Home Incompressible limit for the compressible viscoelastic fluids in critical space
Article Open Access

Incompressible limit for the compressible viscoelastic fluids in critical space

  • Bin Han and Dan Wu EMAIL logo
Published/Copyright: January 23, 2025

Abstract

In this article, we consider the incompressible limit of global-in-time strong solutions with arbitrary large initial velocity for the compressible viscoelastic fluids in the sense of critical Besov framework. We decouple our compressible system into two coupling sub-systems by introducing a skew symmetric matrix, which is related to the deformation tensor. This work generalizes the similar result obtained by Hu et al. (Incompressible limit for compressible viscoelastic flows with large velocity, Advances in Nonlinear Analysis 12 (2023), 20220324) to the critical functional space with respective to the natural scaling of the system. The proof relies on the dispersive property of the linear system on the high-frequency regime and the parabolic property on the low-frequency regime. The dispersion tends to disappear when λ tends to infinite, but having large λ provides strong dissipation on the potential part of the velocity and thus makes the flow almost incompressible. In addition, by exploiting the intrinsic structure of the viscoelastic system, we obtain the global uniform estimates of the solutions near equilibrium.

MSC 2010: 35A05; 76A10; 76D03

1 Introduction

It is well known that viscoelasticity is a material property that exhibits both viscous and elastic characteristics when undergoing deformation. Viscoelastic fluids show intermediate behavior with some remarkable phenomena due to their elastic nature. They exhibit a combination of both fluid and solid characteristics and have received a great deal of interest. In this article, we focus on the Cauchy problem of the following compressible viscoelastic fluids in R N :

(1.1) t ρ + ( ρ v ) = 0 , t ( ρ v ) + ( ρ v v ) + P ( ρ ) = μ Δ v + ( λ + μ ) v + W F ˜ ( F ˜ ) F ˜ T det F ˜ , t F ˜ + v F ˜ = v F ˜ , ( ρ , v , F ˜ ) t = 0 = ( ρ 0 , v 0 , F ˜ 0 ) .

The unknowns are the fluid density ρ , velocity field v , and the deformation tensor F ˜ . The pressure P = P ( ρ ) is a given smooth function of ρ . Moreover, we denote by W F ˜ ( F ˜ ) the Piola-Kirchhoff tensor and W F ˜ ( F ˜ ) F ˜ T det F ˜ the Cauchy-Green tensor, respectively. Here, F ˜ T is the transpose of F ˜ . The viscous coefficients μ and λ are assumed to satisfy the following physical restrictions:

μ > 0 , 2 μ + N λ 0 ,

with N 2 the spatial dimension. As for F ˜ , we further assume that the initial data satisfy

(1.2) F ˜ 0 T det F ˜ 0 = ( ρ 0 F ˜ 0 ) = 0 , ρ 0 det F ˜ 0 T = 1 , F ˜ 0 l k l F ˜ 0 i j = F ˜ 0 l j l F ˜ 0 i k .

It has been proved in [22] (also see [29]) that (1.2) is preserved by the flow, i.e.,

(1.3) F ˜ T det F ˜ = ( ρ F ˜ T ) = 0 , ρ det F ˜ T = 1 ,

and

(1.4) F ˜ l k l F ˜ i j = F ˜ l j l F ˜ i k .

In this work, we consider a special case of the Hookean elasticity: W ( F ˜ ) = F ˜ 2 . Then, with the aid of (1.3), we can rewrite the compressible viscoelastic flows in a nonconservative form. Indeed, I denotes the identity matrix and after the changes of functions

b ( t , x ) = ρ ( t , x ) 1 , E ˜ ( t , x ) = F ˜ ( t , x ) I ,

system (1.1) reads

(1.5) t b + v b + div v = b div v , t v i + v v i ( A v ) i + i b j E ˜ i j = E ˜ j k j E ˜ i k I ( b ) ( A v ) i k ( b ) i b , t E ˜ + v E ˜ v = v E ˜ , ( b , v , E ˜ ) t = 0 = ( b 0 , v 0 , E ˜ 0 ) ,

where we denote A μ Δ + ( λ + μ ) , and

I ( b ) b 1 + b , k ( b ) P ( 1 + b ) 1 + b 1 .

Before going any further, we first recall some known results on the viscoelastic fluids, including incompressible and compressible systems. For the incompressible system, Lin et al. [25] first proved the local and global existence (with small initial data) of classical solutions to the two-dimensional (2D) case (1.8). Zhang et al. [4,26] generalized this result to the three-dimensional (3D) case via the curl free quantity F ˜ 1 . Liu et al. [22] presented a new proof by showing that curl F ˜ is actually of higher order. In the proof of the global part, they captured the damping mechanism on F ˜ through very subtle energy estimates. As for the scaling invariant approach, Qian [28] studied the well posedness of system (1.8) in the L 2 -type critical space (the critical L p framework, please refer to Zhang and Fang [32]). Fang et al. [9] proved global existence in critical space for density-dependent incompressible viscoelastic fluids by means of the Friedrichs method and compactness arguments. Besides, for the weak solutions, Hu and Lin [12] obtained the global existence of the 2D Leray-Hopf-type solutions to (1.8) in the physical energy space. Combined with a similar structure for the density, it is worth noting that the global existence of weak solutions to incompressible viscoelastic fluids with small perturbations near the rest state was established by Hu and Lin [13]. In addition, mathematicians also investigated other topics on viscoelastic fluids such as the large time behavior of solutions [17], the incompressible limit [23], the regularity of [17], and so on.

For compressible viscoelastic fluids, motivated by the seminal works on the compressible Navier-Stokes equations [7], Qian and Zhang [29] established the global existence of system (1.1) near equilibrium in the critical space. Hu and Wu [16] proved the global existence of the strong solutions of (1.1) by the standard energy method under the condition that the initial data are close to the constant equilibrium state in H 2 . By deeply exploiting the intrinsic structure of the compressible system under some physical restrictions, the first author of this article and Zi [10] proved that the compressible viscoelastic fluid admit a unique global solution with a class of large initial data. We mention here that it is still an open problem whether a global solution of the equations of viscoelastic fluids exists for any general large initial data in 2D or 3D space, even for the incompressible viscoelastic fluids. Similar to the compressible Navier-Stokes system, different viscosities make contributions to the flow of fluids. Nevertheless, there are still some literature on the stability theories about the compressible viscoelastic fluids equations near the incompressible system. Previous studies [15,22,25,29] showed that the presence of shear viscosity μ prevents the formation of shocks near the equilibrium. While the shear viscosity μ vanishes, the shock formation is expected even for small initial perturbations of incompressible models. Moreover, the volume viscosity can also prevent the shock formation viscoelastic fluids. Motivated by the work on the incompressible limit for the compressible Navier-Stokes equations [8], recently, Hu et al. [14] studied the incompressible limit of the compressible viscoelastic system near the equilibrium when shear viscosity μ > 0 and the value of volume viscosity is large with the initial data in Sobolev spaces H 2 (the case when μ tends to 0 and λ tends to + , we refer to [5,6] by Cui and Hu for 2D and 3D cases). For more studies on compressible viscoelastic fluids, we refer to [3,11,13,1820,24,27,30,31] and references therein.

Motivated by [14], in this article, we are going to study the incompressible limit of system (1.1) in a larger functional framework. Before stating our main result, we use scaling transformation for (1.1) to guess which space may be critical. It is not difficult to verify that (1.1) is invariant by the transformation

(1.6) ( ρ ( t , x ) , v ( t , x ) , F ˜ ( t , x ) ) ( ρ ( l 2 t , l x ) , l v ( l 2 t , l x ) , F ˜ ( l 2 t , l x ) ) , ( ρ 0 ( t , x ) , v 0 ( t , x ) , F ˜ 0 ( t , x ) ) ( ρ 0 ( l x ) , l v 0 ( l x ) , F ˜ 0 ( l x ) ) ,

provided that P ( ρ ) and W ( F ˜ ) have been changed to l 2 P ( ρ ) and l 2 W ( F ˜ ) , respectively.

Definition 1.1

A function space S ( R N ) × ( S R N ) × S ( R N ) is called critical space if the associated norm is invariant under the transformation ( ρ ( x ) , v ( x ) , F ˜ ( x ) ) ( ρ ( l x ) , l v ( l x ) , F ˜ ( l x ) ) .

We can directly verify that H ˙ N 2 × H ˙ N 2 1 × H ˙ N 2 is critical space for initial data. However, if ρ 0 I H ˙ N 2 , F ˜ 0 I H ˙ N 2 , then it is difficult to obtain L control on the density and deformation tensor since H ˙ N 2 is not included in L . Fortunately, B ˙ 2 , 1 N 2 (the exact definition will be given in Section 2) is an algebra embedded in L . Therefore, we shall choose suitable homogeneous Besov space as our working space (see [7] for compressible Navier-Stokes). We note that, Qian and Zhang [29] studied the global well posedness result, provided the initial data ( ρ 0 , v 0 , F ˜ 0 ) are close to a stable equilibrium, say, ( 1 , 0 , I ) in

(1.7) ( B ˙ 2 , 1 N 2 1 B ˙ 2 , 1 N 2 ) × B ˙ 2 , 1 N 2 1 × ( B ˙ 2 , 1 N 2 1 B ˙ 2 , 1 N 2 ) .

Our main goal here is to prove the global existence of strong solution to (1.5) with the initial data belonging to the critical space defined in (1.7). More precisely, for any initial velocity field v 0 with critical regularity and ( ρ 0 , F ˜ 0 ) sufficiently close to the equilibrium as λ tends to + . This result will strongly rely on the fact that the limit velocity and deformation for λ + satisfy the incompressible viscoelastic fluid system in R N :

(1.8) t V + V V μ Δ V + P = ( F F T ) , t F + V F = V F , V = 0 , ( V , F ) t = 0 = ( V 0 , F 0 ) .

As mentioned earlier, the global existence issue of strong solution to (1.8) supplemented with general data is open in R N ( N = 2 , 3 ). The first step we need to assume is that ( V 0 , F 0 ) generates a global strong solution to (1.8), and then to analyze the stability of that solution in the setting of the compressible model (1.5) with large λ in critical space defined in (1.7). We first admit the following result.

  • Suppose that ( V , F ) be smooth strong solution to the incompressible viscoelastic system (1.8) with initial data ( V 0 , M 0 F 0 I ) B ˙ 2 , 1 N 2 1 × ( B ˙ 2 , 1 N 2 1 B ˙ 2 , 1 N 2 ) . Denote M = F I , and assume that system (1.8) with initial data can generate a unique global solution and satisfy

    V C b ( R + ; B ˙ 2 , 1 N 2 1 ( R N ) ) L T 1 ( R + ; B ˙ 2 , 1 N 2 + 1 ( R N ) ) ,

    and

    M C b ( R + ; B ˙ 2 , 1 N 2 B ˙ 2 , 1 N 2 1 ( R N ) ) L T 2 ( R + ; B ˙ 2 , 1 N 2 B ˙ 2 , 1 N 2 + 1 ( R N ) ) .

    In addition, under the assumptions that if ( V 0 , M 0 ) B ˙ 2 , 1 N 2 and

    (1.9) M 0 B ˙ 2 , 1 N 2 1 B ˙ 2 , 1 N 2 ε 0

    holds true for some small positive ε 0 , then there exist two constants D 0 and δ 0 such that

    (1.10) V d ( T ) V L T ( R + ; B ˙ 2 , 1 N 2 1 ) + V t , 2 V L T 1 ( R + ; B ˙ 2 , 1 N 2 1 ) D 0

    and

    (1.11) M d ( T ) M L T ( R + ; B ˙ 2 , 1 N 2 1 B ˙ 2 , 1 N 2 ) + M L T 2 ( R + ; B ˙ 2 , 1 N 2 + 1 B ˙ 2 , 1 N 2 ) δ 0 ,

    where δ 0 can be chosen as a sufficiently small constant.

Based on the aforementioned statement, we want to justify the incompressible limit for the compressible viscoelastic system (1.5) around the incompressible state ( 1 , V , F ) in critical space when the volume viscosity λ tends to infinity. The global existence relies on the dispersion properties of the acoustic wave equations. In this situation, dispersion tends to disappear as λ + , but large λ provides a strong dissipation on the potential part of the velocity and thus makes the flow almost incompressible. In fact, our main result can be regarded as the stability theory of incompressible flow in compressible flow and can be stated as the following theorem.

Theorem 1.2

Let ( v 0 , ρ 0 , F ˜ 0 ) be initial data for (1.1) satisfying the constraint (1.3) with v 0 B ˙ 2 , 1 N 2 1 , ( b 0 , E ˜ 0 ) ( ρ 0 1 , F ˜ 0 I ) B ˙ 2 , 1 N 2 1 B ˙ 2 , 1 N 2 . Suppose that ( V 0 , F 0 ) is initial data of the incompressible viscoelastic fluids system satisfying, for some constant C > 0 such that

C e C ( 1 + D 0 2 ) ( b B ˙ 2 , 1 N 2 1 + ν b 0 B ˙ 2 , 1 N 2 + d 0 B ˙ 2 , 1 N 2 1 + D 0 2 + μ 2 ) μ ν ,

holds with ν λ + 2 μ satisfying ν μ , then (1.5) has a unique global-in-time solution ( ρ , v , E ˜ ) such that

v C b ( R + ; B ˙ 2 , 1 N 2 1 ) , v t , 2 v L 1 ( R + ; B ˙ 2 , 1 N 2 1 ) , b ( ρ 1 ) C ( R + ; B ˙ 2 , 1 N 2 1 B ˙ 2 , 1 N 2 ) L 2 ( R + ; B ˙ 2 , 1 N 2 ) , E ˜ F ˜ I C ( R + ; B ˙ 2 , 1 N 2 1 B ˙ 2 , 1 N 2 ) L 2 ( R + ; B ˙ 2 , 1 N 2 ) ,

where V 0 Λ 1 curl v 0 , d 0 Λ 1 div v 0 , and F ˜ 0 F 0 . Moreover, there holds

d L T ( R + ; B ˙ 2 , 1 N 2 1 ) + ν 2 d L T 1 ( R + ; B ˙ 2 , 1 N 2 1 ) + b L T ( R + ; B ˙ 2 , 1 N 2 1 ) + ν b L T ( R + ; B ˙ 2 , 1 N 2 ) C e C ( 1 + D 0 2 ) ( b 0 B ˙ 2 , 1 N 2 1 + ν b 0 B ˙ 2 , 1 N 2 + d 0 B ˙ 2 , 1 N 2 1 + D 0 2 + μ 2 ) ,

where d Λ 1 div v . Denote Ω Λ 1 curl ( v V ) , and if in addition b 0 = 0 , then ( ρ , v , F ˜ ) converges to ( 1 , V , F ) as follows:

ν μ ( ρ 1 L T ( R + ; B ˙ 2 , 1 N 2 ) + 2 d L T 1 ( R + ; B ˙ 2 , 1 N 2 1 ) ) + Ω L T ( R + ; B ˙ 2 , 1 N 2 1 ) + F ˜ F L T ( R + ; B ˙ 2 , 1 N 2 ) + μ 2 Ω L T 1 ( R + ; B ˙ 2 , 1 N 2 1 ) μ ν ,

provided that λ tends to + .

We should make the following remarks here.

Remark 1.3

The initial values of equations (1.5) and (1.8) are assumed so that the estimate of the perturbation quantity ( Ω , E ¯ ) can be closed by using energy method with zero initial value ( 0 , 0 ) . On the other hand, the equations coupled with d and b can reach a global estimate under the assumption that ν is sufficiently large.

Remark 1.4

As justified in [29], when V 0 B ˙ 2 , 1 N 2 1 + M 0 B ˙ 2 , 1 N 2 B ˙ 2 , 1 N 2 1 is sufficiently small, conditions (1.10) and (1.11) can be justified with ( D 0 , δ 0 ) replaced by a small parameter that depends on the initial data. However, for a large volume D 0 , the global-in-time existence of (1.8) is still out of reach.

Remark 1.5

Our future work will focus on the stability problems of the compressible viscoelastic fluids in the largest functional space. Indeed, for p 2 and s R , the inclusion relation B ˙ 2 , 1 s + N ( 1 2 1 p ) B ˙ p , 1 s motivates us to consider the incompressible limit of (1.5) in the framework of L p -type critical space. However, in view of the fact that one order terms predominate the system, L p approach is not applicable to the low-frequency regime. On the other hand, system (1.5) is not strictly invariant under the transformation (1.6). In the low-frequency regime, the system behaves like wave equations when the viscous terms vanish. This implies that we should balance between the parabolic and hyperbolic effects and consider the problem in the following space:

( ( ρ 0 1 ) l , v 0 l , ( F ˜ 0 I ) l ) B ˙ 2 , 1 N 2 1 + α × B ˙ 2 , 1 N 2 1 + α × B ˙ 2 , 1 N 2 1 + α ,

for some α > 0 and

( ( ρ 0 1 ) h , v 0 h , ( F ˜ 0 I ) h ) B ˙ p , 1 N p × B ˙ p , 1 N p 1 × B ˙ p , 1 N p .

This space described earlier was used by the first author and Zi in [10] for research of the global well posedness of the compressible viscoelastic fluids with small data.

1.1 Difficulties and key ideas

Formally, to study the limit of solutions of (1.5) by passing λ , we naturally define that u = v V and E = F ˜ F , where ( V , F ) satisfies (1.8). We now briefly explain how to construct the global estimates of the solutions in the selected functional space. First of all, we give a reformulation of system (1.5). To this end, denote

(1.12) Λ Δ , Λ s u 1 ( ξ s u ˆ ) , for s R ,

where u ˆ is the Fourier transform of u . Denote

(1.13) d Λ 1 u .

From (1.4), we infer that

(1.14) Λ 1 ( j k E ˜ i k ) = Λ E ˜ i j Λ 1 k ( E ˜ l j l E ˜ i k E ˜ l k l E ˜ i j ) .

Applying the operator Λ 1 j to (1.5)2, using (1.13) and (1.14), we find that it is difficult to achieve the coupling estimates of ( b , E ) in spite of ignoring the second-order terms. To overcome these difficulties, we turn our attention to the compatibility conditions (1.3) and (1.4) and deeply explore the coupling structure between b and E ¯ . This will be done by two steps.

1.1.1 Step 1. Reduction to two coupling hyperbolic systems

Motivated by interesting work on the isentropic compressible Navier-Stokes equations [7], u can also be split into the compressible part d Λ 1 u and the incompressible part

Ω Λ 1 curl u ,

with ( curl z ) i j = j z i i z j for any vector fields z . We first derive the system of ( b , d ) . In fact, by condition (1.3), we find the relation that

(1.15) E ˜ = Δ b ( b E ˜ ) .

Then, applying Λ 1 to ( 8 ) 2 , and using (1.15), the pair ( b , d ) solves

(1.16) b t + v b + Λ d = b div u , d t 2 Λ b ν Δ d = Λ 1 div ( ( b ( E + M ) ) + ( E + M ) ( E + M ) I ( b ) A v k ( b ) b v v ) ,

where on the right-hand side of (1.16), we have used E ˜ = E + M by the definition of E and M .

On the other hand, one cannot directly expect the coupling between Ω and E ˜ . Thanks to the skew symmetric property on Ω , it is naturally introducing a skew symmetric quantity, which is related to the deformation tensor F ˜ . Indeed, condition (1.4) can be rewritten in terms of E ˜ as

k E ˜ i j j E ˜ i k = E ˜ l j l E ˜ i k E ˜ l k l E ˜ i j .

It follows immediately that

(1.17) j k E ˜ i k i k E ˜ j k = k ( j E ˜ i k i E ˜ j k ) = k ( k E ˜ i j k E ˜ j i E ˜ l j l E ˜ i k ) + k ( E ˜ l k l E ˜ i j + E ˜ l i E ˜ l k E ˜ l k l E ˜ j i ) = Δ ( E ˜ i j E ˜ j i ) + k ( E ˜ l i l E ˜ j k E ˜ l k l E ˜ j i ) k ( E ˜ l j l E ˜ i k k E ˜ l k l E ˜ i j ) .

In the same way, we can obtain easily that

(1.18) j k M i k i k M j k = k ( j M i k i M j k ) = k ( k M i j k M j i M l j l M i k ) + k ( M l k l M i j + M l i M l k M l k l M j i ) = Δ ( M i j M j i ) + k ( M l i l M j k M l k l M j i ) k ( M l j l M i k k M l k l M i j ) .

Applying Λ 1 curl to (1.5) and (1.8), then making a difference, and using formulas (1.17) and (1.18), we find that the skew symmetric matrix E ¯ ( E T E ) solves

(1.19) E ¯ t + v E ¯ + Λ Ω = v E ¯ + u ( M T M ) u ( M T M ) , Ω t μ Δ Ω Λ E ¯ = Λ 1 curl ( v u + u V ( E + M ) ( E + M ) + M M T + I ( b ) A v ) + Λ 1 H ,

where

H i j k ( E ˜ l i l E ˜ j k E ˜ l k l E ˜ j i ) k ( E ˜ l j l E ˜ i k k E ˜ l k l E ˜ i j ) ( k ( M l i l M j k M l k l M j i ) k ( M l j l M i k k M l k l M i j ) ) .

To simplify the presentation, we assume from now on that shear viscosity μ is 1. The hyperbolic sub-systems (1.16) and (1.19) were first introduced by Hu and Wu in [16] to study the decay rates of the strong solutions of (1.5) in some Hilbert space. From the short review above, key observation emerges: in this article, (1.16) and (1.19) have almost the same structure. Roughly speaking, the linearized system of (1.16) and (1.19) is the following form:

t a + Λ c = f , t c Δ c Λ a = g ,

which is the same as one linearized from the isentropic compressible Navier-Stokes equations. In order to bound b , v , and E ¯ , the general approach is adapted from [8]: we localize equations (1.16) and (1.19) in frequency space by means of the dyadic operators Δ j and perform suitable energy estimates. In order to obtain optimal bounds for the two coupling sub-systems, the key point is that we have to use the same regularity for u , b , and E ¯ (or, equivalently, to use different regularities for the low and high frequencies of b and E ¯ ). From a quick glance at equations (1.16) and (1.19), it is more natural to estimate d t 2 Λ b and Ω t Λ E ¯ rather than only d t and Ω t .

1.1.2 Step 2. Recovering the estimates of E

We would like to emphasize that in (1.19), the incompressible part Ω is coupled with E ¯ . In order to obtain the estimates of E , we need to find the potential relation between E and E ¯ . Different from Hu and Wu in [16], we will first rewrite ρ E ˜ in terms of E ¯ and b (up to some nonlinear terms). In fact, from (1.3), one deduces that

(1.20) k ( ρ E i k ) = k E ¯ k i i b + k ( b E ¯ k i ) k ( b M k i ) .

Multiplying (1.4) by ρ , and using (1.3) yield

l ( ρ ( F ˜ l k F ˜ i j F l k F i j ) ) = l ( ρ ( F ˜ l j F ˜ i k F l j F i k ) ) ,

i.e.,

l ( ρ E l k ( E ˜ i j + ε i j ) + ρ ( M l k + ε l k ) E i j ) = l ( ρ E l j ( E ˜ i k + ε i k ) + ρ ( M l j + ε l j ) E i k ) .

Here, we use ε i j to denote the diagonal matrix, which means that

ε i j = 1 , if i = j , 0 , if i j .

Using (1.3) again, we infer that

l ( ρ E l k E i j + ρ M l k E i j + k ( ρ E i j ) ) = l ( ρ E l j E i k + ρ M l j E i k + k ( ρ E i k ) ) .

Therefore,

(1.21) k ( ρ E i j ) j ( ρ E i k ) = l ( ρ E l j E i k ) l ( ρ E l k E i j ) + l ( ρ E l j M i k ) l ( ρ E l k M i j ) + l ( ρ M l j E i k ) l ( ρ M l k E i j ) .

Applying the div-curl decomposition to ρ E , and in view of (1.20) and (1.21), we obtain

ρ E i j = Λ 2 k ( k ( ρ E i j ) j ( ρ E i k ) ) Λ 2 j k ( ρ E i k ) = Λ 2 k l ( E l j E i k E l k E i j + E l j M i k E l k M i j + M l j E i k M l k E i j ) Λ 2 k l ( b ( E l j E i k E l k E i j ) + b ( E l j M i k E l k M i j ) + b ( M l j E i k M l k E i j ) ) Λ 2 j k E ¯ k i + Λ 2 j k b Λ 2 j k ( b E ¯ k i ) + Λ 2 j k ( b M k i ) .

Finally, noting that E = ρ E b E , we find that

(1.22) E i j = Λ 2 k l ( E l j E i k E l k E i j + E l j M i k E l k M i j + M l j E i k M l k E i j ) Λ 2 k l ( b ( E l j E i k E l k E i j ) + b ( E l j M i k E l k M i j ) + b ( M l j E i k M l k E i j ) ) Λ 2 j k E ¯ k i + Λ 2 j k b Λ 2 j k ( b E ¯ k i ) + Λ 2 j k ( b M k i ) b E i j .

It can be observed from (1.22) that all the derivatives in nonlinear terms are outside products. These key observations make full use of the compatibility conditions and allow us to accomplish the estimates of E from E ¯ and b without losing derivative.

The rest of this article unfolds as follows. In the next section, we introduce Besov space and recall basic facts about them. Section 3 is devoted to proving Theorem 1.2.

1.2 Notations

  1. For u S ( R N ) , we denote u j Δ j u , and

    (1.23) u l ν 2 j < 1 Δ j u and u h ν 2 j > 1 Δ j u ,

    (1.24) u L 2 j < 1 Δ j u and u H 2 j > 1 Δ j u .

  2. Throughout this article, C denotes various “harmless” positive constants, which may change line by line. The notation A B means that A C B .

  3. In what follows, we denote by b l and b h the low- and high-frequency parts of b and use E ¯ L and E ¯ H to represent the low- and high-frequency parts of E ¯ , respectively, and set

    (1.25) X d ( T ) d , b , ν Λ b L T B ˙ 2 , 1 N 2 1 , Y d ( T ) d t 2 Λ b , ν 2 d , ν Λ 2 b l , Λ b h L T 1 B ˙ 2 , 1 N 2 1 , Z d ( T ) Ω , E ¯ , Λ E ¯ L T B ˙ 2 , 1 N 2 1 , W d ( T ) Ω t Λ E ¯ , 2 Ω , Λ 2 E ¯ L , Λ E ¯ H L T 1 B ˙ 2 , 1 N 2 1 ,

    and initial quantities are given by

    X d ( 0 ) d 0 , b 0 , ν Λ b 0 B ˙ 2 , 1 N 2 1 , Z d ( 0 ) 0 .

2 Littlewood-Paley decomposition results

The proof of most of the results presented in this article requires a dyadic decomposition of Fourier variables, which is called the Littlewood-Paley decomposition. Let us briefly explain how it may be built in the case x R N , which the readers may see in [1,2].

Let ( χ , φ ) be a couple of C functions satisfying

Supp χ ξ R N : ξ 4 3 , Supp φ ξ R N : 3 4 ξ 8 3 ,

and

χ ( ξ ) + j 0 φ ( 2 j ξ ) = 1 ,

j Z φ ( 2 j ξ ) = 1 , for ξ 0 .

Set φ j ( ξ ) = φ ( 2 j ξ ) , h j = 1 ( φ j ) , and h ˜ = 1 ( χ ) . The dyadic blocks and the low-frequency cutoff operators are defined for all j Z by

Δ j u = φ ( 2 j D ) u = 1 ( φ ( 2 j ) u ) = 2 j N h ( 2 j ) u , with h = 1 φ ,

S j u = χ ( 2 j D ) u = 1 ( χ ( 2 j ) u ) = 2 j N h ˜ ( 2 j ) u , with h ˜ = 1 φ .

Then,

(2.1) u = j Z Δ j u

holds for tempered distributions modulo polynomials. As working modulo polynomial is not appropriate for nonlinear problems, we shall restrict our attention to the set S h ( R N ) of tempered distributions u such that

lim j S j u L = 0 .

Note that (2.1) holds true whenever u is in S h and that one may write

S j u = q j 1 Δ q u .

The homogeneous decomposition has nice properties of quasi-orthogonality:

Δ p Δ j u 0 , if p j 2 , and Δ p ( S j 1 u Δ j u ) 0 , if p j 5 .

A series of functional space may be characterized in terms of Littlewood-Paley decomposition. Let us now give the definition of homogeneous Besov space:

Definition 2.1

[1] For s R and u S h ( R N ) , we set

u B ˙ 2 , 1 s = j Z 2 j s Δ j u L 2 .

Then, the homogeneous Besov space is defined as

B ˙ 2 , 1 s { u S h ( R N ) : u B ˙ 2 , 1 s < } .

The following lemmas will be needed to estimate the nonlinear terms of (1.16) and (1.19). One of them is just a consequence of Bony decomposition and of continuity results for the paraproduct and remainder operators.

Lemma 2.2

(Bernstein inequality [1]) For all 1 p q and k N , we have

Δ j k u L q ( R N ) C 2 j ( k + N ( 1 p 1 q ) ) Δ j u L p ( R N )

and

Δ j u L p ( R N ) C 2 j Δ j u L p ( R N ) .

The following properties of continuity for the product of two functions in Besov space will be constantly used in this article.

Lemma 2.3

Let g B ˙ 2 , 1 s 1 ( R N ) and h B ˙ 2 , 1 s 2 ( R N ) for some couple constants ( s 1 , s 2 ) satisfying

s 1 N 2 , s 2 N 2 , a n d s 1 + s 2 > 0 .

Then, g h B ˙ 2 , 1 s 1 + s 2 N 2 ( R N ) , and we have

g h B ˙ 2 , 1 s 1 + s 2 N 2 C g B ˙ 2 , 1 s 1 h B ˙ 2 , 1 s 2 .

For the details of the proof on the aforementioned lemma, please refer to [1] (Chapter 2).

Lemma 2.4

(Interpolation inequality [1]) Let z B ˙ 2 , 1 N 2 1 ( R N ) , with 2 z B ˙ 2 , 1 N 2 1 ( R N ) . Then, z B ˙ 2 , 1 N 2 ( R N ) , and the following inequality holds true:

z B ˙ 2 , 1 N 2 C z B ˙ 2 , 1 N 2 1 1 2 2 z B ˙ 2 , 1 N 2 1 1 2 .

The following lemma is used in (3.24).

Lemma 2.5

(Commutator estimate [21]) For s > 0 , Λ ( Δ ) 1 2 , and 1 p = 1 p 1 + 1 q 1 = 1 p 2 + 1 q 2 , we have

Λ s ( f g ) f Λ s g C ( f L p 1 Λ s 1 g L q 1 + Λ s L p 2 g L q 2 ) .

We end this section by introducing the Chemin-Lener space [1], which will be used for studying the evolution equations in the Besov framework.

Definition 2.6

For m [ 1 , + ] , s R , and T ( 0 , + ] , we set

u L T m ( B ˙ 2 , 1 s ) = j Z 2 j s Δ j u L T m ( L 2 )

when T = + , the index T is omitted.

3 Proof of the main results

In this section, we aim to construct the global a priori estimates on the perturbation equations in low- and high-frequency regimes. We first assume that T * be the maximal existence time of the solution ( ρ , v , E ˜ ) to (1.5) corresponding to data ( ρ 0 , v 0 , E ˜ 0 ) defined on the time interval T < T * and satisfying regularity assumption, and we fix some D 0 0 and δ 0 ( 0 , 1 ) so that the incompressible solution ( V , M ) to (1.8) satisfies (1.10) and (1.11). We claim that if ν is large enough, then one may find some (large) D and some δ so that for all T < T * , the following bounds are valid:

(3.1) X d ( T ) + Y d ( T ) D and Z d ( T ) + W d ( T ) δ .

Then, by a standard continuation principle, we may extend the solution beyond T * . For the purpose of (3.1), we first assume that for some large constant D and small constant δ ( 0 , 1 ) ,

(3.2) X d ( T ) + Y d ( T ) 2 D , Z d ( T ) + W d ( T ) δ 1 2 ,

which indicates that

(3.3) b L T B ˙ 2 , 1 N 2 1 .

Now, we derive the estimates on E ˜ in our original perturbation system (1.5). Using (1.22), it is easy to see that

E L T 2 B ˙ 2 , 1 N 2 C ( b , E ¯ ) L T 2 B ˙ 2 , 1 N 2 + E E L T 2 B ˙ 2 , 1 N 2 + b ( E E ) L T 2 B ˙ 2 , 1 N 2 + b E L T 2 B ˙ 2 , 1 N 2 + M E L T 2 B ˙ 2 , 1 N 2 + b ( E M ) L T 2 B ˙ 2 , 1 N 2 + b M L T 2 B ˙ 2 , 1 N 2 .

In what follows, we will show the desired norm of the right-hand side in the aforementioned inequality that

(3.4) b L T 2 B ˙ 2 , 1 N 2 b l L T B ˙ 2 , 1 N 2 1 1 2 b l L T 1 ( B ˙ 2 , 1 N 2 + 1 ) 1 2 + b h L T B ˙ 2 , 1 N 2 1 2 b h L T 1 B ˙ 2 , 1 N 2 1 2 ,

(3.5) E ¯ L T 2 B ˙ 2 , 1 N 2 E ¯ l L T B ˙ 2 , 1 N 2 1 1 2 E ¯ l L T 1 ( B ˙ 2 , 1 N 2 + 1 ) 1 2 + E ¯ h L T B ˙ 2 , 1 N 2 1 2 E ¯ h L T 1 B ˙ 2 , 1 N 2 1 2 ,

(3.6) E E , M E L T 2 B ˙ 2 , 1 N 2 ( E L T B ˙ 2 , 1 N 2 + M L T B ˙ 2 , 1 N 2 ) E L T 2 B ˙ 2 , 1 N 2 ,

(3.7) b ( E E ) , b ( M E ) L T 2 B ˙ 2 , 1 N 2 E E , M E L T 2 B ˙ 2 , 1 N 2 b L T B ˙ 2 , 1 N 2 ,

(3.8) b E L T 2 B ˙ 2 , 1 N 2 + b M L T 2 B ˙ 2 , 1 N 2 ( E L T 2 B ˙ 2 , 1 N 2 + M L T 2 B ˙ 2 , 1 N 2 ) b L T B ˙ 2 , 1 N 2 .

In order to obtain the key estimates of E L T B ˙ 2 , 1 N 2 and E L T 2 B ˙ 2 , 1 N 2 on the aforementioned inequalities, we first subtract the equation (1.8)2 from (1.1)3 and derive that

(3.9) t E + v E + u M u = v E + u M .

Now, localizing (3.9) by the Littlewood-Paley operators Δ j , we discover that

(3.10) t E j + v E j = ( v E + u M u M + u ) j [ Δ j , v ] E .

The commutator term of (3.10) may be bounded with the help of Lemma 2.100 in [1]. Testing (3.10) by E j , using Hölder inequality, and integrating in time variable, we obtain that

(3.11) E B ˙ 2 , 1 N 2 0 T div v B ˙ 2 , 1 N 2 E B ˙ 2 , 1 N 2 d τ + u L T 2 B ˙ 2 , 1 N 2 M L T 2 B ˙ 2 , 1 N 2 + u L T 1 B ˙ 2 , 1 N 2 1 + M L T B ˙ 2 , 1 N 2 + 0 T v B ˙ 2 , 1 N 2 E B ˙ 2 , 1 N 2 d τ .

Using Gronwall’s inequality, we conclude that

(3.12) E L T B ˙ 2 , 1 N 2 C e C 0 T d , Ω , V B ˙ 2 , 1 N 2 d τ u L T 2 B ˙ 2 , 1 N 2 M L T 2 B ˙ 2 , 1 N 2 + u L T 1 B ˙ 2 , 1 N 2 + u L T 1 B ˙ 2 , 1 N 2 M L T B ˙ 2 , 1 N 2 C e C 2 ν 1 D + δ 1 2 + D 0 ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) M d ( T ) + ν 1 Y d ( T ) + W d ( T ) + ( ν 1 Y d ( T ) + W d ( T ) ) M d ( T ) C e C 2 ν 1 D + δ 1 2 + D 0 2 ν 1 2 D δ 0 + 2 δ 1 2 δ 0 + 2 ν 1 D + δ 1 2 + 2 ν 1 D δ 0 ,

which implies that E L T B ˙ 2 , 1 N 2 is a small quantity with sufficiently small δ 0 , δ 1 2 and large ν . Then, we can obtain that the estimates of (3.6), (3.7), and b E L T 2 B ˙ 2 , 1 N 2 can be absorbed by the estimates of E L T 2 B ˙ 2 , 1 N 2 . By virtue of (3.4)–(3.8), we arrive at

(3.13) E L T 2 B ˙ 2 , 1 N 2 ( b , E ¯ ) L T 2 B ˙ 2 , 1 N 2 + M L T 2 B ˙ 2 , 1 N 2 b L T B ˙ 2 , 1 N 2 ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) .

Thus, from (3.13), we have

E ˜ L T 2 B ˙ 2 , 1 N 2 E L T 2 B ˙ 2 , 1 N 2 + M L T 2 B ˙ 2 , 1 N 2 ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) + M d ( T ) .

3.1 Estimates on ( b , d )

To estimate the potential part of the velocity d , we are required to consider coupling system about the momentum and continuity equations. Now, localizing (1.16) by the Littlewood-Paley operators Δ j , we discover that

(3.14) b j , t + ( u + V ) b j + Λ d j = g j , d j , t 2 Λ b j ν Δ d j + Λ 1 ( ( u + V ) d j ) = f j ,

where

(3.15) b j Δ j b , d j Δ j d , g j Δ j ( b div u ) [ Δ j , ( u + V ) ] b , f j Δ j Λ 1 div ( ( b ( E + M ) ) + ( E + M ) ( E + M ) I ( b ) A v k ( b ) b v ( V + Ω ) ) + Λ 1 [ Δ j , ( u + V ) ] d .

We follow an energy method to bound ( b j , d j ) . More precisely, testing (3.14)1 and (3.14)2 by b j and d j , respectively, yields

(3.16) 1 2 d d t b j 2 d x + Λ d j b j d x = 1 2 div u b j 2 d x + g j b j d x

and

(3.17) 1 2 d d t d j 2 d x + ( ν d j 2 2 Λ b j d j ) d x = 1 2 div u d j 2 + f j d j d x .

To obtain an estimate on the high-frequency part of b , applying Λ to (3.14)1, we then have

(3.18) Λ b j , t + ( u + V ) Λ b j Δ d j = Λ g j [ Λ , ( u + V ) b j ] ,

where [ Λ , ( u + V ) b j ] = Λ ( ( u + V ) b j ) ( u + V ) Λ b j .

Performing the inner product with Λ b j to the both sides of the aforementioned equation yields

(3.19) 1 2 d d t Λ b j 2 d x + ( u + V ) Λ b j Λ b j d x Δ d j Λ b j d x = ( Λ g j [ Λ , ( u + V ) b j ] ) Λ b j d x .

To eliminate the highest order term, namely, the one with Δ d j Λ b j , it is suitable to combine the aforementioned equality with a relation involving d j Λ b j d x . Now, testing (3.18) by d j and the momentum equation by Λ b j , we obtain

(3.20) d d t d j Λ b j d x + ( u + V ) ( Λ b j d j ) d x ν Δ d j Λ b j d x 2 Λ b j 2 d x Δ d j d j d x = ( Λ g j [ Λ , ( u + V ) b j ] ) d j d x + f j b j d x .

By integration by parts, we have

( u + V ) ( Λ b j d j ) d x = ( Λ b j d j ) div u d x .

Hence, adding ν times (3.19) to (3.20), the highest order terms cancel out, and we write

1 2 d d t ( ν Λ b j 2 2 d j b j ) d x + ( 2 Λ b j 2 d j 2 ) d x = ν 2 Λ b j 2 Λ b j d j div u d x + ν ( Λ g j [ Λ , ( u + V ) b j ] ) Λ b j d x ( Λ g j [ Λ , ( u + V ) b j ] ) d j d x f j Λ b j d x .

Multiplying the aforementioned equality by ν and adding up four times of (3.16) and twice of (3.17) imply that

(3.21) 1 2 d d t j 2 + ν ( d j 2 + 2 Λ b j 2 ) d x = ( 1 2 ν Λ b j 2 ν Λ b j d j + 2 b j 2 + d j 2 ) div u d x + 4 g j b j d x + ν 2 ( Λ g j [ Λ , ( u + V ) b j ] ) Λ b j d x ν f j Λ b j d x ν ( Λ g j [ Λ , ( u + V ) b j ] ) d j d x + f j d j d x ,

with

j 2 ( 4 b j 2 + 2 d j 2 2 ν d j Λ b j + ν Λ b j 2 ) d x ( d j ν Λ b j 2 + 4 b j 2 + d j 2 ) d x .

According to the definition of j , obviously, we have the following two facts:

(3.22) j d j , b j , ν Λ b j L 2 , for all j Z ,

and second,

(3.23) ν ( d j 2 + 2 Λ b j 2 ) d x c min ( ν 2 2 j , ν 1 ) j 2 .

In addition, using Lemma 2.5, we have

(3.24) [ Λ , ( u + V ) b j ] L 2 = Λ ( ( u + V ) b j ) ( u + V ) Λ b j L 2 C ( ( u + V ) L b j L 2 + Λ ( u + V ) L b j L 2 ) .

Therefore, (3.21)–(3.24) lead to

1 2 d d t j 2 + c min ( ν 2 2 j , ν 1 ) j 2 1 2 div u L + C Λ ( u + V ) L j 2 + C [ g j , f j , ν Λ g j ] L 2 j ,

whence, integrating over [ 0 , t ] in time,

(3.25) j ( t ) + c min ( ν 2 2 j , ν 1 ) 0 t j d τ j ( 0 ) + 0 t Λ ( u + V ) L j d τ + 0 t [ g j , ν Λ g j , f j ] L 2 d τ .

According to (3.25), we note that there will be loss of the expected parabolic smoothing effect of d since min ( ν 2 2 j , ν 1 ) = ν 1 for the high-frequency part. However, it may be recovered directly from (3.17). In fact, after using the Bernstein inequality in Lemma 2.2 and the Hölder inequality, we arrive at

1 2 d d t d j L 2 2 + c ν 2 2 j d j L 2 2 2 Λ b j L 2 d j L 2 + 1 2 div u L d j L 2 2 + f j L 2 d j L 2 ,

whence, integrating in time variable,

d j ( t ) L 2 + c ν 2 2 j 0 t d j L 2 d τ d j ( 0 ) L 2 + 1 2 0 t div u L d j L 2 d τ + 0 t f j L 2 d τ + 0 t 2 Λ b j L 2 d τ .

Together with (3.25), by using (3.22) and multiplying by 2 j N 2 1 , and eventually summing up over j Z , we end up with the following fundamental inequality:

(3.26) ( b , ν Λ b , d ) ( t ) B ˙ 2 , 1 N 2 1 + ν 0 t Λ b l , d B ˙ 2 , 1 N 2 d τ + 0 t b h B ˙ 2 , 1 N 2 d τ ( b , ν Λ b , d ) ( 0 ) B ˙ 2 , 1 N 2 1 + 0 t Λ ( u + V ) L ( b , ν Λ b , d ) ( t ) B ˙ 2 , 1 N 2 1 d τ + 0 t j Z 2 j N 2 1 [ g j , f j , ν Λ g j ] L 2 d τ ,

where notations b l and b h have been defined in (1.23). To complete the proof of estimates for b and d , we now have to obtain suitable bounds for the last term in (3.26). Let us start with the estimate of g j defined by (3.15). First, we see that, by virtue of Lemma 2.3, we have

(3.27) b div u B ˙ 2 , 1 N 2 1 div u B ˙ 2 , 1 N 2 b B ˙ 2 , 1 N 2 1 Λ d B ˙ 2 , 1 N 2 b B ˙ 2 , 1 N 2 1 ,

and using the Leibniz rule and, again, Lemma 2.3,

(3.28) ν Λ ( b div u ) B ˙ 2 , 1 N 2 1 2 d B ˙ 2 , 1 N 2 1 ν b B ˙ 2 , 1 N 2 + ν Λ b B ˙ 2 , 1 N 2 1 d B ˙ 2 , 1 N 2 .

The commutator term may be bounded as follows (the first inequality stems from Lemma 2.100 in [1] and the second one may be deduced from that lemma and Leibniz rule):

(3.29) j Z 2 j ( N 2 1 ) [ Δ j , ( u + V ) ] b L 2 ( u + V ) B ˙ 2 , 1 N 2 b B ˙ 2 , 1 N 2 1 ,

(3.30) j Z 2 j ( N 2 1 ) ν Λ [ Δ j , ( u + V ) ] b L 2 ( u + V ) B ˙ 2 , 1 N 2 ν Λ b B ˙ 2 , 1 N 2 1 .

Hence, putting (3.27)–(3.30) together, we obtain

(3.31) j Z 2 j N 2 1 ( g j , ν g j ) L 2 d , Ω , V B ˙ 2 , 1 N 2 ( b B ˙ 2 , 1 N 2 1 + ν Λ b B ˙ 2 , 1 N 2 1 ) .

Now, let us bound f j defined in (3.15). In fact, by Lemma 2.3, we have the following estimates:

(3.32) Λ 1 div ( v ( V + Ω ) ) B ˙ 2 , 1 N 2 1 d , Ω , V B ˙ 2 , 1 N 2 1 V , Ω B ˙ 2 , 1 N 2 ,

(3.33) Λ 1 div ( M E , E E ) B ˙ 2 , 1 N 2 1 M , E B ˙ 2 , 1 N 2 E B ˙ 2 , 1 N 2 1 ,

(3.34) Λ 1 div ( M M , E M ) B ˙ 2 , 1 N 2 1 M , E B ˙ 2 , 1 N 2 M B ˙ 2 , 1 N 2 1 ,

(3.35) Λ 1 div ( ( b E ) , ( b M ) ) B ˙ 2 , 1 N 2 1 b B ˙ 2 , 1 N 2 E , M B ˙ 2 , 1 N 2 ,

(3.36) Λ 1 div ( I ( b ) A v ) B ˙ 2 , 1 N 2 1 b B ˙ 2 , 1 N 2 ( 2 d , 2 Ω , 2 V B ˙ 2 , 1 N 2 1 + ν 2 d B ˙ 2 , 1 N 2 1 ) ,

and using the stability of the Besov space B ˙ 2 , 1 N 2 by left composition

(3.37) Λ 1 div k ( b ) b B ˙ 2 , 1 N 2 1 b B ˙ 2 , 1 N 2 2 .

As for g , the commutator term of f may be bounded according to Lemma 2.100 in [1]:

(3.38) j Z 2 j N 2 1 [ Δ j , ( u + V ) ] d L 2 ( u + V ) B ˙ 2 , 1 N 2 d B ˙ 2 , 1 N 2 1 .

From (3.32) to (3.38) and (3.13), applying the Hölder inequality in time variable and recalling the setting of X d ( T ) , Y d ( T ) , Z d ( T ) , and W d ( T ) defined in (1.25), we conclude that

(3.39) j Z 2 j N 2 1 f j L 2 L T 1 0 T d , Ω , V B ˙ 2 , 1 N 2 d B ˙ 2 , 1 N 2 1 d τ + M d 2 ( T ) + ( Z d ( T ) + V d ( T ) ) ( V d ( T ) + W d ( T ) ) + ν 1 X d ( T ) ( ν 1 Y d ( T ) + Y d ( T ) + W d ( T ) + V d ( T ) ) + M d ( T ) ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) 2 .

Returning to equation (3.14)2, it is worth mentioning that

d t 2 Λ b L T 1 B ˙ 2 , 1 N 2 1 ( u + V ) d L T 1 B ˙ 2 , 1 N 2 1 + f L T 1 B ˙ 2 , 1 N 2 1 + ν 2 d L T 1 B ˙ 2 , 1 N 2 1 .

The first term of the right-hand side satisfies

( u + V ) d B ˙ 2 , 1 N 2 1 d + Ω B ˙ 2 , 1 N 2 1 d B ˙ 2 , 1 N 2 + V B ˙ 2 , 1 N 2 d B ˙ 2 , 1 N 2 1 .

To bound ( u + V ) d , we apply Lemma 2.4 interpolation inequality to obtain that

( u + V ) d L T 1 B ˙ 2 , 1 N 2 1 C d + Ω L T B ˙ 2 , 1 N 2 1 d L T 1 B ˙ 2 , 1 N 2 + ε 2 d L T 1 B ˙ 2 , 1 N 2 1 + C ε 0 T V B ˙ 2 , 1 N 2 2 d B ˙ 2 , 1 N 2 1 d τ .

Assuming that δ is small enough and X d ( T ) ν , then putting (3.31) and (3.39) together in (3.26), and using equation (1.16), together with the fact that B ˙ 2 , 1 N 2 1 ( R N ) is continuously embedded in L ( R N ) gives us for all 0 T < T * , we obtain that

X d ( T ) + Y d ( T ) X d ( 0 ) + 0 T d , Ω , V B ˙ 2 , 1 N 2 + V B ˙ 2 , 1 N 2 2 X d ( T ) d τ + ( Z d ( T ) + V d ( T ) ) ( V d ( T ) + W d ( T ) ) + ν 1 X d ( T ) ( ν 1 Y d ( T ) + Y d ( T ) + W d ( T ) + V d ( T ) ) + M d ( T ) ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) 2 + ν 1 Z d ( T ) Y d ( T ) + M d 2 ( T ) .

Using Gronwall’s inequality, we conclude that

(3.40) X d ( T ) + Y d ( T ) C e C 0 T ( d , Ω , V B ˙ 2 , 1 N 2 + V B ˙ 2 , 1 N 2 2 ) d τ ( X d ( 0 ) + ν 1 Z d ( T ) Y d ( T ) + ( Z d ( T ) + V d ( T ) ) ( V d ( T ) + W d ( T ) ) + ν 1 X d ( T ) ( ν 1 Y d ( T ) + Y d ( T ) + W d ( T ) + V d ( T ) ) + M d ( T ) ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) 2 + M d 2 ( T ) .

3.2 Estimate on Ω and E ¯

We are required to consider the momentum and continuity equations altogether. Now, localizing (1.19) according to Littlewood-Paley operators, we discover that

(3.41) E ¯ j , t + ( u + V ) E ¯ j + Λ Ω j = m j ,

(3.42) Ω j , t Λ E ¯ j Δ Ω j + Λ 1 curl ( ( u + V ) Ω j ) = n j ,

where

(3.43) E ¯ j Δ j E ¯ , Ω j Δ j Ω , m j Δ j ( v E ¯ + u ( M T M ) u ( M T M ) ) [ Δ j , ( u + V ) ] E ¯ n j Δ j ( Λ 1 curl ( u V + v d ( E + M ) ( E + M ) + M M T + I ( b ) A v ) + Λ 1 H ) + Λ 1 curl ( [ Δ j , ( u + V ) ] Ω ) .

By introducing E ¯ , we observe that (1.16) and (1.19) have the same structure. So we deal with the estimates of ( Ω , E ¯ ) in the same way and can obtain

(3.44) ( E ¯ , Λ E ¯ , Ω ) ( t ) B 2 , 1 N 2 1 + 0 t Λ 2 E ¯ L , Λ E ¯ H , 2 Ω B 2 , 1 N 2 1 d τ 0 t ( Λ ( u + V ) L E ¯ , Λ E ¯ , Ω B 2 , 1 N 2 1 + j Z 2 j N 2 1 [ m j , n j , Λ m j ] L 2 ) d τ ,

where notations E ¯ L and E ¯ H have been defined in (1.24). To complete the proof of estimates for E ¯ and Ω , we now have to obtain suitable bounds for the last term in (3.44). Let us start with the bound of m j defined by (3.43). First, by virtue of Lemma 2.3, we have

(3.45) v E ¯ , Λ ( v E ¯ ) B ˙ 2 , 1 N 2 1 E ¯ B ˙ 2 , 1 N 2 1 + E ¯ B ˙ 2 , 1 N 2 d , Ω , V B ˙ 2 , 1 N 2 ,

(3.46) j Z 2 j N 2 1 [ Δ j , ( u + V ) ] E ¯ L 2 ( u + V ) B ˙ 2 , 1 N 2 E ¯ B ˙ 2 , 1 N 2 1 ,

(3.47) j Z 2 j N 2 1 Λ ( [ Δ j , ( u + V ) ] E ¯ ) L 2 ( u + V ) B ˙ 2 , 1 N 2 Λ E ¯ B ˙ 2 , 1 N 2 1 ,

(3.48) u ( M T M ) , Λ ( u ( M T M ) ) B ˙ 2 , 1 N 2 1 d , Ω B ˙ 2 , 1 N 2 M B ˙ 2 , 1 N 2 1 + M B ˙ 2 , 1 N 2 ,

(3.49) u ( M T M ) , Λ ( u ( M T M ) ) B ˙ 2 , 1 N 2 1 M B ˙ 2 , 1 N 2 + M B ˙ 2 , 1 N 2 + 1 d , Ω B ˙ 2 , 1 N 2 .

Therefore, according to (3.45)–(3.49), we can obtain

j Z 2 j N 2 1 ( m j , Λ m j ) L 2 L T 1 0 T d , Ω , V B ˙ 2 , 1 N 2 ( E ¯ B ˙ 2 , 1 N 2 1 + Λ E ¯ B ˙ 2 , 1 N 2 1 ) d τ + ν 1 Y d ( T ) M d ( T ) + W d ( T ) M d ( T ) + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) M d ( T ) .

Now, let us bound n j defined in (3.43).

(3.50) Λ 1 curl ( v d ) B ˙ 2 , 1 N 2 1 d , Ω , V B ˙ 2 , 1 N 2 1 d B ˙ 2 , 1 N 2 ,

(3.51) j Z 2 j N 2 1 [ Δ j , ( u + V ) ] Ω L 2 ( u + V ) B ˙ 2 , 1 N 2 Ω B ˙ 2 , 1 N 2 1 ,

(3.52) Λ 1 curl ( u V ) B ˙ 2 , 1 N 2 1 Ω B ˙ 2 , 1 N 2 1 V B ˙ 2 , 1 N 2 + d B ˙ 2 , 1 N 2 V B ˙ 2 , 1 N 2 1 ,

(3.53) Λ 1 curl ( E E , E M ) B ˙ 2 , 1 N 2 1 E , M B ˙ 2 , 1 N 2 1 E B ˙ 2 , 1 N 2 ,

(3.54) Λ 1 curl ( M E , M M M M T ) B ˙ 2 , 1 N 2 1 M B ˙ 2 , 1 N 2 E B ˙ 2 , 1 N 2 1 + M B ˙ 2 , 1 N 2 2 ,

(3.55) Λ 1 curl I ( b ) A v B ˙ 2 , 1 N 2 1 b B ˙ 2 , 1 N 2 ( 2 d , 2 Ω , 2 V B ˙ 2 , 1 N 2 1 + ν 2 d B ˙ 2 , 1 N 2 1 ) .

Combining the aforementioned inequalities (3.50)–(3.55) and (3.13) and using the Hölder inequality in time variable, we conclude that

j Z 2 j N 2 1 n j L 2 L T 1 0 T ( u + V ) B ˙ 2 , 1 N 2 Ω B ˙ 2 , 1 N 2 1 d τ + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) V d ( T ) + ν 1 X d ( T ) Y d ( T ) + ν 1 Z d ( T ) Y d ( T ) + ν 1 Y d ( T ) V d ( T ) + M d 2 ( T ) + ν 1 X d ( T ) ( W d ( T ) + V d ( T ) + ν 1 Y d ( T ) ) + M d ( T ) ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) 2 .

Then, bounding ( u + V ) Ω , we obtain the following simplification:

(3.56) Z d ( T ) + W d ( T ) 0 T ( d , Ω , V B 2 , 1 N 2 + d , V B 2 , 1 N 2 2 ) Z d ( T ) d τ + M d 2 ( T ) + ν 1 Y d ( T ) ( Z d ( T ) + V d ( T ) + M d ( T ) ) + ν 1 X d ( T ) ( W d ( T ) + V d ( T ) + Y d ( T ) + ν 1 Y d ( T ) ) + M d ( T ) ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) 2 + W d ( T ) M d ( T ) + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) V d ( T ) .

Hence, by the Gronwall lemma,

(3.57) Z d ( T ) + W d ( T ) C e C 0 T d , Ω , V B ˙ 2 , 1 N 2 + d , V B ˙ 2 , 1 N 2 2 d τ ( M d ( T ) ( M d ( T ) + W d ( T ) ) + ν 1 Y d ( T ) ( Z d ( T ) + V d ( T ) + M d ( T ) ) + ν 1 X d ( T ) ( W d ( T ) + V d ( T ) + Y d ( T ) + ν 1 Y d ( T ) ) + M d ( T ) ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) + ν 1 X d ( T ) M d ( T ) 2 + ν 1 2 X d 1 2 ( T ) Y d 1 2 ( T ) V d ( T ) .

3.3 Global-in-time closure of the estimates

In this section, we will complete the proof of (3.1). Assuming that

(3.58) ν 1 2 D ( 4 ( 1 + D 0 2 ) ) 1 1

and

(3.59) δ 0 1 2 , δ 1 2 1 2 , C 2 e 2 C 2 ( 1 + D 0 2 ) δ 1 2 1 2 .

These assumptions combined with (3.40) imply that

X d ( T ) + Y d ( T ) C 1 e C 1 ( D 0 + D 0 2 + 2 ν 1 D + δ 1 2 ) X d ( 0 ) + 2 ν 1 D δ 1 2 + ( δ 1 2 + D 0 ) 2 + 2 ν 1 D 2 ν 1 D + 2 D + δ 1 2 + D 0 + δ 0 2 ν 1 2 D + δ 1 2 + 2 ν 1 D δ 0 + 2 ν 1 2 D + δ 1 2 + 2 ν 1 D δ 0 2 + δ 0 2 C 1 e C 1 1 + 2 D 0 2 + 1 2 ν 1 2 + 1 2 X d ( 0 ) + D 0 2 + D 0 + 1 + 5 ν 2 D 2 + 4 ν 3 2 D 2 + 8 ν 1 D 2 + 7 2 ν 1 D + 2 ν 1 D D 0 + 3 ν 1 2 D ,

which gives that

X d ( T ) + Y d ( T ) C 1 e 2 C 1 ( 1 + D 0 2 ) ( X d ( 0 ) + 2 D 0 2 + 2 ) .

Then, we choose δ 0 satisfying

(3.60) C 2 e 2 C 2 ( 1 + D 0 2 ) ( δ 0 + δ 0 2 ) 1 16 ,

while (3.57), (3.59), and the fact that

d L T 2 ( 0 , T ; B ˙ 2 , 1 N 2 ) 2 d L T ( 0 , T ; B ˙ 2 , 1 N 2 1 ) d L T 1 ( 0 , T ; B ˙ 2 , 1 N 2 ) 4 ν 1 D 2

yield

Z d ( T ) + W d ( T ) C 2 e C 2 D 0 + D 0 2 + 2 ν 1 D + δ 1 2 + 4 ν 1 D 2 ( δ 0 2 + M d ( T ) W d ( T ) + 2 ν 1 D ( D 0 + δ 0 ) + 2 ν 1 D ( D 0 + 4 D + 2 ν 1 D ) + 2 ν 1 D ( Z d ( T ) + W d ( T ) ) + 2 ν 1 2 D δ 0 + δ 0 Z d 1 2 ( T ) W d 1 2 ( T ) + 2 ν 1 D δ 0 2 + 4 ν 2 D 2 δ 0 2 + Z d ( T ) W d ( T ) + 4 ν 1 2 D + 4 ν 1 D δ 0 Z d 1 2 ( T ) W d 1 2 ( T ) + 8 ν 3 2 D 2 δ 0 + 2 ν 1 2 D D 0 C 2 e C 2 ( 1 + 2 D 0 2 + 1 2 ν 1 2 + 1 2 + 1 4 ) δ 0 2 ν 1 2 D + δ 0 + ν 1 2 D 4 ν 1 2 D 0 + 8 ν 1 2 D + 5 ν 3 2 D + 4 ν 1 D + 2 D 0 + 2 ν 1 D ( Z d ( T ) + W d ( T ) ) + δ 0 W d ( T ) + Z d 1 2 ( T ) W d 1 2 ( T ) δ 0 + 4 ν 1 2 D + 4 ν 1 D δ 0 + Z d 1 2 ( T ) W d 1 2 ( T ) ,

which implies that

Z d ( T ) + W d ( T ) C 2 e 2 C 2 ( 1 + D 0 2 ) δ 0 2 ν 1 2 D + δ 0 + ν 1 2 D ( 4 + 3 D 0 ) .

Next, assuming that for a suitably large (universal) constant C and we take ν satisfying

(3.61) C e C ( 1 + D 0 2 ) ( X d ( 0 ) + 2 D 0 2 + 2 ) ν ,

then set

(3.62) D = C 1 e 2 C 1 ( 1 + D 0 2 ) ( X d ( 0 ) + 2 D 0 2 + 2 ) .

Thus, we can derive that

C 2 e 2 C 2 ( 1 + D 0 2 ) D ( 4 + 3 D 0 ) ν 1 2 16 .

Then, we choose

(3.63) δ = C 2 e 2 C 2 ( 1 + D 0 2 ) ( δ 0 + δ 0 2 + ν 1 2 D ( 4 + 3 D 0 ) ) .

Hence, we end up with

X d ( T ) + Y d ( T ) D and Z d ( T ) + W d ( T ) δ .

From (3.61)–(3.63), we see that assumption (3.2) is recovered, from which we see that assumptions (3.58) and (3.59) are recovered. This also ensures that b ( t ) L is small for all T 0 , as we have

b ( t ) L C ν 1 ( ν Λ b ( t ) B ˙ 2 , 1 N 2 1 ) C ν 1 D .

If ν and the compressible part of the data fulfill (3.61), then defining D and δ according to (3.62) and (3.63) ensures that (3.1) is fulfilled for all T < T * . Then, by combining with the continuation criterion, one can conclude that T * = + and that (3.1) is satisfied for all time.

Acknowledgement

The authors would like to thank the anonymous referees for their valuable comments and suggestions that help to improve and clarify the article greatly.

  1. Funding information: The first author was partially supported by Zhejiang Province Science Fund (LY21A010009).

  2. Author contributions: All authors contributed equally to the manuscript and read and approved the final manuscript.

  3. Conflict of interest: The authors state no conflict of interest.

  4. Data availability statement: No data were used for the research described in the article.

References

[1] H. Bahouri, J. Y. Chemin, and R. Danchin, Fourier Analysis and Nonlinear Partial Differential Equations, Grundlehren der Mathematischen Wissenschaften (Fundamental Principles of Mathematical Sciences), vol. 343, Springer, Heidelberg, 2011, DOI: https://doi.org/10.1007/978-3-642-16830-7.10.1007/978-3-642-16830-7Search in Google Scholar

[2] J. Y. Chemin, Localization in Fourier space and Navier-Stokes system. Phase space analysis of partial differential equations, Pisa 1 (2004), 53–136. Search in Google Scholar

[3] J. Chemin and N. Masmoudi, About lifespan of regular solutions of equations related to viscoelastic fluids, SIAM J. Math. Anal. 33 (2001), 84–112, DOI: https://doi.org/10.1137/S0036141099359317.https://doi.org/10.1137/S0036141099359317Search in Google Scholar

[4] Y. M. Chen and P. Zhang, The global existence of small solutions to the incompressible viscoelastic fluid system in 2 and 3 space dimensions, Commun. Partial Differ. Equ. 31 (2006), 1793–1810, DOI: https://doi.org/10.1080/03605300600858960.https://doi.org/10.1080/03605300600858960Search in Google Scholar

[5] X. Cui and X. Hu, Incompressible limit of three dimensional compressible viscoelastic systems with vanishing shear viscosity, Archive Rational Mech. Anal. 245 (2022), 753–807, DOI: https://doi.org/10.1007/s00205-022-01795-z.https://doi.org/10.1007/s00205-022-01795-zSearch in Google Scholar

[6] X. Cui and X. Hu, Vanishing shear viscosity limit and incompressible limit of two dimensional compressible dissipative elastodynamics, Commun. Math. Phys. 402 (2023), 3253–3336, DOI: https://doi.org/10.1007/s00220-023-04805-7.https://doi.org/10.1007/s00220-023-04805-7Search in Google Scholar

[7] R. Danchin, Global existence in critical spaces for compressible Navier-Stokes equations, Invent. Math. 141 (2000), 579–614, DOI: https://doi.org/10.1007/s002220000078.https://doi.org/10.1007/s002220000078Search in Google Scholar

[8] R. Danchin and P. Mucha, Compressible Navier-Stokes system: large solutions and incompressible limit, Adv. Math. 320 (2017), 904–925, DOI: https://doi.org/10.1016/j.aim.2017.09.025.https://doi.org/10.1016/j.aim.2017.09.025Search in Google Scholar

[9] D. Fang, B. Han, and T. Zhang, Global existence in critical spaces for density-dependent incompressible viscoelastic fluids, Acta Appl. Math. 130 (2014), 51–80, DOI: https://doi.org/10.1007/s10440-013-9838-z.https://doi.org/10.1007/s10440-013-9838-zSearch in Google Scholar

[10] B. Han and R. Zi, Dispersive effect and global well posedness of the compressible viscoelastic fluids, J. Differ. Equ. 269 (2020), 9254–9296, DOI: https://doi.org/10.1016/j.jde.2020.06.059.https://doi.org/10.1016/j.jde.2020.06.059Search in Google Scholar

[11] X. Hu, Global existence of weak solutions to two dimensional compressible viscoelastic flows, J. Differ. Equ. 265 (2018), 3130–3167, DOI: https://doi.org/10.1016/j.jde.2018.05.001.https://doi.org/10.1016/j.jde.2018.05.001Search in Google Scholar

[12] X. Hu and F. Lin, On the Cauchy problem for two dimensional incompressible viscoelastic flows, 2016, arXiv :1601.03497, DOI: https://doi.org/1601.03497, 2016 - arxiv.org.Search in Google Scholar

[13] X. Hu and F. Lin, Global solutions of two-dimensional incompressible viscoelastic flows with discontinuous initial data, Commun. Pure Appl. Math. 69 (2016), 372–404, DOI: https://doi.org/10.1002/cpa.21561.https://doi.org/10.1002/cpa.21561Search in Google Scholar

[14] X. Hu, Y. Ou, D. Wang, and L. Yang, Incompressible limit for compressible viscoelastic flows with large velocity, Adv. Nonl. Anal. 12 (2023), 20220324, DOI: https://doi.org/10.1515/anona-2022-0324.https://doi.org/10.1515/anona-2022-0324Search in Google Scholar

[15] X. Hu and D. Wang, Global existence for the multi-dimensional compressible viscoelastic flows, J. Differ. Equ. 250 (2011), 1200–1231, DOI: https://doi.org/10.1016/j.jde.2010.10.017.https://doi.org/10.1016/j.jde.2010.10.017Search in Google Scholar

[16] X. Hu and G. Wu, Global existence and optimal decay rates for three-dimensional compressible viscoelastic flows, SIAM J. Math. Anal. 45 (2013), 2815–2833, DOI: https://doi.org/10.1137/120892350.https://doi.org/10.1137/120892350Search in Google Scholar

[17] X. Hu and H. Wu, Long-time behavior and weak-strong uniqueness for incompressible viscoelastic flows, Discrete Contin. Dyn. Syst 35 (2015), 3437–3461, DOI: https://doi.org/10.3934/dcds.2015.35.3437.https://doi.org/10.3934/dcds.2015.35.3437Search in Google Scholar

[18] X. Hu and W. Zhao, Global existence of compressible dissipative elastodynamics systems with zero shear viscosity in two dimensions, Archive Ration. Mech. Anal. 235 (2020), 1177–1243, DOI: https://doi.org/10.1007/s00205-019-01443-z.https://doi.org/10.1007/s00205-019-01443-zSearch in Google Scholar

[19] X. Huo and W. A. Yong, Structural stability of a 1D compressible viscoelastic fluid model, J. Differ. Equ. 261 (2016), 1264–1284, DOI: https://doi.org/10.1016/j.jde.2016.03.041.https://doi.org/10.1016/j.jde.2016.03.041Search in Google Scholar

[20] F. Jiang and J. Song, Strong solutions of the equations for viscoelastic fluids in some classes of large data, J. Differ. Equ. 282 (2021), 148–183, DOI: https://doi.org/10.1016/j.jde.2021.02.020.https://doi.org/10.1016/j.jde.2021.02.020Search in Google Scholar

[21] T. Kato and G. Ponce, Commutator estimates and the Euler and Navier-Stokes equations, Commun. Pure Appl. Math. 41 (1988), 891–907, DOI: https://doi.org/10.1002/cpa.3160410704.https://doi.org/10.1002/cpa.3160410704Search in Google Scholar

[22] Z. Lei, C. Liu, and Y. Zhou, Global solutions for incompressible viscoelastic fluids, Archive Ration. Mech. Anal. 188 (2008), 371–398, DOI: https://doi.org/10.1007/s00205-007-0089-x.https://doi.org/10.1007/s00205-007-0089-xSearch in Google Scholar

[23] Z. Lei and Y. Zhou, Global existence of classical solutions for the two-dimensional Oldroyd model via the incompressible limit, SIAM J. Math. Anal. 37 (2005), 797–814, DOI: https://doi.org/10.1137/040618813.https://doi.org/10.1137/040618813Search in Google Scholar

[24] Y. Li, R. Wei, and Z. Yao, Optimal decay rates for the compressible viscoelastic flows, J. Math. Phys. 57 (2016), 111506, DOI: https://doi.org/10.1063/1.4967975.https://doi.org/10.1063/1.4967975Search in Google Scholar PubMed PubMed Central

[25] F. Lin, C. Liu, and P. Zhang, On hydrodynamics of viscoelastic fluids, Commun. Pure Appl. Math 58 (2005), 1437–1471, DOI: https://doi.org/10.1002/cpa.20074.https://doi.org/10.1002/cpa.20074Search in Google Scholar

[26] F. Lin and P. Zhang, On the initial-boundary value problem of the incompressible viscoelastic fluid system, Commun. Pure Appl. Math 61 (2008), 539–558, DOI: https://doi.org/10.1002/cpa.20219.https://doi.org/10.1002/cpa.20219Search in Google Scholar

[27] C. Liu, Walkington and Noel J, An Eulerian description of fluids containing viscoelastic particles, Archive Ration. Mech. Anal. 159 (2001), no. 3, 229–252, DOI: https://doi.org/10.1007/s002050100158.https://doi.org/10.1007/s002050100158Search in Google Scholar

[28] J. Qian, Well-posedness in critical spaces for incompressible viscoelastic fluid system, Nonl. Anal. Theory Meth. Appl. 72 (2010), 3222–3234, DOI: https://doi.org/10.1016/j.na.2009.12.022.https://doi.org/10.1016/j.na.2009.12.022Search in Google Scholar

[29] J. Qian and Z. Zhang, Global well posedness for compressible viscoelastic fluids near equilibrium, Archive Ration. Mech. Anal. 198 (2010), 835–868, DOI: https://doi.org/10.1007/s00205-010-0351-5.https://doi.org/10.1007/s00205-010-0351-5Search in Google Scholar

[30] W. Wang and Y. Zhao, On the Rayleigh-Taylor instability in compressible viscoelastic fluids, J. Math. Anal. Appl. 463 (2018), 198–221, DOI: https://doi.org/10.1016/j.jmaa.2018.03.018.https://doi.org/10.1016/j.jmaa.2018.03.018Search in Google Scholar

[31] G. Wu, Z. Gao, and Z. Tian, Time decay rates for the compressible viscoelastic flows, J. Math. Anal. Appl. 452 (2017), 990–1004, DOI: https://doi.org/10.1016/j.jmaa.2017.03.044.https://doi.org/10.1016/j.jmaa.2017.03.044Search in Google Scholar

[32] T. Zhang and D. Fang, Global existence of strong solution for equations related to the incompressible viscoelastic fluids in the critical Lp framework, SIAM J. Math. Anal 44 (2013), 2266–2288, DOI: https://doi.org/10.1137/110851742.https://doi.org/10.1137/110851742Search in Google Scholar

Received: 2024-04-06
Revised: 2024-09-23
Accepted: 2025-01-03
Published Online: 2025-01-23

© 2025 the author(s), published by De Gruyter

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

Articles in the same Issue

  1. Research Articles
  2. Incompressible limit for the compressible viscoelastic fluids in critical space
  3. Concentrating solutions for double critical fractional Schrödinger-Poisson system with p-Laplacian in ℝ3
  4. Intervals of bifurcation points for semilinear elliptic problems
  5. On multiplicity of solutions to nonlinear Dirac equation with local super-quadratic growth
  6. Entire radial bounded solutions for Leray-Lions equations of (p, q)-type
  7. Combinatorial pth Calabi flows for total geodesic curvatures in spherical background geometry
  8. Sharp upper bounds for the capacity in the hyperbolic and Euclidean spaces
  9. Positive solutions for asymptotically linear Schrödinger equation on hyperbolic space
  10. Existence and non-degeneracy of the normalized spike solutions to the fractional Schrödinger equations
  11. Existence results for non-coercive problems
  12. Ground states for Schrödinger-Poisson system with zero mass and the Coulomb critical exponent
  13. Geometric characterization of generalized Hajłasz-Sobolev embedding domains
  14. Subharmonic solutions of first-order Hamiltonian systems
  15. Sharp asymptotic expansions of entire large solutions to a class of k-Hessian equations with weights
  16. Stability of pyramidal traveling fronts in time-periodic reaction–diffusion equations with degenerate monostable and ignition nonlinearities
  17. Ground-state solutions for fractional Kirchhoff-Choquard equations with critical growth
  18. Existence results of variable exponent double-phase multivalued elliptic inequalities with logarithmic perturbation and convections
  19. Homoclinic solutions in periodic partial difference equations
  20. Classical solution for compressible Navier-Stokes-Korteweg equations with zero sound speed
  21. Properties of minimizers for L2-subcritical Kirchhoff energy functionals
  22. Asymptotic behavior of global mild solutions to the Keller-Segel-Navier-Stokes system in Lorentz spaces
  23. Blow-up solutions to the Hartree-Fock type Schrödinger equation with the critical rotational speed
  24. Qualitative properties of solutions for dual fractional parabolic equations involving nonlocal Monge-Ampère operator
  25. Regularity for double-phase functionals with nearly linear growth and two modulating coefficients
  26. Uniform boundedness and compactness for the commutator of an extension of Riesz transform on stratified Lie groups
  27. Normalized solutions to nonlinear Schrödinger equations with mixed fractional Laplacians
  28. Existence of positive radial solutions of general quasilinear elliptic systems
  29. Low Mach number and non-resistive limit of magnetohydrodynamic equations with large temperature variations in general bounded domains
  30. Sharp viscous shock waves for relaxation model with degeneracy
  31. Bifurcation and multiplicity results for critical problems involving the p-Grushin operator
  32. Asymptotic behavior of solutions of a free boundary model with seasonal succession and impulsive harvesting
  33. Blowing-up solutions concentrated along minimal submanifolds for some supercritical Hamiltonian systems on Riemannian manifolds
  34. Stability of rarefaction wave for relaxed compressible Navier-Stokes equations with density-dependent viscosity
  35. Singularity for the macroscopic production model with Chaplygin gas
  36. Global strong solution of compressible flow with spherically symmetric data and density-dependent viscosities
  37. Global dynamics of population-toxicant models with nonlocal dispersals
  38. α-Mean curvature flow of non-compact complete convex hypersurfaces and the evolution of level sets
  39. High-energy solutions for coupled Schrödinger systems with critical growth and lack of compactness
  40. On the structure and lifespan of smooth solutions for the two-dimensional hyperbolic geometric flow equation
  41. Well-posedness for physical vacuum free boundary problem of compressible Euler equations with time-dependent damping
  42. On the existence of solutions of infinite systems of Volterra-Hammerstein-Stieltjes integral equations
  43. Remark on the analyticity of the fractional Fokker-Planck equation
  44. Continuous dependence on initial data for damped fourth-order wave equation with strain term
  45. Unilateral problems for quasilinear operators with fractional Riesz gradients
  46. Boundedness of solutions to quasilinear elliptic systems
  47. Existence of positive solutions for critical p-Laplacian equation with critical Neumann boundary condition
  48. Non-local diffusion and pulse intervention in a faecal-oral model with moving infected fronts
Downloaded on 29.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/anona-2024-0062/html
Scroll to top button