AIAA JOURNAL

Vol. 48, No. 11, November 2010

Analytical Solution of Converging Shock

Wave in Magnetogasdynamics

L. P. Singh∗ and M. Singh†

Banaras Hindu University, Varanasi 221 005, India
and
Bishun D. Pandey‡
Ohio State University, Marion, Ohio 43302
DOI: 10.2514/1.J050244
In this paper, a similarity solution to a problem in magnetogasdynamics with strong converging cylindrical shock
wave is examined. An analytical description of converging shock waves is presented by replacing the previous
approach for numerical solutions of the ordinary differential equations with a theoretical study of the singular points
of the differential equations. A study of singular points of the system of differential equations leads to an analytic
description of the flowfield and a determination of the similarity exponent. The influences of adiabatic heat exponent
and magnetic field strength on the flow pattern for various cases are assessed. The general behavior of the velocity
and density distribution remains unaffected. However, the pressure profiles are greatly affected by the magnetic field
interaction.

Nomenclature Similarity exponent in non uniform medium has been determined by

a = sound speed many researchers including Whitham [13], Hafner [14], Radha and
H = magnetic field strength Sharma [15], and Toque [16]. Gundersen [17] discussed the problem
h = magnetic pressure of cylindrical and spherical MHD shock waves and derived a relation
p = pressure of the gas between strengths of converging cylindrical and spherical magneto-
T = temperature hydrodynamic shocks near the point of collapse and the distance
t = time from the point of collapse. Chisnell [1] provided an analytical
u = velocity description of converging shock waves by replacing the previous
x = spatial coordinate approach for numerical solutions of the ordinary differential equa-

= specific heat ratio tions with a theoretical study of the singular points of the differential
 = magnetic permeability equations. This theoretical approach leads to an approximate deter-
 = density of the gas mination of the similarity exponent, which is in close agreement with
R = gas constant previously obtained values. These extremely good, though approxi-
mate, values of the similarity exponent led to simple analytical
description of the flowfield behind the converging shock waves.
In the present investigation, we determine the similarity exponent
I. Introduction
and an analytical description of the flowfield behind the converging

I N THE present paper we analyze the self-similar motion of

converging cylindrical shock wave in ideal plasma of varying
density. The plasma is assumed to be an ideal gas with an infinite
shock wave in magnetogasdynamics. The study of singular points of
the system of differential equations leads to an approximate deter-
mination of the similarity exponent and an analytical description of
electrical conductivity and permeated by an axial magnetic field flowfield behind the shock wave. The influence of a change in the
orthogonal to the trajectories of gas particles. The medium is initially value of the adiabatic heat exponent
and a change in the magnetic
uniform and at rest. In the final stages of the collapse the shock field strength on the flow pattern for various cases are assessed. The
becomes strong and the pressure ahead of the shock is neglected in general behavior of the velocity and density distribution remains
comparison with the pressure behind the shock, leading to similarity unaffected but the pressure profiles are greatly affected by the
formulation for the problem. The similarity variable, which is the magnetic field interaction.
ratio of distance to a particular power of the time, is not known
a priori. Chisnell [1] has determined this particular power, known as
similarity exponent, from the solution of a single ordinary differential II. Basic Equations
equation. Many authors, e.g., Guderley [2], Butler [3], Sedov [4], Assuming the electrical conductivity to be infinite and the
Stanyukovich [5], Zeldovich and Raizer [6], Welsh [7], Lazarus [8], direction of the magnetic field to be orthogonal to the trajectories of
Hirschler and Gretler [9], Taylor and Cargill [10], Lock and the gas particles, the basic equations for a one-dimensional cylin-
Mestel [11], and Gurovich et al. [12] have described the methods for drically symmetric motion in magnetogasdynamics can be written in
determination of the similarity exponent in uniform medium. the form [18]
Received 24 September 2009; revision received 3 August 2010; accepted t
ux
ux
ux1  0 (1)
for publication 6 August 2010. Copyright © 2010 by the American Institute of
Aeronautics and Astronautics, Inc. All rights reserved. Copies of this paper
may be made for personal or internal use, on condition that the copier pay the ut
uux
1 px
hx   0 (2)
$10.00 per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood
Drive, Danvers, MA 01923; include the code 0001-1452/10 and $10.00 in
correspondence with the CCC.
∗
Department of Applied Mathematics, Institute of Technology; lpsingh. pt
upx  a2 t
ux   0 (3)
apm@itbhu.ac.in (Corresponding Author).
†
Department of Applied Mathematics, Institute of Technology,
‡
Department of Mathematics. ht
uhx
2hux
ux1   0 (4)
2523
2524 SINGH, SINGH, AND PANDEY

where  is the density, p the pressure, u the particle velocity;
1 G0
2
3
4

a  
p=1=2 is the sound speed with
as the constant specific heat V 0  ;  ; Z0  ; B0  (16)

G



ratio; h  H2 =2 is the magnetic pressure with, H and  being the
magnetic field strength and the magnetic permeability, respectively; t where
is the time and x the single spatial coordinate. The letter subscripts
denote partial differentiation unless stated otherwise. The system of
 Z  B
V  2 (17)
Eqs. (1–4) is supplemented with an equation of state p  RT, in
is the determinant of the left-hand side coefficients and
i , i  1, 2,
which R is the gas constant and T is the temperature.
3, 4 are the determinants obtained by replacing the ith column of

by the right-hand side of Eqs. (12–15). The four determinants
i are
III. Similarity Transformations given by
We introduce the nondimensional variables G, V, Z, and B so


21   2
that the flow variables are written with these new variables in the
1  
2V     VQV
B1    1
following form


(18)
x x2 x2
  0 G; u  V; a2  Z; b2  B (5)
t t2 t2

2
1   B 2
1=2
2   QV
1    1 (19)
where b  2h= is the Alfvén speed.
  V   V

The motion of the shock takes place for t < 0, arriving at the origin
O at time t  0. The Eqs. (1–4) in terms of new nondimensional


variables G, V, Z, and B reduces to the following form Z 1

3  2
  V


 1  VQV
tGt
xVGx
xGVx  2VG (6) V



2
 
 1B1    1 (20)
xZx xZGx xBx xBGx 2Z

tVt
xVVx



 V  V2  B


G 2 2G

(7) 


B 1

4  2
1  V 
  VQV
V

tZt xVZx 
 1


 tGt
xVGx   2  2V (8) 2
Z Z G  B1    1 (21)


where
tBt tGt xVBx xVGx




2xVx  2  6V (9)

B G B G 21   21  
QV  V 2
1  2  V
(22)
Self-similar solutions of Eqs. (6–9) are sought in terms of the variable


  x=Rt, where Rt is the distance of the shock from the origin at
time t (<0) and G, V, Z, and B are functions of  alone. Changing the In Eq. (16)  occurs as the coefficients of its derivatives, and G does
independent variables from x; t to ; t using not occur in the determinants, the system of equations could be
1 @ @ @ R_ @ decoupled to provide a pair of ordinary differential equations
@
 ;   (10)
@x R @ @t @t R @ dZ
3
 (23)
dV
1
where R_  dRdt
.
It is noted from Eqs. (6–9) that the t- variable enters the equations
dB
4
_ arising from the t@=@t term. Thus, the
only in the combination tR=R  (24)
dV
1
condition for the self-similar solution to exist is
_  with two supplementary equations
tR=R or R  At (11)
1 dG
2
where  and A are constants. The Eqs. (6–9) in terms of V, G,  (25)
Z, and B reduces to G dV
1

G0 1 d

V 0
V    2V (12)  (26)
G  dV
1


Z B G0 Z0 B0 2Z
V  V 0



  V  V2  B The system Eqs. (23–26) can be solved after determining ZV

2 G
2
and BV from Eqs. (23–26), respectively. To describe the flow
(13) behind the shock, a solution of the Eq. (16) is required which, at the
shock   1, satisfies the Rankine–Hugoniot conditions for a strong
G0 2Z1  V shock

 1Z  Z0  (14)
G V

1 2 _
 ; u R

1 0

1
G0

2BV 0
V  B
V  B0  2  6VB (15) 2

 1 _ 2

1 2 _2
G a2  R; b2  R (27)


12
1

The system of Eqs. (12–15) can be solved for the derivatives V 0 , Using Eq. (11) in Eq. (5), the flow variable may be written as in the
G , Z0 , and B0 in the following form:
0
terms of , t as following
 SINGH, SINGH, AND PANDEY 2525

_ _2  1
  0 G; u  R V; 0R
p  
 2
2  GZ;
_2 _2 _2
(28)
0R
h  
 2
2  GB; a2  R2 2 Z; b2  R2 2 B
0.8
The transformed boundary conditions at the shock (  1) become Z+B

1 2
Gs  ; Vs  0.6

1

1


2

 1 2

1 2 2 ∆ := 0
Zs   ; B s   (29)


12
1 0.4

behind the shock, where x is large, Eq. (5) shows that

V1  0; Z1  0; B1  0 (30) 0.2

It may be noted that
 0 is a parabola in the V; Z plane

touching the V axis at V  . Equation (28) shows that
has a 0
negative value at the shock   1 and a positive value at   1. 0 0.2 0.4 0.6 0.8 1
Thus, the solution curve has to cross the parabola to reach the origin V V0 Vs α
represent the condition behind the shock. The solution curve and the Fig. 1 The solution curve (Z
B) and the parabola
 0 for
 1:4.
parabola
 0 are shown in Fig. 1 for cylindrical shock with The solution curve has a singular point at V0 , the shock at Vs , and the

 1:4. Equation (16) show that at a point on the parabola
 0, origin corresponds to conditions far behind the shock.
the flow variables will have infinite slopes unless
i simultaneously
vanishes. To get the nonsingular solution of Eq. (16), we choose the

exponent  in such way that
vanishes at those points where the 21  
Q  V  V0  V  (35)
determinant
i vanishes. Here we use the numerical determination
V0
method for the similarity exponent given by Chisnell [1] and
formulated by Zeldovich and Raizer [6].
Using above values of
and Q, Eqs. (31) and (32) reduces in the
following form
IV. Determination of the Similarity Exponent
1 dZ N1 N2 N3
In this section the behavior of the differential equations that 

(36)
determines the similarity exponent is examined at two singular Z dV V V V
q
points. Using Eqs. (18–21), the Eqs. (23) and (24) may be written as where

1 2
1 dZ 2
  V



 1  VQV  
 1B1  
 1
 (31)
Z dV 2
V  1


  V
  V2 QV    VB1  
2  1

1 dB 2
1  V  1


  VQV  B1  
2  1
 (32)
B dV 2
V  1


  V
  V2 QV    VB1  
2  1

   
By taking note of the correct behavior of Z and B at the singular 1 2 2 22
points, two trial functions ZT and BT are constructed which, when q 
1   1  k  1  k

1  2

V0 V0
substituted into
in right-hand side of Eqs. (31) and (32) leads to
extremely accurate values of the similarity exponent  and a simple 2
N1  
analytic description of the flowfield. The integration of Eqs. (31) and qV0 1  2
(32) has to be performed without knowing a priori, the value of V at  
1 2
which the solution crosses the parabola
 0. If this value is N2  K  1  2

 

q1  2 V0
V  V0 , then QV0   0 and ZT V0 
BT V0  must equal  
  V0 2 . We introduce the trial function developed by Chisnell [1] 1 22
N3  1  1  2


K


q1  2 qV0


2
ZT  kV 2 and BT  1  kV 2 (33) K

3

1  k
 3  
V0


21   

1  k 
where   V0  12 and 0 < k < 1 has required value at V0 is finite at
V0
V  . 1 dB N1 K1 K2
Inserting the trial functions into
in the right-hand side of 

(37)
B dV V V V
q
Eqs. (31) and (32) enables the singular behavior at V  V0 be
removed as both
and Q will have the factor V  V0. On using
Eq. (33), expressions for
and Q reduces to where

  
2 1 
2

 V  V0  V1    (34) K1  K  3  2 
V0 
q1  2 V0
2526 SINGH, SINGH, AND PANDEY

 
1 22 It is evident from Eqs. (44) and (45) that as
increases, the value of
K2  K 
q3  2  =V0 increases and the value of  decreases.

q1  2 qV0



2 21    V. Flowfield
K  2
 1 
1
V0
V0 In the previous section, the similarity exponent  was determined

  for various values of
and k by solving differential equations for V, Z
1 k

 k1    1
 1   and V, B, which passed through the singular point V0 , Z0 , B0 , and Vs ,



Zs , Bs . Now the supplementary differential equations for G and  in
To determine  by a simple iteration scheme a solution of the Eqs. (25) and (26) are integrated. These equations also have a
system of simultaneous Eqs. (36) and (37) is required and it must pass singular point at V0 . The trial function ZT
BT given in Eq. (33) is
through points V0 ; Z0 , V0 ; B0  and the shock points Vs ; Zs , used to remove the factor (V  V0 ) from the numerators and denomi-
Vs ; Bs . Integration from V0 to Vs gives nators of the right-hand side of Eqs. (25) and (26). Appropriate

N

 expansions of
and Q are given in Eqs. (34) and (35), respectively,
Zs Vs 1   V0 N 2 Vs
q N 3 and the corresponding expansions for
1 and
2 given by Eqs. (18)
 (38) and (19) are
Z0 V0   Vs V0
q

N


1  V  V0 V1  V
q (46)
Bs Vs 1   V0 K1 Vs
q K2
 (39) 

B0 V0   Vs V0
q 2

2  V  V0  V1    1  
as the condition to be satisfied by the parameters  and V0 . Substi- V0


tuting Vs , Zs and Bs from Eq. (29) in Eqs. (38) and (39), Z0  21  
k  V0 2 and B0  1  k  V0 2 provides    Vx V 

V0

2


2




1 N1 V0 N1 2  2
1 
V1  k  1    1 (47)
V0 2  V0


N
 
2

 1
1 2 V0
q N 3 Substituting the values of
, Q,
1 and
2 in Eqs. (25) and (26)



12 

11  V0  2


1

q yields

N
N 2 
2


1 1 V0 1

1 1 dG L1 L2


(48)
2 
1 G dV V   V
q

K
 

1 1 V0
q K 2
 (40) 1 d N1 L3


11  V0  2


1

q 
(49)
 dV 2V V
q
The relation between  and V0 is given by the condition QV0   0 where
and Q is given by Eq. (22). 
q1 q  q1
By rearranging the terms in Eq. (22), V0 may be written in terms of L1  ; L2 
 as 
q1  2 
q1  2
N1 1
 1 2 L3 


(41) 2 1  2
1   1  V0 =
V0 = 
 

2 
This explicit determination of =1   in terms of V0 = can be q1  1
1    1 1  k  1  2 1



V0
used as the iteration parameter in the solution of Eq. (40). For a given
V0 =, having determined =1  , the parameter  and V0 follow Integration of Eqs. (36), (37), (48), and (49) from the shock   1
and may be inserted in the right side of Eq. (40). Substituting N1  2, provides:
N2  0, N3  0, K1  0, and K2  0 in Eq. (40), we achieve a good
N


initial approximation of =V0 . The results for the similarity exponent Z V 1   Vs N2 V
q N3
 (50)
 for various values of
and k are given in Table 1. Zs Vs V Vs
q
It can be seen from fore going analysis that as
increases from
unity, the value of =1   and  decreases and the two zeros of
N


B V 1   Vs K 1 V
q K 2
quadratic expression QV move closer to each other. Writing  (51)
Eq. (22) in the form Bs Vs V Vs
q



2
QV V 1 1 2


  1
 1 G   V L1 V
q L2
2  2 
 (52)



2  Gs   Vs Vs
q
1 2 8
 2 
1  (42)
N1

4

Vs 2 V
q L3
 (53)
where V Vs
q
  =1   (43) with Vs , Zs , Bs , and Gs given in Eq. (29). Using Eqs. (28) and (29) the
nondimensional form of flow variables are written as
It may be noted that the roots of QV will coincide if

1N1 

u V V 2 V
q L3
  f1
2=
1=2 g2 (44)   (54)
us Vs Vs Vs
q
and this particular location of singular point V0 may be written as



  G   V L1 V
q L2
 1

=21=2 (45)   (55)
V0 s Gs   Vs Vs
q
 SINGH, SINGH, AND PANDEY 2527

1N


p G Z V 1   V L1 N2 V
q N2
L2
2L3 Smaller values of k signify the higher magnetic field effect. The value
 2   corresponding to different values of the specific heat ratio
and k
pS Gs Zs Vs   Vs Vs
q
are given in Table 1. The fact that  is always less than one shows that
(56) the shock wave is continuously accelerated. The computed results
indicate that an increase in the specific heat ratio
or k causes the

1N


h G B V 1   V L1 K1 V
q K2
L2
2L3 similarity exponent  to decrease. In fact, the presence of the
 2  magnetic field ahead of the shock has an important effect on the
hs Gs Bs Vs   Vs Vs
q
flowfield behind the shock.
(57) Numerical computation of Eqs. (54–57) for 1 <   1 has been
performed in terms of a new variable y  1=, and the range of
VI. Results and Discussion computation is 0 < y < 1. In Figs. 2–5, : kWe find that the density
The inverse density, the gas velocity, the pressure, and the increases monotonically behind the shock and has a nonzero limiting
magnetic pressure normalized with respect to its value at the shock value far behind the shock. The inverse density variable, s = as a
are computed using Eqs. (54–57) for different values of k and specific function of 1=, is displayed Fig. 2 for cylindrical shocks with
heat ratio
. The flow and field profiles are presented in Figs. 2–5. It various values of
and k. Also, an increase in
and magnetic field
may be noted here that the magnetic pressure term enters into the strength causes s = to increase. The monotonic decreasing variable
solution through trial functions defined in Eq. (33) with 0 < k  1. u=us is displayed as a function of 1=, using Eq. (54) in Fig. 3 for
For k  1, the analysis reduces to the case described by Chisnell [1]. cylindrical shocks for various value of
and k. The effect of

1.0
1.0

0.8
γ=1.4
0.8

γ=1.4
1.67
7
0.6 0.6
2.0
ρs /ρ

p/ps

8 1.67
5
1
0.4 0.4
4 9 2
6 3 8
3
4 2.0
0.2 2 0.2 6
9

1 7
5

0.0 0.0
1.0 0.8 0.6 0.4 0.2 0.0 1.0 0.8 0.6 0.4 0.2 0.0
1/ξ 1/ξ
Fig. 2 The inverse density, normalized with respect to its values at Fig. 4 The pressure, normalized with respect to its values at shock, is
shock, is shown as function 1= for various values
and k. The numbers shown as function 1= for various values
and k. The numbers on the
on the lines correspond to the following different values of k: 1) 0.4, 2) 0.5, lines correspond to the following different values of k: 1) 0.4, 2) 0.5, 3) 0.6,
3) 0.6, 4) 0.4, 5) 0.5, 6) 0.6, 7) 0.4, 8) 0.5, and 9) 0.6. 4) 0.4, 5) 0.5, 6) 0.6, 7) 0.4, 8) 0.5, and 9) 0.6.

1.0
1.0

0.8
1 0.8
7 4 3
2
6
9 5
0.6 γ=1.4
0.6
8
u/us

h/hs

1.67 γ=1.4

0.4 0.4
2.0 1.67
2 8
5
2.0
3
0.2 0.2 1
6 4 7
9

0.0 0.0
1.0 0.8 0.6 0.4 0.2 0.0 1.0 0.8 0.6 0.4 0.2 0.0
1/ξ 1/ξ
Fig. 3 The fluid velocity, normalized with respect to its values at shock, Fig. 5 The magnetic pressure, normalized with respect to its values at
is shown as function 1= for various values
and k. The numbers on the shock, is shown as function 1= for various values
and k. The numbers
lines correspond to the following different values of k: 1) 0.4, 2) 0.5, 3) 0.6, on the lines correspond to the following different values of k: 1) 0.4, 2) 0.5,
4) 0.4, 5) 0.5, 6) 0.6, 7) 0.4, 8) 0.5, and 9) 0.6. 3) 0.6, 4) 0.4, 5) 0.5, 6) 0.6, 7) 0.4, 8) 0.5, and 9) 0.6.
2528 SINGH, SINGH, AND PANDEY

Table 1 The values of  and =V0 obtained Acknowledgment

from Eqs. (40) and (41) for various values
of
and k Financial support from Institute of Technology, Banaras Hindu
University in Varanasi, India is gratefully acknowledged.

k  =V0
1.3 0.1 .841327 1.502220 References
0.4 .841304 1.502597
[1] Chisnell, R. F., “An Analytic Description of Converging Shock Waves,”
0.5 .841281 1.503468
Journal of Fluid Mechanics, Vol. 354, Jan. 1998, pp. 357–375.
0.6 .841251 1.503468
doi:10.1017/S0022112097007775
0.9 .841105 1.505884
[2] Guderley, G., “Starke Kugelige und Zylindrische, Verdichtungsstosse
1.4 0.1 .830089 1.660501
in Der Nahe Des Kugelmitteelpunktes bzw der Zylinderachse,”
0.4 .830132 1.659623
Luftfahrtforschung, Vol. 19, 1942, pp. 302–312.
0.5 .830132 1.65962
[3] Butler, D., “Converging Spherical and Cylindrical Shocks,” Ministry of
0.6 .830128 1.659776
Supply, Armament Research Establishment, United Kingdom, Rept. 54/
0.9 .830089 1.661314
54, 1954.
1.67 0.1 .814361 1.894759
[4] Sedov, L. I., Similarity and Dimensional Methods in Mechanics,
0.4 .814358 1.894759
Academic Press, Newyork, 1959.
0.5 .814357 1.897624
[5] Stanyukovich, K. P., Unsteady Motion of Continuous Media, Pergamon
0.6 .814356 1.897624
Press, New York, 1960.
0.9 .814352 1.900930
[6] Zeldovich, Y. B., and Raizer, Y. P., Physics of Shock Waves and High
2.0 0.1 .800111 2.054208
Temperature Hydrodynamic Phenomenon, Vol. 2, Academic Press,
0.4 .800128 2.058177
New York, 1967.
0.5 .800135 2.059650
[7] Welsh, R. L., “Imploding Shocks and Detonations,” Journal of Fluid
0.6 .800141 2.061204
Mechanics, Vol. 29, No. 1, 1967, pp. 61–79.
0.9 .800165 2.066400
doi:10.1017/S0022112067000631
[8] Lazarus, R. B., “Self-Similar Solutions for Converging Shocks and
Collapsing Cavities,” SIAM Journal on Numerical Analysis, Vol. 18,
increasing values of
and magnetic field strength (lower values of k) No. 2, 1981, pp. 316–371.
doi:10.1137/0718022
causes u=us to decrease. The pressure distribution p=ps is displayed [9] Hirschler, T., and Gretler, W., “On the Eigenvalue Problem of
using Eq. (56) in Fig. 4 for cylindrical shocks for various values of
Imploding Shock Waves,” Zeitschrift fur Angewandte Mathematik und
and k. The behavior of the gas pressure in the magnetogasdynamics Physik, Vol. 52, No. 1, 2001, pp. 151–166.
case is in contrast to the corresponding nonmagnetic case. The doi:10.1007/PL00001537
monotonic decreasing behavior of pressure is displayed as a function [10] Taylor, M. G. G. T., and Cargill, P. J., “A General Theory of Self-similar
of 1= for different values of
and k in Fig. 4. It is clear from the Expansion Waves in Magnetohydrodynamic flows,” Journal of Plasma
figure that an increase in magnetic pressure causes to slow down the Physics, Vol. 66, No. 4, 2001, pp. 239–257.
process of steepening compared with lower magnetic pressure levels doi:10.1017/S0022377801001398
and nonmagnetic case (Chisnell [1]) for all values of
. This shows [11] Lock, R. M., and Mestel, A. J., “Annular Self-similar Solutions in Ideal
Magnetogasdynamics,” Journal of Plasma Physics, Vol. 74, No. 4,
that an increase of the magnetic field strength has just the reverse 2008, pp. 531–554.
effect in the sense that it effectively resists the compression. Also an doi:10.1017/S0022377808007101
increase in the value of
causes a falling off of the pressure levels. [12] Gurovich, V., Grinenko, A., and Krasik, Y., “Semianalytical Solution of
The magnetic pressure profiles as a function of 1= are displayed the Problem of Converging Shock Waves,” Physical Review Letters,
using Eq. (57) in Fig. 5 for various values of
and magnetic field Vol. 99, 2007, pp. 124503(1)–124503(4).
strength. The effect of increasing values of
and magnetic field [13] Whitham, G. B., Linear and Nonlinear Waves, Interscience Publishers,
strength is to increase the magnetic pressure levels in the flowfield. New York, 1974.
[14] Hafner, P., “Strong Convergent Shock Waves Near the Centre of
Convergence: A Power Series Solution,” SIAM Journal on Applied
VII. Conclusions Mathematics, Vol. 48, No. 6, 1988, pp. 1244–1261.
doi:10.1137/0148076
The self-similar motion of converging cylindrical shock waves in
[15] Radha, C., and Sharma, V. D., “Imploding Cylindrical Shock in a
ideal plasma with varying density has been analyzed. The plasma is Perfectly Conducting and Radiating Gas,” Physics of Fluids B, Vol. 5,
considered to have infinite electrical conductivity and to be No. 12, 1993, pp. 4287–4294.
permeated by an axial magnetic field orthogonal to the trajectories of doi:10.1063/1.860596
gas particles. Using the method developed by Chisnell [1] the [16] Toque, N., “Self Similar Implosion of a Continuous Stratified Medium,”
similarity exponent has been determined and an analytical descrip- Shock Waves, Vol. 11, No. 3, 2001, pp. 157–165.
tion of the flowfield behind the shock has been presented. It has been doi:10.1007/PL00004074
observed that as the magnetic field strength decreases the value of the [17] Gundersen, Roy M., “Cylindrical and Spherical Shock Waves in
similarity exponent decreases. Also, the density distribution, velocity Monatomic Conducting Fluids,” Applied Scientific Research, Section B
distribution, pressure distribution, and magnetic pressure profiles are Vol. 10, No. 2, 1963, pp. 119–128.
[18] Korobeinikov, V. P., Problem in the Theory of Point Explosion in Gases,
presented for various values of the specific heat ratio
and k. It has American Mathematical Society, Providence, RI, 1970.
been observed that the density and velocity distribution profiles have
similar behavior as in nonmagnetic case. However, the pressure X. Zhong
profiles are greatly affected by the magnetic field. Associate Editor