Home A Singular Limit Problem for the Rosenau–Korteweg-de Vries-Regularized Long Wave and Rosenau-regularized Long Wave Equations
Article Open Access

A Singular Limit Problem for the Rosenau–Korteweg-de Vries-Regularized Long Wave and Rosenau-regularized Long Wave Equations

  • Giuseppe Maria Coclite EMAIL logo and Lorenzo di Ruvo
Published/Copyright: April 7, 2016

Abstract

We consider the Rosenau–Korteweg-de Vries-regularized long wave and Rosenau-regularized long wave equations, which contain nonlinear dispersive effects. We prove that by adding small diffusion to the equations, as the diffusion and dispersion parameters tends to zero, the solutions of the duffusive/dispersive equations converge to the unique entropy solution of a scalar conservation law. The proof relies on deriving suitable a priori estimates together with an application of the compensated compactness method in the Lp setting.

MSC 2010: 35G25; 35L65; 35L05

1 Introduction

The study of nonlinear wave phenomena is an important area of scientific research. In the last years several mathematical models describing wave behavior have been proposed. Some of them are the Korteweg-de Vries (KdV), the regularized long wave (RLW), the Rosenau, the Rosenau–Kawahara, the Rosenau–KdV, and the Rosenau–KdV-RLW equations (see [3, 10, 12, 14, 22]).

The KdV equation has a very wide range of applications, such as magnetic fluid waves, ion sound waves, and longitudinal astigmatic waves (see [1, 9]). The RLW equation, which was first proposed by Peregrine (see [20, 21]), is one of models which are encountered in many areas, e.g., ion-acoustic plasma waves, magnetohydrodynamic plasma waves, and shallow water waves.

The wave-wave and wave-wall interactions cannot be described by the KdV equation. Therefore, Rosenau proposed the following equation describing the dynamic of dense discrete systems (see [24, 25]):

(1.1) t u + x u 2 + t x x x x 5 u = 0 .

Existence and uniqueness of solutions for (1.1) have been proved in [18, 19].

For the further considerations of nonlinear wave phenomena, a viscous term -txx3u, and a generic flux need to be included, i.e.,

(1.2) t u - t x x 3 u + t x x x x 5 u + x u n = 0 , n , n 2 .

On the other hand, to capture other behaviors of nonlinear waves, the viscous term xxx3u, and a generic polynomial flux need to be added in (1.1). In this case we obtain the Rosenau–KdV equation (see [11, 12])

(1.3) t u + x x x 3 u + t x x x x 5 u + x u n = 0 , n , n 2 .

By coupling (1.2) and (1.3), we obtain the Rosenau–KdV-RLW equation

(1.4) t u - t x x 3 u + x x x 3 u + t x x x x 5 u + x u n = 0 , n , n 2 .

In particular, if n=2, (1.4) reads

(1.5) t u + x u 2 - t x x 3 u + x x x 3 u + t x x x x 5 u = 0 .

In [22], Razborova, Ahmed and Biswas analyzed (1.5). They got solitary waves, shock waves and singular solitons along with conservation laws.

We consider n=2 in (1.3). Therefore, we have

(1.6) t u + x u 2 + x x x 3 u + t x x x x 5 u = 0 .

In [28], Zuo discussed the solitary wave solutions of (1.6). In [12], a conservative linear finite difference scheme for the numerical solution of an initial-boundary value problem for the Rosenau–KdV equation was considered. Esfahani in [11] and Razborova Triki and Biswas in [23] studied the solitary solutions for (1.6) with the solitary ansatz method, and also gave two invariants for (1.6). In particular, in [23], the two types of soliton solutions were analyzed, one is a solitary wave and the other is a singular soliton. In [27], Zheng and Zhou proposed an average linear finite difference scheme for the numerical approximation of the solutions of the initial-boundary value problem for (1.6).

Choosing n=2 in (1.2), we get

(1.7) t u + x u 2 - t x x 3 u + t x x x x 5 u = 0 .

In this paper, we analyze (1.5) and (1.7). Arguing as in [4, 6, 7], we re-scale the equations as follows

(1.8) t u + x u 2 + β x x x 3 u - β t x x 3 u + β 2 t x x x x 5 u = 0 ,
(1.9) t u + x u 2 - β t x x 3 u + β 2 t x x x x 5 u = 0 ,

where β is the dispersion parameter.

We are interested in the no high frequency limit, we send β0 in (1.8) and (1.9). In this way we pass from (1.8) and (1.9) to the Burgers equation

(1.10) t u + x u 2 = 0 .

Therefore, we add to (1.8) and (1.9) some diffusion and get

(1.11) t u + x u 2 + β x x x 3 u - β t x x 3 u + β 2 t x x x x 5 u = ε x x 2 u ,
(1.12) t u + x u 2 - β t x x 3 u + β 2 t x x x x 5 u = ε x x 2 u ,

and prove that when εβ0, the solutions of (1.11) and (1.12) converge to the unique entropy solution of (1.10).

The paper is organized as follows. In Section 2, we prove the convergence of (1.8) to (1.10), while in Section 3, we show how to modify the argument of Section 2 and prove the convergence of (1.9) to (1.10).

2 The Rosenau–KdV-RLW Equation

In this section, we consider (1.8) and augment it with the initial condition

(2.1) u ( 0 , x ) = u 0 ( x ) ,

on which we assume that

(2.2) u 0 L 2 ( ) L 4 ( ) .

We study the dispersion-diffusion limit for equation (1.8). Therefore, we consider the following fifth order approximation:

(2.3) { t u ε , β + x u ε , β 2 + β x x x 3 u ε , β - β t x x 3 u ε , β + β 2 t x x x x 5 u ε , β = ε x x 2 u ε , β , t > 0 , x , u ε , β ( 0 , x ) = u ε , β , 0 ( x ) , x ,

where uε,β,0 is a C approximation of u0 such that

(2.4) { u ε , β , 0 u 0 in L loc p ( ) , 1 p < 4 , as ε , β 0 , u ε , β , 0 L 2 ( ) 2 + u ε , β , 0 L 4 ( ) 4 + ( β + ε 2 ) x u ε , β , 0 L 2 ( ) 2 C 0 , ε , β > 0 , ( β ε + β ε 2 + β 2 ) x x 2 u ε , β , 0 L 2 ( ) 2 + ( β 2 ε 2 + β 3 ) x x x 3 u ε , β , 0 L 2 ( ) 2 C 0 , ε , β > 0 , β 4 x x x x 4 u ε , β , 0 L 2 ( ) 2 C 0 , ε , β > 0 ,

and C0 is a constant independent of ε and β. Such a family of approximated initial data can be constructed convolving u0 with a mollifier. The well-posedness of (2.3) was proved in [18].

The main difference between (1.8) and (2.3) is the term εxx2uε,β. Such a term is needed not for well-posedness purposes like in the vanishing viscosity problem [2, 13]. However, it is needed for the convergence to entropy solutions of (1.10). Lax and Levermore proved that without such terms we may have convergence to nonclassical solutions [15].

The main result of this section is the following theorem.

Theorem 2.1

Assume that (2.2) and (2.4) hold. If

(2.5) β = 𝒪 ( ε 4 ) ,

then there exist two sequences {εn}n and {βn}n with εn,βn0, and a limit function

u L ( + ; L 2 ( ) L 4 ( ) )

such that

(2.6) u ε n , β n u strongly in L loc p ( + × ) for each 1 p < 4 ,

and

(2.7) u is the unique entropy solution of (1.10) .

Let us prove some a priori estimates on uε,β, denoting with C0 the constants which depend only on the initial data.

Lemma 2.2

For each t>0,

(2.8) u ε , β ( t , ) L 2 ( ) 2 + β x u ε , β ( t , ) L 2 ( ) 2 + β 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 + 2 ε 0 t x u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 .

In particular, we have

(2.9) u ε , β ( t , ) L ( ) C 0 β - 1 4 ,
(2.10) x u ε , β ( t , ) L ( ) C 0 β - 3 4 .

Proof.

Multiplying (2.3) by uε,β, we have

(2.11) u ε , β t u ε , β + 2 u ε , β 2 x u ε , β + β u ε , β x x x 3 u ε , β - β u ε , β t x x 3 u ε , β + β 2 u ε , β t x x x x 5 u ε , β = ε u ε , β x x 2 u ε , β .

Since

u ε , β t u ε , β d x = 1 2 d d t u ε , β ( t , ) L 2 ( ) 2 ,
2 u ε , β 2 x u ε , β d x = 0 ,
β u ε , β x x x 3 u ε , β d x = β x u ε , β x x 2 u ε , β d x = 0 ,

- β u ε , β t x x 3 u ε , β d x = β 2 d d t x u ε , β ( t , ) L 2 ( ) 2 ,
β 2 u ε , β t x x x x 5 u ε , β d x = - β 2 x u ε , β t x x x 4 u ε , β d x = β 2 2 d d t x x 2 u ε , β ( t , ) L 2 ( ) 2 ,
ε u ε , β x x 2 u ε , β d x = - ε x u ε , β ( t , ) L 2 ( ) 2 ,

integrating (2.11) on , we get

(2.12) d d t u ε , β ( t , ) L 2 ( ) 2 + β d d t x u ε , β ( t , ) L 2 ( ) 2 + β 2 d d t x x 2 u ε , β ( t , ) L 2 ( ) 2 + 2 ε x u ε , β ( t , ) L 2 ( ) 2 = 0 .

Estimate (2.8) follows from (2.4), (2.12) and an integration on (0,t).

We now prove (2.9). Due to (2.8) and the Hölder inequality,

u ε , β 2 ( t , x ) = 2 - x u ε , β x u ε , β d x 2 | u ε , β | | x u ε , β | d x 2 u ε , β ( t , ) L 2 ( ) x u ε , β ( t , ) L 2 ( ) C 0 β - 1 2 .

Therefore,

| u ε , β ( t , x ) | C 0 β - 1 4 ,

which gives (2.9).

Finally, we prove (2.10). Thanks to (2.8) and the Hölder inequality,

x u ε , β 2 ( t , x ) = 2 - x x u ε , β x x 2 u ε , β d x
2 | x u ε , β | | x x 2 u ε , β | d x
2 x u ε , β ( t , ) L 2 ( ) x x 2 u ε , β ( t , ) L 2 ( )
C 0 β - 1 2 C 0 β - 1
C 0 β - 3 2 .

Hence,

x u ε , β ( t , ) L ( ) C 0 β - 3 4 ,

that is, (2.10). ∎

Following [4, Lemma 2.2] or [7, Lemma 2.2] or [8, Lemma 4.2], we prove the following result.

Lemma 2.3

Assume (2.5). Then, the following hold:

  1. The family {uε,β}ε,β is bounded in L(+;L4()).

  2. The families {εxuε,β}ε,β, {βεxx2uε,β}ε,β, {εβxx2uε,β}ε,β, {εβxxx3uε,β}ε,β, {ββxxx3uε,β}ε,β and {βxxxx4uε,β}ε,β are bounded in L(+;L2()).

  3. The families {βεtx2uε,β}ε,β, {βεtxx3uε,β}ε,β, {ββεtxxx4uε,β}ε,β, {εββxxx3uε,β}ε,β, {εuε,βxuε,β}ε,β and {εεxx2uε,β}ε,β are bounded in L2(+×).

Moreover,

(2.13) β 0 t x u ε , β ( s , ) x x 2 u ε , β ( s , ) L 1 ( ) 𝑑 s C 0 ε 2 , t > 0 ,
(2.14) β 2 0 t x x 2 u ε , β ( t , ) L 2 ( ) 2 𝑑 s C 0 ε 5 , t > 0 .

Proof.

Let A,B,C be some positive constants which will be specified later. Multiplying (2.3) by

u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ,

we have

( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) t u ε , β
+ 2 ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) u ε , β x u ε , β
+ β ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) x x x 3 u ε , β
- β ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) t x x 3 u ε , β
+ β 2 ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) t x x x x 5 u ε , β
(2.15) = ε ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) x x 2 u ε , β .

We observe that

( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) t u ε , β d x
= 1 4 d d t u ε , β ( t , ) L 4 ( ) 4 + A β ε t x 2 u ε , β ( t , ) L 2 ( ) 2
(2.16)     + B ε 2 2 d d t x u ε , β ( t , ) L 2 ( ) 2 + C β 2 2 d d t x x 2 u ε , β ( t , ) L 2 ( ) 2 .

We have that

2 ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) u ε , β x u ε , β d x
= - 2 A β ε u ε , β x u ε , β t x x 3 u ε , β d x - 2 B ε 2 u ε , β x u ε , β x x 2 u ε , β d x
(2.17)     - 2 C β 2 ( x u ε , β ) 2 x x x 3 u ε , β d x - 2 C β 2 u ε , β x x 2 u ε , β x x x 3 u ε , β d x .

Since

- 2 C β 2 ( x u ε , β ) 2 x x x 3 u ε , β d x - 2 C β 2 u ε , β x x 2 u ε , β x x x 3 u ε , β d x = 5 C β 2 ( x x 2 u ε , β ) 2 x u ε , β d x
(2.18) = - 5 C β 2 2 ( x u ε , β ) 2 x x x 3 u ε , β d x ,

from (2.17) and (2.18), it follows that

2 ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) u ε , β x u ε , β d x
= - 2 A β ε u ε , β x u ε , β t x x 3 u ε , β d x - 2 B ε 2 u ε , β x u ε , β x x 2 u ε , β d x
(2.19)     - 5 C β 2 2 ( x u ε , β ) 2 x x x 3 u ε , β d x .

We observe that

β ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) x x x 3 u ε , β d x
(2.20) = - 3 β u ε , β 2 x u ε , β x x 2 u ε , β d x - A β 2 ε x x x 3 u ε , β t x x 3 u ε , β d x .

We get

- β ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) t x x 3 u ε , β d x
= 3 β u ε , β 2 x u ε , β t x 2 u ε , β d x + A β 2 ε t x x 3 u ε , β ( t , ) L 2 ( ) 2
(2.21)     + B β ε 2 2 d d t x x 2 u ε , β ( t , ) L 2 ( ) 2 + C β 3 2 d d t x x x 3 u ε , β ( t , ) L 2 ( ) 2 .

We have

β 2 ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) t x x x x 5 u ε , β d x
= - 3 β 2 u ε , β 2 x u ε , β t x x x 4 u ε , β d x + A β 3 ε t x x x 4 u ε , β ( t , ) L 2 ( ) 2
(2.22)     + B β 2 ε 2 2 d d t x x x 3 u ε , β ( t , ) L 2 ( ) 2 + C β 4 2 d d t x x x x 4 u ε , β ( t , ) L 2 ( ) 2 .

Moreover,

ε ( u ε , β 3 - A β ε t x x 3 u ε , β - B ε 2 x x 2 u ε , β + C β 2 x x x x 4 u ε , β ) x x 2 u ε , β d x
= - 3 ε u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 - A β ε 2 d d t x x 2 u ε , β ( t , ) L 2 ( ) 2
(2.23)     - ε 3 B x x 2 u ε , β ( t , ) L 2 ( ) 2 - β 2 ε C x x x 3 u ε , β ( t , ) L 2 ( ) 2 .

From (2.16), (2.19)–(2.23), and an integration of (2.15) on , it follows that

d d t ( 1 4 u ε , β ( t , ) L 4 ( ) 4 + A β ε + B β ε 2 + C β 2 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 )
+ d d t ( B ε 2 2 x u ε , β ( t , ) L 2 ( ) 2 + B β 2 ε 2 + C β 3 2 x x x 3 u ε , β ( t , ) L 2 ( ) 2 )
+ C β 4 2 d d t x x x x 4 u ε , β ( t , ) L 2 ( ) 2 + A β ε t x 2 u ε , β ( t , ) L 2 ( )
+ A β 2 ε t x x 3 u ε , β ( t , ) L 2 ( ) 2 + A β 3 ε t x x x 4 u ε , β ( t , ) L 2 ( ) 2
+ 3 ε u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + ε 3 B x x 2 u ε , β ( t , ) L 2 ( ) 2 + β 2 ε C x x x 3 u ε , β ( t , ) L 2 ( ) 2
= 2 A β ε u ε , β x u ε , β t x x 3 u ε , β d x + 2 B ε 2 u ε , β x u ε , β x x 2 + u ε , β d x 5 C β 2 2 ( x u ε , β ) 2 x x x 3 u ε , β d x
- 3 β u ε , β 2 x u ε , β x x 2 u ε , β d x + A β 2 ε x x x 3 u ε , β t x x 3 u ε , β d x
(2.24) - 3 β u ε , β 2 x u ε , β t x 2 u ε , β d x + 3 β 2 u ε , β 2 x u ε , β t x x x 4 u ε , β d x .

Due to the Young inequality,

2 A β ε | u ε , β x u ε , β t x x 3 u ε , β d x | A ε | 2 u ε , β x u ε , β | | β t x x 3 u ε , β | d x
2 A ε u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + A β 2 ε 2 t x x 3 u ε , β ( t , ) L 2 ( ) 2 ,
2 B ε 2 | u ε , β x u ε , β x x 2 u ε , β d x | | ε 1 2 u ε , β x u ε , β | | ε 3 2 2 B x x 2 u ε , β | d x
(2.25) ε 2 u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + 2 B 2 ε 3 x x 2 u ε , β ( t , ) L 2 ( ) 2 .

Hence, from (2.24),

d d t ( 1 4 u ε , β ( t , ) L 4 ( ) 4 + A β ε + B β ε 2 + C β 2 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 )
+ d d t ( B ε 2 2 x u ε , β ( t , ) L 2 ( ) 2 + B β 2 ε 2 + C β 3 2 x x x 3 u ε , β ( t , ) L 2 ( ) 2 )
+ C β 4 2 d d t x x x x 4 u ε , β ( t , ) L 2 ( ) 2 + A β ε t x 2 u ε , β ( t , ) L 2 ( ) 2
+ A β 2 ε 2 t x x 3 u ε , β ( t , ) L 2 ( ) 2 + A β 3 ε t x x x 4 u ε , β ( t , ) L 2 ( ) 2
+ ( 5 2 - 2 A ) ε u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + β 2 ε C x x x 3 u ε , β ( t , ) L 2 ( ) 2
+ ( B - 2 B 2 ) ε 3 x x 2 u ε , β ( t , ) L 2 ( ) 2
5 C β 2 2 ( x u ε , β ) 2 | x x x 3 u ε , β | d x + 3 β u ε , β 2 | x u ε , β | x x 2 u ε , β | d x + A β 2 ε | x x x 3 u ε , β | | t x x 3 | d x
(2.26) + 3 β u ε , β 2 | x u ε , β | | t x 2 u ε , β | d x + 3 β 2 u ε , β 2 | x u ε , β | | t x x x 4 u ε , β | 𝑑 x .

From (2.5), we have

(2.27) β D 2 ε 4 ,

where D is a positive constant which will be specified later. From (2.10), (2.27) and the Young inequality, it follows that

5 C β 2 2 ( x u ε , β ) 2 | x x x 3 u ε , β | 𝑑 x = β 2 C 5 2 ε 1 2 ( x u ε , β ) 2 | ε 1 2 x x x 3 u ε , β |
25 C β 2 8 ε ( x u ε , β ) 4 𝑑 x + C β 2 ε 2 x x x 3 u ε , β ( t , ) L 2 ( ) 2
25 C 8 ε β 2 x u ε , β ( t , ) L ( ) 2 x u ε , β ( t , ) L 2 ( ) 2 + C β 2 ε 2 x x x 3 u ε , β ( t , ) L 2 ( ) 2
C 0 β 1 2 ε x u ε , β ( t , ) L 2 ( ) 2 + C β 2 ε 2 x x x 3 u ε , β ( t , ) L 2 ( ) 2
(2.28) C 0 D ε x u ε , β ( t , ) L 2 ( ) 2 + C β 2 ε 2 x x x 3 u ε , β ( t , ) L 2 ( ) 2 .

Due to (2.9), (2.27) and the Young inequality,

3 β u ε , β 2 | x u ε , β | x x 2 u ε , β | d x 3 β u ε , β ( t , ) L ( ) 2 | x u ε , β | x x 2 u ε , β | d x
C 0 β 1 2 | x u ε , β | | x x 2 u ε , β | d x
| ε 1 2 x u ε , β | | C 0 D ε 3 2 x x 2 u ε , β | d x
(2.29) ε x u ε , β ( t , ) L 2 ( ) 2 + D 2 C 0 2 ε 3 x x 2 u ε , β ( t , ) L 2 ( ) 2 .

Thanks to the Young inequality,

A β 2 ε | x x x 3 u ε , β | | t x x 3 u ε , β | d x = A β 2 ε | 2 x x x 3 u ε , β | | 1 2 t x x 3 u ε , β | d x
(2.30) 2 A β 2 ε x x x 3 u ε , β ( t , ) L 2 ( ) 2 + A β 2 ε 8 t x x 3 u ε , β ( t , ) L 2 ( ) 2 .

From (2.9), (2.27) and the Young inequality, it follows that

3 β u ε , β 2 | x u ε , β | | t x 2 u ε , β | d x = β | 3 u ε , β 2 x u ε , β ε 1 2 A 1 2 | | ε 1 2 A 1 2 t x 2 u ε , β | d x
9 β 2 ε A u ε , β 4 ( x u ε , β ) 2 𝑑 x + β ε A 2 t x 2 u ε , β ( t , ) L 2 ( ) 2
9 2 ε A β u ε , β ( t , ) L ( ) 2 u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + β ε A 2 t x 2 u ε , β ( t , ) L 2 ( ) 2
C 0 β 1 2 ε A u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + β ε A 2 t x 2 u ε , β ( t , ) L 2 ( ) 2
(2.31) C 0 D A ε u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + β ε A 2 t x 2 u ε , β ( t , ) L 2 ( ) 2 .

Again by (2.9), (2.27) and the Young inequality,

3 β 2 u ε , β 2 | x u ε , β | | t x x x 4 u ε , β | d x = | 3 β 1 2 u ε , β 2 x u ε , β ε 1 2 A 1 2 | | β 3 2 ε 1 2 A 1 2 t x x x 4 u ε , β | d x
9 β 2 ε A u ε , β 4 ( x u ε , β ) 2 𝑑 x + β 3 ε A 2 t x x x 4 u ε , β ( t , ) L 2 ( ) 2
9 2 ε A β u ε , β ( t , ) L ( ) 2 u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2
+ β 3 ε A 2 t x x x 4 u ε , β ( t , ) L 2 ( ) 2
C 0 β 1 2 ε A u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + β 3 ε A 2 t x x x 4 u ε , β ( t , ) L 2 ( ) 2
(2.32) C 0 D A ε u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + β 3 ε A 2 t x x x 4 u ε , β ( t , ) L 2 ( ) 2 .

From (2.26), (2.28)–(2.32), we get

d d t ( 1 4 u ε , β ( t , ) L 4 ( ) 4 + A β ε + B β ε 2 + C β 2 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 )
+ d d t ( B ε 2 2 x u ε , β ( t , ) L 2 ( ) 2 + B β 2 ε 2 + C β 3 2 x x x 3 u ε , β ( t , ) L 2 ( ) 2 ) + C β 4 2 d d t x x x x 4 u ε , β ( t , ) L 2 ( ) 2
+ β ε A 2 t x 2 u ε , β ( t , ) L 2 ( ) 2 + 3 A β 2 ε 8 t x x 3 u ε , β ( t , ) L 2 ( ) 2 + A β 3 ε 2 t x x x 4 u ε , β ( t , ) L 2 ( ) 2
+ β 2 ε ( C 2 - 2 A ) x x x 3 u ε , β ( t , ) L 2 ( ) 2 + ( 5 2 - 2 A - C 0 D A ) ε u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2
+ ( B - 2 B 2 - D 2 C 0 2 ) ε 3 x x 2 u ε , β ( t , ) L 2 ( ) 2
(2.33) C 0 ε x u ε , β ( t , ) L 2 ( ) 2 .

We search for A, B and C such that

5 2 - 2 A - C 0 D A > 0 , B - 2 B 2 - D 2 C 0 2 > 0 and C 2 - 2 A > 0 ,

that is,

(2.34) 4 A 2 - 5 A + 2 C 0 D < 0 , 2 B 2 - B - D 2 C 0 2 < 0 and C > 4 A .

We choose

(2.35) C = 6 A .

The first inequality of (2.34) admits a solution if

25 - 32 C 0 D > 0 ,

that is,

(2.36) D < 25 32 C 0 .

The second inequality of (2.34) admits a solution if

1 - 8 D 2 C 0 2 > 0 ,

that is,

(2.37) D < 2 4 C 0 .

From (2.36) and (2.37), it follows that

(2.38) D < min { 25 32 C 0 , 2 4 C 0 } = 2 4 C 0 .

Therefore, from (2.34), (2.35) and (2.38), we have that there exist 0<A1<A2 and 0<B1<B2, such that by choosing

(2.39) A 1 < A < A 2 , B 1 < B < B 2 and C = 6 A ,

(2.34) holds.

Now (2.33) and (2.34) give

d d t ( 1 4 u ε , β ( t , ) L 4 ( ) 4 + A β ε + B β ε 2 + 6 A β 2 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 )
+ d d t ( B ε 2 2 x u ε , β ( t , ) L 2 ( ) 2 + B β 2 ε 2 + 6 A β 3 2 x x x 3 u ε , β ( t , ) L 2 ( ) 2 ) + 3 A β 4 d d t x x x x 4 u ε , β ( t , ) L 2 ( ) 2
+ β ε A 2 t x 2 u ε , β ( t , ) L 2 ( ) 2 + 3 A β 2 ε 8 t x x 3 u ε , β ( t , ) L 2 ( ) 2 + A β 3 ε 2 t x x x 4 u ε , β ( t , ) L 2 ( ) 2
+ β 2 ε A x x x 3 u ε , β ( t , ) L 2 ( ) 2 + ε K 1 u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + ε 3 K x x 2 u ε , β ( t , ) L 2 ( ) 2
C 0 ε x u ε , β ( t , ) L 2 ( ) 2

for some K1,K2>0.

The assumptions in (2.4), (2.8) and an integration on (0,t) give

1 4 u ε , β ( t , ) L 4 ( ) 4 + A β ε + B β ε 2 + 6 A β 2 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 + B ε 2 2 x u ε , β ( t , ) L 2 ( ) 2
+ B β 2 ε 2 + 6 A β 3 2 x x x 3 u ε , β ( t , ) L 2 ( ) 2 + 3 A β 4 x x x x 4 u ε , β ( t , ) L 2 ( ) 2 + β ε A 2 0 t t x 2 u ε , β ( s , ) L 2 ( ) 2 𝑑 s
+ 3 A β 2 ε 8 0 t t x x 3 u ε , β ( s , ) L 2 ( ) 2 𝑑 s + A β 3 ε 2 0 t t x x x 4 u ε , β ( s , ) L 2 ( ) 2 𝑑 s + β 2 ε A 0 t x x x 3 u ε , β ( s , ) L 2 ( ) 2 𝑑 s
+ ε K 1 0 t u ε , β ( s , ) x u ε , β ( t , ) L 2 ( ) 2 𝑑 s + ε 3 K 2 0 t x x 2 u ε , β ( s , ) L 2 ( ) 2 𝑑 s
C 0 + C 0 ε 0 t x u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 .

Hence,

u ε , β ( t , ) L 4 ( ) C 0 ,
ε x u ε , β ( t , ) L 2 ( ) C 0 ,
β ε x x 2 u ε , β ( t , ) L 2 ( ) C 0 ,
ε β x x 2 u ε , β ( t , ) L 2 ( ) C 0 ,
β x x 2 u ε , β ( t , ) L 2 ( ) C 0 ,
β ε x x x 3 u ε , β ( t , ) L 2 ( ) C 0 ,
β β x x x 3 u ε , β ( t , ) L 2 ( ) C 0 ,
β x x x x 4 u ε , β ( t , ) L 2 ( ) C 0 ,

β ε 0 t t x 2 u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 ,
β 2 ε 0 t t x x 3 u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 ,
β 3 ε 0 t t x x x 4 u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 ,
β 2 ε 0 t x x x 3 u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 ,
ε 0 t u ε , β ( s , ) x u ε , β ( t , ) L 2 ( ) 2 𝑑 s C 0 ,
ε 3 0 t x x 2 u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0

for every t>0.

Arguing as in [4, Lemma 2.2], we have (2.13) and (2.14). ∎

To prove Theorem 2.1, we consider the following definition.

Definition 2.4

A pair of functions (η,q) is called an entropy-entropy flux pair if η: is a C2 function and q: is defined by

q ( u ) = 2 0 u ξ η ( ξ ) 𝑑 ξ .

An entropy-entropy flux pair (η,q) is called convex if, in addition, η is convex, (η,q) is called compactly supported if, in addition, η is compactly supported, i.e., there exists R>0 such that if |u|R, then η(u)=0.

Moreover, the following technical lemmas are needed [17, 26].

Lemma 2.5

Let Ω be a bounded open subset of 2. Assume that the sequence {n}n of distributions is bounded in W-1,(Ω). Assume also that

n = 1 , n + 2 , n ,

where {1,n}n lies in a compact subset of Hloc-1(Ω) and {2,n}n lies in a bounded subset of loc(Ω). Then, {n}n lies in a compact subset of Hloc-1(Ω).

Lemma 2.6

Let {vν}ν>0 be a family of functions defined on +× such that

v ν L p ( ( 0 , T ) × ) M T , T , ν > 0 ,

for some p>2, and the family

{ t η ( v ν ) + x q ( v ν ) } ν > 0

is compact in Hloc-1(+×) for every convex ηC2(), where q(u)=2uη(u). Then, there exist a sequence {νn}n(0,) with νn0 and a map vLp((0,T)×),T>0 such that

v ν n v a.e. and in L loc r ( + × ) ,  1 r < 2 .

Following [16], we prove Theorem 2.1.

Proof of Theorem 2.1.

Let us consider a compactly supported entropy-entropy flux pair (η,q). Multiplying (2.3) by η(uε,β), we have

t η ( u ε , β ) + x q ( u ε , β ) = ε η ( u ε , β ) x x 2 u ε , β - β η ( u ε , β ) x x x 3 u ε , β - β η ( u ε , β ) t x x 3 u ε , β + β 2 η ( u ε , β ) t x x x x 5 u ε , β
= I 1 , ε , β + I 2 , ε , β + I 3 , ε , β + I 4 , ε , β + I 5 , ε , β + I 6 , ε , β + I 7 , ε , β + I 8 , ε , β ,

where

I 1 , ε , β = x ( ε η ( u ε , β ) x u ε , β ) ,
I 2 , ε , β = - ε η ′′ ( u ε , β ) ( x u ε , β ) 2 ,
I 3 , ε , β = x ( - β η ( u ε , β ) x x 2 u ε , β ) ,
I 4 , ε , β = β η ′′ ( u ε , β ) x u ε , β x x 2 u ε , β ,
I 5 , ε , β = x ( - β η ( u ε , β ) t x 2 u ε , β ) ,
I 6 , ε , β = β η ′′ ( u ε , β ) x u ε , β t x 2 u ε , β ,
I 7 , ε , β = x ( β 2 η ( u ε , β ) t x x x 4 u ε , β ) ,
(2.40) I 8 , ε , β = - β 2 η ′′ ( u ε , β ) x u ε , β t x x x 4 u ε , β .

Fix T>0. Arguing as in [5, Lemma 3.2], we have that I1,ε,β0 in H-1((0,T)×), and {I2,ε,β}ε,β>0 is bounded in L1((0,T)×). Arguing as in [4, Theorem 1.1], we get I3,ε,β0 in H-1((0,T)×), and I4,ε,β0 in L1((0,T)×). We claim that

I 5 , ε , β 0 in H - 1 ( ( 0 , T ) × ) , T > 0 , as ε 0 .

By (2.5) and Lemma 2.3,

β η ( u ε , β ) t x 2 u ε , β L 2 ( ( 0 , T ) × ) 2 η L ( ) 2 β 2 0 T t x 2 u ε , β ( t , ) L 2 ( ) 2 𝑑 t
= η L ( ) 2 β 2 ε ε 0 T t x 2 u ε , β ( t , ) L 2 ( ) 2 𝑑 t
C 0 η L ( ) 2 β ε
C 0 η L ( ) 2 ε 3 0 .

We have that

I 6 , ε , β 0 in L 1 ( ( 0 , T ) × ) , T > 0 , as ε 0 .

Due to (2.5), Lemmas 2.2 and 2.3, and the Hölder inequality,

β η ′′ ( u ε , β ) x u ε , β t x 2 u ε , β L 1 ( ( 0 , T ) × ) η ′′ L ( ) β 0 T | x u ε , β t x 2 u ε , β | 𝑑 t 𝑑 x
η ′′ L ( ) β ε ε x u ε , β L 2 ( ( 0 , T ) × ) t u ε , β L 2 ( ( 0 , T ) × )
C 0 η ′′ L ( ) β 1 2 ε
C 0 η ′′ L ( ) ε 0 .

We claim that

I 7 , ε , β 0 in H - 1 ( ( 0 , T ) × ) , T > 0 , as ε 0 .

By (2.5) and Lemma 2.3,

β 2 η ( u ε , β ) t x x x 4 u ε , β L 2 ( ( 0 , T ) × ) 2 β 4 η L ( ) t x x x 4 u ε , β L 2 ( ( 0 , T ) × ) 2
= η L ( ) β 4 ε ε t x x x 4 u ε , β L 2 ( ( 0 , T ) × ) 2
C 0 η L ( ) β ε
C 0 η L ( ) ε 3 0 .

We have that

I 8 , ε , β 0 in L 1 ( ( 0 , T ) × ) , T > 0 , as ε 0 .

Thanks to (2.5), Lemmas 2.2 and 2.3, and the Hölder inequality,

β 2 η ′′ ( u ε , β ) x u ε , β t x x x 4 u ε , β L 1 ( ( 0 , T ) × ) β 2 η ′′ L ( ) 0 T | x u ε , β t x x x 4 u ε , β | d s d x
η ′′ L ( ) β 2 ε ε x u ε , β L 2 ( ( 0 , T ) × ) t x x x 4 u ε , β L 2 ( ( 0 , T ) × )
C 0 η ′′ L ( ) β 1 2 ε
C 0 η ′′ L ( ) ε 0 .

Therefore, (2.6) follows from Lemmas 2.5 and 2.6.

To have (2.7), we begin by proving that u is a distributional solution of (1.10). Let ϕC(2) be a test function with compact support. We have to prove that

(2.41) 0 ( u t ϕ + u 2 x ϕ ) 𝑑 t 𝑑 x + u 0 ( x ) ϕ ( 0 , x ) 𝑑 x = 0 .

We define

(2.42) u ε n , β n := u n .

We have that

0 ( u n t ϕ + u n 2 x ϕ ) 𝑑 t 𝑑 x + u 0 , n ( x ) ϕ ( 0 , x ) 𝑑 x + ε n 0 u n x x 2 ϕ d t d x
+ β n 0 u n x x x 3 ϕ d t d x - β n 0 u n t x x 3 ϕ d t d s - β n 0 u 0 , n ( x ) x x 2 ϕ ( 0 , x ) 𝑑 x
+ β n 2 0 u n t x x x x 5 ϕ d t d s + β n 0 u 0 , n ( x ) x x x x 4 ϕ ( 0 , x ) 𝑑 x = 0 .

Therefore, (2.41) follows from (2.4) and (2.6).

We conclude by proving that u is the unique entropy solution of (1.10). Fix T>0. Let us consider a compactly supported entropy-entropy flux pair (η,q) and a non-negative function ϕCc((0,)×). We have to prove that

(2.43) 0 ( t η ( u ) + x q ( u ) ) ϕ 𝑑 t 𝑑 x 0 .

We have

0 ( t η ( u n ) + x q ( u n ) ) ϕ 𝑑 t 𝑑 x
= ε n 0 x ( η ( u n ) x u n ) ϕ d t d x - ε n 0 η ′′ ( u n ) ( x u n ) 2 ϕ 𝑑 t 𝑑 x - β n 0 x ( η ( u n ) x x 2 u n ) ϕ d t d x
    + β n 0 η ′′ ( u n ) x u n x x 2 u n ϕ d t d x - β n 0 x ( η ( u n ) t x 2 u n ) ϕ d t d x
    + β n 0 η ′′ ( u n ) x u n t x 2 u n ϕ d t d x + β n 2 0 x ( η ( u n ) t x x x 4 u n ) ϕ d t d x
    - β n 2 0 η ′′ ( u n ) x u n t x x x 4 u n ϕ d t d x
- ε n 0 η ( u n ) x u n x ϕ d t d x + β n 0 η ( u n ) x x 2 u n x ϕ d t d x + β n 0 η ′′ ( u n ) x u n x x 2 u n ϕ d t d x
    + β n 0 η ( u n ) t x 2 u n x ϕ d t d x + β n 0 η ′′ ( u n ) x u n t x 2 u n ϕ d t d x
    - β n 2 0 η ( u n ) t x x x 4 u n x ϕ d t d x - β n 2 0 η ′′ ( u n ) x u n t x x x 4 u n ϕ d t d x
ε n 0 | η ( u n ) | | x u n | | x ϕ | d t d x + β n 0 | η ( u n ) | | x x 2 u n | | x ϕ | d t d x
    + β n 0 | η ′′ ( u n ) | | x u n | | x x 2 u n | | ϕ | d t d x + β n 0 | η ( u n ) | | t x 2 u n | | x ϕ | d t d x
    + β n 0 | η ′′ ( u n ) | | x u n | | t x 2 u n | | ϕ | d t d x + β n 2 0 | η ( u n ) | | t x x x 4 u n | | x ϕ | d t d x
    + β n 2 0 | η ′′ ( u n ) | | x u n | | t x x x 4 u n | ϕ d t d x
ε n η L ( ) x u n L 2 ( supp ( x ϕ ) ) x ϕ L 2 ( supp ( x ϕ ) )
    + β n η L ( ) x x 2 u n L 2 ( supp ( x ϕ ) ) x ϕ L 2 ( supp ( x ϕ ) )
    + β n η ′′ L ( ) ϕ L ( + × ) x u n x x 2 u n L 1 ( supp ( ϕ ) )
    + β n η L ( ) t x 2 u n L 2 ( supp ( x ϕ ) ) x ϕ L 2 ( supp ( x ϕ ) )
    + β n η ′′ L ( ) ϕ L ( + × x u n t x 2 u n L 1 ( supp ( ϕ ) )
    + β n 2 η L ( ) t x x x 4 u n L 2 ( supp ( x ϕ ) ) x ϕ L 2 ( supp ( x ϕ ) )
    + β n 2 η ′′ L ( ) ϕ L ( + × ) x u n t x x x 4 u n L 1 ( supp ( ϕ ) )
ε n η L ( ) x u n L 2 ( ( 0 , T ) × ) x ϕ L 2 ( ( 0 , T ) × )
    + β n η L ( ) x x 2 u n L 2 ( ( 0 , T ) × ) x ϕ L 2 ( ( 0 , T ) × )
    + β n η ′′ L ( ) ϕ L ( + × ) x u n x x 2 u n L 1 ( ( 0 , T ) × )
    + β n η L ( ) t x 2 u n L 2 ( ( 0 , T ) × ) x ϕ L 2 ( ( 0 , T ) × )
    + β n η ′′ L ( ) ϕ L ( + × ) x u n t x 2 u n L 1 ( ( 0 , T ) × )
    + β n 2 η L ( ) t x x x 4 u n L 2 ( ( 0 , T ) × ) x ϕ L 2 ( ( 0 , T ) × )
    + β n 2 η ′′ L ( ) ϕ L ( + × ) x u n t x x x 4 u n L 1 ( ( 0 , T ) × ) .

Inequality (2.43) follows from (2.5), (2.6), Lemmas 2.2 and 2.3. ∎

3 The Rosenau-RLW Equation

In this section, we consider (1.9) and augment (1.9) with the initial condition

(3.1) u ( 0 , x ) = u 0 ( x ) ,

on which we assume that

(3.2) u 0 L 2 ( ) L 4 ( ) .

We study the dispersion-diffusion limit for (1.9). Therefore, we consider the following fifth order problem:

(3.3) { t u ε , β + x u ε , β 2 - β t x x 3 u ε , β + β 2 t x x x x 5 u ε , β = ε x x 2 u ε , β , t > 0 , x , u ε , β ( 0 , x ) = u ε , β , 0 ( x ) , x ,

where uε,β,0 is a C approximation of u0 such that

(3.4) { u ε , β ,  0 u 0 in L loc p ( ) , 1 p < 4 , as ε , β 0 , u ε , β , 0 L 2 ( ) 2 + u ε , β , 0 L 4 ( ) 4 + β ε 2 x x 2 u ε , β , 0 L 2 ( ) 2 C 0 , ε , β > 0 ,

and C0 is a constant independent on ε and β. The well-posedness of (3.3) was proved in [18].

The main result of this section is the following theorem.

Theorem 3.1

Assume that (3.2) and (3.4) hold. If

(3.5) β = 𝒪 ( ε 4 ) ,

then there exist two sequences {εn}n and {βn}n with εn,βn0, and a limit function

u L ( + ; L 2 ( ) L 4 ( ) )

such that

(3.6) u ε n , β n u strongly in L loc p ( + × ) for each 1 p < 4 ,

and

(3.7) u is the unique entropy solution of (1.10) .

Let us prove some a priori estimates on uε,β, denoting with C0 the constants which depend on the initial data.

We begin by observing that Lemma 2.2 holds also for (3.3).

Following [4, Lemma 2.2] or [7, Lemma 2.2] or [8, Lemma 4.2], we prove the following result.

Lemma 3.2

Assume (3.5). For each t>0, the following hold:

  1. The family {uε,β}ε,β is bounded in L(+;L4()).

  2. The family {εβxx2uε,β}ε,β is bounded in L(+;L2()).

  3. The families {βεtx2uε,β}ε,β, {βεtxx3uε,β}ε,β, {ββεtxxx4uε,β}ε,β and {εuε,βxuε,β}ε,β are bounded in L2(+×).

Proof.

Let A be a positive constant which will be specified later. Multiplying (3.3) by uε,β3-Aβεtxx3uε,β, we have

( u ε , β 3 - A β ε t x x 3 u ε , β ) t u ε , β + 2 ( u ε , β 3 - A β ε t x x 3 u ε , β ) u ε , β x u ε , β
- β ( u ε , β 3 - A β ε t x x 3 u ε , β ) t x x 3 u ε , β + β 2 ( u ε , β 3 - A β ε t x x 3 u ε , β ) t x x x x 5 u ε , β
(3.8) = ε ( u ε , β 3 - A β ε t x x 3 u ε , β ) x x 2 u ε , β .

Since

( u ε , β 3 - A β ε t x x 3 u ε , β ) t u ε , β d x = 1 4 d d t u ε , β ( t , ) L 4 ( ) 4 + A β ε t x 2 u ε , β ( t , ) L 2 ( ) 2 ,
2 ( u ε , β 3 - A β ε t x x 3 u ε , β ) u ε , β x u ε , β d x = - 2 A β ε u ε , β x u ε , β t x x 3 u ε , β d x ,
- β ( u ε , β 3 - A β ε t x x 3 u ε , β ) t x x 3 u ε , β d x = 3 β u ε , β 2 x u ε , β t x 2 u ε , β d x + A β 2 ε t x x 3 u ε , β ( t , ) L 2 ( ) 2 ,
β 2 ( u ε , β 3 - A β ε t x x 3 u ε , β ) t x x x x 5 u ε , β d x = - 3 β 2 u ε , β 2 x u ε , β t x x x 4 u ε , β d x + A β 3 ε t x x x 4 u ε , β ( t , ) L ( ) 2

and

ε ( u ε , β 3 - A β ε t x x 3 u ε , β ) x x 2 u ε , β d x = - 3 ε u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 - A β ε 2 2 d d t x x 2 u ε , β ( t , ) L 2 ( ) 2 ,

integrating (3.8) on , we get

d d t ( 1 4 u ε , β ( t , ) L 4 ( ) 4 + A β ε 2 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 ) + A β ε t x 2 u ε , β ( t , ) L 2 ( ) 2 + A β 2 ε t x x 3 u ε , β ( t , ) L 2 ( ) 2
+ A β 3 ε t x x x 4 u ε , β ( t , ) L 2 ( ) 2 + 3 ε u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2
(3.9) = 2 A β ε u ε , β x u ε , β t x x 3 u ε , β d x - 3 β u ε , β 2 x u ε , β t x 2 u ε , β d x + 3 β 2 u ε , β 2 x u ε , β t x x x 4 u ε , β d x .

From (2.25), (2.27), (2.31), (2.32) and (3.9), it follows that

d d t ( 1 4 u ε , β ( t , ) L 4 ( ) 4 + A β ε 2 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 ) + A β ε 2 t x 2 u ε , β ( t , ) L 2 ( ) 2 + A β 2 ε 2 t x x 3 u ε , β ( t , ) L 2 ( ) 2
+ ε ( 3 - 2 C 0 D A - 2 A ) u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + A β 3 ε 2 t x x x 4 u ε , β ( t , ) L 2 ( ) 2 0 ,

where D is a positive constant which will be specified later.

We search for a constant A such that

3 - 2 C 0 D A - 2 A > 0 ,

that is,

(3.10) 2 A 2 - 3 A + 2 C 0 D < 0 .

Such A exists if and only if 9-16C0D>0. So choosing D=116C0, it follows that there exist 0<A1<A2 such that (3.10) holds for every A1<A<A2. Hence, we get

d d t ( 1 4 u ε , β ( t , ) L 4 ( ) 4 + A β ε 2 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 ) + A β ε 2 t x 2 u ε , β ( t , ) L 2 ( ) 2 + A β 2 ε 2 t x x 3 u ε , β ( t , ) L 2 ( ) 2
(3.11) + ε K 1 u ε , β ( t , ) x u ε , β ( t , ) L 2 ( ) 2 + A β 3 ε 2 t x x x 4 u ε , β ( t , ) L 2 ( ) 2 0 ,

where K1 is a fixed positive constant. Integrating (3.11) on (0,t), from (3.4) we have

1 4 u ε , β ( t , ) L 4 ( ) 4 + A 3 β ε 2 2 x x 2 u ε , β ( t , ) L 2 ( ) 2 + A 3 β ε 2 0 t t x 2 u ε , β ( s , ) L 2 ( ) 2 𝑑 s
+ A 3 β 2 ε 2 0 t t x x 3 u ε , β ( s , ) L 2 ( ) 2 𝑑 s + ε K 1 0 t u ε , β ( s , ) x u ε , β ( s , ) L 2 ( ) 2 𝑑 s
+ A 3 β 3 ε 2 0 t t x x x 4 u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 .

Hence,

u ε , β ( t , ) L 4 ( ) C 0 ,
β ε x x 2 u ε , β ( t , ) L 2 ( ) C 0 ,
β ε 0 t t x 2 u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 ,
β 2 ε 0 t t x x 3 u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 ,
ε 0 t u ε , β ( s , ) x u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0 ,
β 3 ε 0 t t x x x 4 u ε , β ( s , ) L 2 ( ) 2 𝑑 s C 0

for every t>0. ∎

Now, we are ready for the proof of Theorem 3.1.

Proof of Theorem 3.1.

Let us consider a compactly supported entropy–entropy flux pair (η,q). Multiplying (3.3) by η(uε,β), we have

t η ( u ε , β ) + x q ( u ε , β ) = ε η ( u ε , β ) x x 2 u ε , β - β η ( u ε , β ) t x x 3 u ε , β + β 2 η ( u ε , β ) t x x x x 5 u ε , β
= I 1 , ε , β + I 2 , ε , β + I 3 , ε , β + I 4 , ε , β + I 5 , ε , β + I 6 , ε , β ,

where

I 1 , ε , β = x ( ε η ( u ε , β ) x u ε , β ) ,
I 2 , ε , β = - ε η ′′ ( u ε , β ) ( x u ε , β ) 2 ,
I 3 , ε , β = x ( - β η ( u ε , β ) t x 2 u ε , β ) ,
I 4 , ε , β = β η ′′ ( u ε , β ) x u ε , β t x 2 u ε , β ,
I 5 , ε , β = x ( β 2 η ( u ε , β ) t x x x 4 u ε , β ) ,
I 6 , ε , β = - β 2 η ′′ ( u ε , β ) x u ε , β t x x x 4 u ε , β .

Following Theorem 2.1, we have that I1,ε,β, I3,ε,β, I5,ε,β0 in H-1((0,T)×), {I2,ε,β}ε,β>0 is bounded in L1((0,T)×), I4,ε,β, I6,ε,β0 in L1((0,T)×).

Arguing as in Theorem 2.1, we get (3.6) and (3.7). ∎

The authors are members of the Gruppo Nazionale per l’Analisi Matematica, la Probabilità e le loro Applicazioni (GNAMPA) of the Istituto Nazionale di Alta Matematica (INdAM).

References

[1] Bahadir A. R., Exponential finite-difference method applied to Korteweg–de Vries equation for small times, Appl. Math. Comput. 160 (2005), no. 3, 675–682. 10.1016/j.amc.2003.11.025Search in Google Scholar

[2] Bianchini S. and Bressan A., Vanishing viscosity solutions of nonlinear hyperbolic systems, Ann. of Math. (2) 161 (2005), no. 1, 223–342. 10.4007/annals.2005.161.223Search in Google Scholar

[3] Biswas A., Triki H. and Labidi M., Bright and dark solitons of the Rosenau–Kawahara equation with power law nonlinearity, Phys. Wave Phenom. 19 (2011), no. 1, 24–29. 10.3103/S1541308X11010067Search in Google Scholar

[4] Coclite G. M. and di Ruvo L., A singular limit problem for the Kudryashov–Sinelshchikov equation, ZAMM Z. Angew. Math. Mech., to appear. 10.1002/zamm.201500146Search in Google Scholar

[5] Coclite G. M. and di Ruvo L., Convergence of the Ostrovsky Equation to the Ostrovsky–Hunter one, J. Differential Equations 256 (2014), no. 9, 3245–3277. 10.1016/j.jde.2014.02.001Search in Google Scholar

[6] Coclite G. M. and di Ruvo L., A singular limit problem for conservation laws related to the Kawahara–Korteweg–de Vries equation, Netw. Heterog. Media 11 (2016), 281–300. 10.3934/nhm.2016.11.281Search in Google Scholar

[7] Coclite G. M., di Ruvo L., Ernest J. and Mishra S., Convergence of vanishing capillarity approximations for scalar conservation laws with discontinuous fluxes, Netw. Heterog. Media 8 (2013), no. 4, 969–984. 10.3934/nhm.2013.8.969Search in Google Scholar

[8] Coclite G. M. and Karlsen K. H., A singular limit problem for conservation laws related to the Camassa–Holm shallow water equation, Comm. Partial Differential Equations 31 (2006), no. 8, 1253–1272. 10.1080/03605300600781600Search in Google Scholar

[9] Cui Y. and Mao D.-K., Numerical method satisfying the first two conservation laws for the Korteweg–de Vries equation, J. Comput. Phys. 227 (2007), no. 1, 376–399. 10.1016/j.jcp.2007.07.031Search in Google Scholar

[10] Ebadi G., Mojaver A., Triki H., Yildirim A. and Biswas A., Topological solitons and other solutions of the Rosenau–KdV equation with power law nonlinearity, Rom. Journ. Phys 58 (2013), no. 1–2, 3–14. Search in Google Scholar

[11] Esfahani A., Solitary wave solutions for generalized Rosenau–KdV equation, Commun. Theor. Phys. 55 (2011), no. 3, 396–398. 10.1088/0253-6102/55/3/04Search in Google Scholar

[12] Hu J., Xu Y. and Hu B., Conservative linear difference scheme for Rosenau–Kdv equation, Adv. Math. Phys. (2013), Article ID 423718. 10.1155/2013/423718Search in Google Scholar

[13] Kružkov S. N., First order quasilinear equations in several independent variables, Math. USSR Sb. 10 (1970), no. 2, 217–243. 10.1070/SM1970v010n02ABEH002156Search in Google Scholar

[14] Labidi M. and Biswas A., Application of He’s principles to Rosenau–Kawahara equation, Math. Eng. Sci. Aerospace 2 (2011), no. 2, 183–197. Search in Google Scholar

[15] Lax P. and Levermore C. D., The zero dispersion limit for the Korteweg–de Vries equation, Proc. Natl. Acad. Sci. USA 2 (1979), 3602–3606. 10.1073/pnas.76.8.3602Search in Google Scholar PubMed PubMed Central

[16] LeFloch P. G. and Natalini R., Conservation laws with vanishing nonlinear diffusion and dispersion, Nonlinear Anal. 36 (1992), no. 2, 212–230. Search in Google Scholar

[17] Murat F., L’injection du cône positif de H-1 dans W-1,q est compacte pour tout q<2, J. Math. Pures Appl. (9) 60 (1981), 309–322. Search in Google Scholar

[18] Park M. A., On the Rosenau equation, Mat. Apl. Comput. 9 (1990), no. 2, 145–152. Search in Google Scholar

[19] Park M. A., Pointwise decay estimates of solutions of the generalized Rosenau equation, J. Korean Math. Soc. 29 (1992), no. 2, 261–280. Search in Google Scholar

[20] Peregrine D. H., Calculations of the development of an undular bore, J. Fluid Mech. 25 (1966), no. 2, 321–330. 10.1017/S0022112066001678Search in Google Scholar

[21] Peregrine D. H., Long waves on a beach, J. Fluid Mech. 27 (1967), no. 4, 815–827. 10.1017/S0022112067002605Search in Google Scholar

[22] Razborova P., Ahmed B. and Biswas A., Solitons, shock waves and conservation laws of Rosenau–KdV–RLW equation with power law nonlinearity, Appl. Math. Inf. Sci. 8 (2014), no. 2, 485–491. 10.12785/amis/080205Search in Google Scholar

[23] Razborova P., Triki H. and Biswas A., Perturbation of dispersive shallow water waves, Ocean Eng. 63 (2013), 1–7. 10.1016/j.oceaneng.2013.01.014Search in Google Scholar

[24] Rosenau P., A quasi-continuous description of a nonlinear transmission line, Phys. Scripta 34 (1986), no. 6B, 827–829. 10.1088/0031-8949/34/6B/020Search in Google Scholar

[25] Rosenau P., Dynamics of dense discrete system, Progr. Theoret. Phys. 79 (1988), no. 5, 1028–1042. 10.1143/PTP.79.1028Search in Google Scholar

[26] Schonbek M. E., Convergence of solutions to nonlinear dispersive equations, Comm. Partial Differential Equations 7 (1982), 959–1000. 10.1080/03605308208820242Search in Google Scholar

[27] Zheng M. and Zhou J., An average linear difference scheme for the generalized Rosenau–KdV equation, J. Appl. Math. (2014), Article ID 202793. 10.1155/2014/202793Search in Google Scholar

[28] Zuo J. M., Solitons and periodic solutions for the Rosenau–KdV and Rosenau–Kawahara equations, Appl. Math. Comput. 215 (2009), no. 2, 835–840. 10.1016/j.amc.2009.06.011Search in Google Scholar

Received: 2015-12-24
Revised: 2016-01-18
Accepted: 2016-01-19
Published Online: 2016-04-07
Published in Print: 2016-08-01

© 2016 by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.

Downloaded on 25.10.2025 from https://www.degruyterbrill.com/document/doi/10.1515/ans-2015-5034/html
Scroll to top button