Next Article in Journal
Decentralized Stochastic Recursive Gradient Method for Fully Decentralized OPF in Multi-Area Power Systems
Previous Article in Journal
Reinforcement Learning with Value Function Decomposition for Hierarchical Multi-Agent Consensus Control
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Scaling-Invariant Serrin Criterion via One Row of the Strain Tensor for the Navier–Stokes Equations

1
School of Mathematics and Informational Technology, Yuncheng University, Yuncheng 044000, China
2
College of Science, Nanchang Institute of Technology, Nanchang 330099, China
*
Author to whom correspondence should be addressed.
Mathematics 2024, 12(19), 3063; https://doi.org/10.3390/math12193063 (registering DOI)
Submission received: 29 August 2024 / Revised: 27 September 2024 / Accepted: 28 September 2024 / Published: 30 September 2024
(This article belongs to the Section Difference and Differential Equations)

Abstract

:
Miller (Arch. Rational Mech. Anal., 2020) posed the question of whether it is possible to prove the Navier–Stokes regularity criterion using only one entry of the strain tensor S i j . Although this paper does not fully address this question, we do establish several scaling-invariant Serrin-type criteria based on one row of the strain tensor.

1. Introduction

The Navier–Stokes equation (NSE for short) is one of the fundamental equations of fluid dynamics. The three-dimensional viscous incompressible NSE reads as
t u + u · u Δ u + π = 0 , x R 3 , t > 0 · u = 0 , x R 3 , t > 0 u t = 0 = u 0 ( x ) , x R 3 ,
where u = ( u 1 , u 2 , u 3 ) —the velocity field—and π —the pressure—are the unknowns, and u 0 is the initial velocity field. We will denote h = ( x 1 , x 2 ) as the horizontal gradient operator and Δ h = x 1 2 + x 2 2 as the horizontal Laplacian, where Δ and ∇ are the Laplacian and the gradient operators, respectively. Here, we utilize the following notations:
( u · ) v = i = 1 3 u i x i v k , ( k = 1 , 2 , 3 ) , · u = i = 1 3 x i u i , S = S i j = 1 2 x i u x j + x j u x i
and for the sake of simplicity, we denote the partial derivative with respect to x i as i .
In the pioneering work of Leray [1] and Hopf [2], they found that, for any u 0 L σ 2 ( R 3 ) , one can construct a global weak solution to (1), namely, a function u that, for each T > 0 , is in the class
u L ( 0 , T ; L σ 2 ( R 3 ) ) L 2 ( 0 , T ; H 1 ( R 3 ) )
and solves (1) in a distributional sense. Here, L σ 2 ( R 3 ) is the subspace of L 2 ( R 3 ) of divergence-free vector functions. In addition, such a solution u satisfies the so-called energy inequality:
u ( t ) L 2 2 + 2 0 t u ( τ ) L 2 2 d τ u 0 L 2 2 , t 0 .
As we all know, for strong solutions, this inequality passes to the corresponding equality, while Leray–Hopf weak solutions, as these solutions are commonly referred to, must exist globally in time. They are not known to be either regular or unique.
It is known that if a weak solution u of NSE (1) satisfies the Ladyzhenskaya–Prodi–Serrin criteria [3,4,5,6]
u L q ( 0 , T ; L p ( R 3 ) ) , 2 q + 3 p = 1 , p [ 3 , ] ,
then u must be regular in R 3 × ( 0 , T ] . From a physical point of view, the velocity describes how a fluid is advected, the vorticity describes how it is rotated, and the strain describes how a parcel of fluid is deformed. Therefore, the vorticity and the strain tensor are also important physical quantities to describe fluid behaviors. Beale–Kato–Majda [7] and Beirão da Veiga [8] extended the Ladyzhenskaya–Prodi–Serrin criteria to the vorticity ω = × u and showed that if
ω L q ( 0 , T ; L p ( R 3 ) ) , 2 q + 3 p = 2 , 3 2 < p < ,
then u is regular. More generally, Ponce in [9] proved that the weak solution u became regular if the strain tensor satisfies
S L 1 ( 0 , T ; L ( R 3 ) ) .
We should mention that the in above cases, 2 q + 3 p = 1 and 2 q + 3 p = 2 , the function space L t q L x p is invariant under the following scaling:
u ( x , t ) u λ ( x , t ) = λ u λ x , λ 2 t , λ > 0 ,
where u λ is still a solution to (1) with initial data u 0 λ : = λ u 0 ( λ x ) .
There have also been several scale-critical, component-reduction-type regularity criteria that require control over only specific parts of the solution. These include the regulation of just two components of the vorticity ( ω 1 , ω 2 , 0 ) [10,11], the derivative in a single direction 3 u [12,13,14], and only one component of the velocity u 3 [15,16,17]. In addition, for the literature related to the regularity criteria of NSE in BMO spces, we refer to [18,19].
Recently, the study of the regularity criterion, which involves the positive part of the middle eigenvalue of the strain tensor, has gained popularity. Specifically, it has been proven that the regularity of weak solutions exists within the class of
λ 2 + L q ( 0 , T ; L p ( R 3 ) ) with λ 2 + = max 0 , λ 2 , 2 q + 3 p = 2 , 3 2 < p .
This was first proven by Neustupa and Penel [20,21,22] and independently by Miller [23] using somewhat different methods. Subsequently, Miller also proved regularity criteria that only require control of the velocity, vorticity, or the positive part of the second eigenvalue of the strain matrix in the sum space of two scale critical spaces [24]. In particular, in [23], the author posed the question of whether it is possible to establish the NSE regularity criterion based solely on one entry of the strain tensor S i j . This paper does not fully answer this question; however, we will demonstrate several regularity criteria using one row of the strain tensor in certain critical spaces.
First, we will prove the following.
Theorem 1. 
Let u 0 H 1 , and suppose that u be a weak solution in ( 0 , T ) of NSE (1) if one row of the strain tensor S = S i j 1 2 i u j + j u i ( 1 i , j 3 ) satisfies
S 3 j L 2 2 r 0 , T ; B ˙ , r , r ( 0 , 1 ) , j = 1 , 2 , 3 .
Then,
u L 0 , T ; H 1 R 3 L 2 0 , T ; H 2 R 3 ,
and thus, u C ( 0 , T ] × R 3 .
Remark 1. 
Theorem 1 addresses the case of the strain tensor component, and it partially answers the question raised by [23]. In addition, due to the embedding
L 3 r ( R 3 ) B ˙ , r ( R 3 ) ,
with 0 < r 3 2 , we know that the condition (8) improves and develops the Ladyzhenskaya–Prodi–Serrin criteria.
Next, we will observe that we can also articulate the regularity criterion for one row of the strain tensor S 3 j in terms of the sum space L T p L x q + L T 1 L x .
Theorem 2. 
Let u 0 H 1 , and suppose that u be a weak solution in ( 0 , T ) of NSE (1) with initial data u 0 if one row of the strain tensor S 3 j 1 2 3 u j + j u 3 = S ^ + S ˇ ( 1 j 3 ) satisfies
0 T S ^ ( t ) L p q d t + 0 T S ˇ ( t ) L d t < , 2 q + 3 p = 2 , 3 2 < p < .
Then, u is regular.
Remark 2. 
We will note that this is a significant advance because the regularity criterion in the sum space L T p L x q + L T 1 L x contains within it the whole family of regularity criteria in the spaces L T p L x q , where 2 p + 3 q = 2 , and q q + . This theorem is a generalization of the Navier–Stokes regularity criteria in sum spaces in [24].
Finally, we will present the regularity criteria for the 3D NSE (1) on the framework of mixed-norm Lorentz spaces. This approach is particularly motivated by the physical interpretation that fluid behavior can vary in different directions. Consequently, understanding the solutions of the NSE in anisotropic functional spaces appears to be a topic of independent interest.
Theorem 3. 
Let u 0 H 1 , and suppose that u be a weak solution in ( 0 , T ) of NSE (1) with initial data u 0 . If one row of the strain tensor S = S i j 1 2 i u j + j u i ( 1 i , j 3 ) satisfies
0 T S 3 j L x 1 p , L x 2 q , L x 3 r , 2 2 1 p + 1 q + 1 r d t < ,
where 2 < p , q , r with 1 1 p + 1 q + 1 r 0 , then the solution u is regular on ( 0 , T ] .
Remark 3. 
Theorem 3 can be considered a Serrin-type criterion that involves only one row of the strain tensor. In contrast, our global regularity criterion pertains to the components of the strain tensor, which scale like u . From this perspective, all of our results are optimal.
Remark 4. 
To the best of our knowledge, Theorem 3 is the first result that proves the regularity criterion by imposing one row of the strain tensor in the mixed-norm Lorentz space.
Remark 5. 
All the spaces mentioned in Theorem 1, Theorem 2, and Theorem 3 are scaling-invariant spaces under the natural 3D Navier–Stokes scaling. Furthermore, it appears that a slightly modified version of the techniques used in the aforementioned theorems can be applied to other incompressible fluid equations, such as the fractional NSE and the micropolar fluid equations.

2. Definitions and Notations

Before proceeding with the proofs of our results, we need to define several spaces. First, we will introduce some notation. Let S ( R 3 ) denote the Schwartz class of rapidly decreasing functions. Given f S ( R 3 ) , its Fourier transform F f = f ^ is defined as
f ^ ( ξ ) = R 3 f ( x ) e i x · ξ d x .
Let ( χ , φ ) be a pair of smooth functions taking values in the interval [ 0 , 1 ] , where χ is supported in the set B = { ξ R 3 : | ξ | 4 3 } , and φ is supported in C = { ξ R 3 : 3 4 | ξ | 8 3 } such that
χ ( ξ ) + j 0 φ ( 2 j ξ ) = 1 , ξ R 3 ,
j Z φ ( 2 j ξ ) = 1 , ξ R 3 { 0 } .
Denote h = F 1 φ and h ˜ = F 1 χ . We then define the homogeneous dyadic blocks Δ ˙ j and the homogeneous low-frequency cut-off operator S ˙ j as follows:
Δ ˙ j u = φ ( 2 j D ) u = 2 3 j R 3 h ( 2 j y ) u ( x y ) d y ,
and
S ˙ j u = χ ( 2 j D ) = 2 3 j R 3 h ˜ ( 2 j y ) u ( x y ) d y .
Definition 1 
(Homogeneous Besov spaces [25]). Let
S h = f S ; lim j S ˙ j f = 0 .
The homogeneous Besov space B ˙ p , r s is defined as
B ˙ p , r s = f S h ; f B ˙ p , r s <
where
f B ˙ p , r s = j Z 2 j s r Δ ˙ j f p r 1 r , if r < , f B ˙ p , r s = sup j Z 2 j s Δ ˙ j f p , r = .
Lemma 1 
([26]). Let 0 < r < 1 . For f B ˙ , r R 3 and g , h H 1 R 3 , then there exists a constant C > 0 such that
R 3 f g h d x C f B ˙ r , g L 2 h L 2 1 r h L 2 r + h L 2 g L 2 1 r g L 2 r .
We will further define the Sobolev spaces.
Definition 2. 
For all α R , let
u H α 2 = R 3 1 + | ξ | 2 α u ^ ( ξ ) 2 d ξ ,
and let
H α R 3 = u S R 3 : u H α < + ,
where S R 3 is the space of tempered distributions.
We have defined the space H α R 3 ; now, we will define the next space H ˙ α R 3 .
Definition 3. 
For all α R , let
u H ˙ α 2 = R 3 | ξ | 2 α u ^ ( ξ ) 2 d ξ ,
and let
H ˙ α R 3 = u S R 3 : u H ˙ α < + .
Next, we define the sum spaces, which plays an essential role our in Theorem 2.
Definition 4 
(Lebesgue sum spaces). Let X and Y be Banach spaces, and let V be a vector space such that X , Y V . Then,
X + Y = x + y : x X , y Y .
Furthermore, X + Y is a Banach space with norm
f X + Y = inf g + h = f g X + h Y .
We now introduce Lorentz spaces and mixed-norm Lorentz spaces.
Given a measurable function f : R n R on X, its distribution function d f defined on [ 0 , ) is as follows:
d f ( α ) = μ ( { x X : | f ( x ) | > α } ) .
We now define its decreasing rearrangement f * : [ 0 , ) [ 0 , ] as
f * ( t ) = inf α : d f ( α ) t ,
with the convention that inf = . The point of this definition is that f and f * have the same distribution function,
d f * ( α ) = d f ( α ) ,
but f * is a positive non-increasing scalar function.
Definition 5 
(Lorentz spaces). Let ( p , q ) [ 1 , ] 2 be the Lorentz space L p , q ( R 3 ) that consists of all measurable functions f for which the quantity
f L p , q : = 0 [ t 1 p f * ( t ) ] q d t t 1 q q < , sup 0 < t < t 1 p f * ( t ) q = ,
is finite.
In order to lead the following definition involving anisotropic Lorentz space, we denote f = f ( x 1 , x 2 , x 3 ) as a measurable function defined on R 3 , f * ( t ) = f * 1 , * 2 , * 3 t 1 , t 2 , t 3 .
Definition 6 
(Mixed-norm Lorentz spaces). Let multi-indexes p = ( p 1 , p 2 , p 3 ) , q = ( q 1 , q 2 , q 3 ) be such that if 0 < p i < , then 0 < q i , and if p i = , then q i = for every i = 1 , 2 , 3 . An anisotropic Lorentz space L p 1 , q 1 R x 1 ; L p 2 , q 2 R x 2 ; L p 3 , q 3 R x 3 is the set of functions for which the following norm is finite:
f L x 1 p 1 , q 1 L x 2 p 2 , q 2 L x 3 p 3 , q 3 : = 0 0 0 [ t 1 1 p 1 t 2 1 p 2 t 3 1 p 3 f * 1 , * 2 , * 3 ( t 1 , t 2 , t 3 ) ] q 1 d t 1 t 1 q 2 q 1 d t 2 t 2 q 3 q 2 d t 3 t 3 1 q 3 .
Lemma 2. 
There exists a positive constant C such that
f L x 1 2 p p 2 , 2 L x 2 2 q q 2 , 2 L x 3 2 r r 2 , 2 C 1 f L 2 1 p 2 f L 2 1 q 3 f L 2 1 r f L 2 1 1 p + 1 q + 1 r ,
for every f C 0 R 3 , where 2 < p , q , r , 1 1 p + 1 q + 1 r 0 .
Proof. 
Let Λ 1 p be the Fourier multiplier defined as
F 1 Λ 1 p f ξ 1 , x 2 , x 3 = ξ 1 p F 1 f ξ 1 , x 2 , x 3
with
F 1 f ξ 1 , x 2 , x 3 = R e i ξ 1 x 1 f x 1 , x 2 , x 3 d x 1 ,
Λ 2 p and Λ 3 p can be defined analogously. Then, through the Sobolev’s embedding theorem, the Minkowski’s inequality and the Hölder’s inequality to obtain
f L x 1 2 p p 2 , 2 L x 2 2 q q 2 , 2 L x 3 2 r r 2 , 2 C Λ 1 1 p f L x 1 2 L x 2 2 q q 2 , 2 L x 3 2 r r 2 , 2 Λ 1 1 p f L x 2 2 q q 2 , 2 L x 1 2 L x 3 2 r r 2 , 2 C Λ 2 1 q Λ 1 1 p f L x 1 , x 2 2 L x 3 2 r r 2 , 2 C Λ 2 1 q Λ 1 1 p f L x 3 2 r r 2 , 2 L x 1 , x 2 2 C Λ 3 1 r Λ 2 1 q Λ 1 1 p f L 2 .
Combining the Fourier–Plancherel formula and the Hölder’s inequality, we have
C Λ 3 1 r Λ 2 1 q Λ 1 1 p f L 2 C R 3 ξ 1 2 p ξ 2 2 q ξ 3 2 r F f ξ 1 , ξ 2 , ξ 3 2 d ξ 1 d ξ 2 d ξ 3 1 2 = C R 3 ξ 1 2 p | F f ( ξ ) | 2 p ξ 2 2 q | F f ( ξ ) | 2 q ξ 3 2 r | F f ( ξ ) | 2 r | F f ( ξ ) | 2 2 p + 2 q + 2 r d ξ 1 d ξ 2 d ξ 3 1 2 C F f L 2 1 1 p 1 q 1 r R 3 ξ 1 2 | F f | 2 d ξ 1 2 p R 3 ξ 2 2 | F f | 2 d ξ 1 2 q R 3 ξ 3 2 | F f | 2 d ξ 1 2 r C 1 f L 2 1 p 2 f L 2 1 q 3 f L 2 1 r f L 2 1 1 p + 1 q + 1 r .

3. Proof of Theorem 1

In this section, we will consider the regularity criteria for one row of strain tensors in critical Besov spaces.

3.1. h u L 2 Estimates

First, we obtain some estimates of the horizontal gradient. Taking the inner product of NSE (1) with Δ h u in L 2 ( R 3 ) , one has
1 2 d d t h u L 2 2 + h u L 2 2 = R 3 [ ( u · ) u ] · Δ h u d x = R 3 i , j = 1 3 k = 1 2 k u i i u j k u j d x = R 3 i = 1 3 j = 1 2 k = 1 2 k u i i u j k u j d x + R 3 i = 1 3 k = 1 2 k u i i u 3 k u 3 d x = ( R 3 i = 1 2 j = 1 2 k = 1 2 k u i i u j k u j d x + R 3 j = 1 2 k = 1 2 k u 3 3 u j k u j d x + R 3 i = 1 2 k = 1 2 k u i i u 3 k u 3 d x + R 3 k = 1 2 k u 3 3 u 3 k u 3 d x ) = I ( t ) + J ( t ) + K ( t ) + L ( t ) .
Next, we will estimate the right-hand side of the above term I ( t ) , J ( t ) , K ( t ) , and L ( t ) by using the incompressibility condition and Lemma 1.
I ( t ) = R 3 i = 1 2 j = 1 2 k = 1 2 k u i i u j k u j d x = R 3 i = 1 2 j = 1 2 1 u i i u j 1 u j d x + R 3 i = 1 2 j = 1 2 2 u i i u j 2 u j d x = ( R 3 i = 1 2 1 u i i u 1 1 u 1 d x + R 3 i = 1 2 1 u i i u 2 1 u 2 d x + R 3 i = 1 2 2 u i i u 1 2 u 1 d x + R 3 i = 1 2 2 u i i u 2 2 u 2 d x ) = ( R 3 ( 1 u 1 ) 3 + ( 2 u 2 ) 3 d x + R 3 ( 1 u 1 + 2 u 2 ) ( 1 u 2 ) 2 d x + R 3 ( 1 u 1 + 2 u 2 ) 1 u 2 2 u 1 d x + R 3 ( 1 u 1 + 2 u 2 ) 1 u 2 2 u 1 d x ) = R 3 3 u 3 ( 1 u 1 ) 2 1 u 1 2 u 2 + ( 2 u 2 ) 2 d x + R 3 3 u 3 ( 1 u 2 ) 2 d x + R 3 3 u 3 1 u 2 2 u 1 d x + R 3 3 u 3 1 u 2 2 u 1 d x C R 3 | S 33 | | h u | 2 d x C S 33 B ˙ , r h u L 2 2 r h u L 2 r C S 33 B ˙ , r 2 2 r h u L 2 2 + ϵ h u L 2 2 .
As to J ( t ) + K ( t ) , we take the same trick as I ( t ) to obtain
J ( t ) + K ( t ) = R 3 k = 1 2 j = 1 2 k u 3 3 u j k u j d x + R 3 k = 1 2 i = 1 2 k u i i u 3 k u 3 d x = R 3 k = 1 2 k u 3 k u 1 ( 3 u 1 + 1 u 3 ) + k = 1 2 k u 3 k u 2 ( 3 u 2 + 2 u 3 ) d x C R 3 | S 31 + S 32 | | h u | 2 d x C S 3 j B ˙ , r h u L 2 2 r h u L 2 r C S 3 j B ˙ , r 2 2 r h u L 2 2 + ϵ h u L 2 2 , j = 1 , 2 .
and
L ( t ) C S 33 B ˙ , r 2 2 r h u L 2 2 + ϵ h u L 2 2 .
Plugging estimates (21)–(23) into (20) give us
1 2 d d t h u L 2 2 + 1 2 h u L 2 2 C S 3 j B ˙ , r 2 2 r h u L 2 2 , j = 1 , 2 , 3 .
Gronwall’s inequality implies that
h u L 2 2 + 0 T h u L 2 2 d τ C .

3.2. u L 2 Estimates

Taking the inner product of the Equation (1) with Δ u in L 2 , one obtains
1 2 d d t u L 2 2 + Δ u L 2 2 = i , j , k = 1 3 R 3 u i i u j k k u j d x = j = 1 3 R 3 u 3 3 u j Δ h u j d x + i = 1 2 j = 1 3 R 3 u i i u j Δ u j d x + j = 1 3 R 3 u 3 3 u j 33 u j d x = M 1 ( t ) + M 2 ( t ) + M 3 ( t ) .
The proof of the bound (26) has been shown in [17] and concludes that
| M i ( t ) | C R 3 | h u | | u | 2 d x C h u L 2 u L 4 2 C h u L 2 u L 2 1 2 h u L 2 Δ u L 2 1 2 , i = 1 , 2 , 3 .
Gathering the above two estimates, we obtain
1 2 d d t u L 2 2 + Δ u L 2 2 C h u L 2 u L 2 1 2 h u L 2 Δ u L 2 1 2 ϵ h u L 2 4 3 Δ u L 2 2 + C u L 2 2 .
Hence,
1 2 d d t u L 2 2 + 1 2 Δ u L 2 2 C u L 2 2 .
Theorem 1 follows from the Grönwall inequality.

4. The Proof of Theorem 2

In this section, we will prove Theorem 2. Based on the proofs of Theorem 1, it is sufficient to demonstrate that there exists a positive constant K such that
0 T S ^ ( t ) L p q d t + 0 T S ˇ ( t ) L d t K , with 2 q + 3 p = 2 and 3 2 < p < ,
and then we derive
h u L ( 0 , T ; L 2 ( R 3 ) ) L 2 ( 0 , T ; H 1 ( R 3 ) ) .
Recall that in Theorem 1, the h u satisfies
1 2 d d t h u L 2 2 + h u L 2 2 C R 3 | S 3 j | | h u | 2 d x ( j = 1 , 2 , 3 ) = C R 3 | h u | 2 ( S ^ + S ˇ ) d x C S ^ L p h u L 2 p p 1 2 + C S ˘ L h u L 2 2 C S ^ L p h u L 2 2 3 p h u L 2 3 p + C S ˇ L h u L 2 2 C S ^ L p q h u L 2 2 + C S ˇ L h u L 2 2 + ϵ h u L 2 2 ,
where q = 2 p 2 p 3 , 3 2 < p < . Therefore, we can conclude that
1 2 d d t h u L 2 2 + h u L 2 2 C S ^ L p q + S ˇ L h u L 2 2 .
Applying Grönwall’s inequality, this implies that
h u L ( 0 , T ; L 2 ( R 3 ) ) L 2 ( 0 , T ; H 1 ( R 3 ) ) ,
so this completes the proof of Theorem 2.

5. Proof of Theorem 3

The proof of Theorem 3 is not significantly different from the previous one. We begin with Equation (30), applying Hölder’s inequality and Young’s inequality, and combining these with Lemma 2 yields that
1 2 d d t h u L 2 2 + h u L 2 2 C R 3 | S 3 j | | h u | 2 d x ( j = 1 , 2 , 3 ) C S 3 j L x 1 p , L x 2 q , L x 3 r , h u L x 1 2 p p 2 , 2 L x 2 2 q q 2 , 2 L x 3 2 r r 2 , 2 h u L 2 C S 3 j L x 1 p , L x 2 q , L x 3 r , 1 h u L 2 1 p 2 h u L 2 1 q 3 h u L 2 1 r h u L 2 1 1 p + 1 q + 1 r h u L 2 C S 3 j L x 1 p , L x 2 q , L x 3 r , h u L 2 1 p + 1 q + 1 r h u L 2 2 1 p + 1 q + 1 r C S 3 j L x 1 p , L x 2 q , L x 3 r , 2 2 1 p + 1 q + 1 r h u L 2 2 + ϵ h u L 2 2 .
Next, integrating in time interval ( 0 , T ] and applying Gronwall’s lemma gives
h u ( t ) L 2 2 + 0 T h u ( t ) d t h u 0 L 2 2 exp C 0 T S 3 j L x 1 p , L x 2 q , L x 3 r , 2 2 1 p + 1 q + 1 r d t .
The same arguments as in the conclusion of the previous theorem lead to the desired result. Theorem 3 is proved.

Discussion and Conclusions

As is well known in fluid mechanics, velocity describes the advection of a fluid, vorticity characterizes its rotation, and strain quantifies the deformation of a fluid parcel. Consequently, the strain tensor is a crucial physical quantity for understanding fluid behavior. From a mathematical perspective, our results partially address the questions raised by Miller in [23]. Furthermore, our findings suggest that the strain tensor S plays a less significant role than the individual components of the strain tensor in the regularity theory of solutions to the incompressible Navier–Stokes equations. In a certain sense, our results represent a preliminary step toward resolving Miller’s open problem.

Author Contributions

Formal analysis and writing—original draft, J.D.; Review and editing, F.W. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

There is no mention of data availability anywhere in the paper.

Conflicts of Interest

The authors declare that there are no conflicts of interest.

References

  1. Leray, J. Sur le mouvement d’un liquide visqueux emplissant l’espace. Acta Math. 1934, 63, 193–248. [Google Scholar] [CrossRef]
  2. Hopf, E. Über die Anfangswertaufgabe für die hydrodynamischen Grundgleichungen. Erhard Schmidt zu seinem 75. Geburtstag gewidmet. Math. Nachrichten 1950, 4, 213–231. [Google Scholar] [CrossRef]
  3. Escauriaza, L.; Seregin, G. L3,∞-solutions of the Navier-Stokes equations and backward uniqueness. Nonlinear Probl. Math. Phys. Relat. Top. II 2003, 18, 353–366. [Google Scholar] [CrossRef]
  4. Ladyzhenskaya, O.A. Uniqueness and smoothness of generalized solutions of Navier-Stokes equations. Zap. Naucn. Sem. Leningrad. Otdel. Mat. Inst. Steklov. (LOMI) 1967, 5, 169–185. [Google Scholar]
  5. Prodi, G. Un teorema di unicita per le equazioni di Navier-Stokes. Ann. Mat. Pura Appl. 1959, 48, 173–182. [Google Scholar] [CrossRef]
  6. Serrin, J. On the interior regularity of weak solutions of the Navier-Stokes equations. Arch. Ration. Mech. Anal. 1962, 9, 187–195. [Google Scholar] [CrossRef]
  7. Beale, J.T.; Kato, T.; Majda, A. Remarks on the breakdown of smooth solutions for the 3-D Euler equations. Commun. Math. Phys. 1984, 94, 61–66. [Google Scholar] [CrossRef]
  8. Beirao da Veiga, H. A new regularity class for the Navier-Stokes equations in Rn. Chin. Ann. Math. 1995, 16, 407–412. [Google Scholar]
  9. Ponce, G. Remarks on a paper by J T Beale, T. Kato, and A. Majda. Commun. Math. Phys. 1985, 98, 349–353. [Google Scholar] [CrossRef]
  10. Chae, D.; Choe, H.J. Regularity of solutions to the Navier-Stokes equations. Electron. J. Differ. Equ. 1999, 1999, 1–7. [Google Scholar]
  11. Zhang, Z.F.; Chen, Q.L. Regularity criterion via two components of vorticity on weak solutions to the Navier-Stokes equations in R3. J. Differ. Equ. 2005, 216, 470–481. [Google Scholar]
  12. Chen, H.; Le, W.J.; Qian, C.Y. Prodi-Serrin condition for 3D Navier-Stokes equations via one directional derivative of velocity. J. Differ. Equ. 2021, 298, 500–527. [Google Scholar] [CrossRef]
  13. Kukavica, I.; Ziane, M. Navier-Stokes equations with regularity in one direction. J. Math. Phys. 2007, 48, 065203. [Google Scholar] [CrossRef]
  14. Ragusa, M.A.; Wu, F. Regularity criteria via one directional derivative of the velocity in anisotropic Lebesgue spaces to the 3D Navier-Stokes equations. J. Math. Anal. Appl. 2021, 502, 125286. [Google Scholar] [CrossRef]
  15. Cao, C.; Titi, E.S. Regularity criteria for the three-dimensional Navier-Stokes equations. INdiana Univ. Math. J. 2008, 57, 2643–2661. [Google Scholar] [CrossRef]
  16. Guo, Z.; Caggio, M.; Skalák, Z. Regularity criteria for the Navier-Stokes equations based on one component of velocity. Nonlinear Anal. Real World Appl. 2017, 35, 379–396. [Google Scholar] [CrossRef]
  17. Zhou, Y.; Pokorný, M. On the regularity of the solutions of the Navier-Stokes equations via one velocity component. Nonlinearity 2010, 23, 1097–1107. [Google Scholar] [CrossRef]
  18. Hirata, D. Regularity criterion in terms of BMO–type norm of the pressure gradient for the Navier-Stokes equations on unbounded domains. J. Math. Anal. Appl. 2023, 521, 126949. [Google Scholar] [CrossRef]
  19. Sin, C.; Baranovskii, E.S. Regularity criterion for 3D generalized Newtonian fluids in BMO. J. Differ. Equ. 2023, 377, 859–872. [Google Scholar] [CrossRef]
  20. Neustupa, J.; Penel, P. Regularity of a weak solution to the Navier-Stokes equation in dependence on eigenvalues and eigenvectors of the rate of deformation tensor. In Trends in Partial Differential Equations of Mathematical Physics; Birkhäuser: Basel, Swizerland, 2005; pp. 197–212. [Google Scholar]
  21. Neustupa, J.; Penel, P. On regularity of a weak solution to the Navier-Stokes equation with generalized impermeability boundary conditions. Nonlinear Anal. Theory Methods Appl. 2007, 66, 1753–1769. [Google Scholar] [CrossRef]
  22. Neustupa, J.; Penel, P. On regularity of a weak solution to the Navier-Stokes equations with the generalized Navier Slip boundary conditions. Adv. Math. Phys. 2018, 2018, 4617020. [Google Scholar] [CrossRef]
  23. Miller, E. A regularity criterion for the Navier-Stokes equation involving only the middle eigenvalue of the strain tensor. Arch. Ration. Mech. Anal. 2020, 235, 99–139. [Google Scholar] [CrossRef]
  24. Miller, E. Navier-Stokes regularity criteria in sum spaces. Pure Appl. Anal. 2021, 3, 527–566. [Google Scholar] [CrossRef]
  25. Bahouri, H.; Chemin, J.Y.; Danchin, R. Fourier Analysis and Nonlinear Partial Differential Equations; vol. 343 of Grundlehren der Mathematischen Wissenschaften [Fundamental Principles of Mathematical Sciences]; Springer: Berlin/Heidelberg, Germany, 2011. [Google Scholar]
  26. Ines, B.O.; Sadek, G.; Ragusa, M.A. A new regularity criterion for the 3D incompressible Boussinesq equations in terms of the middle eigenvalue of the strain tensor in the homogeneous Besov spaces with negative indices. Evol. Equ. Control Theory 2023, 12, 1688–1701. [Google Scholar] [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Du, J.; Wu, F. Scaling-Invariant Serrin Criterion via One Row of the Strain Tensor for the Navier–Stokes Equations. Mathematics 2024, 12, 3063. https://doi.org/10.3390/math12193063

AMA Style

Du J, Wu F. Scaling-Invariant Serrin Criterion via One Row of the Strain Tensor for the Navier–Stokes Equations. Mathematics. 2024; 12(19):3063. https://doi.org/10.3390/math12193063

Chicago/Turabian Style

Du, Juan, and Fan Wu. 2024. "Scaling-Invariant Serrin Criterion via One Row of the Strain Tensor for the Navier–Stokes Equations" Mathematics 12, no. 19: 3063. https://doi.org/10.3390/math12193063

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.
Back to TopTop