Startseite Fujita-type theorems for a quasilinear parabolic differential inequality with weighted nonlocal source term
Artikel Open Access

Fujita-type theorems for a quasilinear parabolic differential inequality with weighted nonlocal source term

  • Yuepeng Li und Zhong Bo Fang EMAIL logo
Veröffentlicht/Copyright: 16. März 2023

Abstract

This work is concerned with the nonexistence of nontrivial nonnegative weak solutions for a quasilinear parabolic differential inequality with weighted nonlocal source term in the whole space, which involves weighted polytropic filtration operator or generalized mean curvature operator. We establish the new critical Fujita exponents containing the first and second types. The key ingredient of the technique in proof is the test function method developed by Mitidieri and Pohozaev. No use of comparison and maximum principles or assumptions on symmetry or behavior at infinity of the solutions are required.

MSC 2010: 35K59; 35R45; 35B33; 35B53

1 Introduction

We consider a quasilinear parabolic differential inequality with weighted nonlocal source

(1.1) u t L A u a ( x ) u q K 1 r u r s , ( x , t ) S ,

where S R N × R + , R + ( 0 , + ) , N 1 , q 1 , K 1 r u r s R N K ( x ) u r ( x , t ) d x s r , r 1 , s > 0 , and the nonnegative initial data u 0 ( x ) L loc 1 ( R N ) . The differential operator L A could be both the weighted polytropic filtration operator and the generalized mean curvature operator, which are related to filtration theory and differential geometry etc., see [24,26]. The positive weights a ( x ) and K ( x ) are measurable functions and satisfy

(1.2) a ( x ) x α , K ( x ) x β , x R N \ { 0 } , α , β R ,

which may be singular or degenerate at the origin. Note that the differential inequality (1.1) is called nonlocal differential inequality in the sense that it is not defined pointwise. Meanwhile, norm type nonlocal terms appear in population dynamics and theory of biological populations, see [7,22].

We are interested in finding sufficient conditions for the nonexistence of nonnegative nontrivial weak solutions for (1.1), especially the results related to critical Fujita exponents. The pioneering article in this subject was by Fujita [6], where he considered the Cauchy problem for semilinear parabolic equation u t Δ u = u q and obtained the critical exponent q F = 1 + 2 / N , on the existence versus nonexistence of nonnegative nontrivial solutions. Precisely, nonexistence of solutions, that is blow-up, holds when 1 < q < 1 + 2 / N , while blow-up can occur when q > 1 + 2 / N depending on the size of u 0 . Since then, there have been a number of extensions of Fujita results in many directions. Note that, Hayakawa [10] and Weissler [25] proved that the critical case q = q F belongs to the blow-up case. In fact, for the Cauchy problem of parabolic equations with pure power like nonlinearity, there have been extensive literature and results on the critical (Fujita) exponents for the existence and nonexistence of the solutions, see monographs [24,26] and survey papers [3,8,14].

Motivated by previous results of Mitidieri and Pohozaev in [17,18] and Kartsatos and Kurta in [12], we first study the Fujita-type nonexistence of nonnegative nontrivial weak solutions for a wide class of nonlocal quasilinear parabolic differential inequalities in the framework of the test function. No use of comparison and maximum principles or assumptions on symmetry or behavior at infinity of the solutions are required.

We assume throughout this article that L A is defined as follows:

L A u div [ A ( x , u , u ) ] ,

where A ( x , z , ξ ) is a usual Carathéodory function in R N × R + × R N satisfying weak ellipticity, namely,

(1.3) A ( x , z , ξ ) ξ 0 ,

(1.4) A ( x , z , 0 ) = 0 ,

for all ( x , z , ξ ) R N × R + × R N , and the large radius conditions, namely, for almost all x R N with x R 0 > 0 and z R + , ξ R N , there exists an exponent p > 1 such that

(1.5) A ( x , z , ξ ) p K 0 x σ z m [ A ( x , z , ξ ) ξ ] p 1 ,

where K 0 is a positive constant and σ , m R .

For the case m = 0 , condition (1.5) is assumed for all z R 0 + [ 0 , + ) .

The general class of weakly coercive operators satisfying the structural condition (1.5) was introduced by Farina and Serrin in [4,5] for studying the ellipticproblem, while the subcase of (1.5) with σ = m = 0 appears in [2], which is known as p -weak coercivity.

Operators satisfying the aforementioned assumptions can be given by

L A u = div ( x σ u m u p 2 u ) , p > 1 ,

or

L A u = div ( x σ u m u / 1 + u 2 ) .

Thus, our analysis includes weighted polytropic filtration operator and generalized mean curvature operator. In particular, the aforementioned generalized mean curvature operator satisfies (1.5) for all 1 < p 2 .

We point out that the presence of various components such as the singular coefficient in the operator on the left-hand side of (1.1) and singular coefficient, weighted norm type nonlocal term, power like source, etc. on the right-hand side makes the study of Fujita-type nonexistence results very delicate. In particular, the derivation of a priori estimates is much more complicated because of the fact that several factors appear.

The main results are stated as follows.

Theorem 1.1

Let N 1 , q 1 , r 1 , s > 0 , and operator L A satisfy (1.3)–(1.5) with σ R , m = 0 , p > max 1 , 2 N + σ 1 + N , and p σ < N . Assume (1.2) with

(1.6) α + s r β < p σ + min p 2 + s r N , s r ( p 2 ) N .

  1. If

    (1.7) max 1 , 1 + s r ( p 1 ) q + s < q F , 1 p 1 + s r + p σ α s r β N ,

    then (1.1) does not admit nontrivial nonnegative weak solutions in

    S d 0 u ( x , t ) W loc 1 , p ( S ) a ( x ) u q d K 1 r u r s , x σ u p u 1 d L loc 1 ( S )

    for 0 < d < min { 1 , p 1 } small enough.

  2. If both q and r are not equal to 1, and

    max 1 , 1 + s r ( p 1 ) < q + s = q F , 1 ,

    then (1.1) does not admit nontrivial nonnegative weak solutions in S d 0 for 0 < d < min 1 , p 1 , r ( q + s ) ( r + s ) ( p 1 ) ( r + s ) ( p 1 ) small enough.

Theorem 1.2

Let N 1 , q 1 , r 1 , s > 0 , and operator L A satisfy (1.3)–(1.5) with σ R , m 0 , p > max 1 , 1 m , ( 2 m ) N + σ 1 + N and p σ < N . Assume (1.2) with

(1.8) α + s r β < s r N + p σ + min ( p + m 2 ) N , s r ( p + m 1 ) N .

  1. If

    (1.9) max 1 , 1 + s r ( p + m 1 ) q + s < q F , 2 p + m 1 + s r + p σ α s r β N ,

    then (1.1) does not admit nontrivial positive weak solutions in

    S d { u ( x , t ) W loc 1 , p ( S ) a ( x ) u q d K 1 r u r s , x σ u p u m 1 d L loc 1 ( S ) }

    for 0 < d < min { 1 , p + m 1 } small enough.

  2. If both q and r are not equal to 1, and

    max 1 , 1 + s r ( p + m 1 ) < q + s = q F , 2 ,

    then (1.1) does not admit nontrivial positive weak solutions in S d for 0 < d < min 1 , p + m 1 , r ( q + s ) ( r + s ) ( p + m 1 ) ( r + s ) ( p 1 ) small enough.

Remark 1.1

Conditions (1.6) and (1.8) force that (1.7) and (1.9) are not empty intervals, respectively.

As a consequence of our main Theorems 1.1 and 1.2, by using the polytropic filtration type operator in problem (1.1) and rewriting it as

(1.10) u t div ( u m u p 2 u ) a ( x ) u q K 1 r u r s , ( x , t ) S ,

where N 1 , m R , p > 1 , we obtain the following corollary.

Corollary 1.1

Let N 1 , q 1 , r 1 , s > 0 , m 0 (or m = 0 ), and max 1 , 1 m , ( 2 m ) N 1 + N < p < N . Assume (1.2) with

α + s r β < s r N + p + min ( p + m 2 ) N , s r ( p + m 1 ) N .

  1. If

    max 1 , 1 + s r ( p + m 1 ) q + s < q F , 3 p + m 1 + s r + p α s r β N ,

    then (1.10) has no nontrivial positive (or nonnegative) weak solution in

    u ( x , t ) W l o c 1 , p ( S ) a ( x ) u q d K 1 r u r s , u p u 1 d L l o c 1 ( S )

    for 0 < d < min { 1 , p + m 1 } small enough;

  2. If both q and r are not equal to 1, and

    max 1 , 1 + s r ( p + m 1 ) < q + s = q F , 3 ,

    then (1.10) has no nontrivial positive (or nonnegative) weak solution in

    u ( x , t ) W l o c 1 , p ( S ) a ( x ) u q d K 1 r u r s , u p u 1 d L l o c 1 ( S ) ,

    for 0 < d < min 1 , p + m 1 , r ( q + s ) ( r + s ) ( p + m 1 ) ( r + s ) ( p 1 ) small enough.

Corollary 1.1 implies the critical Fujita exponent for the power of the source at the right-hand side of inequality (1.10) is p + m 1 + s r + p α s r β N . When s = 0 , it covers the existing critical Fujita exponents classified by the parameter q for diffusion equations [6,10,14,25] and inequalities [11,12,17] with multipower-like local source terms. When s 0 , taking α = 0 and considering the quasilinear parabolic equations with nonlocal source terms result in Corollary 1.1 covers the critical Fujita exponents classified by the parameter s given in the recent literature [8,2830], and is in full consistency with that given in [1] classified by q .

The analogous of Corollary 1.1 holds for the mean curvature case, which can be considered morally for the case p = 2 of the aforementioned corollary. Let

L A u = u m u 1 + u 2 ,

and rewrite (1.1) as

(1.11) u t div u m u 1 + u 2 a ( x ) u q K 1 r u r s , ( x , t ) S ,

where N 1 , m R . Then we obtain the following new corollary for the mean curvature problem (1.11).

Corollary 1.2

Let N > 2 , q 1 , r 1 , s > 0 , m > 2 N , and m 0 (or m = 0 ). Assume (1.2) with

α + s r β < s r N + 2 + min m N , s r ( m + 1 ) N .

  1. If

    max 1 , 1 + s r ( m + 1 ) q + s < q F , 4 m + 1 + s r + 2 α s r β N ,

    then (1.11) has no nontrivial positive (or nonnegative) weak solution in

    u ( x , t ) W l o c 1 , 2 ( S ) a ( x ) u q d K 1 r u r s , u 2 u m 1 d L l o c 1 ( S )

    for 0 < d < min { 1 , m + 1 } sufficiently small.

  2. If both q and r are not equal to 1 and

    1 + s r < q + s = q F , 4 ,

    then (1.11) has no nontrivial positive (or nonnegative) solution in

    u ( x , t ) W l o c 1 , 2 ( S ) a ( x ) u q d K 1 r u r s , u 2 u m 1 d L l o c 1 ( S )

    for 0 < d < min { 1 , m + 1 , r ( q + s ) ( r + s ) ( m + 1 ) r + s } sufficiently small.

From Corollary 1.2, the critical Fujita exponent for the power q + s of nonlocal source in (1.11) is m + 1 + s r + 2 α s r β N . Taking s = 0 and m = α = 0 , we can deduce the existing critical Fujita exponent q F = 1 + 2 N for local mean curvature equation [14] and the corresponding local inequality [17] when N > 2 .

In conclusion, comparison of the main results on the first critical Fujita exponents obtained earlier with the existing results is presented in Table 1.

Table 1

First critical Fujita exponents

Operators in (1.1) First critical Fujita exponents for q + s Comparison with existing results
L A u is the generalized weakly coercive operator p + m 1 + s r + p σ α s r β N . See Theorems 1.1 and 1.2 New
Polytropic filtration operator L A u = div ( u m u p 2 u ) p + m 1 + s r + p α s r β N . See Corollary 1.1 s = 0 , obtained in [6,1012,14,17,25]; s 0 , α = 0 , obtained in [1,8,28,29].
Mean curvature operator L A u = u m u 1 + u 2 m + 1 + s r + 2 α s r β N . See Corollary 1.2 s = m = α = 0 , obtained in [14,17]

On the other hand, it is obvious from [10,25] that the classical critical Fujita exponent q F is not optimal for the Cauchy problem of parabolic equation u t u = u q . Thus, to identify the global and nonglobal solutions in the coexistence region, q > q F becomes really interesting and challenging. We refer the reader to [13] for the pioneering work in this subject. Here, for the nonlocal inequality (1.1), we derive the new second critical exponents corresponding to all the aforementioned Fujita-type results by virtue of the slow decay behavior of the initial data at spatial infinity.

Theorem 1.3

Let N 1 , q 1 , r 1 , s > 0 , and operator L A satisfy (1.3)–(1.5), where σ R , m = 0 , p > 1 and q + s 1 + s r ( p 1 ) . Assume (1.2) with

α + s r β < s r N + p σ .

If there exists μ > 0 such that

(1.12) liminf x + u 0 ( x ) x μ > 0 ,

and

(1.13) μ < μ 1 s r N α s r β + p σ q + s ( p 1 ) ,

then (1.1) does not admit nonnegative nontrivial solutions belonging to the class S d 0 for 0 < d < min { 1 , p 1 } sufficiently small.

Theorem 1.4

Let N 1 , q 1 , r 1 , s > 0 , and operator L A satisfy (1.3)–(1.5), where σ R , m 0 , p > max { 1 , 1 m } and q + s 1 + s r ( p + m 1 ) . Assume (1.2) with

α + s r β < s r N + p σ .

If there exists μ > 0 such that (1.12) holds and

(1.14) μ < μ 2 s r N α s r β + p σ q + s ( p + m 1 ) ,

then (1.1) does not admit nontrivial positive solutions belonging to the class S d for 0 < d < min { 1 , p + m 1 } sufficiently small.

Remark 1.2

It can be compared between the first critical Fujita exponents given in Theorems 1.1 and 1.2 and the second critical exponents given in Theorems 1.3 and 1.4 under appropriate conditions. Actually, if

1 + s r ( p 1 ) < q + s < p 1 + N s r N + p σ α s r β N ( p 1 ) + s r N + p σ α s r β ,

or

1 + s r ( p + m 1 ) < q + s < p + m 1 + N s r N + p σ α s r β N ( p + m 1 ) + s r N + p σ α s r β ,

we have q F , 1 < μ 1 or q F , 2 < μ 2 .

In addition, we give the second critical Fujita exponent results corresponding with Corollaries 1.1 and 1.2.

Corollary 1.3

Let N 1 , q 1 , r 1 , s > 0 , m 0 (or m = 0 ), p > max { 1 , 1 m } , and q + s 1 + s r ( p + m 1 ) . Assume (1.2) with

α + s r β < s r N + p .

If there exists μ > 0 such that (1.12) holds and

μ < μ 3 s r N α s r β + p q + s ( p + m 1 ) ,

then (1.10) has no nontrivial positive (or nonnegative) weak solution belonging to the class

u ( x , t ) W loc 1 , p ( S ) a ( x ) u q d K 1 r u r s , u p u 1 d L loc 1 ( S )

for 0 < d < min { 1 , p + m 1 } sufficiently small.

We obtain a new second critical exponent s r N α s r β + p q + s ( p + m 1 ) for nonlocal inequality (1.10) in Corollary 1.3. By taking s = 0 , we can derive the second critical exponents for diffusion equations with multipower like local source in [9,13,1921,27]. If s 0 and considering nonlocal problems, we can deduce the second critical exponents for quasilinear parabolic equations with nonlocal source given in [15,16,2830].

Corollary 1.4

Let N 1 , q 1 , r 1 , s > 0 , m > 1 , m 0 (or m = 0 ), and q + s 1 + s r ( m + 1 ) . Assume (1.2) with

α + s r β < s r N + 2 .

If there exists μ > 0 such that (1.12) holds and

μ < μ 4 s r N α s r β + 2 q + s ( m + 1 ) ,

then (1.11) has no nontrivial positive (or nonnegative) weak solution belonging to the class

u ( x , t ) W loc 1 , 2 ( S ) a ( x ) u q d K 1 r u r s , u p u 1 d L loc 1 ( S )

for 0 < d < min { 1 , m + 1 } sufficiently small.

Remark 1.3

Similar to Remark 1.2, the first critical exponents in Corollaries 1.1 and 1.2 can be compared with the second ones in Corollaries 1.3 and 1.4 under appropriate conditions. By taking σ = 0 in Remark 1.2, we have q F , 3 < μ 3 , furthermore, q F , 4 < μ 4 when p = 2 .

Indeed, the second critical Fujita exponent s r N α s r β + 2 q + s ( m + 1 ) is new for the mean curvature type nonlocal inequality (1.11).

To sum up, we compare the main results on the second critical Fujita exponents with the existing results and obtain Table 2.

Table 2

Second critical Fujita exponents

Operators in (1.1) Second critical Fujita exponents for q + s Comparison with existing results
L A u is the generalized weakly coercive operator s r N α s r β + p σ q + s ( p + m 1 ) . See Theorems 1.3–1.4 New
Polytropic filtration operator L A u = div ( u m u p 2 u ) s r N α s r β + p q + s ( p + m 1 ) . See Corollary 1.3 s = 0 , obtained in [9,13,1921,27]; s 0 , α = 0 , obtained in [15,16,2830].
Mean curvature operator
L A u = u m u 1 + u 2 s r N α s r β + 2 q + s ( m + 1 ) . See Corollary 1.4 New

This article is organized as follows. In Section 2, we introduce the preparatory knowledge. In Section 3, we present several technical lemmas and obtain fine a priori estimates. Finally, we present the detailed proofs of main results in Section 4.

2 Preliminaries

In this section, we introduce some notations, definitions, and the careful selections of test functions.

Throughout this article, we denote C various constants independent of u , which may be different from line to line.

First, we define the weak solution of (1.1).

Definition 2.1

For a weak solution of (1.1), we mean a nonnegative function u ( x , t ) defined on S , given by those functions u W loc 1 , p ( S ) with

  1. A ( x , u , u ) [ L loc p ( S ) ] N , here 1 p + 1 p = 1 ;

  2. K ( x ) u r ( x , t ) L 1 ( R N ) ;

  3. a ( x ) u q K 1 r u r s L loc 1 ( S ) ;

such that for any nonnegative test function φ C 0 1 ( S ) , we have

(2.1) S a ( x ) u q K 1 r u r s φ d x d t S u t φ d x d t + S A ( x , u , u ) φ d x d t .

Obviously, when the equal sign holds in (1.1), we can consider test functions φ C 0 1 not necessarily nonnegative.

When necessary, we make use of the following weak formulation of (1.1)

(2.2) S a ( x ) u q K 1 r u r s φ d x d t R N u 0 ( x ) φ ( x , 0 ) d x S u φ t d x d t + S A ( x , u , u ) φ d x d t ,

for any nonnegative test function φ C 0 1 ( S ) .

Next, we construct the test function in the form of separated variables with parameters carefully. Let B R ( 0 ) be the ball of R N , centered at the origin and with radius R > 0 . We define a function ξ 0 ( s ) C 1 ( [ 0 , + ) ) such that

0 ξ 0 ( s ) 1 , s 0 ; ξ 0 ( s ) = 1 , 0 s 1 ; ξ 0 ( s ) = 0 , s 2 .

Moreover,

ξ 0 ( s ) C , s 0 ,

where C > 0 is a constant.

For the space variable, we take

χ ( x ) ξ 0 x R ,

thus, χ ( x ) C 0 1 ( R N ) and

χ ( x ) = 1 , x B R ( 0 ) ; χ ( x ) = 0 , x R N \ B 2 R ( 0 ) ; 0 χ ( x ) 1 , x R N ; χ ( x ) C R , x R N ,

where C > 0 is the constant given in (2.3).

For the time variable, consider

η ( t ) ξ 0 t R γ ,

with γ > 0 to be chosen later.

Now, we define, for R > R 0 > 0 , a nonnegative cutoff function in S , given by

ζ ( x , t ) χ ( x ) η ( t ) .

Obviously ζ ( x , t ) C 0 1 ( S ) . Meanwhile, we give the notations for the supports by

P ( 2 R ) supp ζ = B 2 R ( 0 ) × [ 0 , 2 R γ ] , P 1 ( R ) supp ζ = { x R x 2 R } × [ 0 , 2 R γ ] , P 2 ( R ) supp ζ t = B 2 R ( 0 ) × [ R γ , 2 R γ ] .

Then, we give the test functions in two cases.

Case 1. For m = 0 , we take

(2.3) φ ( x , t ) = u ˜ ε d ζ k ,

where

u ˜ ε ( x , t ) τ + R N ξ ε ( x y , t ) u ( y , t ) d y , u τ ( x , t ) τ + u ( x , t ) .

Here, u ( x , t ) is a nonnegative weak solution of the differential inequality (1.1) on S , ( ξ ε ) ε > 0 is a standard family of mollifiers, and

ε > 0 , τ > 0 , d > 0 , k > 0 .

Clearly for all ( x , t ) S , u ˜ ε , u τ τ > 0 , u τ ( x , t ) is increasing with respect to variable τ , and φ ( x , t ) C 0 1 ( S ) .

Case 2. For m 0 , there is no need to introduce the parameter τ given in case 1, since we deal with positive solutions, namely,

φ ( x , t ) = u ε d ζ k ,

with

u ε ( x , t ) R N ξ ε ( x y , t ) u ( y , t ) d y .

Similarly, u ε > 0 and φ ( x , t ) C 0 1 ( S ) for all ( x , t ) S .

3 A priori estimates

In this section, we present a priori estimates of the solution needed in the proofs of the main results. To this end, we will give several technical preliminaries in Subsection 3.1 and derive the corresponding a priori estimates in two cases in Subsection 3.2.

3.1 Technical lemmas

In this subsection, we prepare some technical preparatory lemmas in order to obtain a priori estimates.

Lemma 3.1

Let q 1 , r 1 and s > 0 . If u ( x , t ) is a nonnegative (or positive) weak solution to differential inequality (1.1) in S with u ( x , t ) S d 0 (or u ( x , t ) S d ), then for 0 < d < 1 and k > 0 , the following inequality holds:

(3.1) P 2 ( R ) u 1 d ζ t ζ k 1 d x d t ε 1 P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + C R γ 2 R γ x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x σ 1 σ 2 d t ,

where

σ 1 = q + s d 1 d > 1 , σ 1 = σ 1 σ 1 1 = q + s d q + s 1 , σ 2 = r σ 1 r σ 1 r s > 1 ,

and ε 1 , C > 0 are constants.

Proof

Rewrite the integrand in the integral on the left-hand side of (3.1) as

u 1 d ζ t ζ k 1 = a ( x ) 1 σ 1 u q d σ 1 ζ k σ 1 K ( x ) s r σ 1 u s σ 1 a ( x ) 1 σ 1 K ( x ) s r σ 1 u 1 d q + s d σ 1 ζ t ζ k σ 1 1 .

We take σ 1 = q + s d 1 d such that the exponent of u in the third term on the right-hand side is 0, and σ 2 satisfying

1 σ 1 + s r σ 1 + 1 σ 2 = 1 .

It follows that

r σ 1 s = r ( q + s d ) s ( 1 d ) , σ 2 = r ( q + s d ) r ( q + s d ) ( r + s ) ( 1 d ) .

We claim σ 1 , r σ 1 s and σ 2 > 1 . In fact, by calculating directly, we obtain

σ 1 1 = q + s 1 1 d , r σ 1 s 1 = r ( q + s d ) s ( 1 d ) s ( 1 d ) = s d + r ( q 1 ) + s ( r 1 ) + r ( 1 d ) s ( 1 d ) , σ 2 1 = ( r + s ) ( 1 d ) r ( q + s d ) ( r + s ) ( 1 d ) = ( r + s ) ( 1 d ) s d + r ( q 1 ) + s ( r 1 ) .

Due to q 1 , r 1 , s > 0 and 0 < d < 1 , we arrive at σ 1 > 1 , r σ 1 s > 1 , and σ 2 > 1 obviously.

Thus, by Hölder inequality with exponents σ 1 , r σ 1 s and σ 2 , we obtain

x 2 R u 1 d ζ t ζ k 1 d x x 2 R a ( x ) u q d ζ k d x 1 σ 1 x 2 R K ( x ) u r ( x , t ) d x s r σ 1 × x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x 1 σ 2 x 2 R a ( x ) u q d K 1 r u r s ζ k d x 1 σ 1 x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x 1 σ 2 .

By substituting it into the left-hand side of (3.1), we have

P 2 ( R ) u 1 d ζ t ζ k 1 d x d t R γ 2 R γ x 2 R a ( x ) u q d K 1 r u r s ζ k d x 1 σ 1 × x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x 1 σ 2 d t .

Then, by Young’s inequality with exponents σ 1 , σ 1 , and parameter ε 1 , we derive

P 2 ( R ) u 1 d ζ t ζ k 1 d x d t ε 1 P 2 ( R ) a ( x ) u q d K 1 r u r s ζ k d x d t + C R γ 2 R γ x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x σ 1 σ 2 d t ε 1 P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + C R γ 2 R γ x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x σ 1 σ 2 d t ,

which completes the proof of Lemma 3.1.□

Lemma 3.2

Let q 1 , r 1 , s > 0 , σ , m R , p > max { 1 , 1 m } , and

q + s 1 + s r ( p + m 1 ) .

If u ( x , t ) is a nonnegative (or positive) weak solution of differential inequality (1.1) on S with u ( x , t ) S d 0 (or u ( x , t ) S d ), then for all 0 < d < p + m 1 and k > 0 , the following inequality holds:

(3.2) P 1 ( R ) u p + m 1 d x σ ζ p ζ k p d x d t ε 2 P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + C 0 2 R γ R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x σ 3 σ 4 d t ,

where

σ 3 = q + s d p + m 1 d > 1 , σ 3 = σ 3 σ 3 1 = q + s d q + s ( p + m 1 ) , σ 4 = r σ 3 r σ 3 r s > 1 ,

and ε 2 , C > 0 are constants.

Proof

The integrand in the integral on the left-hand side of (3.2) can be rewritten as follows:

u p + m 1 d x σ ζ p ζ k p = a ( x ) 1 σ 3 u q d σ 3 ζ k σ 3 K ( x ) s r σ 3 u s σ 3 a ( x ) 1 σ 3 K ( x ) s r σ 3 u p + m 1 d q + s d σ 3 x σ ζ p ζ k σ 3 p .

Let σ 3 = q + s d p + m 1 d such that the exponent of u in the third term on the right-hand side is 0, and take σ 4 satisfying

1 σ 3 + s r σ 3 + 1 σ 4 = 1 .

It implies that

r σ 3 s = r ( q + s d ) s ( p + m 1 d ) , σ 4 = r ( q + s d ) r [ q + s ( p + m 1 ) ] s ( p + m 1 d ) .

We claim σ 3 , r σ 3 s , and σ 4 > 1 . Indeed, by the fact r 1 , s > 0 , p > 1 m , and 0 < d < p + m 1 , we compute directly and obtain that

σ 3 1 = q + s ( p + m 1 ) p + m 1 d > q + s 1 + s r ( p + m 1 ) p + m 1 d , r σ 3 s 1 = r ( q + s d ) s ( p + m 1 d ) s ( p + m 1 d ) > r ( q + s d ) ( r + s ) ( p + m 1 d ) s ( p + m 1 d ) = s d + r q + s 1 + s r ( p + m 1 ) s ( p + m 1 d ) , σ 4 1 = ( r + s ) ( p + m 1 d ) r [ q + s ( p + m 1 ) ] s ( p + m 1 d ) = ( r + s ) ( p + m 1 d ) s d + r q + s 1 + s r ( p + m 1 ) .

It follows from q + s 1 + s r ( p + m 1 ) that, σ 3 > 1 , r σ 3 s > 1 and σ 4 > 1 obviously.

Hence, by using Hölder inequality with exponents σ 3 , r σ 3 s , and σ 4 we derive

R x 2 R u p + m 1 d x σ ζ p ζ k p d x R x 2 R a ( x ) u q d ζ k d x 1 σ 3 R x 2 R K ( x ) u r ( x , t ) d x s r σ 3 R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x 1 σ 4 R x 2 R a ( x ) u q d K 1 r u r s ζ k d x 1 σ 3 R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x 1 σ 4 .

By inserting the aforementioned formula into the left-hand side of (3.2), we have

P 1 ( R ) u p + m 1 d x σ ζ p ζ k p d x d t 0 2 R γ R x 2 R a ( x ) u q d K 1 r u r s ζ k d x 1 σ 3 R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x 1 σ 4 d t .

Then, by applying Young’s inequality with exponents σ 3 , σ 3 , and parameter ε 2 to the right-hand side of the last inequality, we obtain

P 1 ( R ) u p + m 1 d x σ ζ p ζ k p d x d t ε 2 P 1 ( R ) a ( x ) u q d K 1 r u r s ζ k d x d t + C 0 2 R γ R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x σ 3 σ 4 d t ε 2 P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + C 0 2 R γ R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x σ 3 σ 4 d t .

Lemma 3.2 is proved.□

Lemma 3.3

Let q 1 , r 1 , s > 0 , σ , m R , p > max { 1 , 1 m } , and

q + s > 1 + s r ( p + m 1 ) .

If u ( x , t ) is a nonnegative (or positive) weak solution of differential inequality (1.1) and u ( x , t ) S d 0 (or u ( x , t ) S d ), then for 0 < d < r ( q + s ) ( r + s ) ( p + m 1 ) ( r + s ) ( p 1 ) and k > 0 , the following inequality holds:

(3.3) P 1 ( R ) u ( p 1 ) ( 1 + d ) + m x σ ζ p ζ k p d x d t P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 5 0 2 R γ R < x 2 R a ( x ) σ 6 σ 5 K ( x ) s σ 6 r σ 5 x σ σ 6 ζ p σ 6 ζ k σ 6 σ 5 p σ 6 d x σ 5 σ 6 d t 1 σ 5 ,

where

σ 5 = q + s m + ( p 1 ) ( 1 + d ) > 1 , σ 5 = σ 5 σ 5 1 = q + s q + s [ m + ( p 1 ) ( 1 + d ) ] , σ 6 = r σ 5 r σ 5 r s > 1 .

Proof

We rewrite the integrand in the integral on the left-hand side of (3.3) as follows:

u ( p 1 ) ( 1 + d ) + m x σ ζ p ζ k p = a ( x ) 1 σ 5 u q σ 5 ζ k σ 5 K ( x ) s r σ 5 u s σ 5 a ( x ) 1 σ 5 K ( x ) s r σ 5 u ( p 1 ) ( 1 + d ) + m q + s σ 5 x σ ζ p ζ k σ 5 p .

Choose σ 5 = q + s m + ( p 1 ) ( 1 + d ) such that the exponent of u in the third term on the right-hand side of the aforementioned formula is 0, and an appropriate σ 6 satisfying

1 σ 5 + s r σ 5 + 1 σ 6 = 1 .

It follows that

r σ 5 s = r ( q + s ) s [ m + ( p 1 ) ( 1 + d ) ] , σ 6 = r ( q + s ) r ( q + s ) ( r + s ) [ m + ( p 1 ) ( 1 + d ) ] .

We claim σ 5 , r σ 5 s and σ 6 > 1 . Indeed, with r 1 , s > 0 and p > 1 m , simple calculations show that

σ 5 1 = q + s [ m + ( p 1 ) ( 1 + d ) ] m + ( p 1 ) ( 1 + d ) = q + s ( p + m 1 ) ( p 1 ) d m + ( p 1 ) ( 1 + d ) , r σ 5 s 1 = r ( q + s ) s [ m + ( p 1 ) ( 1 + d ) ] s [ m + ( p 1 ) ( 1 + d ) ] > r ( q + s ) ( r + s ) [ m + ( p 1 ) ( 1 + d ) ] s [ m + ( p 1 ) ( 1 + d ) ] = r ( q + s ) ( r + s ) ( p + m 1 ) d ( r + s ) ( p 1 ) s [ m + ( p 1 ) ( 1 + d ) ] , σ 6 1 = ( r + s ) [ m + ( p 1 ) ( 1 + d ) ] r ( q + s ) ( r + s ) [ m + ( p 1 ) ( 1 + d ) ] = ( r + s ) [ m + ( p 1 ) ( 1 + d ) ] r ( q + s ) ( r + s ) ( p + m 1 ) d ( r + s ) ( p 1 ) .

Then, r σ 5 s > 1 and σ 6 > 1 hold since 0 < d < r ( q + s ) ( r + s ) ( p + m 1 ) ( r + s ) ( p 1 ) . Moreover, for all 0 < d < r ( q + s ) ( r + s ) ( p + m 1 ) ( r + s ) ( p 1 ) , we have

d < r ( q + s ) ( r + s ) ( p + m 1 ) ( r + s ) ( p 1 ) = q + s ( p + m 1 ) p 1 s ( q + s ) ( r + s ) ( p 1 ) < q + s ( p + m 1 ) p 1 ,

which gives σ 5 > 1 .

So we can use Hölder inequality with exponents σ 5 , r σ 5 s , and σ 6 to obtain

R x 2 R u ( p 1 ) ( 1 + d ) + m x σ ζ p ζ k p d x R x 2 R a ( x ) u q ζ k d x 1 σ 5 R x 2 R K ( x ) u r ( x , t ) d x s r σ 5 R x 2 R a ( x ) σ 6 σ 5 K ( x ) s σ 6 r σ 5 x σ σ 6 ζ p σ 6 ζ k σ 6 σ 5 p σ 6 d x 1 σ 6 R x 2 R a ( x ) u q K 1 r u r s ζ k d x 1 σ 5 R x 2 R a ( x ) σ 6 σ 5 K ( x ) s σ 6 r σ 5 x σ σ 6 ζ p σ 6 ζ k σ 6 σ 5 p σ 6 d x 1 σ 6 .

By substituting it into the left-hand side of (3.3), we have

P 1 ( R ) u ( p 1 ) ( 1 + d ) + m x σ ζ p ζ k p d x d t 0 2 R γ R x 2 R a ( x ) u q K 1 r u r s ζ k d 1 σ 5 R x 2 R a ( x ) σ 6 σ 5 K ( x ) s σ 6 r σ 5 x σ σ 6 ζ p σ 6 ζ k σ 6 σ 5 p σ 6 d x 1 σ 6 d t .

Then, by applying Hölder inequality again with exponents σ 5 and σ 5 to the right-hand side of the aforementioned formula, we derive

P 1 ( R ) u ( p 1 ) ( 1 + d ) + m x σ ζ p ζ k p d x d t P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 5 0 2 R γ R x 2 R a ( x ) σ 6 σ 5 K ( x ) s σ 6 r σ 5 x σ σ 6 ζ p σ 6 ζ k σ 6 σ 5 p σ 6 d x σ 5 σ 6 d t 1 σ 5 ,

which yields (3.3).□

Lemma 3.4

Let q 1 , r 1 , s > 0 , and both q, r are not equal to 1. If u ( x , t ) is a nonnegative (or positive) weak solution of differential inequality (1.1) and u ( x , t ) S d 0 (or u ( x , t ) S d ), then for 0 < d < 1 and k > 0 , the following inequality holds:

(3.4) P 2 ( R ) u ζ t ζ k 1 d x d t P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 × R γ 2 R γ x 2 R a ( x ) σ 8 σ 7 K ( x ) s σ 8 r σ 7 ζ t σ 8 ζ k σ 8 σ 7 σ 8 d x σ 7 σ 8 d t 1 σ 7 ,

where

σ 7 = q + s > 1 , σ 7 = σ 7 σ 7 1 = q + s q + s 1 , σ 8 = r σ 7 r σ 7 r s > 1 .

Proof

Rewrite the integrand in the integral on the left-hand side of (3.4) as follows:

u ζ t ζ k 1 = a ( x ) 1 σ 7 u q σ 7 ζ k σ 7 K ( x ) s r σ 7 u s σ 7 a ( x ) 1 σ 7 K ( x ) s r σ 7 u 1 q + s σ 7 ζ t ζ k σ 7 1 .

We take σ 7 = q + s such that the exponent of u in the third term on the right-hand side of the aforementioned formula is 0, and σ 8 satisfying

1 σ 7 + s r σ 7 + 1 σ 8 = 1 .

By simple calculations, we have

r σ 7 s = r ( q + s ) s , σ 8 = r ( q + s ) r ( q + s ) r s .

We claim σ 7 , r σ 7 s , and σ 8 > 1 . In fact, σ 7 > 1 holds clearly due to q 1 , r 1 , and s > 0 . Meanwhile, we obtain by a simple calculation that

r σ 7 s 1 = r ( q + s ) s s = s ( r 1 ) + r q s , σ 8 1 = r + s r ( q + s ) r s = r + s r ( q 1 ) + s ( r 1 ) .

Since q 1 , r 1 , s > 0 , and both q and r do not equal to 1, we obtain r σ 7 s > 1 and σ 8 > 1 obviously.

Thus, by Hölder inequality with exponents σ 7 , r σ 7 s and σ 8 , we obtain

x 2 R u ζ t ζ k 1 d x x 2 R a ( x ) u q ζ k d x 1 σ 7 x 2 R K ( x ) u r ( x , t ) d x s r σ 7 × x 2 R a ( x ) σ 8 σ 7 K ( x ) s σ 8 r σ 7 ζ t σ 8 ζ k σ 8 σ 7 σ 8 d x 1 σ 8 x 2 R a ( x ) u q K 1 r u r s ζ k d x 1 σ 7 x 2 R a ( x ) σ 8 σ 7 K ( x ) s σ 8 r σ 7 ζ t σ 8 ζ k σ 8 σ 7 σ 8 d x 1 σ 8 .

By substituting it into the left-hand side of (3.4), we have

P 2 ( R ) u ζ t ζ k 1 d x d t 0 R x 2 R a ( x ) u q K 1 r u r s ζ k d x 1 σ 7 x 2 R a ( x ) σ 8 σ 7 K ( x ) s σ 8 r σ 7 ζ t σ 8 ζ k σ 8 σ 7 σ 8 d x 1 σ 8 d t .

Then, by applying again Hölder inequality with exponents σ 7 and σ 7 to the right-hand side of the aforementioned formula, we derive

P 2 ( R ) u ζ t ζ k 1 d x d t P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 R γ 2 R γ x 2 R a ( x ) σ 8 σ 7 K ( x ) s σ 8 r σ 7 ζ t σ 8 ζ k σ 8 σ 7 σ 8 d x σ 7 σ 8 d t 1 σ 7 .

Lemma 3.4 is proved.□

3.2 A priori estimates

In this subsection, by using the preparatory lemmas obtained in Subsection 3.1, we give fine a priori estimates for the two cases m = 0 and m 0 , namely, Propositions 3.1 and 3.3. Also, we obtain the additional a priori estimates, Propositions 3.2 and 3.4, to study the nonexistence of the solutions to problem (1.1) in the critical case. They play a crucial role in the proofs of our main results.

Proposition 3.1

Let N 1 , q 1 , r 1 , s > 0 , and L A satisfy (1.3)–(1.5), where σ R , m = 0 , p > 1 and

q + s 1 + s r ( p 1 ) .

Assume k > max q + s d q + s 1 , p ( q + s d ) q + s ( p 1 ) and 0 < d < min { 1 , p 1 } . If u ( x , t ) S d 0 is a nonnegative weak solution to (1.1), then we have the inequalities

(3.5) P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t C ( R δ 1 + R δ 2 ) ,

(3.6) P 1 ( R ) ( A u ) u d 1 ζ k d x d t C ( R δ 1 + R δ 2 ) ,

(3.7) B 2 R ( 0 ) u 0 1 d ( x ) ζ k ( x , 0 ) d x C ( R δ 1 + R δ 2 ) ,

where

δ 1 = 1 d q + s 1 γ + 1 d q + s 1 α + s ( 1 d ) r ( q + s 1 ) β + r N ( q + s d ) ( r + s ) ( 1 d ) N r ( q + s 1 ) , δ 2 = γ + p 1 d q + s p + 1 α + s ( p 1 d ) r ( q + s p + 1 ) β ( p σ ) q + s d q + s p + 1 + r N ( q + s p + 1 ) s N ( p 1 d ) r ( q + s p + 1 ) .

Proof

Let φ ( x , t ) = u ˜ ε d ζ k , then φ ( x , t ) C 1 0 ( S ) clearly and can be used as a test function in the weak formulation of (1.1), given by (2.1), so that

P ( 2 R ) a ( x ) u q u ˜ ε d K 1 r u r s ζ k d x d t P ( 2 R ) u t u ˜ ε d ζ k d x d t + P ( 2 R ) A ( x , u , u ) ( u ˜ ε d ζ k ) d x d t .

Since u ˜ ε u τ in L l o c 1 ( S ) as ε 0 , by applying Lebesgue dominated convergence theorem [23, p. 26], passing to the limit as ε 0 , we arrive at

(3.8) P ( 2 R ) a ( x ) u q u τ d K 1 r u r s ζ k d x d t P ( 2 R ) u t u τ d ζ k d x d t + P ( 2 R ) A ( x , u , u ) ( u τ d ζ k ) d x d t .

Being u τ = u and ( u τ ) t = u t by definition of u τ , we have

u t u τ d ζ k = 1 1 d u τ 1 d t ζ k ,

and (3.8) gives

P ( 2 R ) a ( x ) u q u τ d K 1 r u r s ζ k d x d t k 1 d P ( 2 R ) u τ 1 d ζ k 1 ζ t d x d t 1 1 d B 2 R ( 0 ) u τ 1 d ( x , 0 ) ζ k ( x , 0 ) d x d P ( 2 R ) [ A ( x , u , u ) u ] u τ 1 d ζ k d x d t + k P ( 2 R ) [ A ( x , u , u ) ζ ] u τ d ζ k 1 d x d t .

Since u τ > 0 and ζ 0 , we have

(3.9) P ( 2 R ) a ( x ) u q u τ d K 1 r u r s ζ k d x d t + 1 1 d B 2 R ( 0 ) u τ 1 d ( x , 0 ) ζ k ( x , 0 ) d x + d P 1 ( R ) [ A ( x , u , u ) u ] u τ 1 d ζ k d x d t k 1 d P ( 2 R ) u τ 1 d ζ k 1 ζ t d x d t + k P 1 ( R ) [ A ( x , u , u ) ζ ] u τ d ζ k 1 d x d t .

From (1.5), we have for all R > R 0 that

k P 1 ( R ) [ A ( x , u , u ) ζ ] u τ d ζ k 1 d x d t k P 1 ( R ) A ( x , u , u ) ζ u τ d ζ k 1 d x d t C P 1 ( R ) x σ p ( A u ) p 1 p u τ d ζ ζ k 1 d x d t .

Further, by Young’s inequality with exponents p , p p 1 and parameter ε 3 , we have

k P 1 ( R ) [ A ( x , u , u ) ζ ] u τ d ζ k 1 d x d t ε 3 P 1 ( R ) ( A u ) u τ d 1 ζ k d x d t + C P 1 ( R ) u τ p 1 d x σ ζ p ζ k p d x d t .

Since ( A u ) u τ d 1 < K 0 x σ u p u 1 d L loc 1 ( S ) when R > R 0 , by combining with (3.9), we obtain

(3.10) P ( 2 R ) a ( x ) u q u τ d K 1 r u r s ζ k d x d t + 1 1 d B 2 R ( 0 ) u τ 1 d ( x , 0 ) ζ k ( x , 0 ) d x + ( d ε 3 ) P 1 ( R ) ( A u ) u τ 1 d ζ k d x d t k 1 d P 2 ( R ) u τ 1 d ζ t ζ k 1 d x d t + C P 1 ( R ) u τ p 1 d x σ ζ p ζ k p d x d t .

Since all the exponents of u τ on the right-hand side are positive, being 0 < d < min { 1 , p 1 } , we let τ 0 and apply Beppo-Levi theorem ([23, p. 21]) to obtain

P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + 1 1 d B 2 R ( 0 ) u 0 1 d ( x , 0 ) ζ k ( x , 0 ) d x + ( d ε 3 ) P 1 ( R ) ( A u ) u 1 d ζ k d x d t k 1 d P 2 ( R ) u 1 d ζ t ζ k 1 d x d t + C P 1 ( R ) u p 1 d x σ ζ p ζ k p d x d t .

Then, we estimate the last two terms by Lemmas 3.1 and 3.2 with m = 0 and derive

P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + 1 1 d B 2 R ( 0 ) u 0 1 d ( x ) ζ k ( x , 0 ) d x + ( d ε 3 ) P 1 ( R ) ( A u ) u d 1 ζ k d x d t ε 1 P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + C R γ 2 R γ x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x σ 1 σ 2 d t + ε 2 P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + C 0 2 R γ R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x σ 3 σ 4 d t ,

which gives, since u ( x , t ) S d 0 ,

(3.11) ( 1 ε 1 ε 2 ) P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + 1 1 d B 2 R ( 0 ) u 0 1 d ( x ) ζ k ( x , 0 ) d x + ( d ε 3 ) P 1 ( R ) ( A u ) u d 1 ζ k d x d t C R γ 2 R γ x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x σ 1 σ 2 d t + C 0 2 R γ R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x σ 3 σ 4 d t .

We next estimate the last two integrals in (3.11). The exponents of ζ in the two integrals are both positive since k > max q + s d q + s 1 , p ( q + s d ) q + s ( p 1 ) , so we have

0 ζ k σ 2 σ 1 σ 2 , ζ k σ 4 σ 3 p σ 4 1 .

By the definition of ζ and (1.2), we arrive at

ζ t C R γ , ζ C R 1 ; a ( x ) x α , K ( x ) x β , x R N \ { 0 } .

Thus, we have for the first one that

R γ 2 R γ x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x σ 1 σ 2 d t C R γ 2 R γ x 2 R x σ 2 σ 1 α x s σ 2 r σ 1 β R γ σ 2 d x σ 1 σ 2 d t .

On the basis of the scale transformation ξ x R , θ t R γ , we obtain

x σ 2 σ 1 α = ( R ξ ) σ 2 σ 1 α = R σ 2 σ 1 α ξ σ 2 σ 1 α , x s σ 2 r σ 1 β = ( R ξ ) s σ 2 r σ 1 β = R s σ 2 r σ 1 β ξ s σ 2 r σ 1 β ,

and

R γ 2 R γ x 2 R x σ 2 σ 1 α x s σ 2 r σ 1 β R γ σ 2 d x σ 1 σ 2 d t C R ( σ 2 σ 1 α + s σ 2 r σ 1 β γ σ 2 + N ) σ 1 σ 2 + γ 1 2 ξ 2 ξ σ 2 σ 1 α ξ s σ 2 r σ 1 β d ξ σ 1 σ 2 d θ C R ( σ 2 σ 1 α + s σ 2 r σ 1 β γ σ 2 + N ) σ 1 σ 2 + γ C R δ 1 .

In a similar way, we estimate the last integral in (3.11)

0 2 R γ R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x σ 3 σ 4 d t C R σ 4 σ 3 α + s σ 4 r σ 3 β p σ 4 + σ σ 4 + N σ 3 σ 4 + γ C R δ 2 .

Choosing ε 1 < 1 2 , ε 2 < 1 2 , and ε 3 < d , by the nonnegativity of all the three integrals on the left-hand side of (3.11), estimates (3.5)–(3.7) are proved.□

Under some extra assumptions on the parameters, we obtain a refined a priori estimate for the critical case when m = 0 .

Proposition 3.2

Let N 1 , q 1 , r 1 , s > 0 , both q and r do not equal to 1, and L A satisfy (1.3)–(1.5) with σ R , m = 0 , p > 1 and

q + s > 1 + s r ( p 1 ) .

For k > max p ( q + s ) q + s ( p 1 ) ( 1 + d ) , q + s q + s 1 and 0 < d < min { 1 , p 1 , r ( q + s ) ( r + s ) ( p 1 ) ( r + s ) ( p 1 ) } , if u ( x , t ) S d 0 is a nonnegative weak solution to (1.1), then we have

(3.12) P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t C P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p α 5 ( R δ 1 + R δ 2 ) p 1 p R δ 3 p + C P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 R δ 4 ,

where

δ 3 = γ σ 5 + N σ 6 + α σ 5 + s β r σ 5 ( p σ ) , δ 4 = γ σ 7 + N σ 8 + α σ 7 + s β r σ 7 ,

and here σ 5 σ 8 are given in Lemmas 3.3 and 3.4 with m = 0 , namely,

σ 5 = q + s ( p 1 ) ( 1 + d ) , σ 6 = r ( q + s ) r ( q + s ) ( r + s ) ( p 1 ) ( 1 + d ) , σ 7 = q + s , σ 8 = r ( q + s ) r ( q + s ) r s .

Proof

By choosing φ ( x , t ) = ξ k ( x , t ) in the weak formulation (2.2), since u 0 ( x ) , φ ( x , t ) 0 , we derive

P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t k P ( 2 R ) u ζ t ζ k 1 d x d t + k P ( 2 R ) [ A ( x , u , u ) ζ ] ζ k 1 d x d t k P 2 ( R ) u ζ t ζ k 1 d x d t + k P 1 ( R ) A ( x , u , u ) ζ ζ k 1 d x d t .

We take R > R 0 > 0 and use (1.5) to obtain

P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t k P 2 ( R ) u ζ t ζ k 1 d x d t + k P 1 ( R ) K 0 1 p x σ p ( A u ) p 1 p ζ ζ k 1 d x d t .

By applying Hölder inequality with exponents p and p p 1 to the last term in the aforementioned formula, we obtain

(3.13) P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t k P 2 ( R ) u ζ t ζ k 1 d x d t + C P 1 ( R ) ( A u ) u d 1 ζ k d x d t p 1 p P 1 ( R ) x σ u ( p 1 ) ( 1 + d ) ζ p ζ k p d x d t 1 p .

By combining it with Lemmas 3.3, 3.4, and (3.6), we have

P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t C ( R δ 1 + R δ 2 ) p 1 p P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p σ 5 × 0 2 R γ R x 2 R x σ σ 6 a ( x ) σ 6 σ 5 K ( x ) s σ 6 r σ 5 ζ p σ 6 ζ k σ 6 σ 5 p σ 6 d x σ 5 σ 6 d t 1 p σ 5 + C P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 R γ 2 R γ x 2 R a ( x ) σ 8 σ 7 K ( x ) s σ 8 r σ 7 ζ t σ 8 ζ k σ 8 σ 7 σ 8 d x σ 7 σ 8 d t 1 σ 7 .

Since k > max p ( q + s ) q + s ( p 1 ) ( 1 + d ) , q + s q + s 1 , we can estimate the second and fourth terms on the right-hand side of the aforementioned inequality in a similar approach as in Proposition 3.1 for (3.11) and obtain

P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t C P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p α 5 ( R δ 1 + R δ 2 ) p 1 p R δ 3 p + C P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 R δ 4 ,

and inequality (3.12) is proved.□

Similar to Proposition 3.1, we derive a priori estimate of the positive solution to the differential inequality (1.1).

Proposition 3.3

Let N 1 , q 1 , r 1 , s > 0 , and L A satisfy (1.3)–(1.5) with σ R and m 0 , p > max { 1 , 1 m } and

q + s 1 + s r ( p + m 1 ) .

Assume k > max q + s d q + s 1 , p ( q + s d ) q + s ( p + m 1 ) and 0 < d < min { 1 , p + m 1 } . If u ( x , t ) S d is a positive weak solution to (1.1), then we have

(3.14) P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t C ( R δ 1 + R δ 5 ) ,

(3.15) P 1 ( R ) ( A u ) u d 1 ζ k d x d t C ( R δ 1 + R δ 5 ) ,

(3.16) B 2 R ( 0 ) u 0 1 d ( x ) ζ k ( x , 0 ) d x C ( R δ 1 + R δ 5 ) ,

where δ 1 is the same as in Proposition 3.1 and

δ 5 = γ + p + m 1 d q + s ( p + m 1 ) α + s ( p + m 1 d ) r [ q + s ( p + m 1 ) ] β ( p σ ) q + s d q + s ( p + m 1 ) + r N [ q + s ( p + m 1 ) ] s N ( p + m 1 d ) r [ q + s ( p + m 1 ) ] .

Proof

The proof is analogous to that of Proposition 3.1 where we deal with positive solutions and take the test function as φ ( x , t ) = u ε d ξ k since m 0 . Repeat the proof of Proposition 3.1, we obtain, in place of (3.10)

P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + 1 1 d B 2 R ( 0 ) u 0 1 d ( x ) ζ k ( x , 0 ) d x + ( d ε 3 ) P 1 ( R ) [ A ( x , u , u ) u ] u 1 d ζ k d x d t k 1 d P 2 ( R ) u 1 d ζ t ζ k 1 d x d t + C P 1 ( R ) u p + m 1 d x σ ζ p ζ k p d x d t .

By applying Lemmas 3.1 and 3.2 with m 0 into the last two integrals, we have

(3.17) ( 1 ε 1 ε 2 ) P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t + 1 1 d B 2 R ( 0 ) u 0 1 d ( x ) ζ k ( x , 0 ) d x + ( d ε 3 ) P 1 ( R ) ( A u ) u 1 d ζ k d x d t C R γ 2 R γ x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x σ 1 σ 2 d t + C 0 2 R γ R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x σ 3 σ 4 d t .

We choose k > max q + s d q + s 1 , p ( q + s d ) q + s ( p + m 1 ) , similar to the estimation of (3.11) in Proposition 3.1, we obtain

R γ 2 R γ x 2 R a ( x ) σ 2 σ 1 K ( x ) s σ 2 r σ 1 ζ t σ 2 ζ k σ 2 σ 1 σ 2 d x σ 1 σ 2 d t C R σ 2 σ 1 α + s σ 2 r σ 1 β γ σ 2 + N σ 1 σ 2 + γ C R δ 1

and

0 2 R γ R x 2 R a ( x ) σ 4 σ 3 K ( x ) s σ 4 r σ 3 x σ σ 4 ζ p σ 4 ζ k σ 4 σ 3 p σ 4 d x σ 3 σ 4 d t C R σ 4 σ 3 α + s σ 4 r σ 3 β p σ 3 + σ σ 4 + N σ 3 σ 4 + γ C R δ 5 .

Taking ε 1 < 1 2 , ε 2 < 1 2 and ε 3 < d in (3.17), and by the nonnegativity of all the three integrals on the left-hand side, estimates (3.14)–(3.16) are proved.□

The analogous of Proposition 3.2 holds when m 0 , and the precise statement of the refined a priori estimate is the following.

Proposition 3.4

Let N 1 , q 1 , r 1 , s > 0 , both q and r do not equal to 1, and L A satisfy (1.3)–(1.5), where σ R , m 0 , p > max { 1 , 1 m } , and

q + s > 1 + s r ( p + m 1 ) .

Assume k > max p ( q + s ) q + s [ m + ( p 1 ) ( 1 + d ) ] , q + s q + s 1 and 0 < d < r ( q + s ) ( r + s ) ( p + m 1 ) ( r + s ) ( p 1 ) . If u ( x , t ) S d is a positive weak solution to (1.1), then we have

(3.18) P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t C P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p α 5 ( R δ 1 + R δ 2 ) p 1 p R δ 3 p + C P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 R δ 4 ,

where

δ 3 = γ σ 5 + N σ 6 + α σ 5 + s β r σ 5 ( p σ ) , δ 4 = γ σ 7 + N σ 8 + α σ 7 + s β r σ 7 ,

and here, σ 5 σ 8 are given in Lemmas 3.3 and 3.4 with m 0 .

Proof

The proof here is similar to that of Proposition 3.2, but now we deal with the positive solution since m 0 so that (3.13) is replaced by

P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t k P 2 ( R ) u ζ t ζ k 1 d x d t + C P 1 ( R ) ( A u ) u d 1 ζ k d x d t p 1 p P 1 ( R ) x σ u m + ( p 1 ) ( 1 + d ) ζ p ζ k p d x d t 1 p .

By substituting (3.15) into the second term on the right-hand side and applying Lemmas 3.4 and 3.3 with m 0 for the first and third terms, we derive

P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t C ( R δ 1 + R δ 5 ) p 1 p P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p σ 5 × 0 2 R γ R x 2 R x σ σ 6 a ( x ) σ 6 σ 5 K ( x ) s σ 6 r σ 5 ζ p σ 6 ζ k σ 6 σ 5 p σ 6 d x σ 5 σ 6 d t 1 p σ 5 + C P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 R γ 2 R γ x 2 R a ( x ) σ 8 σ 7 K ( x ) s σ 8 r σ 7 ζ t σ 8 ζ k σ 8 σ 7 σ 8 d x σ 7 σ 8 d t 1 σ 7 .

By taking k > max p ( q + s ) q + s [ m + ( p 1 ) ( 1 + d ) ] , q + s q + s 1 , similar to the estimation of (3.11) in Proposition 3.1, we derive

P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t C P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p σ 5 ( R δ 1 + R δ 5 ) p 1 p R δ 3 p + C P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 R δ 4 ,

inequality (3.18) is proved.□

4 Proofs of main theorems

In this section, we show the process of the proofs for the main Theorems 1.1–1.4 in detail.

4.1 First critical Fujita exponent

This subsection is devoted to prove Theorems 1.1 and 1.2 on the first critical Fujita exponents.

Proof of Theorem 1.1

By N 1 , q 1 , r 1 , s > 0 , and q + s 1 + s r ( p 1 ) , we can verify that the conditions in Proposition 3.1 are fulfilled. Suppose u ( x , t ) S d 0 is a nonnegative weak solution of the differential inequality (1.1), where 0 < d < min { 1 , p 1 } , we obtain from (3.5) in Proposition 3.1 that

P ( 2 R ) a ( x ) u q d K 1 r u r s ζ k d x d t C ( R δ 1 + R δ 2 ) ,

which follows, since ζ 1 in P ( R ) ,

(4.1) P ( R ) a ( x ) u q d K 1 r u r s d x d t C ( R δ 1 + R δ 2 ) .

To make it easier to discuss the exponents of R in (4.1), we rewrite δ 1 and δ 2 as follows:

δ 1 = N q + s 1 [ q + s q 1 ( γ ) ] + d r γ + s N r α s β r ( q + s 1 ) ,

where

q 1 ( γ ) 1 + s r r α + s β r N + γ N

and

δ 2 = N + γ ( p σ ) q + s ( p 1 ) [ q + s q 2 ( γ ) ] + d s N + r ( p σ ) r α s β r [ q + s ( p 1 ) ] ,

where

q 2 ( γ ) ( p 1 ) 1 + s N + r ( p σ ) r α s β r ( N + γ p + σ ) .

Thanks to q + s 1 + s r ( p 1 ) and p σ < N , we have

N q + s 1 > 0 and N + γ ( p σ ) q + s ( p 1 ) > 0

for γ > 0 . Considering the intersection of q 1 ( γ ) and q 2 ( γ ) , for which we need to solve equation q 1 ( γ ) = q 2 ( γ ) , that is,

1 + s r r α + s β r N + γ N = ( p 1 ) 1 + s N + ( p σ ) r α s β r ( N + γ p + σ ) ,

which is equivalent to

( r N + r γ + s N r α s β ) [ N + γ p + σ N ( p 1 ) ] = 0 .

Condition α + s r β < p σ s r ( p 2 ) N gives α + s r β < N + s r N , which implies r N + r γ + s N r α s β > 0 for γ > 0 . Hence, the unique solution to equation q 1 ( γ ) = q 2 ( γ ) is

γ = N ( p 1 ) N + p σ ,

and we have γ > 0 from p > 2 N + σ 1 + N . By direct calculation, we obtain

q 1 ( γ ) = q 2 ( γ ) = p 1 + s r + p σ α s r β N = q F , 1 .

We claim that when α + s r β < p σ + min p 2 + s r N , s r ( p 2 ) N , we have

q F , 1 > max 1 , 1 + s r ( p 1 ) .

Indeed,

q F , 1 1 = 1 N p 2 + s r N + p σ α s r β ,

from α + s r β < p σ + p 2 + s r N , we know, p 2 + s r N + p σ α s r β > 0 , namely, q F , 1 > 1 . Similarly, computing directly, we obtain

q F , 1 1 + s r ( p 1 ) = 1 N s r ( p 2 ) N + p σ α s r β .

Due to α + s r β < p σ s r ( p 2 ) N , we have s r ( p 2 ) N + p σ α s r β > 0 , so that, q F , 1 > 1 + s r ( p 1 ) . Hence, q F , 1 > max 1 , 1 + s r ( p 1 ) holds.

In the following, we discuss the exponents of R on the right-hand side of (4.1) in two cases.

Case 1. If max 1 , 1 + s r ( p 1 ) q + s < q F , 1 , then γ > 0 such that q + s < min { q 1 ( γ ) , q 2 ( γ ) } . In this case, taking d > 0 small enough, we have δ 1 , δ 2 < 0 . Passing to the limit as R + in (4.1), we obtain

S a ( x ) u q d K 1 r u r s d x d t = 0 ,

which implies u ( x , t ) 0 for a.e. ( x , t ) S , namely, there is no nontrivial nonnegative weak solution in S d 0 with 0 < d < min { 1 , p 1 } .

Case 2. If max 1 , 1 + s r ( p 1 ) < q + s = q F , 1 , both q and r are not equal to 1. Combining with q F , 1 > max 1 , 1 + s r ( p 1 ) , we verify that conditions given in Proposition 3.2 hold. Assume u ( x , t ) S d 0 is a nonnegative weak solution for (1.1), where

0 < d < min 1 , p 1 , r ( q + s ) ( r + s ) ( p 1 ) ( r + s ) ( p 1 ) .

By using Proposition 3.2, we obtain another a priori estimate, namely, (3.12)

P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t C P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p σ 5 ( R δ 1 + R δ 2 ) p 1 p R δ 3 p + C P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 R δ 4 .

Taking γ = N ( p 1 ) N + p σ , a simple calculation yields

δ 1 = δ 2 , δ 3 + ( p 1 ) δ 1 = 0 and δ 4 = 0 .

Thus, we obtain from (3.12) that

(4.2) P ( R ) a ( x ) u q K 1 r u r s d x d t P ( 2 R ) a ( x ) u q K 1 r u r s ζ k d x d t C P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p σ 5 + C P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 .

Since a ( x ) u q K 1 r u r s L loc 1 ( S ) by Definition 2.1, we have

P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p σ 5 C ,

P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 C .

Then, let R + in (4.1), and we obtain

S a ( x ) u q K 1 r u r s d x d t C .

By the absolute continuity of the Lebesgue integral ([23, p. 20]), let R + , and we derive

P 1 ( R ) a ( x ) u q K 1 r u r s d x d t 0 , P 2 ( R ) a ( x ) u q K 1 r u r s d x d t 0 .

Now, by passing to the limit as R + in (4.2), we obtain

S a ( x ) u q K 1 r u r s d x d t = 0 ,

which completes the proof of the theorem.□

Proof of Theorem 1.2

The proof is exactly the same of Theorem 1.1 but now u ( x , t ) is any positive weak solution of (1.1) for 0 < d < min { 1 , p + m 1 } small enough. We use Proposition 3.3 to obtain

(4.3) P ( 2 R ) a ( x ) u q d K 1 r u r s d x d t C ( R δ 1 + R δ 5 ) ,

where δ 5 can be rewritten as follows:

δ 5 = N + γ ( p σ ) q + s ( p + m 1 ) [ q + s q 5 ( γ ) ] + d s N + r ( p σ ) r α s β r [ q + s ( p + m 1 ) ]

and

q 5 ( γ ) ( p + m 1 ) 1 + s N + r ( p σ ) r α s β r ( N + γ p + σ ) .

Now we solve equation q 1 ( γ ) = q 5 ( γ ) , that is,

1 + s r r α + s β r N + γ N = ( p + m 1 ) 1 + s N + r ( p σ ) r α s β r ( N + γ p + σ ) .

Due to p > ( 2 m ) N + σ 1 + N , the unique positive solution is

γ = N ( p + m 1 ) N + p σ .

A straightforward computation shows us that

q 1 ( γ ) = q 5 ( γ ) = p + m 1 + s r + p σ α s r β N = q F , 2 .

By arguing as in the proof of Theorem 1.1, we obtain, if

α + s r β < s r N + p σ + min ( p + m 2 ) N , s r ( p + m 1 ) N ,

then q F , 2 > max 1 , 1 + s r ( p + m 1 ) .

Next, we discuss the exponents of R on the right-hand side of (4.3) in two cases.

Case 1. If max 1 , 1 + s r ( p + m 1 ) q + s < q F , 2 , choosing d > 0 small enough, then we have δ 1 , δ 2 < 0 . Let R + in (4.3), and we obtain

S a ( x ) u q d K 1 r u r s d x d t = 0 .

Case 2. If max 1 , 1 + s r ( p + m 1 ) < q + s = q F , 2 , and both q and r are not equal to 1. Let u ( x , t ) S d be a positive weak solution for (1.1), where

0 < d < min 1 , p + m 1 r ( q + s ) ( r + s ) ( p + m 1 ) ( r + s ) ( p 1 ) .

By (3.18) in Proposition 3.4, we obtain

(4.4) P ( R ) a ( x ) u q K 1 r u r s d x d t C P 1 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 p σ 5 + C P 2 ( R ) a ( x ) u q K 1 r u r s ζ k d x d t 1 σ 7 .

Then, similar to the proof of Theorem 1.1, we derive

S a ( x ) u q K 1 r u r s d x d t = 0 ,

which completes the proof of the theorem.□

4.2 Second critical Fujita exponent

This subsection shows the proofs of Theorems 1.3 and 1.4 related to the second critical Fujita exponents.

Proof of Theorem 1.3

By N 1 , q 1 , r 1 , s > 0 , σ R , m = 0 , p > 1 , and q + s 1 + s r ( p 1 ) , all the conditions in Proposition 3.1 hold. Thus, we assume u S d 0 is a nonnegative weak solution to (1.1) and apply Proposition 3.1 to obtain

B 2 R ( 0 ) u 0 1 d ( x ) ζ k ( x , 0 ) d x C ( R δ 1 + R δ 2 ) .

By (1.12), we have

(4.5) R N μ ( 1 d ) C ( R δ 1 + R δ 2 ) ,

where

δ 1 = N + γ + r α + s β s N r ( q + s ) γ r ( q + s 1 ) + d s N r α s β + r γ r ( q + s 1 ) , δ 2 = N + γ + ( p 1 ) ( r α + s β s N ) r ( p σ ) ( q + s ) r [ q + s ( p 1 ) ] + d s N r α s β + r ( p σ ) r [ q + s ( p 1 ) ] .

For 0 < d < min { 1 , p 1 } small enough, (4.5) gives

(4.6) R μ C R γ + r α + s β s N r ( q + s ) γ r ( q + s 1 ) + R γ + ( p 1 ) ( r α + s β s N ) r ( p σ ) ( q + s ) r [ q + s ( p 1 ) ] .

One can see that the expression on the right-hand side in (4.6) arrives its minimum at

C R ( r α + s β s N ) r ( p σ ) r [ q + s ( p 1 ) ] .

By substituting it into (4.6), we obtain

R μ C R ( r α + s β s N ) r ( p σ ) r [ q + s ( p 1 ) ] .

Meanwhile, by (1.13), μ < s r N α s r β + p σ q + s ( p 1 ) , let R + , and the contradiction can be derived.□

Proof of Theorem 1.4

The proof is analogous to that of Theorem 1.3. Suppose u S d is a positive weak solution to (1.1), and by (3.16) in Proposition 3.3 and (1.12), we derive

(4.7) R N μ ( 1 d ) C ( R δ 1 + R δ 5 ) ,

where

δ 1 = N + γ + r α + s β s N r ( q + s ) γ r ( q + s 1 ) + d s N r α s β + r γ r ( q + s 1 ) , δ 2 = N + γ + ( p + m 1 ) ( r α + s β s N ) r ( p σ ) ( q + s ) r [ q + s ( p + m 1 ) ] + d s N r α s β + r ( p σ ) r [ q + s ( p + m 1 ) ] .

Taking 0 < d < min { 1 , p + m 1 } small enough, then (4.7) follows

(4.8) R μ C R γ + r α + s β s N r ( q + s ) γ r ( q + s 1 ) + R γ + ( p 1 ) ( r α + s β s N ) r ( p σ ) ( q + s ) r [ q + s ( p + m 1 ) ] .

The expression on the right-hand side in (4.8) arrives its minimum at

C R ( r α + s β s N ) r ( p σ ) r [ q + s ( p + m 1 ) ] .

By substituting it into (4.8), we obtain

R μ C R ( r α + s β s N ) r ( p σ ) r [ q + s ( p + m 1 ) ] .

By (1.14), let R + , the contradiction can be derived. So (1.1) has no nontrivial positive weak solution belonging to S d . Theorem 1.4 is proved.□


tel: +86 0532 66787153; fax: +86 0532 66787153

Acknowledgment

The authors would like to deeply thank all the reviewers for their insightful and constructive comments.

  1. Funding information: The work was supported by the Natural Science Foundation of Shandong Province of China (No. ZR2019MA072).

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

  3. Conflict of interest: The authors declare that there is no conflict of interests regarding the publication of this article.

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

References

[1] N. V. Afanas’eva and A. F. Tedeev, Theorems on the existence and nonexistence of solutions of the Cauchy problem for degenerate parabolic equations with nonlocal source, Ukr. Math. J. 57 (2005), no. 11, 1687–1711. 10.1007/s11253-006-0024-6Suche in Google Scholar

[2] M. F. Bidaut-Veron and S. Pohozaev, Nonexistence results and estimates for some nonlinear elliptic problems, J. Anal. Math. 84 (2001), no. 1, 1–49. 10.1007/BF02788105Suche in Google Scholar

[3] K. Deng and H. A. Levine, The role of critical exponents in blow-up theorems: The sequel, J. Math. Anal. Appl. 243 (2000), no. 1, 85–126. 10.1006/jmaa.1999.6663Suche in Google Scholar

[4] A. Farina and J. Serrin, Entire solutions of completely coercive quasilinear elliptic equations, J. Differ. Equations 250 (2011), no. 12, 4367–4408. 10.1016/j.jde.2011.02.007Suche in Google Scholar

[5] A. Farina and J. Serrin, Entire solutions of completely coercive quasilinear elliptic equations, II, J. Differ. Equations 250 (2011), no. 12, 4409–4436. 10.1016/j.jde.2011.02.016Suche in Google Scholar

[6] H. Fujita, On the blowing up of solutions to the Cauchy problem for ut=△u+u1+α, J. Fac. Sci. Univ. Tokyo Sect. I. 13 (1966), no. 2, 109–124. Suche in Google Scholar

[7] J. Furter and M. Grinfeld, Local vs. nonlocal interactions in population dynamics. J. Math. Biol. 27 (1989), no. 1, 65–80. 10.1007/BF00276081Suche in Google Scholar

[8] V. A. Galaktionov and H. A. Levine, A general approach to critical Fujita exponents in nonlinear parabolic problems, Nonlinear Anal.-Theor. 34 (1998), no. 7, 1005–1027. 10.1016/S0362-546X(97)00716-5Suche in Google Scholar

[9] J. S. Guo and Y. J. Guo, On a fast diffusion equation with source, Tohoku. Math. J. 53 (2001), no. 4, 571–579. 10.2748/tmj/1113247801Suche in Google Scholar

[10] K. Hayakawa, On non existence of global solutions of some semilinear parabolic differential equations, P. Jpn. Acad. A-Math. 49 (1973), no. 7, 503–505. 10.3792/pja/1195519254Suche in Google Scholar

[11] Z. X. Jiang and S. N. Zheng, A Liouville-type theorem for a doubly degenerate parabolic inequality, Acta. Math. Sci. 30 (2010), no. 3, 639–643. Suche in Google Scholar

[12] A. G. Kartsatos and V. V. Kurta, On the critical Fujita exponents for solutions of quasilinear parabolic inequalities, J. Math. Anal. Appl. 269 (2002), no. 1, 73–86. 10.1016/S0022-247X(02)00005-7Suche in Google Scholar

[13] T. Y. Lee and W. M. Ni, Global existence, large time behavior and life span of solutions of a semilinear parabolic Cauchy problem, T Am. Math. Soc. 333 (1992), no. 1, 365–378. 10.1090/S0002-9947-1992-1057781-6Suche in Google Scholar

[14] H. A. Levine, The role of critical exponents in blow-up theorems, SIAM Rev. 32 (1990), no. 2, 262–288. 10.1137/1032046Suche in Google Scholar

[15] L. W. Ma and Z. B. Fang, A new second critical exponent and life span for a quasilinear degenerate parabolic equation with weighted nonlocal sources, Commun. Pur. Appl. Anal. 16 (2017), no. 5, 1697–1706. 10.3934/cpaa.2017081Suche in Google Scholar

[16] L. W. Ma, and Z. B. Fang, Secondary critical exponent and life span for a nonlocal parabolic p-Laplace equation, Appl. Anal. 97 (2017), no. 13, 1–12. 10.1080/00036811.2017.1286646Suche in Google Scholar

[17] E. Mitidieri and S. I. Pohozaev, A priori estimates and the absence of solutions of nonlinear partial differential equations and inequalities, P. Strklov I. Math. 234 (2001), no. 3, 1–362. Suche in Google Scholar

[18] E. Mitidieri and S. I. Pohozaev, Fujita-type theorems for quasi-linear parabolic inequalities with nonlinear gradient, Dokl. Math. 66 (2002), no. 2, 187–191. Suche in Google Scholar

[19] C. L. Mu, Y. H. Li, and Y. Wang, Life span and a new critical exponent for a quasilinear degenerate parabolic equation with slow decay initial values, Nonlinear Anal.-Real. 11 (2010), no. 1, 198–206. 10.1016/j.nonrwa.2008.10.048Suche in Google Scholar

[20] C. L. Mu, R. Zeng, and S. M. Zhou, Life span and a new critical exponent for a doubly degenerate parabolic equation with slow decay initial values, J. Math. Anal. Appl. 384 (2011), no. 2, 181–191. 10.1016/j.jmaa.2010.12.042Suche in Google Scholar

[21] K. Mukai, K. Mochizuki, and Q. Huang, Large time behavior and life span for a quasilinear parabolic equation with slowly decaying initial values, Nonlinear Anal.-Theor. 39 (2000), no. 1, 33–45. 10.1016/S0362-546X(98)00161-8Suche in Google Scholar

[22] C. V. Pao, Nonlinear Parabolic and Elliptic Equations, Plenum Press, New York, 1992. 10.1007/978-1-4615-3034-3Suche in Google Scholar

[23] W. Rudin, Real and Complex Analysis, 3rd edn, McGraw-Hill, New York, 1987. Suche in Google Scholar

[24] A. A. Samarskii, V. A. Galaktionov, S. P. Kurdyumov, and A. P. Mikhailov, Blow-up in Quasilinear Parabolic Equations, Walter de Gruyter, Berlin, 1995. 10.1515/9783110889864Suche in Google Scholar

[25] F. B. Weissler, Existence and nonexistence of global solutions for a semilinear heat equation, Isr. J. Math. 38 (1981), no. 1–2, 29–40. 10.1007/BF02761845Suche in Google Scholar

[26] Z. Q. Wu, J. N. Zhao, J. X. Yin, and H. L. Li, Nonlinear Diffusion Equations, World Scientific, Singapore, 2001. 10.1142/4782Suche in Google Scholar

[27] J. G. Yang, C. X. Yang, and S. N. Zheng, Second critical exponent for evolution p-Laplacian equation with weighted source, Math. Comput. Model. 56 (2012), no. 11–12, 247–256. 10.1016/j.mcm.2011.12.036Suche in Google Scholar

[28] Y. D. Zheng and Z. B. Fang, New critical exponents, large time behavior, and life span for a fast diffusive p-Laplacian equation with nonlocal source, Z. Angew. Math. Phys. 70 (2019), no. 5, 1–17. 10.1007/s00033-019-1191-2Suche in Google Scholar

[29] Y. D. Zheng and Z. B. Fang, Critical exponents for a non-Newtonian polytropic filtration equation with weighted nonlocal inner sources, Math. Method. Appl. Sci. 43 (2020), no. 5, 2403–2420. 10.1002/mma.6051Suche in Google Scholar

[30] Y. D. Zheng and Z. B. Fang, New critical exponents for a doubly singular parabolic equation, Appl. Anal. 100 (2021), no. 11, 2386–2404. 10.1080/00036811.2019.1687885Suche in Google Scholar

Received: 2022-07-19
Revised: 2022-11-21
Accepted: 2023-02-09
Published Online: 2023-03-16

© 2023 the author(s), 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 16.9.2025 von https://www.degruyterbrill.com/document/doi/10.1515/anona-2022-0303/html
Button zum nach oben scrollen