• No results found

Existence and multiplicity of positive solutions for p-Kirchhoff type problem with singularity

N/A
N/A
Protected

Academic year: 2020

Share "Existence and multiplicity of positive solutions for p-Kirchhoff type problem with singularity"

Copied!
17
0
0

Loading.... (view fulltext now)

Full text

(1)

R E S E A R C H

Open Access

Existence and multiplicity of positive

solutions for

p

-Kirchhoff type problem with

singularity

Dechen Wang and Baoqiang Yan

*

*Correspondence:

Yanbqcn@aliyun.com

School of Mathematical Sciences, Shandong Normal University, Jinan, 250014, P.R. China

Abstract

In this paper, we consider a class ofp-Kirchhoff type problems with a singularity in a bounded domain inRN. By using the variational method, the existence and

multiplicity of positive solutions are obtained.

Keywords: p-Kirchhoff type problem; singularity; Nehari manifold

1 Introduction and main results

The purpose of this paper is to investigate the existence of multiple positive solutions to the following problem:

⎧ ⎨ ⎩

M(|∇u|pdx)

pu=λf(x)ur+g(x)uq– in,

u=  on∂,

(.)

wherepu=div(|∇u|p–∇u),is a smooth bounded domain inRN,  <r<  <p<q<p

(p∗=NNpp ifN>pandp∗=∞ifNp),M(s) =asp–+banda,b,λ> .f,gC() are nontrivial nonnegative functions.

Problem (.) is related to the stationary problem introduced by Kirchhoff in []. More precisely, it is the model

ρ∂

u

∂t –

ρ

h + E

L

L

uxdx u

∂x = ,

where ρ,ρ,E,L are constants, which was proposed as an extension of the classical

D’Alembert’s wave equation for free vibrations of elastic strings to describe transversal oscillations of a stretched string. For more details and backgrounds, we refer to [, ].

The existence and multiplicity of solutions for the following problem: ⎧

⎨ ⎩

M(|∇u|pdx)

pu=h(x,u) in,

u=  on∂, (.)

on a smooth bounded domainRN has been studied in many papers. Liu and Zhao

[] proved (.) has at least two nontrivial weak solutions by Morse theory under some

(2)

restriction onM(s) andh(x,u). In [], the authors considered the following problem: ⎧

⎨ ⎩

M(|∇u|pdx)pu=λf(x)|u|q–u+g(x)|u|r–u in,

u=  on∂, (.)

whereM(s) =as+b,  <q<p<rp∗, they proved the existence of multiplicity nontrivial solutions by using the Nehari manifold when the weight functionsf(x) andg(x) change their signs. For more results, we refer to [–] and the references therein.

Whenp=  andN= , problem (.) reduces to the following singular Kirchhoff type problem:

⎧ ⎨ ⎩

–(a+b|∇u|dx)u=λf(x)ur+μg(x)uq in,

u=  on∂, (.)

where  <q≤, the existence of solutions for problem (.) has been widely studied (see [–]). When  <q<  and λ= , Liu and Sun [] proved that problem (.) has at least two positive solutions forμ>  small enough. Liaoet al.showed the multiplicity of positive solutions by the Nehari mainfold in the case ofq=  in []. Whenqis a critical exponents, at least two positive solutions are obtained by variational and perturbation methods in [].

However, the singularp-Kirchhoff type problems have few been considered, especially

p= ,N= . Here we focus on extending the results in [] and []. In fact, the extension

is nontrivial and requires a more careful analysis. Our method is based on the Nehari manifold; see [, , , ].

Before starting our main theorems, we make use of the following notations:

•LetW,p() be the Sobolev space with norm u = (|∇u|pdx)p, the norm inLp() is denoted by p;

•LetSzbe the best Sobolev constant for the embedding ofW,p() inLz() with  <

z<p∗. Then, for alluW,p()\{},

u zS

–p

z u .

In general, we say that a functionuW,p() is a weak solution of problem (.) if

M u p

|∇u|p–∇uϕdxλ

f|u|–rϕdx

g|u|q–ϕdx= 

for allϕW,p(). Thus, the functional corresponding to problem (.) is defined by

J(u) = 

pM

u pλ  –r

f|u|–rdx–

q

g|u|qdx, ∀uW,p(),

whereM(s) =sM(t)dt.

To obtain the existence results, we introduce the Nehari manifold:

=

uW,p()\ {}:M u p u pλ

f|u|–rdx

g|u|qdx= 

(3)

and we define

K(u) = (p– )M u p u p+pM u p up+λr

f|u|–rdx– (q– )

g|u|qdx.

Now we splitinto three disjoint parts as follows:

+=

u:K(u) > 

;

=

u:K(u) = 

;

–=

u:K(u) < 

.

Letλ∗=max{(–r)λ(a)

p

p+

p

,(–r)λ

p }, whereλ(a) andλare given by

λ(a) = pS

–r p

–r p

abp–(qp)(qp

p–)p–

(q+r– ) f

pS

q p

q p

abp–(p+r– )

(q+r– ) g

p+r–

q–p+

and

λ= bS

–r p

–r(qp)

(q+r– ) f

bS

q p

q(p+r– )

(q+r– ) gp+r–

qp ,

then we state the main theorems.

Theorem . Assume that p<q<pand N< p.Then,for each a> and <λ<λ,the problem(.)has at least two positive solutions u+λ+and uλ–.

Define

=inf

u p:uW,p(),

g|u|pdx= 

, (.)

then>  is obtained by someφW,p() with

g|φ|

p

dx= . In particular,

g|u|pdxu p (.)

and ⎧ ⎨ ⎩

u p

pu=μg|u|p

–

in,

u=  on∂, (.)

whereμis an eigenvalue of (.),uW,p() is nonzero and an eigenvector corresponding toμsuch that

u p

|∇u|p–∇u∇(uϕ)dx=μ

(4)

we write

I(u) = u p, foruE=

uW,p() :

g|u|pdx= 

,

and all distinct eigenvalues of (.) denoted by  <μ<μ<· · ·, we have

μ=inf

uEI(u) > ,

whereμis simple, isolated and can be obtained at someψEandψ>  in(see []).

Theorem . Assume that p=q<pand N< p.Then

(i) for eachaandλ> ,the problem(.)has at least one positive solution; (ii) for eacha<and <λ<–prλ,ˆ where

ˆ λ= bS

–r p

–r(p–p)

f ∞(p+r– )

b(p+r– ) ( –a)(p+r– )

p+r–

p–p ,

the problem(.)has at least two positive solutionsu+λ+,uλand

lim

a→

uinfN

λ

J(u) =∞.

This paper is organized as follows: In Section , we present some lemmas which will be used to prove our main results. In Section  and Section , we will prove Theorems . and ., respectively.

2 Preliminaries

Lemma . (i)If qp,then the energy functional J(u)is coercive and bounded below in ;

(ii)if q<p,then the energy functional J(u)is coercive and bounded below in W,p

 ().

Proof (i) Foru, we have

M u p u pλ

f|u|–rdx

g|u|qdx= .

By the Sobolev inequality,

J(u) = 

pM

u pλ  –r

f|u|–rdx–

q

g|u|qdx

=

pM

u p–

qM

u p u pλq+r– 

q( –r)

f|u|–rdx

u p

pq

a(qp)

p u

p–p+b(qp) λq+r– 

q( –r) fS r–

p

–r u –r

b(qp)

pq u

pλq+r– 

q( –r) fS r–

p

–r u –r.

(5)

(ii) ForuW,p(), we have

J(u) = 

pM

u pλ  –r

f|u|–rdx–

q

g|u|qdx

a

pu

p+b

p u

pλ fS

r–

p

–r

 –r u

–r gS

qp

q

q u

q

=

a pu

pq

gS

qp

q

q u

q+

b p u

p+r–λ fS

r–

p

–r

 –r u

–r.

Thus,J(u) is coercive and bounded below inW,p().

Lemma . If q>pand <λ<max{λ(a),λ},then,for all a> ,

(i) the submanifoldN

λ=∅; (ii) the submanifoldNλ±=∅.

Proof (i) SupposeN

λ=∅. Then, foru, we have

(q+r– ) gS

qp

q u q≥(q+r– )

g|u|qdx

=ap+r–  u p+b(p+r– ) u p

≥ ⎧ ⎨ ⎩

ppabp–(p+r– ) up–,

b(p+r– ) u p, (.)

and

λ(q+r– ) fS

––pr

–r u –rλ(q+r– )

f|u|–rdx

=aqpu p+b(qp) u p

≥ ⎧ ⎨ ⎩

ppabp–(qp)(qp

p–)p– up–,

b(qp) u p. (.)

By (.) and (.), for allu, we have

pS

q p

q p

abp–(p+r– )

(q+r– ) g

q–p+

u

λ(q+r– ) f

pS

–r p

–r p

abp–(qp)(qp

p–)p–

 p+r–

and

bS

q p

q(p+r– )

(q+r– ) g

qp

u

λ(q+r– ) f

bS

–r p

–r(qp)

p+r–

.

(6)

(ii) FixuW,p(). Let

ha(t) =atp

–(–r)

u p+btp–(–r) u ptq–(–r)

g|u|qdx fora,t≥.

We see thatha() =  andha(t)→–∞ast→ ∞. Sinceq>pand

ha(t) =tp+r–

ap+r– tp–p u p+b(p+r– ) u p

– (q+r– )tqp

g|u|qdx ,

there is a uniqueta,max>  such thatha(t) reaches its maximum atta,max, increasing for t∈[,ta,max) and decreasing fort∈(ta,max,∞) withlimt→∞ha(t) = –∞. Clearly, iftu, thentu+(or–) if and only ifha(t) >  (or < ). Moreover,

t,max=

b(p+r– ) u p

(q+r– )g|u|qdx

qp

and

h(t,max)

=b

q+r–

qp

p+r– 

q+r–  p+r–

qp

p+r– 

q+r–  q+r–

qp u p(q+r–)

qp

(g|u|qdx)p+qr–pb(qp)

q+r– 

bS

q p

q(p+r– )

(q+r– ) gp+r–

qp

u –r.

On the other hand, since

ha() =  <λ

f|u|–rdxλ fS

––pr –r u –r

<b(qp)

q+r– 

bS

q p

q(p+r– )

(q+r– ) gp+r–

qp

u –r

h(t,max) <ha(ta,max), (.)

there exist uniquet+andtsuch that  <t+<t

a,max<t–, ha

t+=λ

f|u|–rdx=ha

t

and

hat+>  >hat–.

(7)

Lemma . (i)If q=pand a

,then,for allλ> ,N

+

λ==∅; (ii)if q=p,a<

and <λ<λ,ˆ then Nλ=N

+

λand Nλ±=∅.

Proof First, we show that+=. Indeed, for allu, we have

ap+r–  u p+b(p+r– ) u pp+r– 

g|u|pdx

≥(a– )(p+r– ) u

p+b(p+r– ) u p> .

Therefore,uN+

λ. Next, we declare+=∅. FixuW,p(). Let

¯

h(t) =tp+r–

a u p–

g|u|pdx +btp+r– u p fora,t≥.

Obviously,h¯() =  andlimt→∞h¯(t) =∞. Since

¯

h(t) =p+r– tp+r–

a u p–

g|u|pdx

+b(p+r– )tp+r– u p,

we can deduce thath¯(t) is increasing fort∈[,∞). Thus, there is a uniquet+>  such that

¯

h(t+) =λ

f|u|

–rdxandh¯(t+) > . That is,t+uN+

λ.

(ii) The proof is similar to Lemma ., we omit it here.

We write=+∪–and define

α+= inf

uNλ+J(u); α

= inf

uNλJ(u),

then we have the following lemma.

Lemma . Suppose that q>pand <λ<λ,then we have

(i) α+< ; (ii) α>C

,for someC> . In particularα+=infuJ(u).

Proof (i) Letu+, it follows that

M u p u pλ

f|u|–rdx

g|u|qdx= 

and

λ(q+r– )

(8)

Substituting this intoJ(u), we have

J(u) = 

pM

u pλ  –r

f|u|–rdx–

q

g|u|qdx

=

pM

u p–

qM

u p u pλq+r– 

q( –r)

f|u|–rdx

<a(qp

)( –rp)

pq( –r) u

p+b(qp)( –rp)

pq( –r) u

p< ,

and thenα+< . (ii) LetuN

λ. We divide the proof into two cases.

Case(A):λ∗=(–r)λ

p . Sinceu–, and by the Sobolev inequality,

b(p+r– ) u pap+r–  u p+b(p+r– ) u p

< (q+r– )S

q p g

u q, which implies

u >

bS

q p

q(p+r– )

(q+r– ) g

qp

for allu–.

Hence,

J(u)≥a(qp

) u p

pq +

b(qp) u p pqλ

q+r– 

q( –r) fS

––pr –r u –r

u –r

b(qp)

pq u

p+r–λq+r–  q( –r) fS

––pr –r

>

bS

q p

q(p+r– )

(q+r– ) g

–r

qpb(qp)

pq

bS

q p

q(p+r– )

(q+r– ) gp+r–

qp

λq+r– 

q( –r) fS

––pr –r

=C.

Thus, if  <λ<(–r)λ

p , thenα–>C> .

Case(B):λ∗=(–r)λ(a)

p

p+

p

. By (.), one has

pp

abp–p+r–  up–(q+r– ) g

S

qp

q u q,

which implies

u >

pS

q p

q p

abp–(p+r– )

(q+r– ) g

q–p+

for allu–.

Repeating the argument of case (A), we conclude ifλ< (–r)λ(a)

p

p+

p

, thenα>C

for some

(9)

Lemma . Suppose that q=p,a<

and <λ<

–r

p λ,ˆ then we have

(i) αˆ+< ;

(ii) αˆ–>C,for someC> . In particularαˆ+=inf

uNλJ(u).

Proof (i) Repeating the same argument of Lemma .(i), we conclude thatαˆ+< .

(ii) LetuN

λ. By (.), one has

b(p+r– ) u p<p+r– 

g|u|pdxa u p

≤( –a)(p+r– ) u

p,

which implies that

u >

b(p+r– ) ( –a)(p+r– )

p–p

for alluNλ–. (.)

Then we have

J(u) = 

pM

u pλ  –r

f|u|–rdx–

q

g|u|qdx

u –r

(p– )b

pu

p+r–λp+r–  p( –r) fS

––pr

–r

>

b(p+r– ) ( –a)(p+r– )

–r

p–p(p– )b

p

b(p+r– ) ( –a)(p+r– )

p+r–

p–p

λp

+r– 

p( –r) fS ––pr

–r

. (.)

Thus, ifλ<–prλ, thenˆ αˆ–>C

for someC> .

Lemma . For each u+(resp.u–),there existε> and a continuous function

f :B(;ε)W,p()→R+such that

f() = ,f(ω) > ,f(ω)(u+ω)Nλ+resp. uNλ–, for allωB(;ε),

where B(;ε) ={ωW,p() : ω <ε}.

Proof Foru+, defineF:W

,p

 ()×RRas follows:

F(ω,t) =atp+r–

(u+ω)pdx

p

+btp+r–

(u+ω)pdx

tq+r–

g|u|qdxλ

f|u|–rdx.

SinceuN+

λ, it is easily seen thatF(, ) =  andFt(, ) > . Then by the implicit function

(10)

f :B(;ε)W,p()→R+such that

f() = ,f(ω) > ,f(ω)(u+ω)Nλ+, for allωB(;ε).

In the same way, we can prove the caseuN

λ.

Remark . The proof of Lemma . is inspired by [].

3 Proof of Theorem 1.1

By Lemma . and the Ekeland variational principle [], there exists a minimizing se-quence{un} ⊂+such that

(i) J(un) <α++

n;

(ii) J(u) >J(un) –

n uun , ∀uN +

λ.

Note thatJ(|un|) =J(un). We may assume thatun≥ in. Using Lemma . again, we can

see that there is a constantC>  such that, for allnN+, unC. Thus, there exist a

subsequence (still denoted by{un}) andu+λinW

,p

 () such that

unu+λweakly inW

,p

 (),

unu+λstrongly inL–r(),

unu+λstrongly inLq(),

unu+λa.e. in.

Now we conclude thatu+λ+is a positive solution of (.). The proof is inspired by Liu and Sun []. In order to prove the claim, we divide the arguments into six steps.

Step:u+

λis not identically zero.

Indeed, it is an immediate conclusion of the following inequalities:

Ju+λ

≤ lim

n→∞J(un) =α

+< .

Step: There existsCsuch that up to a subsequence we have

ap+r–  un p

+b(p+r– ) un p– (q+r– )

gu+λqdx>C. (.)

In order to prove (.), it suffices to verify

ap+r–  lim

n→∞ un

p+b(p+r– ) lim n→∞ un

p> (q+r– )

gu+λ

q

dx. (.)

Sinceun+,

ap+r–  un p

+b(p+r– ) un p≥(q+r– )

gu+λ

q

(11)

It follows that

ap+r–  lim

n→∞ un

p+b(p+r– ) lim n→∞ un

p(q+r– )

gu+λ

q

dx.

Suppose by contradiction that

ap+r–  lim

n→∞ un

p

+b(p+r– ) lim

n→∞ un

p= (q+r– )

gu+λqdx. (.)

Then, from (.) and (.), one has

ap+r–  lim

n→∞ un

p+b(p+r– ) lim n→∞ un

p= (q+r– )

gu+λ

q

dx.

Thus un pconverges to a positive numberAthat satisfies

ap+r– Ap+b(p+r– )A= (q+r– )

gu+λqdx

and

aqpAp+b(qp)A=λ(q+r– )

fu+λ

–r

dx.

On the other hand, by (.), we have

≤λ∗–λ

f|un|–rdx

<bq+qr–p

p+r– 

q+r–  p+r–

qp qp

q+r– 

un

p(q+r–)

qp

(g|un|qdx)

p+r–

qpλ

f|un|–rdx

b

q+r–

qp

p+r– 

q+r–  p+r–

qp qp

q+r– 

Aqq+–r–p

(a(p+r–)qA+pr+–b(p+r–)A)pq+–r–pa(qp

)Ap+b(qp)A

q+r– 

<bq+qr–p

p+r– 

q+r–  p+r–

qp qp

q+r– 

A

q+r–

qp

(b(qp++rr–)–A)pq+–r–pa(qp

)Ap+b(qp)A

q+r– 

= –a(qp

)

q+r– A

p< ,

which is impossible. Hence, (.) and (.) must hold.

Step: For nonnegative ϕW,p() andt>  small, we can findfn(t) :=fn(tϕ) such

thatfn() =  andfn(t)(un+tϕ)+for eachun+by Lemma ..fn+()∈[–∞,∞] is

(12)

fn+() > –Cfor allnN+. Sinceun,fn(t)(un+tϕ), we deduce that

 =a un p

+b un pλ

f|un|–rdx

g|un|qdx

and

 =afnp(t) un+ p

+bfnp(t) un+ pλfn–r(t)

f|un+|–rdx

fnq(t)

g|un+|qdx.

Thus

 =afnp(t) –  un+ p

+a un+ p

un p

+bfnp(t) –  un+ p+b

un+ pun p

λfn–r(t) – 

f|un+|–rdxλ

f|un+|–r–|un|–r

dx

fnq(t) – 

g|un+|qdx

g|un+|q–|un|q

dx

afnp(t) –  un+ p

+a un+ p

un p

+bfnp(t) –  un+ p+b

un+ pun p

λfn–r(t) – 

f|un+|–rdx

fnq(t) – 

g|un+|qdx.

Then, dividing byt>  and lettingt→, we have

≤apun p

fn+() +apun p

p

|∇un|p–∇unϕdx

+bp un pfn+() +bp

|∇un|p–∇unϕdx

λ( –r)fn+()

f|un|–rdxqfn+()

g|un|qdx

=fn+()

apun p

+bp un pλ( –r)

f|un|–rdxq

g|un|qdx

+apun p

p

|∇un|p–∇unϕdx+bp

|∇un|p–∇unϕdx

=fn+()

ap+r–  un p

+b(p+r– ) un p– (q+r– )

g|un|qdx

+apun p

p

|∇un|p–∇unϕdx+bp

|∇un|p–∇unϕdx.

One deduces from (.)

fn+()≥–ap

u

n p

p

|∇un|p–∇unϕdx+bp

|∇un|p–∇unϕdx

a(p+r– ) u

n p

+b(p+r– ) un p– (q+r– )

(13)

Therefore, by the boundedness of{un}, we conclude that{fn+()}is bounded from below. Step: Choosen∗large enough such that (–r)C

n < C

 for alln>n∗. Then we claim that

there existsCsuch thatfn+() <Cfor eachn>n∗. Without loss of generality, we may

supposefn+()≥. Then from condition (ii), we have fn(t) – 

un

n +tfn(t)

ϕ

n

≥ 

nfn(t)(un+tϕ) –un

J(un) –J

fn(t)(un+tϕ)

=a(p

+r– )

p( –r)

fnp(t) –  un+ p

+a(p

+r– )

p( –r)

un+ p

un p

+b(p+r– )

p( –r)

fnp(t) –  un+ p+

b(p+r– )

p( –r)

un+ pun p

q+r– 

q( –r)

fnq(t) – 

g|un|qdx

q+r– 

q( –r)f

q n(t)

g|un+|q–|un|q

dx.

Then, dividing byt>  and lettingt→, we deduce

fn+() un

n +

ϕ

n

a(p+r– )  –r f

n+() un p

+a(p

+r– )

 –r un

p–p

|∇un|p–∇unϕdx

+b(p+r– )  –r f

n+() un p+

b(p+r– )  –r

|∇un|p–∇unϕdx

q+r–   –r f

n+()

g|un|qdx

q+r–   –r

g|un|q–ϕdx. (.)

From (.) and the choice ofn∗, we have

ϕ

nC

( –r)f

n+() +

a(p+r– )

 –r un

p–p

|∇un|p–∇unϕdx

+b(p+r– )  –r

|∇un|p–∇unϕdx

q+r–   –r

g|un|q–ϕdx.

Namely,

C

( –r)f

n+()≤

ϕ

n

a(p+r– )

 –r un

p–p

|∇un|p–∇unϕdx

b(p+r– )  –r

|∇un|p–∇unϕdx+

q+r–   –r

g|un|q–ϕdx.

Therefore, by the boundedness of{un}, we conclude{fn+()}n>n∗is bounded from above.

Step:u+

λ>  a.e. inand for nonnegativeϕW

,p

 (), we have

au+λp–p+b

u+λp–∇u+λϕdxλ

fu+λrϕdx

gu+λq–ϕdx

(14)

Similar to the argument in Step , one can obtain

fn+() un

n +

ϕ

n

≥–fn+()

a un p

+b un p

g|un|qdxλ

f|un|–rϕdx

a un p

p

|∇un|p–∇unϕdxb

|∇un|p–∇unϕdx

+

g|un|q–ϕdx+ lim t→+

λ  –r

f(|un+|–r–|un|–r)

t dx

= –a un p

p

|∇un|p–∇unϕdxb

|∇un|p–∇unϕdx

+

g|un|q–ϕdx+ lim t→+

λ  –r

f(|un+|–r–|un|–r)

t dx. (.)

Sincef(|un+|–r–|un|–r)≥,∀t> , by Fatou’s lemma, we obtain

f|un|–rϕdx≤ lim t→+

  –r

f(|un+|–r–|un|–r)

t dx. (.)

It follows from (.) and (.) that

λ

f|un|–rϕdx

≤ 

n

fn+() un + ϕ

+a un p

p

|∇un|p–∇unϕdx

+b

|∇un|p–∇unϕdx

g|un|q–ϕdx

C·max{C,C}+ ϕ

n +a un

p–p

|∇un|p–∇unϕdx

+b

|∇un|p–∇unϕdx

g|un|q–ϕdx,

for alln>n∗.

Passing to the limit asn→ ∞, one has

lim

n→∞λ

f|un|–rϕdxa lim n→∞ un

p–p

u+λ

p–

u+λϕdx

+b

u+λp–∇u+λϕdx

gu+λq–ϕdx.

Then using Fatou’s lemma again, we infer that

λ

fu+λ

r

ϕdx

a lim

n→∞ un

p–p

u+

λ

p–

u+λϕdx

+b

u+λ

p–

u+λϕdx

gu+λ

q–

(15)

Sinceunu+λa.e. in, we getu+λ≥ a.e. in. Thus, one infers from (.) that

λ

fu+λ

–r

dxa lim

n→∞

un p

p

u+λ

p

+bu+λ

p

gu+λ

q

dx. (.)

On the other hand

a lim

n→∞

un p

p

u+λp+bu+λpa lim

n→∞ un

p

+b lim

n→∞ un

p

=λ

fu+λ

–r

dx+

gu+λ

q

dx. (.)

Combining (.) and (.), we have

lim

n→∞

un p= lim n→∞ un

p=u+

λ

p

. (.)

Thus, (.) can be obtained by inserting (.) into (.). Moreover, from (.), one has

u+λ

p–

u+λϕdx≥, ∀ϕW

,p

 (),ϕ≥.

Therefore, using the strong maximum principle for weak solutions (see []), we obtain

u+

λ>  a.e. in.

Step:u+λis a weak solution of (.), andu+λ+. By (.), we haveunu+λstrongly inW,p(), and so+∈+. AssumeφW

,p

 () andε> , defineW ,p

 () by:=

(u+λ+εφ)+. Then from Step  it follows

≤

au+λ

p–p

+bu+λ

p–

u+λλfu+λ

r

gu+λ

q–

dx

=

[u+λ+εφ>]

au+λp–p+buλ+p–∇u+λu+λ+εφ

λfu+λ

r

u+λ+εφ

gu+λ

q– u+λ+εφ

dx =

[u+λ+εφ≤]

au+λ

p–p

+bu+λ

p– u+λ

u+λ+εφ

λfu+λ

r

u+λ+εφ

gu+λ

q– u+λ+εφ

dx

=au+λp+bu+λpλ

fu+λ–rdx

gu+λqdx

+ε

au+λ

pp

+bu+λ

p–

u+λφλfu+λ

r

φgu+λ

q–

φdx

[u+λ+εφ≤]

au+λp–p+buλ+p–∇u+λu+λ+εφ

λfu+λ

r

u+λ+εφ

gu+λ

q– u+λ+εφ

dx =ε

au+λ

pp

+bu+λ

p–

u+λφλfu+λ

r

φgu+λ

q–

φdx

[u+

λ+εφ≤]

au+λ

p–p

+bu+λ

p–

u+λ

(16)

λfu+λ

r

u+λ+εφ

gu+λ

q– u+λ+εφ

dx

ε

au+λ

pp

+bu+λ

p–

u+λφλfu+λ

r

φgu+λ

q–

φdx

εau+λ

p–p

+b

[u+λ+εφ≤]

u+λ

p–

u+λφdx.

Since the measure of the domain of integration [u+λ+εφ≤] tends to zero asε→, it follows[u+

λ+εφ≤]|∇u

+

λ|p–∇u+λφdx→. Dividing byεand lettingε→, we have

au+λ

pp

+b

u+λ

p–

u+λφdxλ

fu+λ

r

φdx

gu+λ

q–

φdx≥.

Notice thatφis arbitrary, the inequality also holds for –φ, so it follows thatu+

λis a weak solution of (.). Moreover, from (.) and (.), we deduce thatu+

λ+. A similar argument shows that there exists another solutionuλ–.

4 Proof of Theorem 1.2

(i) By Lemma .(i), we write=+and define

θ+= inf

uNλ+J(u).

Similar to Lemma .(i), we haveθ+< . Applying Lemma .(i) and the Ekeland varia-tional principle, we see that there exists a minimizing sequence{un}forJ(u) in+such that

(i) J(un) <θ++

n;

(ii) J(u) >J(un) –

n uun , ∀uN +

λ.

Repeating the same argument as Theorem ., we can see that+is a positive solution of the problem (.).

(ii) Similar to the proof of Theorem ., we know that the problem (.) has at least two positive solutionsu+

λ+anduλ–. Moreover, combining (.) with (.), we have

lim

a–

uλ=∞

and

lim

a–

inf

uNλJ(u) =∞.

This completes the proof of Theorem ..

(17)

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

All authors contributed equally and significantly in writing this paper. All authors read and approved the final manuscript.

Acknowledgements

This work was supported by Young Award of Shandong Province (ZR2013AQ008), NNSF (61603226), the Fund of Science and Technology Plan of Shandong Province (2014GGH201010) and NSFC (11671237).

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Received: 11 December 2016 Accepted: 9 March 2017

References

1. Kirchhoff, G: Mechanik. Teubner, Leipzig (1883)

2. Alves, CO, Corrêa, FJSA: On existence of solutions for a class of the problem involving a nonlinear operator. Commun. Appl. Nonlinear Anal.8, 43-56 (2001)

3. Chipot, M, Lovat, B: Some remarks on nonlocal elliptic and parabolic problems. Nonlinear Anal.30, 4619-4627 (1997) 4. Liu, DC, Zhao, PH: Multiple nontrivial solutions to ap-Kirchhoff equation. Nonlinear Anal.75, 5032-5038 (2012) 5. Li, YX, Mei, M, Zhang, KJ: Existence of multiple nontrivial solutions for ap-Kirchhoff type elliptic problem involving

sign-changing weight functions. Discrete Contin. Dyn. Syst., Ser. B21, 883-908 (2016)

6. Chen, CY, Kuo, YC, Wu, TF: The Nehari manifold for a Kirchhoff type problem involving sign-changing weight functions. J. Differ. Equ.250, 1876-1908 (2011)

7. Alves, CO, Corrêa, FJSA, Figueiredo, GM: On a class of nonlocal elliptic problems with critical growth. Differ. Equ. Appl.

2, 409-417 (2010)

8. Cheng, BT, Wu, X: Existence results of positive solutions of Kirchhoff type problems. Nonlinear Anal.71, 4883-4892 (2009)

9. Ma, TF, Munoz Rivera, JE: Positive solutions for a nonlinear nonlocal elliptic transmission problem. Appl. Math. Lett.16, 243-248 (2003)

10. Perera, K, Zhang, Z: Nontrivial solutions of Kirchhoff-type problems via the Yang index. J. Differ. Equ.221, 246-255 (2006)

11. Liu, X, Sun, YJ: Multiple positive solutions for Kirchhoff type problems with singularity. Commun. Pure Appl. Anal.12, 721-733 (2013)

12. Liao, JF, Zhang, P, Liu, J, Tang, CL: Existence and multiplicity of positive solutions for a class of Kirchhoff type problems with singularity. J. Math. Anal. Appl.430, 1124-1148 (2015)

13. Lei, CY, Liao, JF, Tang, CL: Multiple positive solutions for Kirchhoff type of problems with singularity and critical exponents. J. Math. Anal. Appl.421, 521-538 (2015)

14. Brown, KJ, Wu, TF: A fibering map approach to a semilinear elliptic boundary value problem. Electron. J. Differ. Equ.

200769, (2007)

15. Brown, KJ, Wu, TF: A fibering map approach to a potential operator equation and its applications. Differ. Integral Equ.

22, 1097-1114 (2009)

16. Zhang, Z, Perera, K: Sign changing solutions of Kirchhoff type problems via invariant sets of descent flow. J. Math. Anal. Appl.317, 456-463 (2006)

17. Ekeland, I: On the variational principle. J. Math. Anal. Appl.47, 324-353 (1974)

References

Related documents

The purpose of this study was to analyze the optimal dose of vermicompost fertilizer to produce high quality of agar rendement, viscosity and gel strength

Clearly, each of the principles of our approach to develop- ing and designing the architecture of complex systems — automated detection of goal compliance, verification against

In [9], consistency of the interval valued fuzzy relational equations and the complete set of solutions of xA = b where A is an interval valued fuzzy matrix and b is

The absorption edge shifts towards the higher wavelength with the increase in film thickness.The nonlinear absorption coefficient is inversely proportional to the thickness

The paper has presented some unexplored design paradigms of software evaluation while designing the embedded systems well particular like polar satellite antenna

Given the increased volatility in the price of energy, especially crude oil, and strong evidence from previous studies that the impact of oil price shocks differs between

 Public relations is “a management function supplying maintenance of relations based on common benefits between an institution and the target communities” (Pohl,1995:1)..

In our data compression based test method, we change the scan-cell architecture to reduce power consumption by both decreasing the number of CUT’s switching activities and