- About this Journal ·
- Abstracting and Indexing ·
- Aims and Scope ·
- Annual Issues ·
- Article Processing Charges ·
- Articles in Press ·
- Author Guidelines ·
- Bibliographic Information ·
- Citations to this Journal ·
- Contact Information ·
- Editorial Board ·
- Editorial Workflow ·
- Free eTOC Alerts ·
- Publication Ethics ·
- Reviewers Acknowledgment ·
- Submit a Manuscript ·
- Subscription Information ·
- Table of Contents

Abstract and Applied Analysis

Volume 2014 (2014), Article ID 502756, 9 pages

http://dx.doi.org/10.1155/2014/502756

## Three Solutions Theorem for -Laplacian Problems with a Singular Weight and Its Application

^{1}Department of Mathematics, Pusan National University, Busan 609-735, Republic of Korea^{2}Department of Mathematics, University of Alabama at Birmingham, Birmingham, AL 35294-1170, USA^{3}Department of Mathematics Education, Pusan National University, Busan 609-735, Republic of Korea

Received 3 September 2013; Revised 18 December 2013; Accepted 1 January 2014; Published 18 February 2014

Academic Editor: Adem Kılıçman

Copyright © 2014 Yong-Hoon Lee et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

#### Abstract

We prove Amann type three solutions theorem for one dimensional
*p*-Laplacian problems with a singular weight function. To prove this theorem, we
define a strong upper and lower solutions and compute the Leray-Schauder degree
on a newly established weighted solution space. As an application, we consider the
combustion model and show the existence of three positive radial solutions on an
exterior domain.

#### 1. Introduction

Let us consider the following -Laplacian problem with a sign-changing singular weight: where ,, , and may change sign. Moreover, and satisfies , for for some , where a class of weight functions is given as It is well known that .

If , then all solutions of are in and based on the Leray-Schauder degree argument on -space: Ben-Naoum and de Coster [1] proved three solutions theorem for . On the other hand, if , then solutions of may not be in ; for example, take , , and , then and the solution for corresponding problem of is given by which is not in . Thus if , the three solutions theorem in [1] can not be applied. Our main interest in this paper is to establish three solutions theorem for problem for those weights satisfying .

Main step for the proof of three solutions theorem, in general, is to compute the Leray-Schauder degree on a sector in solution space made from strong sense of upper and lower solutions. Since the sector needs to be open in the space, strong sense of ordering and the sector made from the ordering are closely related to the topology of solution space. A typical situation in application usually happens as follows.

It is comparatively easy to find a lower solution and an upper solution of satisfying , for all and . Denote , . In -analysis, that is, , we see that is nonempty and open in by providing additional conditions like and which implies a strong sense of ordering. On the other hand, in -analysis, that is, , we see so that the Leray-Schauder degree on is not even defined. Three solutions theorem with no use of such is very restrictive in application. To overcome this difficulty in our problem, we introduce a weighted solution space, say , which is finer than and also introduce a strong sense of ordering suitable to -space which makes the degree computation effective (see Section 2 in detail).

As to a question of triple multiplicity of solutions, besides the Amann type three solutions theorems, many works are done by using the variational method (see [2–5] and the references therein) and by using several extensions of the Liggett-Williams fixed point theorem and Guo-Krasnoselskii fixed point theorem, especially for positive solutions (see [6–8] and the references therein). For the problem we are concerned with in this paper, the variational setup is not possible due to lack of regularity of solutions. Three solutions theorem proved in this paper is for the case that is not only but also possibly sign-changing. By this aspect, it is new as far as the authors know. For the case , one may refer to [9] for a partial result about the theorem.

As an application, we study the existence of triple positive solutions for certain nonlinear -Laplacian problems with positive singular coefficient function and give an example of a combustion model defined on an exterior domain. Applying this three solutions theorem to the case having sign-changing coefficient function could be an interesting and challenging problem.

We organize this paper as follows. In Section 1, we introduce a weighted solution space , the strong upper and lower solutions of , and prove three solutions theorem for problem . In Section 3, we prove a multiplicity result for certain nonlinear -Laplacian problems by using three solutions theorem introduced in Section 2. In Section 4, we apply the result in Section 3 to a combustion model to show the existence of three positive radial solutions on exterior domain.

#### 2. Preliminaries

In this section, we introduce a weighted solution space and prove three solutions theorem for on . Let where with If , then , for , and is integrable on . More precisely, if , then and if , then .

For , define by And also define ; then is a Banach space. We give a proof for reader's convenience.

Lemma 1. *Let ; then is a Banach space with a norm .*

*Proof. *Let be a Cauchy sequence in . Then and are Cauchy sequences in so that there exist such that and in . It is sufficient to show that on . Since for , there exists such that for all . For , we know in . This implies in . Since is arbitrary, pointwise in . Therefore, on . Since converges uniformly to on , we have
Thus and on . This implies in and the proof is completed.

Define

Then we see that for given , being a solution of implies . In fact, if and is a solution of , then for , and by using L’Hospital’s rule, we have For the cases that and (or ), by the same argument, we have

*Example 2. *Let us consider the following example:
where , and is given by
Then we see that sign-changing and every solution satisfies by using (9). We also see by calculation that can be given as
and by using (10).

To establish corresponding integral operator for problem , let us first consider the problem where . We remind the reader that needs not be integrable near or . Integrating on for , we have Denoting , Since and is a fixed constant, we can see that so that we may integrate on . Using a boundary condition , we get We note that in (18) is a solution of only on the interval . Doing similar computation on the interval , we get It is known by Lemma 2.2 in [10] that the equation has a unique zero in for each . Therefore it is natural to paste ’s in (18) and (19) in a continuous way. Now let us define a function by Then satisfies and is a unique solution of problem .

Based on this setup, we now introduce corresponding integral operator for problem . For , define Then in if and only if is a solution of .

*Remark 3. *We understand the number in the above as a function of defined on . That is, . It is known that maps bounded sets in into bounded sets in ([10, Lemma 3.1]). It is also known that is completely continuous on ([10, Theorem 3.4]).

As mentioned in Introduction, the regularity of solutions of problem sensitively depends on the shape of nonlinear term even if , and we are concerned with the case that problem does not have -solutions. Therefore, it is interesting to consider operator restricted on to complete three solutions theorem for those problems with no -solutions.

Define the restriction of on .

In what is to follow, we assume and we now prove the complete continuity of on the solution space . Before doing that, we give a remark useful to calculate -Laplacians.

*Remark 4. *If , then
where

Theorem 5. * is completely continuous.*

*Proof. *Let be a bounded subset of . Then for any sequence , we need to show the relative compactness of with respect to -norm. We know by Remark 3 that is completely continuous on so that there exists and a subsequence of , say again such that in . To complete the proof, we need to show the following.

and there is a subsequence of such that as in and is continuous on .*Claim 1*. is uniform bounded in .

Since is bounded in , there exists such that and , for all . We also know by Remark 3 that there is such that , for all . For , by using Remark 4, we get
where . From the fact that and the definition of , we see
thus we have
for . For , by similar calculation, we get the same upper bound of as in (27) and by taking limt and in (27), we have the same upper bound of and as in (27). This proves that is bounded in .*Claim 2*. is equicontinuous on .

If , then since , for all and , we get
for all . This implies that is equicontinuous in and by Arzela-Ascoli theorem, there exist a subsequence of and such that converges uniformly to on as . Thus using Lebesgue Dominated Convergence theorem, we obtain
uniformly on . This implies that is equicontinuous in .

On the other hand, for the case of , suppose that is not equicontinuous on . Then there exists such that we may choose a subsequence of and sequences , satisfying
As for sequences and , it is easy to see that . We show that or . Suppose it is not true so let . Then taking satisfying , we see that and uniformly on . By the same argument of the above case of , we can prove that is equicontinuous on . Thus there is sufficiently large such that
and this contradicts with (30). Now we consider the case . The argument for the case is similar. It is easy to see that this case implies . Since , , for all and we get
Now we want to calculate .

Since is bounded and , we have
On the other hand,
We show .

Indeed, is close to for sufficiently large , since and . Therefore, without loss of generality, we may assume that so that . Thus using the fact in , we have
Next we show
Indeed, using the fact for sufficiently large , we get
If , then we can easily verify . Since , we see that the limit is . On the other hand, if . We note that , for given . By using L’Hospital’s rule, we get the conclusion.

Therefore we get
By the same argument, also has the same limit and this contradicts with (30). Consequently, is equicontinuous in . By Arzela-Ascoli theorem, there exists a subsequence of and such that
*Claim 3*. and in .

It is enough to show that . Since for , let ; then and for , uniformly in and thus on . Since is arbitrary, on and from (39), we have
Therefore on . By similar argument near , we get on and thus .*Claim 4*. is continuous on .

Assume that in . By compactness of , there is a subsequence of and such that in . It is suffice to see that , for all . Since is continuous in and in , we have in . Thus and by standard limit argument, we see that . This completes the proof.

Now we define a strong sense of ordering in .

*Definition 6. *For , one says that if and only if(i) for all ,(ii)either or ,(iii)either or .

*Definition 7. *One says that is a lower solution of if and only if , and
We also say that is an upper solution of if and only if with and it satisfies the reverse of the above inequalities.

*Definition 8. *One says that is a strict lower solution of if and only if is a lower solution of and satisfies where is a solution of such that , .

We say that is a strict upper solution of if and only if is an upper solution of and satisfies where is a solution of such that , .

Theorem 9. *Assume that there exist a strict lower solution and a strict upper solution of such that . Then problem has at least one solution such that . Moreover, for large enough, the Leray-Schauder degree can be computed as
**
where .*

*Proof. *Consider the modified problem
where is defined by
Then it is well known that if is a solution of , then and thus is solution of . Define by . Then is bounded and thus there exists such that for all . By the homotopy invariance property of degree, we have
where . Thus has a solution and has a solution satisfying . Since and are strict lower and upper solutions, respectively, by the definition of strict lower and upper solution, we have . Moreover, by using the fact that on , (44) and excision property, we conclude that there exists large enough such that

Theorem 10 (three solutions theorem). *Assume that there exist a lower solution , an upper solution , a strict lower solution , and a strict upper solution of such that
**
and there exists with . Then problem has at least three solutions , and such that
*

*Proof. *Consider the modified problem,
where is defined by
For any , and are strict lower solution and strict upper solution of . In fact, if is a solution of , then we have . By Theorem 5, there is a sufficient large :
where is defined by and
Then by excision and additive property, we have
Thus there are three solutions of , , , and . Since all solutions of satisfy , they are solutions of and the proof is done.

#### 3. Application

In this section, we prove the existence of triple positive solutions for a problem of the form where and . Let us assume with ,, , is nondecreasing.

The existence of two positive solutions for problem was proved in [11] under a stronger condition on such as for some . Since , we can easily see that any solution of problem is in but not in so that with the aid of three solutions theorem in Section 2, we prove the following theorem.

Theorem 11. *Assume , , and , and also assume that there exist and such that and
**
where and the unique solution of
**
Then for satisfying
**
problem has at least three positive solutions.*

*Proof. *For , , it is trivial that is a lower solution of . Let . Then
Thus is an upper solution of . To show that is a strict upper solution of , assume that is a solution of such that . We first show that , for all . Suppose it is not true, then there exist and with in such that and . Integrate from to , by (55) and monotonicity of , we have the following contradiction:
Since , it suffices to show that . The inequality can be proved similarly. For the case that , we know and exist. Since
we know that there exists such that . Indeed, otherwise, for all ; then by integrating this from to 1, we have the contradiction
Integrating (57) from to 1, we have
and thus and
For the case that ,
From the monotonicity of and choice of with help of L’Hospital’s rule, we have
Thus we proved that is a strict upper solution of . Now, since , we may choose satisfying ; then since , we may choose such that . Let be the solution of
where when
and , for . We note that and let us show that for . It is clear that for . We note that from the symmetry of . For , by integrating (63) from to , from the choice of and , we have
Thus for and it is clear that for , by the symmetry of and . From the monotonicity of , we have
This implies that is a lower solution of and by using the similar argument as of , we can show that for all solution of such that . Now to show that is a strict lower solution of , we need to show that and . For the case that , can be proved by similar argument as of . Now, for the case of , from (65) and = , we have = = and
Similarly, we can prove and thus for all solution of such that . This implies that is a strict lower solution of . Since , there exists such that . Define
Then from , there exists sufficiently large such that
and , . Thus we have
This implies that is an upper solution of and the proof is complete by three solutions theorem.

#### 4. Example

As an example, let us consider the following combustion model defined on an exterior domain: where , , , and . Moreover . For the radial solutions of , by changes of variables, , , can be transformed into with Let us define then it is easy to check and if , then corresponding in (71) satisfies .

As an example of , take for ; then . Moreover, in (71) can be calculated as , for some and given as and we see that , when .

Finally, we have a multiplicity result of positive solutions for combustion model .

Corollary 12. *Assume . If is sufficiently large, then has at least three positive radial solutions for , where
*

*Proof. *Let and , . Then since is nondecreasing and
as , all hypotheses of Theorem 11 are satisfied for sufficiently large