Nonlinear Analysis: Real World Applications 11 (2010) 3744–3751

Contents lists available at ScienceDirect

Nonlinear Analysis: Real World Applications

journal homepage: www.elsevier.com/locate/nonrwa

Exact solutions for generalized Maxwell fluid flow due to oscillatory and
constantly accelerating plate
Liancun Zheng a,∗ , Fangfang Zhao a , Xinxin Zhang b
a
Department of Mathematics and Mechanics, University of Science and Technology Beijing, Beijing 100083, China
b
Mechanical Engineering School, University of Science and Technology Beijing, Beijing 100083, China

article info abstract

Article history: This paper deals with the analytical solutions for generalized Maxwell fluid flow due to
Received 20 July 2009 oscillatory and constantly accelerating plate. The fractional calculus approach is used to
Accepted 3 February 2010 establish the constitutive relationship of fluid model. Exact solutions are presented for the
velocity field and the corresponding shear stress in series forms in terms of generalized G
Keywords: and R functions by using the discrete inverse Laplace transform method.
Generalized Maxwell fluid
© 2010 Elsevier Ltd. All rights reserved.
Exact solutions
Laplace transform
G and R functions

1. Introduction

Recently, considerable attention has been devoted to the problem of how to predict the behavior of non-Newtonian
fluids. The main reason for this is probably that fluids (such as molten plastics, pulps, slurries, emulsions, petroleum
drilling, manufacturing of food and other similar activities), which do not obey the Newtonian postulate that the stress
tensor is directly proportional to the rate of deformation tensor, are produced industrially in increasing quantities. The flow
characteristics of non-Newtonian fluids are quite different from the Newtonian fluids, numbers of constitutive equations
have been proposed. These models are usually classified as fluids of differential, rate and integral types. The differential type
and rate type have received much attention. Among them, the viscoelastic rate type model, which is used widely, is the
Maxwell model.
Fractional calculus has achieved much success in the description of the complex dynamics [1,2]; it is proving to be a
valuable tool for handling viscoelastic properties. The starting point of the fractional derivative model of viscoelastic fluids
is usually a classical differential equation which is modified by replacing the time derivative of an integer order by the so-
called Riemann–Liouville fractional calculus operator. This generalization allows us to define precisely non-integer order
derivatives. A very good fit of experimental data was achieved when the fractional Maxwell model has been used instead of
the Maxwell model [3,4]. Later on, this model was employed to study various problems due to its relatively simple structure.
Moreover, the Maxwell fluid is the subclass of non-Newtonian fluids for which one can reasonably hope to obtain exact
solutions. The exact solutions are important due to a few reasons. They provide a standard for checking the accuracies of
many approximate solutions. They can also be used as tests for verifying numerical schemes that are developed for studying
more complex flow problems. Therefore, many researchers [5–10] in the field are engaged in obtaining such solutions. The
aim of this paper is to establish exact solutions for the unsteady flow of an incompressible generalized Maxwell fluid due to
an infinite constantly accelerating and oscillating plate.

∗ Corresponding author. Tel.: +86 1062332891.

E-mail address: liancunzheng@sina.com (L. Zheng).

1468-1218/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.nonrwa.2010.02.004
 L. Zheng et al. / Nonlinear Analysis: Real World Applications 11 (2010) 3744–3751 3745

2. Governing equations

Unsteady incompressible flow is governed by the following equations

dV
div V = 0, ρ
= div T, (1)
dt
where T is the Cauchy stress tensor, V is the velocity vector.
We select the velocity of the following form:
V = (u (y, t ) , 0, 0) , (2)
u (y, t ) is the velocity in the x-coordinate direction. Together with the initial condition (the fluid being at rest up to the
moment t = 0)
S(y, 0) = 0, y > 0. (3)
The generalized Maxwell fluid being at rest up to the time t = 0, we get [11]

∂ u (y, t ) ∂ 2 u (y, t )
1 + λDαt =ν , (4)
∂t ∂ y2
where v = µ/ρ is the kinematic viscosity of the fluid, λ is the material constant and Dαt is the Riemann–Liouville fractional
differentiation operator of order α with respect to t defined as [12]
t
f (τ )
Z
α 1 d
Dt f (t ) = dτ , 0 ≤ α < 1, (5)
0 (1 − α) dt 0 (t − τ )α
where 0 (·) is the Gamma function.

3. Statement of the problem

Consider an incompressible generalized Maxwell fluid over an infinitely extended flat plate. Initially the fluid as well as
the plate is at rest. After time t = 0, the plate begins to oscillate and slide with constant acceleration in the x-direction and
induces the motion into the fluid. Owing to the shear the fluid is gradually moved, its velocity being of the form (2) and the
governing equation is given by Eq. (4). Accordingly, the boundary and initial conditions are
u (0, t ) = At + U cos (ωt ) or u (0, t ) = U sin (ωt ) ; t > 0, (6)
where A is the constant acceleration, U the amplitude and ω the frequency of oscillation of the plate, and
∂ u (y, t )
u (y, t ) = = 0 when t = 0, y > 0. (7)
∂t
Moreover, the natural conditions are:
∂ u (y, t )
u (y, t ) , → 0 as y → ∞, t > 0. (8)
∂y

4. Calculation of the velocity field

4.1. The case u(0, t ) = At + U cos(ωt )

√
Multiplying both sides of Eq. (4) by 2/π sin (ξ y), integrating the result with respect to y from 0 to ∞ and having in
mind the initial and boundary conditions (6)–(8), gives

∂ us (ξ , t )
r
α 2
1 + λDt = vξ (At + U cos ωt ) − vξ 2 us (ξ , t ) , (9)
∂t π
where the Fourier sine transform us (ξ , t ) of u (y, t ) has to satisfy the conditions
∂ us (ξ , 0)
us (ξ , 0) = = 0 for ξ > 0. (10)
∂t
Applying the Laplace transform to Eq. (9) and using the Laplace transform formula for sequential fractional derivatives [6],
we obtain the image function us (ξ , s) of us (ξ , t ) under the form
r r !
1 1 2 2 s
us (ξ , s) = vξ A + vξ U . (11)
s + λsα+1 + vξ 2 s2 π π s2 + ω 2
3746 L. Zheng et al. / Nonlinear Analysis: Real World Applications 11 (2010) 3744–3751

In order to obtain us (ξ , t ) = L−1 us (ξ , s) and to avoid the lengthy calculations of residues and contour integrals, we apply
the discrete inverse Laplace transform method. However, for a more suitable presentation of the final results, we firstly
rewrite Eq. (11) in the equivalent forms
r r r  
2 1 2 U s A 2 1 1 1
us (ξ , s) = A + · − −
π ξ s2 π ξ s2 + ω 2 v π s s + vξ 2 ξ3
ξ ∞ 
−1 k
r
sk+α−1

2 X
− Av ·
π s + vξ 2 k=0 λ
 k+1
vξ 2
sα+1 + λ

ξ
k+1
∞ 
−1 k ∞
r r
sk+α−1
 
X2 2 U X −1
− Av · + ·
π s + vξ 2 k=0 λ 2 k+1 π ξ λ
 
vξ
sα+1 + λ k =0

sα+k+1
r ∞ k
sk+1

s 2 U X −1 s
× · − · · . (12)
2 k+1 s2 +ω 2 π ξ λ 2 k+1 s2 + ω2
  
vξ vξ
sα+1 + λ k=0 sα+1 + λ

Inverting this result by means of the Fourier sine formula [13], we find that
sin (yξ ) 2Av ξ sin (yξ ) X (−1)k
Z ∞   Z ∞ ∞
A s 2A 1 1
u (y, s) = +U · − − · dξ −
s2 s2 + ω 2 vπ 0 s s + vξ 2 ξ3 π 0 s + vξ 2 k=0 λk

sin (yξ ) −1 k+1

k+α−1 Z ∞ ∞   k+1
s 2U X s s
× k+1 dξ + k+1 · dξ
π ξ λ s2 + ω2

vξ 2 vξ 2
sα+1 + λ
0 k=0 sα+1 + λ

sin (yξ ) X sα+k+1

Z ∞ ∞  k
2U −1 s
− k+1 · dξ . (13)
π ξ λ s2 + ω 2

vξ 2
0 k=0 sα+1 + λ

Finally, applying the inverse Laplace transform to Eq. (13) and taking into account all the previous results, it is not difficult
to show that
2A  sin (yξ )
Z ∞  2t
u(y, t ) = At + U cos(ωt ) − · dξ 1 − e−vξ
vπ 0 ξ3
2Av (−1)k vξ 2
Z ∞Z t ∞  
−vξ 2 (t −s)
X
− ξ sin (yξ ) · e · Gα+1,k+α−1,k+1 − , s dsdξ
π 0 0 k =0
λk λ
sin (yξ ) X (−1)k+1 vξ 2
Z ∞Z t ∞  
2U
+ · · Gα+1,k+1,k+1 − , s · cos (ω (t − s)) dsdξ
π 0 0 ξ k=0
λk+1 λ

sin (yξ ) X (−1)k vξ 2

Z ∞Z t ∞  
2U
− · · Gα+1,α+k+1,k+1 − , s · cos (ω (t − s)) dsdξ , (14)
π 0 0 ξ k=0
λk λ
where [14]

sb
 
Ga,b,c (p, t ) = L −1
, (15)
(sa − p)c
or equivalently
2A
Z ∞   sin (yξ ) 2Av
Z ∞Z t
−vξ 2 t
u(y, t ) = At + U cos(ωt ) − 1−e · dξ − ξ sin (yξ )
vπ 0 ξ3 π 0 0
k 
(k + 1)j skα+jα+j+1
j
∞ X ∞ 
−1 vξ 2
× e−vξ (t −s)
2
X
− dsdξ
k=0 j=0
λ λ 0 (1 + j) 0 (kα + jα + j + 2)
k+1 
skα+jα+α+j−1
j
sin (yξ ) X X −1
∞ ∞ 
vξ 2 (k + 1)j
Z ∞Z t
2U
+ · −
π 0 0 ξ k=0 j=0
λ λ 0 (1 + j) 0 (kα + jα + α + j)
k  j
sin (yξ ) X X −1 vξ 2
Z ∞Z t ∞ ∞ 
2U
× cos (ω (t − s)) dsdξ − −
π 0 0 ξ k=0 j=0
λ λ
 L. Zheng et al. / Nonlinear Analysis: Real World Applications 11 (2010) 3744–3751 3747

(k + 1)j skα+jα+j−1
× cos (ω (t − s)) dsdξ . (16)
0 (1 + j) 0 (kα + jα + j)

4.2. The case u(0, t ) = U sin(ωt )

In the same way, we can obtain the solution

sin (yξ ) X (−1)k+1 vξ 2

Z ∞ Z t ∞  
2U
u (y, t ) = U sin (ωt ) + · Gα+ 1 , k+ 1, k + 1 −· , s · sin (ω (t − s)) dsdξ
π 0 0 ξ k=0
λk+1 λ

sin (yξ ) X (−1)k vξ 2

Z ∞Z t ∞  
2U
− · · Gα+1,α+k+1,k+1 − , s · sin (ω (t − s)) dsdξ , (17)
π 0 0 ξ k=0
λk λ
the similar solutions corresponding to generalized Maxwell fluids have been obtained as limiting cases of general solutions
in [15]. The present solutions are written in terms of the generalized G-functions, while in the reference they are written in
terms of the Mittag-Leffler functions.

5. Calculation of the shear stress

5.1. The case u(0, t ) = At + U cos(ωt )

To determine the shear stress τ (y, t ), we use the relation

µ ∂ u (y, s)
τ (y, s) = , (18)
λsα + 1 ∂y
where, in view of Eq. (13),

∂ u (y, s) ∞
cos (yξ )
ξ 2 cos (yξ ) X ∞
(−1)k 2Av ∞
Z   Z
2A 1 1
=− − · dξ −
∂y vπ 0 s s + vξ 2 ξ 2 π 0 s + vξ 2
k =0
λk
∞ 
−1 k+1
Z ∞
sk+α−1 sk+1

2U X s
× d ξ + cos ( yξ ) k+1 · 2 dξ
k+1
π 0 λ s + ω2
 
vξ 2 vξ 2
sα+1 + λ k=0 sα+1 + λ

−1 k sα+k+1
Z ∞ ∞  
2U X s
− cos (yξ ) k+1 · dξ . (19)
π λ s2 + ω2

vξ 2
0 k=0 sα+1 + λ

Introducing (19) into (18), it results that

2ρ A ∞
cos (yξ ) 2µA ∞
sα−1
Z   Z
1 1
τ (y, s) = − − · dξ +
cos (yξ ) dξ
π s + vξ 2 ξ2 π s + vξ 2 sα + λ1


0 s 0

2Av µ ξ 2 cos (yξ ) X (−1)k

∞ ∞
sk+α−1
Z
1
− · dξ
π λ s + vξ 2 k=0 λk 2 k+1 sα 1

0
sα+1 + λ
vξ + λ

2U µ −1 k+1
∞ ∞ 
sk+1
Z 
X 1 s
+ cos (yξ ) k+1 · dξ
π λ λ sα + λ1 s2 + ω2

vξ 2
0 k=0 sα+1 + λ

2U µ −1 k sα+k+1
Z ∞ ∞  
X 1 s
− cos (yξ ) k+1 · dξ , (20)
π λ λ sα + λ1 s2 + ω2

vξ 2
0 k=0 sα+1 + λ

or equivalently
2ρ A ∞
cos (yξ ) 2µA ∞
sα−1
Z   Z
1 1
τ (y, s) = − − · dξ +
cos (yξ ) dξ
π s + vξ ξ π s + vξ 2 sα + λ1
2 2


0 s 0
k  j
2Av µ ∞ ∞ ∞
XX vξ 2 (k + 1)j
Z
1 1
− ξ 2 cos (yξ ) − − · dξ
π λ λ λ 0 (1 + j) s + vξ s + λ1 sα k+α j+j+2
α



2
0 k=0 j=0
3748 L. Zheng et al. / Nonlinear Analysis: Real World Applications 11 (2010) 3744–3751

k+1  j
2U µ ∞ ∞ X ∞ 
vξ 2 (k + 1)j
Z X 1 1 s
+ cos (yξ ) − − dξ
π λ λ λ 0 (1 + j) s + λ s
α 1 α s + ω2

2
k+α j + j +α
0 k=0 j=0

k  j
2U µ ∞ ∞ X ∞ 
vξ 2 (k + 1)j
Z X 1 1 s
− cos (yξ ) − − dξ . (21)
π λ λ λ 0 ( ) s +λ s
α 1 α ω2

2
1 + j k+α j+ j s +
0 k=0 j=0

Inverting this result, we find the shear stress under the form
2ρ A ∞  cos (yξ ) 2µA ∞ t −vξ 2 (t −s)
Z  Z Z
2t
τ (y, t ) = − 1 − e−vξ
· d ξ + e cos (yξ )
π 0 ξ2 π 0 0
k  j
2Av µ ∞ 2 vξ 2
  Z ∞ X ∞ 
1 X 1
× Rα,α−1 − , 0, s dsdξ − ξ cos (yξ ) − −
λ π λ 0 k=0 j=0
λ λ
Z tZ σ
(k + 1)j sα k+α j+j+1
 
1
e−vξ (σ −s)
2
× Rα,0 − , 0, t − σ dsdσ dξ
0 (1 + j) 0 0 0 (α k + α j + j + 2) λ
Z ∞Z tZ σ k+1  2 j
2U µ vξ
∞ X ∞  
X 1
+ cos (yξ ) − −
π λ 0 0 0 k=0 j=0
λ λ
(k + 1)j sα k+α j+α+j−1
 
1
× Rα,0 − , 0, t − σ cos (ω (σ − s)) dsdσ dξ
0 (1 + j) 0 (α k + α j + α + j) λ
k 
2U µ ∞ t σ
j
∞ X ∞ 
vξ 2
Z Z Z X 1
− cos (yξ ) − −
π λ 0 0 0 k=0 j=0
λ λ

(k + 1)j sα k+α j+j−1

 
1
× Rα,0 − , 0, t − σ cos (ω (σ − s)) dsdσ dξ . (22)
0 (1 + j) 0 (α k + α j + j) λ

5.2. The case u(0, t ) = U sin(ωt )

Proceeding in a similar manner as before, we find the corresponding solution of the following form:
σ k+1 
sα k+α j+α+j−1
j
2U µ ∞ ∞ X ∞ 
vξ 2 (k + 1)j
Z Z tZ
X 1
τ (y, t ) = − cos (yξ )− ·
π λ 0 0 0 k=0 j=0
λ λ 0 (1 + j) 0 (α k + α j + α + j)
Z ∞Z tZ σ k
2U µ
  ∞ X ∞ 
1 X 1
× Rα,0 − , 0, t − σ sin (ω (σ − s)) dsdσ dξ − cos (yξ ) −
λ π λ 0 0 0 k=0 j=0
λ

sα k+α j+j−1
j
vξ 2 (k + 1)j
  
1
× − · Rα,0 − , 0, t − σ sin (ω (σ − s)) dsdσ dξ . (23)
λ 0 (1 + j) 0 (α k + α j + j) λ

6. Limiting cases

(i) The case u(0, t ) = At + U cos(ωt )

By letting α → 1 into Eqs. (16) and (22), we obtain the similar solutions

2A
Z  sin (yξ )
∞  2t
u(y, t ) = At + U cos(ωt ) − · 1 − e−vξ dξ
vπ 0 ξ3
k j
2Av
∞ X ∞
vξ 2 (k + 1)j sk+2j+1
Z ∞Z t  
−vξ 2 (t −s)
X −1
− ξ sin (yξ ) · e − dsdξ
π 0 0 k=0 j=0
λ λ 0 (1 + j) 0 (k + 2j + 2)
k+1  j
sin (yξ ) X X −1
∞ ∞ 
vξ 2 (k + 1)j
Z ∞Z t
2U sk+2j
+ · − cos (ω (t − s)) dsdξ
π 0 0 ξ k=0 j=0
λ λ 0 (1 + j) 0 (k + 2j + 1)
k  j
sin (yξ ) X X −1
∞ ∞ 
vξ 2 (k + 1)j sk+2j−1
Z ∞Z t
2U
− − cos (ω (t − s)) dsdξ , (24)
π 0 0 ξ k=0 j=0
λ λ 0 (1 + j) 0 (k + 2j)
 L. Zheng et al. / Nonlinear Analysis: Real World Applications 11 (2010) 3744–3751 3749

2ρ A ∞  cos (yξ ) 2µA ∞

cos (yξ )  −vξ 2 t
Z  Z 
2t t
τ (y, t ) = − 1 − e−vξ · dξ − λ e − e− λ dξ
π 0 ξ 2 π 0 vλξ − 1
2
σ
2v A µ − t ∞ t
Z Z Z
σ
e−vξ (σ −s)+ λ
2
− e λ ξ 2 cos (yξ )
π λ 0
( k  j 0 0 )
X∞ X ∞ 
1 vξ 2 (k + 1)j sk+2j+1
× − − dsdσ dξ
k =0 j =0
λ λ 0 (1 + j) 0 (k + 2j + 2)
Z tZ σ
2U µ − t
Z ∞
σ
+ e λ cos (yξ ) e λ cos (ω (σ − s))
(π λ 0
k+1  0 2  0
j
)
X∞ X ∞ 
1 vξ (k + 1)j sk+2j
× − − dsdσ dξ
k =0 j =0
λ λ 0 (1 + j) 0 (k + 2j + 1)
Z tZ σ
2U µ − t
Z ∞
σ
− e λ cos (yξ ) e λ cos (ω (σ − s))
π λ 0 0
( j 0 )
vξ 2 (k + 1)j sk+2j−1
∞ X ∞  k 
X 1
× − − dsdσ dξ , (25)
k =0 j =0
λ λ 0 (1 + j) 0 (k + 2j)
corresponding to the classical Maxwell model. Similar solutions to those given by (24) and (25) have also been obtained
in [16] by a different technique.

(ii) The case u(0, t ) = U sin(ωt )

As in the previous case, the corresponding velocity field and tangential stress for classical Maxwell fluid are, respectively,
given by
u(y, t ) = U sin(ωt )
j
∞ t
sin (yξ ) X X
∞ ∞
−1 k+1 vξ 2 (k + 1)j sk+2J
Z Z   
2U
+ · − sin (ω(t − s)) dsdξ
π 0 0 ξ k=0 j=0
λ λ 0 (1 + j) 0 (k + 2j + 1)
k  j
sin (yξ ) X X −1
∞ ∞ 
vξ 2 (k + 1)j sk+2j−1
Z ∞Z t
2U
− − sin (ω (t − s)) dsdξ , (26)
π 0 0 ξ k=0 j=0
λ λ 0 (1 + j) 0 (k + 2j)
Z tZ σ
2U µ − t
Z ∞
σ
τ (y, t ) = e λ cos (yξ ) e λ sin (ω (σ − s))
π(λ 0
k+1  0 02 j )
X∞ X ∞ 
1 vξ (k + 1)j sk+2j
× − − dsdσ dξ
k =0 j =0
λ λ 0 (1 + j) 0 (k + 2j + 1)
Z tZ σ
2U µ − t
Z ∞
σ
− e λ cos (yξ ) e λ sin (ω (σ − s))
π λ 0 0
( j0 )
vξ 2 (k + 1)j sk+2j−1
∞ X ∞  k 
X 1
× − − dsdσ dξ . (27)
k =0 j =0
λ λ 0 (1 + j) 0 (k + 2j)

(iii) Taking the limits of (16) and (22), as U → 0, we find

2A
Z ∞   sin (yξ ) 2Av
Z ∞Z t
ξ sin (yξ ) · e−vξ (t −s)
2t 2
u(y, t ) = At − 1 − e−vξ
· d ξ −
vπ 0 ξ 3 π 0 0
k 
(k + 1)j skα+jα+j+1
j
X∞ X ∞ 
−1 vξ 2
× − dsdξ , (28)
k=0 j=0
λ λ 0 (1 + j) 0 (kα + jα + j + 2)

2ρ A ∞   cos (yξ ) 2µA ∞ t −vξ 2 (t −s)

Z Z Z  
−vξ 2 t 1
τ (y, t ) = − 1−e · dξ + e cos (yξ ) Rα,α−1 − , 0, s dsdξ
π 0 ξ2 π 0 0 λ
k 2 j
2Av µ
∞ X ∞
vξ ( )
Z ∞    
X 1 k + 1 j
− ξ 2 cos (yξ ) − − ·
π λ 0 k=0 j=0
λ λ 0 (1 + j)
Z tZ σ α k+α j+j+1  
−vξ 2 (σ −s) s 1
× e Rα,0 − , 0, t − σ dsdσ dξ , (29)
0 0 0 (α k + α j + j + 2) λ
which are identical to solutions Eqs. (21) and (26) of [11], respectively, Eqs. (48) and (49) of [17].
3750 L. Zheng et al. / Nonlinear Analysis: Real World Applications 11 (2010) 3744–3751

(a) u(0, t ) = At + U cos(ωt ). (b) u(0, t ) = U sin(ωt ).

Fig. 1. Velocity u(y, t ) versus y for different values of ω when other parameters are fixed.

(a) u(0, t ) = At + U cos(ωt ). (b) u(0, t ) = U sin(ωt ).

Fig. 2. Velocity u(y, t ) versus y for different values of α when other parameters are fixed.

7. Conclusions

In this paper, the closed-form expressions for the velocity field u (y, t ) and the shear stress τ (y, t ), corresponding to the
unsteady flow induced by an infinite constantly accelerating and oscillating plate in an incompressible generalized Maxwell
fluid, have been determined by using the Fourier sine and Laplace transform. The exact solutions are presented for the
velocity field and the corresponding shear stress in series forms in terms of generalized G and R functions by using the
discrete inverse Laplace transform method.
Furthermore, some diagrams are presented to show the characteristics of solutions for values of parameters, for the sake
of simplicity, all the diagrams are plotted by taking U = 1, A = 1, λ = 8.
As is to be expected, the velocity goes to zero as y increases. In Fig. 1, it can be seen that the velocity is an increasing func-
tion of ω (the frequency of oscillation of the plate) near the plate for the cosine oscillations, however, decreases for the sine
oscillations. In the special cases when α → 1, the general solutions (16) and (22) reduce to the classical solutions for Maxwell
fluid performing the same motion. The differences between the velocity fields (16) and (24) as well as (17) and (26), are pre-
sented by graphical illustrations by Fig. 2. In Fig. 3, as a check of the present results, this provides a useful mathematical check.

Acknowledgements

The work is supported by the National Natural Science Foundations of China (No. 50936003), the Open Project of Institute
of Rhe. Mech. & Material Eng. (Cent. South Univ. of Forestry and Tech., No. 09RM04); The open Project of State Key Lab. for
Adv. Metals and Materials (2009Z-02) and Research Foundation of Engineering Research Institute of USTB.
 L. Zheng et al. / Nonlinear Analysis: Real World Applications 11 (2010) 3744–3751 3751

(a) u(0, t ) = At + U cos(ωt ). (b) u(0, t ) = U sin(ωt ).

Fig. 3. Velocity u(y, t ) versus y for different values of t when other parameters are fixed.

References

[1] L.I. Palade, P. Attane, R.R. Huilgol, B. Mena, Anomalous stability behavior of a properly invariant constitutive equation which generalizes fractional
derivative models, Int. J. Eng. Sci. 37 (1999) 315–329.
[2] Y.A. Rossikhin, M.V. Shitikova, A new method for solving dynamic problems of fractional derivative viscoelasticity, Int. J. Eng. Sci. 39 (2001) 149–176.
[3] N. Makris, Theoretical and experimental investigation of viscous dampers in applications of seismic and vibration isolation, Ph.D. Thesis, State Univ.
of New York at Buffalo, Buffalo, NY, 1991.
[4] N. Makris, M.C. Constantinou, Fractional derivative model for viscous dampers, J. Struct. Eng. ASCE 117 (1991) 2708–2724.
[5] S. Hyder Ali Muttaqi Shah, M. Khan, Haitao Qi, Exact solutions for a viscoelastic fluid with the generalized Oldroyd-B model, Nonlinear Anal. RWA 10
(2009) 2590–2599.
[6] D. Vieru, Corina Fetecau, C. Fetecau, Flow of a generalized Oldroyd-B fluids due to a constantly accelerating plate, Appl. Math. Comput. 201 (2008)
834–842.
[7] M. Khan, S. Hyder Ali, Haitao Qi, Exact solutions for some oscillating flows of a second grade fluid with a fractional derivative model, Math. Comput
Model. (submitted in publication).
[8] Masood Khan, Tasawar Hayat, Some exact solutions for fractional generalized Burgers’ fluid in a porous space, Nonlinear Anal. RWA 9 (2008)
1952–1965.
[9] S. Hyder Ali Muttaqi Shah, Some accelerated flows of generalized Oldroyd-B fluid between two side walls perpendicular to the plate, Nonlinear Anal.
RWA 10 (2009) 2146–2150.
[10] M. Nazar, Corina Fetecau, A.U. Awan, A note on the unsteady flow of a generalized second-grade fluid through a circular subject to a time dependent
shear stress, Nonlinear Anal. RWA (2008), doi:10.1016/j.nonrwa.2008.10.055.
[11] Corina Fetecau, M. Athar, C. Fetecau, Unsteady flow of a generalized Maxwell fluid with fractional derivative due to a constantly accelerating plate,
Comput. Math. Appl. 57 (2009) 596–603.
[12] I. Podlubny, Fractional Differential Equations, Academic Press, San Diego, 1999.
[13] I.N. Sneddon, Fourier Transforms, McGraw Hill Company, New York, Toronto, London, 1951.
[14] C.F. Lorenzo, T.T. Hartley, Generalized functions for the fractional calculus, NASA/ TP-1999-209424, 1999.
[15] M. Kan, Asia Anjum, Haitao Qi, C. Fetecau, On exact solutions for some oscillating motions of a generalized Oldroyd-B fluid, Z. Angew. Math. Phys.,
doi:10.1007/s00033-009-0004-4.
[16] M. Khan, Maryam Saleem, C. Fetecau, T. Hayat, Transient oscillatory and constantly accelerated non-Newtonian flow in a porous medium, Int. J. Non-
Linear Mech. 42 (2007) 1224–1239.
[17] Constantin Fetecau, Corina Fetecau, M. Kamran, D. Vieru, Exact solutions for the flow of a generalized Oldroyd-B fluid induced by a constantly
accelerating plate between two side walls perpendicular to the plate, J. Non-Newtonian Fluid Mech. 156 (2009) 189–201.