Miskolc Mathematical Notes HU e-ISSN 1787-2413

Vol. 21 (2020), No. 2, pp. 1019–1030 DOI: 10.18514/MMN.2020.2345

A CLASS OF FOURTH-ORDER KIRCHHOFF TYPE ELLIPTIC

PROBLEMS WITH FOUR SOLUTIONS

S. SHOKOOH
Received 03 June, 2017

Abstract. In this paper, we establish the existence of at least four solutions for a perturbed fourth-
order Kirchhoff type elliptic problem by using a recent variational principle due to Ricceri.

2010 Mathematics Subject Classification: 47H25; 35J66; 35A15

Keywords: nonlinear sciences, Kirchhoff type elliptic problem, four solutions

1. I NTRODUCTION
This paper is concerned with the existence of weak solutions to the following
fourth-order Kirchhoff type elliptic problem
  Z  p−1
 ∆(|∆u| p−2 ∆u) − M |∇u| p dx ∆ p u + η|u| p−2 u


Ω (1.1)

 = λa(x) f (u) + b(x)g(u) in Ω,
u = ∆u = 0 on ∂Ω,


where Ω is a bounded smooth domain in RN (N ≥ 1), p > max 1, N2 , ∆ p u =



div(|∇u| p−2 ∇u) is the p-Laplacian operator, η > 0, f , g : R → R are continuous non-
constant functions, λ ∈ R, a, b ∈ L1 (Ω) are non-negative non-constant functions and
the function M : [0, +∞[→ R is a continuous function.
The fourth-order equation furnishes a model to study travelling waves in suspen-
sion bridges, so it is important to physics. Due to this, many researchers have dis-
cussed the existence of at least one solution, or multiple solutions, or even infinitely
many solutions for fourth-order boundary value problems by using lower and up-
per solution methods, Morse theory, the mountain-pass theorem, constrained minim-
ization and concentration-compactness principle, fixed point theorems and degree
theory, and variational methods and critical point theory. We refer the reader to
[1, 8, 10, 13, 22, 23].

This research was supported by a grant from Gonbad Kavous University.

c 2020 Miskolc University Press

1020 S. SHOKOOH

This problem models several physical and biological systems where u describes a
process which depends on the average of itself, such as the population density, see
[2].
The problem (1.1) is a generalization of the stationary problem of a model intro-
duced by Kirchhoff [9]. More precisely, Kirchhoff proposed a model given by the
equation !
∂2 u ∂u 2 ∂2 u
Z L
P0 E
ρ − + dx = 0, (1.2)
∂t 2 h 2L 0 ∂x ∂x2
which L, E, h, ρ, P0 are constants, which extends the classical D’Alembert’s wave
equation by considering the effects of the changes in the length of the string during
the vibrations.
However, to be accurate, the problem (1.1) is related to the models of extensible
beam and plates by Woinowsky-Krieger [20] and Berger [4]. In addition, the first
stationary study of such fourth order nonlocal boundary value problem was given by
Ma [11]. In [13], it was named first time fourth order problem of Kirchhoff type. The
problem is also related to the so-called p-Kirchhoff problems.
In recent years, problems involving Kirchhoff type operators have been studied in
many papers, we refer interested readers to [2, 3, 7, 8, 12, 16, 19, 21, 24]. For instance,
B. Ricceri in an interesting paper [16] established the existence of at least three weak
solutions to a class of Kirchhoff-type doubly eigenvalue boundary value problems
using Theorem A of [14].
The starting point of our approach to problem (1.1) has been [15], in which the
author considers the following ordinary Neumann problem
−u00 + u = λa(x) f (u) + b(x)g(u) in [0, 1],

(1.3)
u0 (0) = u0 (1) = 0,
where λ ∈ R and f , g : R → R, a, b : [0, 1] → [0, +∞[ are four continuous non-constant
functions. Ricceri has proved that the above problem admits at least three non-zero
solutions.
Our results here are motivated by the recent papers [5, 15, 17, 18]. We establish some
sufficient conditions under which the problem (1.1) possesses four weak solutions.
The paper is arranged as follows. In Section 2, we give preliminaries and our main
tool, that is, Theorem 1, while in Section 3, we present our main results.

2. F UNCTIONAL SETTING
As already said in the introduction, our approach in facing problem (1.1) is based
on the multiplicity result established in [15], that we recall below for the reader’s
convenience (see also Theorem 1 of [17] for a similar result).
Theorem 1 ([15, Theorem 1]). Let X be a reflexive real Banach space; Φ : X → R
be a coercive and sequentially weakly lower semicontinuous C1 functional whose
 A CLASS OF FOURTH-ORDER PROBLEMS WITH FOUR SOLUTIONS 1021

derivative admits a continuous inverse on X ∗ , Ψ1 , Ψ2 : X → R two C1 functionals with

compact derivative. Assume that there exist two points u0 , v0 ∈ X with the following
properties:
(i) u0 is a strict local minimum of Φ and Φ(u0 ) = Ψ1 (u0 ) = Ψ2 (u0 ) = 0;
(ii) Φ(v0 ) ≤ Ψ1 (v0 ) and Ψ2 (v0 ) > 0.
Moreover, suppose that, for some ρ ∈ R, one has either
sup inf (λ(Φ(u) − Ψ1 (u) − ρ) − Ψ2 (u)) < inf sup(λ(Φ(u) − Ψ1 (u) − ρ) − Ψ2 (u))
λ>0 u∈X u∈X λ>0
(2.1)
or
sup inf (Φ(u) − Ψ1 (u) − λ(ρ + Ψ2 (u))) < inf sup(Φ(u) − Ψ1 (u) − λ(ρ + Ψ2 (u))).
λ>0 u∈X u∈X λ>0
(2.2)
Finally, assume that
 
Ψ1 (u) Ψ1 (u)
max lim sup , lim sup <1 (2.3)
kuk→+∞ Φ(u) u→u0 Φ(u)

and
 
Ψ2 (u) Ψ2 (u)
max lim sup , lim sup ≤ 0. (2.4)
kuk→+∞ Φ(u) u→u0 Φ(u)

Under such hypotheses, there exists λ∗ > 0 such that the equation Φ0 (u) = Ψ01 (u) +
λ∗ Ψ02 (u) has at least four solutions in X. More precisely, among them, one is u0
as a strict local, not global minimum and two are global minima of the functional
Φ − Ψ1 − λ∗ Ψ2 .
Remark 1. It is important to remark that, in view of Theorem 1 of [6], condition
(2.1) is equivalent to the existence of u1 , v1 ∈ X satisfying
Φ(u1 ) − Ψ1 (u1 ) < ρ < Φ(v1 ) − Ψ1 (v1 )
and
sup(Φ−Ψ1 )−1 (]−∞,ρ]) Ψ2 − Ψ2 (u1 ) sup(Φ−Ψ1 )−1 (]−∞,ρ]) Ψ2 − Ψ2 (v1 )
< .
ρ − Φ(u1 ) + Ψ1 (u1 ) ρ − Φ(v1 ) + Ψ1 (v1 )
Likewise, condition (2.2) is equivalent to the existence of u1 , v1 ∈ X satisfying
Ψ2 (v1 ) < ρ < Ψ2 (u1 )
and
Φ(u1 ) − Ψ1 (u1 ) − infΨ−1 ([ρ,+∞[) (Φ − Ψ1 ) Φ(v1 ) − Ψ1 (v1 ) − infΨ−1 ([ρ,+∞[) (Φ − Ψ1 )
2 2
< .
Ψ2 (u1 ) − ρ Ψ2 (v1 ) − ρ
1022 S. SHOKOOH

Here and in the sequel, X will denote the space W 2,p (Ω) ∩W01,p (Ω) endowed with
the norm
Z  1p
p p p
kuk := (|∆u(x)| + |∇u(x)| + |u(x)| ) dx .
Ω
Put
maxx∈Ω |u(x)|
c = sup .
u∈X\{0} kuk
For p > max 1, N2 , since the embedding X ,→ C0 (Ω) is compact, one has c < +∞.


Let M : [0, +∞[→ R be a continuous function such that there is a positive constant m
with M(t) ≥ m for all t ≥ 0. Set
Z t
M̃(t) = (M(s)) p−1 ds for all t ≥ 0,
0
M − := min{1, m p−1 , η}.
The following lemma is useful in the proof of our results.
Lemma 1. If γ : Ω → [0, +∞[ be a non-zero function in L1 (Ω) and h : R → R a
continuous non-zero function, we denote by Tγ,h the functional defined on X by putting
Z
Tγ,h (u) = γ(x)h(u(x))dx
Ω
for all u ∈ X. Then, one has
 
Tγ,h (u) p h(ξ)
lim sup ≤ c kγkL1 (Ω) max 0, lim sup p (2.5)
u→0 kuk p ξ→0 |ξ|

and  
Tγ,h (u) p h(ξ)
lim sup p
≤ c kγkL1 (Ω) max 0, lim sup p . (2.6)
kuk→+∞ kuk |ξ|→+∞ |ξ|
Rξ
Also, the functional Tγ,H , where H is defined by H(ξ) = 0 h(t) dt for all ξ ∈ R, turns
out to be in C1 (X, R) and its derivative is given by
Z
0
Tγ,H (u)(v) = γ(x)h(u(x)) v(x)dx
Ω
0 :
for all u, v ∈ X. Moreover, the compact embedding of X in C0 (Ω̄) implies that Tγ,H
∗
X → X is compact.
Proof of the previous lemma is similar to proof of Lemma 2.1 in [5]. The following
lemma is Proposition 2.3 in [8], it is needed in the proof of Theorem 2.
Lemma 2. Let Λ : X → X ∗ be the operator defined by
Z
Λ(u)v = |∆u(x)| p−2 ∆u(x)∆v(x) dx
Ω
 A CLASS OF FOURTH-ORDER PROBLEMS WITH FOUR SOLUTIONS 1023

 Z  p−1 Z
p
+ M |∇u| dx |∇u(x)| p−2 ∇u(x)∇v(x) dx
Ω Ω
Z
+η |u(x)| p−2 u(x)v(x) dx
Ω
for every u, v ∈ X. Then Λ admits a continuous inverse on X ∗ .
We can deduce that u ∈ X is a weak solution to problem (1.1) if and only if u is a
critical point of the functional
Z 
1 1
Z Z
p p η
|∆u(x)| dx + M̃ |∇u| dx + |u(x)| p dx − λTa,F − Tb,G
p Ω p Ω p Ω
which represents, therefore, the energy functional related to problem (1.1).

3. M AIN RESULTS
Now, we present our main result.
Theorem 2. Let f , g : R → R be two continuous non-constant functions and let
a, b : Ω → [0, +∞[ be two non-constant functions in L1 (Ω). Assume that
Z ξ Z ξ
f (t) dt f (t) dt
( )
0 0
( f1 ) max lim sup , lim sup ≤ 0,
|ξ|→+∞ |ξ| p ξ→0 |ξ| p
Z ξ
g(t) dt
M−
Z ξ
0
(g1 ) supξ∈R g(t) dt < +∞, lim sup < .
0 ξ→0 |ξ| p pc p kbkL1 (Ω)
q  
1
p p
Finally, suppose that there exist σ > c M− max 1, (kbkL1 (Ω) supR G) p and
ξ1 ∈ R such that
Z ξ1 Z ξ Z ξ
( f2 ) 0 < f (t) dt = sup f (t) dt < sup f (t) dt,
0 |ξ|≤σ 0 ξ∈R 0
p
Z ξ1
(g2 ) |ξ1 | p ≤ kbkL1 (Ω) g(t) dt.
η|Ω| 0
Under such hypotheses, there exists λ∗ > 0 such that the problem
  Z  p−1
p−2 p
 ∆(|∆u| ∆u) − M |∇u| dx ∆ p u + η|u| p−2 u


Ω

 = λ∗ a(x) f (u) + b(x)g(u) in Ω,
u = ∆u = 0 on ∂Ω,


has at least three non-zero solutions, two of which are global minima of the associ-
ated energy functional.
1024 S. SHOKOOH

Proof. Our aim is to apply Theorem 1. Take Ψ1 , Ψ2 equal, respectively, to Tb,G , Ta,F
defined in Section 2. For each u ∈ X, let the functional Φ : X → R be defined by
Z 
1 1
Z Z
p p η
Φ(u) := |∆u(x)| dx + M̃ |∇u| dx + |u(x)| p dx.
p Ω p Ω p Ω
The function Φ is continuously differentiable whose differential at the point u ∈ X is
Z
Φ0 (u)h = |∆u(x)| p−2 ∆u(x)∆v(x) dx
Ω
 Z  p−1 Z
p
+ M |∇u| dx |∇u(x)| p−2 ∇u(x)∇v(x) dx
Ω Ω
Z
+η |u(x)| p−2 u(x)v(x) dx
Ω

for every v ∈ X, while Lemma 2 gives that Φ0 admits a continuous inverse on X ∗ .

Furthermore, Φ is sequentially weakly lower semicontinuous.
Now, take u0 = 0 and v0 = ξ1 ; of course, assumption (i) of Theorem 1 is evident.
Moreover, thanks to ( f2 ) and (g2 ), one has
Z ξ1
η
Φ(v0 ) = |ξ1 | p |Ω| ≤ kbkL1 (Ω) g(t) dt = Tb,G (v0 )
p 0

and
Z ξ1
Ta,F (v0 ) = kakL1 (Ω) f (t) dt > 0,
0
−
so assumption (ii) of Theorem 1 is satisfied. Now, since Mp kuk p ≤ Φ(u) for each
u ∈ X, we see that
Tb,G (u) pTb,G (u)
≤ − ,
Φ(u) M kuk p
and by Lemma 1, we get
Rξ
pc p
 
Tb,G (u) 0 g(t) dt
lim sup ≤ − kbkL1 (Ω) max 0, lim sup ≤ 0.
kuk→+∞ Φ(u) M |ξ|→+∞ |ξ| p

Also, in view of Lemma 1 and (g1 ), we have

Rξ
pc p
 
Tb,G (u) 0 g(t) dt
lim sup ≤ − kbkL1 (Ω) max 0, lim sup < 1.
u→0 Φ(u) M ξ→0 |ξ| p

So, (2.3) is fulfilled. Likewise, from Lemma 1 and ( f1 ), one has

Rξ
pc p
 
Ta,F (u) 0 f (t) dt
lim sup ≤ − kakL1 (Ω) max 0, lim sup ≤0
kuk→+∞ Φ(u) M |ξ|→+∞ |ξ| p
 A CLASS OF FOURTH-ORDER PROBLEMS WITH FOUR SOLUTIONS 1025

and
Rξ
pc p
 
Ta,F (u) 0 f (t) dt
lim sup ≤ − kakL1 (Ω) max 0, lim sup ≤ 0,
u→0 Φ(u) M ξ→0 |ξ| p
q
which satisfy (2.4). Finally, let us check that (2.1) holds. Since σ > c p Mp− , we see
that
σ pM−
1 − kbkL1 (Ω) sup G < − kbkL1 (Ω) sup G
R pc p R
and so it is possible to choose ρ ∈ R such that
 σ pM−
max 0, 1 − kbkL1 (Ω) sup G < ρ < − kbkL1 (Ω) sup G.
R pc p R

Let u ∈ X such that Φ(u) − Tb,G (u) ≤ ρ. For each u ∈ X, by the definition of σ and
the choice of ρ, we get
M−
kuk p ≤ Φ(u) ≤ ρ + kbkL1 (Ω) sup G
p R

and then, being ρ + kbkL1 (Ω) supR G > 1, we obtain

 1
p
p σ
kuk ≤ ρ + kbkL1 (Ω) sup G ≤ .
M− R c
Thus, owing to the embedding of X in C0 (Ω̄) we have the inclusion
 

u ∈ X : Φ(u) − Tb,G (u) ≤ ρ ⊆ u ∈ X : sup |u(x)| ≤ σ . (3.1)
x∈Ω̄

Now, in order to fulfil the equivalent formulation of (2.1) recalled in Remark 1,

choose u1 = v0 and take as v1 any constant d such that F(d) > sup[−σ,σ] F. Such a
d exists by ( f2 ). Thanks to (g2 ), we have
Z ξ1
η|Ω|
Φ(u1 ) − Tb,G (u1 ) ≤ |ξ1 | p − kbkL1 (Ω) g(t) dt ≤ 0 < ρ.
p 0

Moreover, Φ(v1 ) − Tb,G (v1 ) has to be necessarily strictly greater than ρ, otherwise,
by (3.1), we would have |d| ≤ σ and F(d) ≤ sup[−σ,σ] F, a contradiction. Then, due
to ( f2 ) and to the choice of d, we easily obtain that
sup Ta,F ≤ Ta,F (u1 )
(Φ−Tb,G )−1 (]−∞,ρ])

and
sup Ta,F ≤ Ta,F (v1 ).
(Φ−Tb,G )−1 (]−∞,ρ])
1026 S. SHOKOOH

Thus, the following inequalities hold

sup(Φ−Tb,G )−1 (]−∞,ρ]) Ta,F − Ta,F (u1 ) sup(Φ−Tb,G )−1 (]−∞,ρ]) Ta,F − Ta,F (v1 )
<0<
ρ − Φ(u1 ) + Tb,G (u1 ) ρ − Φ(v1 ) + Tb,G (v1 )
and, each assumption of Theorem 1 being satisfied, our problem admits at least three
non-zero weak solutions, two of which are global minima of the energy functional.

Remark 2. Suppose that α ≥ 0 and β > 0 be two real numbers and suppose M :
[α, β] ⊆ [0, +∞[→ R be a function defined by M(t) = θ1 + θ2t for each t ∈ [α, β]
where θ1 , θ2 > 0. So, M − = min{1, (θ1 + θ2 α) p−1 , η} and Theorem 2 holds for the
following problem
 Z
p−2

 ∆(|∆u| ∆u) − (θ1 + θ2 |∇u| p ) p−1 ∆ p u + η|u| p−2 u
Ω
 = λa(x) f (u) + b(x)g(u) in Ω,
u = ∆u = 0

on ∂Ω,
where f , g, a and b satisfy in the conditions of Theorem 2.
Remark 3. It is worth pointing out that, if assumption ( f2 ) of Theorem 2 is replaced
by the existence of
r  
p
p 1
σ>c max 1, (kbkL1 (Ω) sup G) p
M− R

and ξ1 , ξ2 ∈ R, with ξ1 ξ2 > 0, such that

Z ξ1 Z ξ Z ξ2
0< f (t) dt = sup f (t) dt < f (t) dt, (3.2)
0 |ξ|≤σ 0 0

M −
under the additional assumption |Ω| ≥ ηc p , we can ensure that the three non-zero

solutions of the thesis of Theorem 2 are non-negative (respectively non-positive)

provided ξ1 > 0 (respectively ξ2 < 0). To see this, it suffices to apply Theorem 2
to the functions f0 , g0 : R → R defined by
(
f (ξ) if ξ ≥ 0,
f0 (ξ) =
0 if ξ < 0,
(
g(ξ) if ξ ≥ 0,
g0 (ξ) =
0 if ξ < 0,
when ξ1 > 0 or by
(
f (ξ) if ξ ≤ 0,
f0 (ξ) =
0 if ξ > 0,
 A CLASS OF FOURTH-ORDER PROBLEMS WITH FOUR SOLUTIONS 1027

(
g(ξ) if ξ ≤ 0,
g0 (ξ) =
0 if ξ > 0,
when ξ1 < 0. In fact, conditions ( f1 ) and (g1 ) ensure that f (0) = 0 and g(0) = 0,
M−
namely, that f0 and g0 are continuous functions. Therefore, condition |Ω| ≥ ηc p,

guarantees that |ξ1 | ≤ σ, in fact by (g2 ) one has

pkbkL1 (Ω) supR G M−σ p
|ξ1 | p ≤ ≤ p ≤ σp.
η|Ω| ηc |Ω|
Hence, the function f0 satisfies in the assumption ( f2 ).
From Theorem 2 and Remark 3, we get:
M−
Corollary 1. Let |Ω| ≥ ηc p , and f : R → R be a continuous function such that
Z ξ
Z ξ f (t) dt
0
( f3 ) supξ∈R f (t) dt < +∞, lim sup ≤ 0.
0 ξ→0 |ξ| p
Moreover, suppose that there exist σ > 0 and ξ1 , ξ2 ∈ R, with ξ1 ξ2 > 0, such that
Z ξ1 Z ξ Z ξ2
( f4 ) 0 < f (t) dt = sup f (t) dt < f (t) dt
0 |ξ|≤σ 0 0

and
η|Ω||ξ1 | p M−σ p
( f5 ) R ξ < Rξ .
1
f (t) dt c p sup f (t) dt
0 ξ∈R 0
Under such hypotheses, for every non-constant function a : Ω → [0, +∞[ in L1 (Ω)
satisfying
( )
1 η|Ω||ξ1 | p M−σ p
( f6 ) max , Rξ ≤ kakL1 (Ω) < ,
supR F p 1 f (t) dt pc p sup
Rξ
f (t) dt
0 ξ∈R 0

there exists bλ > 1 such that the problem

  Z  p−1
p−2 p
 ∆(|∆u| ∆u) − M |∇u| dx ∆ p u + η|u| p−2 u


Ω


 = bλa(x) f (u) in Ω,
u = ∆u = 0 on ∂Ω
has at least three non-zero solutions which are non-negative or non-positive accord-
ing to whether ξ1 > 0 or ξ1 < 0.
Another consequence of Theorem 2 is as follows:
1028 S. SHOKOOH

Proposition 1. Let f : R → R be a continuous function and z1 , z2 , z3 , z4 four pos-

itive constants, with z1 < z2 < z3 < z4 such that f (ξ) ≥ 0 for all ξ ∈] − ∞, −z4 ] ∪
[−z3 , 0] ∪ [z1 , z2 ], while f (ξ) ≤ 0 for all ξ ∈] − z4 , −z3 ] ∪ [0, z1 ] ∪ [z2 , +∞[, and
Z z2 Z −z4
0< f (t) dt < f (t) dt.
0 0

Moreover, let g : R → R be a continuous function and let b : Ω → [0, +∞[ be a function

in L1 (Ω) such that

M−
Z z2 Rξ
g(t) dt
Z ξ
0
g(t) dt > 0, sup g(t) dt < +∞, lim sup <
0 ξ∈R 0 ξ→0 |ξ| p pc p kbkL1 (Ω)
and
η|Ω||z2 | p , M − z3p
 
1
max , R z2 ≤ kbkL1 (Ω) < .
supR G p 0 g(t) dt pc p supξ∈R
Rξ
g(t) dt
0

Under such hypotheses, for each non-zero function a : Ω → [0, +∞[ in L1 (Ω), there
exists λ∗ > 0 such that the problem
  Z  p−1
p−2 p
 ∆(|∆u| ∆u) − M |∇u| dx ∆ p u + η|u| p−2 u


Ω

 = λ∗ a(x) f (u) + b(x)g(u) in Ω,
u = ∆u = 0 on ∂Ω


has at least three non-zero solutions, two of which are global minima of the associ-
ated energy functional.
Rξ
Proof. The same assumption also implies that 0 f (t) dt ≤ 0 for all ξ ∈ [−z3 , z1 ],
and so ( f1 ) holds. Consequently, if we take ξ1 = z2 , the assumptions of Theorem 2
are satisfied and the conclusion follows. 

We now exhibit an example in which the hypotheses of Proposition 1 are satisfied.

Example 1. Suppose that M : [0, +∞[→ R is a continuous function such that
M − > 162 and suppose Ω = {(x, y) ∈ R2 ; x2 +y2 < 9}. Define
f , g : R → R as follows:
f (t) = −t 5 − 7t 4 + 7t 3 + 43t 2 − 42t,


 0 if t < 0,
t 2

if 0 ≤ t ≤ 2,
g(t) =

 −2t + 8 if 2 < t ≤ 4,

0 if t > 4.


Also, put a(x, y) = b(x, y) = x2 + y2 for all (x, y) ∈ Ω.

 A CLASS OF FOURTH-ORDER PROBLEMS WITH FOUR SOLUTIONS 1029
q
3 4
Then, by choosing z1 = 1, z2 = 2, z3 = 3, z4 = 7, p = 3, and since c = π, the
conclusion of Proposition 1 holds for the following problem
  Z 2
 ∆(|∆u|∆u) − M 3
|∇u| dxdy

 ∆3 u + η|u|u
Ω

 = λ∗ a(x, y) f (u) + b(x, y)g(u) in Ω,
u = ∆u = 0 on ∂Ω.


R EFERENCES
[1] G. Afrouzi and S. Shokooh, “Existence of infinitely many solutions for quasilinear problems with
a p(x)-biharmonic operator,” Electron. J. Diff. Eqs., vol. 2015, no. 317, pp. 1–14, 2015.
[2] C. Alves, F. Corrêa, and T. Ma, “Positive solutions for a quasilinear elliptic equation of Kirchhoff
type,” Comput. Math. Appl., vol. 49, pp. 85–93, 2005, doi: 10.1016/j.camwa.2005.01.008.
[3] A. Bensedik and M. Bouchekif, “On an elliptic equation of Kirchhoff-type with a potential
asymptotically linear at infinity,” Math. Comput. Modelling, vol. 49, pp. 1089–1096, 2009, doi:
10.1016/j.mcm.2008.07.032.
[4] H. Berger, “A new approach to the analysis of large deflections of plates,” J. Appl. Mech., vol. 22,
pp. 465–472, 1995.
[5] F. Cammaroto and L. Vilasi, “Multiplicity results for a Neumann boundary value problem in-
volving the p(x)-Laplacian,” Taiwanese J. Math., vol. 16, no. 2, pp. 621–634, 2012, doi:
10.11650/twjm/1500406606.
[6] G. Cordaro, “On a minimax problem of Ricceri,” J. Inequal. Appl., vol. 6, pp. 261–285, 2001, doi:
10.1155/S1025583401000169.
[7] F. Corrêa and G. Figueiredo, “On an elliptic equation of p-Kirchhoff type via variational methods,”
Bull. Aust. Math. Soc., vol. 74, no. 02, pp. 263–277, 2006, doi: 10.1017/S000497270003570X.
[8] S. Heidarkhani, S. Khademloo, and A. Solimaninia, “Multiple solutions for a perturbed fourth-
order Kirchhoff type elliptic problem,” Portugal. Math., vol. 1, no. 1, pp. 39–61, 2014, doi:
10.4171/PM/1940.
[9] G. Kirchhoff, Mechanik. Leipzig, Germany: Teubner, 1876.
[10] C. Li and C. Tang, “Three solutions for a Navier boundary value problem involving the p-
biharmonic,” Nonlinear Anal., vol. 72, pp. 1339–1347, 2010, doi: 10.1016/j.na.2009.08.011.
[11] T. Ma, “Existence results for a model of nonlinear beam on elastic bearings,” Appl. Math. Lett.,
vol. 13, pp. 11–15, 2000, doi: 10.1016/S0893-9659(00)00026-4.
[12] T. Ma, “Remarks on an elliptic equation of Kirchhoff type,” Nonlinear Anal., vol. 63, pp. 1967–
1977, 2005, doi: 10.1016/j.na.2005.03.021.
[13] T. Ma, “Positive solutions for a nonlocal fourth order equation of Kirchhoff type,” Discrete Contin.
Dyn. Syst, pp. 694–703, 2007, doi: 10.1016/S0893-9659(00)00026-4.
[14] B. Ricceri, “A further three critical points theorem,” Nonlinear Anal., vol. 71, pp. 4151–4157,
2009, doi: 10.1016/j.na.2009.02.074.
[15] B. Ricceri, “A class of nonlinear eigenvalue problems with four solutions,” J. Nonlinear Convex
Anal., vol. 11, no. 3, pp. 503–511, 2010.
[16] B. Ricceri, “On an elliptic Kirchhoff-type problem depending on two parameters,” J. Global Op-
timization, vol. 46, p. 543–549, 2010, doi: 10.1007/s10898-009-9438-7.
[17] B. Ricceri, “Another four critical points theorem,” J. Nonlinear Convex Anal., vol. 2, pp. 163–171,
2011.
[18] S. Shokooh, “Multiplicity results for a non-homogeneous Neumann problem via a variational
principle of Ricceri,” Minimax Theory Appl., vol. 2, no. 2, pp. 249–263, 2017.
1030 S. SHOKOOH

[19] J. Sun and C. Tang, “Existence and multiplicity of solutions for Kirchhoff type equations,” Non-
linear Anal., vol. 74, no. 4, pp. 1212–1222, 2011, doi: 10.1016/j.na.2010.09.061.
[20] S. Woinowsky-Krieger, “The effect of an axial force on the vibration of hinged bars,” J. Appl.
Mech., vol. 17, pp. 35–36, 1950.
[21] M. Yang and Z. Han, “Existence and multiplicity results for Kirchhoff type problems with
four superlinear potential,” Applicable Anal., vol. 91, no. 11, pp. 2045–2055, 2012, doi:
10.1080/00036811.2011.587808.
[22] H. Yin and Y. Liu, “Existence of three solutions for a Navier boundary value problem involving
the p(x)-biharmonic,” Bull. Korean. Math. Soc., vol. 50, no. 6, pp. 1817–1826, 2013, doi:
10.4134/BKMS.2013.50.6.1817.
[23] H. Yin and M. Xu, “Existence of three solutions for a Navier boundary value problem involving
the p(x)-biharmonic operator,” Ann. Polon. Math., vol. 109, pp. 47–54, 2013, doi: 10.4064/ap109-
1-4.
[24] Z. Zhang and K. Perera, “Existence of three solutions for a Navier boundary value problem in-
volving the p(x)-biharmonic operator,” J. Math. Anal. Appl., vol. 317, no. 2, pp. 456–463, 2006,
doi: 10.1016/j.jmaa.2005.06.102.

Author’s address
S. Shokooh
Department of Mathematics, Faculty of Basic Sciences, Gonbad Kavous University, Gonbad Ka-
vous, Iran
E-mail address: [email protected]