Startseite On regular solutions to compressible radiation hydrodynamic equations with far field vacuum
Artikel Open Access

On regular solutions to compressible radiation hydrodynamic equations with far field vacuum

  • Hao Li und Shengguo Zhu EMAIL logo
Veröffentlicht/Copyright: 20. August 2022

Abstract

The Cauchy problem for three-dimensional (3D) isentropic compressible radiation hydrodynamic equations is considered. When both shear and bulk viscosity coefficients depend on the mass density ρ in a power law ρ δ (with 0 < δ < 1 ), based on some elaborate analysis of this system’s intrinsic singular structures, we establish the local-in-time well-posedness of regular solution with arbitrarily large initial data and far field vacuum in some inhomogeneous Sobolev spaces by introducing some new variables and initial compatibility conditions. Note that due to the appearance of the vacuum, the momentum equations are degenerate both in the time evolution and viscous stress tensor, which, along with the strong coupling between the fluid and the radiation field, make the study on corresponding well-posedness challenging. For proving the existence, we first introduce an enlarged reformulated structure by considering some new variables, which can transfer the degeneracies of the radiation hydrodynamic equations to the possible singularities of some special source terms, and then carry out some singularly weighted energy estimates carefully designed for this reformulated system.

MSC 2010: 35A01; 35A09; 35Q35; 35M11

1 Introduction

It is well known that the radiation effects become remarkable in some regime when the temperature is high. Radiation sometimes contributes largely to energy density, momentum density, and pressure, for instance, in astrophysics and inertial confinement fusion. Radiation transfer is usually the most effective mechanism that affects the energy exchange in fluids, so it is necessary to take effects of the radiation field into consideration in the classical hydrodynamic framework. The equations of radiation hydrodynamics result from the balances of particles, momentum, and energy. From a microscopic point of view, the radiation field is composed of photons. We first introduce some basic concepts necessary for describing the radiation field and its interaction with matter. At any time t , we need 2 d variables to specify the state of a photon in phase space, namely, d position variables and d velocity (or momentum) variables. Usually, we can denote by x the d position variables, and replace the d momentum variables equivalently with frequency ν and the travel direction Ω of the photon. Via these variables, we then define the phase-space distribution function f = f ( t , x , ν , Ω ) for photons such that

d n = f ( t , x , ν , Ω ) d x d ν d Ω ,

where n is the number of photons; d n is the number of photons (at time t ) at space point x in a volume element d x , with local frequency ν in a frequency interval d ν , and traveling in a direction Ω in the cubic angel element d Ω . In the radiation transport, we usually use the specific radiation intensity I = I ( t , x , ν , Ω ) to replace the distribution function f . The specific radiation intensity is defined as follows:

I ( t , x , ν , Ω ) = c h ν f ( t , x , ν , Ω ) ,

where c is the vacuum speed of light and h is Planck constant. The physical interpretation of I is contained in the relationship

d E 1 = I ( t , x , ν , Ω ) cos Θ d Σ d ν d Ω d t ,

where d E 1 is the amount of the radiation energy in d ν centered at ν , travelling in a direction Ω confined to a solid angle element d Ω , which crosses, in a time element d t , an area d Σ oriented such that Θ is the angel, which the direction Ω makes with the normal to d Σ ( cos Θ = Ω n , and here n is the outward unit normal vector of d Σ ).

Regarding the three basic interactions between photons and matter, namely, absorption, scattering, and emission, we have transport equation in the general form (see [23]):

(1.1) 1 c I t + Ω I = A r ,

where A r is a collision term given by

(1.2) A r = σ e σ a I + 0 S 2 ν ν σ s I σ s I d Ω d ν ,

I = I ( t , x , ν , Ω ) , I = I ( t , x , ν , Ω ) , t 0 is the time, x R 3 is the Eulerian spatial coordinate, ν , ν 0 are the frequency of photons, Ω , Ω S 2 are the travel direction of photons, and S 2 R 3 denotes the unit sphere in R 3 . σ e ( t , x , ν , Ω , ρ ) 0 is the rate of energy emission due to spontaneous process, and σ a ( t , x , ν , Ω , ρ ) 0 denotes the absorption coefficient that may depend on the mass density ρ .

Similarly to absorption, a photon can undergo scattering interactions with matter, and the scattering interaction serves to change the photon’s characteristics ν and Ω to a new set of characteristics ν and Ω . To quantitatively describe the scattering event, one requires a probabilistic statement concerning this change, which leads to the definition of the “differential scattering coefficient" σ s ( ν ν , Ω Ω ) σ s ( ν ν , Ω Ω , ρ ) that may depend on ρ such that the probability of a photon being scattered from ν to ν contained in d ν , from Ω to Ω contained in d Ω , and traveling a distance d s is given by σ s ( ν ν , Ω Ω , ρ ) d ν d Ω d s . Therefore, the time rates of outscattering and inscattering within a unit volume element are expressed as follows:

outscattering = 0 S 2 σ s ( ν ν , Ω Ω , ρ ) I d Ω d ν , inscattering = 0 S 2 σ s ( ν ν , Ω Ω , ρ ) I d Ω d ν ,

where σ s = O ( ρ ) .

Concerning the effect of radiation on the dynamic properties of the fluid is very significant, we introduce the following two quantities to describe this effect:

(1.3) F r = 0 S 2 I ( t , x , ν , Ω ) Ω d Ω d ν , P r = 0 S 2 I ( t , x , ν , Ω ) Ω Ω d Ω d ν ,

which are called the radiation flux and the radiation pressure tensor, respectively.

Now we take radiation effect into consideration for viscous (barotropic) fluids to have the following isentropic radiation hydrodynamics equations in 3D space:

(1.4) 1 c I t + Ω I = A r , ρ t + div ( ρ u ) = 0 , ρ u + 1 c 2 F r t + div ( ρ u u + P r ) + P m = div T ,

where A r is defined by (1.2). The unknown functions ρ , u = ( u ( 1 ) , u ( 2 ) , u ( 3 ) ) represent the density and the velocity, respectively. P m is the material pressure with the following equation of state for the polytropic fluid:

(1.5) P m = A ρ γ , A > 0 , γ > 1 ,

where A is a constant and γ is the adiabatic exponent. T denotes the viscous stress tensor with the form

(1.6) T = 2 μ ( ρ ) D ( u ) + λ ( ρ ) div u I 3 ,

where D ( u ) = 1 2 ( u + ( u ) ) is the deformation tensor and I 3 is the 3 × 3 identity matrix,

(1.7) μ ( ρ ) = α ρ δ , λ ( ρ ) = β ρ δ ,

for some constant δ 0 , μ ( ρ ) is the shear viscosity coefficient, λ ( ρ ) + 2 3 μ ( ρ ) is the bulk viscosity coefficient, and ( α , β ) are both constants satisfying

(1.8) α > 0 , 2 α + 3 β 0 .

According to (1.1) and (1.3), system (1.4) can be rewritten as follows:

(1.9) 1 c I t + Ω I = A r , ρ t + div ( ρ u ) = 0 , ( ρ u ) t + div ( ρ u u ) + P m = div T 1 c 0 S 2 A r Ω d Ω d ν .

In the current paper, we are concerned with the local-in-time well-posedness of regular solution to the Cauchy problem (1.9) with the following initial data and far field behavior:

(1.10) ( I , ρ , u ) t = 0 = ( I 0 ( x , ν , Ω ) , ρ 0 ( x ) , u 0 ( x ) ) for ( x , ν , Ω ) R 3 × R + × S 2 , ( I , ρ , u ) ( 0 , 0 , 0 ) as x for t 0 , ( ν , Ω ) R + × S 2 .

Throughout this paper, we will adopt the following simplified notations, and most of them are for the homogeneous and inhomogeneous Sobolev spaces:

f p = f L p ( R 3 ) , f m , r = f W m , r ( R 3 ) , f s = f H s ( R 3 ) , D k , r = { f L loc 1 ( R 3 ) : f D k , r = f r < } , D k = D k , 2 , D 1 = { f L 6 ( R 3 ) : f D 1 = f 2 < } , f D k , r = f D k , r ( R 3 ) , f X 1 X 2 = f X 1 + f X 2 , f d x = R 3 f d x , X ( [ 0 , T ] ; Y ) = X ( [ 0 , T ] ; Y ( R 3 ) ) , ( f , g ) X = f X + g X .

A detailed study of homogeneous Sobolev space can be found in the study by Galdi [9].

When radiation effects are not considered, there is a lot of literature on the well-posedness of strong/classical solutions for the isentropic compressible Navier-Stokes equations. For the constant viscous flow (i.e., δ = 0 in (1.7)), when inf x ρ 0 ( x ) > 0 , the local well-posedness of classical solutions to the Cauchy problem follows from the standard symmetric hyperbolic-parabolic structure satisfying the well-known Kawashima’s condition, cf. [13,22,27], which has been extended to be a global one by Matsumura-Nishida [21] for initial data close to a nonvacuum equilibrium in some Sobolev space H s ( R 3 ) s > 5 2 . However, these approaches do not work when inf x ρ 0 ( x ) = 0 , which occurs when some physical requirements are imposed, such as finite total initial mass and energy in the whole space. One of the main issues in the presence of vacuum is the degeneracy of the time evolution operator, which makes it hard to understand the behavior of the velocity field near the vacuum. Via imposing some initial compatibility conditions, Cho et al. [4] established the local well-posedness of strong solutions with vacuum, which, recently, has been shown to be a global one with small energy by Huang et al. [11]. We also refer to Lions [19], and Feireisl et al. [8] to readers and the references therein for the existence theory of global weak solutions with finite energy.

For degenerate viscous flow (i.e., δ > 0 in (1.7)) without considering radiation, when inf x ρ 0 ( x ) = 0 , instead of the uniform elliptic structure in the constant viscous flow, the viscosity degenerates when density vanishes, which raises the difficulty of the problem to another level. Recently, some significant progress has been made on the well-posedness of smooth solutions in 3D space. In the study by Li et al. [16], via introducing a “quasi-symmetric hyperbolic”–“degenerate elliptic” coupled structure to control the behavior of the velocity u near the vacuum, they showed that the unique 3D regular solution with vacuum exists locally in time, which has been extended to be a global one with small density but possibly large velocity by Xin and Zhu [31]. Recently, for the case 0 < δ < 1 , by using an elaborate elliptic approach on the operators L ( ρ δ 1 u ) and some initial compatibility conditions, the existence of 3D local regular solution with far field vacuum has also been obtained by Xin and Zhu [32]. Some other interesting results can also be found in [3,5,10].

When the effect of radiation is taken into consideration, studying the radiation hydrodynamics equations becomes more complicated for both inviscid and viscous fluids because of the complexity and mathematical difficulty. For Euler-Boltzmann equations of inviscid fluids, the local well-posedness and finite time blow up of classical solutions away from the vacuum were studied by Jiang and Zhong [33] in multi-dimensional (M-D) space, and see also the study by Jiang and Wang [12] for a simplified 3D isentropic model. Later, Li and Zhu [18] considered the system discussed in [12] and proved the local existence of a unique regular solution with vacuum by means of the theory of quasi-linear symmetric hyperbolic systems and some technique tools. They also showed that the regular solution will blow up if the initial density vanishes in some local domain. For Navier-Stokes-Boltzmann equations of viscous fluids, Ducomet and Nečasová [7] showed that the global existence theory for the compressible viscous fluids developed in [21] can be generalized to the radiation hydrodynamics. Li and Zhu [17] established the existence of local strong solutions for the isentropic flow in homogeneous Sobolev space for large initial data satisfying the initial compatibility conditions. Ducomet et al. [6] obtained the existence of global weak solutions for some radiation hydrodynamic system, where the velocity u may develop uncontrolled time oscillations on the hypothetical vacuum zones. Some other interesting results for the related radiation hydrodynamics models can also be seen in [2,24,25] and so on.

Our goal in the present paper is to show the local existence and uniqueness of the 3D regular solution with far field vacuum to the Cauchy problem (1.9)–(1.10) for 0 < δ < 1 . It is worth pointing out that when vacuum appears, one would encounter some difficulties as follows. First, (1.9) is a system of fluid equations coupled with a nonlinear integro-differential hyperbolic equation, which makes the corresponding calculation very complicated. Second, the degeneracy of the time evolution and elliptic operator in momentum equations caused by the vacuum makes it very difficult to determine and control the behavior of velocity u near the vacuum, which is the key point in the vacuum-related problems (see Xin and Zhu [32]). Finally, besides the difficulties mentioned earlier, additional difficulties in the proof lie in dealing with the nonlinear terms and the coupled cross terms between radiation field and fluid field, which prevent us obtaining the uniform a priori estimates. To this end, we shall employ the delicate energy estimates based on the well-designed reformulated structure obtained during the linearization and some physically reasonable assumptions on the radiation quantities.

The rest this paper is organized as follows. In Section 2, we first present some physically reasonable assumptions on the radiation quantities and then conclude this section by stating the main result. Section 3 is devoted to establishing the local-in-time well-posedness of regular solution to the Cauchy problem (1.9)–(1.10). Finally, we give an appendix to list some lemmas that are frequently used in our proof.

2 Hypotheses and main result

In this section, we first make some physical assumptions on the radiation quantities and then state the main result.

2.1 Hypotheses on radiation quantities

The general form of radiation coefficients is usually not known because it is difficult to evaluate these physical coefficients in quantum mechanics. Physically speaking, the radiation coefficients can be written as σ e = ρ σ ¯ e , σ a = ρ σ ¯ a , σ s = ρ σ ¯ s , σ s = ρ σ ¯ s , and then A r can be written as A r = ρ A ¯ r , where

(2.1) A ¯ r = σ ¯ e σ ¯ a I + 0 S 2 ν ν σ ¯ s I σ ¯ s I d Ω d ν .

Next we will make some physically reasonable assumptions on the radiation coefficients σ ¯ e , σ ¯ a , and σ ¯ s , which are similar as in [15,17,30].

H1 (Differential scattering coefficient). Let σ s = ρ σ ¯ s ( ν ν , Ω Ω ) = ρ σ ¯ s and σ s = ρ σ ¯ s ( ν ν , Ω Ω ) = ρ σ ¯ s , with the functions σ ¯ s 0 and σ ¯ s 0 satisfying

(2.2) 0 S 2 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν λ 1 d Ω d ν Λ , 0 S 2 0 S 2 σ ¯ s d Ω d ν λ 2 d Ω d ν + 0 S 2 σ ¯ s d Ω d ν Λ ,

where λ 1 = 1 or 1/2, λ 2 = 1 or 2, and Λ is a fixed constant.

H2 (Emission and absorption coefficient). Let σ e = σ e ( t , x , ν , Ω , ρ ) = ρ σ ¯ e ( t , x , ν , Ω , ρ ) and σ a = σ a ( t , x , ν , Ω , ρ ) = ρ σ ¯ a ( t , x , ν , Ω , ρ ) , where σ ¯ e 0 , σ ¯ a 0 . Due to ϕ = A γ γ 1 ρ γ 1 , for s = 1 , 2 , 3 , we assume

(2.3) σ ¯ e L 1 L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H s ) ) M ( ϕ 3 ) ( 1 + ϕ s ) , ( σ ¯ e ) t L 1 L 2 ( R + × S 2 ; C ( [ 0 , T ] ; L 2 ) ) M ( ϕ 3 ) ( 1 + ϕ t 2 ) , ( σ ¯ e ) t L 1 L 2 ( R + × S 2 ; C ( [ 0 , T ] ; D 1 ) ) M ( ϕ 3 ) ( 1 + ϕ t D 1 ) , ( σ ¯ e ) t t L 1 L 2 ( R + × S 2 ; C ( [ 0 , T ] ; L 2 ) ) M ( ϕ 3 ) ( 1 + ϕ t t 2 ) , σ ¯ a L L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H s ) ) M ( ϕ 3 ) ( 1 + ϕ s ) , ( σ ¯ a ) t L 2 ( R + × S 2 ; C ( [ 0 , T ] ; L 2 ) ) M ( ϕ 3 ) ( 1 + ϕ t 2 ) , ( σ ¯ a ) t L 2 ( R + × S 2 ; C ( [ 0 , T ] ; D 1 ) ) M ( ϕ 3 ) ( 1 + ϕ t D 1 ) , ( σ ¯ a ) t t L 2 ( R + × S 2 ; C ( [ 0 , T ] ; L 2 ) ) M ( ϕ 3 ) ( 1 + ϕ t t 2 ) ,

and for s = 0 , 1 , 2 ,

(2.4) σ ¯ e ( , , , , ρ 1 ) σ ¯ e ( , , , , ρ 2 ) L 1 L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H s ) ) M ( ( ϕ 1 , ϕ 2 ) 3 ) ( 1 + ϕ 1 ϕ 2 s ) , σ ¯ a ( , , , , ρ 1 ) σ ¯ a ( , , , , ρ 2 ) L L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H s ) ) M ( ( ϕ 1 , ϕ 2 ) 3 ) ( 1 + ϕ 1 ϕ 2 s ) ,

where M = M ( ) denotes a strictly increasing function from [ 0 , ) to [ 1 , ) .

Remark 2.1

It follows from (2.3)–(2.4) that for s = 0 , 1 , 2 , 3 ,

(2.5) σ e L 1 L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H s ) ) M ( ϕ 3 ) ϕ 3 1 γ 1 ( 1 + ϕ s ) , σ a L L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H s ) ) M ( ϕ 3 ) ϕ 3 1 γ 1 ( 1 + ϕ s ) ,

and for s = 0 , 1 , 2 ,

(2.6) σ e ( , , , , ρ 1 ) σ e ( , , , , ρ 2 ) L 1 L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H s ) ) M ( ( ϕ 1 , ϕ 2 ) 3 ) ( ϕ 1 , ϕ 2 ) 3 1 γ 1 ( 1 + ϕ 1 ϕ 2 s ) , σ a ( , , , , ρ 1 ) σ a ( , , , , ρ 2 ) L L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H s ) ) M ( ( ϕ 1 , ϕ 2 ) 3 ) ( ϕ 1 , ϕ 2 ) 3 1 γ 1 ( 1 + ϕ 1 ϕ 2 s ) .

Remark 2.2

A general expression of σ a and σ s used for describing Compton Scattering process are given by (see [23])

(2.7) σ a ( t , x , ν , Ω , ρ , θ ) = D 1 ρ θ 1 2 exp D 2 θ 1 2 ν ν 0 ν 0 2 , σ s = σ ¯ s ( t , x , ν ν , Ω Ω ) ρ ,

where ν 0 is the fixed frequency and D i ( i = 1 , 2 ) are positive constants. It is not difficult to see that, together with the Boyle and Gay-Lussac laws for polytropic gas, σ a also satisfies assumptions (2.3)–(2.6).

2.2 Main result

Before stating the main result, let us first introduce the definition of regular solution to the Cauchy problem (1.9)–(1.10).

Definition 2.1

Let T > 0 be a finite constant. The triple ( I , ρ , u ) is called a regular solution to the Cauchy problem (1.9)–(1.10), if ( I , ρ , u ) satisfies this problem in the sense of distributions and

( 1 ) I L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 3 ) ) , I t L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 2 ) ) ; ( 2 ) inf x R 3 ρ ( t , x ) = 0 for 0 t T , 0 < ρ γ 1 C ( [ 0 , T ] ; H 3 ) , ρ δ 1 L ( [ 0 , T ] ; L D 2 ) ; ( 3 ) u C ( [ 0 , T ] ; H 3 ) L 2 ( [ 0 , T ] ; H 4 ) , u t C ( [ 0 , T ] ; H 1 ) L 2 ( [ 0 , T ] ; D 2 ) , ρ δ 1 2 u C ( [ 0 , T ] ; L 2 ) , ρ δ 1 2 u t L ( [ 0 , T ] ; L 2 ) , ρ δ 1 u L ( [ 0 , T ] ; D 1 ) , ρ δ 1 2 u C ( [ 0 , T ] ; H 1 ) L 2 ( [ 0 , T ] ; D 2 ) .

Via denoting σ ¯ e 0 = σ ¯ e ( t = 0 , x , ν , Ω , ρ 0 ) and

A ¯ r 0 = σ ¯ e 0 σ ¯ a ( t = 0 , ν , Ω , ρ 0 ) I 0 + 0 S 2 ν ν σ ¯ s I 0 σ ¯ s I 0 d Ω d ν ,

then our main result can be stated as follows.

Theorem 2.1

Let parameters ( γ , δ , α , β ) satisfy

(2.8) 1 < γ 4 3 , 0 < δ < 1 , α > 0 , 2 α + 3 β 0 .

If the initial data ( I 0 , ρ 0 , u 0 ) satisfies

(2.9) I 0 L 2 ( R + × S 2 ; H 3 ) , ( 0 < ρ 0 γ 1 , u 0 ) H 3 , ρ 0 δ 1 D 1 D 2 , ρ 0 δ 1 2 L 4 ,

and the initial compatibility conditions

(2.10) u 0 = ρ 0 1 δ 2 g 1 , L u 0 = ρ 0 1 δ g 2 , a ρ 0 δ 1 L u 0 + 1 c 0 S 2 A ¯ r 0 Ω d Ω d ν = ρ 0 1 δ 2 g 3 ,

for some g i L 2 ( i = 1 , 2 , 3 ) , then there exist a time T > 0 and a unique regular solution ( I , ρ , u ) in [ 0 , T ] × R 3 × R + × S 2 to the Cauchy problem (1.9)–(1.10), satisfying

(2.11) t 1 2 u L ( [ 0 , T ] ; D 4 ) , t 1 2 u t L ( [ 0 , T ] ; D 2 ) L 2 ( [ 0 , T ] ; D 3 ) , ρ δ 1 2 u t L ( [ 0 , T ] ; L 2 ) , ρ δ 1 2 u t L 2 ( [ 0 , T ] ; L 2 ) , u t t L 2 ( [ 0 , T ] ; L 2 ) , t 1 2 u t t L ( [ 0 , T ] ; L 2 ) L 2 ( [ 0 , T ] ; D 1 ) , ρ 1 δ L ( [ 0 , T ] ; L D 1 , 6 D 2 , 3 D 3 ) , ρ δ 1 C ( [ 0 , T ] ; D 1 D 2 ) , log ρ L ( [ 0 , T ] ; L L 6 D 1 , 3 D 2 ) .

Moreover, ( I , ρ , u ) is also a classical solution to (1.9)–(1.10) for t ( 0 , T ] .

Remark 2.3

The condition γ 4 3 in (2.8) is a technical assumption used to derive the high-order uniform a priori estimates for the radiation intensity (see the proof of Lemma 3.4).

Remark 2.4

One can find the following class of initial data ( I 0 , ρ 0 , u 0 ) satisfying the conditions (2.9)–(2.10):

I 0 L 2 ( R + × S 2 ; C 0 3 ( R 3 ) ) , ρ 0 ( x ) = 1 1 + x 2 ϰ , u 0 C 0 3 ( R 3 ) ,

where 3 4 ( γ 1 ) ϰ 1 4 ( 1 δ ) .

Remark 2.5

As mentioned in the abstract, the compatibility conditions (2.10) are also necessary for the existence of the unique regular solutions obtained in Theorem 2.1. Indeed, the one shown that in the second one of (2.10) 1 (resp. (2.10) 2 ) plays a key role in the derivation of u t L ( [ 0 , T ] ; L 2 ( R 3 ) ) (resp. ρ δ 1 2 u t L ( [ 0 , T ] ; L 2 ( R 3 ) ) ), which will be used in the uniform estimates for u D 2 (resp. u D 3 ).

3 Local-in-time well-posedness

The purpose of this section is to prove Theorem 2.1. We first reformulate the original problems (1.9)–(1.10) to (3.2)–(3.4) by introducing some new variables and then establish the corresponding local existence and uniqueness of classical solutions to (3.2)–(3.4). Finally, we show that the existence result of the problem (3.2)–(3.4) implies Theorem 2.1.

3.1 Reformulation

Via introducing the following new variables:

(3.1) ϕ = A γ γ 1 ρ γ 1 , ψ = δ δ 1 ρ δ 1 = δ δ 1 A γ γ 1 1 δ γ 1 ϕ δ 1 γ 1 = ( ψ ( 1 ) , ψ ( 2 ) , ψ ( 3 ) ) ,

then the Cauchy problem (1.9) and (1.10) can be equivalently rewritten as follows:

(3.2) 1 c I t + Ω I = A r , ϕ t + u ϕ + ( γ 1 ) ϕ div u = 0 , u t + u u + ϕ + a ϕ 2 e L u = ψ Q ( u ) 1 c 0 S 2 A ¯ r Ω d Ω d ν , ψ t + ( u ψ ) + ( δ 1 ) ψ div u + δ a ϕ 2 e div u = 0 , ( I , ϕ , u , ψ ) t = 0 = ( I 0 , ϕ 0 , u 0 , ψ 0 ) = I 0 ( x , ν , Ω ) , A γ γ 1 ρ 0 γ 1 ( x ) , u 0 ( x ) , δ δ 1 ρ 0 δ 1 ( x ) for ( x , ν , Ω ) R 3 × R + × S 2 , ( I , ϕ , u , ψ ) ( 0 , 0 , 0 , 0 ) as x + for t 0 ,

where A r = ρ A ¯ r , A ¯ r is defined by (2.1),

(3.3) a = A γ γ 1 1 δ γ 1 , e = δ 1 2 ( γ 1 ) < 0 ,

and L ( u ) and Q ( u ) are given by

(3.4) L u = α Δ u ( α + β ) div u , Q ( u ) = α ( u + ( u ) ) + β div u I 3 .

To solve the Cauchy problem (1.9) and (1.10) locally in time, we first establish the local well-posedness of classical solution to the problem (3.2)–(3.4).

Theorem 3.1

Let (2.8) hold. Assume the initial data ( I 0 , ϕ 0 , u 0 , ψ 0 ) satisfy

(3.5) I 0 L 2 ( R + × S 2 ; H 3 ) , ( 0 < ϕ 0 , u 0 ) H 3 , ψ 0 D 1 D 2 , ϕ 0 e L 4 ,

and the initial compatibility conditions

(3.6) u 0 = ϕ 0 e g 1 , L u 0 = ϕ 0 2 e g 2 , a ϕ 0 2 e L u 0 + 1 c 0 S 2 A ¯ r 0 Ω d Ω d ν = ϕ 0 e g 3 ,

for some g i L 2 ( i = 1 , 2 , 3 ) , then there exist a time T > 0 and a unique classical solution I , ϕ , u , ψ = a δ δ 1 ϕ 2 e in [ 0 , T ] × R 3 × R + × S 2 to the Cauchy problem (3.2)–(3.4), satisfying

(3.7) I L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 3 ) ) , I t L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 2 ) ) , ϕ C ( [ 0 , T ] ; H 3 ) , log ϕ L ( [ 0 , T ] ; L L 6 D 1 , 3 D 2 ) , ψ C ( [ 0 , T ] ; D 1 D 2 ) , ϕ 2 e L ( [ 0 , T ] ; L D 1 , 6 D 2 , 3 D 3 ) , u C ( [ 0 , T ] ; H 3 ) L 2 ( [ 0 , T ] ; H 4 ) , ϕ 2 e u L ( [ 0 , T ] ; D 1 ) , ϕ 2 e 2 u C ( [ 0 , T ] ; H 1 ) L 2 ( [ 0 , T ] ; D 2 ) , ϕ e u C ( [ 0 , T ] ; L 2 ) , ϕ e u t L ( [ 0 , T ] ; L 2 ) , ϕ 2 e 2 u t L 2 ( [ 0 , T ] ; L 2 ) , u t C ( [ 0 , T ] ; H 1 ) L 2 ( [ 0 , T ] ; D 2 ) , ( ϕ 2 e 2 u ) t L 2 ( [ 0 , T ] ; L 2 ) , u t t L 2 ( [ 0 , T ] ; L 2 ) , t 1 2 u L ( [ 0 , T ] ; D 4 ) , t 1 2 u t L ( [ 0 , T ] ; D 2 ) L 2 ( [ 0 , T ] ; D 3 ) , t 1 2 u t t L ( [ 0 , T ] ; L 2 ) L 2 ( [ 0 , T ] ; D 1 ) .

3.2 Linearization

To prove Theorem 3.1, we begin by considering the following linearized problem:

(3.8) 1 c I t ε , η + Ω I ε , η = A ˜ r ε , η , ϕ t ε , η + v ϕ ε , η + ( γ 1 ) ϕ ε , η div v = 0 , u t ε , η + v v + ϕ ε , η + a ( h ε , η ) 2 + ε 2 L u ε , η = ψ ε , η Q ( v ) 1 c 0 S 2 A ¯ r ε , η Ω d Ω d ν , h t ε , η + v h ε , η + ( δ 1 ) g div v = 0 , ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) t = 0 = ( I 0 η , ϕ 0 η , u 0 η , h 0 η ) = ( I 0 , ϕ 0 + η , u 0 , ( ϕ 0 + η ) 2 e ) for ( x , ν , Ω ) R 3 × R + × S 2 , ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) ( 0 , η , 0 , η 2 e ) as x + for t 0 ,

where both ε and η are positive constants and

ψ ε , η = a δ δ 1 h ε , η , A ˜ r ε , η = σ e σ a I ε , η + 0 S 2 ν ν σ s χ σ s I ε , η d Ω d ν ,

χ , v = ( v ( 1 ) , v ( 2 ) , v ( 3 ) ) and g are known functions satisfying ( χ , v , g ) ( t = 0 ) = ( I 0 , u 0 , ( ϕ 0 ) 2 e ) and

(3.9) χ L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 3 ) ) , χ t L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 2 ) ) , g L C ( [ 0 , T ] × R 3 ) , g C ( [ 0 , T ] ; H 2 ) , g t C ( [ 0 , T ] ; H 2 ) , g t t L 2 ( [ 0 , T ] ; L 2 ) , v C ( [ 0 , T ] ; H 3 ) L 2 ( [ 0 , T ] ; H 4 ) , v t C ( [ 0 , T ] ; H 1 ) L 2 ( [ 0 , T ] ; D 2 ) , v t t L 2 ( [ 0 , T ] ; L 2 ) , t 1 2 v L ( [ 0 , T ] ; D 4 ) , t 1 2 v t L ( [ 0 , T ] ; D 2 ) L 2 ( [ 0 , T ] ; D 3 ) , t 1 2 v t t L ( [ 0 , T ] ; L 2 ) L 2 ( [ 0 , T ] ; D 1 ) ,

where T > 0 is an arbitrary constant.

Next, by using standard argument ([4,14,20]), the global well-posedness of classical solutions ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) in [ 0 , T ] × R 3 × R + × S 2 to (3.8) can be obtained for ε and η are both positive.

Lemma 3.1

Let (2.8) hold and ε > 0 , η > 0 . Assume ( I 0 , ϕ 0 , u 0 ) satisfies (3.5) and (3.6). Then for any T > 0 , there exists a unique classical solution ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) in [ 0 , T ] × R 3 × R + × S 2 to (3.8), satisfying

(3.10) I ε , η L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 3 ) ) , I t ε , η L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 2 ) ) , ϕ ε , η η C ( [ 0 , T ] ; H 3 ) , h ε , η L C ( [ 0 , T ] × R 3 ) , h ε , η C ( [ 0 , T ] ; H 2 ) , h t ε , η C ( [ 0 , T ] ; H 2 ) , u ε , η C ( [ 0 , T ] ; H 3 ) L 2 ( [ 0 , T ] ; H 4 ) , u t ε , η C ( [ 0 , T ] ; H 1 ) L 2 ( [ 0 , T ] ; D 2 ) , u t t ε , η L 2 ( [ 0 , T ] ; L 2 ) , t 1 2 u ε , η L ( [ 0 , T ] ; D 4 ) , t 1 2 u t ε , η L ( [ 0 , T ] ; D 2 ) L 2 ( [ 0 , T ] ; D 3 ) , t 1 2 u t t ε , η L ( [ 0 , T ] ; L 2 ) L 2 ( [ 0 , T ] ; D 1 ) .

3.3 A priori estimates independent of ( ε , η )

Let ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) be a classical solution obtained in Lemma 3.1 and the initial data satisfy (3.5) and (3.6). Now, we are going to establish the uniform a priori estimates for ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) , which are independent of ε , η . For this purpose, we first choose a positive constant c 0 independent of η such that

(3.11) 2 + η + ϕ 0 η η 3 + I 0 η L 2 ( R + × S 2 ; H 3 ) + h 0 η D 1 D 2 + u 0 η 3 + g 1 η 2 + g 2 η 2 + g 3 η 2 + ( h 0 η ) 1 L D 1 , 6 D 2 , 3 D 3 + h 0 η / h 0 η L L 6 D 1 , 3 D 2 c 0 ,

where

(3.12) g 1 η = ( ϕ 0 η ) e u 0 , g 2 η = a ( ϕ 0 η ) 2 e L u 0 η , g 3 η = ( ϕ 0 η ) e a ( ϕ 0 η ) 2 e L u 0 η + 1 c 0 S 2 A ¯ r 0 η d Ω d ν .

Remark 3.1

It follows from (3.12) and the definition of Lamé operator that

(3.13) a L ( ( ϕ 0 η ) 2 e u 0 η ) = g 2 η δ 1 δ G ( ψ 0 η , u 0 η ) , ( ϕ 0 η ) 2 e u 0 η 0 as x + ,

where ψ 0 η = a δ δ 1 h 0 η , and

(3.14) G = α ψ 0 η u 0 η + α div ( u 0 η ψ 0 η ) + ( α + β ) ( ψ 0 η div u 0 η + ψ 0 η ( u 0 η ) + u 0 η ψ 0 η ) .

As a consequence of Lemma A.6 and (3.11) and (3.12), one can deduce that

(3.15) ( ϕ 0 η ) 2 e u 0 η D 2 C ( g 2 η 2 + G ( ψ 0 η , u 0 η ) 2 ) C 1 , ( ϕ 0 η ) 2 e 2 u 0 η 2 C ( ( ϕ 0 η ) 2 e u 0 η D 2 + ψ 0 η 6 u 0 η 3 + ψ 0 η u 0 η 2 ) C 1 , ( ϕ 0 η ) e 2 ϕ 0 η 2 + ( ϕ 0 η ) e ( ψ 0 η Q ( u 0 η ) ) 2 C 1 ,

where C 1 is a positive constant depending on c 0 , A , Λ , α , β , γ , and δ , but independent of ε , η .

We assume there exist some time T ( 0 , T ] and positive constants c i ( i = 1 , , 7 ) such that

1 < c 0 c 1 c 2 c 3 c 4 c 5 c 6 c 7 ,

and

(3.16) sup 0 t T χ ( t ) L 2 ( R + × S 2 ; H 3 ) 2 c 1 2 , sup 0 t T χ t ( t ) L 2 ( R + × S 2 ; H 2 ) 2 c 2 2 , sup 0 t T g ( t ) D 1 D 2 2 c 3 2 , sup 0 t T v ( t ) 1 2 + 0 T ( v ( t ) D 2 2 + v t ( t ) 2 2 ) d t c 4 2 , sup 0 t T ( v D 2 2 + v t 2 2 + g 2 v 2 2 ) ( t ) + 0 T ( v ( t ) D 3 2 + v t ( t ) D 1 2 ) d t c 5 2 , sup 0 t T ( v D 3 2 + v t D 1 2 + g t D 1 2 ) ( t ) + 0 T ( v ( t ) D 4 2 + v t ( t ) D 2 2 + v t t ( t ) 2 2 ) d t c 6 2 , sup 0 t T ( g 2 v D 1 2 + g t 2 ) ( t ) + 0 T ( ( g 2 v ) t ( t ) 2 2 + g 2 v ( t ) D 2 2 ) d t c 6 2 , ess sup 0 t T t ( v t D 2 2 + v D 4 2 + v t t 2 2 ) ( t ) + 0 T t ( v t t ( t ) D 1 2 + v t ( t ) D 3 2 ) d t c 7 2 ,

where T and c i ( i = 1 , , 7 ) will be determined later, and depend only on c 0 and the fixed constants ( A , α , β , γ , δ , Λ , T ) .

Hereinafter, M = M ( c 0 ) denotes a strictly increasing function from [ 0 , ) to [ 1 , ) , and C 1 denotes a generic positive constant. Both M ( c 0 ) and C depend only on ( c 0 , A , α , β , γ , δ , Λ , T ) and may be different from line to line. Moreover, in the rest of Section 3.3, without causing ambiguity, we simply denote

( I 0 η , ϕ 0 η , u 0 η , h 0 η ) as ( I 0 , ϕ 0 , u 0 , h 0 ) , ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) as ( I , ϕ , u , h ) , ( g 1 η , g 2 η , g 3 η ) as ( g 1 , g 2 , g 3 ) .

3.3.1 The a priori estimates for ϕ and ψ

The following two lemmas give the estimates for ϕ and ψ .

Lemma 3.2

ϕ ( t ) η 3 C c 0 , ϕ t ( t ) 2 C c 0 c 4 , ϕ t ( t ) D 1 C c 0 c 5 , ϕ t ( t ) D 2 C c 0 c 6 , ϕ t t ( t ) 2 C c 6 3 , 0 t ϕ t t ( s ) 1 2 d s C c 0 2 c 6 2 ,

for 0 t T 1 = min { T , ( 1 + C c 6 ) 2 } .

Proof

First, the standard arguments for the transport equation and (3.16) yield that for 0 t T 1 ,

ϕ η 3 C ϕ 0 η 3 + η 0 t v 3 d s exp 0 t C v 4 d s C c 0 .

Second, according to the equation (3.8) 2 , one can obtain that for 0 t T 1 ,

ϕ t ( t ) 2 C ( v 3 ϕ 6 + ϕ v 2 ) C c 0 c 4 , ϕ t ( t ) D 1 C ( v + ϕ 6 v 3 + ϕ 2 v 2 ) C c 0 c 5 , ϕ t ( t ) D 2 C v 3 ( ϕ 2 + ϕ ) C c 0 c 6 , ϕ t t ( t ) 2 = v t ϕ v v t ( γ 1 ) ( ϕ t div v + ϕ div v t ) 2 C ( v 3 ϕ t 2 + ϕ 1 v t 1 ) C c 6 3 .

Moreover, one also has that for 0 t T 1 ,

0 t ϕ t t ( s ) 1 2 d s C 0 t ( v t ϕ 1 + v ϕ t 1 + ϕ t div v 1 + ϕ div v t 1 ) 2 d s C c 0 2 c 6 2 .

The proof of Lemma 3.2 is complete.□

Lemma 3.3

ψ ( t ) 2 + ψ ( t ) D 1 D 2 2 C c 0 2 , ψ t ( t ) 2 2 C c 5 4 , h t ( t ) 2 C c 5 3 c 6 , ψ t ( t ) D 1 2 + 0 t ( ψ t t ( s ) 2 2 + h t t ( s ) 6 2 ) d s C c 6 4 , for 0 t T 1 .

Proof

According to the equation (3.8) 4 and ψ = a δ δ 1 h , one can deduce that

(3.17) ψ t + l = 1 3 A l ( v ) l ψ + B ( v ) ψ + a δ ( g div v + g div v ) = 0 ,

where A l ( v ) = ( a i j l ) 3 × 3 ( i , j , l = 1 , 2 , 3 ) are symmetric with

a i j l = v l for i = j ; otherwise a i j l = 0 ,

and B ( v ) = ( v ) .

First, let ζ = ( ζ 1 , ζ 2 , ζ 3 ) ( 1 ζ 2 and ζ i = 0 , 1 , 2 ). Applying x ζ to (3.17), multiplying by 2 x ζ ψ and integrating over R 3 , one obtains

(3.18) d d t x ζ ψ 2 C ( v x ζ ψ 2 + Θ ζ 2 ) ,

where

Θ ζ = x ζ ( B ψ ) B x ζ ψ + l = 1 3 ( x ζ ( A l l ψ ) A l l x ζ ψ ) + a δ x ζ ( g div v + g div v ) .

For Θ ζ 2 , it can be estimated as follows:

Θ ζ 2 C ( 2 v 2 ( ψ + g ) + v ( ψ 2 + 2 g 2 ) + g 2 v D 1 ) for ζ = 1 , Θ ζ 2 C ( v ( 2 ψ 2 + 3 g 2 ) + 2 v 3 ( ψ 6 + 2 g 6 ) + 3 v 2 ( ψ + g ) + g div v D 2 ) for ζ = 2 .

Plugging the aforementioned estimates for Θ ζ 2 into (3.18) and using the Gagliardo-Nirenberg inequality, one obtains

d d t ψ ( t ) D 1 D 2 C ( c 6 ψ ( t ) D 1 D 2 + g div v D 2 + c 6 2 ) ,

which, along with the Gronwall inequality, implies that for 0 t T 1 ,

ψ ( t ) D 1 D 2 C c 0 + c 6 2 t + 0 t g div v D 2 d s exp ( C c 6 t ) C c 0 .

Second, according to equations (3.17), one has that for 0 t T 1 ,

ψ t ( t ) 2 C ( v ψ D 1 + v 2 ψ + g 2 v 2 + g v 2 ) C c 5 2 , ψ t ( t ) 2 C ( v 3 ( ψ D 1 D 2 + g D 1 D 2 ) + g 2 v D 1 ) C c 6 2 , 0 t ψ t t ( s ) 2 2 d s C 0 t ( v t 6 2 ψ 3 2 + v 2 ψ t 2 2 + v 2 ψ t 2 2 + ψ 2 v t 2 2 + ( g div v ) t 2 2 + g 2 v t 2 2 + v 2 g t 2 2 ) d s C c 6 4 .

On the other hand, Lemma A.1 (Appendix A) and (3.16) yield that

(3.19) g div v C g div v D 1 1 2 g div v D 2 1 2 C ( g v 2 + g 2 v 2 ) 1 2 ( 2 g 2 v + g 2 v 2 + g 2 v D 1 ) 1 2 C c 5 3 2 c 6 1 2 .

Finally, it follows from (3.19) and the equation (3.8) 4 that for 0 t T 1 ,

h t ( t ) C ( v ψ + g div v ) C c 5 3 2 c 6 1 2 , 0 t h t t ( s ) 6 2 d s C 0 t ( v ψ t 6 + v t 6 ψ + g t v 6 + g v t 6 ) 2 d s C c 6 4 ,

where one has used

g v t 6 C ( g v t 2 + g 2 v t 2 ) C ( g v t 2 + ( g 2 v ) t 2 + g t 2 v 2 ) C c 6 2 .

The proof of Lemma 3.3 is complete.□

3.3.2 The a priori estimates for I

Lemma 3.4

I L 2 ( R + × S 2 ; C ( [ 0 , T 2 ] ; H 3 ) ) C c 0 , I t L 2 ( R + × S 2 ; C ( [ 0 , T 2 ] ; H 2 ) ) M ( c 0 ) c 1 κ + 2 , I t t L 2 ( R + × S 2 ; C ( [ 0 , T 2 ] ; L 2 ) ) M ( c 0 ) c 4 2 κ + 3 ,

for 0 t T 2 = min { T , ( 1 + M ( c 0 ) c 6 ) 2 ( κ + 1 ) } with κ = 1 / ( γ 1 ) 3 .

Proof

First, multiplying (3.8) 1 by 2 I and integrating over R 3 , one has

(3.20) d d t I 2 2 C σ e 2 I 2 + I 2 ϕ κ χ L 2 ( R + × S 2 ; L 2 ) 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν 1 2 C σ e 2 2 + I 2 2 + c 0 2 κ c 1 2 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν ,

where one has used the fact σ a 0 and σ s 0 .

Second, by applying x ζ ( 1 ζ 3 ) to (3.8) 1 , multiplying the resulting equation by x ζ I , and integrating over R 3 , one can obtain

(3.21) 1 2 d d t x ζ I 2 d x + σ a + 0 S 2 σ s d Ω d ν ( x ζ I ) 2 d x = ( x ζ ( σ a I ) σ a x ζ I ) x ζ I d x 0 S 2 σ ¯ s ( x ζ ( ϕ κ I ) ϕ κ x ζ I ) d Ω d ν x ζ I d x + x ζ σ e x ζ I d x + 0 S 2 ν ν σ ¯ s x ζ ( ϕ κ χ ) d Ω d ν x ζ I d x i = 1 4 R i .

According to Lemma 3.2, Hölder’s inequality, and Gagliardo-Nirenberg inequality, one obtains

R 1 C σ a I 2 I 2 for ζ = 1 , R 1 C ( 2 σ a 2 I + σ a 3 I 6 ) 2 I 2 for ζ = 2 , R 1 C ( 3 σ a 2 I + 2 σ a 2 I + σ a 3 2 I 6 ) 3 I 2 for ζ = 3 , R 2 C ϕ κ 3 I 6 I 2 for ζ = 1 , R 2 C ( 2 ϕ κ 2 I + ϕ κ 3 I 6 ) 2 I 2 for ζ = 2 , R 2 C ( 3 ϕ κ 2 I + 2 ϕ κ 2 I + ϕ κ 3 2 I 6 ) 3 I 2 for ζ = 3 , R 3 C ( x ζ σ e 2 2 + x ζ I 2 2 ) , R 4 C c 0 κ x ζ I 2 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν 1 2 0 S 2 χ 1 2 d Ω d ν 1 2 C I 2 2 + C c 0 2 κ c 1 2 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν for ζ = 1 , R 4 C 2 I 2 2 + C c 0 2 κ c 1 2 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν for ζ = 2 , R 4 C 3 I 2 2 + C c 0 2 κ c 1 2 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν for ζ = 3 .

Plugging the aforementioned estimates for R i ( i = 1 , , 4 ) into (3.21), summing up all ζ ( 1 ζ 3 ) , and combining (3.20), one arrives at

(3.22) d d t I 3 2 M ( c 0 ) c 0 κ + 1 I 3 2 + C σ e 3 2 + C c 0 2 κ c 1 2 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν .

Integrating (3.22) over R + × S 2 , using the hypotheses H1H2 and the Gronwall inequality, one has that for 0 t T 2 ,

(3.23) I L 2 ( R + × S 2 ; H 3 ) 2 C ( I 0 L 2 ( R + × S 2 ; H 3 ) 2 + M ( c 0 ) c 0 2 κ c 1 2 t ) exp ( M ( c 0 ) c 0 κ + 1 t ) C c 0 2 .

According to the equation (3.8) 1 and (3.23), one can obtain that for 0 t T 2 ,

I t L 2 ( R + × S 2 ; H 2 ) C ( I L 2 ( R + × S 2 ; H 3 ) + σ e L 2 ( R + × S 2 ; H 2 ) + σ a L 2 ( R + × S 2 ; H 2 ) I L 2 ( R + × S 2 ; H 2 ) + ϕ κ 2 χ L 2 ( R + × S 2 ; H 2 ) 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν 1 2 + ϕ κ 2 I 2 0 S 2 σ ¯ s d Ω d ν M ( c 0 ) c 1 κ + 2 .

Finally, differentiating (3.8) 1 with respect to t , one can similarly obtain

I t t L 2 ( R + × S 2 ; L 2 ) C ( I t L 2 ( R + × S 2 ; L 2 ) + ( σ e ) t L 2 ( R + × S 2 ; L 2 ) + ( σ a ) t L 2 ( R + × S 2 ; L 2 ) I L 2 ( R + × S 2 ; L ) + σ a L 2 ( R + × S 2 ; L ) I t L 2 ( R + × S 2 ; L 2 ) + ( ( ϕ κ ) t 2 I L 2 ( R + × S 2 ; L ) + ϕ κ I t L 2 ( R + × S 2 ; L 2 ) ) 0 S 2 σ ¯ s d Ω d ν + ( ( ϕ κ ) t 2 χ L 2 ( R + × S 2 ; L ) + ϕ κ χ t L 2 ( R + × S 2 ; L 2 ) ) × 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν 1 2 M ( c 0 ) c 4 2 κ + 3 .

The proof of Lemma 3.4 is complete.□

3.3.3 The a priori estimates for auxiliary variables φ and f

(3.24) φ = h 1 = ϕ 2 e , f = ψ φ = 2 a e δ δ 1 log ϕ .

Lemma 3.5

φ ( t ) D 1 , 6 D 2 , 3 D 3 2 + f ( t ) L L 6 D 1 , 3 D 2 2 C c 0 4 , h ( t , x ) > 1 2 c 0 , 2 3 η 2 e < φ ( t , x ) < 2 φ 0 2 c 0 , φ t ( t ) L 6 D 1 , 3 D 2 2 + f t ( t ) L 3 D 1 2 C c 6 10 ,

for 0 t T 3 = min { T , ( 1 + M ( c 0 ) c 6 ) 2 ( κ + 2 ) } .

Proof

Step 1: Estimate on φ . First, according to (3.8) 2 and the definition of φ in (3.24), one deduces that φ satisfies the following transport equation:

(3.25) φ t + v φ ( δ 1 ) g φ 2 div v = 0 ,

which, along with the standard characteristic method, yields that

(3.26) 2 3 η 2 e < φ ( t , x ) < 2 φ 0 2 c 0 for ( t , x ) [ 0 , T 3 ] × R 3 .

Second, by the standard energy estimates for the equation (3.25), one can obtain

d d t φ 6 C ( ( v + φ g div v ) φ 6 + φ 2 ( g 2 v 6 + v g 6 ) ) , d d t 2 φ 3 C ( ( v + φ g div v ) φ 3 + φ 6 ( 2 v 6 + φ 6 g div v ) + φ 2 ( g 2 div v 3 + 2 g 3 v + g 2 v 3 ) + φ φ 6 ( g 2 v 6 + g v 6 ) ) , d d t 3 φ 2 C ( ( v + φ g div v ) 3 φ 2 + φ 6 3 v 2 + 2 φ 3 2 v 6 + φ 2 ( g 3 div v 2 + v 3 g 2 + 2 g 6 2 v 3 + g 3 v 2 ) + g div v φ 6 2 φ 3 + φ 6 2 ( g v 6 + g 2 v 6 ) + φ 2 φ 3 ( g v 6 + g 2 v 6 ) + φ φ 6 ( g 2 v 3 + 2 g 6 v 6 + g 2 div v 3 ) ) ,

which, along with (3.26) and the Gronwall inequality, yield that

(3.27) φ ( t ) D 1 , 6 + φ ( t ) D 2 , 3 + φ ( t ) D 3 C c 0 for 0 t T 3 .

Finally, according to equations (3.25)–(3.27), one can similarly obtain

(3.28) φ t ( t ) L 6 D 1 , 3 D 2 C c 6 4 for 0 t T 3 .

Step 2: Estimate on f . First, by the definition of f in (3.24), (3.27)–(3.28), and Lemma 3.3, one has that for 0 t T 3 ,

f ( t ) C c 0 2 , f ( t ) 6 ψ 6 φ C c 0 2 , f ( t ) 3 C ( φ 6 ψ 6 + φ ψ 3 ) C c 0 2 , 2 f ( t ) 2 C ( φ 2 ψ 2 + ψ 6 2 φ 3 + φ 6 ψ 3 ) C c 0 2 , f t ( t ) 3 C ( ψ 6 φ t 6 + φ ψ t 3 ) C c 6 5 , f t ( t ) 2 C ( ψ 6 φ t 3 + φ ψ t 2 + ψ 3 φ t 6 + φ 6 ψ t 3 ) C c 6 5 .

Second, it follows from the definition that f satisfies the following equations:

f t + l = 1 3 A l ( v ) l f + B ( v ) f + a δ ( g φ div v + φ g div v ) ( δ 1 ) g φ f div v = 0 .

The proof of Lemma 3.5 is complete.□

3.3.4 The a priori estimates for u

On the basis of estimates obtained in Lemmas 3.23.5, now we are ready to give the lower order estimates for u .

Lemma 3.6

h u ( t ) 2 2 + u ( t ) 1 2 + 0 t ( u ( s ) 1 2 + u t ( s ) 2 2 ) d s C c 0 4 , ( u D 2 2 + h 2 u 2 2 + u t 2 2 ) ( t ) + 0 t ( u ( s ) D 3 2 + u t ( s ) D 1 2 ) d s M ( c 0 ) c 4 9 c 5 ,

for 0 t T 4 = min { T , ( 1 + M ( c 0 ) c 6 ) 2 ( κ + 8.25 ) } .

Proof

Step 1: Estimate on u 2 . First, multiplying (3.8) 3 by u and integrating over R 3 , according to Hölder’s inequality, Young’s inequality, and Gagliardo-Nirenberg inequality, one has

(3.29) 1 2 d d t u 2 2 + a α ( h 2 + ε 2 ) 1 4 u 2 2 + a ( α + β ) ( h 2 + ε 2 ) 1 4 div u 2 2 = v v + ϕ + a h 2 + ε 2 Q ( u ) ψ Q ( v ) + 1 c 0 S 2 A ¯ r Ω d Ω d ν u d x , C u 2 v v 2 + ϕ 2 + ψ h u 2 φ 1 2 + ψ v 2 + 0 S 2 A ¯ r 2 d Ω d ν , C ( 1 + v 2 + v 2 2 + φ ψ 2 ) u 2 2 + a α 2 h u 2 2 + C c 5 2 + C u 2 0 S 2 A ¯ r 2 d Ω d ν R 5 .

Next we estimate the last term R 5 on the right-hand side of (3.29). By Hölder’s inequality and the hypotheses H1H2, one obtains

(3.30) R 5 = C u 2 0 S 2 σ ¯ e σ ¯ a I + 0 S 2 ν ν σ ¯ s I σ ¯ s I d Ω d ν 2 d Ω d ν C u 2 ( σ ¯ e L 1 ( R + × S 2 ; L 2 ) + σ ¯ a L 2 ( R + × S 2 ; L ) I L 2 ( R + × S 2 ; L 2 ) + I L 2 ( R + × S 2 ; L 2 ) 0 S 2 0 S 2 ν ν 2 σ ¯ s 2 d Ω d ν 1 2 d Ω d ν + I L 2 ( R + × S 2 ; L 2 ) 0 S 2 0 S 2 σ ¯ s d Ω d ν 2 d Ω d ν 1 2 M ( c 0 ) ( u 2 2 + c 0 4 ) .

Substituting (3.30) into (3.29), one arrives at

(3.31) d d t u 2 2 + h u 2 2 M ( c 0 ) ( c 5 3 u 2 2 + c 5 4 ) ,

which, along with the Gronwall inequality, yields that for 0 t T 4 ,

(3.32) u ( t ) 2 2 + 0 t h u ( s ) 2 2 d s C ( u 0 2 2 + M ( c 0 ) c 5 4 t ) exp ( M ( c 0 ) c 5 3 t ) C c 0 2 .

Step 2: Estimate on u 2 . Multiplying (3.8) 3 by u t and integrating over R 3 , one can obtain

(3.33) 1 2 d d t a α ( h 2 + ε 2 ) 1 4 u 2 2 + a ( α + β ) ( h 2 + ε 2 ) 1 4 div u 2 2 + u t 2 2 = v v + ϕ + a h 2 + ε 2 Q ( u ) ψ Q ( v ) + 1 c 0 S 2 A ¯ r Ω d Ω d ν u t d x a h h t h 2 + ε 2 ( α u 2 + ( α + β ) div u 2 ) d x C u t 2 v v 2 + ϕ 2 + ψ h u 2 φ 1 2 + σ ¯ e L 1 ( R + × S 2 ; L 2 ) + ( 1 + σ ¯ a L 2 ( R + × S 2 ; L ) ) I L 2 ( R + × S 2 ; L 2 ) + C h t φ 1 2 h u 2 2 1 8 u t 2 2 + C 1 + ( v 2 + ψ 2 ) v 2 2 + ϕ 2 2 + ( φ ψ 2 + h t φ 1 2 ) h u 2 2 + σ ¯ e L 1 ( R + × S 2 ; L 2 ) 2 + σ ¯ a L 2 ( R + × S 2 ; L ) 2 + I L 2 ( R + × S 2 ; L 2 ) 2 ,

which, along with the Gronwall inequality, yields that for 0 t T 4 ,

(3.34) h u ( t ) 2 2 + 0 t u t ( s ) 2 2 d s C ( c 0 2 + M ( c 0 ) c 5 4 t ) exp ( C c 6 3 t ) C c 0 2 , u ( t ) D 1 2 C c 0 3 .

Step 3: Estimate on u D 2 . First, it follows from (3.8) 3 and the definition of Lamé operator that

(3.35) a L ( h 2 + ε 2 u ) = u t v v ϕ + ψ Q ( v ) G ( h 2 + ε 2 , u ) 1 c 0 S 2 A r ¯ Ω d Ω d ν .

Then according to (3.14) for the definition of G , (3.30), and Lemma A.6 (Appendix A), one can obtain that

(3.36) h 2 + ε 2 u D 2 C u t + v v + ϕ ψ Q ( v ) + G ( h 2 + ε 2 , u ) 2 + 0 S 2 A ¯ r 2 d Ω d ν C u t 2 + M ( c 0 ) c 4 3 2 c 5 1 2 , h 2 + ε 2 2 u 2 C ( h 2 + ε 2 u D 2 + ψ 3 u 6 + ψ u 2 + ψ 2 u 2 φ ) C u t 2 + M ( c 0 ) c 4 7 2 c 5 1 2 .

It thus follows from (3.34) and (3.36) that for 0 t T 4 ,

(3.37) 0 t ( h 2 u ( s ) 2 2 + 2 u ( s ) 2 2 ) d s C c 0 4 .

Second, differentiating the equations (3.8) 3 with respect to t , one has

(3.38) u t t + a h 2 + ε 2 L u t = ( v v ) t ϕ t a h h t h 2 + ε 2 L u + ( ψ Q ( v ) ) t 1 c 0 S 2 ( A ¯ r ) t Ω d Ω d ν .

Multiplying (3.38) by u t and integrating over R 3 , one arrives at

(3.39) 1 2 d d t u t 2 2 + a α ( h 2 + ε 2 ) 1 4 u t 2 2 + a ( α + β ) ( h 2 + ε 2 ) 1 4 div u t 2 2 = ( v v ) t ϕ t a h 2 + ε 2 Q ( u ) t + ( ψ Q ( v ) ) t a h h t h 2 + ε 2 L u 1 c 0 S 2 ( A ¯ r ) t Ω d Ω d ν u t d x C u t 2 v v t 2 + v t 2 v + ϕ t 2 + φ 1 2 ψ h u t 2 + ψ v t 2 + ψ t 2 v + h t 2 u 2 + ( σ ¯ e ) t L 1 ( R + × S 2 ; L 2 ) + I L 2 ( R + × S 2 ; L ) ( σ ¯ a ) t L 2 ( R + × S 2 ; L 2 ) + ( 1 + σ ¯ a L 2 ( R + × S 2 ; L ) ) I t L 2 ( R + × S 2 ; L 2 ) ) 1 8 h u t 2 2 + C ( ( 1 + φ ψ 2 + h t 2 ) u t 2 2 + ( ψ 2 + v 2 2 ) v t 1 2 + ϕ t 2 2 + ψ t 2 2 v 2 + 2 u 2 2 + ( σ ¯ e ) t L 1 ( R + × S 2 ; L 2 ) 2 + I L 2 ( R + × S 2 ; L ) 2 ( σ ¯ a ) t L 2 ( R + × S 2 ; L 2 ) 2 + ( 1 + σ ¯ a L 2 ( R + × S 2 ; L ) 2 ) I t L 2 ( R + × S 2 ; L 2 ) 2 ) ,

which, along with the hypotheses H1H2 and Lemmas 3.23.5, yields that

(3.40) d d t u t 2 2 + ( h 2 + ε 2 ) 1 4 u t 2 2 C ( c 6 4 u t 2 2 + 2 u 2 2 ) + M ( c 0 ) c 6 2 κ + 6 .

Integrating (3.40) over ( τ , t ) ( τ ( 0 , t ) ) and using (3.37), one has

(3.41) u t ( t ) 2 2 + τ t h u t ( s ) 2 2 d s u t ( τ ) 2 2 + C c 6 4 τ t u t ( s ) 2 2 d s + C c 0 4 + M ( c 0 ) c 6 2 κ + 6 t .

It follows from (3.8) 3 that

u t ( τ ) 2 C ( v v 2 + ϕ 2 + ( h + ε ) L u 2 + ψ v 2 + σ ¯ e L 1 ( R + × S 2 ; L 2 ) + σ ¯ a L 2 ( R + × S 2 ; L ) I L 2 ( R + × S 2 ; L 2 ) + I L 2 ( R + × S 2 ; L 2 ) ) ( τ ) ,

which, together with the time continuity of ( I , ϕ , u , ψ ) , (2.3) and (3.11), implies that

lim sup τ 0 u t ( τ ) 2 M ( c 0 ) c 0 2 .

Letting τ 0 in (3.41) and using the Gronwall inequality, one can obtain

(3.42) u t ( t ) 2 2 + 0 t h u t ( s ) 2 2 d s M ( c 0 ) c 0 4 and 0 t u t ( s ) 2 2 d s M ( c 0 ) c 0 5 ,

for 0 t T 4 .

It thus follows from (3.36) that for 0 t T 4 ,

(3.43) h 2 u ( t ) 2 + h 2 + ε 2 u ( t ) D 2 M ( c 0 ) c 4 7 2 c 5 1 2 and u ( t ) D 2 M ( c 0 ) c 4 9 2 c 5 1 2 .

As a consequence of Lemma A.6 and (3.35), one also has

(3.44) h 2 + ε 2 u D 3 C u t + v v + ϕ ψ Q ( v ) D 1 + G ( h 2 + ε 2 , u ) D 1 + 0 S 2 A ¯ r D 1 d Ω d ν C u t 2 + M ( c 0 ) c 5 6 ,

which implies that

(3.45) h 2 + ε 2 3 u 2 C ( h 2 + ε 2 u D 3 + h 2 + ε 2 2 u 2 + 2 h 2 + ε 2 3 u 6 + 3 h 2 + ε 2 2 u ) C ( h 2 + ε 2 u D 3 + ( ψ + φ ψ 6 2 + ψ 3 ) 2 u 2 + ( φ 2 ψ 6 3 + φ ψ 2 ψ 2 + 2 ψ 2 ) u 2 1 2 2 u 2 1 2 C u t 2 + M ( c 0 ) c 5 8.25 .

Consequently, it follows from (3.42) and (3.45) that for 0 t T 4 ,

(3.46) 0 t ( h 3 u ( s ) 2 2 + h 2 u ( s ) D 1 2 + u ( s ) D 3 2 ) d s M ( c 0 ) c 0 7 .

The proof of Lemma 3.6 is complete.□

The following lemma gives the higher order estimates for u .

Lemma 3.7

( h u t 2 2 + u t D 1 2 + u D 3 2 + h 2 u D 1 2 ) ( t ) + 0 t u t ( s ) D 2 2 d s M ( c 0 ) c 5 18.5 , 0 t ( u t t ( s ) 2 2 + u ( s ) D 4 2 + h 2 u ( s ) D 2 2 + ( h 2 u ) t ( s ) 2 2 ) d s M ( c 0 ) c 0 10 ,

for 0 t T 5 = min { T , ( 1 + M ( c 0 ) c 6 ) 2 ( κ + 10.25 ) } .

Proof

First, multiplying (3.38) by u t t and integrating over R 3 , using Hölder’s inequality, Young’s inequality, and Gagliardo-Nirenberg inequality, one has

(3.47) 1 2 d d t a α ( h 2 + ε 2 ) 1 4 u t 2 2 + a ( α + β ) ( h 2 + ε 2 ) 1 4 div u t 2 2 + u t t 2 2 = ( v v ) t ϕ t a h h 2 + ε 2 h t L u + a h h 2 + ε 2 h Q ( u t ) + ( ψ Q ( v ) ) t 1 c 0 S 2 ( A ¯ r ) t Ω d x d Ω d ν u t t d x + a h h 2 + ε 2 h t ( α u t 2 + ( α + β ) div u t 2 ) d x C u t t 2 v t 1 v 2 + ϕ t 2 + h t 2 u 2 + ψ φ 1 2 h u t 2 + ψ t 2 v + ψ v t 2 + ( σ ¯ e ) t L 1 ( R + × S 2 ; L 2 ) + I L 2 ( R + × S 2 ; L ) ( σ ¯ a ) t L 2 ( R + × S 2 ; L 2 ) + ( 1 + σ ¯ a L 2 ( R + × S 2 ; L ) ) I t L 2 ( R + × S 2 ; L 2 ) ) + h t φ h u t 2 2 1 8 u t t 2 2 + M ( c 0 ) c 6 3 h u t 2 2 + M ( c 0 ) c 6 2 κ + 14 .

Integrating (3.47) over ( τ , t ) ( τ ( 0 , t ) ) , one can obtain

(3.48) h u t ( t ) 2 2 + τ t u t t ( s ) 2 2 d s ( h 2 + ε 2 ) 1 4 u t ( τ ) 2 2 + M ( c 0 ) c 6 3 0 t h u t 2 2 d s + M ( c 0 ) c 6 2 κ + 14 t .

It follows from the equations (3.8) 3 that

(3.49) h u t ( τ ) 2 C h ( v v + ϕ ψ Q ( v ) ) 2 + h a h 2 + ε 2 L u + 1 c 0 S 2 A ¯ r d Ω d ν 2 ( τ ) ,

which, along with the hypotheses H1H2, (3.11)–(3.12) and (3.15), implies that

(3.50) lim sup τ 0 h u t ( τ ) 2 C ( ϕ 0 e ( u 0 u 0 ) 2 + ϕ 0 e 2 ϕ 0 2 + ϕ 0 e ( ψ 0 Q ( u 0 ) ) 2 + g 3 2 + ε L u 0 2 ) C ( u 0 g 1 2 + ϕ 0 e u 0 6 2 u 0 3 + ϕ 0 e 2 ϕ 0 2 + ϕ 0 e u 0 6 ψ 0 3 + ϕ 0 e g 2 2 ψ 0 + g 3 2 + u 0 3 ) M ( c 0 ) c 0 3 .

Letting τ 0 in (3.48) and using the Gronwall inequality, one concludes that for 0 t T 5 ,

(3.51) h u t ( t ) 2 2 + 0 t u t t ( s ) 2 2 d s M ( c 0 ) c 0 6 and u t ( t ) D 1 2 M ( c 0 ) c 0 7 ,

which, together with (3.44)–(3.45), yields that

h 2 + ε 2 u D 3 + h 2 + ε 2 3 u 2 + h 2 u D 1 M ( c 0 ) c 5 8.25 and 3 u 2 M ( c 0 ) c 5 9.25 .

Second, it follows from (3.38) and Lemma A.6 that for 0 t T 5 ,

(3.52) h 2 + ε 2 u t D 2 C u t t + ( v v ) t + ϕ t ( ψ Q ( v ) ) t + a h h 2 + ε 2 h t L u 2 + G ( h 2 + ε 2 , u t ) 2 + 0 S 2 ( A ¯ r ) t 2 d Ω d ν M ( c 0 ) ( u t t 2 + c 6 κ + 7 ) , h 2 + ε 2 2 u t 2 C ( h 2 + ε 2 u t D 2 + 2 h 2 + ε 2 u t 2 + h 2 + ε 2 u t 2 ) M ( c 0 ) ( u t t 2 + c 6 κ + 7 ) , ( h 2 u ) t 2 C ( h 2 u t 2 + h t 2 u 2 ) M ( c 0 ) ( u t t 2 + c 6 κ + 7 ) , u D 4 C ( h 2 + ε 2 ) 1 2 ( u t + v v + ϕ ψ Q ( v ) ) D 2 + ( h 2 + ε 2 ) 1 2 0 S 2 A ¯ r Ω d Ω d ν D 2 M ( c 0 ) ( c 0 2 u t t 2 + c 6 κ + 9 ) .

On the other hand, according to equation (3.8) 3 , for multi-index ζ R 3 ( ζ = 2 ) , one has

a L ( h 2 + ε 2 ζ u ) = h 2 + ε 2 ζ ( h 2 + ε 2 ) 1 2 ( u t + v v + ϕ ψ Q ( v ) ) h 2 + ε 2 ζ ( h 2 + ε 2 ) 1 2 0 S 2 A ¯ r Ω d Ω d ν G ( h 2 + ε 2 , 2 u ) ,

which, along with Lemma A.6 (Appendix A), yields that

(3.53) h 2 + ε 2 2 u D 2 C h 2 + ε 2 2 ( h 2 + ε 2 ) 1 2 ( u t + v v + ϕ ψ Q ( v ) ) 2 + h 2 + ε 2 2 ( h 2 + ε 2 ) 1 2 0 S 2 A ¯ r Ω d Ω d ν 2 + ψ u D 3 + ψ 3 2 u 6 + 2 u 2 ψ 2 φ ) M ( c 0 ) ( u t D 2 + c 6 κ + 10.25 ) .

It thus follows from (3.51)–(3.53) that for 0 t T 5 ,

0 t ( h 2 u t ( s ) 2 2 + u t ( s ) D 2 2 + u ( s ) D 4 2 + h 2 u ( s ) D 2 2 + ( h 2 u ) t ( s ) 2 2 ) d s M ( c 0 ) c 0 10 .

The proof of Lemma 3.7 is complete.□

Finally, we establish the time-weighted energy estimates for u .

Lemma 3.8

t ( u t D 2 2 + u t t 2 2 + u D 4 2 ) ( t ) + 0 t s ( u t t ( s ) D 1 2 + u t ( s ) D 3 2 ) d s M ( c 0 ) c 6 2 κ + 23.5 ,

for 0 t T 6 = min { T , ( 1 + M ( c 0 ) c 7 ) 2 ( κ + 12.25 ) } .

Proof

First, differentiating (3.38) with respect to time t , one has

(3.54) u t t t + a h 2 + ε 2 L u t t = ( v v ) t t ϕ t t + ( ψ Q ( v ) ) t t 2 a h h 2 + ε 2 h t L u t a h h 2 + ε 2 h t t L u a ε 2 h t 2 ( h 2 + ε 2 ) 3 2 L u 1 c 0 S 2 ( A ¯ r ) t t Ω d Ω d ν .

Second, multiplying (3.54) by u t t and integrating over R 3 , one can obtain

(3.55) 1 2 d d t u t t 2 2 + a α ( h 2 + ε 2 ) 1 4 u t t 2 2 + a ( α + β ) ( h 2 + ε 2 ) 1 4 div u t t 2 2 = ( v v ) t t ϕ t t a h h 2 + ε 2 h Q ( u ) t t + ( ψ Q ( v ) ) t t 2 a h h 2 + ε 2 h t L u t a h h 2 + ε 2 h t t L u + a ε 2 h t 2 ( h 2 + ε 2 ) 3 2 L u 1 c 0 S 2 ( A ¯ r ) t t Ω d Ω d ν u t t d x C u t t 2 ( ( v t 3 + ψ t 3 ) v t 6 + ( v t t 2 + ψ t t 2 ) v + ( v + ψ ) v t t 2 ) + C h u t t 2 φ 1 2 ( ϕ t t 2 + ψ u t t 2 ) + C u t t 2 ( h t 2 u t 2 + h t t 6 2 u 3 + h t 2 φ 2 u 2 + ( σ ¯ e ) t t L 1 ( R + × S 2 ; L 2 ) + I L 2 ( R + × S 2 ; L ) ( σ ¯ a ) t t L 2 ( R + × S 2 ; L 2 ) + ( σ ¯ a ) t L 2 ( R + × S 2 ; L ) I t L 2 ( R + × S 2 ; L 2 ) + ( 1 + σ ¯ a L 2 ( R + × S 2 ; L ) ) I t t L 2 ( R + × S 2 ; L 2 ) ) .

Multiplying (3.55) by s and integrating over ( τ , t ) ( τ ( 0 , t ) ) , one arrives at

(3.56) t u t t ( t ) 2 2 + τ t s h u t t 2 2 d s τ u t t ( τ ) 2 2 + M ( c 0 ) c 7 2 κ + 20 τ t s u t t 2 2 d s + M ( c 0 ) c 6 2 κ + 18.5 ( 1 + t ) .

It follows from Lemmas 3.7 and A.5 (Appendix A) that there exists a sequence { s k } k = 1 such that

s k 0 and s k u t t ( s k , x ) 2 2 0 , as k .

Letting τ = s k 0 in (3.56) and using the Gronwall inequality, one has that for 0 t T 6 ,

(3.57) t u t t ( t ) 2 2 + 0 t s h u t t ( s ) 2 2 d s M ( c 0 ) c 6 2 κ + 18.5 ,

and

0 t s u t t ( s ) 2 2 d s M ( c 0 ) c 6 2 κ + 19.5 .

According to (3.52) and (3.57), one obtains

(3.58) t 1 2 2 u t ( t ) 2 M ( c 0 ) c 6 κ + 10.25 , t 1 2 4 u ( t ) 2 M ( c 0 ) c 6 κ + 11.25 .

Finally, according to (3.38) and Lemma A.6, one obtains that for 0 t T 6 ,

(3.59) h 2 + ε 2 u t D 3 C u t t + ( v v ) t + ϕ t a h h 2 + ε 2 h t L u + ( ψ Q ( v ) ) t D 1 + G ( h 2 + ε 2 , u t ) D 1 + 0 S 2 ( A ¯ r ) t D 1 d Ω d ν M ( c 0 ) ( u t t 2 + c 6 ( 2 u t 2 + 2 v t 2 ) + c 6 κ + 11.25 ) , h 2 + ε 2 3 u t 2 C ( h 2 + ε 2 u t D 3 + u t 2 ψ 2 + 2 u t 2 ψ L D 1 , 3 + u t 2 φ ψ ψ L D 1 , 3 + u t 2 ψ 3 φ 2 ) M ( c 0 ) ( u t t 2 + c 6 ( 2 u t 2 + 2 v t 2 ) + c 6 κ + 12.25 ) ,

which, along with (3.57)–(3.58), yield that

0 T 6 s ( h 2 + ε 2 u t D 3 2 + h 3 u t 2 2 + 3 u t 2 2 ) d t M ( c 0 ) c 6 2 κ + 23.5 .

The proof of Lemma 3.8 is complete.□

From Lemmas 3.23.8, for 0 t T 6 = min { T , ( 1 + M ( c 0 ) c 7 ) 2 ( κ + 12.25 ) } , one has

( ϕ ϕ 3 2 + ϕ t 2 2 + ϕ t t 2 2 ) ( t ) + 0 t ϕ t t ( s ) 1 2 d s C c 6 6 , ψ ( t ) D 1 D 2 2 C c 0 2 , ψ t ( t ) 2 C c 5 2 , I ( t ) L 2 ( R + × S 2 ; H 3 ) C c 0 , I t ( t ) L 2 ( R + × S 2 ; H 2 ) M ( c 0 ) c 1 κ + 2 , h ( t , x ) > 1 2 c 0 , ψ t ( t ) D 1 2 + 0 t ( ψ t t ( s ) 2 2 + h t t ( s ) 6 2 ) d s C c 6 4 , φ ( t ) L D 1 , 6 D 2 , 3 D 3 2 + f ( t ) L L 6 D 1 , 3 D 2 2 C c 0 4 , φ > 2 3 η 2 e , φ t ( t ) L 6 D 1 , 3 D 2 2 + f t ( t ) L 3 D 1 2 C c 6 10 , ( h u 2 2 + u 1 2 ) ( t ) + 0 t ( u ( s ) 1 2 + u t ( s ) 2 2 ) d s C c 0 4 , ( u D 2 2 + h 2 u 2 2 + u t 2 2 ) ( t ) + 0 t ( u ( s ) D 3 2 + h 2 u ( s ) D 1 2 + u t ( s ) D 1 2 ) d s M ( c 0 ) c 4 9 c 5 , ( u t D 1 2 + h u t 2 2 + u D 3 2 + h 2 u D 1 2 ) ( t ) + 0 t u t ( s ) D 2 2 d s M ( c 0 ) c 5 18.5 , 0 t ( u t t ( s ) 2 2 + u ( s ) D 4 2 + h u ( s ) D 1 2 + h 2 u ( s ) D 2 2 + ( h 2 u ) t ( s ) 2 2 ) d s M ( c 0 ) c 0 10 , t ( u t D 2 2 + u t t 2 2 + u D 4 2 ) ( t ) + 0 t s ( u t t ( s ) D 1 2 + u t ( s ) D 3 2 ) d s M ( c 0 ) c 6 2 κ + 23.5 .

Then, defining the constants c i ( i = 1 , , 7 )

(3.60) c 1 = C c 0 , c 2 = c 3 = c 4 = M ( c 0 ) c 1 κ + 2 = M ( c 0 ) ( C c 0 ) κ + 2 , c 5 = M ( c 0 ) c 4 9 = M ( c 0 ) 10 ( C c 0 ) 9 ( κ + 2 ) , c 6 = M ( c 0 ) 1 2 c 5 9.25 = M ( c 0 ) 93 ( C c 0 ) 83.25 ( κ + 2 ) , c 7 = M ( c 0 ) 1 2 c 6 κ + 11.75 = M ( c 0 ) 93 κ + 1093.25 ( C c 0 ) 83.25 ( κ 2 + 13.75 κ + 23.5 ) ,

and the time T

(3.61) T = min { T , ( 1 + M ( c 0 ) c 7 ) 2 ( κ + 12.25 ) } ,

one can conclude that for 0 t T ,

(3.62) ( ϕ ϕ 3 2 + ϕ t 2 2 + ϕ t t 2 2 ) ( t ) + 0 t ϕ t t ( s ) 1 2 d s c 7 2 , ψ ( t ) D 1 D 2 2 c 3 2 , h t ( t ) 2 + ψ t ( t ) 2 c 6 2 , ψ t ( t ) D 1 2 + 0 t ( ψ t t ( s ) 2 2 + h t t ( s ) 2 2 ) d s c 7 2 , I ( t ) L 2 ( R + × S 2 ; H 3 ) 2 c 1 2 , I t ( t ) L 2 ( R + × S 2 ; H 2 ) 2 c 2 2 , h > 1 2 c 0 , φ ( t ) L D 1 , 6 D 2 , 3 D 3 2 + f ( t ) L L 6 D 1 , 3 D 2 2 c 4 2 , φ > 2 3 η 2 e , φ t ( t ) L 6 D 1 , 3 D 2 2 + f t ( t ) L 3 D 1 2 c 7 2 , ( h u 2 2 + u 1 2 ) ( t ) + 0 t ( u ( s ) 1 2 + u t ( s ) 2 2 ) d s c 4 2 , ( u D 2 2 + h 2 u 2 2 + u t 2 2 ) ( t ) + 0 t ( u ( s ) D 3 2 + u t ( s ) D 1 2 ) d s c 5 2 , ( h t D 1 2 + u t D 1 2 + u D 3 2 + h u D 1 2 + h 2 u D 1 2 ) ( t ) + 0 t u t ( s ) D 2 2 d s c 6 2 , 0 t ( u t t ( s ) 2 2 + u ( s ) D 4 2 + h 2 u ( s ) D 2 2 + ( h 2 u ) t ( s ) 2 2 ) d s c 6 2 , t ( u t D 2 2 + u t t 2 2 + u D 4 2 ) ( t ) + 0 t s ( u t t ( s ) D 1 2 + u t ( s ) D 3 2 ) d s c 7 2 .

3.4 Passing to the limit ε 0

With the help of ( ε , η ) -independent estimates obtained in (3.62), we will establish the local-in-time existence result for the following linearized problem (3.63) without an artificial viscosity (i.e., ε = 0 ) under the assumption that ϕ 0 η η .

(3.63) 1 c I t η + Ω I η = A ˜ r η , ϕ t η + v ϕ η + ( γ 1 ) ϕ η div v = 0 , u t η + v v + ϕ η + a h η L u η = ψ η Q ( v ) 1 c 0 S 2 A ¯ r η Ω d Ω d ν , h t η + v h η + ( δ 1 ) g div v = 0 , ( I η , ϕ η , u η , h η ) t = 0 = ( I 0 η , ϕ 0 η , u 0 η , h 0 η ) = ( I 0 , ϕ 0 + η , u 0 , ( ϕ 0 + η ) 2 e ) for ( x , ν , Ω ) R 3 × R + × S 2 , ( I η , ϕ η , u η , h η ) ( 0 , η , 0 , η 2 e ) as x + for t 0 ,

where ψ η = a δ δ 1 h η .

Theorem 3.2

Let (2.8) hold. Assume the initial data ( I 0 , ϕ 0 , u 0 , h 0 ) satisfies (3.5) and (3.6), then there exist a time T > 0 independent of η and a unique classical solution

I η , ϕ η , u η , h η , ψ η = a δ δ 1 h η

in [ 0 , T ] × R 3 × R + × S 2 to (3.63) satisfying (3.10) with T replaced by T . Moreover, ( I η , ϕ η , u η , h η ) satisfies the estimates in (3.62) independent of η .

Proof

Step 1: Existence. First, according to Lemmas 3.13.8, one can see that for every fixed ε > 0 and η > 0 , there exist a time T independent of ( ε , η ) and a unique strong solution ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) in [ 0 , T ] × R 3 × R + × S 2 to the linearized problem (3.8) satisfying the estimates in (3.62), which are independent of ( ε , η ).

Second, using the characteristic method and standard energy estimates for the equation (3.8) 4 , one can obtain that for 0 t T ,

(3.64) h ε , η ( t ) + h ε , η ( t ) 2 + h t ε , η ( t ) 2 C ( η , α , β , γ , δ , T , ϕ 0 , u 0 ) .

Then, based on the uniform estimates in (3.62) independent of ( ε , η ), estimates in (3.64) independent of ε , and Lemma A.3 (Appendix A), one concludes that for any R > 0 , there exists a subsequence of solutions (still denoted by ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) ), which converges to a limit ( I η , ϕ η , u η , h η ) in the following strong sense:

(3.65) I ε , η I η in L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 2 ( B R ) ) ) as ε 0 , ( ϕ ε , η , u ε , η , h ε , η ) ( ϕ η , u η , h η ) in C ( [ 0 , T ] ; H 2 ( B R ) ) as ε 0 ,

where B R is a ball centered at origin with radius R .

On the other hand, the uniform estimates in (3.62) independent of ( ε , η ) and estimates in (3.64) independent of ε imply that there exists a subsequence (of subsequence chosen earlier) of solutions (still denoted by ( I ε , η , ϕ ε , η , u ε , η , h ε , η ) ), which converges to a limit ( I η , ϕ η , u η , h η ) as ε 0 in the following weak or weak sense:

(3.66) I ε , η I η in L 2 ( R + × S 2 ; L ( [ 0 , T ] ; H 3 ) ) , I t ε , η I t η in L 2 ( R + × S 2 ; L ( [ 0 , T ] ; H 2 ) ) , ( ϕ ε , η η , u ε , η ) ( ϕ η η , u η ) in L ( [ 0 , T ] ; H 3 ) , ( ϕ t ε , η , ψ ε , η , h t ε , η ) ( ϕ t η , ψ η , h t η ) in L ( [ 0 , T ] ; H 2 ) , u t ε , η u t η in L ( [ 0 , T ] ; H 1 ) , ( ϕ t t ε , η , 3 φ ε , η , 2 f ε , η ) ( ϕ t t η , 3 φ η , 2 f η ) in L ( [ 0 , T ] ; L 2 ) , ( 2 φ t ε , η , f t ε , η ) ( 2 φ t η , f t η ) in L ( [ 0 , T ] ; L 2 ) , t 1 2 ( 2 u ε , η , u t t ε , η , 4 u ε , η ) t 1 2 ( 2 u η , u t t η , 4 u η ) in L ( [ 0 , T ] ; L 2 ) , ( φ ε , η , φ t ε , η , f ε , η ) ( φ η , φ t η , f η ) in L ( [ 0 , T ] ; L 6 ) , ( h ε , η , h t ε , η , φ ε , η , f ε , η ) ( h η , h t η , φ η , f η ) in L ( [ 0 , T ] ; L ) , ( 2 φ ε , η , φ t ε , η ) ( 2 φ η , φ t η ) in L ( [ 0 , T ] ; L 3 ) , ( f ε , η , f t ε , η ) ( f η , f t η ) in L ( [ 0 , T ] ; L 3 ) , u ε , η u η in L 2 ( [ 0 , T ] ; H 3 ) , u t ε , η u t η in L 2 ( [ 0 , T ] ; H 2 ) , ϕ t t ε , η ϕ t t η in L 2 ( [ 0 , T ] ; H 1 ) , ( ψ t t ε , η , h t t ε , η , u t t ε , η ) ( ψ t t η , h t t η , u t t η ) in L 2 ( [ 0 , T ] ; L 2 ) , t 1 2 ( u t t ε , η , 3 u t ε , η ) t 1 2 ( u t t η , 3 u t η ) in L 2 ( [ 0 , T ] ; L 2 ) ,

which, along with the lower semi-continuity of weak or weak convergence, imply that ( I η , ϕ η , u η , h η ) satisfies the corresponding estimates in (3.62) and (3.64) except those weighted estimates for u η .

It follows from the strong convergence in (3.65) and the weak or weak convergence in (3.66) that

(3.67) h ε , η ( u ε , η , u t ε , η ) h η ( u η , u t η ) in L ( [ 0 , T ] ; L 2 ) , h ε , η 2 u ε , η h η 2 u η in L ( [ 0 , T ] ; H 1 ) , ( h ε , η 2 u ε , η ) t ( h η 2 u η ) t in L 2 ( [ 0 , T ] ; L 2 ) , h ε , η 2 u ε , η h η 2 u η in L 2 ( [ 0 , T ] ; D 1 D 2 ) .

Moreover, combining (3.65)–(3.67), one can easily show that ( I η , ϕ η , u η , h η ) is a weak solution in the sense of distributions to (3.63), satisfying

(3.68) I η L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 3 ) ) , I t η L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 2 ) ) , ϕ η η L ( [ 0 , T ] ; H 3 ) , h η L ( [ 0 , T ] × R 3 ) , h η L ( [ 0 , T ] ; H 2 ) , h t η L ( [ 0 , T ] ; H 2 ) , u η L ( [ 0 , T ] ; H 3 ) L 2 ( [ 0 , T ] ; H 4 ) , u t η L ( [ 0 , T ] ; H 1 ) L 2 ( [ 0 , T ] ; D 2 ) , u t t η L 2 ( [ 0 , T ] ; L 2 ) , t 1 2 u η L ( [ 0 , T ] ; D 4 ) , t 1 2 u t η L ( [ 0 , T ] ; D 2 ) L 2 ( [ 0 , T ] ; D 3 ) , t 1 2 u t t η L ( [ 0 , T ] ; L 2 ) L 2 ( [ 0 , T ] ; D 1 ) ,

which imply that the weak solution ( I η , ϕ η , u η , h η ) of (3.63) is actually a strong one.

Step 2: Uniqueness and time continuity. Since h η > 1 2 c 0 , the uniqueness and time continuity of the strong solutions can be obtained by using same arguments as in Lemma 3.1. Finally, Theorem 3.2 is proved.□

3.5 Local slovability of nonlinear problem away from vacuum

In this section, under the assumption that ϕ 0 η η , we will prove the local well-posedness of classical solutions to the following nonlinear problem:

(3.69) 1 c I t η + Ω I η = A r η , ϕ t η + u η ϕ η + ( γ 1 ) ϕ η div u η = 0 , u t η + u η u η + ϕ η + a h η L u η = ψ η Q ( u η ) 1 c 0 S 2 A ¯ r η Ω d Ω d ν , h t η + u η h η + ( δ 1 ) ( ϕ η ) 2 e div u η = 0 , ( I η , ϕ η , u η , h η ) t = 0 = ( I 0 η , ϕ 0 η , u 0 η , h 0 η ) = ( I 0 , ϕ 0 + η , u 0 , ( ϕ 0 + η ) 2 e ) for ( x , ν , Ω ) R 3 × R + × S 2 , ( I η , ϕ η , u η , h η ) ( 0 , η , 0 , η 2 e ) as x + for t 0 ,

where ψ η = a δ δ 1 h η .

Theorem 3.3

Let (2.8) hold and η > 0 . Assume the initial data ( I 0 , ϕ 0 , u 0 , h 0 ) satisfies (3.5) and (3.6), then there exist a time T > 0 independent of η and a unique classical solution:

I η , ϕ η , u η , h η = ( ϕ η ) 2 e , ψ η = a δ δ 1 h η

in [ 0 , T ] × R 3 × R + × S 2 to (3.69) with T is independent of η . Moreover, ( I η , ϕ η , u η , h η ) satisfies the estimates in (3.62) with T replaced by T .

The proof of Theorem 3.3 is based on the classical iteration scheme and the conclusions obtained in Sections 3.23.4. Next, let ( I 0 , ϕ 0 , u 0 , h 0 ) be the solution to the following problem:

(3.70) 1 c W t + Ω W = 0 , in ( 0 , ) × R 3 × R + × S 2 , X t + u 0 X = 0 , in ( 0 , ) × R 3 , Y t Z Y = 0 , in ( 0 , ) × R 3 , Z t + u 0 Z = 0 , in ( 0 , ) × R 3 , ( W , X , Y , Z ) t = 0 = ( I 0 η , ϕ 0 η , u 0 η , h 0 η ) = ( I 0 , ϕ 0 + η , u 0 , ( ϕ 0 + η ) 2 e ) for ( x , ν , Ω ) R 3 × R + × S 2 , ( W , X , Y , Z ) ( 0 , η , 0 , η 2 e ) as x + for t 0 .

Choosing a time T ˆ ( 0 , T ] small enough such that

(3.71) sup 0 t T ˆ I 0 ( t ) L 2 ( R + × S 2 ; H 3 ) 2 c 1 2 , sup 0 t T ˆ I t 0 ( t ) L 2 ( R + × S 2 ; H 2 ) 2 c 2 2 , sup 0 t T ˆ h 0 ( t ) D 1 D 2 2 c 3 2 , sup 0 t T ˆ u 0 ( t ) 1 2 + 0 T ˆ ( u 0 ( t ) D 2 2 + u t 0 ( t ) 2 2 ) d t c 4 2 , sup 0 t T ˆ ( u 0 D 2 2 + u t 0 2 2 + h 0 2 u 0 2 2 ) ( t ) + 0 T ˆ ( u 0 ( t ) D 3 2 + u t 0 ( t ) D 1 2 ) d t c 5 2 , sup 0 t T ˆ ( u 0 D 3 2 + u t 0 D 1 2 + h t 0 D 1 2 ) ( t ) + 0 T ˆ ( u 0 ( t ) D 4 2 + u t 0 ( t ) D 2 2 + u t t 0 ( t ) 2 2 ) d t c 6 2 , sup 0 t T ˆ ( h 0 2 u 0 D 1 2 + h t 0 2 ) ( t ) + 0 T ˆ ( ( h 0 2 u 0 ) t ( t ) 2 2 + h 0 2 u 0 ( t ) D 2 2 ) d t c 6 2 , ess sup 0 t T ˆ t ( u t 0 D 2 2 + u 0 D 4 2 + u t t 0 2 2 ) ( t ) + 0 T ˆ t ( u t t 0 ( t ) D 1 2 + u t 0 ( t ) D 3 2 ) d t c 7 2 .

Proof

Step 1: Existence. Let the beginning step of our iteration be ( χ , v , g ) = ( I 0 , u 0 , h 0 ) , then one can obtain a classical solution ( I 1 , ϕ 1 , u 1 , h 1 ) to (3.63). Inductively, one constructs the approximate sequences of solutions ( I k + 1 , ϕ k + 1 , u k + 1 , h k + 1 ) as follows: given ( I k , ϕ k , u k , h k ) for k 1 , define ( I k + 1 , ϕ k + 1 , u k + 1 , h k + 1 ) by solving the following problem:

(3.72) 1 c I t k + 1 + Ω I k + 1 = A r k , ϕ t k + 1 + u k ϕ k + 1 + ( γ 1 ) ϕ k + 1 div u k = 0 , u t k + 1 + u k u k + ϕ k + 1 + a h k + 1 L u k + 1 = ψ k + 1 Q ( u k ) 1 c 0 S 2 A ˜ r k Ω d Ω d ν , h t k + 1 + u k h k + 1 + ( δ 1 ) h k div u k = 0 , ( I k + 1 , ϕ k + 1 , u k + 1 , h k + 1 ) t = 0 = ( I 0 η , ϕ 0 η , u 0 η , h 0 η ) = ( I 0 , ϕ 0 + η , u 0 , ( ϕ 0 + η ) 2 e ) for ( x , ν , Ω ) R 3 × R + × S 2 , ( I k + 1 , ϕ k + 1 , u k + 1 , h k + 1 ) ( 0 , η , 0 , η 2 e ) as x + for t 0 ,

where

A r k = σ e k + 1 σ a k + 1 I k + 1 + 0 S 2 ν ν σ s k + 1 I k ( σ s ) k + 1 I k + 1 d Ω d ν , A ˜ r k = σ ¯ e k + 1 σ ¯ a k + 1 I k + 1 + 0 S 2 ν ν σ ¯ s I k + 1 σ ¯ s I k + 1 d Ω d ν , σ e k + 1 = ( ϕ k + 1 ) κ σ ¯ e k + 1 , σ ¯ e k + 1 = σ ¯ e ( t , x , ν , Ω , ( ϕ k + 1 ) κ ) , σ a k + 1 = ( ϕ k + 1 ) κ σ ¯ a k + 1 , σ ¯ a k + 1 = σ ¯ a ( t , x , ν , Ω , ( ϕ k + 1 ) κ ) , σ s k + 1 = ( ϕ k + 1 ) κ σ ¯ s ( ν ν , Ω Ω ) , σ s k + 1 = ( ϕ k + 1 ) κ σ ¯ s ( ν ν , Ω Ω ) .

By replacing ( χ , v , g ) = ( I k , u k , h k ) , and ( I k , ϕ k , u k , h k ) ( k = 1 , ) satisfying the uniform estimates (3.62), one can solve problem (3.72).

Let

(3.73) ψ k + 1 = a δ δ 1 h k + 1 , φ k + 1 = ( h k + 1 ) 1 , f k + 1 = φ k + 1 ψ k + 1 = a δ δ 1 log h k + 1 ,

then direct calculation shows that problem (3.72) can be rewritten as follows:

(3.74) 1 c I t k + 1 + Ω I k + 1 = A r k , ϕ t k + 1 + u k ϕ k + 1 + ( γ 1 ) ϕ k + 1 div u k = 0 , φ k + 1 ( u t k + 1 + u k u k + ϕ k + 1 ) + a L u k + 1 = f k + 1 Q ( u k ) 1 c φ k + 1 0 S 2 A ˜ r k Ω d Ω d ν , f t k + 1 + l = 1 3 A l ( u k ) l f k + 1 + B ( u k ) f k + 1 + a δ ( φ k ) 1 φ k + 1 div u k = a δ φ k + 1 h k div u k + ( δ 1 ) ( φ k ) 1 φ k + 1 f k + 1 div u k , φ t k + 1 + u k φ k + 1 ( δ 1 ) ( φ k ) 1 ( φ k + 1 ) 2 div u k = 0 , ( I k + 1 , ϕ k + 1 , u k + 1 , f k + 1 , φ k + 1 ) t = 0 = ( I 0 η , ϕ 0 η , u 0 η , f 0 η , φ 0 η ) = I 0 , ϕ 0 + η , u 0 , 2 e a δ δ 1 ϕ 0 ϕ 0 + η , ( ϕ 0 + η ) 2 e ( x , ν , Ω ) R 3 × R + × S 2 , ( I k + 1 , ϕ k + 1 , u k + 1 , f k + 1 , φ k + 1 ) ( 0 , η , 0 , 0 , η 2 e ) as x + t 0 .

Next we are going to show the sequence of approximate solutions ( I k , ϕ k , u k , f k , φ k ) converges strongly to a limit ( I η , ϕ η , u η , f η , φ η ) . Let

I ¯ k + 1 = I k + 1 I k , ϕ ¯ k + 1 = ϕ k + 1 ϕ k , u ¯ k + 1 = u k + 1 u k , f ¯ k + 1 = f k + 1 f k , φ ¯ k + 1 = φ k + 1 φ k .

Then, it follows from (3.74) that

(3.75) 1 c I ¯ t k + 1 + Ω I ¯ k + 1 + σ a k + 1 + 0 S 2 σ s k + 1 d Ω d ν I ¯ k + 1 = L 1 , ϕ ¯ t k + 1 + u k ϕ ¯ k + 1 + u ¯ k ϕ k + ( γ 1 ) ( ϕ ¯ k + 1 div u k + ϕ k div u ¯ k ) = 0 , φ k + 1 ( u ¯ t k + 1 + u k u ¯ k + u ¯ k u k 1 ) + φ k + 1 ϕ ¯ k + 1 + φ ¯ k + 1 ϕ k + a L u ¯ k + 1 = φ ¯ k + 1 ( u t k + u k 1 u k 1 ) + f k + 1 Q ( u ¯ k ) + f ¯ k + 1 Q ( u k 1 ) 1 c φ ¯ k + 1 0 S 2 A ˜ r k Ω d Ω d ν + 1 c φ k L 2 , f ¯ t k + 1 + l = 1 3 A l ( u k ) l f ¯ k + 1 + B ( u k ) f ¯ k + 1 + a δ div u ¯ k = Γ 1 k + Γ 2 k + Γ 3 k , φ ¯ t k + 1 + u k φ ¯ k + 1 + u ¯ k φ k + ( 1 δ ) ( φ ¯ k div u k + φ k 1 div u ¯ k + Γ 4 k ) = 0 ,

where L i ( i = 1 , 2 ) and Γ i k ( i = 1 , , 4 ) are given by

L 1 = ( σ e k + 1 σ e k ) I k ( σ a k + 1 σ a k ) 0 S 2 ( σ s k + 1 σ s k ) I k + ν ν ( σ s k I ¯ k + I k ( σ s k + 1 σ s k ) ) d Ω d ν , L 2 = 1 c 0 S 2 ( ( σ ¯ e k + 1 σ ¯ e k ) σ ¯ a k + 1 I ¯ k + 1 I k ( σ ¯ a k + 1 σ ¯ a k ) ) Ω d Ω d ν 1 c 0 S 2 0 S 2 ν ν σ ¯ s I ¯ k + 1 σ ¯ s I ¯ k + 1 Ω d Ω d ν d Ω d ν , Γ 1 k = ( 1 δ ) ( f ¯ k div u k + f k 1 div u ¯ k + φ ¯ k + 1 f k h k div u k φ ¯ k f k 1 h k 1 div u k 1 ) , Γ 2 k = l = 1 3 ( A l ( u k ) A l ( u k 1 ) ) l f k ( B ( u k ) B ( u k 1 ) ) f k a δ ( φ ¯ k + 1 h k div u k φ ¯ k h k 1 div u k 1 ) , Γ 3 k = ( δ 1 ) ( f k + 1 div u k φ ¯ k + 1 h k f k div u k 1 φ ¯ k h k 1 + f ¯ k + 1 div u k + f k div u ¯ k ) , Γ 4 k = φ ¯ k + 1 φ k + 1 h k div u k φ ¯ k φ k h k 1 div u k 1 + φ ¯ k + 1 div u k φ ¯ k div u k 1 .

Now we are ready to give several estimates, which will be used later. To facilitate the discussion, we will adopt the following notations in the rest of this subsection

R k ( t ) = ( u k + φ k + 1 h k u k ) , F k ( t ) = u t k 3 + u k 1 u k 1 3 , G k ( t ) = ( f k + f k 1 ) h k 1 div u k 1 .

First, from Remark 3.2 at the end of this subsection, one has

Lemma 3.9

φ ¯ k + 1 L ( [ 0 , T ˆ ] ; H 3 ) , f ¯ k + 1 L ( [ 0 , T ˆ ] ; H 2 ) for k = 1 , 2 , .

Then multiplying (3.75) 5 by 2 φ ¯ k + 1 and integrating over R 3 , one arrives at

(3.76) d d t φ ¯ k + 1 2 2 C φ ¯ k + 1 2 ( R k ( t ) φ ¯ k + 1 2 + u ¯ k 3 φ k 6 + R k 1 ( t ) φ ¯ k 2 + u ¯ k 2 φ k 1 + φ ¯ k 2 u k ) .

Applying x ζ ( ζ = 1 ) to (3.75) 5 , multiplying the resulting equation by 2 x ζ φ ¯ k + 1 , and integrating over R 3 , one has

d d t x ζ φ ¯ k + 1 2 2 C φ ¯ k + 1 2 ( ( R k + h k u k 6 φ k + 1 6 + φ k + 1 h k 2 u k 3 + 2 u k 3 ) φ ¯ k + 1 2 + φ k 1 6 u ¯ k 3 + φ ¯ k + 1 2 φ k + 1 ψ k u k + u ¯ k 2 2 φ k 3 + φ ¯ k 2 ( 2 u k 3 + u k ) + u ¯ k 3 ( φ k 6 + φ k 1 6 ) + φ k 1 2 u ¯ k 2 + C φ ¯ k 2 ( R k 1 + φ k ψ k 1 u k 1 3 + 2 u k 1 3 + φ k h k 1 2 u k 1 3 + φ k 6 h k 1 u k 1 6 ) ) ,

which, along with (3.76), yields that

(3.77) d d t φ ¯ k + 1 1 2 σ ( u ¯ k 1 2 + φ ¯ k 1 2 ) + C σ 1 φ ¯ k + 1 1 2 ,

where σ ( 0 , 1 ) is a constant to be determined later.

For f ¯ k + 1 , multiplying (3.75) 4 by 2 f ¯ k + 1 and integrating over R 3 , one can obtain

(3.78) d d t f ¯ k + 1 2 2 C φ ¯ k + 1 1 2 + C σ 1 f ¯ k + 1 2 2 + σ ( u ¯ k 1 2 + φ ¯ k 1 2 + f ¯ k 2 2 ) ,

where one has used

Γ 1 k 2 C ( f ¯ k 2 u k + u ¯ k 2 f k 1 + f k φ ¯ k + 1 2 h k u k + f k 1 h k 1 u k 1 φ ¯ k 2 ) , Γ 2 k 2 C ( f k 3 u ¯ k 6 + f k u ¯ k 2 + φ ¯ k + 1 2 h k div u k 3 + φ ¯ k 2 h k 1 div u k 1 3 ) , Γ 3 k 2 C ( f k + 1 φ ¯ k + 1 2 h k div u k + f k h k 1 div u k 1 φ ¯ k 2 + u k f ¯ k + 1 2 + f k u ¯ k 2 ) .

For ϕ ¯ k + 1 , multiplying (3.75) 2 by 2 ϕ ¯ k + 1 and integrating over R 3 , one has

(3.79) d d t ϕ ¯ k + 1 2 2 C ϕ ¯ k + 1 2 ( u k ϕ ¯ k + 1 2 + u ¯ k 2 ϕ k 3 + u ¯ k 2 ϕ k ) .

Applying x ζ ( ζ = 1 ) to (3.75) 2 and integrating the resulting equation over R 3 , one obtains

d d t x ζ ϕ ¯ k + 1 2 2 C ϕ ¯ k + 1 2 ( u k ϕ ¯ k + 1 2 + u ¯ k 2 2 ϕ k 3 + ϕ k u ¯ k 2 + 2 u k 3 ϕ ¯ k + 1 6 + u ¯ k 2 ϕ k + ϕ k 2 u ¯ k 2 ) ,

which, along with (3.79), yields that

(3.80) d d t ϕ ¯ k + 1 1 2 C σ 1 ϕ ¯ k + 1 1 2 + σ u ¯ k 1 2 .

For I ¯ k + 1 , multiplying (3.75) 1 by I ¯ k + 1 and integrating over R + × S 2 × R 3 , one has

(3.81) d d t I ¯ k + 1 L 2 ( R + × S 2 ; L 2 ) 2 C 0 S 2 σ e k + 1 σ e k 2 I ¯ k + 1 2 + I k σ a k + 1 σ a k 2 I ¯ k + 1 2 + 0 S 2 ( σ s I k ( ϕ k + 1 ) κ ( ϕ k ) κ 2 I ¯ k + 1 2 + ν ν σ ¯ s ( ( ϕ k ) κ I ¯ k 2 + I k ( ϕ k + 1 ) κ ( ϕ k ) κ 2 ) I ¯ k + 1 2 d Ω d ν d Ω d ν C ϕ ¯ k + 1 2 2 + σ I ¯ k L 2 ( R + × S 2 ; L 2 ) 2 + C σ 1 I ¯ k + 1 L 2 ( R + × S 2 ; L 2 ) 2 ,

where one has used (2.6).

For u ¯ k + 1 , multiplying (3.75) 3 by 2 u ¯ k + 1 and integrating over R 3 , one has

(3.82) d d t φ k + 1 u ¯ k + 1 2 2 + 2 a α u ¯ k + 1 2 2 + 2 a ( α + β ) div u ¯ k + 1 2 2 = ( φ t k + 1 u ¯ k + 1 2 φ k + 1 ( u k u ¯ k + u ¯ k u k 1 ) 2 ( φ ¯ k + 1 ( u t k + u k 1 u k 1 ) + φ k + 1 ϕ ¯ k + 1 + φ ¯ k + 1 ϕ k ) + 2 ( f k + 1 Q ( u ¯ k ) + f ¯ k + 1 Q ( u k 1 ) ) 2 c φ ¯ k + 1 0 S 2 A ˜ r k Ω d Ω d ν + 2 c φ k L 2 u ¯ k + 1 d x i = 1 5 J i .

Using Hölder’s inequality, (3.73), and equation (3.74) 5 , we estimate terms J i ( i = 1 , , 5 ) as follows:

J 1 = ( φ t k + 1 u ¯ k + 1 2 φ k + 1 ( u k u ¯ k + u ¯ k u k 1 ) ) u ¯ k + 1 d x C φ k + 1 u ¯ k + 1 2 u ¯ k + 1 2 u k φ k + 1 1 2 + φ k + 1 u ¯ k + 1 2 ( u k + φ k + 1 h k u k ) + φ k + 1 1 2 ( u k u ¯ k 2 + u ¯ k 6 u k 1 3 ) , J 2 = 2 ( φ ¯ k + 1 ( u t k + u k 1 u k 1 ) + φ k + 1 ϕ ¯ k + 1 + φ ¯ k + 1 ϕ k ) u ¯ k + 1 d x C φ ¯ k + 1 2 u ¯ k + 1 2 ( F k ( t ) + ϕ k ) + φ k + 1 1 2 φ k + 1 u ¯ k + 1 2 ϕ ¯ k + 1 2 , J 3 = 2 ( f k + 1 Q ( u ¯ k ) + f ¯ k + 1 Q ( u k 1 ) ) u ¯ k + 1 d x C φ k + 1 u ¯ k + 1 2 φ k + 1 1 2 ψ k + 1 u ¯ k 2 + f ¯ k + 1 2 u k 1 2 u ¯ k + 1 2 , J 4 = 2 c 0 S 2 φ ¯ k + 1 A ˜ r k Ω u ¯ k + 1 d Ω d ν d x C φ ¯ k + 1 2 u ¯ k + 1 6 A ˜ r k L 1 ( R + × S 2 ; L 3 ) , J 5 = 2 c φ k L 2 u ¯ k + 1 d x = 2 c ( φ k + 1 φ ¯ k + 1 ) L 2 u ¯ k + 1 d x C φ k + 1 1 2 φ k + 1 u ¯ k + 1 2 L 2 2 + φ ¯ k + 1 2 u ¯ k + 1 6 L 2 3 .

Substituting the estimates for J i ( i = 1 , , 5 ) into (3.82) and using Young’s inequality, one obtains

(3.83) d d t φ k + 1 u ¯ k + 1 2 2 + a α u ¯ k + 1 2 2 σ u ¯ k 2 2 + C σ 1 φ k + 1 u ¯ k + 1 2 2 + C ( ϕ ¯ k + 1 1 2 + φ ¯ k + 1 1 2 + f ¯ k + 1 2 2 + I ¯ k + 1 L 2 ( R + × S 2 ; L 2 ) 2 ) ,

where one has used

(3.84) A ˜ r k L 1 ( R + × S 2 ; L 3 ) C ( σ ¯ e k + 1 L 1 ( R + × S 2 ; L 3 ) + ( 1 + σ ¯ a k + 1 L 2 ( R + × S 2 ; L ) ) I k + 1 L 2 ( R + × S 2 ; L 3 ) ) C , L 2 p C ( σ ¯ e k + 1 σ ¯ e k L 1 ( R + × S 2 ; L p ) + I k L 2 ( R + × S 2 ; L ) σ ¯ a k + 1 σ ¯ a k L 2 ( R + × S 2 ; L p ) + ( 1 + σ ¯ a k + 1 L 2 ( R + × S 2 ; L ) ) I ¯ k + 1 L 2 ( R + × S 2 ; L p ) ) , p = 2 , 3 .

Next multiplying (3.75) 3 by 2 u ¯ t k + 1 and integrating over R 3 , one has

(3.85) d d t ( α u ¯ k + 1 2 2 + ( α + β ) div u ¯ k + 1 2 2 ) + 2 φ k + 1 u ¯ t k + 1 2 2 = 2 ( φ k + 1 ( u k u ¯ k + u ¯ k u k 1 ) φ ¯ k + 1 ( u t k + u k 1 u k 1 ) φ k + 1 ϕ ¯ k + 1 φ ¯ k + 1 ϕ k + f k + 1 Q ( u ¯ k ) + f ¯ k + 1 Q ( u k 1 ) 1 c φ ¯ k + 1 0 S 2 A ˜ r k Ω d Ω d ν + 1 c φ k L 2 u ¯ t k + 1 d x i = 6 11 J i .

Then the terms J i ( i = 6 , , 9 ) on the right-hand side of (3.85) can be estimated as follows:

J 6 = 2 φ k + 1 ( u k u ¯ k + u ¯ k u k 1 ) u ¯ t k + 1 d x C φ k + 1 1 2 φ k + 1 u ¯ t k + 1 2 ( u k u ¯ k 2 + u ¯ k 6 u k 1 3 ) , J 7 = 2 φ ¯ k + 1 ( u t k + u k 1 u k 1 ) u ¯ t k + 1 d x = 2 d d t φ ¯ k + 1 ( u t k + u k 1 u k 1 ) u ¯ k + 1 d x + 2 ( φ ¯ k + 1 ( u t k + u k 1 u k 1 ) ) t u ¯ k + 1 d x 2 d d t φ ¯ k + 1 ( u t k + u k 1 u k 1 ) u ¯ k + 1 + C u ¯ k + 1 6 ( F k ( t ) ( u k φ ¯ k + 1 2 + u ¯ k 3 φ k 6 + u ¯ k 2 φ k 1 + φ ¯ k + 1 2 R k ( t ) + φ ¯ k 2 R k 1 ( t ) ) + φ ¯ k + 1 3 ( u t t k 2 + u t k 1 2 u k 1 + u k 1 u t k 1 2 ) ) , J 8 = 2 ( φ k + 1 ϕ ¯ k + 1 + φ ¯ k + 1 ϕ k ) u ¯ t k + 1 d x = 2 d d t φ ¯ k + 1 ϕ k u ¯ k + 1 d x 2 ( φ k + 1 ϕ ¯ k + 1 u ¯ t k + 1 ( φ ¯ k + 1 ϕ k ) t u ¯ k + 1 ) d x 2 d d t φ ¯ k + 1 ϕ k u ¯ k + 1 d x + C φ k + 1 1 2 φ k + 1 u ¯ t k + 1 2 ϕ ¯ k + 1 2 + C u ¯ k + 1 2 ( φ ¯ k + 1 2 ϕ t k 3 + ϕ k 3 ( u k φ ¯ k + 1 2 + u ¯ k 3 φ k 6 + φ k 1 u ¯ k 2 + φ ¯ k + 1 2 R k ( t ) + φ ¯ k 2 u k + φ ¯ k 2 R k 1 ( t ) ) ) , J 9 = 2 ( f k + 1 Q ( u ¯ k ) + f ¯ k + 1 Q ( u k 1 ) ) u ¯ t k + 1 d x = 2 d d t f ¯ k + 1 Q ( u k 1 ) u ¯ k + 1 d x + 2 ( f k + 1 Q ( u ¯ k ) u ¯ t k + 1 ( f ¯ k + 1 Q ( u k 1 ) ) t u ¯ k + 1 ) d x 2 d d t f ¯ k + 1 Q ( u k 1 ) u ¯ k + 1 d x + C φ k + 1 1 2 φ k + 1 u ¯ t k + 1 2 ψ k + 1 u ¯ k 2 + C u ¯ k + 1 2 ( u k 1 3 u ¯ k 2 f k 3 + u k 1 u ¯ k 2 + 2 u k 1 3 u ¯ k 2 + f ¯ k + 1 2 ( u k u k 1 + u k 6 2 u k 1 6 + u k 6 u k 1 6 + u t k 1 3 ) + u k 1 3 ( φ ¯ k + 1 6 h k 2 u k 3 + φ ¯ k 6 h k 1 2 u k 1 3 + u k ( f ¯ k 2 + f ¯ k + 1 2 ) + u ¯ k 2 ( f k + f k 1 ) + φ ¯ k + 1 2 G k + 1 ( t ) + φ ¯ k 2 G k ( t ) ) ) ,

where one has used the equations (3.75) 4 –(3.75) 5 .

For the terms J i ( i = 10 , 11 ) , according to (3.75) 5 and Hölder’s inequality, one has

J 10 = 2 c 0 S 2 φ ¯ k + 1 A ˜ r k Ω u ¯ t k + 1 d Ω d ν d x = 2 c d d t 0 S 2 φ ¯ k + 1 A ˜ r k Ω u ¯ k + 1 d Ω d ν d x 2 c 0 S 2 ( φ ¯ k + 1 A ˜ r k ) t Ω u ¯ k + 1 d Ω d ν d x 2 c d d t 0 S 2 φ ¯ k + 1 A ˜ r k Ω u ¯ k + 1 d Ω d ν d x , + C u ¯ k + 1 2 ( A ˜ r k L 1 ( R + × S 2 ; L 3 ) ( u k φ ¯ k + 1 2 + u ¯ k 3 φ k 6 + φ ¯ k 6 u k 3 + φ ¯ k + 1 2 R k ( t ) + φ ¯ k 2 R k 1 ( t ) + φ k 1 u ¯ k 2 ) + φ ¯ k + 1 2 ( A ˜ r k ) t L 1 ( R + × S 2 ; L 3 ) ) , J 11 = 2 c φ k L 2 u ¯ t k + 1 d x = 2 c ( φ ¯ k + 1 + φ k + 1 ) L 2 u ¯ t k + 1 d x = 2 c d d t φ ¯ k + 1 L 2 u ¯ k + 1 d x + 2 c ( ( φ ¯ k + 1 L 2 ) t u ¯ k + 1 + φ k + 1 L 2 u ¯ t k + 1 ) d x 2 c d d t φ ¯ k + 1 L 2 u ¯ k + 1 d x + C u ¯ k + 1 6 L 2 3 ( u k φ ¯ k + 1 2 + u ¯ k 3 φ k 6 + φ ¯ k 2 u k + φ k 1 u ¯ k 2 + φ ¯ k + 1 2 R k ( t ) + φ ¯ k 2 R k 1 ( t ) ) + C φ ¯ k + 1 2 u ¯ k + 1 6 ( L 2 ) t 3 + C φ k + 1 1 2 φ k + 1 u ¯ t k + 1 2 L 2 2 .

From the definition of L 2 and A ˜ r k , one arrives at

(3.86) ( L 2 ) t 3 C ( ( σ ¯ e k + 1 σ ¯ e k ) t L 1 ( R + × S 2 ; L 3 ) + ( σ ¯ a k + 1 ) t L 2 ( R + × S 2 ; L 3 ) I ¯ k + 1 L 2 ( R + × S 2 ; L ) + ( 1 + σ ¯ a k + 1 L 2 ( R + × S 2 ; L ) ) I ¯ t k + 1 L 2 ( R + × S 2 ; L 3 ) + I t k L 2 ( R + × S 2 ; L 3 ) σ ¯ a k + 1 σ ¯ a k L 2 ( R + × S 2 ; L ) + I k L 2 ( R + × S 2 ; L ) ( σ ¯ a k + 1 σ ¯ a k ) t L 2 ( R + × S 2 ; L 3 ) ) C , ( A ˜ r k ) t L 1 ( R + × S 2 ; L 3 ) C ( ( σ ¯ e k + 1 ) t L 1 ( R + × S 2 ; L 3 ) + ( σ ¯ a k + 1 ) t L 2 ( R + × S 2 ; L 3 ) I t k + 1 L 2 ( R + × S 2 ; L ) + ( 1 + σ ¯ a k + 1 L 2 ( R + × S 2 ; L ) ) I t k + 1 L 2 ( R + × S 2 ; L 3 ) ) C .

Substituting the estimates for J i ( i = 6 , , 11 ) into (3.85) and using (3.84) and (3.86), one deduces that

(3.87) d d t ( a α u ¯ k + 1 2 2 + a ( α + β ) div u ¯ k + 1 2 2 ) + φ k + 1 u ¯ t k + 1 2 2 J 12 + E σ k ( t ) u ¯ k + 1 2 2 + C u ¯ k 2 2 + σ ( φ ¯ k 1 2 + f ¯ k 2 2 ) + C ( ϕ ¯ k + 1 1 2 + φ ¯ k + 1 1 2 + f ¯ k + 1 2 2 + I ¯ k + 1 L 2 ( R + × S 2 ; L 2 ) 2 ) ,

where

J 12 = 2 d d t φ ¯ k + 1 ( u t k + u k 1 u k 1 + ϕ k ) f ¯ k + 1 Q ( u k 1 ) 1 c 0 S 2 φ ¯ k + 1 A ˜ r k Ω d Ω d ν + 1 c φ ¯ k + 1 L 2 u ¯ k + 1 d x ,

and E σ k ( t ) satisfies

0 t E σ k ( t ) d s C + C σ 1 t for 0 t T ˆ .

It follows from (3.77)–(3.78), (3.80)–(3.81), (3.83), and (3.87) that

(3.88) d d t ( φ k + 1 u ¯ k + 1 2 2 + ϕ ¯ k + 1 1 2 + φ ¯ k + 1 1 2 + f ¯ k + 1 2 2 + I ¯ k + 1 L 2 ( R + × S 2 ; L 2 ) 2 + a α ν u ¯ k + 1 2 2 + a ( α + β ) ν div u ¯ k + 1 2 2 ) + ( a α u ¯ k + 1 2 2 + ν φ k + 1 u ¯ t k + 1 2 2 ) ν J 12 + C ( ν + σ ) a α u ¯ k 2 2 + C ( N σ k ( t ) + ν ) ( φ k + 1 u ¯ k + 1 2 2 + a α u ¯ k + 1 2 2 + ϕ ¯ k + 1 1 2 + φ ¯ k + 1 1 2 + f ¯ k + 1 2 2 + I ¯ k + 1 L 2 ( R + × S 2 ; L 2 ) 2 ) + C σ ( ϕ ¯ k 1 2 + 2 u ¯ k 2 2 + ( 1 + ν ) ( φ ¯ k 1 2 + f ¯ k 2 2 + I ¯ k L 2 ( R + × S 2 ; L 2 ) 2 ) ) ,

where ν , σ ( 0 , 1 ) are sufficiently small constants, which will be determined later and N σ k ( t ) satisfies

(3.89) 0 t N σ k ( s ) d s C ( 1 + σ 1 t ) for 0 t T ˆ .

Conversely, for J 12 , one has

(3.90) 0 t J 12 d s C ( u ¯ k + 1 2 2 + φ ¯ k + 1 1 2 + f ¯ k + 1 2 2 ) .

For 2 u ¯ k , according to (3.75) 3 and Lemma A.6 (Appendix A), one obtains

(3.91) u ¯ k D 2 C ( φ k u ¯ t k 2 + u ¯ k 1 2 + ϕ ¯ k 1 + φ ¯ k 1 + I ¯ k L 2 ( R + × S 2 ; L 2 ) + f ¯ k 2 ) .

Denote

Γ k + 1 ( t , ν ) = sup 0 s t ϕ ¯ k + 1 ( s ) 1 2 + sup 0 s t φ ¯ k + 1 ( s ) 1 2 + sup 0 s t I ¯ k + 1 ( s ) L 2 ( R + × S 2 ; L 2 ) 2 + sup 0 s t f ¯ k + 1 ( s ) 2 2 + sup 0 s t φ k + 1 u ¯ k + 1 ( s ) 2 2 + a ν sup 0 s t ( α u ¯ k + 1 ( s ) 2 2 + ( α + β ) div u ¯ k + 1 ( s ) 2 2 ) .

According to the aforementioned estimates (3.88)–(3.91) and the Gronwall inequality, one concludes that

Γ k + 1 ( t , ν ) + 0 t ( a α u ¯ k + 1 2 2 + ν φ k + 1 u ¯ t k + 1 2 2 ) d s C 0 t ( a α ( σ + ν ) ( u ¯ k 2 2 + u ¯ k 1 2 2 ) + σ φ k u ¯ t k 2 2 ) d s exp ( C + C σ 1 t ) + C σ t Γ k ( t , ν ) exp ( C + C σ 1 t ) .

Choose ν 0 , σ 0 ( 0 , 1 ) and 0 < T T ˆ small enough such that

C σ 0 exp C ν 0 32 , C ν 0 exp C 1 32 , ( T + 1 ) exp ( C σ 0 1 T ) 4 .

Then one can easily obtain

(3.92) k = 1 Γ k + 1 ( T , ν 0 ) + 0 T ( a α u ¯ k + 1 2 2 + ν 0 φ k + 1 u ¯ t k + 1 2 2 ) d s C < ,

which, along with the uniform (with respect to k ) estimates (3.62), yields that

(3.93) lim k ( I ¯ k + 1 L 2 ( R + × S 2 ; L 2 ) + f ¯ k + 1 6 + φ ¯ k + 1 ) = 0 .

Then there exists a subsequence (still denoted by ( I k , ϕ k , u k , φ k , f k ) ) and limit functions ( I η , ϕ η , u η , φ η , f η ) such that for any s [ 1 , 3 ) ,

(3.94) I k I η in L 2 ( R + × S 2 ; L ( [ 0 , T ] ; H s ( R 3 ) ) ) , ϕ k η ϕ η η in L ( [ 0 , T ] ; H s ( R 3 ) ) , f k f η in L ( [ 0 , T ] ; L 6 ( R 3 ) ) , φ k φ η in L ( [ 0 , T ] ; L ( R 3 ) ) , u k u η in L ( [ 0 , T ] ; D 1 ( R 3 ) D s ( R 3 ) ) .

Conversely, by virtue of the uniform estimates in (3.62), there exists a subsequence (still denoted by ( I k , ϕ k , u k , φ k , ψ k , f k ) ) converging to the limit ( I η , ϕ η , u η , φ η , ψ η , f η ) in the weak or weak sense. According to the lower semi-continuity of norms, the corresponding estimates in (3.62) for the limit function ( I η , ϕ η , u η , φ η , ψ η , f η ) still hold except those weighted estimates for u η . Therefore, one concludes that ( I η , ϕ η , u η , φ η , f η ) is a weak solution in the sense of distributions to the following Cauchy problem:

(3.95) 1 c I t η + Ω I η = A r η , ϕ t η + u η ϕ η + ( γ 1 ) ϕ η div u η = 0 , φ η ( u t η + u η u η + ϕ η ) = a L u η + f η Q ( u η ) 1 c φ η 0 S 2 A ¯ r η Ω d Ω d ν , f t η + l = 1 3 A l ( u η ) l f + B ( u η ) f + a δ div u η = 0 , φ t η + u η φ η ( δ 1 ) φ η div u η = 0 , ( I η , ϕ η , u η , f η , φ ) t = 0 = ( I 0 η , ϕ 0 η , u 0 η , f 0 η , φ 0 η ) = ( I 0 , ϕ 0 + η , u 0 , ( ϕ 0 + η ) 2 e ψ 0 , ( ϕ 0 + η ) 2 e ) ( x , ν , Ω ) R 3 × R + × S 2 , ( I η , ϕ η , u η , f η , φ η ) ( 0 , η , 0 , 0 , ( η ) 2 e ) as x + for t 0 .

It should be emphasized that the conclusions obtained earlier are not sufficient to show the existence of desired strong solution to the Cauchy problem (3.69).

To this end, one first shows the strong convergence of ψ k . In fact, from (3.62), one obtains

ψ k + 1 ψ k 6 = f k + 1 φ k f k φ k + 1 φ k + 1 φ k 6 C ( φ k f ¯ k + 1 6 + f k 6 φ ¯ k + 1 ) ,

which, along with (3.94), yields that

(3.96) ψ k ψ η in L ( [ 0 , T ] ; L 6 ( R 3 ) ) as k .

Then, according to (3.73), (3.94), and (3.96), one has

f k φ η ψ η 6 C ( φ k φ η ψ k 6 + ψ k ψ η 6 φ η ) 0 as k ,

which implies that

(3.97) f η ( t , x ) = ψ η φ η ( t , x ) a.e. on [ 0 , T ] × R 3 .

Second, one also needs to show the following relationship between ϕ η and ( f η , φ η , ψ η ) :

(3.98) f η = 2 a e δ δ 1 log ϕ η , φ η = ( ϕ η ) 2 e , ψ η = a δ δ 1 ( ϕ η ) 2 e .

Indeed, (3.98) can be obtained via the same argument as used in [32].

Moreover, denote h η = ( φ η ) 1 , it follows from (3.94), the uniform positivity for ( φ k , φ η ) , and the upper bounds of the norms of ( ϕ η , u η , φ η , f η ) that

(3.99) h k 2 u k h η 2 u η in L ( [ 0 , T ] ; H 1 ) , h k ( u k , u t k ) h η ( u η , u t η ) in L ( [ 0 , T ] ; L 2 ) , h k 2 u k h η 2 u η in L 2 ( [ 0 , T ] ; D 1 D 2 ) , ( h k 2 u k ) t ( h η 2 u η ) t in L 2 ( [ 0 , T ] ; L 2 ) .

Thus, the corresponding weighted estimates for u η shown in (3.62) still hold for the limit functions. Based on the estimates (3.62), the strong/weak convergence shown in (3.94), (3.96), and (3.99) and the relations (3.97) and (3.98), one can easily show that

I η , ϕ η , u η , h η = ( ϕ η ) 2 e , ψ η = a δ δ 1 ( ϕ η ) 2 e

is a weak solution in the sense of distributions to the Cauchy problem (3.69), satisfying

I η L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 3 ) ) , I t η L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 2 ) ) , ϕ η η L ( [ 0 , T ] ; H 3 ) , ψ η L ( [ 0 , T ] ; D 1 D 2 ) , 2 3 η 2 e < φ η L D 1 , 6 D 2 , 3 D 3 , f η L L 6 D 1 , 3 D 2 , u η L ( [ 0 , T ] ; H 3 ) L 2 ( [ 0 , T ] ; H 4 ) , u t η L ( [ 0 , T ] ; H 1 ) L 2 ( [ 0 , T ] ; D 2 ) , u t t η L 2 ( [ 0 , T ] ; L 2 ) , t 1 2 u η L ( [ 0 , T ] ; D 4 ) , t 1 2 u t η L ( [ 0 , T ] ; D 2 ) L 2 ( [ 0 , T ] ; D 3 ) , t 1 2 u t t η L ( [ 0 , T ] ; L 2 ) L 2 ( [ 0 , T ] ; D 1 ) .

Step 2: Uniqueness and time continuity. The uniqueness and time continuity follows easily from the same procedure as in Lemma 3.1. Finally, Theorem 3.3 is proved.□

Remark 3.2

We conclude this subsection by giving the proof of Lemma 3.9.

Proof

Let X ( x ) C 0 ( R 3 ) be a truncation function satisfying

(3.100) 0 X ( x ) 1 and X ( x ) = 1 if 0 x 1 , 0 if x 2 .

Define, for any R > 0 , X R ( x ) = X x R , φ ¯ k + 1 , R = φ ¯ k + 1 X R . Then from (3.75) 5 ,

(3.101) φ ¯ t k + 1 , R + u k φ ¯ k + 1 , R φ ¯ k + 1 u k X R + u ¯ k φ k X R = ( 1 δ ) ( φ ¯ k div u k + φ k 1 div u ¯ k + Γ 4 k ) X R .

Multiplying (3.101) by 2 φ ¯ k + 1 , R and integrating over R 3 , one has

(3.102) d d t φ ¯ k + 1 , R 2 C ( ( u k + φ k + 1 h k div u k ) φ ¯ k + 1 , R 2 + φ ¯ k u k 1 2 + φ ¯ k φ k h k 1 u k 1 2 + u ¯ k 3 φ k 6 + φ ¯ k u k 2 + φ k 1 u ¯ k 2 ) C ˜ φ ¯ k + 1 , R 2 + C ˜ ,

where C ˜ > 0 is a generic constant depending on C and η but independent of R . Then applying the Gronwall inequality to (3.102), one obtains

φ ¯ k + 1 , R ( t ) 2 C ˜ T ˆ exp ( C ˜ T ˆ ) for ( t , R ) [ 0 , T ˆ ] × [ 0 , ) .

It follows from Fatou’s lemma (Lemma A.2 in Appendix A) that

(3.103) φ ¯ k + 1 L ( [ 0 , T ˆ ] ; L 2 ) ,

which, along with

φ ¯ k + 1 L ( [ 0 , T ˆ ] ; L D 1 , 6 D 2 , 3 D 3 ) ,

yields that

φ ¯ k + 1 L ( [ 0 , T ˆ ] ; H 3 ) .

Similarly, one can also show that f ¯ k + 1 L ( [ 0 , T ˆ ] ; H 2 ) .

3.6 Limit from the nonvacuum flow to the flow with far field vacuum

Based on the local-in-time uniform estimates in (3.62), now we are ready to give the proof of Theorem 3.1.

Proof

The proof will be divided into four steps.

Step 1: The locally uniform positivity of ϕ . For η ( 0 , 1 ) , one sets

ϕ 0 η = ϕ 0 + η , ψ 0 η = a δ δ 1 ( ϕ 0 + η ) 2 e , h 0 η = ( ϕ 0 + η ) 2 e , A ¯ r 0 η = A ¯ r ( t = 0 , x , ν , ϕ 0 η , I 0 η , I 0 η ) .

Then the initial compatibility conditions of the perturbed initial data can be given as follows:

(3.104) u 0 = ( ϕ 0 + η ) e g 1 η , L u 0 = ( ϕ 0 + η ) 2 e g 2 η , a ( ϕ 0 + η ) 2 e L u 0 + 1 c 0 S 2 A ¯ r 0 η Ω d Ω d ν = ( ϕ 0 + η ) e g 3 η ,

where

g 1 η = ϕ 0 e ( ϕ 0 + η ) e g 1 , g 2 η = ϕ 0 2 e ( ϕ 0 + η ) 2 e g 2 , g 3 η = ϕ 0 3 e ( ϕ 0 + η ) 3 e g 3 a η ϕ 0 2 e ϕ 0 e L u 0 ( ϕ 0 + η ) + 1 c 0 S 2 ϕ 0 e ϕ 0 2 e ( ϕ 0 + η ) 2 e A ¯ r 0 η A ¯ r 0 Ω d Ω d ν .

It follows from (3.5) and (3.6) that there exists a η 1 > 0 such that if 0 < η < η 1 , then

(3.105) 1 + η + ϕ 0 η η 3 + ψ 0 η D 1 D 2 + I 0 L 2 ( R + × S 2 ; H 3 ) + u 0 3 + g 1 η 2 + g 2 η 2 + g 3 η 2 + ( h 0 η ) 1 L D 1 , 6 D 2 , 3 D 3 + h 0 η / h 0 η L L 6 D 1 , 3 D 2 c ¯ 0 ,

where c ¯ 0 is a positive constant independent of η . Thus, for the initial data ( I 0 η , ϕ 0 η , u 0 η , ψ 0 η ) , the problem (3.69) admits a unique classical solution ( I η , ϕ η , u η , ψ η ) satisfying the local uniform estimates in (3.62) with c 0 replaced by c ¯ 0 , and the life span T is also independent of η .

Moreover, one has the following:

Lemma 3.10

For any R 0 > 0 and η ( 0 , 1 ) , there exists a constant a R 0 > 0 such that

(3.106) ϕ η ( t , x ) a R 0 > 0 , ( t , x ) [ 0 , T ] × B R 0 ,

where a R 0 is independent of η .

Proof

One only needs to consider the case when R 0 is suitable large.

First, due to

ϕ 0 H 3 , ψ 0 = a δ δ 1 ϕ 0 2 e D 1 D 2 ,

one can see that ϕ 0 2 e L , which means that the initial vacuum occurs if and only if in the far field. Then there exists a constant C R independent of η such that for any R > 2 ,

(3.107) ϕ 0 η ( x ) C R + η > 0 , x B R .

Second, let x ( t ; x 0 ) be the solution to the following initial value problem:

(3.108) d d t x ( t ; x 0 ) = u ( t , x ( t , x 0 ) ) , x ( 0 ; x 0 ) = x 0 .

Let B ( t , R ) = { x ( t ; x 0 ) x 0 B R } be the closed regions that are the images of B R under the flow map (3.108). It follows from (3.2) 2 that

(3.109) ϕ η ( t , x ) = ϕ 0 η ( x 0 ) exp ( γ 1 ) 0 t div u η ( s ; x ( s ; x 0 ) ) d s .

According to (3.62), one has

(3.110) 0 t div u η ( t , x ( t ; x 0 ) ) d s 0 t u η 2 d s t 1 2 0 t u η 2 2 d s 1 2 c 5 T 1 2 .

It thus follows from (3.107) and (3.109) and (3.110) that

(3.111) ϕ η ( t , x ) C ( C R + η ) > 0 , x B ( t , R ) ,

where C = exp ( γ 1 ) c 3 T 1 2 .

Moreover, it follows from (3.60)–(3.62) and (3.108) that

x 0 x = x 0 x ( t ; x 0 ) 0 t u η ( τ , x ( τ ; x 0 ) ) d τ c 5 t 1 R / 2 ,

for all ( t , x ) [ 0 , T ] × B R , which implies that B R / 2 B ( t , R ) . Thus, the desired conclusion can be achieved by taking R = 2 R 0 and a R 0 = C C R .□

Step 2: Existence. First, by virtue of the uniform (with respect to η ) estimates in (3.62), one can see that there exists a subsequence (still denoted by ( I η , ϕ η , u η , ψ η ) ) converging to a limit ( I , ϕ , u , ψ ) in the following weak or weak sense:

(3.112) I η I , in L 2 ( R + × S 2 ; L ( [ 0 , T ] ; H 3 ) ) , I t η I t , in L 2 ( R + × S 2 ; L ( [ 0 , T ] ; H 2 ) ) , ( ϕ η η , u η ) ( ϕ , u ) in L ( [ 0 , T ] ; H 3 ) , u η u in L 2 ( [ 0 , T ] ; H 4 ) , φ η φ in L ( [ 0 , T ] ; L D 1 , 6 D 2 , 3 D 3 ) , f η f in L ( [ 0 , T ] ; L L 6 D 1 , 3 D 2 ) , ψ η ψ in L ( [ 0 , T ] ; D 1 D 2 ) , ϕ t η ϕ t in L ( [ 0 , T ] ; H 2 ) , ( u t η , ψ t η ) ( u t , ψ t ) in L ( [ 0 , T ] ; H 1 ) , φ t η φ t in L ( [ 0 , T ] ; L 6 D 1 , 3 D 2 ) , f t η f t in L ( [ 0 , T ] ; L 3 D 1 ) .

According to the lower semi-continuity of norms, then the corresponding estimates (3.62) still hold for the limit functions ( I , ϕ , u , ψ ) except those weighted estimates for u .

Second, Lemmas A.3 (Appendix A) and 3.10 guarantee that there exists a subsequence (still denoted by ( I η , ϕ η , u η , ψ η ) ) satisfying

(3.113) I η I in L 2 ( R + × S 2 ; C ( [ 0 , T ] ; H 2 ( B R ) ) ) , ( ϕ η η , ψ η , φ η ) ( ϕ , ψ , φ ) in C ( [ 0 , T ] ; H 1 ( B R ) ) , u η u in C ( [ 0 , T ] ; H 2 ( B R ) ) .

Arguing as in the proof of (3.98), one can also verify that

(3.114) f = 2 a e δ δ 1 ϕ ϕ , φ = ϕ 2 e , ψ = a δ δ 1 ϕ 2 e .

Conversely, according to the estimates (3.62) except those weighted estimates for u , Lemma 3.10, the weak or weak convergence in (3.112) and the strong convergence in (3.113), one can show that for h = φ 1 ,

(3.115) h η ( u η , u t η ) h ( u , u t ) in L ( [ 0 , T ] ; L 2 ) , h η 2 u η h 2 u in L ( [ 0 , T ] ; H 1 ) , h η 2 u η h 2 u in L 2 ( [ 0 , T ] ; D 1 D 2 ) , ( h η 2 u η ) t ( h 2 u ) t in L 2 ( [ 0 , T ] ; L 2 ) .

Then the corresponding weighted estimates for u shown in (3.62) still hold for the limit functions. Therefore, one can conclude that ( I , ϕ , u , ψ ) is the weak solution in the sense of distributions to the Cauchy problem (3.2)–(3.4).

Step 3: Uniqueness. The uniqueness can be easily obtained by the same argument as in Theorem 3.3.

Step 4: Time continuity. First, the time continuity of ( I , ϕ , u , ψ ) follows from the similar argument used in Lemma 3.1.

For the velocity u , the uniform estimates obtained earlier and the Sobolev embedding theorem imply that

(3.116) u C ( [ 0 , T ] ; H 2 ) C ( [ 0 , T ] ; weak- H 3 ) and ϕ e u C ( [ 0 , T ] ; L 2 ) .

Then it follows from (3.2) 3 that

φ u t L 2 ( [ 0 , T ] ; H 2 ) , ( φ u t ) t L 2 ( [ 0 , T ] ; L 2 ) ,

which, along with the Lemma A.3, yields that φ u t C ( [ 0 , T ] ; H 1 ) .

Thus, the continuity of u C ( [ 0 , T ] ; H 3 ) can be obtained by using the classical elliptic estimates and

a L u = φ u t + u u + ϕ ψ Q ( u ) 1 c 0 S 2 A ¯ r Ω d Ω d ν .

For h 2 u , due to

h 2 u L ( [ 0 , T ] ; H 1 ) L 2 ( [ 0 , T ] ; D 2 ) , ( h 2 u ) t L 2 ( [ 0 , T ] ; L 2 ) ,

and the Soblev embedding theorem, one obtains

h 2 u C ( [ 0 , T ] ; H 1 ) ,

which, along with (3.2) 3 , yields that u t C ( [ 0 , T ] ; H 1 ) .

The proof of Theorem 3.1 is complete.□

3.7 Proof of Theorem 2.1

With Theorem 3.1 at hand, now we turn to the proof of Theorem 2.1.

Step 1: The local well-posedness of regular solutions. First, Theorem 3.1 guarantees that there exists a time T > 0 such that the problem (3.2)–(3.4) has a unique regular solution ( I , ϕ , u , ψ ) satisfying the regularities (3.7), which implies that

(3.117) ( I , I t ) L 2 ( R + × S 2 ; C ( [ 0 , T ] × R 3 ) ) , ϕ C 1 ( [ 0 , T ] × R 3 ) , ( u , u ) C ( [ 0 , T ] × R 3 ) .

Second, from the transformation in (3.1), one has

(3.118) ρ ( t , x ) = γ 1 A γ ϕ 1 γ 1 ( t , x ) and ρ ϕ ( t , x ) = 1 γ 1 γ 1 A γ 1 γ 1 ϕ 2 γ γ 1 ( t , x ) .

Then multiplying (3.2) 2 by ρ ϕ ( t , x ) yields the continuity equation (1.9) 2 ; and multiplying (3.2) 3 by ρ gives the momentum equations (1.9) 3 .

Thus, we have shown that ( I , ρ , u ) is a weak solution in the sense of distributions to the Cauchy problem (1.9) and (1.10) and satisfying the regularities in Definition 2.1 and (2.11). Moreover, it follows from the continuity equation that ρ ( t , x ) > 0 for [ 0 , T ] × R 3 . In summary, the Cauchy problem (1.9) and (1.10) has a unique regular solution ( I , ρ , u ) .

Step 2: The local well-posedness of classical solutions. Now we show that the regular solution obtained above is indeed a classical one in positive time.

First, according to (3.117) and (3.118), one concludes that

( I , I t ) L 2 ( R + × S 2 ; C ( [ 0 , T ] × R 3 ) ) , ( ρ , ρ , ρ t , u , u ) C ( [ 0 , T ] × R 3 ) .

Second, it follows from the classical Sobolev embedding:

(3.119) L 2 ( [ 0 , T ] ; H 1 ) W 1 , 2 ( [ 0 , T ] ; H 1 ) C ( [ 0 , T ] ; L 2 ) ,

and the regularities in (2.11) that

(3.120) t u t C ( [ 0 , T ] ; H 2 ) and u t C ( [ τ , T ] × R 3 ) .

It remains to show that 2 u C ( [ τ , T ] × R 3 ) . According to (3.2) 3 , one has

(3.121) a L u = ϕ 2 e u t + u u + ϕ ψ Q ( u ) 1 c 0 S 2 A ¯ r Ω d Ω d ν ϕ 2 e H ,

which, along with (2.11), yields that

(3.122) t ϕ 2 e H L ( [ 0 , T ] ; H 2 ) .

On the other hand, one has

(3.123) ( t ϕ 2 e H ) t = ϕ 2 e H + t ( ϕ 2 e ) t H + t ϕ 2 e H t L 2 ( [ 0 , T ] ; L 2 ) .

It thus follows from the classical embedding theorem:

(3.124) L ( [ 0 , T ] ; H 1 ) W 1 , 2 ( [ 0 , T ] ; H 1 ) C ( [ 0 , T ] ; L q ) for q [ 2 , 6 ) ,

and (3.121)–(3.123) that

t ϕ 2 e H C ( [ 0 , T ] ; W 1 , 4 ) and t 2 u C ( [ 0 , T ] ; W 1 , 4 ) .

Again the Sobolev embedding theorem implies that 2 u C ( ( 0 , T ] × R 3 ) .

Finally, the proof of Theorem 2.1 is complete.□

Acknowledgments

The authors would like to thank the anonymous referees for their careful reading and several valuable comments.

  1. Funding information: The research was supported in part by National Natural Science Foundation of China under the grant 12101395. The research of Shengguo Zhu was also supported in part by The Royal Society–Newton International Fellowships Alumni AL/201021 and AL/211005.

  2. Conflict of interest: The authors declare that they have no conflict of interest. The authors also declare that this manuscript has not been previously published and will not be submitted elsewhere before your decision.

  3. Data availability statement: Data sharing is not applicable to this article as no data sets were generated or analysed during the current study.

Appendix A Basic lemmas

This appendix is devoted to listing some useful lemmas that were used frequently in the previous sections. The first one is the well-known Gagliardo-Nirenberg inequality.

Lemma A.1

[9] Let 1 q , r , and p satisfies

1 p = j d + ζ 1 r m d + ( 1 ζ ) 1 q , and j m ζ 1 .

Then there exists a generic constant C > 0 depending only on j , m , d , q , r , and ζ such that for f L q ( R d ) D m , r ( R d ) ,

D j f L p C D m f L r ζ f L q 1 ζ ,

with the exceptions that if j = 0 , r m < d , q = we assume that f vanishes at infinity or f L q ˜ for some finite q ˜ > 0 , while if 1 < r < and m j d / r is a nonnegative integer we take j / m ζ < 1 .

The second one is the well-known Fatou’s lemma, which can be found in [27].

Lemma A.2

[27] Given a measure space ( V , , ν ) and a set X , let { f n } be a sequence of ( , R 0 ) -measurable nonnegative functions f n : X [ 0 , ] . Define the function f : X [ 0 , ] by setting

f ( x ) = liminf n f n ( x ) ,

for every x X . Then f is ( , R 0 ) -measurable, and

X f ( x ) d ν liminf n X f n ( x ) d ν .

The third one shows some compactness results obtained via the Aubin-Lions Lemma.

Lemma A.3

[29] Let X 0 X X 1 be three Banach spaces. Suppose that X 0 is compactly embedded in X and X is continuously embedded in X 1 . Then the following statements hold.

  1. If J is bounded in L p ( [ 0 , T ] ; X 0 ) for 1 p < + , and J t is bounded in L 1 ( [ 0 , T ] ; X 1 ) , then J is relatively compact in L p ( [ 0 , T ] ; X ) ;

  2. If J is bounded in L ( [ 0 , T ] ; X 0 ) and J t is bounded in L p ( [ 0 , T ] ; X 1 ) for p > 1 , then J is relatively compact in C ( [ 0 , T ] ; X ) .

The following calculus inequalities can be found in Majda [21].

Lemma A.4

[21] Let r , a and b be constants such that

1 r = 1 a + 1 b , and 1 a , b , r .

s 1 , if f , g W s , a W s , b ( R 3 ) , then it holds that

s ( f g ) f s g r C s ( f a s 1 g b + s f b g a ) , s ( f g ) f s g r C s ( f a s 1 g b + s f a g b ) ,

where C s > 0 is a constant depending only on s , and s f ( s > 1 ) is the set of all x ζ f with ζ = s . Here, ζ = ( ζ 1 , ζ 2 , ζ 3 ) R 3 is a multi-index.

The following lemma is used to obtain the time-weighted estimates of the velocity u .

Lemma A.5

[1] If f ( t , x ) L 2 ( [ 0 , T ] ; L 2 ) , then there exists a sequence s k such that

s k 0 and s k f ( s k , x ) 2 2 0 as k + .

The last one gives the regularity estimates for the Lamé operator in R 3 .

Lemma A.6

[30] If u D 1 , q ( R 3 ) with 1 < q < + is a weak solution to the problem

α u ( α + β ) div u = Z , u 0 as x + ,

then it holds that

u D k + 2 , q C Z D k , q ,

where the constant C > 0 depends only on α , β , and q .

References

[1] J. L. Boldrini, M. A. Rojas-Medar, and E. Fernández-Cara, Semi-Galerkin approximation and regular solutions to the equations of the nonhomogeneous asymmetric fluids, J. Math. Pures Appl. 82 (2003), 1499–1525. 10.1016/j.matpur.2003.09.005Suche in Google Scholar

[2] J. Cao and Y. Qin, Global existence of strong solutions in H^2 to the diffusion approximation model in radiation hydrodynamics, Math. Methods Appl. Sci. 41 (2018), 1509–1517. 10.1002/mma.4681Suche in Google Scholar

[3] G. Chen, G.-Q. Chen, and S. Zhu, Vanishing viscosity limit of the three-dimensional isentropic compressible Navier-Stokes equations with degenerate viscosities and far field vacuum, Ann. Inst. H. Poincaré Anal. Non Linéaire 39 (2022), 121–170. 10.4171/aihpc/4Suche in Google Scholar

[4] Y. Cho, H. Choe, and H. Kim, Unique solvability of the initial boundary value problems for compressible viscous fluids, J. Math. Pures Appl. 83 (2004), 243–275. 10.1016/j.matpur.2003.11.004Suche in Google Scholar

[5] M. Ding and S. Zhu, Vanishing viscosity limit of the Navier-Stokes equations to the Euler equations for compressible fluid flow with far field vacuum, J. Math. Pures Appl. 107 (2017), 288–314. 10.1016/j.matpur.2016.07.001Suche in Google Scholar

[6] B. Ducomet, E. Feireisl, and Š. Nečasová, On a model in radiation hydrodynamics, Ann. Inst. H. Poincaré Anal. Non Linéaire 28 (2011), 797–812. 10.1016/j.anihpc.2011.06.002Suche in Google Scholar

[7] B. Ducomet and Š. Nečasová, Global smooth solution of the Cauchy problem for a model of radiative flow, Ann. Sci. Norm. Super. Pisa Cl. Sci. (5) 14 (2015), 1–36. 10.2422/2036-2145.201208_007Suche in Google Scholar

[8] E. Feireisl, A. Novotný, and H. Petzeltová, On the existence of globally defined weak solutions to the Navier-Stokes equations, J. Math. Fluid Mech. 3 (2001), 358–392. 10.1007/PL00000976Suche in Google Scholar

[9] G. Galdi, An Introduction to the Mathematical Theory of the Navier-Stokes Equations, Springer, New York, 1994. 10.1007/978-1-4612-5364-8Suche in Google Scholar

[10] Y. Geng, Y. Li, and S. Zhu, Vanishing viscosity limit of the Navier-Stokes equations to the Euler equations for compressible fluid flow with vacuum, Arch. Ration. Mech. Anal. 234 (2019), 727–775. 10.1007/s00205-019-01401-9Suche in Google Scholar

[11] X. Huang, J. Li, and Z. Xin, Global well-posedness of classical solutions with large oscillations and vacuum to the three-dimensional isentropic compressible Navier-Stokes equations, Comm. Pure Appl. Math. 65 (2012), 549–585. 10.1002/cpa.21382Suche in Google Scholar

[12] P. Jiang and D. Wang, Formation of singularities of solutions to the three-dimensional Euler- Boltzmann equations in radiation hydrodynamics, Nonlinearity 23 (2010), 809–821. 10.1088/0951-7715/23/4/003Suche in Google Scholar

[13] S. Kawashima, Systems of a Hyperbolic-Parabolic Composite Type, with Applications to the Equations of Magnetohydrodynamics, Ph.D. thesis, Kyoto University, Kyoto, Japan, 1983, https://doi.org/10.14989/doctor.k3193. Suche in Google Scholar

[14] O. A. Ladyženskaja, V. A. Solonnikov, and N. N. Ural’ceva, Linear and Quasilinear Equations of Parabolic Type, American Mathematical Society, Providence, R.I., 1968. 10.1090/mmono/023Suche in Google Scholar

[15] H. Li and Y. Li, Local well-posedness of compressible radiation hydrodynamic equations with density-dependent viscosities and vacuum, Math. Methods Appl. Sci. 44 (2020), 4715–4744. 10.1002/mma.7064Suche in Google Scholar

[16] Y. Li, R. Pan, and S. Zhu, On classical solutions for viscous polytropic fluids with degenerate viscosities and vacuum, Arch. Ration. Mech. Anal. 234 (2019), 1281–1334. 10.1007/s00205-019-01412-6Suche in Google Scholar

[17] Y. Li and S. Zhu, Existence results for compressible radiation hydrodynamic equations with vacuum, Commun. Pure Appl. Anal. 14 (2015), 1023–1052. 10.3934/cpaa.2015.14.1023Suche in Google Scholar

[18] Y. Li and S. Zhu, On regular solutions of the 3D compressible isentropic Euler-Boltzmann equations with vacuum, Discrete Contin. Dyn. Syst. 35 (2015), 3059–3086. 10.3934/dcds.2015.35.3059Suche in Google Scholar

[19] P.-L. Lions, Mathematical Topics in Fluid Mechanics: Compressible Models, vol. 2, Oxford University Press, New York, 1998. Suche in Google Scholar

[20] A. Majda, Compressible fluid flow and systems of conservation laws in several space variables, in: Applied Mathematical Sciences, vol. 53, Springer Berlin Heidelberg, New York, 1984. 10.1007/978-1-4612-1116-7Suche in Google Scholar

[21] A. Matsumura and T. Nishida, The initial value problem for the equations of motion of viscous and heat-conductive gases, J. Math. Kyoto Univ. 20 (1980), 67–104. 10.1215/kjm/1250522322Suche in Google Scholar

[22] J. Nash, Le problème de Cauchy pour les équations différentielles daun fluide général, Bull. Soc. Math. France 90 (1962), 487–497. 10.24033/bsmf.1586Suche in Google Scholar

[23] G. C Pomraning, The Equations of Radiation Hydrodynamics, Pergamon Press, Oxford, UK, 1973. Suche in Google Scholar

[24] Y. Qin, B. Feng, and M. Zhang, Large-time behavior of solutions for the one-dimensional infra relativistic model of a compressible viscous gas with radiation, J. Differ. Equ. 252 (2012), 6175–6213. 10.1016/j.jde.2012.02.022Suche in Google Scholar

[25] Y. Qin and J. Zhang, Global existence and asymptotic behavior of cylindrically symmetric solutions for the 3D infrarelativistic model with radiation, J. Math. Fluid Mech. 20 (2018), 35–81. 10.1007/s00021-016-0312-3Suche in Google Scholar

[26] W. Rudin, Real and Complex Analysis, McGraw-Hill, Inc, Singapore, 1987. Suche in Google Scholar

[27] J. Serrin, On the uniqueness of compressible fluid motions, Arch. Rational Mech. Anal. 3 (1959), 271–288. 10.1007/BF00284180Suche in Google Scholar

[28] J. Simon, Compact sets in the space Lp(0,T;B), Ann. Mat. Pura Appl. 146 (1987), 65–96. 10.1007/BF01762360Suche in Google Scholar

[29] E. M. Stein. Singular Integrals and Differentiability Properties of Functions, Princeton University Press, Princeton, 1970. 10.1515/9781400883882Suche in Google Scholar

[30] Z. Wang, Existence results for the radiation hydrodynamic equations with degenerate viscosity coefficients and vacuum, J. Differ. Equ. 265 (2018), 354–388. 10.1016/j.jde.2018.02.035Suche in Google Scholar

[31] Z. Xin and S. Zhu, Global well-posedness of regular solutions to the three-dimensional isentropic compressible Navier-Stokes equations with degenerate viscosities and vacuum, Adv. Math. 393 (2021), 108072. 10.1016/j.aim.2021.108072Suche in Google Scholar

[32] Z. Xin and S. Zhu, Well-posedness of three-dimensional isentropic compressible Navier-Stokes equations with degenerate viscosities and far field vacuum, J. Math. Pures Appl. 152 (2021), 94–144. 10.1016/j.matpur.2021.05.004Suche in Google Scholar

[33] X. Zhong and S. Jiang, Local existence and finite-time blow-up in multidimensional radiation hydrodynamics, J. Math. Fluid Mech. 9 (2007), 543–564. 10.1007/s00021-005-0213-3Suche in Google Scholar

Received: 2022-03-06
Accepted: 2022-07-14
Published Online: 2022-08-20

© 2023 Hao Li and Shengguo Zhu, published by De Gruyter

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

Artikel in diesem Heft

  1. Regular Articles
  2. On sufficient “local” conditions for existence results to generalized p(.)-Laplace equations involving critical growth
  3. On the critical Choquard-Kirchhoff problem on the Heisenberg group
  4. On the local behavior of local weak solutions to some singular anisotropic elliptic equations
  5. Viscosity method for random homogenization of fully nonlinear elliptic equations with highly oscillating obstacles
  6. Double-phase parabolic equations with variable growth and nonlinear sources
  7. Logistic damping effect in chemotaxis models with density-suppressed motility
  8. Bifurcation diagrams of one-dimensional Kirchhoff-type equations
  9. Standing wave solution for the generalized Jackiw-Pi model
  10. Weak-strong uniqueness and energy-variational solutions for a class of viscoelastoplastic fluid models
  11. Eigenvalue inequalities for the buckling problem of the drifting Laplacian of arbitrary order
  12. Homoclinic solutions for a differential inclusion system involving the p(t)-Laplacian
  13. Equivalence between a time-fractional and an integer-order gradient flow: The memory effect reflected in the energy
  14. Bautin bifurcation with additive noise
  15. Small solitons and multisolitons in the generalized Davey-Stewartson system
  16. Nonstationary Poiseuille flow of a non-Newtonian fluid with the shear rate-dependent viscosity
  17. A global compactness result with applications to a Hardy-Sobolev critical elliptic system involving coupled perturbation terms
  18. On a strongly damped semilinear wave equation with time-varying source and singular dissipation
  19. Singularity for a nonlinear degenerate hyperbolic-parabolic coupled system arising from nematic liquid crystals
  20. Stability of stationary solutions to the three-dimensional Navier-Stokes equations with surface tension
  21. Uniform decay estimates for the semi-linear wave equation with locally distributed mixed-type damping via arbitrary local viscoelastic versus frictional dissipative effects
  22. Symmetry and nonsymmetry of minimal action sign-changing solutions for the Choquard system
  23. Periodic and quasi-periodic solutions of a four-dimensional singular differential system describing the motion of vortices
  24. Multiple nontrivial solutions of superlinear fractional Laplace equations without (AR) condition
  25. Existence and blow-up of solutions in Hénon-type heat equation with exponential nonlinearity
  26. Existence and concentration behavior of positive solutions to Schrödinger-Poisson-Slater equations
  27. On the fractional Korn inequality in bounded domains: Counterexamples to the case ps < 1
  28. Gradient estimates for nonlinear elliptic equations involving the Witten Laplacian on smooth metric measure spaces and implications
  29. Dynamical analysis of a reaction–diffusion mosquito-borne model in a spatially heterogeneous environment
  30. Existence and multiplicity of solutions for a quasilinear system with locally superlinear condition
  31. Nontrivial solution for Klein-Gordon equation coupled with Born-Infeld theory with critical growth
  32. Modeling Wolbachia infection frequency in mosquito populations via a continuous periodic switching model
  33. Infinitely many localized semiclassical states for nonlinear Kirchhoff-type equation
  34. Fujita-type theorems for a quasilinear parabolic differential inequality with weighted nonlocal source term
  35. Approximations of center manifolds for delay stochastic differential equations with additive noise
  36. Periodic solutions to a class of distributed delay differential equations via variational methods
  37. Groundstate for the Schrödinger-Poisson-Slater equation involving the Coulomb-Sobolev critical exponent
  38. Existence of nontrivial solutions for the Klein-Gordon-Maxwell system with Berestycki-Lions conditions
  39. Global Sobolev regular solution for Boussinesq system
  40. Normalized solutions for the p-Laplacian equation with a trapping potential
  41. Nonlinear elliptic–parabolic problem involving p-Dirichlet-to-Neumann operator with critical exponent
  42. Blow-up for compressible Euler system with space-dependent damping in 1-D
  43. High energy solutions of general Kirchhoff type equations without the Ambrosetti-Rabinowitz type condition
  44. On the dynamics of grounded shallow ice sheets: Modeling and analysis
  45. A survey on some vanishing viscosity limit results
  46. Blow-up for logarithmic viscoelastic equations with delay and acoustic boundary conditions
  47. Generalized Liouville theorem for viscosity solutions to a singular Monge-Ampère equation
  48. Front propagation in a double degenerate equation with delay
  49. Positive solutions for a class of singular (pq)-equations
  50. Higher integrability for anisotropic parabolic systems of p-Laplace type
  51. The Poincaré map of degenerate monodromic singularities with Puiseux inverse integrating factor
  52. On a system of multi-component Ginzburg-Landau vortices
  53. Existence and concentration of solutions to Kirchhoff-type equations in ℝ2 with steep potential well vanishing at infinity and exponential critical nonlinearities
  54. Multiplicity results for fractional Schrödinger-Kirchhoff systems involving critical nonlinearities
  55. On double phase Kirchhoff problems with singular nonlinearity
  56. Estimates for eigenvalues of the Neumann and Steklov problems
  57. Nodal solutions with a prescribed number of nodes for the Kirchhoff-type problem with an asymptotically cubic term
  58. Dirichlet problems involving the Hardy-Leray operators with multiple polars
  59. Incompressible limit for compressible viscoelastic flows with large velocity
  60. Characterizations for the genuine Calderón-Zygmund operators and commutators on generalized Orlicz-Morrey spaces
  61. Non-local gradients in bounded domains motivated by continuum mechanics: Fundamental theorem of calculus and embeddings
  62. Noncoercive parabolic obstacle problems
  63. Touchdown solutions in general MEMS models
  64. Existence and nonexistence of nontrivial solutions for the Schrödinger-Poisson system with zero mass potential
  65. Short-time existence of a quasi-stationary fluid–structure interaction problem for plaque growth
  66. Fine bounds for best constants of fractional subcritical Sobolev embeddings and applications to nonlocal PDEs
  67. Symmetries of Ricci flows
  68. Asymptotic behavior of solutions of a second-order nonlinear discrete equation of Emden-Fowler type
  69. On the topological gradient method for an inverse problem resolution
  70. Supersolutions to nonautonomous Choquard equations in general domains
  71. Uniform complex time heat Kernel estimates without Gaussian bounds
  72. Global existence for time-dependent damped wave equations with nonlinear memory
  73. Normalized solutions for a critical fractional Choquard equation with a nonlocal perturbation
  74. Reversed Hardy-Littlewood-Sobolev inequalities with weights on the Heisenberg group
  75. Lamé system with weak damping and nonlinear time-varying delay
  76. Global well posedness for the semilinear edge-degenerate parabolic equations on singular manifolds
  77. Blowup in L1(Ω)-norm and global existence for time-fractional diffusion equations with polynomial semilinear terms
  78. Boundary regularity results for minimisers of convex functionals with (p, q)-growth
  79. Parametric singular double phase Dirichlet problems
  80. Special Issue on Nonlinear analysis: Perspectives and synergies
  81. Editorial to Special issue “Nonlinear analysis: Perspectives and synergies”
  82. Identification of discontinuous parameters in double phase obstacle problems
  83. Global boundedness to a 3D chemotaxis-Stokes system with porous medium cell diffusion and general sensitivity
  84. On regular solutions to compressible radiation hydrodynamic equations with far field vacuum
  85. On Cauchy problem for fractional parabolic-elliptic Keller-Segel model
  86. The evolution of immersed locally convex plane curves driven by anisotropic curvature flow
  87. Stability of combination of rarefaction waves with viscous contact wave for compressible Navier-Stokes equations with temperature-dependent transport coefficients and large data
  88. On concave perturbations of a periodic elliptic problem in R2 involving critical exponential growth
  89. Existence of solutions for a double-phase variable exponent equation without the Ambrosetti-Rabinowitz condition
Heruntergeladen am 11.9.2025 von https://www.degruyterbrill.com/document/doi/10.1515/anona-2022-0264/html
Button zum nach oben scrollen