SCIENCE CHINA

Mathematics
. ARTICLES . doi: 10.1007/s11425-014-4792-4

Convergence of compressible Navier-Stokes-Maxwell

equations to incompressible Navier-Stokes equations

YANG JianWei1,2,∗ & WANG Shu3

1Instituteof Applied Physics and Computational Mathematics, Beijing 100088, China;

2College of Mathematics and Information Science, North China University of Water Resources and
Electric Power, Zhengzhou 450046, China;
3College of Applied Sciences, Beijing University of Technology, Beijing 100022, China

Email: [email protected], [email protected]

Received November 8, 2012; accepted September 6, 2013

Abstract The combined quasi-neutral and non-relativistic limit of compressible Navier-Stokes-Maxwell equa-
tions for plasmas is studied. For well-prepared initial data, it is shown that the smooth solution of compressible
Navier-Stokes-Maxwell equations converges to the smooth solution of incompressible Navier-Stokes equations
by introducing new modulated energy functional.

Keywords Navier-Stokes-Maxwell equations, incompressible Navier-Stokes equations, asymptotic limit, mod-

ulated energy function

MSC(2010) 35B40, 35C20, 35Q35, 76Y05

Citation: Yang J W, Wang S. Convergence of compressible Navier-Stokes-Maxwell equations to incompressible

Navier-Stokes equations. Sci China Math, 2014, 57, doi: 10.1007/s11425-014-4792-4

1 Introduction

The compressible Navier-Stokes-Maxwell (NSM) system reads

∂t n + div(n u ) = 0, (1.1)
           
∂t (n u ) + div(n u ⊗ u ) + ∇p(n ) = μΔu + (μ + ν)∇divu − n (E + ηu × B ), (1.2)
1 1
∂t E  − ∇ × B  = n u , ∂t B  + ∇ × E  = 0, (1.3)
η η
εdivE  = 1 − n , divB  = 0, (1.4)

with initial data

(n , u , E  , B  )(t = 0) = (n0 , u0 , E0 , B0 ), (1.5)
for x ∈ Ω and t > 0. Here Ω is a torus T3 . The non-dimensionless parameters  and η can be chosen
independently on each other, according to the desired scaling [2]. Physically,  and η are proportional
1
to the Debye length and 1c , respectively, where c = (0 ν0 )− 2 is the speed of light, with 0 and ν0 being
the vacuum permittivity and permeability. Thus, the limit  → 0 is called the quasi-neutral limit while

∗ Corresponding author


c Science China Press and Springer-Verlag Berlin Heidelberg 2014 math.scichina.com link.springer.com
2 Yang J W et al. Sci China Math

the limit η → 0 is called the non-relativistic limit. μ and ν are the (scaled) constant viscosity coefficients
satisfying
μ > 0, 2μ + 3ν > 0. (1.6)
The unknowns n , u , E  and B  denote the density, velocity vector of the electric particles in a plasma,
the electric field and magnetic field, respectively. In this paper, we suppose that

p(s) = asγ , s > 0, a > 0, γ > 1. (1.7)

Equations (1.1)–(1.2) are the mass and momentum balances for the electrons, respectively, while equa-
tions (1.3)–(1.4) are the Maxwell system. It is obvious that (1.4) is redundant with (1.3), as soon as they
are satisfied by the initial conditions.
In this paper, we will concentrate on the so-called quasi-neutral regime. Precisely, we consider the
combined quasi-neutral and non-relativistic limit of Navier-Stokes-Maxwell system (1.1)–(1.5) in the
following scaling case,
η =  → 0. (1.8)
This scaling corresponds to a limit, without any magnetic field, coupled with a quasi-neutral limit (the
electron density instantaneously adjust itself to the unit background density). Given a converging se-
quence of solutions (n , u , E  , B  ), we are concerned with what system the limit (n, u, E, B) satisfies. At
the formal level, a necessary condition for the convergence of the Maxwell equations (1.3)–(1.4) is that

∇ × B  → 0,
n → 1 and ∇ × E  → 0,
divB  = 0,

which implies that n = 1, B = 0 and E = ∇p0 (∇ × E = 0 implies the existence of a potential

function p0 such that E = ∇p0 ). Then passing to the limit in the mass conservation equation (1.1),
we get divu → 0. Coming back to the momentum equation (1.2), we conclude that u must satisfy the
incompressible Navier-Stokes equation of ideal fluids:

divu0 = 0, (1.9)
∂t u + u · ∇u = μΔu − ∇p .
0 0 0 0 0
(1.10)

The main purpose of this paper is to prove rigorously the above limit for smooth solutions to the Navier-
Stokes-Maxwell system (1.1)–(1.4) on time intervals, on which a smooth solution to the incompressible
Navier-Stokes equations of ideal fluids exists.
It is well known that the asymptotic limit is a challenging and physically very complex modeling
problem for fluid dynamics and kinetic models for semiconductors and plasmas. The quasi-neutral limit
has been performed in Vlasov-Poisson system by Brenier [1], Grenier [5] and Masmoudi [17], in drift-
diffusion system by Gasser et al. [6,7], Wang et al. [24], in Euler-Poisson system by Cordier and Grenier [4],
Wang [22], Li [15], Peng and Wang [21] and in Navier-Stokes-Poisson equations by Ju et al. [11] and in
viscous quantum hydrodynamic model by Li [13]. Moreover, we note that Jiang et al. rigorously proved
that the quasi-neutral limit of the multi-dimensional two-fluid Euler-Poisson system is compressible Euler
equations in [10] recently.
Because of the complicated coupling of the Lorentz force, the Euler-Maxwell equations are more intri-
cate than the Euler-Poisson equations. Then, there have been less study on the Euler-Maxwell equations
and their asymptotic analysis than that on the Euler-Poisson equations. The first mathematical study of
the Euler-Maxwell equations with extra relaxation terms is due to Chen et al. [3], where a global weak so-
lution is proved in one spatial dimension. Since then, little progress has been made on the Euler-Maxwell
equations. Jerome [9] established a local smooth solution theory for the Cauchy problem of compressible
Hydrodynamic-Maxwell systems via a modification of the classical semigroup-resolvent approach of Kato.
Recently, the convergence of one-fluid (isentropic) Euler-Maxwell system to an incompressible e-MHD
system was proven by Peng and Wang in [19] via the non-relativistic limit. Peng and Wang also proved
 Yang J W et al. Sci China Math 3

that the combined non-relativistic and quasi-neutral limit is the (isentropic) incompressible Euler equa-
tion in a uniform background of non-moving ions with fixed unit density [20]. Yang and Wang [25, 26]
studied the asymptotic limit of the two-fluid Euler-Maxwell system and non-isentropic Euler-Maxwell
system via a non-relativistic regime.
To the authors’ best knowledge, there are few results on quasi-neutral limit of the Navier-Stokes-
Maxwell equations (1.1)–(1.4). The objective of this paper is to make this limit rigorous. In contrast with
the results in [8, 14, 20, 23], where the limiting equations are incompressible Euler equations, our limiting
equations are the incompressible Navier-Stokes equations (1.9)–(1.10). First, we must deal with viscosity
terms carefully. Second, in order to obtain the uniform estimate of magnetic field, we must introduce a
correction term ηb0 . Finally, we use the refined energy analysis to obtain the desired convergence results.
The rest of this paper is organized as follows. In the next section, we state some useful known results
and our main result. Finally, Section 3 is devoted to the proof of our main theorem.

2 Main results
Before stating our results, we give some notation used in the sequel. We denote by C, Ci (i = 1, 2, . . .)
and CT various positive constants independent of (η). Moreover, CT may depend on T . In this paper
 
we denote f dx = T3 f dx for convenience.
It is well known that the incompressible Navier-Stokes equations (1.9)–(1.10) enjoy a unique local
smooth solution (see [12, 18]), i.e.,
Proposition 2.1. Assume that the initial vector u0 (x, 0) = u00 (x) satisfies
3
u00 (x) ∈ H s , s> +3 and div u00 = 0.
2
Then there exists T the maximal existence time, such that the incompressible Navier-Stokes equations
(1.9)–(1.10) with the initial data enjoy a unique solution u0 ∈ L∞ (0, T ; H s ) satisfying, for any 0 < T0 <
T , div u0 = 0 and

sup (u0 H s + ∂t u0 H s−2 + ∇p0 H s−2 + ∂t ∇p0 H s−3 )  CT0 , (2.1)
0tT0

for some positive constant C(T0 ), depending only on T0 .

In [19,20], Peng and Wang obtained the local smooth solution to the problem (1.1)–(1.4) with μ = ν = 0
for the well-prepared initial data (n0 , u0 , E0 , B0 ) ∈ H s (T3 ) with
3
s> +2 and min n0  δ > 0.
2 x∈T3

And they also obtained that the local existence time interval is uniform in  by using the standard
continuous induction method. It is a straightforward modification of their arguments to obtain the local
smooth solution to the problem (1.1)–(1.4) with μ > 0 and 2μ + 3ν > 0. We omit the details here.
The main result of this paper reads as follows.
Theorem 2.2. Let η =   1. Assume that the initial data (n0 , u0 , E0 , B0 ) ∈ H s with
3
s> +3 and min n0 (x) > 0
2 x∈T3

satisfying the following compatibility conditions,

div E0 = 1 − n0 , min n0  δ > 0, div B0 = 0. (2.2)

x∈T3

Assume further that they satisfy

  
1   a  γ    1 
n |u − u0 | +
0 2
((n ) − 1 − γ(n0 − 1)) + |E0 − ∇p0 | + |B0 − ηb0 | dx  C,
0 2 0 2
(2.3)
2 0 0 γ−1 0 2 2
4 Yang J W et al. Sci China Math

where the function b0 satisfies

−∇ × b0 = u0 and div b0 = 0.

Let (n , u , E  , B  ) be the smooth solution to the problem (1.1)–(1.5) satisfying (2.2)–(2.3) and let (n0 , p0 )
be the classical solution to (1.9)–(1.10), both on the time interval [0, T ]. Then, for all t ∈ [0, T ], it holds
that
  
1   a  γ    1 
n |u − u | + 0 2
((n ) − 1 − γ(n − 1)) + |E − ∇p | + |B − ηb | dx  CT , (2.4)
0 2 0 2
2 γ−1 2 2

where CT > 0 is a constant independent of .

Remark 2.3. It should be pointed out that in Theorem 2.2, we need to assume that  and η go to
zero simultaneously. For simplicity, we take η = .
Remark 2.4. In order to obtain the uniform estimate of magnetic field, we introduce a correction
term ηb0 (see equation (3.6)). In fact, using Maxwell equation (1.3), we have that
   
−η b · ∂t B dx = (∇ × E ) · b dx = (∇ × b ) · E dx = − u0 · E  dx,
0   0 0 
(2.5)

which shows the interaction between electric field and magnetic field.

3 Proof of Theorem 2.2

In this section we shall give the proofs of our results by the modulated energy method which helps us
obtain a stability inequality.
Now we first define the total energy E(t) to the problem (1.1)–(1.4):
   
1 a  1
E(t) = n |u |2 dx + ((n )γ − 1 − γ(n − 1))dx + |E  |2 dx + |B  |2 dx
2 γ−1 2 2
=: E1 (t) + E2 (t) + E3 (t) + E4 (t). (3.1)

Then we have the following estimate on E(t).

Lemma 3.1. For the smooth solution to the problem (1.1)–(1.4), we have
 
d
E(t) = −μ |∇u | dx − (μ + ν) |div u |2 dx  0.
 2
(3.2)
dt

Moreover, the following inequality holds,

E(t)  E(0). (3.3)

Proof. Taking the time derivatives of Ei (t) (i = 1, 2, 3, 4) directly, and using equations (1.1)–(1.4) and
the periodic boundary condition, we obtain that
   
d
E1 (t) = − u · ∇p(n )dx − n u · E  dx − μ |∇u |2 dx − (μ + ν) |divu |2 dx,
dt

d
E2 (t) = u · ∇p(n )dx,
dt
 
d 1
E3 (t) = ∇ × B  · E  dx + n u · E  dx,
dt η

d 1
E4 (t) = − ∇ × E  · B  dx.
dt η

Summing up the above equations, we have derived (3.2). Here, we have used the vector analysis formula
div(f × g) = (∇ × f ) · g − (∇ × g) · f.
 Yang J W et al. Sci China Math 5

From (1.9)–(1.10) we find the energy identity of the impressible Navier-Stokes equations:
 
1 d
|u | dx + μ
0 2
|∇u0 |2 dx = 0. (3.4)
2 dt T3 T3

Next, we define the modulated energy functional H(t) as follows:

  
1   a  1
H(t) = n |u − u0 |2 + ((n )γ − 1 − γ(n − 1)) + |E  − ∇p0 |2 + |B  − ηb0 |2 dx. (3.5)
2 γ −1 2 2

By using equations (1.1)–(1.4), (2.5) and (3.2), we obtain that

  
d d 1 
H(t) = E(t) − ∂t (n u ) · u0 dx + ∂t (n |u0 |2 )dx − ∂t (2E  · ∇p0 − |∇p0 |2 )dx
dt dt 2 2
 
− η b0 · ∂t B  dx − η (B  − ηb0 ) · ∂t b0 dx
   
= −μ |∇u |2 dx − μ |∇u0 |2 dx − (μ + ν) |divu |2 dx − ∂t (n u ) · u0 dx
  
1 1
− (n u ) · ∂t u0 dx + ∂t n |u0 |2 dx + (n − 1)∂t |u0 |2 dx
2 2
 
1 d
+ |u | dx + μ |∇u0 |2 dx
0 2
2 dt
 
=0 from equation (3.4)
 
−  ∂t E · ∇p dx −  (E  − ∇p0 ) · ∂t ∇p0 dx
 0

 
+ ∇ × E  · b0 dx − η (B  − ηb0 ) · ∂t b dx
   
= −μ |∇u | dx − μ |∇u | dx − (μ + ν) |divu | dx − ∂t (n u ) · u0 dx
 2 0 2  2

  
1 
− (n u ) · ∂t u0 dx + ∂t n |u0 |2 dx + E  ∇ · (∂t |u0 |2 )dx
2 2
 
−  ∂t E  · ∇p0 dx −  (E  − ∇p0 ) · ∂t ∇p0 dx
 
− E  · u0 dx − η (B  − ηb0 ) · ∂t b dx
   5
= −μ |∇u |2 dx − μ |∇u0 |2 dx − (μ + ν) |divu |2 dx + Ii
i=1
  
 
− (E − ∇p ) · ∂t ∇p dx −
0 0
E · u dx − η
0
(B  − ηb0 ) · ∂t b0 dx, (3.6)

where the integrals

 
I1 = − ∂t (n u ) · u0 dx, I2 = − (n u ) · ∂t u0 dx,
  
1 
I3 = ∂t n |u0 |2 dx, I4 = E  ∇ · (∂t |u0 |2 )dx, I5 = − ∂t E  · ∇p0 dx
2 2
will be treated separately below. First, using div u0 = 0 and integrating by parts, we have that
  
I1 = u0 · div(n u ⊗ u )dx + u0 · ∇p(n )dx − μ u0 · Δu dx
  
− (μ + ν) u · ∇divu dx + u · (n E )dx + η u0 · (n u × B  )dx
0  0  

   
= u0 · div(n u ⊗ u )dx − μ u · Δu0 dx + u0 · (n E  )dx + η u0 · (n u × B  )dx
6 Yang J W et al. Sci China Math


=− (n u ⊗ u ) : ∇u0 dx + W,

where, W is defined as
  
W = −μ u · Δu0 dx + u0 · (n E  )dx + η u0 · (n u × B  )dx,

and the symbol “ : ”means the summation over both matrix indices. By a simple computation, we have
that
  
− (n u ⊗ u ) : ∇u0 dx = − n (u − u0 ) ⊗ (u − u0 ) : ∇u0 dx − (n u ⊗ u0 ) : ∇u0 dx
 
+ (n u ⊗ u ) : ∇u dx − (n u0 ⊗ u ) : ∇u0 dx
 0 0 0

 
1
= − n (u − u0 ) ⊗ (u − u0 )dx + div(n u )|u0 |2 dx
2
 
+ n u · ((u · ∇)u )dx − n u · ((u0 · ∇)u0 )dx
 0 0 0


= − n (u − u0 ) ⊗ (u − u0 )dx − I3
 
+ n u0 · ((u0 · ∇)u0 )dx − n u · ((u0 · ∇)u0 )dx.

Therefore, we obtain that

 
I1 + I3 = − n (u − u0 ) ⊗ (u − u0 )dx + n u0 · ((u0 · ∇)u0 )dx

− n u · ((u0 · ∇)u0 )dx + W. (3.7)

Using (1.10), we have that

  
I2 = −μ n u · Δu0 dx + n u · ((u0 · ∇)u0 )dx + n u · ∇p0 dx. (3.8)

For I4 , by integration by parts, we have that

 
 
I4 = (E  − ∇p0 ) · ∇(∂t |u0 |2 )dx + ∇p0 · ∇(∂t |u0 |2 )dx. (3.9)
2 2
For I5 , it is easy to find that
   
I5 =  ∂t divE  · p0 dx = − ∂t n · p0 dx = div(n u ) · p0 dx = − (n u ) · ∇p0 dx. (3.10)

So, from the above equations, we get

5  
  
Ii = − n (u − u ) ⊗ (u − u ) : ∇u dx +
0 0 0
n u0 · ((u0 · ∇)u0 )dx
i=1
 

+ (E  − ∇p0 ) · ∇(∂t |u0 |2 )dx + μ n u · Δu0 dx
2


+ ∇p0 · ∇(∂t |u0 |2 )dx + W. (3.11)
2
Therefore, we get that
  
d
H(t) = −μ |∇u |2 dx − μ |∇u0 |2 dx − (μ + ν) |divu |2 dx
dt
 Yang J W et al. Sci China Math 7
 
− n (u − u0 ) ⊗ (u − u0 ) : ∇u0 dx + n u0 · ((u0 · ∇)u0 )dx
 

+ (E − ∇p ) · ∇(∂t |u | )dx − μ n u · Δu0 dx
 0 0 2
2
  

+ ∇p0 · ∇(∂t |u0 |2 )dx − μ u · Δu0 dx + (n − 1)E  · u0 dx
2
  
+ η u · (n u × B )dx −  (E − ∇p ) · ∂t ∇p dx − η (B  − ηb0 ) · ∂t b0 dx
0     0 0

 
= −μ |∇u − ∇u0 |2 dx − (μ + ν) |divu |2 dx
 
− n (u − u ) ⊗ (u − u ) : ∇u dx + n u0 · ((u0 · ∇)u0 )dx
  0  0 0

 

+ (E  − ∇p0 ) · ∇(∂t |u0 |2 )dx − μ (n − 1)u · Δu0 dx
2
 

+ ∇p · ∇(∂t |u | )dx + (n − 1)E  · u0 dx
0 0 2
2
  
+ η u0 · (n u × B  )dx −  (E  − ∇p0 ) · ∂t ∇p0 dx − η (B  − ηb0 ) · ∂t b0 dx
  9
=: −μ |∇u − ∇u0 |2 dx − (μ + ν) |divu |2 dx + Jj , (3.12)
j=1

where
 
J1 = − n (u − u0 ) ⊗ (u − u0 ) : ∇u0 dx,
J2 = n u0 · ((u0 · ∇)u0 )dx,
 

J3 = (E − ∇p ) · ∇(∂t |u | )dx, J4 = −μ (n − 1)u · Δu0 dx,
 0 0 2
2
  

J5 = ∇p0 · ∇(∂t |u0 |2 )dx, J6 = (n − 1)E  · u0 dx, J7 = η u0 · (n u × B  )dx,
2
 
J8 = − (E − ∇p ) · ∂t ∇p dx, J9 = −η (B  − ηb0 ) · ∂t b0 dx.
 0 0

Now we begin to treat Jj term by term in (3.12). Using inequality (2.1), the first integral can be
bounded by the modulated energy:

J1  ∇u0 L∞ n |u − u0 |2  CH(t). (3.13)

Using div u0 = 0, Cauchy-Schwartz’s inequality and inequality (2.1), we have that


J2 = (n − 1)u0 · ((u0 · ∇)u0 )dx
 

=  div(E − ∇p )u · ((u · ∇)u )dx +  Δp0 u0 · ((u0 · ∇)u0 )dx
0 0 0 0


  |E  − ∇p0 |2 dx + C

 CH(t) + C, (3.14)

where, we have used

  
1
u · ((u · ∇)u )dx = (u · ∇)u · u dx = −
0 0 0 0 0 0
div u0 |u0 |2 dx = 0.
2
For J3 , with the help of the Cauchy-Schwartz’s inequality we obtain

J3  − |E  − ∇p0 |2 dx + C  CH(t) + C. (3.15)
8 Yang J W et al. Sci China Math

From the Maxwell equation (1.4) and the inequality (2.1), we get
 
J4 = −μ (E  − ∇p0 )∇(u · Δu0 )dx − μ Δp0 u · Δu0 dx

 − |E  − ∇p0 |2 dx + C

 CH(t) + C, (3.16)

where we have used the uniform bounds on ∇u due to the energy inequality (3.2) and the bounds on u
which can be deduced as in [20]. A direct computation yields

J5  C. (3.17)

Applying the Cauchy-Schwartz’s inequality and integration by parts, we get that

 
J6 = (n − 1)(E  − ∇p0 ) · u0 dx − (n − 1)∇p0 · u0 dx
 
= − (E  − ∇p0 )div(E  − ∇p0 ) · u0 dx +  div(E  − ∇p0 )∇p0 · u0 dx

+  Δp0 ∇p0 · u0 dx

 ∇u0 L∞ |E  − ∇p0 |2 dx + C

 CH(t) + C. (3.18)

For J7 , we have that


J7 = η u0 · (n u × B  )dx

= η u0 · (n (u − u0 ) × B  )dx
 
  
= η u · (n (u − u ) × (B − ηb ))dx − η
0 0 0 2
u0 · (n (u − u0 ) × b0 )dx
  
    
 ηu L
0
∞ |n (u − u ) × (B − ηb )|dx + η |n (u − u ) × b )|dx
0 0 0 0

  
  
 Cη n |u − u | + |B − ηb | dx + Cη
0 2 0 2

 CH(t) + C (η = ). (3.19)

Using Cauchy-Schwartz’s inequality and inequality (2.1), we have that


J8   |E  − ∇p0 |2 dx + C  CH(t) + C. (3.20)

With the help of Cauchy-Schwartz’s inequality, we have that


J9  |B  − ηb0 |2 dx + Cη 2  CH(t) + C2 . (3.21)

Putting (3.13)–(3.21) together, pulling them into (3.12), and using the inequality (2.3), we obtain
 
d
H(t) + μ |∇u − ∇u0 |2 dx + (μ + ν) |divu |2 dx  CH(t) + C. (3.22)
dt
The Gronwall’s inequality and well-prepared initial data (2.4) imply that

H(t)  CT , ∀ t ∈ [0, T ], (3.23)

with CT > 0 a constant, which completes the proof. 2

 Yang J W et al. Sci China Math 9

Remark 3.2 (Viscosity vanishing limit). From (3.22) and (3.23), we can deduce that
 t  t

H(t) + μ |∇u − ∇u | dx + (μ + ν)
0 2
|div u |2 dx  C, ∀ t ∈ [0, T ],
0 0

which implies the estimate in Theorem 2.2 is uniform with respect to μ. Therefore Theorem 2.2 is a
stability result not only with respect to  and η but also with respect to μ if 0 < 2μ + 3ν  cμ for some
fixed constant c. In other words, we can also obtain a viscosity vanishing limit result from Navier-Stokes-
Maxwell system to the compressible Euler-Maxwell system. Moreover, we also can prove the convergence
of the incompressible Navier-Stokes equation to the incompressible Euler equations in a similar way. Thus,
the viscosity vanishing limit μ → 0 and the combined quasi-neutral and non-relativistic limit  = η → 0
can commute each other. In fact, we can show that the combined quasi-neutral, non-relativistic and
vanishing viscosity limit of the Navier-Stokes-Maxwell system (1.1)–(1.4) is the incompressible Euler
equation.

Acknowledgements This work was supported by the Joint Funds of National Natural Science Foundation of
China (Grant No. U1204103), China Postdoctoral Science Foundation Funded Project (Grant No. 2013M530032)
and the Science and Technology Research Projects of Education Department of Henan Province (Grant No.
13A110731). The authors are very grateful to both referees for their constructive comments and helpful sugges-
tions, which considerably improved the presentation of the paper.

References
1 Brenier Y. Convergence of the Vlasov-Poisson system to the incompressible Euler equations. Comm Partial Differential
Equations, 2000, 25: 737–754
2 Brenier Y, Mauser N J, Puel M. Incompressible Euler and e-MHD as scaling limits of the Vlasov-Maxwell system.
Commun Math Sci, 2003, 1: 437–447
3 Chen G Q, Jerome J W, Wang D H. Compressible Euler-Maxwell equations. Transport Theory Statist Phys, 2000, 29:
311–331
4 Cordier S, Grenier E. Quasineutral limit of an Euler-Poisson system arising from plasma physics. Comm Partial
Differential Equations, 2000, 25: 1099–1113
5 Grenier E. Oscillations in quasineutral plasmas. Comm Partial Differential Equations, 1996, 21: 363–394
6 Gasser I, Hsiao L, Markowich P A, et al. Quasineutral limit of a nonlinear drift-diffusion model for semiconductor
models. J Math Anal Appl, 2002, 268: 184–199
7 Gasser I, Levermore C D, Markowich P, et al. The initial time layer problem and the quasineutral limit in the
semiconductor drift-diffusion model. European J Appl Math, 2001, 12: 497–512
8 Hsiao L, Li F C, Wang S. Convergence of the Vlasov-Poisson-Fokker-Planck system to the incompressible Euler
equations. Sci China Ser A, 2006, 49: 255–266
9 Jerome J W. The Cauchy problem for compressible hydrodynamic-Maxwell systems: A local theory for smooth solu-
tions. Differential Integral Equations, 2003, 16: 1345–1368
10 Jiang S, Ju Q C, Li H L, et al. Quasi-neutral limit of the full bipolar Euler-Poisson system. Sci China Math, 2010,
53: 3099–3114
11 Ju Q C, Li F C, Li H L. The quasineutral limit of compressible Navier-Stokes-Poisson system with heat conductivity
and general initial data. J Differential Equations, 2009, 247: 203–224
12 Kato T. Nonstationary flow of viscous and ideal fluids in R3 . J Funct Anal, 1972, 9: 296–305
13 Li F C. Quasineutral limit of the viscous quantum hydrodynamic model for semiconductors. J Math Anal Appl, 2009,
352: 620–628
14 Li H L, Lin C K. Zero Debye length asymptotic of the quantum hydrodynamic model for semiconductors. Comm Math
Phys, 2005, 256: 195–212
15 Li Y. Convergence of the non-isentropic Euler-Poisson equations to incompressible type Euler equations. J Math Anal
Appl, 2008, 342: 1107–1125
16 Lions P L, Masmoudi N. Incompressible limit for a viscous compressible fluid. J Math Pure Appl, 1998, 77: 585–627
17 Masmoudi N. From Vlasov-Poisson system to the incompressible Euler system. Comm Partial Differential Equations,
2001, 26: 1913–1928
18 McGrath F J. Nonstationary plane flow of viscous and ideal fluids. Arch Ration Mech Anal, 1968, 27: 229–348
10 Yang J W et al. Sci China Math

19 Peng Y J, Wang S. Rigorus derivation of incompressible e-MHD equations from compressible Euler-Maxwell equations.
SIAM J Math Anal, 2008, 40: 540–565
20 Peng Y J, Wang S. Convergence of compressible Euler-Maxwell equations to incompressible Eule equations. Comm
Partial Differential Equations, 2008, 33: 349–376
21 Peng Y J, Wang Y G. Convergence of compressible Euler-Poisson equations to incompressible type Euler equations.
Asymptotic Anal, 2005, 41: 141–160
22 Wang S. Quasineutral limit of Euler-Poisson system with and without viscosity. Comm Partial Differential Equations,
2004, 29: 419–456
23 Wang S, Jiang S. The convergence of the Navier-Stokes-Poisson system to the incompressible Euler equations. Comm
Partial Differential Equations, 2006, 31: 571–591
24 Wang S, Xin Z P, Markowich P A. Quasi-neutral limit of the drift-diffusion models for semiconductors: The case of
general sign-changing doping profile. SIAM J Math Anal, 2006, 37: 1854–1889
25 Yang J W, Wang S. Convergence of the nonisentropic Euler-Maxwell equations to compressible Euler-Poisson equations.
J Math Phys, 2009, 50: 123508
26 Yang J W, Wang S. The non-relativistic limit of Euler-Maxwell equations for two-fluid plasma. Nonlinear Anal, 2010,
72: 1829–1840