\documentclass[reqno]{amsart}
\usepackage{hyperref}

\AtBeginDocument{{\noindent\small
\emph{Electronic Journal of Differential Equations},
Vol. 2012 (2012), No. 84, pp. 1--15.\newline
ISSN: 1072-6691. URL: http://ejde.math.txstate.edu or http://ejde.math.unt.edu
\newline ftp ejde.math.txstate.edu}
\thanks{\copyright 2012 Texas State University - San Marcos.}
\vspace{9mm}}

\begin{document}
\title[\hfilneg EJDE-2012/84\hfil Existence of solutions]
{Existence of solutions for a two-point boundary-value problem of
a fourth-order Sturm-Liouville type}

\author[S. Heidarkhani \hfil EJDE-2012/84\hfilneg]
{Shapour Heidarkhani}

\address{Shapour Heidarkhani \newline
Department of Mathematics, Faculty of Sciences,
 Razi University,  67149 Kermanshah, Iran\newline
School of Mathematics, Institute for Research in Fundamental Sciences,\\
P.O. Box 19395-5746, Tehran, Iran}
\email{sh.heidarkhani@yahoo.com\quad s.heidarkhani@razi.ac.ir}

\thanks{Submitted January 20, 2012. Published May 23, 2012.}
\subjclass[2000]{35J35, 47J10, 58E05}
\keywords{Fourth-order Sturm-Liouville problem;  critical point;
\hfill\break\indent  multiplicity of solutions}

\begin{abstract}
 In this work, we establish the existence of two intervals for
 a parameter $\lambda$ for which a two-point boundary-value 
 problem of fourth-order Sturm-Liouville type admits three weak
 solutions whose norms are uniformly bounded with respect to $\lambda$.
 Employing two three critical point theorems, existence of at least
 three weak solutions is ensured. This approach is based on variational
 methods and critical point theory.
\end{abstract}

\maketitle
\numberwithin{equation}{section}
\newtheorem{theorem}{Theorem}[section]
\newtheorem{proposition}[theorem]{Proposition}
\newtheorem{corollary}[theorem]{Corollary}
\newtheorem{remark}[theorem]{Remark}
\newtheorem{example}[theorem]{Example}
\allowdisplaybreaks

\section{Introduction}

Consider the  fourth-order Sturm-Liouville type problem
 \begin{equation}\label{1}
\begin{gathered}
(p(x) u''(x))'' - (q(x) u'(x))' + r(x) u(x) =\lambda f(x,u) \quad x \in (0,1), \\
u(0)=u(1)=u''(0)=u''(1)=0
\end{gathered}
\end{equation}
where $p$, $q$, $r \in L^{\infty}([0,1])$ with
$p^-:=\operatorname{ess\,inf}_{x \in [0,1]} p(x) > 0$, $\lambda$ is a positive
parameter and $f: [0,1] \times \mathbb{R} \to \mathbb{R} $ is an
$L^1$-Carath\'eodory function.

 Due to the importance of fourth-order two-point boundary value problems 
in describing  a large class of elastic deflection,
many researchers have studied the existence and multiplicity of solutions 
for such a problem, we refer the reader
 to \cite{AHO,B,BD1,BD2,C,L,SR,TC} and references therein.
In this article we are interested in establishing the existence of
two intervals for the positive real parameter $ \lambda$ for which 
 \eqref{1} admits three weak solutions whose norms are
uniformly bounded with respect to $\lambda$  belonging to one of
the two intervals. A basic tool in our arguments is a three
critical points theorem due to Bonanno \cite{B1}, which extends
previous results established by   Ricceri \cite{R1,R2}. The
starting point for such properties is the pioneering \emph{three
critical points} theorem due to Pucci and Serrin
\cite{pucser1,pucser2} (see also \cite{KRV}), which asserts that a
function $f\in C^1(E;{\mathbb R})$ has at least three critical
points, provided that $f$ satisfies the Palais-Smale condition and
has two local minima, where $E$ is a Banach space. Critical point
methods of the type appealed to in this paper originate in the
fundamental extension by Leggett and Williams \cite{LW} of the
multiple fixed-point technique pioneered by Krasnosell'ski and
Stecenko \cite{KS}.

 For basic notation and definitions, and for a through on the subject,
 we refer the reader to \cite{BA1,BA2,BM,BMV,GR,HAO,KRV,R2,R1,Z}.

 The rest of this article  is organized as follows. Section 2
contains our main tools, that are, some three critical points
results and some basic definitions. Whereas, Section 3 is devoted
to the existence of three solutions for the problem \eqref{1}. To
be precise, our main results (Theorems \ref{t4}-\ref{t6}),
Corollaries \ref{c1}-\ref{c3} and Corollaries \ref{c4}-\ref{c6}
which present verifiable criteria for applying our main results,
and the proof of the corollaries are there presented. Again in
Section 3, the applicability of our results is illustrated by an
example. Section 4 consists of the proofs of our main results.

 We note that some of the ideas used here were motivated
 by corresponding ideas in \cite{BD2}.

\section{Preliminaries and basic notation}

 First we here recall for the reader's convenience
our main tools to prove the results; in the first one and the
second one the coercivity of the functional $\Phi-\lambda\Psi$ is
required, in the third one a suitable sign hypothesis is assumed.
The first result has been obtained in \cite{B1}, the second one in
\cite{AB2} and the third one in \cite{AB1}. We recall the second
and the third as given in \cite{BC}.

\begin{theorem}[{\cite[Theorem 3.1]{B1}}] \label{t1}
Let $X$ be a separable and reflexive real Banach space,
 $ \Phi:X \to \mathbb{R}$
a nonnegative continuously G\^{a}teaux differentiable and
sequentially weakly lower semi-continuous functional whose
G\^{a}teaux derivative admits a continuous inverse on $X^{*}$,
$\Psi:X\to \mathbb{R}$ a continuously
 G\^{a}teaux differentiable functional whose
 G\^{a}teaux derivative is compact. Assume that there
 exists $x_{0}\in X$ such that $\Phi(x_{0})=\Psi(x_{0})=0$ and that
$$  
\lim_{  \|x\| \to +\infty} (\Phi(x)-\lambda \Psi(x))=+\infty
\quad \text{for all }  \lambda\in[0,+\infty[.
$$ 
Further, assume that there are $r>0$, $x_1\in X$ such that 
$ r <\Phi(x_1)$ and
$$
\sup_{x\in\overline{\Phi^{-1}(]-\infty,r[)}^{w}}\Psi(x) <
\frac{r}{r+\Phi(x_1)}\Psi(x_1);
$$
 here $\overline{\Phi^{-1}(]-\infty,r[)}^{w}$ denotes the closure of
$\Phi^{-1}(]-\infty,r[)$ in the weak topology (in particular
note $\Psi(x_1) \geq 0$ since 
$x_0 \in \overline{\Phi^{-1}(]-\infty,r[)}^{w}$ (note $\Psi(x_0)=0$) so
$\sup_{x\in\overline{\Phi^{-1}(]-\infty,r[)}^{w}}\Psi(x) \geq 0$).
Then, for each
$$ 
\lambda\in \Lambda_1=]
\frac{\Phi(x_1)}{\Psi(x_1)-\sup_{x\in\overline{\Phi^{-1}(]-\infty,r[)}^{w}}\Psi(x)},
\frac{r}{\sup_{x\in\overline{\Phi^{-1}(]-\infty,r[)}^{w}}\Psi(x)}[,
$$
the equation
\begin{equation}\label{2}
\Phi'(u)-\lambda \Psi'(u)=0
\end{equation}
has at least three solutions in
$X$ and, moreover, for each $h>1$, there exist an open interval
$$
\Lambda_2\subseteq[ 0,\frac{hr}{r
\frac{\Psi(x_1)}{\Phi(x_1)}-
\sup_{x\in\overline{\Phi^{-1}(-\infty,r[)}^{w}}\Psi(x)}]
$$
and a positive real number $\sigma$ such that, for each
$\lambda\in\Lambda_2$, the equation \eqref{2} has at least three
solutions in $X$ whose norms are less than $\sigma$.
\end{theorem}

\begin{theorem}[{\cite[Theorem 3.2]{BC}}]\label{t2}
 Let $X$ be a reflexive real Banach space,
$ \Phi:X \to \mathbb{R}$ be a coercive and
continuously G\^ateaux differentiable functional whose derivative
admits a continuous inverse on $X^\ast$,
 $\Psi:X \to \mathbb{R}$ be a continuously G\^ateaux differentiable functional
whose derivative is compact, such that
$$
\inf_{X}\Phi=\Phi(0)=\Psi(0)=0.
$$
Assume that there is a positive constant $r$ and $\overline{v}\in
X$, with $2r<\Phi(\overline{v})$, such that
\begin{itemize}
\item[(a1)]
 $\dfrac{\sup_{u\in\Phi^{-1}(]-\infty,r[)}\Psi(u)}{r}<
 \frac{2}{3}\frac{\Psi(\overline{v})}{\Phi(\overline{v})}$;

\item[(a2)] for all $\lambda\in
\Big]\dfrac{2\Phi(\overline{v})}{3\Psi(\overline{v})},
\frac{r}{\sup_{u\in\Phi^{-1}(]-\infty,r[)}\Psi(u)}\Big[$, the
functional $ \Phi-\lambda
\Psi$ is coercive.
\end{itemize}
Then, for each
$\lambda\in\Big]\frac{3}{2}\frac{\Phi(\overline{v})}{\Psi(\overline{v})},
\frac{r}{\sup_{u\in\Phi^{-1}(]-\infty,r[)}\Psi(u)}\Big[$ the
functional $ \Phi-\lambda \Psi$ has at least three distinct
critical points.
\end{theorem}

\begin{theorem}[{\cite[Theorem 3.3]{BC}}] \label{t3}
Let $X$ be a reflexive real Banach space,
$\Phi:X \to \mathbb{R}$ be a convex, coercive and
continuously G\^ateaux differentiable functional whose derivative
admits a continuous inverse on $X^\ast$, $\Psi:X \to
\mathbb{R}$ be a continuously G\^ateaux differentiable functional
whose derivative is compact, such that
\begin{enumerate}
 \item $\inf_{X}\Phi=\Phi(0)=\Psi(0)=0$;

 \item for each $\lambda>0$ and for every $u_1,\ u_2$ which are local
 minimum for the functional $\Phi-\lambda\Psi$ and such that $\Psi(u_1)\geq
 0$ and $\Psi(u_2)\geq
 0$, one has
  $$
\inf_{s\in[0,1]}\Psi(su_1+(1-s)u_2)\geq 0.
$$
\end{enumerate}
Assume that there are two positive constants $r_1,r_2$ and
$\overline{v}\in X$, with $2r_1<\Phi(\overline{v})<\frac{r_2}{2}$, such that
 \begin{itemize}

\item[(b1)]  $\frac{\sup_{u\in\Phi^{-1}(]-\infty,r_1[)}\Psi(u)}{r_1}<
 \frac{2\Psi(\overline{v})}{3\Phi(\overline{v})}$;

\item[ (b2)] $\frac{\sup_{u\in\Phi^{-1}(]-\infty,r_2[)}\Psi(u)}{r_2}<
 \frac{1}{3}\frac{\Psi(\overline{v})}{\Phi(\overline{v})}$.
\end{itemize}
 Then, for each
$\lambda\in\Big]\frac{3}{2}\frac{\Phi(\overline{v})}{\Psi(\overline{v})},\
\min\{ \frac{r_1}{\sup_{u\in\Phi^{-1}(]-\infty,r_1[)}\Psi(u)},\
\frac{\frac{r_2}{2}}{\sup_{u\in\Phi^{-1}(]-\infty,r_2[)}\Psi(u)}\}\Big[$,
the functional $\Phi-\lambda \Psi$ has at least three distinct
critical points which lie in $\Phi^{-1}(]-\infty,r_2[)$.
\end{theorem}

 Let us introduce some notation which will be used
later. Assume that
\begin{equation}\label{3}
    \min \big\{ \frac{q^-}{\pi^2}, \frac{r^-}{\pi^4}, \frac{q^-}{\pi^2} 
+ \frac{r^-}{\pi^4} \big\} > - p^-,
\end{equation}
where $p^- := \operatorname{ess\,inf}_{x \in [0,1]} p(x) > 0$, 
$q^- := \operatorname{ess\,inf}_{x \in [0,1]} q(x)$, 
$r^- := \operatorname{ess\,inf}_{x \in [0,1]} r(x)$.
Moreover, set
\begin{gather*}
\sigma := \min \big\{ \frac{q^-}{\pi^2}, \frac{r^-}{\pi^4},
\frac{q^-}{\pi^2} + \frac{r^-}{\pi^4},0 \big\},
\\
\delta := \sqrt{ p^- + \sigma}.
\end{gather*}
Let $X:= H^2([0,1])\cap H_0^1([0,1])$ be the Sobolev space endowed with
 the usual norm.
We recall the  following Poincar\'{e} type inequalities 
(see, for instance, \cite[Lemma 2.3]{PTV}):
\begin{gather}\label{4}
   \|u'\|_{L^2([0,1])}^2 \leq \frac{1}{\pi^2} \|u''\|_{L^2([0,1])}^2,
\\  \label{5}
 \|u\|_{L^2([0,1])}^2 \leq \frac{1}{\pi^4} \|u''\|_{L^2([0,1])}^2
\end{gather}
for all $u \in X$. Therefore, taking into account
\eqref{3}-\eqref{5}, the norm
$$
\|u\|= ( \int_0^1 ( p(x) |u''(x)|^2 + q(x) |u'(x)|^2 + r(x)
|u(x)|^2) dx )^{1/2}
$$
is equivalent to the usual norm, and, in particular, 
\begin{equation}\label{6}
\|u''\|_{L^2([0,1])} \leq \frac{1}{\delta} \|u\|.
\end{equation}


\begin{proposition}\label{p1}
  Let $u \in X$. Then
$$
\|u\|_{\infty} \leq \frac{1}{2 \pi \delta} \|u\|.
$$
\end{proposition}

\begin{proof}
  Taking \eqref{4} and \eqref{6} into account, the conclusion follows from the
 well-known inequality
  $\|u\|_{\infty} \leq \frac{1}{2} \|u'\|_{L^2([0,1])}$.
\end{proof}

\section{Main results}

 We say that $u$ is a weak solution of \eqref{1} if $u\in X$ 
and
 $$
\int_{0}^1 ( p(x) u''(x)
v''(x)+q(x)u'(x)v'(x)+r(x)u(x)v(x))dx-\lambda\int_0^1
f(x,u(x))v(x)dx=0
$$ 
for every $v\in X$. If $f$ is continuous in
$[0,1]\times \mathbb{R}$, then the weak and the classical
solutions of the problem \eqref{1} coincide.
Put  
\begin{equation} \label{f}
F(x,t)=\int_{0}^{t}f(x,\xi)d\xi
\end{equation} 
for  $(x,t)\in [0,1]\times  \mathbb{R}$.
We state our main results as follows:

\begin{theorem}\label{t4}
 Assume that there exists a function $w\in X$ and a positive constants $r$ 
such that
\begin{itemize}
\item[(A1)] $\|w\|^2> 2r$;

\item[(A2)] for $F(x,t)$  given in \eqref{f},
\[
\int_0^1\sup_{t \in [-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}, \frac{1}{\delta \pi}\sqrt{\frac{r}{2}}]} 
F(x,t) dx < r \frac{\int_0^1 F(x,w(x))dx}{r+\|w\|^2/2 };
\]

\item[(A3)]  $\frac{2}{\delta^2\pi^{4}}\limsup_{|t|\to
+\infty}\frac{F(x,t)}{t^2}<\Theta_1$ uniformly with respect to
$x\in [0,1]$ where
\begin{align*}
\Theta_1:=\max\Big\{&\frac{\int_0^1\sup_{t \in
[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}, \frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]} F(x,t) dx}{r},\\
& \frac{\frac{2r}{\|w\|^2}  \int_0^1 F(x,w(x))dx-
\int_0^1\sup_{t \in [-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},
\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}]} 
 F(x,t) dx}{hr }\Big\}
\end{align*}
 with $h>1$.
\end{itemize}
Then, for each
\begin{align*}
\lambda\in \Lambda_1
&=\Big] \frac{\frac{1}{2}\|w\|^2}{\int_0^1 F(x,w(x))dx-\int_0^1\sup_{t \in
[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}, \frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]} F(x,t) dx},
\\
&\quad \frac{r}{\int_0^1\sup_{t \in [-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}, \frac{1}{\delta \pi}\sqrt{\frac{r}{2}}]} F(x,t) dx
}\Big[,
\end{align*}
 problem \eqref{1} admits at least three weak
solutions in $X$ and, moreover, for each $h>1$, there exist an
open interval
$$ 
\Lambda_2\subseteq\Big[ 0,\frac{hr}{2r\frac{\int_{ 0}^1F(x,w(x))dx}{\|w\|^2}
- \int_0^1\sup_{t \in [-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}, 
\frac{1}{\delta\pi}\sqrt{\frac{r}{2}}]} F(x,t) dx}\Big]
$$
and a positive real number $\sigma$ such that, for each
$\lambda\in\Lambda_2$, the problem \eqref{1} admits at least
three weak solutions in $X$ whose norms are less than
$\sigma$.
\end{theorem}

\begin{theorem}\label{t5}
Assume that there exist a function $w\in X$ and a positive constant $r$ such that
\begin{itemize}
\item[(B1)] $\|w\|^2> 4r$;

\item[(B2)] $\frac{\int_0^1\sup_{t \in
[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}, \frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]} F(x,t) dx }{r}
<\frac{4}{3}\frac{\int_0^1 F(x,w(x))dx}{\|w\|^2}$;

\item[(B3)] $ \frac{2}{\delta^2\pi^{4}}\limsup_{|t|\to +\infty
}\frac{F(x,t)}{t^2}<\frac{\int_0^1\sup_{t \in
[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}, \frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]} F(x,t) dx }{r}$.
\end{itemize}
Then, for each
$$
\lambda\in\Big]\frac{3}{4}\frac{\|w\|^2}{\int_0^1 F(x,w(x))dx},\
\frac{r}{\int_0^1\sup_{t \in [-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}, \frac{1}{\delta \pi}\sqrt{\frac{r}{2}}
     ]} F(x,t) dx }\Big[,
$$ 
problem \eqref{1} admits at least three weak solutions.
\end{theorem}

\begin{theorem}\label{t6}
Suppose that $f:[0,1]\times \mathbb{R}\to \mathbb{R}$ satisfies the 
condition $f(x,t)\geq 0$ for all $x\in [0,1]$ and $t\geq 0$.
 Assume that there exist a function $w\in X$
and two positive constants $r_1$ and $r_2$ with
$4r_1<\|w\|^2<r_2$ such that
\begin{itemize}
\item[(C1)] $\frac{\int_0^1\sup_{t \in [-\frac{1}{\delta
\pi}\sqrt{\frac{r_1}{2}}, \frac{1}{\delta
\pi}\sqrt{\frac{r_1}{2}}]} F(x,t) dx }{r_1}<\frac{4}{3}\frac{\int_0^1 F(x,w(x))dx}
{\|w\|^2}$;

\item[(C2)] $\frac{\int_0^1\sup_{t \in [-\frac{1}{\delta
\pi}\sqrt{\frac{r_2}{2}}, \frac{1}{\delta
\pi}\sqrt{\frac{r_2}{2}}]} F(x,t) dx }{r_2}<\frac{2}{3}\frac{\int_0^1 F(x,w(x))dx}
{\|w\|^2}$.
\end{itemize}
Then, for each
$$
\lambda\in\Big]\frac{3}{4}\frac{\|w\|^2}{\int_{0}^1F(x,w(x))dx},\
\Theta_2\Big[,
$$
where 
\begin{align*}
\Theta_2:=\min\Big\{&\frac{r_1}{\int_0^1\sup_{t \in
[-\frac{1}{\delta \pi}\sqrt{\frac{r_1}{2}}, \frac{1}{\delta
\pi}\sqrt{\frac{r_1}{2}}]} F(x,t) dx},\\ 
&\frac{\frac{r_2}{2}}{\int_0^1\sup_{t \in [-\frac{1}{\delta
\pi}\sqrt{\frac{r_2}{2}}, \frac{1}{\delta
\pi}\sqrt{\frac{r_2}{2}}]} F(x,t) dx}\Big{\}}, 
\end{align*}
 problem \eqref{1} admits at least three
non-negative weak solutions $v^1, v^2, v^3$ such that
$$
|v^{j}(x)| < \frac{1}{\delta
\pi}\sqrt{\frac{r_2}{2}} \quad \text{for each} \ x\in [0,1],\;
j=1,2,3.
$$
\end{theorem}

Put
$$
k := \Big( \|p\|_\infty + \frac{1}{\pi^2} \|q\|_\infty +
\frac{1}{\pi^4}\|r\|_\infty \Big)^{1/2}\,.
$$ 
It is easy to see that $k > 0$ and $ \delta < k$.
Let us give a particular consequence of Theorems \ref{t4}-\ref{t6} 
for a fixed test function  $w$.

 \begin{corollary}\label{c1}
Assume that there exist two positive constants $c$ and $d$ with 
$c < \frac{32}{3 \sqrt 3 \pi} d$ such that
\begin{itemize}
\item[(A4)] $F(x,t)\geq 0$ for a.e. $x\in  [ 0, 3/8] \cup [5/8 ,1]$
      and all $t\in [0,d]$;

\item[(A5)] $\int_0^1\sup_{t \in [-c,c]}F(x,t) dx < (\delta\pi c)^2 
\frac{ \int_{3/8}^{5/8} F(x,d)dx}{(\delta\pi  c)^2+\frac{1024}{27}k^2d^2}$;

\item[(A6)] $\frac{2}{\delta^2\pi^{4}}\limsup_{|t|\to
+\infty}\frac{F(x,t)}{t^2}<\Theta_3$ uniformly with respect to
$x\in [0,1]$ where 
\begin{align*}
&\Theta_3\\
&:=\max\Big{\{}\frac{\int_0^1\sup_{t
\in [-c, c]} F(x,t) dx }{2(\delta \pi c)^2},\ \frac{(\delta \pi
c)^2\frac{\int_{3/8}^{5/8} F(x,d)dx}{\frac{256}{27}k^2d^2}-
\int_0^1\sup_{t \in
[-c, c]} F(x,t) dx}{2h(\delta \pi c)^2 }\Big{\}}
\end{align*} 
with $h>1$.
\end{itemize}
Then, for each
$$
\lambda\in \Lambda'_1=\Big]
\frac{\frac{2048}{27}k^2d^2}{\int_{3/8}^{5/8}
F(x,d)dx-\int_0^1\sup_{t \in[-c, c ]} F(x,t) dx
},\frac{2(\delta\pi c)^2}{\int_0^1\sup_{t \in [-c, c ]} F(x,t)
dx }\Big[,
$$ 
problem \eqref{1} admits at least three weak
solutions in $X$ and, moreover, for each $h>1$, there exist an
open interval
$$
 \Lambda'_2\subseteq\Big[ 0,\frac{2(\delta\pi
c)^2h }{(\delta\pi c)^2\frac{\int_{3/8}^{5/8}
F(x,d)dx}{\frac{1024}{27}k^2d^2}- \int_0^1\sup_{t \in [-c, c]}
F(x,t) dx}\Big]
$$
and a positive real number $\sigma$ such that,
for each $\lambda\in\Lambda'_2$,  problem \eqref{1} admits at
least three weak solutions in $X$ whose norms are less than
$\sigma$.
\end{corollary}

\begin{proof}
We claim that all the assumptions of Theorem \ref{t4} are fulfilled with
 $w$ given by
\begin{equation}\label{7}
w(x)= \begin{cases}
-\frac{64 d}{9} ( x^2 - \frac{3}{4}x ) & x\in [0, \frac{3}{8}[,\\
d & x\in [\frac{3}{8}, \frac{5}{8}],\\
-\frac{64 d}{9} ( x^2 - \frac{5}{4}x + \frac{1}{4} ) &
x\in ] \frac{5}{8}, 1 ]
\end{cases}
\end{equation}
and $r = 2(\delta \pi c)^2$. It is easy to verify that $w \in X$, 
and in particular, 
$$
\frac{4096}{27} \delta^2 d^2 \leq \|w\|^2 \leq \frac{4096}{27} k^2
d^2.
$$
Hence, taking into account that $c < \frac{32}{3 \sqrt 3 \pi} d$,
\[
\|w\|^2 > 2r.
\]
Since, $0\leq w(x) \leq d$ for each $x\in[0,1]$, the condition
(A4) ensures that
\[
\int_{0}^{3/8} F(x,w(x))d x + \int_{5/8}^1 F(x,w(x))d x\geq 0,
\] 
so from (A5),
\begin{align*}
\int_0^1\sup_{t \in [-c,c]}F(x,t) dx
 &<  (\delta\pi c)^2 \frac{ \int_{3/8}^{5/8} F(x,d)dx}{(\delta\pi
      c)^2+\frac{1024}{27}k^2d^2}\\
&= 2(\delta\pi c)^2 \frac{ \int_{3/8}^{5/8} F(x,d)dx}{2(\delta\pi
      c)^2+\frac{2048}{27}k^2d^2}\\
&\leq 2(\delta\pi c)^2 \frac{ \int_{0}^1 F(x,w(x))dx}{2(\delta\pi
      c)^2+\frac{2048}{27}k^2d^2   }\\
&\leq r\frac{\int_{ 0}^1F(x,w(x))dx}{r+\frac{1}{2}\|w\|^2},
\end{align*} 
so (A2) holds (note $c^2=r/(\delta\pi )^2$).
 Next notice that
\begin{align*}
&\frac{\frac{1}{2}\|w\|^2}{\int_{0}^1F(x,w(x))dx
 -\int_0^1\sup_{t \in [-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},
  \frac{1}{\delta \pi}\sqrt{\frac{r}{2}}     ]} F(x,t) dx}\\
&\leq \frac{\frac{2048}{27}k^2d^2}{ \int_{3/8}^{5/8}
F(x,d)dx-\int_{0}^1 \sup_{t\in[-c,c]}F(x,t)dx }
\end{align*}
and
$$
\frac{r}{\int_0^1\sup_{t \in [-\frac{1}{\delta \pi}
\sqrt{\frac{r}{2}}, \frac{1}{\delta \pi}\sqrt{\frac{r}{2}}  ]} F(x,t) dx}
=\frac{2(\delta\pi c)^2}{\int_{0}^1 \sup_{t\in[-c,c]}F(x,t)dx}.
$$ 
In addition note
\begin{align*}
&\frac{\frac{2048}{27}k^2d^2}{\int_{3/8}^{5/8}
F(x,d)dx-\int_{0}^1 \sup_{t\in[-c,c]}F(x,t)dx }\\
& <  \frac{\frac{2048}{27}k^2d^2}{ \big(
\frac {2(\delta\pi c)^2+ \frac{2048}{27}k^2d^2}{ 2(\delta\pi
c)^2} -1 \big)
\int_{a}^{b} \sup_{t\in[-c,c]}F(x,t)dx } \\
& =  \frac{2(\delta\pi c)^2}{\int_{0}^1
\sup_{t\in[-c,c]}F(x,t)dx }.
\end{align*}
Finally note that
\begin{align*}
&\frac{hr }{2r\frac{\int_{0}^1F(x,w(x))dx}{\|w\|^2}
- \int_0^1\sup_{t \in [-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}, \frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}      ]} F(x,t) dx }\\
&\leq\frac{ 2(\delta\pi c)^2h }{2(\delta\pi c)^2\frac{\int_{3/8}^{5/8}
F(x,d)dx}{\frac{4096}{27}k^2d^2}-
\int_{0}^1\sup_{t\in[-c,c]}F(x,t)dx},
\end{align*}
 and taking into account that $\Lambda'_1\subseteq \Lambda_1$ 
and $\Lambda_2\subseteq \Lambda'_2$ we have the desired conclusion
directly from Theorem \ref{t4}.
\end{proof}

\begin{corollary}\label{c2}
Assume that there exist two positive constants $c$ and $d$ with 
$c < \frac{16\sqrt{2}}{3 \sqrt 3 \pi}  d$ such that the assumption {\rm (A4)}
in Corollary \ref{c1} holds.
Furthermore, suppose that
\begin{itemize}
\item[(B4)] $\frac{\int_0^1\sup_{t \in [-c,c ]} F(x,t) dx }{2(\delta\pi c)^2}
<\frac{4}{3}\frac{\int_{3/8}^{5/8} F(x,d)dx}{\frac{4096}{27} k^2 d^2}$;

\item[(B5)] $ \frac{2}{\delta^2\pi^{4}}\limsup_{|t|\to +\infty
}\frac{F(x,t)}{t^2}<\frac{\int_0^1\sup_{t \in [-c, c]}
F(x,t) dx }{2(\delta\pi c)^2}$.
\end{itemize}
Then, for each
$$
\lambda\in\Big]\frac{3}{4}\frac{\frac{4096}{27} k^2
d^2}{\int_{3/8}^{5/8} F(x,d)dx},\ \frac{2(\delta\pi
c)^2}{\int_0^1\sup_{t \in [-c, c]} F(x,t) dx }\Big[,
$$
 problem \eqref{1} admits at least three weak solutions.
\end{corollary}

\begin{proof}
All the assumptions of Theorem \ref{t5} are fulfilled by
choosing $w$ as given in \eqref{7} and $r=2(\delta\pi c)^2$.
Indeed, bearing in mind that 
$$ 
\frac{4096}{27} \delta^2 d^2 \leq
\|w\|^2 \leq \frac{4096}{27} k^2 d^2 $$ 
and recalling
 $$ 
\int_{0}^{3/8} F(x,w(x))d x + \int_{5/8}^1 F(x,w(x))d x\geq 0,
$$
we clearly observe that our hypotheses
guarantee that all assumptions of Theorem \ref{t5} are satisfied.
Hence, by applying Theorem \ref{t5} we have the
conclusion.
\end{proof}

\begin{corollary}\label{c3}
Suppose that $f:[0,1]\times \mathbb{R}\to \mathbb{R}$ satisfies the condition
 $f(x,t)\geq 0$ for all $x\in [0,1]$ and $t\geq 0$. Assume that there exist 
three positive constants $c_1$, $c_2$ and $d$ with 
$ c_1<\frac{32}{3\sqrt{3}\pi} d$ and $\frac{64}{3\sqrt{3}\pi}
\frac{k d}{\delta}< c_2$ such that
\begin{itemize}
\item[(C3)]
 $\frac{\int_0^1\sup_{t \in [-c_1, c_1]} F(x,t)
dx }{2(\delta\pi c_1)^2}<\frac{4}{3}\frac{\int_{3/8}^{5/8}
F(x,d)dx}{\frac{4096}{27} k^2 d^2}$;

\item[(C4)] $\frac{\int_0^1\sup_{t \in [-c_2, c_2 ]}
F(x,t) dx }{2(\delta\pi c_2)^2}<\frac{2}{3}\frac{\int_{3/8}^{5/8}
F(x,d)dx}{\frac{4096}{27} k^2 d^2}$.
\end{itemize}
Then, for each
$$
\lambda\in\Big]\frac{3}{4}\frac{\frac{4096}{27} k^2
d^2}{\int_{3/8}^{5/8} F(x,d)dx},\ \Theta_4\Big[,
$$
where
$\Theta_4:=\min\Big{\{}\frac{2(\delta\pi c_1)^2}{\int_0^1\sup_{t
\in [-c_1, c_1 ]} F(x,t) dx},\ \frac{(\delta\pi
c_2)^2}{\int_0^1\sup_{t \in [-c_2,c_2]} F(x,t)
dx}\Big{\}}$,  problem \eqref{1} admits at least three
non-negative weak solutions $v^1,\ v^2,\ v^3$ such that
$$
|v^{j}(x)| < c_2 \quad \text{for each }  x\in [0,1],\;j=1,2,3.
$$
\end{corollary}

 \begin{proof}
Following the same way as in the proof of Corollary \ref{c2}, we achieve 
the stated assertion by applying Theorem  \ref{t6} with $w$ as given in \eqref{7},
$r_1=2(\delta\pi c_1)^2$ and $r_2=2(\delta\pi c_2)^2$.
\end{proof}

 The following results give the existence of at least three classical
solutions in $X$ to  problem \eqref{1} in the autonomous case.
Let $f:\mathbb{R}\to\mathbb{R}$ be a continuous function, and put
 $F(t)=\int_{0}^{t}f(\xi)d\xi$ for all $t\in\mathbb{R}$. We have
 the following results as direct consequences of Corollaries \ref{c1}-\ref{c3},
 respectively.

\begin{corollary}\label{c4}
Assume that there exist two positive constants $c$ and $d$ with
 $c < \frac{32}{3 \sqrt 3 \pi}  d$ such that
\begin{itemize}
\item[(A7)] $f(t)\geq 0$ for all $t\in [0,d]$;

\item[(A8)] $\max_{t \in [-c,c]}F(t) < (\delta\pi c)^2 
\frac{ \frac{1}{4}F(d)}{(\delta\pi c)^2+\frac{1024}{27}k^2d^2 }$;

\item[(A9)] $\frac{2}{\delta^2\pi^{4}}\limsup_{|t|\to
+\infty}\frac{F(t)}{t^2}<\Theta_5$ where
$$
\Theta_5:=\max\Big{\{}\frac{\max_{t \in [-c, c]} F(t)}{2(\delta
\pi c)^2},\ \frac{(\delta \pi c)^2\frac{\frac{1}{4}
F(d)}{\frac{256}{27}k^2d^2}- \max_{t \in [-c, c]}
F(t)}{2h(\delta \pi c)^2 }\Big{\}}
$$ 
with $h>1$.
\end{itemize}
Then, for each
$$\lambda\in \Lambda'_1=\Big]
\frac{\frac{2048}{27}k^2d^2}{\frac{1}{4} F(d)-\max_{t \in[-c,
c ]} F(t) },\frac{2(\delta\pi c)^2}{\max_{t \in [-c, c ]}
F(t)}\Big[,
$$ 
the problem 
\begin{equation}\label{8}
\begin{gathered}
(p(x) u''(x))'' - (q(x) u'(x))' + r(x) u(x) =\lambda f(u) \quad x \in (0,1), \\
u(0)=u(1)=u''(0)=u''(1)=0
\end{gathered}
\end{equation} 
admits at least three classical solutions in $X$ and, moreover, 
for each $h>1$, there exist an open interval
$$ 
\Lambda''_2\subseteq\Big[ 0,\frac{2(\delta\pi c)^2h }{(\delta\pi c)^2
\frac{F(d)}{\frac{256}{27}k^2d^2}-\max_{t \in [-c, c]} F(t)}\Big]
$$
and a positive real number $\sigma$ such that, for
each $\lambda\in\Lambda''_2$,  problem \eqref{8} admits at
least three classical solutions in $X$ whose norms are less than
$\sigma$.
\end{corollary}

\begin{corollary}\label{c5}
Assume that there exist two positive constants $c$ and $d$ with 
$c < \frac{16\sqrt{2}}{3 \sqrt 3 \pi} d$ such that the assumption
{\rm (A7)} in Corollary \ref{c4} holds.
Furthermore, suppose that
\begin{itemize}
\item[(B6)] $\frac{\max{t \in [-c,c ]} F(t)}{2(\delta\pi c)^2}
<\frac{1}{3}\frac{ F(d)}{\frac{4096}{27} k^2 d^2}$;

\item[(B7)] $ \frac{2}{\delta^2\pi^{4}}\limsup_{|t|\to +\infty
}\frac{F(t)}{t^2}<\frac{\max_{t \in [-c, c]} F(t)
}{2(\delta\pi c)^2}$.
\end{itemize}
Then, for each
$$
\lambda\in\Big]3\frac{\frac{4096}{27} k^2
d^2}{F(d)},\ \frac{2(\delta\pi c)^2}{\max_{t \in [-c,
c]} F(t)}\Big[
$$ 
 problem \eqref{8} admits at least three weak solutions.
\end{corollary}

\begin{corollary}\label{c6}
Suppose that $f:\mathbb{R}\to \mathbb{R}$ satisfies the condition $f(t)\geq 0$ 
for all $t\geq 0$. Assume that there exist three positive constants $c_1$, $c_2$
and $d$ with $ c_1<\frac{32}{3\sqrt{3}\pi} d$ and
$\frac{64}{3\sqrt{3}\pi} \frac{k d}{\delta}< c_2$ such that
\begin{itemize}
\item[(C5)] $\frac{\max_{t \in [-c_1, c_1]} F(t)
}{2(\delta\pi c_1)^2}<\frac{1}{3}\frac{ F(d)}{\frac{4096}{27}
k^2 d^2}$;
\item[(C6)] $\frac{\max_{t \in [-c_2, c_2 ]} F(t)}{2(\delta\pi c_2)^2}
<\frac{1}{6}\frac{ F(d)}{\frac{4096}{27} k^2 d^2}$.
\end{itemize}
Then, for each
$$
\lambda\in\Big]3\frac{\frac{4096}{27} k^2 d^2}{F(d)},\ \Theta_6\Big[,
$$
where
$\Theta_6:=\min\big\{\frac{2(\delta\pi c_1)^2}{\max_{t \in
[-c_1, c_1 ]} F(t)},\ \frac{(\delta\pi
c_2)^2}{\max_{t \in [-c_2,c_2]} F(t)}\big\}$,
problem \eqref{8} admits at least three non-negative weak
solutions $v^1,\ v^2,\ v^3$ such that
$$
|v^{j}(x)| < c_2 \quad \text{for each }  x\in [0,1],\; j=1,2,3.
$$
\end{corollary}

 We conclude this section by presenting an example to illustrate our results
 applying by Corollary \ref{c4}.

\begin{example} \label{examp1}\rm
 Put $p(x)=3e^x,\ q(x)=x-\pi^2$
and $r(x)=x^2-\pi^4$ for every $x\in[0,1]$, and
$f(t)=e^{-t}t^{9}(10-t)$ for each $t\in \mathbb{R}$. It is easy to
verify that with $c = 1$ and $d=5$, taking into account that
$\delta=1$ and $k=\sqrt{3e+2}$, since
 $\limsup_{|t|\to+\infty}\frac{F(t)}{t^2}=0$, all assumptions of Corollary
\ref{c4} are satisfied. Then, for each
$$ 
\lambda\in \Lambda'''_1=\Big] \frac{\frac{51200}{27}(3e+2)}
{\frac{1}{4}(e^{-5}5^{10})-e},\ \frac{2\pi^2}{e}\Big[
$$ 
problem \eqref{8}  admits at least three classical
solutions in $X$ and, moreover, for each $h>1$, there exist an
open interval
$$
\lambda\in\Lambda'''_2\subseteq[0,\frac{2\pi^2h}{\pi^2\frac{e^{-5}5^{10}}
{\frac{25600}{27}(3e+2)}-e}]
$$
and a positive real number $\sigma$ such that, 
for each $\lambda\in\Lambda'''_2$, 
problem \eqref{8} admits at least three classical
solutions in $X$ whose norms are less than $\sigma$.
\end{example}

\begin{remark} \label{rmk1}\rm 
The weak solutions of  problem \eqref{1} where $f$ is a
continuous function, by using standard methods, are classical
solutions. Namely, the classical and the weak solutions of  problem \eqref{1}
coincide.
\end{remark}

\section{Proofs the main results}

\begin{proof}[Proof of Theorem \ref{t4}]
 With the purpose of applying Theorem \ref{t1}, arguing as in \cite{HAO},
 we begin by setting
$$
\Phi(u)=\frac{1}{2}\|u\|^2, \quad 
\Psi(u)=\int_{a}^{b}F(x,u(x))dx
$$
for $u \in X$. It is well known that $\Psi$ is a G\^{a}teaux differentiable
functional whose G\^{a}teaux derivative at the point $u\in X$ is
the functional $\Psi'(u)\in X^{*}$, given by
$$
\Psi'(u)(v)=\int_{a}^{b}f(x,u(x))v(x)dx
$$ 
for every $v\in X$, and that $\Psi':X \to X^{*}$
is a continuous and compact operator. Moreover, $\Phi$ is a
continuously G\^{a}teaux differentiable and sequentially weakly
lower semi continuous functional whose G\^{a}teaux derivative at
the point $u\in X$ is the functional $\Phi'(u)\in X^{*}$, given
by
$$
\Phi'(u)(v)=\int_{0}^1 ( p(x) u''(x) v'' (x) + q(x) u'(x)
v'(x) + r(x) u(x) v(x))dx
$$
for  $v\in X$. Moreover,
$\Phi'$ admits a continuous inverse on $X^{*}$. Furthermore from
(A3) there exist two constants $\gamma, \tau\in \mathbb{R}$ with
$\gamma<\Theta_1$ such that
$$
\frac{2}{\delta^2\pi^{4}}F(x,t)\leq \gamma t^2+\tau \
\text{for all} \ x\in (0,1)\ \text{and all}\ t\in
\mathbb{R}.
$$ 
Fix $u\in X$. Then
\begin{equation}\label{9}
F(x,u(x))\leq \frac{\delta^2\pi^{4}}{2}(\gamma |u(x)|^2+\tau)\quad
 \text{for all }  x\in (0,1).
\end{equation} 
Now,  to prove the coercivity of the functional $\Phi-\lambda \Psi$, 
first we assume that $\gamma> 0$. So, for any fixed 
$\lambda\in ]0,\frac{1}{\Theta_1}]$, since
 $\|u\|_{L_2([0,1])}\leq \frac{1}{\delta\pi^2}\|u\|$
 (see \cite[pg 1168]{B1}), using \eqref{9}, we have 
\begin{align*}
\Phi(u)-\lambda \Psi(u)
&= \frac{\|u\|^2}{2}-\lambda\int^1_{0}F(x,u(x))dx\\
&\geq \frac{1}{2}\|u\|^2-\frac{\delta^2\pi^{4}}{2\Theta_1}(\gamma
\int_{0}^1|u(x)|^2dx+\tau)\\
&\geq \frac{1}{2}(1-\frac{\gamma}{\Theta_1})\|u\|^2
-\frac{\delta^2\pi^{4}}{2\Theta_1}\tau,
\end{align*}
and so
$$
\lim_{\mid\mid u\mid\mid\rightarrow+\infty} (\Phi(u)-\lambda \Psi(u))=+\infty.
$$
On the other hand, if $\gamma \leq 0$, Clearly, we obtaain
$\lim_{\|u\| \rightarrow+\infty}(\Phi(u)-\lambda \Psi(u))=+\infty$.
Both cases lead to the coercivity of functional $\Phi-\lambda \Psi$. 
Also according to (A1) we achieve $\Phi(w)>r$. 
Note $\|u\|_{\infty} \leq \frac{1}{2\pi \delta} \|u\|$ for each $u \in X$, 
from the definition of $\Phi$ we observe
\begin{align*}
\Phi^{-1}(]-\infty,r[)
&= \{ u\in X; \Phi(u)<r\}\\
&\subseteq \{ u\in X; \|u\|< \sqrt{2r}\}\\
&\subseteq \{ u\in X; |u(x)|\leq  \frac{1}{2 \pi \delta} \sqrt{2r}, \quad
  \text{for all } x\in [0,1]\}\\
&=  \{ u \in X : |u(x)| < \frac{1}{\delta
\pi}\sqrt{\frac{r}{2}} \quad\text{for all } x \in [0,1]\},
\end{align*} 
so, we have
$$
\sup_{\overline{u\in\Phi^{-1}(]-\infty,r[)}^{w}}\Psi(u)
\leq \int_{0}^1 \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx .
$$ 
Therefore, from (A2) we have
\begin{align*}
\sup_{\overline{u\in\Phi^{-1}(]-\infty,r[)}^{w}}\Psi(u)
&\leq   \int_{0}^1 \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}},\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}]}F(x,t)dx\\
&< \frac{r}{r+ \frac{1}{2}\|w\|^2}\int_{0}^1F(x,w(x))dx \\ 
&= \frac{r}{r+\Phi(w)}\Psi(w).
\end{align*} 
Now, we can apply Theorem \ref{t1}. Note for each $x\in [0,1]$,
\begin{align*}
&\frac{\Phi(w)}{\Psi(w)-\sup_{u\in\overline{\Phi^{-1}(]-\infty,r[)}^{w}}\Psi(u)}\\
&\leq \frac{\frac{1}{2}\|w\|^2}{\int_{
a}^{b}F(x,w(x))dx-\int_{a}^{b} \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx }
\end{align*}
and
$$
\frac{r}{\sup_{u\in\overline{\Phi^{-1}(]-\infty,r[)}^{w}}\Psi(u)}\geq
\frac{r}{\int_{0}^1 \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx }.
$$ 
Note also that (A2)  implies
\begin{align*}
& \frac{\frac{1}{2}\|w\|^2}{\int_{0}^1F(x,w(x))dx
-\int_{0}^1 \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx}
\\ 
&< \frac{\frac{1}{2}\|w\|^2}{ ( \frac{r+
\frac{1}{2}\|w\|^2}{r} \,-\,1 ) \int_{0}^1
\sup_{t\in[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx}
\\
&=\frac{r}{\int_{0}^1
\sup_{t\in[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx }.
\end{align*}
Also
\begin{align*}
&\frac{hr}{r \frac{\Psi(w)}{\Phi(w)}-
\sup_{u\in\overline{\Phi^{-1}(-\infty,r[)}^{w}}\Psi(u)}\\
&\leq\frac{hr }{2r\frac{\int_{0}^1F(x,w(x))dx}{\|w\|^2}-
\int_{0}^1\sup_{t\in[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},
\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}]}F(x,t)dx}=\rho.
\end{align*}
Note from (A2) that
\begin{align*}
&  2r\frac{\int_{ 0}^1F(x,w(x))dx}{\|w\|^2}-
\int_{0}^1\sup_{t\in[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},
\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx \\
& > \Big( \frac{2r}{\|w\|^2} - \frac{r}{ r+
\frac{1}{2}\|w\|^2} \Big) \int_{0}^1F(x,w(x))dx \\
& \geq \Big( \frac{2r}{\|w\|^2} - \frac{2r}{
\|w\|^2} \Big)\int_{0}^1F(x,w(x))dx=0,
\end{align*}
since $\int_{0}^1F(x,w(x))dx \geq 0$ (note $F(x,0)=0$ so
$\int_{0}^1\sup_{t\in[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},
\frac{1}{\delta \pi}\sqrt{\frac{r}{2}}]}F(x,t)dx \geq 0$ 
and now apply (A2)).
 Now with $x_{0}=0,  x_1=w$ from Theorem \ref{t1}
(note $\Psi(0)=0$ ) it follows that, for each $\lambda\in
\Lambda_1$,  problem \eqref{1} admits at least three weak
solutions and there exist an open interval
$\Lambda_2\subseteq[0,\rho]$ and a real positive number
 $\sigma$ such that, for each $\lambda\in \Lambda_2$, problem
\eqref{1} admits at least three weak solutions  whose norms in
$X$ are less than $\sigma$. 
\end{proof}

\begin{proof}[Proof of Theorem \ref{t5}]
 To apply Theorem \ref{t2} to our problem, we take the functionals 
$\Phi, \Psi:X\to \mathbb{R} $ as given in the proof of Theorem \ref{t4}.
 Let us prove that the functionals $\Phi$ and $\Psi$ satisfy the
conditions required in Theorem \ref{t2}. The regularity
assumptions on $\Phi$ and $\Psi$, as requested in Theorem \ref{t2}
hold. According to (B1) we deduce $\Phi(w)>2r$. Note that
$\|u\|_{\infty} \leq \frac{1}{2 \pi \delta} \|u\|$ for each 
$u \in X$, we observe
$$
\Phi^{-1}(]-\infty,r[)= \{ u \in X : |u(x)| <
\frac{1}{\delta \pi}\sqrt{\frac{r}{2}} \text{ for all } x \in
[0,1] \},
$$ 
and it follows that
\begin{align*}
\sup_{(u)\in\Phi^{-1}(]-\infty,r[)}\Psi(u)
&= \sup_{(u_1,\dots,u_{n})\in\Phi^{-1}(]-\infty,r[)}\int_{0}^1F(x,u)dx\\
&\leq  \int_{0}^1 \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx.
\end{align*}
 Therefore, due to  assumption (B2), we have
\begin{align*}
\frac{\sup_{u\in\Phi^{-1}(]-\infty,r[)}\Psi(u)}{r}
&= \frac{\sup_{(u_1,\dots,u_{n})\in\Phi^{-1}(]-\infty,r[)}\int_{0}^1F(x,u(x))dx}{r}
\\ 
&\leq   \frac{\int_{0}^1 \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx}{r}\\
&< \frac{4}{3}\frac{\int_{ a}^{b}F(x,w(x))dx}{\|w\|^2}
=\frac{2\Psi(w)}
 {3\Phi(w)}.\end{align*}
 Furthermore, from (B3) there exist two constants
$\eta, \vartheta\in R$ with
 $$
\eta<\frac{\int_{0}^1
\sup_{t\in[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx}{r}
$$ 
such that
$$
\frac{2}{\delta^2\pi^{4}}F(x,t)\leq \eta
t^2+\vartheta \text{ for all } x\in [0,1] \text{ and for all }
 t\in \mathbb{R}.
$$ 
Fix $u\in X$. Then
\begin{equation}\label{10}
F(x,u(x))\leq \frac{\delta^2\pi^{4}}{2}(\eta
|u(x)|^2+\vartheta)\quad \text{for all } x\in[0,1].
\end{equation}

 Now,  to prove the coercivity of the functional
$\Phi-\lambda \Psi$, first we assume that $\eta> 0$. So, for any
fixed
$$
\lambda\in\Big]\frac{3}{4}\frac{\|w\|^2}{\int_{a}^{b}F(x,w(x))dx},\
\frac{r}{\int_{0}^1 \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx}\Big[,
$$ 
since
$\|u\|_{L_2([0,1])}\leq \frac{1}{\delta\pi^2}\|u\|$
 (see \cite[pg 1168]{B1}), using \eqref{10}, we have 
\begin{align*}
\Phi(u)-\lambda \Psi(u)
&= \frac{\|u\|^2}{2}-\lambda\int^1_{0}F(x,u(x))dx\\
&\geq \frac{1}{2}\|u\|^2-\frac{\lambda\delta^2\pi^{4}}{2}(\gamma
\int_{0}^1|u(x)|^2dx+\tau)\\
&\geq \frac{1}{2}\Big{(}1-\gamma\frac{r}{\int_{0}^1
\sup_{t\in[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx}\Big{)}\|u\|^2
-\frac{\lambda\delta^2\pi^{4}}{2}\tau,
\end{align*}
and thus
$$
\lim_{\|u\| \rightarrow+\infty} (\Phi(u)-\lambda \Psi(u))=+\infty.
$$ 
On the other hand, if $\eta \leq 0$, Clearly, we obtain
 $\lim_{\|u\| \rightarrow+\infty}
(\Phi(u)-\lambda \Psi(u))=+\infty$.
 Both cases lead to the coercivity of functional $\Phi-\lambda \Psi$.

 So, the assumptions (a1) and (a2) in Theorem \ref{t2} are
satisfied. Hence, by using Theorem \ref{t2}, taking into account
that the weak solutions of the problem \eqref{1} are exactly the
solutions of the equation $\Phi'(u)-\lambda \Psi'(u)=0$, then
problem \eqref{1} admits at least three distinct weak solutions in
$X$.
\end{proof}

\begin{proof}[Proof of Theorem \ref{t6}]
 Let $\Phi$ and $\Psi$ be as in
the proof of Theorem \ref{t4}. Let us employ Theorem \ref{t3} to
our functionals. Obviously, $\Phi$ and $\Psi$ satisfy the
condition 1 of Theorem \ref{t3}.

 Now, we show that the functional $\Phi-\lambda\Psi$ satisfies the
assumption 2 of Theorem \ref{t3}. Let $u^\ast$ and
$u^{\star\star}$ be two local minima for $\Phi-\lambda\Psi$. Then
$u^\ast$ and $u^{\star\star}$ are critical points for
$\Phi-\lambda\Psi$, and so, they are weak solutions for 
problem \eqref{1}. Since $f(x,t)\geq 0$ for all $(x,t)\in
[0,1]\times (\mathbb{R}^{+}\cup \{0\})$, from the Weak Maximum
Principle (see for instance \cite{Br} ) we deduce
$u^{\star}(x)\geq 0$ and $u^{\star\star}(x)\geq 0$ for every $x\in
[0,1]$. So, it follows that $su^{\star}+(1-s)u^{\star\star}\geq 0$
for all $s\in [0,1]$, and that
$f(su^{\star}+(1-s)u^{\star\star},t)\geq 0$, and consequently,
$\Psi(su^{\star}+(1-s)u^{\star\star})\geq 0$ for all $s\in [0,1]$.
Moreover, from the condition $4r_1<\|w\|^2<r_2$, we observe
$2r_1<\Phi(w)<\frac{r_2}{2}$. Note that $\|u\|_{\infty} \leq
\frac{1}{2 \pi \delta} \|u\|$ for each $u \in X$, we observe
$$
\Phi^{-1}(]-\infty,r_1[)= \big\{ u \in X : |u(x)| <
\frac{1}{\delta \pi}\sqrt{\frac{r_1}{2}} \text{ for all } x \in
[0,1] \big\},
$$ 
and it follows that
\begin{align*}
\sup_{(u_1,\dots,u_{n})\in\Phi^{-1}(]-\infty,r_1[)}\Psi(u)
&= \sup_{(u\in\Phi^{-1}(]-\infty,r_1[)}\int_{0}^1F(x,u(x))dx\\
&\leq  \int_{0}^1
\sup_{t\in[-\frac{1}{\delta \pi}\sqrt{\frac{r}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r}{2}}]}F(x,t)dx.
\end{align*}
 Therefore, due to assumption (C1), we infer that
\begin{align*}
\frac{\sup_{u\in\Phi^{-1}(]-\infty,r_1[)}\Psi(u)}{r_1}
&= \frac{\sup_{u\in\Phi^{-1}(]-\infty,r_1[)}\int_{0}^1F(x,u(x))dx}{r_1}\\ 
&\leq  \frac{\int_{0}^1 \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r_1}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r_1}{2}}]}F(x,t)dx}{r_1}\\
&< \frac{4}{3}\frac{\int_{ 0}^1F(x,w(x))dx}{\|w\|^2}
= \frac{2}{3}\frac{\Psi(w)}  {\Phi(w)}.
\end{align*}
 As above, bearing  assumption (C2) in mind, we deduce that
\begin{align*}
\frac{\sup_{u\in\Phi^{-1}(]-\infty,r_2[)}\Psi(u)}{r_2}
&= \frac{\sup_{(u_1,\dots,u_{n})\in\Phi^{-1}(]
 -\infty,r_2[)}\int_{0}^1F(x,u(x))dx}{r_2}\\ 
&\leq  \frac{\int_{0}^1 \sup_{t\in[-\frac{1}{\delta
\pi}\sqrt{\frac{r_2}{2}},\frac{1}{\delta
\pi}\sqrt{\frac{r_2}{2}}]}F(x,t)dx}{r_2}\\
 &< \frac{2}{3}\frac{\int_{0}^1F(x,w(x))dx}{\|w\|^2 }
= \frac{2}{3}\frac{\Psi(w)}  {\Phi(w)}.
\end{align*} 
So, the assumptions (b1) and (b2) in Theorem \ref{t3} are satisfied.
Hence, by using Theorem \ref{t3}, taking into account that the
weak solutions of  problem \eqref{1} are exactly the solutions
of the equation $\Phi'(u)-\lambda \Psi'(u)=0$,  problem
\eqref{1} admits at least three distinct weak solutions in $X$. 
This completes the proof.
\end{proof}

\subsection*{Acknowledgements}
This research was partially supported by a grant 90470020 from IPM.

\begin{thebibliography}{99}

\bibitem{AHO} G. A. Afrouzi, S. Heidarkhani,  D. O'Regan;
Existence of three solutions for a doubly eigenvalue fourth-order 
boundary value problem, \emph{Taiwanese J. Math.} \textbf{15} (2011), 201-210.

\bibitem{AB1} D. Averna, G. Bonanno;
 A mountain pass theorem for a suitable
class of functions, \emph{Rocky Mountain J. Math.} \textbf{39} (2009), 707-727.

\bibitem{AB2}D. Averna, G. Bonanno;
 A three critical points theorem and its applications to the ordinary
 Dirichlet problem, \emph{Topol. Methods Nonlinear Anal.} \textbf{22} (2003), 93-103.

\bibitem{B} Z. Bai;
 Positive solutions of some nonlocal fourth-order boundary value problem;
 \emph{Appl. Math. Comput.} \textbf{215} (2010), 4191-4197.

\bibitem{B1} G. Bonanno;
A critical points theorem and nonlinear differential problems;
 \emph{J. Global Optimization} \textbf{28} (2004), 249-258.

\bibitem{BC} G. Bonanno, P. Candito;
 Non-differentiable functionals and applications to elliptic problems 
with discontinuous nonlinearities, \emph{J. Differential Equations} 
\textbf{244} (2008), 3031-3059.

\bibitem{BA1} G. Bonanno, G. D'Agu\`{i};
 A Neumann boundary value problem for the Sturm-Liouville equation, 
\emph{Appl. Math. Comput.} \textbf{15} (2009), 318-327.

\bibitem{BA2}G. Bonanno, G. D'Agu\`{i};
 Multiplicity results for a perturbed elliptic
Neumann problem, \emph{Abstract and Applied Analysis},
2010, Art. ID 564363, 10 pp.

\bibitem{BD1} G. Bonanno, B. Di Bella;
A boundary value problem for fourth-order elastic beam equations, 
\emph{J. Math. Anal. Appl.} \textbf{343} (2008), 1166-1176.

\bibitem{BD2} G. Bonanno, B. Di Bella;
 A fourth-order boundary value problem for a Sturm–Liouville type equation, 
\emph{Appl. Math. Comput.
} \textbf{217} (2010),  3635-3640.

\bibitem{BM} G. Bonanno and S.A. Marano,  On
the structure of the critical set of non-differentiable
functionals with a weak compactness condition, \emph{Appl. Anal.} \textbf{89}
(2010), 1-10.

\bibitem{BMV} G. Bonanno, G. Molica Bisci,  V. R\u{a}dulescu;
 Infinitely many solutions for a class of nonlinear eigenvalue problem 
in Orlicz-Sobolev spaces,
\emph{C. R. Acad. Sci. Paris, Ser. I} \textbf{349} (2011), 263-268.

\bibitem{Br} H. Br\'{e}zis, \emph{Analyse Fonctionelle.
Th\'eorie et Applications,} Masson, Paris, 1983.

 \bibitem{C} G. Chai;
Existence of positive solutions for fourth-order boundary value problem with variable
parameters, \emph{Nonlinear Anal.} \textbf{66} (2007), 870-880.

\bibitem{GR} M. Ghergu, V. R\u{a}dulescu;
\emph{Singular Elliptic Problems. Bifurcation and Asymptotic Analysis,}
 Oxford Lecture Series in Mathematics and Its Applications, vol. 37, 
Oxford University Press, 2008.

\bibitem{HAO} S. Heidarkhani, G. A. Afrouzi, D. O'Regan;
 Existence of three solutions for a Kirchhoff-type boundary value problem,
\emph{Electronic Journal of Differential Equations} \textbf{2011} (2011) No. 91,
 1–11.

\bibitem{KS} M. A. Krasnosel'skii, V. JA. Stecenko;
\emph{Some nonlinear problems with many solutions,} 
Amer. Math. Soc. Transl., Ser. \textbf{2}, 54 (1966), 29-48.

\bibitem{KRV} A. Krist\'{a}ly, V. R\u{a}dulescu,  Cs. Varga;
\emph{Variational Principles in Mathematical Physics, Geometry, and Economics:
Qualitative Analysis of Nonlinear Equations and Unilateral
Problems,} Encyclopedia of Mathematics and its Applications, No.
136, Cambridge University Press, Cambridge, 2010.

\bibitem{LW} R. W. Leggett, L. R. Williams;
 \emph{Multiple positive fixed points of nonlinear operators on ordered Banach
spaces,} Indiana Univ. Math. J., \textbf{28} (1979), 673-688.

\bibitem{L} Y. Li;
 On the existence of positive solutions for the bending elastic beam
equations, \emph{Appl. Math. Comput.} \textbf{189} (2007), 821-827.

\bibitem{PTV} L.A. Peletier, W. C. Troy,  R. C. A. M. Van der Vorst;
Stationary solutions of a fourth order nonlinear diffusion equation,
(Russian) Translated from the English by V. V. Kurt.
\emph{Differentsialnye Uravneniya} \textbf{31} (1995), 327-337. English
translation in \emph{Differential Equations} \textbf{31} (1995), 301-314.

\bibitem{pucser1} P. Pucci, J. Serrin;
 Extensions of the mountain pass theorem,
\emph{J. Funct. Anal.} \textbf{59} (1984), 185-210.

\bibitem{pucser2} P. Pucci, J. Serrin;
 A mountain pass theorem, \emph{J. Diff. Equations} 60 (1985), 142-149.

\bibitem{R1} B. Ricceri;
 On a three critical points theorem,
\emph{Arch. Math. (Basel)} \textbf{75} (2000), 220-226.

\bibitem{R2} B. Ricceri;
 A three critical points theorem revisited, \emph{Nonlinear Anal.} 
\textbf{70} (2009), 3084-3089.

\bibitem{SR} V. Shanthi, N. Ramanujam;
 A numerical method for boundary value problems for singularly perturbed 
fourth-order ordinary differential equations,
 \emph{Appl. Math. Comput.} \textbf{129} (2002), 269-294.

\bibitem{TC} S. Tersian, J. Chaparova;
Periodic and homoclinic solutions of extended Fisher–Kolmogorov equations, 
\emph{J. Math. Anal. Appl.} \textbf{260} (2001), 490-506.

\bibitem{Z} E. Zeidler;
 \emph{Nonlinear functional analysis and its applications,} 
Vol. II, III, Berlin-Heidelberg-New York 1985.

\end{thebibliography}

\end{document}

