Abstract
We consider the following nonlinear Kirchhoff type problem of the form
where \(\Omega \subset \mathbb {R}^{3}\) is a bounded domain with smooth boundary \(\partial \Omega \) and \(a>0\), \(b\ge 0 \). The nonlinearity \(\mu g(x,u)+f(x,u)\) may involve a combination of concave and convex terms. Under some suitable conditions on \(f,g\in C(\overline{\Omega }\times \mathbb {R},\mathbb {R})\) and \(\mu \in \mathbb {R}\), we prove the existence of infinitely many high-energy solutions using Fountain theorem. In particular, using the method of invariant sets of descending flow, we prove the existence of at least one sign-changing solutions.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction and main results
In this paper, we investigate the existence of high-energy solutions and sign-changing solutions to the following Kirchhoff type elliptic equation
where \(\Omega \subset \mathbb {R}^{3}\) is a bounded domain with smooth boundary \(\partial \Omega \) and \(a>0\), \(b\ge 0\). The nonlinearity \(\mu g(x,u)+f(x,u)\) may involve a combination of concave and convex terms. When \(a\equiv 1\), and \(b\equiv \mu \equiv 0\), the problem (1.1) turns out to be the following elliptic equation:
Note that Kirchhoff type problem on a smooth-bounded domain \(\Omega \subset \mathbb {R}^{3}\) takes the form
and has been studied extensively. Indeed, such a class of problems is viewed as being nonlocal because of the presence of the term \((\int _{\Omega }|\nabla u|^{2})\triangle u\), which implies that the equation in (1.3) is no longer a pointwise identity and is very different from classical elliptic equations. That is to say, such a phenomenon provokes some mathematical difficulties, which makes the study of such a class of problems particularly interesting. On the other hand, problem (1.3) has its physical motivation. Moreover, the equation of (1.3) related to the following stationary analogue equation:
where u denotes the displacement, f(x, u) the external force and b the initial tension while a is related to the intrinsic properties of the string (such as Young’s modulus). Equation of type (1.4) was first proposed by Kirchhoff [15] and began to attract researchers’ attention mainly after the work of Lions [16]. Some interesting results can be found, for example, in [1, 5, 7, 9, 10, 14, 21, 22, 24,25,26,27]. Similar nonlocal problems also appear in several fields as biological systems (for example, population density, see [2, 8]).
In the last two decades, there are a large number of papers devoted to problems like (1.3) on the existence of positive solutions, radial and non-radial solutions, ground states, and sign-changing (see, e.g., [2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23]). Recently, in almost all the results concerning equation (1.3), the nonlinear term f is assumed to be superlinear or sublinear, and little has been done for the combination of concave and convex terms. It is the first purpose of our paper to investigate the Kirchhoff type problems with concave and convex nonlinearities.
Another topic which has increasingly received interest in recent years is the existence of sign-changing solutions of problems like (1.3). Mao and Zhang [23] considered the existence of sign-changing and multiple solutions of problems like (1.3) without the (PS) condition. Deng et al. [11] studied the following Kirchhoff problem:
and they proved that, for any positive integer k, the problem has a sign-changing solution \(u^{b}_{k}\). In [13], Figueiredo et al. studied the existence of a nodal solution with minimal energy of (1.3). Using constrained minimization on the sign-changing Nehari manifold, Ye in [31] proved the following nonlinear Kirchhoff equation:
has a least energy nodal solution. In [28] W. S studied the existence of least energy sign-changing solutions for a class of Kirchhoff-type problem in bounded domains using quantitative deformation lemma and degree theory.
Motivated by and different from above-mentioned papers, we expect to use the method of invariant sets of descending flow to prove the existence of sign-changing solutions. Mao and Luan [24] obtained existence of signed and sign-changing solutions with asymptotically three-linear-bounded nonlinearity via variational methods and invariant sets of descent flow. Batkam [4] established a new sign-changing version of the symmetric Mountain Pass theorem and then applied it to prove the existence of a sequence of sign-changing solutions to (1.3) with higher energy. Note that the nonlinear term f in [4] is superlinear. Different from the works in the literature, the second aim of our paper is to study the problem (1.1) with more general nonlinearity involving concave and convex nonlinearities.
Before stating our main results, we introduce some conditions for \(f,g\in C(\overline{\Omega }\times \mathbb {R},\mathbb {R})\).
\((g_{1})\) There exist constants \(1<q_{1}<q_{2}<\cdots<q_{m}<2\) and functions \(h_{i}(x)\in L^{\frac{2}{2-q_{i}}}(\overline{\Omega },\mathbb {R}^{+})(i=1,\ldots ,m)\) such that
\((g_{2})\) There exists \(\nu \in (1,2)\) such that
where \(G(x,u)=\int ^{u}_{0}g(x,t)\mathrm{d}t\).
\((f_{1})\) There exist \(C_{1}>0\) and \(2<p<6\) such that
\((f_{2})\) \(\lim _{|u|\rightarrow 0}\frac{f(x,u)}{u}=0 \) uniformly in \(\Omega \).
\((f_{3})\) There exists \(\theta >4\) such that
where \(F(x,u)=\int ^{u}_{0}f(x,t)\mathrm{d}t\).
\((f_{4})\) There exists \( R>0\) such that
Our main results read as follows.
Theorem 1.1
Let \(a>0\), \(b\ge 0\). Assume \((g_{1}),(g_{2})\) and \((f_{1})-(f_{4})\) hold. Then there exists \(\tilde{\mu }>0\) such that for \( \mu \in [-\tilde{\mu },\tilde{\mu }]\), problem (1.1) has at least one nontrivial solutions.
Theorem 1.2
Let \(a>0\), \(b\ge 0\). Assume \((g_{1}),(g_{2})\) and \((f_{1}){\text {--}}(f_{4})\) hold and f, g is odd in u. Then there exists \(\tilde{\mu }>0\) such that for \( \mu \in [-\tilde{\mu },\tilde{\mu }]\), problem (1.1) has a sequence of solutions \((u_{k})\) with \(I_{\mu }(u_{k})\rightarrow +\infty \) as \(k\rightarrow \infty \).
Remark 1.3
In general, one uses the Mountain Pass theorem to get the \((PS)_{c}\) sequences. However, it is difficult to obtain a \((PS)_{c}\) sequence by this method in the case of sublinear. Moreover, it is hard in the situation of concave and convex nonlinearities. Thus, we adopt a new technique to prove the existence of a Mountain Pass geometry in the case of concave and convex nonlinearities.
The combined effect of concave and convex nonlinearities was first investigated by Ambrosetti et al. [3] on the following elliptic equation:
with \(\Omega \subset \mathbb {R}^{N}\) a bounded domain, \(1<q<2<p<2^{*}\), \(2^{*}=\frac{2N}{N-2}\). They proved the existence of infinitely many solutions with negative energy for \(0<\mu \ll \lambda =1\). Bartsch and Willem extended the results in [29], they proved the existence of infinitely many solutions with high energy for \(\lambda >0,\mu \in \mathbb {R}\) and negative energy for \(\mu >0,\lambda \in \mathbb {R}\) using Fountain theorem and Dual Fountain theorem, respectively. Li et al. also considered an elliptic equation with concave and convex nonlinearities in [19]. In [12], Ding and Liu studied multiplicity of periodic solutions of a Dirac equation with concave and convex nonlinearities.
Theorem 1.4
Let \(a>0\), \(b\ge 0\). Assume \((g_{1}),(g_{2})\) and \((f_{1}){\text {--}}(f_{4})\) hold. Then there exists \(\mu _{0}>0\) such that for \(\mu \in (0,\mu _{0}]\), problem (1.1) has at least one sign-changing solutions.
Remark 1.5
The class of nonlinearities \(\mu g(x,u)+f(x,u)\) satisfying the assumptions of Theorem 1.4 includes the nonlinearity \(\mu |u|^{q-2}u+|u|^{p-2}u\) with \(1<q<2\), \(4<p<6\). Even in this special case, Theorem 1.4 seems to be the first attempt in finding sign-changing solutions to (1.1).
The method of invariant sets of descending flow plays an important role in the study of sign-changing solutions of elliptic problems. We refer to [17, 23, 32] and the references therein. Since the pseudo-gradient flow must be constructed in such a way that keeps the positive and negative cones invariant. However, this invariance property makes the construction of the flow very complicated because of the nonlocal term \((\int _{\Omega }|\nabla u|^{2})\triangle u\). Thus, to construct the pseudo-gradient flows using an auxiliary operator is more important.
The idea of the proofs of Theorems 1.4 is to use suitable minimax arguments in the presence of invariant sets of a descending flow for the variational formulation. In particular, we make use of an abstract critical point theory developed by Wang et al. [17] on the nonlinear Schrödinger systems and by Liu et al. [18] on the Schrödinger–Poisson system
where the pseudo-gradient flows were constructed using an auxiliary operator.
This paper is organized as follows. In Sect. 2, we state the variational framework of our problem and some preliminary setting. Section 3 is devoted to the proof of Theorems 1.1 and 1.2. Finally, the proof of Theorem 1.4 is presented in Sect. 4.
2 Preliminaries and functional setting
Throughout this paper we denote by \(\rightarrow \) (resp.\(\rightharpoonup \)) the strong (resp.weak) convergence. We use \(C_i\) to denote various positive constants which may vary from lines to lines and are not essential to the problem.
We make use of the following notations, let \(H:=H_{0}^{1}(\Omega )\) with the norm and the inner product
As usual, for \(1\le p<+\infty \), \( |u|_{p}:=\left( \int _{\Omega }|u|^{p}\right) ^{\frac{1}{p}}, \ \ u\in L^{p}(\Omega ).\)
Lemma 2.1
[29, Theorem 1.9]. If \(|\Omega |<\infty \), the following embeddings are compact:
and so there exists \(\eta _{s}\) such that
Next we define the energy functional \(I_{\mu }\) on H by
under the assumptions\((g_{1})\) and \((f_{1})\), \(I_{\mu }\in C^{1}(H,\mathbb {R})\) and
It is easy to verify that \(u\in H \) is a solution of system (1.1) if and only if \(u\in H\) is a critical point of \(I_{\mu }\).
Let X be a Banach space with the norm \(\Vert \cdot \Vert \) and \(X=\overline{\bigoplus ^{\infty }_{i=1}X_{i}}\) with \(\dim X_{i}<+\infty \) for each \(i\in \mathbb {N}\). Further, we set
Definition 2.2
[29]. Let X be a Banach space, \(I\in C^{1}(X,\mathbb {R})\), and \(c\in \mathbb {R}\). The function I satisfies the \((PS)_{c}\) condition if any sequence \(\{u_{n}\}\subset X\) such that
has a convergent subsequence.
Lemma 2.3
(Mountain Pass theorem [29]). Let X be a Banach space, \(I\in C^{1}(X,\mathbb {R})\), \(e\in X\) and \(r>0\) such that \(\Vert e\Vert >r\) and
If I satisfies the \((PS)_{c}\) condition with
then c is a critical value of I.
Lemma 2.4
[29]. If \(1\le p<2^{*}\) then we have that
To give the proof of our main result, we need the following critical point theorem.
Lemma 2.5
(Fountain theorem, Bartsch, 1992 [29]). Let \(I\in C^{1}(X,\mathbb {R})\) satisfy \(I(-u)=I(u)\). Assume that, for every \(k\in \mathbb {N}\), there exists \(\rho _{k}>\gamma _{k}>0\) such that
\((A_{1})\ \) \(a_{k}:=\max _{u\in Y_{k},\Vert u\Vert =\rho _{k}} I(u)\le 0\),
\((A_{2})\ \) \(b_{k}:=\inf _{u\in Z_{k},\Vert u\Vert =\gamma _{k}} I(u)\rightarrow +\infty \) as \( k\rightarrow +\infty \),
\((A_{3})\ \) I satisfies the \((PS)_{c}\) condition for every \(c>0\),
then I has an unbounded sequence of critical values.
3 Proof of Theorems 1.1 and 1.2
3.1 Proof of Theorem 1.1
Lemma 3.1
If \((g_{1})\), \((f_{1})-(f_{4})\) hold, then there exist \(e\in H\) and \(r>0\) such that \(\inf _{\Vert u\Vert =r}I_{\mu }(u)>0\) and \(I_{\mu }(e)<0\) with \(\Vert e\Vert >r\).
Proof
Since \(1<q_{1}<q_{2}<\cdots<q_{m}<2\), \((g_{1})\) together with (2.1) and Hölder’s inequality imply that, for \(\Vert u\Vert \) large enough,
\((f_{1})\) and \((f_{2})\) imply that for any \(\delta >0 \) there exists \( C_\delta >0\) such that
and then
It follows from (3.1) and (3.3) that
For \(\delta \le \frac{2a-1}{2\eta ^{2}_{2}}\) satisfying \(\frac{a}{2}-\frac{\delta }{2}\eta ^{2}_{2}\ge \frac{1}{4}\), we have
Let
and then we get \(\lim _{s\rightarrow +\infty } k(s)=\lim _{s\rightarrow 0^{+}} k(s)=+\infty ,\) which implies that k(s) is bounded below, thus k(s) admits a minimizer \(s_{0}\):
From (3.5)
Let \(K= C_{3}\left( \frac{C_{3}p(2-q_{m})}{C_{\delta }\eta ^{p}_{p}(p-2)}\right) ^{\frac{q_{m}-2}{p-q_{m}}}+ \frac{C_{\delta }\eta ^{p}_{p}}{p}\left( \frac{C_{3}p(2-q_{m})}{C_{\delta }\eta ^{p}_{p}(p-2)} \right) ^{\frac{q_{m}-2}{p-q_{m}}}\), then \( k(s_{0})=|\mu |^{\frac{p-2}{p-q_{m}}}K\),
if
then
thus, there exists \(r:=s_{0}>0\) such that \(\inf _{\Vert u\Vert =r}I_{\mu }(u)>0=I(0)\).
After integrating, we obtain from \((f_{1}){\text {--}}(f_{4})\) the existence of \(C_{4}>0\) such that
For \(t>0\), note that \(\theta >4\), we have, for some \(u_{0}\in E\)
so, there exists \(e=t_{0}u_{0}\) such that \(\Vert e\Vert >r\) and \(I_{\mu }(e)< 0\).
Lemma 3.2
Let \(a>0\), \(b\ge 0\) and assume \((g_{1})\) hold. Set
then \( \psi ^{'}\) is weakly continuous and compact.
Proof
The proof is similar to Lemma 2.2 in [30], so we omit it.
Lemma 3.3
Assume \((g_{1}),(g_{2})\) and \((f_{1}){\text {--}}(f_{3})\) hold. Then for every \(\mu \in \mathbb {R}\), \(I_{\mu }\) satisfies \((PS)_{c}\) condition for every c.
Proof
Let \(\{u_{n}\}\subset H\) be such that \(I_{\mu }(u_{n})\rightarrow c\) and \(I_{\mu }^{'}(u_{n})\rightarrow 0\ \ \hbox {as}\ \ n\rightarrow \infty \). First, we verify the boundedness of \(\{u_{n}\}\). For n large enough, by \((g_{2})\) and \((f_{3})\) we have
By \((g_{1})\), Lemma 2.1 and Hölder’s inequality we can obtain
Thus,
where \( 1<q_{i}<2\), \(i=1,\ldots ,m,\) which gives a bound for \(\{u_{n}\}\).
Next, we prove that the sequence \(\{u_{n}\}\) has a convergent subsequence. Going if necessary to a subsequence, we can assume that \(u_{n}\rightharpoonup u\) in H and \(u_{n}\rightarrow u\) in \(L^{s}(\Omega )\), \(2\le s<6 \). Theorem A.2 in [29] implies that \(f(x,u_{n})\rightarrow f(x,u)\) in \(L^{q}(\Omega )\), \(q=\frac{p}{p-1}\). Observe that
It is clear that
Since \(\{u_{n}\}\) is bounded and \(u_{n}\rightharpoonup u\) in H,
By Lemma 3.2, we know that \( \psi ^{'}\) is compact, thus
It follows from the Hölder’s inequality that
By (3.8)–(3.11), we obtain that \(\Vert u_{n}-u\Vert \rightarrow 0\ \ \hbox {as}\ \ n\rightarrow \infty .\) \(\square \)
Proof of Theorem 1.1
By Lemmas 3.1, 3.3 and using Lemma 2.3, we complete the proof of Theorem 1.1. \(\square \)
3.2 Proof of Theorem 1.2
In this section, we prove the existence of infinitely many solutions with high energy to system (1.1).
Proof of Theorem 1.2
On \(Z_{k}\), by Lemma 2.3, we see that
which implies that
Thus, by (3.1), (3.3), Lemma 2.1 and (3.12) we have,
For \(\delta \le \frac{2a-1}{2\eta ^{2}_{2}}\) satisfying \(\frac{a}{2}-\frac{\delta }{2}\eta ^{2}_{2}\ge \frac{1}{4}\), we have
and there exists
such that
By Lemma 2.3 we know that \(\beta _{k}\rightarrow 0\ \hbox {as} \ \ k\rightarrow \infty \). Moreover, \(1<q_{m}<2\), \(2\le p<6\), thus \(b_{k}\rightarrow +\infty \ \ \hbox {as}\ \ k\rightarrow \infty \), \((A_{2})\) is proved.
On \(Y_{k}\), by (3.1) and (3.6) we have
Since on the finite dimensional space \(Y_{k}\) all norms are equivalent, thus there exists \(\rho _{k}>\gamma _{k}>0\) such that \(\Vert u\Vert =\rho _{k}\) and condition \((A_{1})\) is satisfied for \(\rho _{k}\) large enough.
As in the proof of Lemma 3.3 one sees that \(I_{\mu }\) satisfies \((PS)_{c}\) condition, i.e. \((A_{3})\) hold. Therefore, using Lemma 2.5 we obtain that system (1.1) has a sequence of of solutions \(u_{k}\) such that \(I_{\mu }(u_{k})\rightarrow \infty \) as \(k\rightarrow \infty \). \(\square \)
4 Proof of Theorem 1.3
In this section, we prove the existence of sign-changing solutions to problem (1.1).
4.1 Preliminaries of invariant subset of descending flow
To obtain sign-changing solutions, we make use of the positive and negative cones. Precisely, define
For \(\varepsilon >0\), we also define
where \(\hbox {dist}(u,P^{\pm })=\inf _{v\in P^{\pm }}\Vert u-v\Vert .\)
Let X be a Banach space, \(I\in C^{1}(X,R)\), \( P,Q\subset X\) be open sets. We denote by K the set of critical point of I that is \(K=\{u\in X:I^{'}(u)=0\}\) and \(E=X{\setminus } K\). For \(c \in R\), \(K_{c} = \{x \in X : I(x) = c, I^{'}(x) = 0\}\) and \(I_{c} = \{x \in X : I(x)\le c\}\).
Definition 4.1
[18]. \(\{P,Q\}\) is called an admissible family of invariant sets with respect to I at level c, provided that the following deformation property holds: if \(K_{c}{\setminus } (P\cap Q)=\emptyset \), then there exists \(\varepsilon _{0}> 0\) such that for \(\varepsilon \in (0,\varepsilon _{0})\), there exists \(\eta \in C(X,X)\) satisfying
-
(i)
\(\eta (\overline{P})\subset \overline{P}\), \(\eta (\overline{Q})\subset \overline{Q}\);
-
(ii)
\(\eta \mid _{{I}^{c-2\varepsilon }}=id\);
-
(iii)
\(\eta (I^{c+\varepsilon }{\setminus } (P\cap Q))\subset {I}^{c-\varepsilon }\).
4.2 Properties of operator A
In the following, we take \(X=H\), \(P=P^{+}_{\varepsilon }\), \(Q=P^{-}_{\varepsilon }\) and \(I=I_{\mu }\). Importantly, we need an auxiliary operator A to construct a locally Lipschitz continuous operator B on E to define the descending flow. Precisely, the operator A is defined as follows: for any \(u\in H\), \(v=A(u)\) is the unique solution to the equation
Then the set of fixed points of A coincide with the set K of critical points of \(I_{\mu }\). Moreover, the operator A has the following important properties.
Lemma 4.2
The operator A is well defined and is continuous.
Proof
For \(u\in H\) fixed, we consider the functional
Obviously, \(J_{u}(v)\in C^{1}(H,R)\), coercive, bounded below, weakly lower semicontinuous and strictly convex. Therefore, \(J_{u}\) admits a unique minimizer which is the unique solution to the problem (4.1).
In the following, we prove that A is continuous. Let \(\{u_{n}\}\subset H\) such that \(u_{n}\rightarrow u\) in H. Let \(v=A(u)\) and \(v_{n}=A(u_{n})\). We need to prove \(\Vert v_{n}-v\Vert \rightarrow 0\). By the definition of A we have for any \(w\in H\)
Taking \(w=v_{n}-v\) in (4.3) and (4.4) and subtracting we obtain
Since \( \psi ^{'}\) is compact and by Theorem A.2 in [29] , we have \(\Vert v_{n}-v\Vert \rightarrow 0\), that is A is continuous.
Lemma 4.3
For any \(u\in H\) we have
-
(i)
\(\langle I_{\mu }^{'}(u),u-A(u)\rangle \ge a\Vert u-A(u)\Vert ^{2}\);
-
(ii)
\(\Vert I_{\mu }^{'}(u)\Vert \le (a+b\Vert u\Vert ^{2})\Vert u-A(u)\Vert \).
Proof
Since A(u) is a solution of (4.1), then for any \( w\in H\), we have \(\langle J_{u}^{'}(A(u)),w\rangle =0\) and
For any \(w\in H\),
so
Lemma 4.4
There exist \(\varepsilon _{0}>0\) and \(\mu _{0}>0\) such that \(A(P_{\varepsilon }^{\pm })\subset P_{\varepsilon }^{\pm }\), \(\forall \varepsilon \in (0,\varepsilon _{0})\), \(\forall \mu \in (0,\mu _{0}]\).
Proof
We only prove \(A(P_{\varepsilon }^{-})\subset P_{\varepsilon }^{-}\), \(A(P_{\varepsilon }^{+})\subset P_{\varepsilon }^{+}\) is similar. Let \(u\in H\), and \(v=A(u)\). As usual we denote \(v^{+}=\max \{0,v\}\) and \(v^{-}=\min \{0,v\}\), for any \(v\in H\). Obviously, \(\hbox {dist}(v,P^{-})\le \Vert v^{+}\Vert \). Then, by \((g_1)\), (3.2), Lemma 2.1 and Hölder’s inequality, we have
It follows that
For \(\hbox {dist}(v,P^{-})\) small enough, by \(1<q_{1}<q_{2}<\cdots<q_{m}<2\), we have
Let
and then there exist \(\varepsilon _{0}>0\) and \(\mu _{0}>0\) such that for \(\varepsilon \in (0,\varepsilon _{0})\) and \(\mu \in [0,\mu _{0}]\)
so, for any \(u\in P^{-}_\varepsilon \), we have \(A(u)\in P^{-}_\varepsilon \). \(\square \)
Remark 4.5
Note that the operator A is not compact and not locally Lipschitz continuous. However, it can be used to construct a locally Lipschitz continuous vector field which inherits its properties.
Lemma 4.6
There exists a locally Lipschitz continuous operator \(B:E\rightarrow H\) such that
-
(i)
\(\langle I_{\mu }^{'}(u),u-B(u)\rangle \ge \frac{1}{2}\Vert u-A(u)\Vert ^{2}\);
-
(ii)
\(\frac{1}{2}\Vert u-B(u)\Vert \le \Vert u-A(u)\Vert \le 2\Vert u-B(u)\Vert \);
-
(iii)
\(B(P_{\varepsilon }^{\pm })\subset P_{\varepsilon }^{\pm }\), \(\forall \varepsilon \in (0,\varepsilon _{0})\);
-
(iv)
if \(I_{\mu }\) is even, then B is odd.
The proof of this theorem follows the lines of [18].
Remark 4.7
and
4.3 Existence of sign-changing solution
Lemma 4.8
Let \(m_{1} < m_{2}\), \(\alpha >0\). Then, there exists \(\beta >0\) such that \(\Vert u-B(u)\Vert \ge \beta \), if \(u\in H\), \(I_{\mu }(u)\in [m_{1},m_{2}]\) and \(\Vert I_{\mu }^{'}(u)\Vert \ge \alpha .\)
Proof
By the definition of A, for all \(u\in H\),
so, we have
thus, by (3.7) and (ii) of Lemma 4.6,
Arguing toward a contradiction, if there exists \(\{u_{n}\}\subset H\) with \(I_{\mu }(u_{n})\in [m_{1},m_{2}]\) and \(\Vert I_{\mu }^{'}(u_{n})\Vert \ge \alpha \) such that \(\Vert u_n-B(u_{n})\Vert \rightarrow 0\) as \(n\rightarrow \infty \), then it follows from (4.5) that \(\{u_{n}\}\) is bounded, and by Remark 4.7 we see that \(\Vert I_{\mu }^{'}(u_{n})\Vert \rightarrow 0\) as \(n\rightarrow \infty \), which is a contradiction. \(\square \)
Lemma 4.9
(Deformation lemma). Let \(S\subset H\). Assume there exists a \(\varepsilon >0\) such that Lemmas 4.6 and 4.8 is satisfied. Let \(c\in \mathbb {R}\) and \(\varepsilon _0>0\) such that
Then for some \(\varepsilon _{1}\in (0,\varepsilon _0)\) there exists \(\eta \in \mathcal {C}([0,1]\times H,H)\) such that
-
(i)
\( \eta (t,u)=u\), if \(t=0\) or if \(u\notin I_{\mu }^{-1}([c-2\varepsilon _{1},c+2\varepsilon _{1}])\),
-
(ii)
\(\eta (1,I_{\mu }^{c+\varepsilon _{1}}\cap S)\subset I_{\mu }^{c-\varepsilon _{1}}\),
-
(iii)
\(I_{\mu }(\eta (\cdot ,u))\) is not increasing, \(\forall u\in H\),
-
(iv)
\(\eta (t,\overline{P^{+}_{\varepsilon }})\subset \overline{P^{+}_{\varepsilon }}\), \(\eta (t,\overline{P^{-}_{\varepsilon }})\subset \overline{P^{-}_{\varepsilon }}\), \(\forall t\in [0,1]\),
-
(v)
\(\eta (t,\cdot )\) is odd, \(\forall t\in [0,1]\).
7
Proof
The proof is similar to the proof of [5, Lemma 2.1.] and we include it here for completeness. Define
Then \(h(u)=0\ \ \hbox {on} \ \ H\backslash A_1,\) \(h(x)=1 \ \ \hbox {on}\ \ A_2\), \(0\le g\le 1\), and g is Lipschitz continuous on H. Consider the vector field
where V is defined as follows:
Obviously, W is locally Lipschitz continuous and odd in u if \(I_{\mu }\) is even. Moreover, by our choice of \(\varepsilon _1\) above we have
Consider the following initial value problem
The basic existence-uniqueness theorem for ordinary differential equations implies that for each \(u\in H\) (4.7) has a unique solution \(\sigma (\cdot ,u)\in C(\mathbb {R^{+}},H).\)
By (4.6), we have
and by Remark 4.7 we have
Define
(i) If \(t=0\), then \(\eta (t,u)=\eta (0,u)=\sigma (0,u)=u.\)
For any u, we have \(W(u)=0\). Since \(A_1\) is a closed subset of H, so \(H\backslash A_1\) is open, then there exists \(r=r(u)>0\) such that \(B_r(u)\subset H\backslash A_1\). And so, for any \(v\in B_r(u)\), \(W(u)=0\), \(\sigma (\cdot ,u)\in C(\mathbb {R^{+}},H)\) then there exists \(\xi>16\varepsilon _1>0\), such that \(|\sigma (t,u)-\sigma (0,u)|<r\), \(\forall t\in [0,\xi ]\), which implies that \(\sigma (t,u)\in B_r(u)\), \(W(\sigma (t,u))=0\), \(\forall t\in [0,\xi ]\).
Thus,
It follows that
(ii) For any \(u\in I_{\mu }^{c+\varepsilon _{1}\cap S}\), we need to prove \(I_{\mu }(\eta (1,u))=I_{\mu }(\sigma (16\varepsilon _1,u))\le c-\varepsilon _1\).
Let \(u\in I_{\mu }^{c+\varepsilon _{1}}\cap S\). If there is \(t\in [0,16\varepsilon _1]\) such that \(I_{\mu }(\sigma (t,u))\le c-\varepsilon _1,\) then it follows from (4.9) that \(I_{\mu }(\sigma (16\varepsilon _1))<c-\varepsilon _1\) and (ii) is satisfied. If
we obtain from (4.8) and (4.9)
and (ii) is also satisfied. Obviously, (iii) is satisfied by (4.9). Finally, (iv) is a consequence of Lemma 4.6 (see [20] for a detailed proof).
The following theorem is a corollary of [17, Theorem 2.4].
Lemma 4.10
Assume there exists \(\varepsilon >0\) such that Lemmas 4.6 and 4.8 is satisfied. Assume also that there exists a map \(\varphi _{0}:\triangle \rightarrow H\) satisfying
-
1.
\(\varphi _{0}(\partial _{1}\triangle )\subset P^{+}_{\varepsilon }\) and \(\varphi _{0}(\partial _{2}\triangle )\subset P^{-}_{\varepsilon }\),
-
2.
\(\varphi _{0}(\partial _{0}\triangle )\cap P^{+}_{\varepsilon }\cap P^{-}_{\varepsilon }=\emptyset \),
-
3.
\(c_{0}:=\underset{{u\in \varphi _{0}(\partial _{0}\triangle )}}{\sup }I_{\mu }(u)<c_{*}:=\underset{{u\in \partial P^{+}_{\varepsilon }\cap \partial P^{-}_{\varepsilon }}}{\inf }I_{\mu }(u).\)
where \(\triangle =\{(t_{1},t_{2}\in R^{2}:t_{1},t_{2}\ge 0, t_{1}+t_{2}\le 1\}\), \(\partial _{1}\triangle =\{0\}\times [0,1],\) \(\partial _{2}\triangle =[0,1]\times \{0\}\) and \(\partial _{0}\triangle =\{(t_{1},t_{2})\in R^{2}:t_{1},t_{2}\ge 0, t_{1}+t_{2}= 1\}\).
Then there exists a sequence \(\{u_{n}\}\subset \overline{(H\backslash (P^{+}_{\varepsilon }\cap P^{-}_{\varepsilon }))_{2\delta }}\) such that
where
If in addition \(I_{\mu }\) satisfies the \((PS)_{c}\) condition for any \(c>0\), then \(I_{\mu }\) has a sign-changing critical point.
Proof
Lemma 2.1 in [17] implies that \(\varphi (\Delta )\cap \partial P^{+}_{\varepsilon }\cap \partial P^{-}_{\varepsilon }\ne \emptyset \) for any \(\varphi \in \Gamma \). This intersection property implies that \(c\ge c_{*}>c_{0}.\)
We claim that
Arguing indirectly, if there exists \(\varepsilon _0\in (0,\frac{c-c_0}{2})\) such that
Taking \(S=H\backslash (P^{+}_{\varepsilon }\cup P^{-}_{\varepsilon })\) in Lemma 4.9 and we define the map
where \(\varepsilon _1\) and \(\eta \) are given by Lemma 4.9 and \(\varphi \) is chosen such that
Since \(c_0<c-2\varepsilon _1\), by assumption (3) we have \(I_{\mu }(\varphi _{0}(\partial _{0}\triangle ))<c-2\varepsilon _1\), which implies that \(\varphi _{0}(\partial _{0}\triangle )\subset I_{\mu }^{c-2\varepsilon _1}\). It follows that \(\tau \in \Gamma \).
Combining (ii), (iv) of Lemma 4.9 with (4.12), we have
which is in contradiction with the definition of c. So, (4.10) is satisfied.
Taking limit \(\varepsilon _0\rightarrow 0,\) then there exists a sequence \(\{u_n\}\) satisfying (4.10). Since \(I_{\mu }\) satisfies the \((PS)_c\) condition, which implies that \(u_n\rightarrow u\) as \(n\rightarrow \infty \), up to a subsequence. Moreover, \(I_{\mu }\in C^{1}(H,\mathbb {R})\), then \(u\in \overline{(H\backslash (P^{+}_{\varepsilon }\cup P^{-}_{\varepsilon }))_{2\delta }}\), and u is sign-changing.
Lemma 4.11
For \(p\in [2,6]\) there exists \(k>0\) independent of \(\varepsilon \) such that \(\Vert u\Vert _{p}\le k\varepsilon \) for \(u\in P^{+}_{\varepsilon }\cap P^{-}_{\varepsilon }\).
Proof
\(\forall u\in P^{+}_{\varepsilon }\cap P^{-}_{\varepsilon }\), by Lemma 2.1, we can obtain
thus \(\Vert u\Vert _{p}\le k\varepsilon \) for \(u\in P^{+}_{\varepsilon }\cap P^{-}_{\varepsilon }\). \(\square \)
Lemma 4.12
If \(\varepsilon >0\) small enough then there exists \(\mu _{0}>0\) such that \(\mu \in (0,\mu _{0})\) have \(I_{\mu }(u)\ge \frac{a}{8}\varepsilon ^{2}\) for \(u\in \partial P^{+}_{\varepsilon }\cap \partial P^{-}_{\varepsilon }\).
Proof
Let \(u\in \partial P^{+}_{\varepsilon }\cap \partial P^{-}_{\varepsilon }\), we have \(\Vert u^{+}\Vert \ge \hbox {dist}(u,P^{\mp })=\varepsilon \). By Lemma 4.11 and (3.3), we have
For \(\mu \le \frac{a\varepsilon ^{2-\nu }}{8C_{5}k}\) satisfying \( \frac{a}{4}\varepsilon ^{2}-\mu kC_{5}\varepsilon ^{\nu }\ge \frac{a}{8}\varepsilon ^{2}\), we have
\(\square \)
Proof of Theorem 1.4
It suffices to verify assumptions (1)–(3) in applying Lemma 4.10.
Choose \(v_{1},v_{2}\in C^{\infty }_{0}(\Omega ){\setminus }\{0\}\) satisfying \(\sup (v_{1})\cap \sup (v_{2})=\emptyset \), and \(v_{1}\le 0\), \(v_{2}\ge 0\). Let \(\varphi _{0}(t,s):=R(tv_{1}+sv_{2})\) for \((t,s)\in \Delta \), where R is a positive constant to be determined later. Obviously, for \(t,s\in [0,1]\), \(\varphi _{0}(0,s)=Rsv_{2}\in P^{+}_{\varepsilon }\) and \(\varphi _{0}(t,0)=Rtv_{1}\in P^{-}_{\varepsilon }\). Hence, the assumption (1) is satisfied.
Observe that \(\rho =\min \{|tv_{1}+(1-t)v_{2}|_{2}:0\le t\le 1\}>0\). Then, \(|u|_{2}\ge \rho R\) for \(u\in \varphi _{0}(\partial _{0}\Delta )\) and it follows from Lemma 4.11 that \(\varphi _{0}(\partial _{0}\Delta )\cap P^{+}_{\varepsilon }\cap P^{-}_{\varepsilon }=\emptyset \) for R large enough.
By (3.6) and \((g_{2})\), we have
since \(\theta >4\), by Lemma 4.12, for R large enough
By Lemma 3.3, we can know that \(I_{\mu }\) satisfies the \((PS)_{c}\) condition for any \(c>0\). According to Lemma 4.10, \(I_{\mu }\) has at least one sign-changing critical point, which is a sign-changing solution of equation (1.1).
References
Arosio, A., Panizzi, S.: On the well-posedness of the Kirchhoff string. Trans. Am. Math. Soc. 348, 305–330 (1996)
Alves, C.O., Corrêa, F.J.S.A., Ma, T.F.: Positive solutions for a quasilinear elliptic equation of Kirchhoff type. Comput. Math. Appl. 49, 85–93 (2005)
Ambrosetti, A., Brezis, H., Cerami, G.: Combined efects of concave and convex nonlinearities in some elliptic problems. J. Funct. Anal. 122, 519–543 (1994)
Batkam, C.J.: Multiple sign-changing solutions to a class of Kirchhoff type problems. arXiv:1503.07280
Batkam, C.J.: An elliptic equationunder the effect of two nonlocal terms. Math. Meth. Appl. Sci. 39, 1535–1547 (2016)
Bartsch, T., Liu, Z.: On a superlinear elliptic p-Laplacian equation. J. Differ. Equations 198, 149–175 (2004)
Chipot, M., Lovat, B.: Some remarks on nonlocal elliptic and parabolic problems. Nonlinear Anal. 30, 4619–4627 (1997)
Chen, C., Kuo, Y., Wu, T.: The Nehari manifold for a Kirchhoff type problem involving sign-changing weight functions. J. Differ. Equations. 250, 1876–1908 (2011)
Cheng, B., Wu, X.: Existence results of positive solutions of Kirchhoff type problems. Nonlinear Anal. 71, 4883–4892 (2009)
D’Ancona, A., Spagnolo, S.: Global solvability for the degenerate Kirchhoff equation with real analytic data. Invent. Math. 108, 247–262 (1992)
Deng, Y., Peng, S., Shuai, W.: Existence and asymptotic behavior of nodal solutions for the Kirchhoff-type problems in \(\mathbb{R}^{3}\). J. Funct. Anal. 269, 3500–3527 (2015)
Ding, Y., Liu, X.: Periodic solutions of a Dirac equation with concave and convex nonlinearities. J. Differ. Equations 258, 3567–3588 (2016)
Figueiredo, G.M., Nascimento, R.G.: Existence of a nodal solution with minimal energy for a Kirchhoff equation. Math. Nachr. 288, 48–60 (2014)
Jin, J., Wu, X.: Infinitely many radial solutions for Kirchhoff-type problems in \(R^{N}\). J. Math. Anal. Appl. 369, 564–574 (2012)
Kirchhoff, G.: Mechanik. Teubner, Leipzig (1883)
Lions, J.: On some questions in boundary value problems of mathematical physics. North-Holland Math. Stud. 30, 284–346 (1978)
Liu, J., Liu, X., Wang, Z.-Q.: Multiple mixed states of nodal solutions for nonlinear Schröinger systems. Calc. Var. PDEs 5, 565–586 (2015)
Liu, Z., Wang, Z., Zhang, J.: Infinitely many sign-changing solutions for the nonlinear Schrödinger–Poisson system. Ann. di Mat. 195, 775–794 (2016)
Liu, S., Li, S.: An elliptic equation with concave and convex nonlinearities. Nonlinear Anal. 53, 723–731 (2003)
Liu, Z., Sun, J.: Invariant sets of descending flow in critical point theory with applications to nonlinear differential equations. J. Differ. Equations 172, 257–299 (2001)
Liu, W., He, X.: Multiplicity of high energy solutions for superlinear Kirchhoff equations. J. Appl. Math. Comput. 39, 473–487 (2012)
Mao, A., Jing, R., Luan, S., Chu, J., Kong, Y.: Some nonlocal elliptic problem involing positive parameter. Topol. Methods Nonlinear Anal. 42(1), 207–220 (2013)
Mao, A., Zhang, Z.: Sign-changing and multiple solutions of Kirchhoff type problems without the \((PS)\) conditon. Nonlinear Anal. 70, 1275–1287 (2009)
Mao, A., Luan, S.: Sign-changing solutions of a class of nonlocal quasilinear elliptic boundary value problems. J. Math. Anal. Appl. 383, 239–243 (2011)
Nie, J., Wu, X.: Existence and multiplicity of non-trivial solutions for Schröinger–Kirchhoff-type equations with radial potentials. Nonlinear Anal. 75, 3470–3479 (2012)
Perera, K., Zhang, Z.: Nontrivial solutions of Kirchhoff-type problems via the Yang index. J. Differ. Equations 211(1), 246–255 (2006)
Rabinowitz, P.H.: Minimax methods in critical point theory with applications to differential equations, CBMS Regional Conference Series in Mathematics, vol. 65. AMS, Providence, RI (1986)
Shuai, Wei: Sign-changing solutions for a class of Kirchhoff-type problem in bounded domains. J. Differ. Equations 259, 1256–1274 (2015)
Willem, M.: Minimax Theorems. Birkhäuser, Boston (1996)
Xu, L., Chen, H.: Nontrivial solutions for Kirchhoff-type problems with a parameter. J. Math. Anal. Appl. 433, 455–472 (2016)
Ye, H.: The existence of least energy nodal solutions for some class of Kirchhoff equations and Choquard equations in \(\mathbb{R}^{N}\). J. Math. Anal. 431, 935–954 (2015)
Zhang, Z., Perera, K.: Sign changing solutions of Kirchhoff type problems via invariant sets of descent flow. J. Math. Anal. Appl. 317(2), 456–463 (2006)
Author information
Authors and Affiliations
Corresponding author
Additional information
Anmin Mao: Supported by the NSFC(11471187,11571197) and SNSFC(ZR2014AM034).
Rights and permissions
About this article
Cite this article
Shao, M., Mao, A. Signed and sign-changing solutions of Kirchhoff type problems. J. Fixed Point Theory Appl. 20, 2 (2018). https://doi.org/10.1007/s11784-018-0486-9
Published:
DOI: https://doi.org/10.1007/s11784-018-0486-9