Professional Documents
Culture Documents
Nath 2017
Nath 2017
PII: S0094-5765(17)30987-6
DOI: 10.1016/j.actaastro.2017.08.010
Reference: AA 6429
Please cite this article as: T. Nath, R.K. Gupta, L.P. Singh, Resonantly interacting non-linear waves in a
van der waals gas, Acta Astronautica (2017), doi: 10.1016/j.actaastro.2017.08.010.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Abstract
PT
The present paper uses the method of multiple time scales to derive the asymptotic solution of
system of one-dimensional quasilinear hyperbolic equations for the generalized geometry in van
der Waals gas. The transport equation for the amplitude of resonantly interacting high frequency
RI
waves propagating into non-ideal gas is derived. Further, we discuss the cases when the initial
data for the wave amplitude is of 2π periodicity. The evolutionary behavior of non-resonant
wave modes culminating into shock wave and its location are examined in van der Waals fluid.
SC
Keywords: Shock wave; van der Waals gas; Asymptotic solution; Wave interaction.
1. Introduction
U
A large number of physical phenomena such as nuclear explosions, chemical explosion, bomb
AN
blast, collision of two or more galaxies, supersonic flow etc.; happening in nature may be
described in terms of mathematical model described by quasilinear hyperbolic partial differential
equations [1-3]. In nonlinear science and engineering, the study of shock waves, acceleration
M
waves, weakly non-linear waves, interaction of shock waves etc. has been of great interest for
researchers since long time, due to its wide application in nuclear physics, plasma physics,
astrophysical sciences and interstellar gas masses. Asymptotic method for the solution of weakly
non-linear hyperbolic waves has received considerable attention of researchers in the last few
D
decades. A significant contribution on small amplitudes nonlinear waves has been made by
Choquet-Bruhat [4] in which they have considered a shockless solution of system of hyperbolic
TE
equations that depends only on single phase function. Earlier Gemain [5] has studied the single
phase progressive wave solution for the weekly nonlinear waves. Latter, some authors such as
Fusco [6], Fusco and Engenbrecht [7], Sharma et al. [8-9] and Radha et al. [10] have used the
EP
progressive wave approach to analyze the wave propagation problem in various gasdynamic
regimes. In this reference, the contribution of various authors like Becker [11], Ockendon and
Spence [12], Chu [13], Modie et al.[14] and He and Moodie [15] are worth mentioning. Hunter
and Keller [16] proposed a general non-resonant multi wave mode theory based on the weakly
C
non-linear geometrical optics, which has led to several important generalizations by Majda and
Rosales [17] and then Hunter et al. [18] in which they have derived the uniformly valid
AC
asymptotic theory of resonantly interacting high frequency waves for non-linear hyperbolic
system of equations.
If the temperature of the gas is very high and density is too low then the ideal gas model is no
longer valid therefore in this situation the alternative of the ideal gas model is a modified van der
Waals gas model. The study of shock related phenomena in van der Waals gas is more complex
as compared to general ideal gas model. In recent years the study of shock related phenomena in
1
Corresponding Author’s mobile no: +91-9651711507
Email: trilokinath.rs.apm12@itbhu.ac.in
ACCEPTED MANUSCRIPT
van der Waals gas have received great attention of scientist and engineers due its application in
space physical science such as chemical processes, nuclear reaction, aerospace engineering and
sciences etc. [19-23]. Zhao et al. [24] has studied a complete classification of shock waves and
shock splitting phenomena together with their admissibility in van der Waals fluid. Further, the
theory of progressive wave is used to study the finite and moderately small amplitude
disturbances in van der Waals gas see [25]. Singh et al. [26] have used the progressive wave
approach to analyze the main feature of weakly non-linear waves in non-ideal gas. Further Nath
PT
et al. [27] used the same technique to analyze the feature of weakly non-linear waves in van der
Waal gas. Sharma et al. [28] have studied the wave interaction in a non-equilibrium gas flow by
using the method of multiple time scale. Using the same technique Arora et al. [29, 30] have
RI
obtained the asymptotic solution of system of hyperbolic equations in various material media.
Further the method of weakly nonlinear geometrical optics is used to study the evolution of fast
magnetosonic waves of high frequency propagating into an axisymmetric equilibrium plasma
SC
and the geometry of fast magnetosonic rays [31, 32].
The aim of the present paper is to study the propagation of weekly nonlinear small amplitude
high frequency asymptotic waves for one dimensional unsteady, compressible flow in a non-
ideal gas followed by van der Waals equation of state with generalized geometry. The resonantly
U
interacting multi wave theory is used to derive the evolution equation governing the growth and
decay behavior of wave amplitude for the nonplanar waves. Also the interaction coefficients are
AN
determined which measure the strength of coupling between different waves modes. The small
amplitude high frequency asymptotic solution of one dimensional hyperbolic system of
equations for the planar and nonplanar flow in van der Waals gas is obtained. Further, the
existence of shock and its path (parametrically) in non-ideal gas followed by van der Waals
M
2. Governing Equations
D
Let us consider a general class of real gases whose equation of state is given as
TE
a
p + 2 (V − b ) = RT , (1)
V
where p is the pressure, V is the volume, R is the universal gas constant, T is the absolute
EP
temperature. Here a represents the amount of intermolecular force of attraction between the gas
particles and b represents the excluded volume of the gas. It is well known that the gases behave
like real gas at low temperature and high pressure. For the given equation of state (1), the
expression of internal energy, with the help of R = ( γ − 1) Cv where Cv is the specific heat at
C
e=
( )
p + a ρ 2 (1 − b ρ ) − a ρ 2 ( γ − 1)
, (2)
(γ − 1) ρ
where ρ is the density of the gas, γ is the adiabatic exponent and is defined as γ = C p Cv where
C p is the specific heat at constant pressure. Here, if we take a = 0 and b = 0 in equation (1) then
the equation of state of van der Waals gas turns to the equation of state of ideal gas.
Thus the governing equations for the one-dimensional compressible, inviscid, unsteady, planar
and non-planar flow in van der Waals gas may be written as [1-2, 33]
ACCEPTED MANUSCRIPT
∂ρ ∂u ∂ρ n ρ u
+ρ +u + = 0, (3a)
∂t ∂x ∂x x
∂u ∂u ∂p
+u + ρ −1 = 0, (3b)
∂t ∂x ∂x
∂p ∂p ∂u nu
+u + ρ A2 + = 0. (3c)
∂t ∂x ∂x x
PT
where u is the gas velocity, t is the time, x is the spatial coordinate. Here ( n = 0 ) represents
the planar flow, ( n = 1) represents the cylindrically symmetric flow and ( n = 2 ) represents the
RI
spherically symmetric flow. Also A represents the sound velocity and is defined as
γ p + a ρ 2 ( γ − 2 + 2b ρ )
1/ 2
A= . (4)
ρ (1 − b ρ )
SC
Now equation (3) can be written in matrix form as
∂U ∂U
+M + N = 0, (5)
∂t ∂x
U
where U = ( ρ , u , p ) , N = ( n ρ u x , 0, ρ A2 nu x )
tr tr
and M is the coefficient matrix of order
AN
3 × 3 having components M ij and the non zero components of M are as
M 11 = M 22 = M 33 = u ,
M 12 = ρ , M 23 = ρ −1 and M 32 = ρ A2 . (6)
M
The Eigen values of the system (5) are given as λ1 = u + A , λ2 = u , λ3 = u − A , therefore the
system of partial differential equation (5) is strictly hyperbolic in nature and admits 3
characteristics curves, among them two represent the waves propagating in the + x direction
D
and − x with the shock speed u + A and u − A respectively. While other represents the particle
path propagating with the shock speed u . Here we are considering only those waves which are
TE
In this section the method of multiple time scale will be used to derive the asymptotic solution in
the form of small amplitude high frequency waves for the system of equations (5), when the
AC
attenuation time scale ( τ at ) is large as compared to the characteristic time scale ( τ ch ), that is
ε = τ ch τ at << 1 .
The left and right eigenvectors of M 0 corresponding to the eigenvalues λ1′ = A0 , λ2′ = 0 , λ3′ = − A0
are denoted by L( i ) and R ( i ) ( i = 1, 2,3 ) respectively. These eigenvectors satisfy the normalization
conditions L(i ) R ( j ) = δ ij ( 1 ≤ i ≤ 3, 1 ≤ j ≤ 3 ), where δ ij is Krӧnecker delta. Therefore in view of
above conditions the left and right eigenvectors may be written as
( )
L(1) = 0, ρ0 2 A0 , 1 2 A02 , R (1) = (1, A0 ρ 0 , A0 2 ) ,
tr
ACCEPTED MANUSCRIPT
( ) ( ),
L( ) = − A0 2 , 0, 1 , R ( ) = − 1 A0 2 , 0, 0
2 2 tr
(7)
L( ) = ( 0, − ρ 2 A , 1 2 A ) , R( ) = (1, − A
3
0 0 0
2 3
0 ρ0 , A02 ) .
We now seek the asymptotic solution of (5) in the following form with ε → 0
( ) ( )
U ( x, t ) = U + ε U x, t , ξ% + ε 2U x, t , ξ% + O ( ε 3 ) ,
0 1 2 (8)
PT
where U 0 is the constant solution of system of equations (5), U1 is a smooth bounded function
of its arguments while U 2 is a bounded function in ( x, t )
coordinate within some bounded
RI
region of interest having at most sub-linear growth in ξ% as ξ% → ±∞ [ 28]. Here ξ% = (ξ1 , ξ 2 , ξ3 )
denotes the “fast variable” and is characterized as ξ% = ψ% ε , where ψ% = (ψ 1 ,ψ 2 ,ψ 3 ) indeed
ψ i (1 ≤ i ≤ 3) is the phase function of i th wave associated with the characteristic speed λi′ . Using
SC
the Taylor series expansion of M and N , in powers of ε about the constant state U 0 and
equation (8) in equation (5) and replace the partial derivatives ∂ ∂X ( X will be either x or t )
U
3
by ∂ ∂X + ε −1 ∑ ( ∂ψ i ∂X ) ∂ ∂ξi and equating the coefficients of ε 0 and ε 1 to zero, we have
AN
i =1
3
∂ψ i ∂ψ i ∂U1
∑ I ∂t + M 0 ∂x ξ = 0 ,
i =1 i
(9)
∂ψ i ∂ψ i ∂U 2 ∂U1 ∂U1 ∂ψ i ∂U
M
3 3
∑ I
∂t + M = − − M − (U .∇ N ) − ∑ (U1.∇M )0 1 , (10)
∂x ∂ξi ∂t ∂t i =1 ∂x ∂ξi
0 0 1
i =1
0
where I is an identity matrix of order 3 × 3 , ∇ is the gradient operator with respect to the flux
D
vector U . Since all the phase functions ψ i ( i = 1, 2, 3) satisfy the eikonal equation
∂ψ i ∂ψ i
TE
det I + M0 = 0. (11)
∂t ∂x
Here “det” represents the determinant. Now let us take a simplest phase function of the above
equation as
EP
ψ i ( x, t ) = x − λi′t , 1 ≤ i ≤ 3 . (12)
From equation (9) we conclude that for each phase function ψ i the terms ∂U1 ∂ξi are parallel to
the right eigenvectors R i of the matrix M 0 , therefore we have
C
3
U1 = ∑ µi ( x, t , ξ i ) R ( i ) , (13)
AC
i =1
( )
where µi = L(i ) .U1 is a scalar function known as wave amplitude that depends on the i th fast
variable ξi . The dependency of µi on ξi describes the wave form whether it is an oscillatory
wave or a pulse. Here we consider that µi ( x, t , ξi ) has zero mean value with respect to the fast
variables ξi , that is
P
1
µi ( x, t , ξ i ) d ξ i = 0 .
P →∞ 2 P ∫
lim (14)
−P
ACCEPTED MANUSCRIPT
Now we use equation (13) and solve the equation (10) for U 2 . To begin with we write
3
U 2 = ∑ n j R( j ) . (15)
j =1
Using equation (15) in equation (10) and then pre multiplying the resulting equation by L( i )
yields the following decoupled inhomogeneous system of first order partial differential equations
∂ni ∂µi ∂µi ∂U1
( i j ) ∂ξ
3 3
∑ (i )
( ) ∑ L( i ) (U1.∇M )0
PT
λ ′ − λ ′ = − − λ ′ − L U .∇ N − , 1≤ i ≤ 3. (16)
∂t ∂x ∂ξ j
i 1 0
j =1 j j =1
RI
The i th characteristics ordinary differential equation in equation (16) is given as
ξ& j = λi′ − λ ′j for i ≠ j , ξ&i = 0 , n&i = Ki , (17)
∂µi ∂µ 3
∂U1
where K i ( x, t , ξ1 , ξ 2 , ξ 3 ) = −
− λi′ i − L( i ) (U1.∇N )0 − ∑ L( i ) (U1.∇M )0
SC
.
∂t ∂x j =1 ∂ξ j
Now we will determine the asymptotically average of equation (16) along the characteristics and
then appeal to the sub linearity of U 2 in ξ , which makes sure that the expression (8) does not
U
contain any secular terms. Since ξi ’s are constant along the characteristics, thus along with this,
AN
the vanishing asymptotic mean value of m& i along the characteristics implies that the wave
amplitudes µi (1 ≤ i ≤ 3) , satisfy the following system of coupled integro-differential equations
[18]
M
∂µi ∂µ ∂µ
( )
P
µ j ξi + ( λi′ − λ ′j ) s µ% k (ξi + ( λi′ − λk′ ) s ) ds = 0 ,
1
+ λi′ i + α i µi + Π iii µi i + ∑ Π ijk lim
∂t ∂x ∂ξ i i ≠ j ≠ k P →∞ 2 P ∫
−P
D
(18)
where µ% k = ∂µk ∂ξ k and the coefficients αi and Π jk are given as
i
( ) ( )
TE
Here it may be observed that for planar flow ( n = 0 ) , the coefficients αi becomes zero as the
source terms in the system of equations becomes zero. Further, in absence of van der Waals
EP
parameters a, b the equation (18) reduces to the similar equation as derived in [17]. Thus the
coefficients α i in equation (18) describe the growth or decay behavior of the wave amplitude µi
C
due to non-planar wave from the non-idealness of the medium. Further the interaction
coefficients Π ijk , which are asymmetric in j and k , measure the strength of coupling between
AC
the j th and k th wave modes ( j ≠ k ) that can generate a i th wave ( i ≠ j ≠ k ) . Also the
coefficients Π iii refer to the non-linear self-interaction which are non-zero for genuinely
nonlinear waves and are zero for linearly degenerate waves. Further it is observed that, if all the
coupling coefficients Π ijk ( i ≠ j ≠ k ) are zero or the integral in (18) vanishes, then the wave will
not resonate and therefore the equation (18) reduces to the system of coupled Burger’s equation.
The coefficients αi , Π ijk and Π iii , as given in equation (19), gives the qualitative picture of the
nonlinear interaction process present in the system under consideration and can be determined
with the formulae as given in (19). Thus the coefficients are given by
ACCEPTED MANUSCRIPT
nA0 nA0
α1 = , α2 = 0 , α3 = − ,
2x 2x
2a ρ0 ( γ − 2 + 3b ρ0 ) + A02
Π123 = −Π 321 = = ω (Say),
2 A03 ρ0 (1 − b ρ0 )
2a ρ0 A0 ( γ − 2 + 3bρ0 ) + ( γ − 1 + 2bρ0 ) A03
Π13
2
= −Π 31
2
=− , (20)
ρ 0 (1 − bρ 0 )
PT
Π132 = Π12 3
= Π 22
2
= 0,
and
RI
2a ρ0 ( γ − 2 + 3bρ0 ) + ( γ + 1) A02
Π111 = −Π 33 3
= = Π (Say).
2 A0 ρ0 (1 − bρ0 )
SC
Further the resonant asymptotic equation (18) can now be written as after some simplification
∂µ1 ∂µ nA ∂µ ξ +ψ
P
1
+ A0 1 + 0 µ1 + Π µ1 1 + lim ∫ ζ x, t , 1 µ3 ( x, t ,ψ ) dψ = 0 , (21)
∂t ∂x 2x ∂ξ1 P →∞ 2 P − P 2
∂µ 2
U
= 0, (22)
∂t
AN
∂µ3 ∂µ nA ∂µ ξ +ψ
P
1
− A0 3 − 0 µ3 − Π µ3 3 − lim ∫ ζ x, t , 3 µ1 ( x, t ,ψ ) dψ = 0 , (23)
∂t ∂x 2x ∂ξ1 P →∞ 2P − P 2
where ζ is kernel and is defined as
M
ξ +ψ ω ∂µ2 ξ +ψ
ζ x, t , = x, t , . (24)
2 2 ∂ξ 2 2
The integral average term in equation (21) exhibits contribution to the wave amplitude µ1 due to
D
the nonlinear interactions of the wave field µ2 with the wave field µ3 . Similarly the integral
TE
average term in equation (23) exhibits contribution to the wave amplitude µ3 due to the
nonlinear interaction of the wave field µ2 with the wave field µ1 .
Let us suppose the initial value of µi at time t = 0 is µi0 ( x, ξi ) . Therefore from equation (22) we
EP
2 ∂ξ 2
If the initial data µi0 ( x, ξ ) are periodic function of 2π of the phase variable ξ , then the pair of
resonant asymptotic equations takes the following form
π
∂µ1 ∂µ nA ∂µ 1 ξ +ψ
+ A0 1 + 0 µ1 + Πµ1 1 + ∫ ζ x, t , µ3 ( x, t ,ψ ) dψ = 0 , (26)
∂t ∂x 2 x ∂ξ1 2π −π 2
π
∂µ3 ∂µ nA ∂µ 1 ξ +ψ
− A0 3 − 0 µ3 − Πµ3 3 − ∫π ζ x, t , µ1 ( x, t ,ψ ) dψ = 0 , (27)
∂t ∂x 2 x ∂ξ3 2π −
2
ACCEPTED MANUSCRIPT
The approximate asymptotic solution (8) of the system of hyperbolic equations (5) satisfying
small amplitude oscillating initial data
U ( x, 0 ) = U 0 + ε U10 ( x, x ε ) + O ( ε 2 ) ,
PT
(28)
will be non- resonant if the functions U10 ( x, x ε ) are smooth with a compact support [17].
Indeed the expansion (8), with U1 as given by equation (13) is uniformly valid to the leading
RI
order until shock waves have formed in the solution.
The characteristic equations are
dξ i Π µr dt τ i
SC
= , = , (29)
dx A0 dx A0
where τ i takes the value +1 and −1 according as i = 1 or i = 3 respectively.
Now in view of (29), the decoupled equations (26) and (27) may be written as
U
d µi nµ
=− i . (30)
AN
dx 2x
Equation (30) yield on integration
µi = µi0 ( si , χ i )( x si ) ,
−n/ 2
(31)
along the rays si = x − τ i A0t = constant, here the function µi0 is obtained from the initial condition
M
(28), and the “fast’ variable χ i parameterizes the set of characteristics curve (29)1 .
Therefore we obtain from equation (29)
D
χi = ξi − τ i Πµi0 ( si , χ i ) I i( n ) ( t ) , (32)
− n/ 2
TE
( n) τ At t
where I i (t ) = ∫ 1 + i 0 dt .
0
si
Thus the solution of system of equations (5) satisfying equation (28) in view of U10 ( x, x t )
EP
(
ρ ( x, t ) = ρ0 + ε x − n / 2 µ10 ( s1 , χ1 )( x − A0t )
n /2
+ µ30 ( s3 , χ 3 )( x + A0t )
n /2
) − Aε 2
0
µ 20 ( x, x ε ) + O ( ε 2 ) ,
C
(33a)
u ( x, t ) = ε x − n /2
A0
(µ ( s1 , χ1 )( x − A0t )
n/ 2
− µ30 ( s3 , χ 3 )( x + A0t ) ) + O (ε ) ,
n/ 2
AC
0 2
(33b)
ρ0 1
(
p ( x, t ) = p0 + ε x − n /2 A02 µ10 ( s1 , χ1 )( x − A0t )
n /2
+ µ ( s , χ )( x + A t ) ) + O ( ε ) ,
0
3 3 3 0
n /2 2
(33c)
where the “fast” variable χ i , as given in equation (32) are taken in such a way that at t = 0 ,
χ i = x ε , and the initial values for µi (1 ≤ i ≤ 3) are obtained from the above solution (33)
specified at t = 0 as
ρ 1
µ10 ( x, χ1 ) = 0 u10 ( x, χ1 ) − 2 p10 ( x, χ1 ) , (34a)
2 A0 2 A0
ACCEPTED MANUSCRIPT
PT
propagate by separating the portion of the continuous region. At shocks, the solution, which we
have found above, satisfies the R-H jump conditions for the shock location ξis ( t ) in the t − ξi
plane. Thus it can be shown that the shock location ξis ( t ) satisfies the following relation [16]
RI
dξis 1 i ( − )
dt 2
( )
= Π ii µi + µi( + ) , i = 1, 3 , (35)
SC
which is the shock speed in the t − ξ i plane. Here µi( − ) and µi( + ) denotes the value of the µi just
ahead and behind the shock respectively. In the undisturbed region the value of µi( − ) will be zero
for the shock front. In view of equation (31) and omitting the superscripts on ξ is and µi( + ) we get
U
− n /2
dξi Π x
= τ i µi0 ( si , χi ) . (36)
AN
dt 2 si
With the help of equation (36) and equation (32) we have the following relation between χ i and
t on the shock
M
χ
2τ i i 0
I i( m ) ( t ) = − µ ( t ) dt .
Π µ 0 2 ∫0 i
(37)
( )
D
Further in view of equation (32), equation (37) yields the following relation which determines
the path of shock wave parametrically.
TE
χ
2 i 0
ξi = χi − 0 ∫ µi ( t ) dt . (38)
µi 0
Note that if µ i0 ≠ 0 , then shock starts quickly right from the origin [17, 29].
EP
The method of multiple time scale is used to find the small amplitudes high frequency
asymptotic solution to the system of hyperbolic partial differential equations describing one-
AC
[17, 18]. The effect of van der Waals parameter enters into the solution trough the parameter A0 .
It is evident from equations (33) that the increasing values of b (keeping a fixed) causes the
density of the small amplitude high frequency waves to increase and also the increasing values b
have the same effect on the velocity and pressure of the small amplitude high frequency waves.
While increasing the value of a (keeping b fixed) causes to decrease the density of the small
amplitude high frequency waves. Similar effect of a are seen on the velocity and pressure of the
PT
small amplitude high frequency waves. From the above results it is found that the van der Waals
parameters play an important role on the solution of the hyperbolic system of partial differential
equations. In nonmagnetic case and without relaxation effect in ideal gas the results which we
have obtained here are in close agreement with the results obtained previously by some authors
RI
[27-29]. It has been observed here that the wave fields associated with the particle path do not
interact with each other. However the wave fields interact with an acoustic wave field to produce
resonant contribution toward the other acoustic field. The acoustic wave fields may or may not
SC
interact but in either case their total contribution towards entropy field must be zero. Thus the
system of hyperbolic equations reduces to a pair of resonant asymptotic equation for the acoustic
wave fields. For a non-resonant multi wave mode as presented in the paper of Hunter and Keller
U
[16], the reduced system of transport equations get decoupled with vanishing integral average
term and the occurrence of shock in the acoustic wave fields are discussed. Also the existence of
AN
shock and their location in the van der Waals gas is discussed. It is observed that in a contracting
piston motion having spherical symmetry, a shock is always formed before the formation of
focus no matter how small be the initial wave amplitudes, this is in contrast with the
corresponding cylindrical situation where a shock forms before the focus only if the initial
M
6. Conclusions
D
In the present article we use the method of multiple time scales to obtain the small amplitude
TE
high frequency asymptotic solution to the system of hyperbolic partial differential equations
describing one-dimensional unsteady, compressible planar and non-planar flow in a van der
Waals gas. Those conditions are discussed in which wave interaction occur resonantly. Transport
equations for the wave amplitude along the rays of the system of hyperbolic equations are
EP
derived; which constitute a system of inviscid Berger’s equation with quadratic non-linearity
coupled through a linear integral operator with known kernel. The growth and decay behavior of
wave amplitudes for the nonplanar waves are studied. It is observed that the increasing values of
C
van der Waals parameter b is to enhance the values of flow parameters while the increasing
values of van der Waals parameter a have reverse effect on flow parameters. Also the existence
AC
of shock and their location in the van der Waals gas is discussed.
Acknowledgement
First and second authors are thankful to the University Grant Commission (UGC), India for
providing senior research fellowship. Authors are also very thankful to the anonymous reviewers
for their valuable suggestion.
ACCEPTED MANUSCRIPT
References
[1] G.B. Whitham, Linear and non-linear waves, Wiley & Sons, New York, 1974.
[2] R. Courant, D. Hilbert, Methods of Mathematical Physics, vol. II. Interscience, N.Y, 1962.
[3] A.N. Anil, J.K. Hunter, P. Pantano, G. Russo, Ray method for non-linear waves in fluids and
plasmas, Longman, N.Y, 1993.
[4] V. Choquet-Bruhat, Ondes Asymptotic et Approaches pour des systems d’Equations aux
PT
Derivees Partielles non lineaires, J. Math Pures Appl 48 (1969) 119-158.
[5] P. Germain, Progressive waves, Jber DGLR (1992) pp. 11-30.
[6] D. Fusco, Some comments on wave motions described by non homogenous quasilinear first
RI
order hyperbolic system, Meccanica 17 (1982) 128-37.
[7] D. Fusco, J. Engelbrecht, The asymptotic analysis of non linear waves in rate dependent
media, Nuovo Ciment B 80 (1984) 49-61.
SC
[8] V.D. Sharma, L.P. Singh, R. Ram, The progressive wave approach analyzing the decay of
sawtooth profile in magnetogasdynamics, Phys Fluid 30 (1987) 1572-1574.
[9] V.D. Sharma, R.R. Sharma, B.D. Pandey, N. Gupta, Nonlinear analysis of a traffic flow, Z.
Angew. Math. Phys. 40 (1989) 828-837.
U
[10] C.H. Radha, V.D. Sharma, Propagation and interaction of waves in a relaxing gas, Philos.
Trans. R. Soc. London 352A (1995) 169-195.
AN
[11] F. Becker, Chemically reacting flows, Ann. Rev. Fluid Mech. 4 (1972) 155-194.
[12] H. Ockendon, D.A. Spence, Nonlinear wave propagation in a relaxing gas, J. Fluid Mech.
39 (1969) 329-345.
[13] B.T. Chu, Weak nonlinear waves in non equilibrium flow, in: P.P. Wegener (Ed.)
M
Nonequilibrium flows vol 1 part II, Marcel Dekker, Ney York 1970.
[14] T.B. Moodie, Y. He, D.W. Barclay, Wavefront expansions for nonlinear hyperbolic waves,
Wave Motion 14 (1991) 347-367.
D
[15] Y. He, T.B. Moodli, Geometrical optics and post shock behaviour for nonlinear
conservation law, Applicable Analysis 57 (1995) 145-176.
TE
[16] J.K. Hunter, J.B. Keller, Weakly nonlinear high frequency waves, Comm. Pure Appl. Math.
36 (1983) 547-569.
[17] A. Majda, R. Rosales, Resonantly interacting weakly nonlinear hyperbolic waves, Studies in
EP
Detonation onset following shock wave focusing, Acta Astronautica 135 (2017) 114-130.
[20] N.N. Smirnov, V.B. Betelin, V.F. Nikitin, Yu. G. Phylippov, Jaye Koo, Detonation engine
AC
[25] K. Ambika, R. Radha, V.D. Sharma, Progressive waves in non-ideal gases, Int. J. Nonlinear
Mech. 67 (2014) 285-290.
[26] R. Singh, L.P. Singh, Growth and decay behavior of sawtooth profile in non-ideal gas, Int. J.
Appl. Comp. Math. (2015), DOI 10.1007/s40819-015-0055-5.
[27] Triloki Nath, R.K. Gupta, L.P. Singh, Evolution of weak shock waves in non-ideal
Magnetogasdynamics, Acta Astronaut. 133 (2017) 397-402.
[28] V.D. Sharma, G.K. Srinivasan, Wave interaction in a nonequilibrium gas flow, Int. J. Non-
PT
Linear Mech. 40 (2005) 1031-1040.
[29] Rajan Arora, Non-planar shock waves in magnetic field, Comp. Math. with appl. 56 (2008)
2686-2691.
RI
[30] R. Arora, A. Tomar, V.P. Singh, Shock wave in reactive hydrodynamics, Shock waves 19
(2009) 145-150.
[31] Manuel Nunez% , Generation of sheet currents by high frequency fast MHD waves, Physics
SC
Letters A. 380 (2016) 2288-2293.
[32] Manuel Nunez% , Geometry of fast magnetosonic rays, wavefronts and shock waves, Physics
Letters A. 380 (2016) 3817-3822.
[33] P.A. Thompson, A fundamental derivative in gas dynamics. Phys. Fluids 14 (1971) 1843-
U
1849.
AN
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT
Highlights:
1. Method of multiple time scales is used to derive asymptotic solution of the problem.
2. Transport equation for the amplitude is derived in van der Waals gas.
3. Interaction of waves in van der Waals gas is discussed.
4. The existence of shock and its location is examined.
PT
RI
U SC
AN
M
D
TE
C EP
AC