Explosion of soliton in a magnetic field

Katsuhiro Nishinari, Kanji Abe, and Junkichi Satsuma

Citation: Physics of Plasmas (1994-present) 1, 3728 (1994); doi: 10.1063/1.870910

View online: http://dx.doi.org/10.1063/1.870910
View Table of Contents: http://scitation.aip.org/content/aip/journal/pop/1/12?ver=pdfcov
Published by the AIP Publishing

Articles you may be interested in

Direct determination of ion wave fields in a hot magnetized and weakly collisional plasma
Phys. Plasmas 3, 4316 (1996); 10.1063/1.871581

Dispersive properties of a magnetized plasma with a fieldaligned drift and inhomogeneous transverse flow
Phys. Plasmas 3, 3091 (1996); 10.1063/1.871656

Far field ICRF sheath formation on walls and limiters

AIP Conf. Proc. 289, 421 (1994); 10.1063/1.44976

Chaotic particle dynamics in a periodic focusing quadrupole magnetic field

Phys. Plasmas 1, 1328 (1994); 10.1063/1.870731

Radio frequency effects on confinement and equilibrium in a magnetic mirror

Phys. Fluids B 3, 3153 (1991); 10.1063/1.859794

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
128.114.34.22 On: Sun, 30 Nov 2014 00:16:13
 Explosion of soliton in a magnetic field
Katsuhiro Nishinari
Department of Aeronautics. Faculty of Engineering. The University of Tokyo. Hongo 7-3-1. Bunkyo-ku.
Tokyo 113. Japan
Kanji Abe
College of Arts and Sciences. The University of Tokyo. Komaba 3-8-1. Meguroku. Tokyo 153. Japan
Junkichi Satsuma
Department of Mathematical Sciences. The University of Tokyo. Komaba 3-8-1, Meguroku, Tokyo 153.
Japan
(Received 12 July 1994; accepted 9 September 1994)
A dynamics of a solitary pulse of the electrostatic ion cyclotron wave that propagates perpendicular
to an applied magnetic field is considered. It is shown that the solitary wave will be singular in some
range of parameters in the system, such as the plasma density and the magnitude of an applied
magnetic field. This fact shows that there is a possibility of controlling the place where explosion
of the solitary wave occurs. © 1994 American Institute of Physics.

Although nonlinear integrable dynamics in multidimen- A <fJ=exp <fJ-n, (3)

sion has been an object of study for the last few decades, it is
not well understood yet. The soliton theory is mostly power- where n, v=(vx,vy,v z ) are the density and the velocity of
ful only in one spatial dimension. The Davey-Stewartson ions, respectively, <fJ is the electrostatic potential, b=(1,O,O),
(DS) equations 1, however, are one of the few multidimen- and A = a2/ ax 2+ a21ay2 + a2/ az 2. The nondimensional pa-
sional equations that can be solved by the inverse scattering rameter a is given by a=w/Oj, where wi=ZeBolMc and
transformation 2 and bilinear method. 3 In recent years, there 0; = 4 7TnoZe2/ M. Here, no is the initial unperturbed ion
has been a renewal of interest in the DS equations. The dro- density, M and Z are the mass and the charge number of an
mion solution admitting a two dimensionally localized struc- ion, respectively, - e is the charge of an electron, c is the
ture was found for these equations. 4 Recently, we have speed of light, and Bo is the magnitude of the applied mag-
shown that the electrostatic ion cyclotron wave that propa- netic field.
gates perpendicular to an applied magnetic field is well de- When we consider the case of perpendicular propaga-
scribed by the DS equations. s Shimizu et a1. 6 studied in de- tion, we must take into account the fact that electron inertia
tail dynamics of such a short-wavelength mode of an ion should not be neglected. Then, (3) is not valid in an exact
wave in one spatial dimension, and recently we developed sense, because electrons cannot follow the ion motion per-
their work in multidimensions. 5•7 pendicular to the magnetic field because of its small Larmor
In this Letter we consider a solitary wave solution of the radius, Thus, (3) is valid, only in the range that the angle
DS equations given in Ref. 5, and show that the solution has between the vector normal to the direction of the magnetic
singularity in some range of parameters in the system, such field and the wave number vector is larger than vithlveth
as the plasma density and the magnitude of an applied mag- - ~(mTi)/(MTe)' where m is the mass of an electron, and
netic field. This study is motivated by a recent work by Mal- v ith and Veth are ion and electron thermal velocities, respec-
omed et at. 8 They numerically found a localized solution in tively. The magnitude of ~(mTi)/(MTe) is, however, very
the generalized derivative nonlinear Schrodinger (NLS) small usually, so that (3) is still valid for "perpendicular"
equation in one spatial dimension, and showed that the solu- propagation in this sense.
tion exists only in a bounded parametric region. In this Letter The linear dispersion relation of this system is given by 5
we developed their result and may suggest useful applica-
tions. such as an energy transport in a plasma system.
Let us consider a plasma in a magnetic field applied to
4 (2
w -
!k(2) 2 2 2( 1 ) _
a + 1+lk1 2 w +a (b·k) 1+lk1 2 -0. (4)
the x direction. We assume that the plasma is collision less
and described by its fluid behavior. We also assume that the
This dispersion relation has two modes that here we call the
temperature of electrons is so high that the temperature of
higher mode and the lower mode. In the following, we will
ions can be neglected, and electrons are in thermal equilib-
consider the higher mode. The equations describing the dy-
rium. The basic equations in a nondimensional form govern-
namics of a wave packet, which propagates in the direction
ing such plasma dynamics ares.?
perpendicular to the magnetic field are derived by means of
the reductive perturbation method. We expand the physical
an quantities around their stationary values as
-+div(nv)==O, (1)
at
00 '"

av n = 1 + ~ en ~ n)n)( g, 1], T)exp[il(k·x- wt)], (5)

- grad <fJ+a(vxb), (2)
at + (v·grad)v= - n=l l=-oo

3728 Phys. Plasmas 1 (12), December 1994 1070-664X/94/1 (12)/3728/3/$6.00 © 1994 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
128.114.34.22 On: Sun, 30 Nov 2014 00:16:13
 and similarly for <P and v. We calculate, up to the third order (a3+ s / a 4)
of E, and finally obtains
f3= albi/(2alP)2+a2b~/(2a2q)2' (16)

iA r + aIAU+ a2A7]7]+ a31AI2A + a4AQ=O, (6) The equation (15) is generally solved by the elliptic function.
2 For the homoclinic orbit of this system, the solution is ex-
Q~e+ 01 Q 7]7]+ 02(IAI )u=o, (7)
pressed by
where A = n\l) and Q = nb 2
) + 03V~~). The coefficients aj,
i = I, 2, 4 and 0;, i = I, 2 are given by 8ad exp( rat)
R= f3+8ad 2 exp(2Ta?)' (17)
I a2wl2 I a2wl
0'1 =2 ak ,0'2=2 ak2 , where d is an arbitrary constant. If a>O and f3>O, (17) gives
x kx=O Y kx=O a regular solitary wave. The solution (17) will be singular if
f3<O. Existence of this type of solitary wave solution indi-
cates that there is a possibility to realize an explosion. That
is, if a is kept positive and f3 becomes negative, then the
See Ref. 5 for coefficients 0'4 and b':l, which are complicated solitary wave will explode. This is easily imaged by the fact
functions of the wave number of the carrier wave, the plasma that the maximum amplitude of R, J2 al f3, goes to infinity
density, and the magnitude of magnetic field. The system (6) as f3'-,,0.
and (7) is the well-known DS equations. It must be noted here that a and f3 given in (16) are
We now consider the solitary wave solution of the DS functions of the plasma density, and the magnitude of the
equations. Let us assume that the solution is of the form magnetic field, both of which can be controlled from outside
the system. Using this fact, a scenario of the explosion is
A =R(x,y,t)exp i(px+qy+ rt), (9) described as follows: First, consider a situation where a soli-
tary wave is excited and propagates in the region where a>O
where R(x,y,t) is a real function and p, q, and r are real
and f3>O. We expect to have a stable traveling wave in this
constants. We substitute (9) into (6) to obtain
region. Assume that the system is inhomogeneous and the
R t + 2alpRx+2a2qRy=O, (10) magnitude of the plasma density or the magnetic field varies
locally. Then, due to a change in the value of a, we have a
from the imaginary part. A solution of (10) is
change that f3 becomes negative while a is kept positive. The
b1x b 2y solitary wave explodes at that place. Of course, we can arti-
R =R(~), ~= -2-+ -2--(b l +b 2)t. (11) ficially give such an inhomogeneity to the system, and con-
alP a2q
trol the place where explosion occurs. We think that this idea
In (11), b l and b 2 are constants that determine the direction will be useful for an energy transport in a plasma system.
of the wave propagation. Substituting (11) into (7), we find Perturbation analysis may not be applicable to the singu-
that lar case. Moreover, the existence of inhomogeneity may
cause other phenomena, such as deformation and fission of
(12) soliton.9 We consider, however, that the tendency for explo-
sion may appear in a real system. The reason of this is as in
where the prime means the derivative with respect to ~. In- the following. Let us consider "collapse," which was first
tegration of (12) gives discussed by Zakharov. 10 The explosion noted above may
have some resemblance to the collapse. The collapse is ex-
Q=sIR2+CO~+CI , (13) plained by the singularity of a localized solution in the mul-
tidimensional NLS equation, which is derived by the same
where
perturbation analysis as ours. Since the tendency for collapse
is observed in real systems such as plasma, optics, etc.,l1 we
may say that the explosion will also be expected to occur.
The physical mechanism of explosion, however, is dif-
and co' C I are integration constants. The constant Co can be ferent from that of collapse. In the case of collapse, the am-
taken to be zero from a physical point of view. plitude goes to infinity within a finite time, keeping its total
The real part of (6) gives energy. It is apparent, however, that in our explosion the total
energy goes to infinity. This is easily seen by considering the
- rR+ 0'1 (R xx - p2R) + a2(R yy -q 2R)+ a3R3+ a4RQ=0.
(14) area of cross section of the soliton given by f R d ~
= 'TrJ21 f3, which goes to infinity as f3'-,,0. The soliton is
We substitute (II) and (13) into (14) to obtain sustained by the balance between the nonlinear effect of the
R"-aR+f3R 3 =O, (15) ponderomotive force and the dispersion effect. From our re-
sults, we can artificially alter the balance and make the soli-
where ton taller by varying the parameter a. As its cross section of
it represents the wave energy, we can then conclude that the
explosion is caused by injecting energy into the soliton from
outside through a suitable change of the parameter a.

This articlePhys. Plasmas, as

is copyrighted Vol.indicated
1, No. 12, December
in the 1994 of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions.
article. Reuse Letters 3729to IP:
Downloaded
128.114.34.22 On: Sun, 30 Nov 2014 00:16:13
 0.1
ky f3 0.05

0.345 0.375 0.38

-0.05
a
-0.\

-0.15

-0.2

0.2 0.4 0.6 0.8 1.2 1.4

a FIG. 3. Relation between f3 and a for ky=0.8.

FIG. I. Dependence of the sign of ex on a and k, . The shaded parts and {3 becomes negative with a>O, which are shown by three
unshaded parts indicate the region where ex is negative and positive, respec- arrows in Fig. 2. Let us analyze one of these in detail. Figure
tively. On the thick curve, ex goes to infinity.
3 shows the a dependence of {3 for k y = 0.8. From Fig. 1 and
Fig. 3, we find a regular solitary wave exists in the region
Finally, we give an estimation of whether the above situ- 0.359<a. If a becomes smaller than 0.359, {3 becomes
ation really occurs or not in this system. This is done by negative with a>O.
studying the signs of a and {3 using (8). Since there are many In conclusion, we have proposed in this Letter a new
parameters in a and {3, we only give some examples. In the type of explosion, which is induced by changing physical
following calculation, all parameters except a and ky are parameters in the system. We think that this type of explo-
fixed as p=2, q= t, m =n = I, r= -2, and Cl =0. The signs sion occurs also in other fields, such as fluid dynamics, op-
of a and {3 are illustrated in Fig. 1 and Fig. 2, respectively. tics, and so on. Investigation of the possibility will be a
From these figures, we see that there exist three cases where subject of future work.

ACKNOWLEDGMENTS
We are grateful to Dr. P. M. Santini, Dr. M. Tanaka, Dr.
T. Yajima, and Mr. K. Matsui for helpful suggestions and
ky comments. Thanks are due to Ms. Yumiko Nojima for her
1.2
assistance in writing figures.
This work is partially supported by a Grant-in-Aid for
the Encouragement of Young Scientists from the Ministry of
Education, Science. and Culture (No. 2426).

1A. Davey and K. Stewartson, Proe. R. Soc. London Ser. A 338. 101
(1974).
2A. S. Fokas and M. J. Ablowitz. J. Math. Phys. 25. 2494 (I 984}.
3J. Satsuma and M. J. Ablowitz, J. Math. Phys. 20, 1496 (1979).
4M. BoW. J. J.-P. Leon. L. Martina. and F. Pempinelli, Phys. Lett. A 132,
432 (1988).
5K. Nishinari, K. Abe, and J. Satsuma, Phys. Plasmas 1, 2559 (1994).
o 0.2 0.4 0.6 0.8 1.2 1.4 6K. Shimizu and Y. H. Ichikawa, J. Phys. Soc. Jpn. 33, 789 (1972).
a 7K. Nishinari. K. Abe, and J. Satsuma, J. Phys. Soc. Jpn. 62, 2021 (1993).
8B. Malomed, K. Matsui. and J. Satsuma, "Solitons and collapse in the
generalized deri vati ve nonlinear Schriidinger equation," submitted to J.
FIG. 2. Dependence of the sign of f3 on a and k y . The shaded parts and Phys. Soc. Jpn.
unshaded parts indicate the region where f3 is negative and positive, respec- 9y. S. Kivshar and B. A. Malomed, Rev. Mod. Phys. 61. 763 (1989).
tively. On thick curves, f3 goes to infinity. Three arrows show ways of 1Ov. E. Zakharov. Zh. Eksp. Teor. Fiz. 62. 1745 (1972).
changing the sign of {3. which cause the explosion. 11M. V. Goldman. Rev. Mod. Phys. 56, 709 (1984).

3730 Phys. Plasmas, Vol. 1, No. 12, December 1994 Letters

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
128.114.34.22 On: Sun, 30 Nov 2014 00:16:13