Communications

in
Mathematical
Research
27(1)(2011), 37–46

A Pseudo-parabolic Type Equation with

Nonlinear Sources∗
Li Ying-hua1 , Cao Yang2 and Wang Yi-fu3
(1. School of Mathematical Sciences, South China Normal University, Guangzhou, 510631)
(2. School of Mathematical Sciences, Dalian University of Technology, Dalian, 116024)
(3. Department of Mathematics, Beijing Institute of Technology, Beijing, 100081)

Communicated by Yin Jing-xue

Abstract: This paper is concerned with the existence and uniqueness of nonnegative
classical solutions to the initial-boundary value problems for the pseudo-parabolic
equation with strongly nonlinear sources. Furthermore, we discuss the asymptotic
behavior of solutions as the viscosity coefficient k tends to zero.
Key words: pseudo-parabolic equation, existence, uniqueness, asymptotic behavior
2000 MR subject classification: 35G30, 35B40
Document code: A
Article ID: 1674-5647(2011)01-0037-10

1 Introduction
In this paper, we investigate the existence, uniqueness and asymptotic behavior of solutions
to the following initial-boundary value problem for the pseudo-parabolic equation in one
spatial dimension:
∂u ∂D2 u
−k = D2 u + m(x, t)uq , (x, t) ∈ Q, (1.1)
∂t ∂t
u(0, t) = u(1, t) = 0, t ≥ 0, (1.2)
u(x, 0) = u0 (x), x ∈ [0, 1], (1.3)
+
where q > 1, Q ≡ (0, 1) × R , D = ∂/∂x, k > 0 represents the viscosity coefficient,
m(x, t) ∈ C α,α/3 (Q) for some α ∈ (0, 1) and satisfies 0 < m ≤ m(x, t) ≤ m for any
(x, t) ∈ Q, m and m are positive constants.
The pseudo-parabolic equations are characterized by the occurrence of mixed third order
derivatives, more precisely, second order in space and first order in time. Such equations
are used to model heat conduction in two-temperature systems (see [1] and [2]), fluid flow
in porous media (see [3] and [4]), two phase flow in porous media with dynamical capillary
∗ Received date: April 18, 2009.
Foundation item: The NSFC, CPSF, SRFDP and 973 Program (2010CB808002).
38 COMM. MATH. RES. VOL. 27

pressure (see [5]), and the populations with the tendency to form crowds (see [6] and [7]).
Furthermore, according to experimental results, some researchers have recently proposed
modifications to Cahn’s model which incorporate out-of-equilibrium viscoelastic relaxation
effects, and thus obtained this type of equations (see [8]). The pseudo-parabolic equations
with strongly nonlinear sources we considered may arise in the study of nonstationary pro-
cesses in semiconductors with sources of free-charge currents (see [9] and [10]). The processes
can be described by the system that has the explicit form of the constitutive equations con-
necting the electric field strength E with the electric flux density G on the one hand, and
the electric field strength E with the current density J in the semiconductor on the other
hand, as follows:
divG = 4πen, E = −∇u, G = E + 4πP,
∂n
= −divJ + Q, Ji = σi Ei , i = 1, 2, · · · , N,
∂t
where P is the polarization vector and in some models there has the following phenomeno-
logical relation divP = k1 u, k1 > 0, n is the free electron concentration, u is the electric
potential, and σi is the conductivity tensor. Finally, assume that, in a semiconductor, there
are sources of free-charge currents whose distribution in the self-consistent electric field of
the semiconductor is of the form Q = m(x, t)up . By differentiating both sides of the first
equation with respect to t and taking account of the second equation, the above system can
be reduced to the equation (1.1).
The pseudo-parabolic equations have been extensively investigated. In [11]–[13], the
authors investigated the initial-boundary value problem and the Cauchy problem for the
linear pseudo-parabolic equation and established the existence and uniqueness of solutions.
The nonlinear pseudo-parabolic type equations with undefined or uninvertible operator at
the highest derivative with respect to time were studied in [14]. The degenerate and quasi-
linear degenerate pseudo-parabolic type equations were investigated in [15] and [16]. For
the local solvability of the pseudo-parabolic type equations with variety nonlocal boundary
conditions, see [17]–[19].
For pseudo-parabolic equations, classical maximum principle is invalid in general. For
the nonnegativity of a solution, not only nonnegative initial data, but also an extra condition
on the elliptic operator is needed (see [20]–[22]). Due to the special type of the problem
(1.1)–(1.3) which is included in the studies of [22], we can prove the comparison principle of
solutions, which enables us to obtain the existence of nonnegative solutions to the problem
(1.1)–(1.3). For the asymptotic behavior of solutions, we know that in certain cases, the
solution of a parabolic initial-boundary value problem can be obtained as a limit of solutions
to the problem of the corresponding pseudo-parabolic equations, see [11]. In this paper, we
show that the semilinear pseudo-parabolic equations still retain this character, namely, the
solutions of the pseudo-parabolic equations converge to the solution of the parabolic equation
as k → 0.
This paper is organized as follows. In Section 2, we show the existence and uniqueness
of nonnegative classical solutions to the initial-boundary value problem (1.1)–(1.3). Then,
NO. 1 LI Y. H. et al. A PSEUDO-PARABOLIC TYPE EQUATION 39

in Section 3, we discuss the asymptotic behavior of solutions as the viscosity coefficient k

tends to zero.

2 The Initial-boundary Value Problem

In this section, we discuss the solvability of the initial-boundary value problem (1.1)–(1.3).
Firstly, we prove the uniqueness of the solution.

Theorem 2.1 If the initial-boundary value problem (1.1)–(1.3) admits classical solutions,
then for any given initial datum u0 ∈ C 2+α [0, 1] with u0 (0) = u0 (1) = 0, the solution of the
problem (1.1)–(1.3) is unique.

Proof. Let u1 , u2 be two solutions of the problem (1.1)–(1.3). Set v = u1 − u2 . Then we

have
∂v ∂D2 v q−1
−k = D2 v + qm(x, t) (u1 + θ(u2 − u1 )) v, (x, t) ∈ QT ,
∂t ∂t
v(0, t) = v(1, t) = 0, t ∈ [0, T ],
v(x, 0) = 0, x ∈ [0, 1],
where θ ∈ (0, 1), T > 0 is a given constant, and QT = (0, 1) × (0, T ). Multiplying the first
equation by v, we get
1 d 1 2
Z Z 1
v + k|Dv|2 dx + |Dv|2 dx


2 dt 0 0
Z 1
= qm(x, t)(u1 + θ(u2 − u1 ))q−1 v 2 dx.
0
It follows that
1 1
d
Z Z
v 2 + k|Dv|2 dx ≤ C v 2 dx


dt 0 0
Z 1
≤C (v 2 + k|Dv|2 )dx, t ∈ (0, T ).
0
Noticing that
v(x, 0) = Dv(x, 0) = 0,

by the Gronwall inequality, we obtain

Z 1
v 2 (x, t) + k|Dv(x, t)|2 dx ≤ 0,


t ∈ (0, T ).
0
Consequently, Z 1
v 2 (x, t)dx ≤ 0, t ∈ (0, T ),
0
which implies that u1 = u2 a.e. in QT . The proof of this theorem is completed.

Next, we prove that the solutions of the problem (1.1)–(1.3) admit comparison principle.
DiBenedetto and Pierre[22] discussed the maximum principle for pseudo-parabolic equations
including the equation
∂
(u − k∆u) − ∆u = f, (x, t) ∈ Ω × [0, T ]
∂t
40 COMM. MATH. RES. VOL. 27

subject to the conditions

u|∂Ω = g(t), u(0) = u0 ,

where Ω is a bounded open set in Rn , and T > 0 is a given constant. Denote

E = {u ∈ H 2 (Ω ); u0 |∂Ω = g(0)}.
Their argument contains the following conclusion (see [22], P.14).

Lemma 2.1 The following propositions are equivalent:

(i) For any u0 ∈ E, f ∈ L1 (0, T ; L2(Ω )),
u0 ≥ 0, f (t) ≥ 0 a.e. t ∈ [0, T ] ⇒ uk (t, u0 , f ) ≥ 0, t ∈ [0, T ].
(ii) For any t ∈ [0, T ], g(t) ≥ e−t/k g(0).

In virtue of that the boundary value conditions (1.2) imply g(t) = 0 for any t ∈ [0, T ],
the proposition (ii) of Lemma 2.1 holds. Therefore, we obtain the comparison principle as
follows.

Proposition 2.1 Let u1 and u2 be two solutions of the initial-boundary value problem
(1.1)–(1.3) in QT with initial data u10 and u20 respectively, u10 , u20 ∈ C 2+α [0, 1]. If u10 ≥
u20 ≥ 0 in (0, 1), then u1 ≥ u2 ≥ 0 in QT .

Proof. We first prove that u1 and u2 are nonnegative. For this purpose, we consider the
following problem:
∂ û1 ∂D2 û1
−k = D2 û1 + m(x, t)(û1+ )q , (x, t) ∈ QT , (2.1)
∂t ∂t
û1 (0, t) = û1 (1, t) = 0, t ∈ [0, T ],
û1 (x, 0) = u1 0 (x), x ∈ [0, 1],
where s+ = max{s, 0}. Since
m(x, t)(û1+ )q ≥ 0, u1 0 (x) ≥ 0,
and the zero boundary value conditions admit the proposition (ii) of Lemma 2.1, by virtue
of Lemma 2.1 we see that û1 ≥ 0 in QT . Hence, û1+ in the equation (2.1) is û1 itself.
Because the solution of the problem (1.1)–(1.3) is unique, we have
u1 = û1 ≥ 0 in QT .
Similarly, we can also prove
u2 ≥ 0 in QT .

Set
v = u1 − u2 .

Then v satisfies
∂v ∂D2 v
−k = D2 v + qm(x, t)(θu1 + (1 − θ)u2 )q−1 v, (x, t) ∈ QT , (2.2)
∂t ∂t
v(0, t) = v(1, t) = 0, t ∈ [0, T ], (2.3)
v(x, 0) = u10 (x) − u20 (x), x ∈ [0, 1], (2.4)
NO. 1 LI Y. H. et al. A PSEUDO-PARABOLIC TYPE EQUATION 41

where θ ∈ (0, 1). Consider the following equation:

∂v̂ ∂D2 v̂
−k = D2 v̂ + qm(x, t)(θu1 + (1 − θ)u2 )q−1 v̂+ , (x, t) ∈ QT . (2.5)
∂t ∂t
Noticing that u1 ≥ 0, u2 ≥ 0, u1 0 (x) ≥ u2 0 (x) and the boundary value of v equals to zero, by
Lemma 2.1 we have the solution of the problem (2.5)–(2.3)–(2.4) v̂ ≥ 0 in QT . Consequently,
we see that v̂+ = v̂ in the equation (2.5). Similarly to the proof of Theorem 2.1, we can
prove that the solution of the problem (2.2)–(2.4) is unique. Therefore, the solution of the
problem (2.2)–(2.4) v = v̂ ≥ 0 in QT , i.e., u1 ≥ u2 in QT . The proof of this proposition is
completed.

In what follows, we prove the existence of solutions to the problem (1.1)–(1.3). For this
purpose, we first construct a uniformly bounded supersolution of the problem (1.1)–(1.3),
which enables us to obtain the upper bound of solutions. The definition of a supersolution
is as follows.
Definition 2.1 A function u is called a supersolution of the initial-boundary value prob-
lem (1.1)–(1.3) provided that
∂u ∂D2 u
−k ≥ D2 u + m(x, t)uq , (x, t) ∈ Q,
∂t ∂t
u(0, t) ≥ 0, u(1, t) ≥ 0, t ≥ 0,
u(x, 0) ≥ u0 (x), x ∈ [0, 1].
Let u(x, t) = α(2−x2 ), where α is a positive constant to be determined. Then, u satisfies
∂u ∂D2 u
−k − D2 u − m(x, t)uq = 2α − m(x, t)αq (2 − x2 )q , (x, t) ∈ Q,
∂t ∂t
u(0, t) = 2α, u(1, t) = α, t ≥ 0,
u(x, 0) = α(2 − x2 ), x ∈ [0, 1].
 1
 q−1
1 1
We choose α to satisfy 0 < α ≤ , where m is the upper bound of m(x, t). When
2 m
  q−1 1
1 1
max u0 ≤ , it is easy to verify that u is a supersolution of the problem (1.1)–(1.3).
(0,1) 2 m
If u is the solution of the problem (1.1)–(1.3), and set v = u − u, then we have
∂v ∂D2 v
−k − D2 v ≥ m(x, t)(θu + (1 − θ)u)q−1 v, (x, t) ∈ QT ,
∂t ∂t
v(0, t) = 2α, v(1, t) = α, t ∈ [0, T ],
v(x, 0) = α(2 − x2 ) − u0 (x) ≥ 0, x ∈ [0, 1].
Noticing that v(0, t) and v(1, t) satisfy the proposition (ii) of Lemma 2.1, similar to the
proof of Proposition 2.1, we can obtain v(x, t) ≥ 0 in QT , i.e., u(x, t) ≥ u(x, t) ≥ 0 in QT .
Consequently, the following lemma holds.

Lemma 2.2 For any given initial datum u0 ∈ C 2+α [0, 1] with u0 (0) = u0 (1) = 0, u0 ≥ 0,
if u0 is appropriately small, u is the solution of the initial-boundary value problem (1.1)–
(1.3), then
0 ≤ u(x, t) ≤ M2 , (x, t) ∈ QT , (2.6)
42 COMM. MATH. RES. VOL. 27

where M2 is a positive constant independent of u and k.

By means of the above lemma, we show the global existence of solutions to the problem
(1.1)–(1.3).

Theorem 2.2 For any given initial datum u0 ∈ C 2+α [0, 1] with u0 (0) = u0 (1) = 0, u0 ≥
0, if u0 is appropriately small, then the problem (1.1)–(1.3) admits at least one nonnegative
classical solution.

Proof. We prove this theorem by employing the Leray-Schauder fixed point theorem.
Consider a family of relevant equations with parameter, namely
∂u ∂D2 u
−k = D2 u + σm(x, t)uq , (x, t) ∈ QT (2.7)
∂t ∂t
subject to the conditions
u(0, t) = u(1, t) = 0, t ∈ [0, T ], (2.8)
u(x, 0) = σu0 (x), x ∈ [0, 1], (2.9)
where σ ∈ [0, 1] is a parameter. Here and below, we denote by C a constant, whose value
may be different from line to line and is independent of u and σ. Multiplying (2.7) by u and
integrating the result over Qt , we have
1 1 2 k 1
Z Z ZZ
2
u (x, t)dx + |Du(x, t)| dx + |Du|2 dxds
2 0 2 0 Qt

σ2 1 2
Z 1
k
Z ZZ
= u0 (x)dx + σ 2 |Du0 (x)|2 dx + σ m(x, s)uq+1 dxds.
2 0 2 0 Qt
By (2.6), we obtain
Z 1
|Du(x, t)|2 dx ≤ C, t ∈ (0, T ), (2.10)
0
ZZ
|Du|2 dxdt ≤ C. (2.11)
QT
Multiplying (2.7) with D2 u and integrating over Qt , we get
1 1 k 1 2
Z Z ZZ
2 2
|Du(x, t)| dx + |D u(x, t)| dx + |D2 u|2 dxds
2 0 2 0 Qt

σ2 1
Z 1
k
Z ZZ
= |Du0 (x)|2 dx + σ 2 |D2 u0 (x)|2 dx − σ m(x, s)uq D2 udxds
2 0 2 0 Qt
1
ZZ
≤ |D2 u|2 dxds + C,
2 Qt
which implies that
Z 1
|D2 u(x, t)|2 dx ≤ C, t ∈ (0, T ), (2.12)
0
ZZ
|D2 u|2 dxdt ≤ C. (2.13)
QT
Combining (2.10) with (2.12), we conclude that
kDukL∞ (QT ) ≤ C. (2.14)
NO. 1 LI Y. H. et al. A PSEUDO-PARABOLIC TYPE EQUATION 43

∂u
Multiplying (2.7) with and integrating over Qt yield
∂t
2 2
∂u ∂Du 1 1
ZZ ZZ Z
dxds + k dxds + |Du(x, t)|2 dx
Qt ∂t Qt ∂t 2 0
σ2 1 ∂u
Z ZZ
= |Du0 (x)|2 dx + σ m(x, t)uq dxds
2 0 Qt ∂t
2
1 ∂u
ZZ
≤ dxds + C.
2 Qt ∂t
It follows that
2
∂u
ZZ
dxdt ≤ C. (2.15)
QT ∂t
We rewrite the equation (2.7) into the form
∂D2 u 1 ∂u 1 2 σ
= − D u − m(x, t)uq .
∂t k ∂t k k
Then, we have
2
∂D2 u
ZZ
dxdt ≤ C. (2.16)
QT ∂t
In the following, we claim that
|u(x1 , t1 ) − u(x2 , t2 )| ≤ C(|x1 − x2 | + |t1 − t2 |1/3 ) (2.17)
for any (xi , ti ) ∈ QT (i = 1, 2). It is obvious that the above inequality is equivalent to
|u(x1 , t) − u(x2 , t)| ≤ C|x1 − x2 |, t ∈ [0, T ], x1 , x2 ∈ [0, 1], (2.18)
1/3
|u(x, t1 ) − u(x, t2 )| ≤ C|t1 − t2 | , x ∈ [0, 1], t1 , t2 ∈ [0, T ]. (2.19)
We can easily obtain (2.18) from (2.14). To prove (2.19), we need only to consider the case
of 0 ≤ x ≤ 1/2. Let ∆t = t2 − t1 > 0 satisfying (∆t)1/3 ≤ 1/4. Integrating (2.7) over
(y, y + (∆t)1/3 ) × (t1 , t2 ) gives
Z y+(∆t)1/3
(u(z, t2 ) − u(z, t1 ))dz
y
Z y+(∆t)1/3 Z t2 Z t2
=k D2 ut dtdz + (Du(y + (∆t)1/3 , t) − Du(y, t))dt
y t1 t1
Z y+(∆t)1/3 Z t2
+σ m(z, t)uq dtdz.
y t1
Integrating the above equality with respect to y over (x, x + (∆t)1/3 ), by (2.6), (2.14), (2.16)
and the mean value theorem, we have
|u(x∗ , t2 ) − u(x∗ , t1 )| ≤ C|t2 − t1 |1/3 ,
where
x∗ = y ∗ + θ∗ (∆t)1/3 , y ∗ ∈ (x, x + (∆t)1/3 ), θ∗ ∈ (0, 1).

Combining the above inequality with (2.18), we arrive at

|u(x, t1 ) − u(x, t2 )| ≤ |u(x, t1 ) − u(x∗ , t1 )| + |u(x∗ , t1 ) − u(x∗ , t2 )| + |u(x, t2 ) − u(x∗ , t2 )|
≤ C|t1 − t2 |1/3 .
44 COMM. MATH. RES. VOL. 27

Now, we define a linear space X by

X = {u ∈ C 1,1/3 (QT ) : u(0, t) = u(1, t) = 0, ∀t ∈ [0, T ]},
and an operator
F : X −→ X, u 7−→ w,

where w is a solution of the following linear problem:

∂w ∂D2 w
−k = D2 w + m(x, t)uq , (x, t) ∈ QT ,
∂t ∂t
w(0, t) = w(1, t) = 0, t ∈ [0, T ],
w(x, 0) = u0 (x), x ∈ [0, 1].
By the classical theory (see [23]), the above linear problem admits a unique solution w ∈
C 2+β,1+β/3 (QT ), β ∈ (0, 1). So, the operator F is well-defined. We can also obtain that the
operator F is compact by means of the compact embedding theorem. Moreover, if u = σF u
holds for some σ ∈ [0, 1], then u satisfies the problem (2.7)–(2.9). Clearly σ = 0 implies
u ≡ 0. If σ 6= 0, from the above argument we see that kukC 1,1/3 (QT ) is bounded and the
bound is independent of u and σ. By the Leray-Schauder fixed point theorem, the operator
F has a fixed point u, which is the desired classical solution of the problem (1.1)–(1.3) in
QT . As above, we consider the problem in Q(T,2T ) , Q(2T,3T ) , · · · , Q((n−1)T,nT ) in turn.
Then, we infer that the problem (1.1)–(1.3) admits a classical solution in Q. The proof of
this theorem is completed.

3 Asymptotic Behavior
In this section, we discuss the asymptotic behavior of solutions as the viscosity coefficient
k tends to zero. Here, we denote by C a constant independent of u and k, and by C(k) a
constant independent of u.

Theorem 3.1 If uk is a nonnegative classical solution of the initial-boundary value prob-

lem (1.1)–(1.3), then uk (x, t) is uniformly convergent in QT as k → 0, and the limit function
u(x, t) is a nonnegative classical solution of the following initial-boundary value problem:
∂u
= D2 u + m(x, t)uq , (x, t) ∈ QT , (3.1)
∂t
u(0, t) = u(1, t) = 0, t ∈ [0, T ], (3.2)
u(x, 0) = u0 (x), x ∈ [0, 1]. (3.3)

Proof. From Lemma 2.2, we see that

kuk kL∞ (QT ) ≤ M2 .
2
Multiplying (1.1) by D uk and integrating over Qt , we have
Z 1
|Duk (x, t)|2 dx ≤ C + C(k), t ∈ (0, T ), (3.4)
0
ZZ
|D2 uk |2 dxdt ≤ C + C(k). (3.5)
QT
NO. 1 LI Y. H. et al. A PSEUDO-PARABOLIC TYPE EQUATION 45

∂uk
Multiplying (1.1) by and integrating over QT , we get
∂t
2
∂uk
ZZ
dxdt ≤ C. (3.6)
QT ∂t
Rewrite the equation (1.1) into the following form:
∂D2 uk ∂uk
k = − D2 uk − m(x, t)uqk .
∂t ∂t
Then, we arrive at
2
∂D2 uk
ZZ
k dxdt ≤ C + C(k). (3.7)
QT ∂t
By a proof similar to that of Theorem 2.2, from (3.4)–(3.7), we can obtain
|uk (x1 , t1 ) − uk (x2 , t2 )| ≤ (C + C(k))(|x1 − x2 |1/2 + |t1 − t2 |1/6 ) (3.8)
2,1
for any (xi , ti ) ∈ QT (i = 1, 2). Therefore, there exists a function u ∈ H (QT ) ∩
C 1/2,1/6 (QT ) such that
∂uk ∂u
uk → u uniformly in QT , ⇀ , D2 uk ⇀ D2 u in L2 (QT ), (3.9)
∂t ∂t
as k → 0. Recalling the equation (1.1), we see that for any ϕ ∈ C 2 (QT ) satisfying ϕ(0, t) =
ϕ(1, t) = 0, we have
∂uk ∂uk 2
ZZ ZZ
ϕdxdt − k D ϕdxdt
QT ∂t QT ∂t
ZZ ZZ
= D2 uk ϕdxdt + m(x, t)uqk ϕdxdt.
QT QT
Taking k → 0, by (3.9), we have
∂u
ZZ ZZ ZZ
2
ϕdxdt = D uϕdxdt + m(x, t)uq ϕdxdt,
QT ∂t QT QT
which implies that u satisfies the equation (3.1) in the sense of distribution. Noticing that
u ∈ C 1/2,1/6 (Qω ), by the classical theory (see [23]) we can deduce that u ∈ C 2+β,1+β/3 (Qω ),
β ∈ (0, 1/2). It is obvious that u is nonnegative and satisfies the conditions (3.2) and (3.3).
Hence, u(x, t) is a nonnegative classical solution of the problem (3.1)–(3.3). The proof of
this theorem is completed.

References
[1] Chen, P. J. and Gurtin, M. E., On a theory of heat conduction involving two temperatures,
Z. Angew. Math. Phys., 19(1968), 614–627.
[2] Ting, T. W., A cooling process according to two-temperature theory of heat conduction, J.
Math. Anal. Appl., 45(1974), 23–31.
[3] Barenblatt, G. I., Entov, V. M. and Ryzhik, V. M., Theory of Fluid Flow Through Natural
Rocks, Kluwer, Dordrecht, 1990.
[4] Barenblatt, G. I., Zheltov, I. and Kochina, I. N., Basic concepts in the theory of seepage of
homogeneous liquids in fissured rocks, J. Appl. Math. Mech., 24(1960), 1286–1303.
[5] Cuesta, C., van Duijn, C. J. and Hulshof, J., Infiltration in porous media with dynamic
capillary pressure: Traveling waves, European J. Appl. Math., 4(2000), 381–397.
[6] Padrón, V., Sobolev regularization of a nonlinear ill-posed parabolic problem as a model for
aggregating populations, Comm. Partial Differential Equations, 23(1998), 457–486.
46 COMM. MATH. RES. VOL. 27

[7] Padrón, V., Effect of aggregation on population recovery modeled by a forward-backward

pesudoparabolic equation, Trans. Amer. Math. Soc., 356(2004), 2739–2756.
[8] Novick-Cohen, A. and Pego, R. L., Stable patterns in a viscous diffusion equation, Trans.
Amer. Math. Soc., 324(1991), 331–351.
[9] Bonch-Bruevich, V. L. and Kalashnikov, S. G., The Physics of Semiconductors [in Russian],
Nauka, Moscow, 1990.
[10] Korpusov, M. O. and Sveshnikov, A. G., On the “blow-up” of the solution of a strongly non-
linear equation of pseudoparabolic type with a double nonlinearity (Russian), Mat. Zametki,
79(6)(2006), 879–899; translation in Math. Notes, 79(5-6)(2006), 820–840.
[11] Ting, T. W., Parabolic and pseudo-parabolic partial differential equations, J. Math. Soc. Japan,
21(1969), 440–453.
[12] Showalter, R. E. and Ting, T. W., Pseudoparabolic partial differential equations, SIAM J.
Math. Anal., 1(1970), 1–26.
[13] Gopala Rao V. and Ting, T. W., Solutions of pseudo-heat equations in the whole space, Arch.
Rational Mech. Anal., 49(1972/73), 57–78.
[14] Egorov, I. E., Pyatkov, S. G. and Popov, S. V., Nonclassical Operator-Differential Equations,
Nauka, Novosibirsk, 2000.
[15] Ptashnyk, M., Degenerate quasilinear pseudoparabolic equations with memory terms and
vairational inequalities, Nonlinear Anal., 66(2007), 2653–2675.
[16] Scarpini, F., Degenerate and pseudoparabolic variational inequalities, Numer. Funct. Anal.
Optim., 9(1987), 859–879.
[17] Mesloub, S., A nonlinear nonlocal mixed problem for a second order pseudoparabolic equation,
J. Math. Anal. Appl., 316(2006), 189–209.
[18] Dai, D. Q. and Huang, Y., On a nonlocal problem for nonlinear pseudoparabolic equations,
Nonlinear Anal., 64(2006), 499–512.
[19] Bouziani, A., Initial-boundary value problems for a class of pseudoparabolic equations with
integral boundary conditions, J. Math. Anal. Appl., 291(2004), 371–386.
[20] Rundell, W., and Stecher, M., The nonpositivity of solutions to pseudoparabolic equations,
Proc. Amer. Math. Soc., 75(1979), 251–254.
[21] Showalter, R. E., Local regularity, boundary values and maximum principles for pseu-
doparabolic equations, Appl. Anal., 16(1983), 235–241.
[22] DiBenedetto, E. and Pierre, M., On the maximum principle for pseudoparabolic equations,
Indiana Univ. Math. J., 6(30)(1981), 821–854.
[23] Ladyzenskaja, O., Solonnikov, V. and Uraltseva, N., Linear and Quasilinear Equations of
Parabolic Type, Translations of Mathematical Monographs, vol. 23, Amer. Math. Soc., Prov-
idence, 1968.