# Global existence and nonexistence of solutions for quasilinear parabolic equation

- Xianghui Xu
^{1}, - Yong-Hoon Lee
^{1}and - Zhong Bo Fang
^{2}Email author

**2014**:33

https://doi.org/10.1186/1687-2770-2014-33

© Xu et al.; licensee Springer. 2014

**Received: **30 October 2013

**Accepted: **22 January 2014

**Published: **7 February 2014

## Abstract

This work is concerned with the global existence and nonexistence of solutions for a quasilinear parabolic equation with null Dirichlet boundary condition. Based on the Galerkin approximation technique and the theory of a family of potential wells, we obtain the invariant sets and vacuum isolating of global solutions including critical case, and we also give global nonexistence.

**MSC:**35A01, 35B06, 35B08.

## Keywords

## 1 Introduction

*p*-Laplacian parabolic equation:

where $b>0$, $\mathrm{\Omega}\subset {R}^{N}$ ($N\ge 1$) is a bounded domain with smooth boundary, $p<2+q<\mathrm{\infty}$ if $N\le p$; $p<2+q<\frac{Np}{N-p}$ if $N>p$ and $2<m<2+q$. For simplicity, we denote ${\parallel \cdot \parallel}_{{L}^{p}(\mathrm{\Omega})}$ by ${\parallel \cdot \parallel}_{p}$ and $(u,v)={\int}_{\mathrm{\Omega}}uv\phantom{\rule{0.2em}{0ex}}dx$.

Many natural phenomena have been formulated as the nonlinear diffusive equation (1.1) such as the model of non-Newton flux in the mechanics of a fluid, the model of a population, biological species and filtration; we refer to [1, 2] and the references therein. In the non-Newtonian theory, the quantity *p* is a characteristic of the medium. Media with $p>2$ are called dilatant fluids, while media with $p<2$ are called pseudoplastics. If $p=2$, they are Newtonian fluids.

and they obtained the sufficient conditions as regards global existence and nonexistence of solution by using a potential well method.

However, the potential wells used in these works were defined by the same method as Sattinger [7] and their results were similar. Until Liu [20] firstly introduced the theory of a family of potential wells, described the structure of potential wells and the estimates of the depth of potential wells. And he firstly found the phenomenon of vacuum isolating of solutions for nonlinear evolution equations. The study of applications about a family of potential wells has attracted more and more attention [20–25]. For instance, Liu and Zhao [21] not only proved the global existence and nonexistence of solutions, but they also obtained the vacuum isolating of solutions of the initial boundary value problem for semilinear hyperbolic equations and parabolic equations.

As far as we know, there are fewer papers on the global existence and nonexistence of weak solutions for nonlinear parabolic equations by using the theory of a family of potential wells. In particular, for our problem (1.1)-(1.3), the analysis of the structure and depth of the potential well, the invariant sets, the vacuum isolating of global solutions, and the question of the global existence of solutions with critical initial conditions are still open. It is difficult to obtain an *a priori* estimate of the approximate solution for the study of the existence of global solutions by using the general Galerkin approximation method, but the theory of the potential well often makes up for the defect. The combination of the two methods can be used to solve the existence of solutions effectively. Moreover, the study of the phenomenon of vacuum isolating will be helpful for us in studying the distribution of solutions in Sobolev space. But the depth of potential well *d* for the problem (1.1)-(1.3) is usually very small *etc.* Our goal is to improve the theory of a family of potential wells for studying the global existence and nonexistence of solutions for our problem (1.1)-(1.3), including the critical case, and we further generalize the results in [10–12, 20].

The outline of the paper is as follows. In Section 2, we firstly give the definition of the weak solution for problem (1.1)-(1.3), and the definition and properties of a family of potential wells. Then we prove the global existence of solutions for problem (1.1)-(1.3) by using the Galerkin approximation technique and the theory of a family of potential wells in Section 3. The invariant sets of global solutions and vacuum isolating are obtained in Section 4. Then the sufficient condition of global nonexistence of solutions is given in Section 5. Finally, we give the result of global existence with critical initial conditions.

## 2 Preliminaries

Due to the degeneracy of (1.1), problem (1.1)-(1.3) has no classical solutions in general. We need to give the definition of the weak solution firstly.

**Definition 1**A function $u=u(x,t)$ is called a weak solution of problem (1.1)-(1.3) on $\mathrm{\Omega}\times [0,T)$ if it satisfies the following conditions:

- (1)
$u\in {L}^{\mathrm{\infty}}(0,T;{W}_{0}^{1,p}(\mathrm{\Omega}))$, ${u}_{t}\in {L}^{m}(0,T;{L}^{m}(\mathrm{\Omega}))$;

- (2)
${\int}_{0}^{t}(({|{u}_{t}|}^{m-2}{u}_{t},v)+({|\mathrm{\nabla}u|}^{p-2}\mathrm{\nabla}u,\mathrm{\nabla}v))\phantom{\rule{0.2em}{0ex}}d\tau ={\int}_{0}^{t}(b{u}^{q+1},v)\phantom{\rule{0.2em}{0ex}}d\tau $ for $\mathrm{\forall}v\in {W}_{0}^{1,p}(\mathrm{\Omega})$, $t\in [0,T)$;

- (3)
$u(x,0)={u}_{0}(x)$,

where *T* is either infinity or the limit of the existence interval of solution.

where $\phi (u)=\{{u}^{1+q},\text{if}u\ge 0;0,\text{if}u0\}$.

*W*as follows:

where ${u}^{+}=max\{u,0\}$, $d={inf}_{u\in {W}_{0}^{1,p},u\ne 0}({sup}_{\lambda \ge 0}J(\lambda u))$.

where ${C}_{\ast}=sup\frac{{\parallel u\parallel}_{q+2}}{{\parallel \mathrm{\nabla}u\parallel}_{p}}$.

*p*and

*q*satisfy (H):

Before giving our main results, we show some preliminary lemmas which are very important in the following proofs. As for the proofs of these several lemmas, we will not repeat them again (see [11, 20]).

**Lemma 1** ([11], Lemma 2.2)

*For any given*$u\in {W}_{0}^{1,p}(\mathrm{\Omega})$, ${\parallel {u}^{+}\parallel}_{q+2}\ne 0$, $g(\lambda )=J(\lambda u)$

*possesses the following properties*:

- (1)
${lim}_{\lambda \to 0}g(\lambda )=0$, ${lim}_{\lambda \to +\mathrm{\infty}}g(\lambda )=-\mathrm{\infty}$;

- (2)
*There exists a unique*$\overline{\lambda}=\overline{\lambda}(u)>0$*such that*${g}^{\prime}(\overline{\lambda})=0$; - (3)
${g}^{\prime}(\lambda )>0$,

*i*.*e*. $g(\lambda )$*is increasing for*$0<\lambda <\overline{\lambda}$; ${g}^{\prime}(\lambda )<0$,*i*.*e*. $g(\lambda )$*is decreasing for*$\overline{\lambda}<\lambda <+\mathrm{\infty}$; - (4)
${g}^{\u2033}(\lambda )<0$.

**Lemma 2** ([20], Lemmas 2.1-2.3)

*The following sufficient and necessary conditions always hold*:

- (1)
*Let*$J(u)\le d(\delta )$,*then*${J}_{\delta}(u)>0$*if and only if*$0<{\parallel \mathrm{\nabla}u\parallel}_{p}<{(\frac{q+2}{pb{C}_{\ast}^{q+2}}\delta )}^{\frac{1}{q-p+2}}$. - (2)
*Let*$J(u)\le d(\delta )$,*then*${J}_{\delta}(u)<0$*if and only if*${\parallel \mathrm{\nabla}u\parallel}_{p}>{(\frac{q+2}{pb{C}_{\ast}^{q+2}}\delta )}^{\frac{1}{q-p+2}}$. - (3)
*Let*$J(u)=d(\delta )$,*then*${J}_{\delta}(u)=0$*if and only if*${\parallel \mathrm{\nabla}u\parallel}_{p}={(\frac{q+2}{pb{C}_{\ast}^{q+2}}\delta )}^{\frac{1}{q-p+2}}$.

**Lemma 3** ([20], Lemma 2.4)

*The function*$d(\delta )$

*possesses the following properties on the interval*$0\le \delta \le 1$:

- (1)
$d(0)=d(1)=0$;

- (2)
$d(\delta )$

*takes the maximum*$d({\delta}_{0})=\frac{1}{\alpha {b}^{\beta}{C}_{\ast}^{\alpha}}$*at*${\delta}_{0}=\frac{p}{q+2}$,*where*$\alpha =\frac{p(q+2)}{q-p+2}$, $\beta =\frac{p}{q-p+2}$; - (3)
$d(\delta )$

*is increasing on*$[0,{\delta}_{0}]$*and decreasing on*$[{\delta}_{0},1]$; - (4)
*For any given*$e\in (0,d({\delta}_{0}))$,*the equation*$d(\delta )=e$*has exactly two roots*${\delta}_{1}\in (0,{\delta}_{0})$*and*${\delta}_{2}\in ({\delta}_{0},1)$.

**Lemma 4** ([20], Lemma 2.5)

$d(\delta )=infJ(u)$, *where* $u\in {W}_{0}^{1,p}(\mathrm{\Omega})$, ${\parallel \mathrm{\nabla}u\parallel}_{p}\ne 0$, ${J}_{\delta}(u)=0$.

**Proposition 1** $d=d({\delta}_{0})=inf(J(u))$, *where* $u\in {W}_{0}^{1,p}(\mathrm{\Omega})$, ${\parallel \mathrm{\nabla}u\parallel}_{p}\ne 0$, $I(u)=0$.

*Proof* The result can easily be obtained by Lemma 4 and the fact that ${J}_{{\delta}_{0}}(u)=0$ is equivalent to $I(u)=0$. □

Obviously, we have ${W}_{{\delta}_{0}}=W$.

**Remark 1** From $J(u)=\frac{1-\delta}{p}{\parallel \mathrm{\nabla}u\parallel}_{p}^{p}+{J}_{\delta}(u)$, we see that ${J}_{\delta}(u)>0$ implies that $J(u)>0$.

Obviously, we have ${V}_{{\delta}_{0}}=V$.

Note that $J(u)\le \frac{1}{p}{\parallel \mathrm{\nabla}u\parallel}_{p}^{p}$, hence for any given $\delta \in (0,1)$, when $0<{\parallel \mathrm{\nabla}u\parallel}_{p}<{(1-\delta )}^{\frac{1}{2}}{(\frac{q+2}{pb{C}_{\ast}^{q+2}}\delta )}^{\frac{1}{q-p+2}}$, we have $J(u)<d(\delta )$ and ${J}_{\delta}(u)>0$. This implies that ${B}_{\overline{\delta}}\subset {W}_{\delta}$, where $\overline{\delta}$ satisfies ${(\frac{q+2}{pb{C}_{\ast}^{q+2}}\overline{\delta})}^{\frac{1}{q-p+2}}={(1-\delta )}^{\frac{1}{2}}{(\frac{q+2}{pb{C}_{\ast}^{q+2}}\delta )}^{\frac{1}{q-p+2}}$.

**Lemma 5** ([20], Theorem 2.7)

*Suppose that*${W}_{\delta}$, ${V}_{\delta}$, ${B}_{\delta}$, ${B}_{\delta}^{c}$,

*and*$\overline{\delta}$

*are defined as the above*,

*then*

**Lemma 6** ([20], Lemma 2.10)

*Assume that* $0<J(u)<d$ *for some given* $u\in {W}_{0}^{1,p}(\mathrm{\Omega})$, ${\delta}_{1}<{\delta}_{2}$ *are the two roots of the equation* $d(\delta )=J(u)$, *then the sign of* ${J}_{\delta}(u)$ *is not changed for* $\delta \in ({\delta}_{1},{\delta}_{2})$.

**Lemma 7** ([11], Lemma 2.8)

*Let*

*p*

*and*

*q*

*satisfy*(H),

*then the solutions given in Theorem*1

*satisfy*

## 3 Existence of global weak solutions

In this section, we obtain the global existence of solutions for problem (1.1)-(1.3) by combining the Galerkin approximation technique and the theory of a family of potential wells.

**Theorem 1**

*Assume that*

*p*

*and*

*q*

*satisfy*(H), ${u}_{0}(x)\in {W}_{0}^{1,p}(\mathrm{\Omega})$.

*If*$0<J({u}_{0})<d$, ${\delta}_{1}<{\delta}_{2}$

*are the two roots of the equation*$d(\delta )=J({u}_{0})$

*and*${J}_{{\delta}_{2}}({u}_{0})>0$

*or*${\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}=0$,

*then problem*(1.1a)-(1.3)

*admits a global weak solution*$u(x,t)$

*such that*$u\in {L}^{\mathrm{\infty}}(0,\mathrm{\infty};{W}_{0}^{1,p}(\mathrm{\Omega}))$, ${u}_{t}\in {L}^{m}(0,\mathrm{\infty};{L}^{m}(\mathrm{\Omega}))$

*and*$u\in {W}_{\delta}$

*for*$\delta \in ({\delta}_{1},{\delta}_{2})$

*and*$0\le t<\mathrm{\infty}$.

*Furthermore*,

*we have*

- (1)
${\parallel u(x,t)\parallel}_{m}\le {\parallel u(x,s)\parallel}_{m}$

*for*$t\ge s\ge 0$; - (2)
*If*$N<p$,*then the solution is uniquely determined by the initial function*; - (3)
*If*${u}_{0}(x)\ge 0$*a*.*e*.*in*Ω,*the solution*$u(x,t)\ge 0$*a*.*e*.*in*Ω*for any fixed*$t>0$,*hence*$u(x,t)$*is a solution of the problem*(1.1)-(1.3).

*Proof*Let $\{{\omega}_{j}(x)\}$ be a system of base functions of ${W}_{0}^{1,p}(\mathrm{\Omega})$. Construct approximate solutions ${u}_{n}(x,t)$ in the form

*s*and integrating with respect to

*t*, we obtain

Note that ${J}_{{\delta}_{2}}({u}_{0})>0$ implies ${\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}\ne 0$. By Lemma 6, we have ${J}_{\delta}({u}_{0})>0$ for $\delta \in ({\delta}_{1},{\delta}_{2})$. From this and $J({u}_{0})=d({\delta}_{1})=d({\delta}_{2})<d(\delta )$, we obtain ${u}_{0}(x)\in {W}_{\delta}$ for $\delta \in ({\delta}_{1},{\delta}_{2})$. If ${\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}=0$, then ${u}_{0}(x)\in {W}_{\delta}$ for $\delta \in (0,1)$. For any fixed $\delta \in ({\delta}_{1},{\delta}_{2})$, we have ${J}_{\delta}({u}_{n}(0))>0$ and ${J}_{n}({u}_{0})<d(\delta )$ (if ${J}_{{\delta}_{2}}({u}_{0})>0$) or ${u}_{n}(0)\in {B}_{\overline{\delta}}$ (if ${\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}=0$ and $\overline{\delta}$ is defined in Lemma 5), thereby ${u}_{n}(0)\in {W}_{\delta}$ for sufficiently large *n*.

*n*and $t>0$. Otherwise, there must be a ${t}_{0}>0$ such that ${u}_{n}({t}_{0})\in \partial {W}_{\delta}$,

*i.e.*${J}_{\delta}({u}_{n}({t}_{0}))=0$ and ${\parallel \mathrm{\nabla}{u}_{n}({t}_{0})\parallel}_{p}\ne 0$ or $J({u}_{n}({t}_{0}))=d(\delta )$. From (3.1), we have

for $t>0$ and sufficiently large *n*. From these and the compactness method, we can prove that problem (1.1a)-(1.3) admits a global weak solution $u(x,t)$ such that $u\in {L}^{\mathrm{\infty}}(0,\mathrm{\infty};{W}_{0}^{1,p}(\mathrm{\Omega}))$, ${u}_{t}\in {L}^{m}(0,\mathrm{\infty};{L}^{m}(\mathrm{\Omega}))$ and $u\in {W}_{\delta}$ for any $\delta \in ({\delta}_{1},{\delta}_{2})$ and $0\le t<\mathrm{\infty}$.

Furthermore, by Theorem 1 in [10] we can easily get the results (1)-(3), here we omit the proofs. □

Similarly, we can get the following conclusions directly.

**Corollary 1** *Under the conditions of Theorem * 1, *we have* $u\in {\overline{W}}_{{\delta}_{1}}$ *for* $0\le t<\mathrm{\infty}$.

**Corollary 2** *If the assumption* $I({u}_{0})>0$ *or* ${\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}=0$ *is replaced by* ${J}_{{\delta}_{2}}({u}_{0})>0$ *or* ${\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}=0$, *i*.*e*. ${u}_{0}(x)\in W$, *then the conclusion of Theorem * 1 *also holds*.

**Corollary 3** *If the assumption* ${u}_{0}(x)\in {B}_{{\delta}_{2}}$ *is replaced by* ${J}_{{\delta}_{2}}({u}_{0})>0$ *or* ${\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}=0$, *then problem* (1.1)-(1.3) *admits a global weak solution* $u(x,t)$ *such that* $u\in {L}^{\mathrm{\infty}}(0,\mathrm{\infty};{W}_{0}^{1,p}(\mathrm{\Omega}))$, ${u}_{t}\in {L}^{m}(0,\mathrm{\infty};{L}^{m}(\mathrm{\Omega}))$ *and* $u\in {\overline{B}}_{{\delta}_{1}}$ *for* $0\le t<\mathrm{\infty}$.

## 4 Invariant property and vacuum isolating of global solutions

In this section, we discuss the invariance of some sets under the flow of (1.1)-(1.3) and vacuum isolating behavior of solutions for problem (1.1)-(1.3).

### 4.1 Invariant property of global solutions

**Theorem 2**

*Assume that*

*p*

*and*

*q*

*satisfy*(H), $0\le {u}_{0}(x)\in {W}_{0}^{1,p}(\mathrm{\Omega})$.

*If*$0<e<d$, ${\delta}_{1}<{\delta}_{2}$

*are the two roots of the equation*$d(\delta )=e$,

*then the following hold*.

- (1)
*All solutions of problem*(1.1)-(1.3)*with initial energy*$0<J({u}_{0})\le e$*belong to*${\overline{W}}_{\delta}$*for*$\delta \in ({\delta}_{1},{\delta}_{2})$,*provided that*$I({u}_{0})>0$*or*${\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}=0$. - (2)
*All solutions of problem*(1.1)-(1.3)*with initial energy*$0<J({u}_{0})\le e$*belong to*${\overline{V}}_{\delta}$*for*$\delta \in ({\delta}_{1},{\delta}_{2})$,*provided that*$I({u}_{0})<0$.

*Proof* Firstly, we consider the case of $J({u}_{0})=e$. Here we denote $u(x,t)\triangleq u(t)$.

*T*be the existence time of $u(t)$. Since

by Theorem 1, we have ${u}_{0}(x)\in {W}_{\delta}$.

*i.e.*${J}_{\delta}(u({t}_{0}))=0$, ${\parallel \mathrm{\nabla}u({t}_{0})\parallel}_{p}\ne 0$ or $J(u({t}_{0}))=d(\delta )$. From Lemma 7

we see that $J(u({t}_{0}))=d(\delta )$ is impossible. On the other hand, if ${J}_{\delta}(u({t}_{0}))=0$ and ${\parallel \mathrm{\nabla}u({t}_{0})\parallel}_{p}\ne 0$, then by Lemma 4, we have $J(u({t}_{0}))\ge d(\delta )$, which contradicts (4.1).

(2) Let $u(t)$ be any solution of problem (1.1)-(1.3) with initial energy $J({u}_{0})=e$ and $I({u}_{0})<0$, *T* be the existence time of $u(t)$. Since the sign of ${J}_{\delta}({u}_{0})$ is not changed for $\delta \in ({\delta}_{1},{\delta}_{2})$, we have ${J}_{\delta}({u}_{0})<0$ for $\delta \in ({\delta}_{1},{\delta}_{2})$. From this and $J({u}_{0})<d(\delta )$ for $\delta \in ({\delta}_{1},{\delta}_{2})$ we obtain ${u}_{0}(x)\in {V}_{\delta}$ for $\delta \in ({\delta}_{1},{\delta}_{2})$.

Next we prove $u(t)\in {V}_{\delta}$ for $\delta \in ({\delta}_{1},{\delta}_{2})$ and $0<t<T$. Otherwise, there exists a ${t}_{0}\in (0,T)$ such that $u({t}_{0})\in \partial {V}_{\delta}$ for some $\delta \in ({\delta}_{1},{\delta}_{2})$, *i.e.* ${J}_{\delta}(u({t}_{0}))=0$ or $J(u({t}_{0}))=d(\delta )$. From (4.1) we see that $J(u({t}_{0}))=d(\delta )$ is impossible. On the other hand, let ${t}_{0}$ be the first time such that ${J}_{\delta}(u({t}_{0}))=0$, then ${J}_{\delta}(u(t))<0$ for $0\le t<{t}_{0}$. From (4.1) and Lemma 2, we have ${\parallel \mathrm{\nabla}u(t)\parallel}_{p}>{(\frac{q+2}{pb{C}_{\ast}^{q+2}}\delta )}^{\frac{1}{q-p+2}}$ for $0\le t<{t}_{0}$. Hence we have ${\parallel \mathrm{\nabla}u({t}_{0})\parallel}_{p}>{(\frac{q+2}{pb{C}_{\ast}^{q+2}}\delta )}^{\frac{1}{q-p+2}}$, thus by Lemma 4, we get $J(u({t}_{0}))\ge d(\delta )$, which contradicts (4.1).

For the case of $0<J({u}_{0})<e$, we can obtain the same results as the case $J({u}_{0})=e$ by Lemma 4, we omit it here. □

**Remark 2** Assume that $0<J({u}_{0})\le e$, then ${W}_{\delta}$ and ${V}_{\delta}$ are invariant under the flow of (1.1)-(1.3) for any $\delta \in ({\delta}_{1},{\delta}_{2})$.

From the above Theorem 2 and Lemma 2, we can easily get the following conclusions.

**Theorem 3**

*Let*

*p*

*and*

*q*

*satisfy*(H), $0\le {u}_{0}(x)\in {W}_{0}^{1,p}(\mathrm{\Omega})$.

*Assume that*$0<e<d$, ${\delta}_{1}<{\delta}_{2}$

*are the two roots of the equation*$d(\delta )=e$,

*then the following hold*.

- (1)
*All solutions of problem*(1.1)-(1.3)*with initial energy*$0<J({u}_{0})\le e$*and*${u}_{0}(x)\in {B}_{{\delta}_{0}}$*belong to*${\overline{B}}_{{\delta}_{1}}$*for*$\delta \in ({\delta}_{1},{\delta}_{2})$. - (2)
*All solutions of problem*(1.1)-(1.3)*with initial energy*$0<J({u}_{0})\le e$*and*${u}_{0}(x)\in {B}_{{\delta}_{0}}^{c}$*belong to*${\overline{B}}_{{\delta}_{2}}^{c}$*for*$\delta \in ({\delta}_{1},{\delta}_{2})$.

**Remark 3** Let *p* and *q* satisfy (H), $0\le {u}_{0}(x)\in {W}_{0}^{1,p}(\mathrm{\Omega})$. Assume that $0<J({u}_{0})\le e$, then ${B}_{\delta}$ and ${B}_{\delta}^{c}$ are invariant under the flow of (1.1)-(1.3) for any $\delta \in ({\delta}_{1},{\delta}_{2})$.

### 4.2 Vacuum isolating of global solutions

*e*. As the limit case $e=0$, we obtain the biggest vacuum region of solutions (for $J({u}_{0})\ge 0$)

**Theorem 4** *Let* *p* *and* *q* *satisfy* (H), $0\le {u}_{0}(x)\in {W}_{0}^{1,p}(\mathrm{\Omega})$. *All nontrivial solutions of problem* (1.1)-(1.3) *with initial energy* $J({u}_{0})=0$ *lie outside of the ball* ${B}_{1}$ (*maybe in* $\partial {B}_{1}$).

**Theorem 5**

*Let*

*p*

*and*

*q*

*satisfy*(H), $0\le {u}_{0}(x)\in {W}_{0}^{1,p}(\mathrm{\Omega})$.

*All nontrivial solutions of problem*(1.1)-(1.3)

*with initial energy*$J({u}_{0})<0$

*satisfy*

*and*

**Remark 4** The proofs of Theorems 4-5 are similar to Theorems 4.7-4.8 in [20], we omit them.

## 5 Nonexistence of global solutions

In this section, we given the sufficient condition of global nonexistence of solutions.

**Theorem 6** *Assume that* $2<m<2+q$, ${u}_{0}(x)\in {W}_{0}^{1,p}(\mathrm{\Omega})$, $u(x,t)$ *is a local solution of problem* (1.1)-(1.3) *on* $[0,T]$, *then no solution of* (1.1)-(1.3) *can exist on* $[0,\mathrm{\infty})$ *when* $J({u}_{0})<0$.

*Proof* Assume for contradiction that there is a solution of (1.1)-(1.3) on $[0,\mathrm{\infty})$.

*u*and integrating over Ω, we have

This is impossible, since the left hand side is finite and the right hand side goes to ∞ as $t\to \mathrm{\infty}$. □

## 6 Existence of global solution with critical initial conditions

In this section, we prove the result of global existence with critical initial conditions.

**Theorem 7** *Assume that* *p* *and* *q* *satisfy* (H), $0\le {u}_{0}(x)\in {W}_{0}^{1,p}(\mathrm{\Omega})$. *If* $J({u}_{0})=d(\delta )$, ${J}_{{\delta}_{1}}({u}_{0})>0$ *or* ${J}_{{\delta}_{1}}({u}_{0})=0$, $0<J({u}_{0})\le d(\delta )$, ${\delta}_{1}<{\delta}_{2}$ *are the two roots of the equation* $J({u}_{0})=d(\delta )$, *then problem* (1.1)-(1.3) *admits a global solution* $u(x,t)$ *such that* $u\in {L}^{\mathrm{\infty}}(0,\mathrm{\infty};{W}_{0}^{1,p}(\mathrm{\Omega}))$, ${u}_{t}\in {L}^{m}(0,\mathrm{\infty};{L}^{m}(\mathrm{\Omega}))$ *and* $u\in {\overline{W}}_{\delta}$ *for any* $\delta \in ({\delta}_{1},{\delta}_{2})$ *and* $0\le t<\mathrm{\infty}$, *where* ${\overline{W}}_{\delta}={W}_{\delta}\cup \partial {W}_{\delta}=\{u\in {W}_{0}^{1,p}(\mathrm{\Omega})|{J}_{\delta}(u)\ge 0,J(u)\le d(\delta )\}$.

*Proof*Let ${\lambda}_{n}=1-\frac{1}{n}$, ${u}_{0n}(x)={\lambda}_{n}{u}_{0}(x)$, $n=2,3,\dots $ . Consider the initial condition

with the corresponding problem (1.1)-(1.3) and suppose that ${\delta}_{1}^{n}<{\delta}_{2}^{n}$ are two roots of the equation $J({u}_{0n})=d(\delta )$.

*i.e.*${[\frac{{\delta}_{1}(q+2){\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}^{p}}{bp{\parallel {u}_{0}\parallel}_{q+2}^{q+2}}]}^{\frac{1}{q-p+2}}\ge 1$. Thus $\overline{\lambda}={(\frac{{\parallel \mathrm{\nabla}{u}_{0}\parallel}_{p}^{p}}{bp{\parallel {u}_{0}\parallel}_{q+2}^{q+2}})}^{\frac{1}{q-p+2}}\ge {(\frac{1}{{\delta}_{1}}\cdot \frac{p}{q+2})}^{\frac{1}{q-p+2}}\ge 1$. From $0<{\lambda}_{n}<1\le \overline{\lambda}$ and Lemma 1, we get

Obviously, $J(u)=\frac{1-\delta}{p}{\parallel \mathrm{\nabla}u\parallel}_{p}^{p}+{J}_{\delta}(u)$ implies that $J({u}_{0n})>0$. As ${\delta}_{2}^{n}>{\delta}_{0}>{\delta}_{1}$ and ${J}_{\delta}(u)$ is increasing with *δ*, it follows that ${J}_{{\delta}_{2}^{n}}({u}_{0n})>0$.

*i.e.*

*u*,

*ξ*, and subsequence $\{{u}_{\nu}\}$ of $\{{u}_{n}\}$ such that

By using the monotone operator method, we get $\xi =b{u}^{q+1}$.

On the other hand, letting $n=\nu \to \mathrm{\infty}$ in ${u}_{n}(x,0)={u}_{0n}(x)$ we get $u(x,0)={u}_{0}(x)$ in ${W}_{0}^{1,p}(\mathrm{\Omega})$. Also ${\delta}_{1}^{n}\to {\delta}_{1}$, ${\delta}_{2}^{n}\to {\delta}_{2}$ as $n\to \mathrm{\infty}$.

By Lemma 2, we have ${J}_{\delta}(u)\ge 0$ and $u\in {\overline{W}}_{\delta}$ for any $\delta \in ({\delta}_{1},{\delta}_{2})$. □

**Remark 5** The invariant sets and vacuum of solutions for problem (1.1)-(1.3) with critical initial conditions also occur.

**Remark 6** Taking $m=2$ or $p=2$, Theorem 7 is still satisfied and generalizes the results of [11]. Similarly, the invariant sets and vacuum isolating of solutions also occur.

**Remark 7**In fact, all the results in our paper also hold for the homogeneous Dirichlet initial boundary value problem for the more general equation

## Declarations

### Acknowledgements

The second and third authors were supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (No. 2012005767) and the National Science Foundation of Shandong Province of China (ZR2012AM018) and the Fundamental Research Funds for the Central Universities (No. 201362032), respectively. The authors would like to express their sincere gratitude to the anonymous reviewers for their insightful and constructive comments.

## Authors’ Affiliations

## References

- Bebernes J, Eberly D Applied Mathematical Sciences 83. In
*Mathematical problems from combustion theory*. Springer, New York; 1989.View ArticleGoogle Scholar - Pao CV:
*Nonlinear parabolic and elliptic equations*. Plenum Press, New York; 1992.Google Scholar - Bourgault Y, Coudiere Y, Pierre C: Existence and uniqueness of the solution for the bidomain model used in cardiac electrophysiology.
*Nonlinear Anal., Real World Appl.*2009, 10: 458-482. 10.1016/j.nonrwa.2007.10.007MathSciNetView ArticleGoogle Scholar - Klainerman S, Ponce G: Global, small amplitude solutions to nonlinear evolution equations.
*Commun. Pure Appl. Math.*1983, 36: 133-141. 10.1002/cpa.3160360106MathSciNetView ArticleGoogle Scholar - Gazzola F, Weth T: Finite time blow-up and global solutions for semilinear parabolic equations with initial data at high energy level.
*Differ. Integral Equ.*2005, 18: 961-990.MathSciNetGoogle Scholar - Levine HA, Park SR, Serrin J: Global existence and nonexistence theorems for quasilinear evolution equations of formally parabolic type.
*J. Differ. Equ.*1998, 142: 212-229. 10.1006/jdeq.1997.3362MathSciNetView ArticleGoogle Scholar - Sattinger DH: On global solution of nonlinear hyperbolic equations.
*Arch. Ration. Mech. Anal.*1968, 30: 148-172.MathSciNetView ArticleGoogle Scholar - Levine HA, Smith RA: Ames, A potential well theory for the heat equation with a nonlinear boundary condition.
*Math. Methods Appl. Sci.*1987, 9: 127-136. 10.1002/mma.1670090111MathSciNetView ArticleGoogle Scholar - Payne LE, Sattinger DH: Saddle points and instability of nonlinear hyperbolic equations.
*Isr. J. Math.*1975, 22: 273-303. 10.1007/BF02761595MathSciNetView ArticleGoogle Scholar - Tsutsumi M: Existence and nonexistence of global solutions for nonlinear parabolic equations.
*Publ. Res. Inst. Math. Sci.*1972/73, 8: 211-229. 10.2977/prims/1195193108MathSciNetView ArticleGoogle Scholar - Liu YC, Zhao JS:Nonlinear parabolic equations with critical initial conditions $J({u}_{0})=d$ or $I({u}_{0})=0$.
*Nonlinear Anal.*2004, 58: 873-883. 10.1016/j.na.2004.05.019MathSciNetView ArticleGoogle Scholar - Pang JS, Zhang HW: Existence and nonexistence of the global solution on the quasilinear parabolic equation.
*Chin. Q. J. Math.*2007, 22: 444-450.MathSciNetGoogle Scholar - Todorova G: Existence of a solution of the wave equation with nonlinear damping and source terms.
*J. Differ. Equ.*1994, 109: 295-308. 10.1006/jdeq.1994.1051MathSciNetView ArticleGoogle Scholar - Ikehata R: Some remarks on the wave equations with nonlinear damping and source terms.
*Nonlinear Anal. TMA*1996, 27: 1165-1175. 10.1016/0362-546X(95)00119-GMathSciNetView ArticleGoogle Scholar - Ikehata R, Suzuki T: Stable and unstable sets for evolution equations of parabolic and hyperbolic type.
*Hiroshima Math. J.*1996, 26: 475-491.MathSciNetGoogle Scholar - Pucci P, Serrin J: Asymptotic stability for nonautonomous dissipative wave systems.
*Commun. Pure Appl. Math.*1996, 49: 177-216. 10.1002/(SICI)1097-0312(199602)49:2<177::AID-CPA3>3.0.CO;2-BMathSciNetView ArticleGoogle Scholar - Aassila M: Global existence and global nonexistence of solutions to a wave equation with nonlinear damping and source terms.
*Asymptot. Anal.*2002, 30: 301-311.MathSciNetGoogle Scholar - Alfredo J, Avila E: A characterization of global and nonglobal solutions of nonlinear wave and Kirchhoff equations.
*Nonlinear Anal.*2003, 52: 1111-1127. 10.1016/S0362-546X(02)00155-4MathSciNetView ArticleGoogle Scholar - Xu RZ, Cao XY, Yu T: Finite time blow-up and global solutions for a class of semilinear parabolic equations at high energy level.
*Nonlinear Anal., Real World Appl.*2012, 13: 197-202. 10.1016/j.nonrwa.2011.07.025MathSciNetView ArticleGoogle Scholar - Liu YC: On ponential wells and vacuum isolating of solutions for semilinear wave equations.
*J. Differ. Equ.*2003, 192: 155-169. 10.1016/S0022-0396(02)00020-7View ArticleGoogle Scholar - Liu YC, Zhao JS: On potential wells and applications to semilinear hyperbolic equations and parabolic equations.
*Nonlinear Anal.*2006, 64: 2665-2687. 10.1016/j.na.2005.09.011MathSciNetView ArticleGoogle Scholar - Liu YC, Xu RZ, Yu T: Wave equations and reaction-diffusion equations with several nonlinear source terms.
*Appl. Math. Mech.*2007, 28: 1209-1218. 10.1007/s10483-007-0909-yMathSciNetView ArticleGoogle Scholar - Liu YC, Xu RZ: Potential well method for initial boundary value problem of the generalized double dispersion equations.
*Commun. Pure Appl. Anal.*2008, 7: 63-81.MathSciNetGoogle Scholar - Jiang XL, Xu RZ: Global well-posedness for semilinear hyperbolic equations with dissipative term.
*J. Appl. Math. Comput.*2012, 38: 467-487. 10.1007/s12190-011-0491-2MathSciNetView ArticleGoogle Scholar - Wu ST: Global existence, blow-up and asymptotic behavior of solutions for a class of coupled nonlinear Klein-Gordon equations with damping terms.
*Acta Appl. Math.*2012, 119: 75-95. 10.1007/s10440-011-9662-2MathSciNetView ArticleGoogle Scholar

## Copyright

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.