Development Plan and with partial
487: support from the European Regional Development Fund.
488: \end{acknowledgements}
489:
490: % BibTeX users please use one of
491: %\bibliographystyle{spbasic} % basic style, author-year citations
492: %\bibliographystyle{spmpsci} % mathematics and physical sciences
493: %\bibliographystyle{spphys} % APS-like style for physics
494: %\bibliography{} % name your BibTeX data base
495:
496: % Non-BibTeX users please use
497: \begin{thebibliography}{}
498: %
499: % and use \bibitem to create references. Consult the Instructions
500: % for authors for reference list style.
501: %
502: \bibitem{} Andries, J., Tirry, W. J., Goossens, M.: ApJ, {\bf 531}, 561 (200)
503:
504: \bibitem{} Blandford, R. D., Pringle, J. E.: MNRAS, {\bf 176}, 443 (1976)
505:
506: \bibitem{} Birk, G. T., Wiechen, H., Lesch, H., Kronberg, P. P.: A\&A, {\bf 353}, 108 (2000)
507:
508: \bibitem{} Birk, G. T., Wiechen, H.: Physics Plasmas, {\bf 9}, 964 (2002)
509:
510: \bibitem{} Chandrasekhar, S.: Hydrodynamics and Hydromagnetic Stability (Oxford: Oxford Univ. Press) (1961)
511:
512: \bibitem{} Chhajlani, R. K., Vyas, M. K.: Ap\&SS, {\bf 176}, 69 (1991)
513:
514:
515: \bibitem{} Downes, T. P., Ray, T. P.: A\&A, {\bf 331}, 1130 (1998)
516:
517:
518: \bibitem{} Draine, B. T., Roberge, W. G., Dalgarno, A.: ApJ {\bf 264},
519: 485 (1983)
520:
521: \bibitem{} Ershkovich, A. I., Mendis, D. A.: ApJ, {\bf 302}, 849 (1986)
522:
523: \bibitem{} Ershkovich, A. I., Prialnik, D., Eviatar, A.: J. Geophys. Res., {\bf 91}, 8782 (1986)
524:
525: \bibitem{} Ershkovich, A. I., Flammer, K. R., Mendis, D. A.: ApJ, {\bf 311}, 1031 (1986)
526:
527: \bibitem{} Hardee, P. E., Stone, J. M.: ApJ, {\bf 483}, 121 (1997)
528:
529: \bibitem{} Jones, T. W., Gaalaas, J. B., Ryu, D., Frank, A.: ApJ, {\bf 482}, 230 (1997)
530:
531: \bibitem{} Lobanov, A. P., Zensus, J. A.: Science, {\bf 294}, 128 (2001)
532:
533: \bibitem{} Malagoli, A., Bodo, G., Rosner, R.: ApJ, {\bf 456}, 708 (1996)
534:
535: \bibitem{} Massaglia, S., Trussoni, E., Bodo, G., Rossi, P., Ferrari, A.: A\&A, {\bf 260}, 243 (1992)
536:
537: \bibitem{} Michikoshi, S., Inutsuka, S.-I.: ApJ, {\bf 641}, 1131 (2006)
538:
539: \bibitem{} Nagano, H.: Planetray and Space Science, {\bf 27}, 881 (1979)
540:
541: \bibitem{} Ray, T.: Planetary and Space Science, {\bf 30}, 245 (1982)
542:
543: \bibitem{} Rossi, P., Bodo, G., Massaglia, S., Ferrari, A.: A\&A, {\bf 321}, 672 (1997)
544:
545: \bibitem{} Watson, C., Zweibel, E.G., Heitsch, F., Churchwell, E.: ApJ {\bf 608}, 274
546: (2004) (WZHC)
547:
548: \bibitem{} Wiechen, H. M.: Physics of Plasmas, {\bf 13}, 062104 (2006)
549:
550:
551:
552: \end{thebibliography}
553:
554: \end{document}
555: % end of file template.tex
556: