Chapter 1 Introduction

1.1 Blasius Equation

If a fluid flows past a solid, a fluid layer is formed adjacent to the
boundary of the solid. This layer is called a boundary layer and strong
viscous effects exist within this layer.
Consider a uniform flow over a flat surface, y=0, 0, . x z < <
Equations of the flow in the boundary layer are the continuity equation
0 (1.1)
u v
x y

+ =


and the reduced Navier- Stokes equation
2
2
(1.2)
u v u
u v
x y y

+ =


where u and v are respectively the components of the velocity vector and
represents the viscosity of the fluid.
Boundary conditions are
( , 0) 0 0 (1.3 )
( , 0) 0 0 (1.3 )
( , ) as (1.3 )
u x x a
v x y b
u x y U y c
=
=


where U is the constant speed of the flow outside the boundary Layer.
Define a stream function ( , ) x y such that
, (1.4) u v
y x

= =


then equation (1.1) is satisfied identically and equation (1.2) becomes
2 2 3
2 3
(1.5) v
y x y x y y

=


Blasius used a similarity transformation to reduce (1.5) to an ordinary
differential equation.

A similarity transformation is based on the symmetry analysis of a
differential equation [11,23]. When a symmetry property of a differential
equation is identified it can be exploited to achieve a simplification. If it is
an ordinary differential equation then usually the order of the equation can
be reduced. If it is a partial differential equation then usually the dependent
and independent variables can be combined to achieve a reduction of order
or a reduction of the partial differential equation to an ordinary differential
equation.
In the case of (1.5) symmetry analysis leads to the following
transformation [11]
( , ) ( )
y
a
x
x y b x f

=
=

where a and b are constants and are chosen to make
and ( ) f dimensionless. They are taken as
,
U
a
b U

=
=

With this choice, is called the dimensionless similarity variable and ( ) f
is called the dimensionless stream function. Now
2
2
2
3
2
3 2
3
1 ( )
( )
2 2
( )
( )
( ) ( )
2 2
( ).
U y f
f U
x x x
Uf
y
a
Uf
y x
U U y U
f f
x y x
x
U
f
y x

= +

= =

A substitution of the above derivatives in equation (1.5) reduces it to
3 2
3 2
1
( ) 0 (1.6)
2
d f d f
f
d d

+ =
Equation (1.6) is known as the Blasius equation. The boundary condition
(1.3a) transforms to
(0) 0 (1.7 ) f a =

while (1.3b) becomes
(0) 0 (1.7 ) f b =
and (1.3c) reduces to
( ) 1 (1.7 ) f as c
Equation (1.6) together with the boundary conditions (1.7a), (1.7b) and
(1.7c) is called the Blasius problem.

1.2 Numerical solution
A numerical solution of the Blasius problem usually uses the shooting
method. In this method it is assumed that
(0) (1.8) f =
and the problem is solved with different values of . Such values of will
lead to different values of as .
df
d

We seek that value of which

will yield an f which satisfies
lim 1.
df
d

=
First accurate numerical solution was obtained by Howarth [17]. More
recently Asaithambi [8], and Cortell[12] have also solved the Blasius
problem by the shooting method.

In practice it is impossible to carry out calculations up to infinity. Hence
an

= is arbitrarily fixed and we demand that

1 when .
df
d

=
We solve the Blasius equation by the shooting method. We start with
0.1 = and find that ( ) f between 12 and 13 = is practically a constant
=0.449287. With 0.6 = . We find the slope for 12 13 < < equal to 1.48352.
This means the actual value should lie between 0.1 and 0.6 . Therefore our
next choice is
0.1 0.6
0.35
2
+
= and so on. We present the results in Table. 1
and 2.



















(12.5) f
0.1 0.449287
0.6 1.48352
0.35 1.03571
0.225 0.771459
0.2875 0.908413
0.31875 0.973101
0.334375 1.00465
0.326562 0.988937
0.330468 0.996808
0.3324218 1.00073
0.3314453 0.998771
0.3319335 0.999751
0.3321771 1.00024
0.3320553 0.999996
0.3321162 1.00012
0.3320857 1.00006
0.3320705 1.00003
0.3320629 1.00001
0.3320591 1
Table 1: Sequence of values of converging to







( ) f ( ) f ( ) f
0.0 0 0 0.332059
0.2 0.00664105 0.0664081 0.331985
0.4 0.02656 0.132765 0.331468
0.6 0.059735 0.198938 0.330081
0.8 0.106109 0.264711 0.327391
1.0 0.165573 0.329782 0.323009
1.2 0.23795 0.393778 0.31659
1.4 0.322983 0.456264 0.307867
1.6 0.420323 0.516759 0.296665
1.8 0.529521 0.574761 0.282932
2.0 0.650028 0.629769 0.266753
2.2 0.781197 0.681314 0.248352
2.4 0.922295 0.728985 0.228092
2.6 1.07251 0.772459 0.206455
2.8 1.23098 0.811513 0.184007
3.0 1.39682 0.846048 0.161359
3.2 1.5691 0.876085 0.139129
3.4 1.74696 0.901765 0.117876
3.6 1.92953 0.923334 0.0980867
3.8 2.11604 0.941122 0.0801258
4.0 2.30576 0.955522 0.0642341
4.2 2.49805 0.966961 0.0505193
4.4 2.69237 0.975875 0.0389731
4.6 2.88826 0.982687 0.0294829
4.8 3.08533 0.987793 0.0218711
5.0 3.28329 0.991546 0.0159068

5.2 3.48188 0.994249 0.0113414
5.4 3.68094 0.996159 0.00792786
5.6 3.88031 0.997481 0.00543169
5.8 4.0799 0.998379 0.00364835
6.0 4.27964 0.998976 0.00240198
6.2 4.47948 0.999366 0.00155022
6.4 4.67938 0.999615 0.000980419
6.6 4.87932 0.999711 0.000608191
6.8 5.07928 0.999867 0.000369599
7.0 5.27926 0.999925 0.000220207
7.2 5.47925 0.999959 0.000128431
7.4 5.67924 0.999979 0.0000737176
7.6 5.87924 0.99999 0.0000413441
7.8 6.07924 0.999996 0.0000227099
8.0 6.27924 1 0.0000122503
8.2 6.47924 1
6
6.4618 10

8.4 6.67924 1
6
3.28575 10

Table2: Numerical values of ( ), ( ) and ( ) f f f

0 2 4 6 8 10 12 14
h
0
2
4
6
8
10
12
f
H
h
L

Fig1: ( ) f as a function of




0 2 4 6 8
h
0.5
0.6
0.7
0.8
0.9
1
f
'
H
h
L

Fig2: ( ) f as a function of







0 2 4 6 8
h
0
0.05
0.1
0.15
0.2
0.25
0.3
f
'
'
H
h
L

Fig3: ( ) f as a function of













1.3 Analytical methods
Blasius [10] found a series solution of Eq.(1.6)
1
3 2
0
1
( ) ( ) (1.9)
2 (3 2)!
k
k k k
k
A
f
k

+
+
=
=
+

where
0 1
1 A A = = and
1
1,
0
3 1
2 (1.10)
3
k
k r k r
r
k
A A A k
r

=
| |
=
|
\

However the series (1.9) converges slowly and its radius of convergence is
approximately 5.90 , therefore it cannot be used to find an accurate value
of by using the condition
( ) 1 f

=
for sufficiently large

. Blasius found an asymptotic expression and

estimated by matching the two expressions for a suitable value of .
In recent years a few authors have tried to find alternate methods to
find . J. H. He has proposed a perturbation approach to solve this
problem [16] and Abbasbandy has combined this approach with Adomian
decomposition method to find an improved solution of the problem [3].
Liao [19] has used the homotopy analysis method to find an accurate value
of .
Recently Wang has used a transformation to transform the Blasius
equation into a second order differential equation in which the condition at
infinity is transformed into
1
lim ( ) 0 (1.11)
x
y x

=
[24]. Several authors have used Eq. (1.11) to find a good estimate of
,[9,15,24].



1.4 Iteration perturbation method of He
Following J. H. He [16] we construct an iteration formula for
Blasius equation
1 1
1
0, (1.12)
2
n n n
f f f
+ +
+ =
where we denote by
n
f the nth approximate solution. We will use the
perturbation method to find approximately
1
.
n
f
+

We begin with
0
2 , (1.13) f b =
Substituting
0
f into (1.12) results in
1 1
0. (1.14) f bf + =
The solution of Eq. (1.14) is
1
1
( ) (1 ), (1.15)
b
f x e
b

=
where b is an unknown constant.
Substituting
1
f into (1.12), we obtain
2 2
1 1
(1 ) 0. (1.16)
2
b
f e f
b

+ =



We re-write Eq. (1.16) in the form
2 2 2
1 1
(1 ) 2 0, (1.17)
2
b
f bf e b f
b

+ + =



and embed an artificial parameter in Eq.(1.17):
2 2 2
1 1
(1 ) 2 0. (1.18)
2
b
f bf e b f
b

+ + =



Supposing that the solution of Eq. (1.18) can be expressed as
(0) (1)
2 2 2
..., (1.19) f f f = + +
we have the following linear equations:

(0) (0) (0) (0) (0)
2 2 2 2 2
(1) (1) (0)
2 2 2
(1) (1) (1)
2 2 2
[ ] [ ] 0, (0) [ ] (0) 0, and [ ] ( ) 1, (1.20)
1 1 1
[ ] [ ] 2 [ ] 0,
2
(0) [ ] 0, and [ ] ( ) 0. (1.21)
b
f b f f f f
f b f e b f
b b
f f f

+ = = = + =
| |

+ + + =
|
\
= = + =

The solution of (1.20) is
(0)
2
( ) (1 ) / ,
b
f e b

= substituting it into (1.21)
results in
(1) (1) 2
2 2
1
[ ] [ ] ( 2 1) (1.22)
2
b b
f b f b e b e

+ = +
The constant b can be identified by the following expression:
2
0
( 2 1) 0. (1.23)
b b b
e b e b e d


+

+ =

So we have
2
1 1 1 2
(2) 0, (1.24)
4 3 2
b
b b b
+
+ =
which leads to the result
1
0.28867. (1.25)
12
b = =
The exact solution of Eq. (1.22) is not required, the expression, Eq. (1.23),
requires no terms of
n
e

(n1,2,3,) in
(1)
2
f . So we can assume that the
approximate solution of Eq. (1.22) can be expressed as
(1) 2
2
1
( ) , (1.26)
8
b b
f Ae e B

= + +
where
1 1
and ,
4 8
A B

= = which are identified from the initial conditions

(1) (1)
2 2
(0) [ ] (0) 0. f f = =
By setting 1, = we obtain first-order approximate solution, i.e.,
(0) (1)
2 2 2
( ) ( ) ( ) (1.27) f f f = +

A highly accurate numerical solution of Blasius equation has been provided
by Howarth [17], who gives the initial slope (0) 0.332057
ex
f = [17].
Comparing the approximate initial slope
(0) (1)
2 2 2
(0) [ ] (0) [ ] (0) 0.3095, f f f = + =
we find that the relative error is 6.8%.
We can improve the accuracy of approximate solution to an high-order
by iteration technique. Substituting
2
f into (1.12), and by parallel
operation, the constant b can be identified by the following relation:
2 2
0
1 1 1
( 2 1 ) 0, (1.28)
4 8 8
b b b b b
e b e b e e e d


+

+ + + =

which leads to the result
0.3062. (1.29) b =
It is obvious that
2
0
(0) 0.25 0.3296 f b b = + = reaches a very high accuracy, the
0.73% accuracy is remarkably good.


1.5 Homotopy analysis method
Recently Liao has developed a method for solving nonlinear
problems which does not depend on the existence of a small parameter ,
also it generalizes the ideas involved in the decomposition method of
Adomian or the artificial small parameter method of Lyapunov [21]. This
method has been used by Liao and other authors to find analytical
expressions for several nonlinear problems.
Suppose the problem to be solved is
[ ( )] ( ) (1.30) N u r f r =
Introduce a new variable q and consider the equation
0
(1 ) [ ( , ) ( )] { [ ( , )] ( )} (1.31) q L r q u r hq N r q f r =

where L is a suitably chosen linear operator ,
0
( ) u r is an initial approximate
solution and h is a parameter to be chosen later. In (1.31) put q=0. We get
0
( , 0) ( ) (1.32) r u r =
If we let q =1 in (1.31) , we get
[ ( ,1)] ( ) (1.33) N r f r =
Whose solution is the required function u(r). Thus as q varies continuously
from 0 to 1 , the initial approximate solution
0
( ) u r evolves to the desired
solution u(r) of the problem Liao gives this reason for calling the method as
the homotopy analysis method.
Define
0,
1
( ) ( , ) (1.34)
!
n
n n q
u r r q
n q
=

then
0
1
( , ) ( ) ( ) . (1.35)
n
n
n
r q u r u r q

=
= +

If the series converges at q =1 , then
0
1
( ) ( ,1) ( ) . (1.36)
n
u r r u r u

=
= = +

The functions ( )
n
u r are found successively from the equations
1 0
[ ( )] { [ ( )] ( )} (1.37) L u r h N u r f r =
when n=1 and
1
[ ( ) ( )] ( ), (1.38)
n n n
L u r u r hR r

=
when 2, where n
1
1 0
1
( ) [ ( , ) ] (1.39)
( 1)!
n
n n q
R r N r q
n q

=


It has been shown by Liao [19] that if we choose h=-1 and L =
0
L , the
linear operator chosen for the Adomian method , then the solution

produced by the homotopy analysis method is identical to the one given by
the Adomian method.

1.6 Liao's solution of the Blasius problem
Liao [18] wrote the Blasius equation in the form
0
3 2
3 2
(1 ) [ ( , , , ) ( )]
( , , , ) 1 ( , , , )
[ ( , , , ) ] (1.40)
2
q L F h q f
F h q F h q
qh F h q




=

+


where [0, ), 0, 0, [0,1] h q >
with the boundary conditions
(0, , , ) (0, , , ) 0, ( , , , ) 1 (1.41) F h q F h q F h q = = =
Here h and are free parameters to be chosen later. Note that at q=0, we
have
0
( , , , 0) ( ) F h f =
and at q=1
( , , ,1) ( ) F h f =
where the initial approximation
0
( ) f can be chosen freely Liao chooses the
linear operator L as
3 2
3 2
, 0 L


= + >


and
0
1 exp( )
( ) f

=
as the initial approximation which satisfies the boundary conditions
f(0)= (0) f =0 and ( ) 1 f = . Successive elements of the series
1 2
( , , ) , ( , , ) ,...... f h f h are calculated recursively and the solution is
formally written as

0
1
( , , )
( , , , ) ( ) , (1.42)
!
k k
k
f h
f h q f q
k

=
= +

Liao also finds a sequence of approximations for (0, , ) f h as
1
2
2
2 3
2 3 3
3
3 3 5 3
(1 )
4
1 5
(1 )
2 24
3 5 275 5
(1 )
4 8 576 24
h
h
h h
h
h h h h
h

= +
= +
= + +

He finds that with
30 35
9
3and , 0.33205and 0.33206
10
h = = = = . Thus
9
3,
10
h = = are the appropriate values of these parameters since they
yield the correct value of (0) f found by Howarth [17]. When we set
9
3,
10
h = = and q=1 in (1.42) we find that the solution with the series
truncated after 35 terms gives numerical values on 0 100 which agree
very well with the solution of Howarth [17].

1.7 Generalized Blasius problem
Cortell numerically studied the problem
0 (1.43 )
(0) (0) 0, ( ) 1 (1.43 )
af ff a
f f f b
+ =
= = =

for various values of a and observed that an increase in a decreases
( )and ( ) also ( ) f f f near 0 = decreases at first with an increase in a but
further away it increases with a. Fang et al. [13] explained the above
behavior analytically by making the following transformation
( ) ( )
a
f aF

=
=


Then the problem(1.43) transforms to
3 2
3 2
0
0 (1.44)
(0) 0, 1
d F d F
F
d d
dF dF
F
d d



=
+ =
= = =

whose solution can be easily obtained. They showed that
[ ]
2
3 2
2
2
3 2
2
( , ) 1
( ) (1.45)
2
( , )
(1.46)
2
( , ) 1
( ) 1 (1.47)
2
f a dF
F
a d a
f a d F
a d
a
f a d F
F
a d
a

Now from (1.45)and (1.46) we see that 0, 0,
f f
a a

<

therefore and f f
decrease with an increase in a . From (1.47) we see that the right side is
negative to start with but becomes positive later on. This explains the
behavior of f noted by Cortell [12].














Chapter 2 Wang's Equation

2.1 Introduction
The two dimensional constant laminar viscous flow over a semi-infinite
flat plate is modeled by the nonlinear differential equation
0
( ) ( ) ( ) 0, [0, ), (2.1 ) f f f a + =
With the boundary conditions
(0) (0) 0, ( ) 1 (2.1 ) f f f b = = =
The above boundary value problem is called the Blasius problem [10] and
it successfully describes the velocity distribution and other physical
parameters associated with the fluid motion in the boundary layer along the
plane [22].The main hurdle in the solution of the above problem is the
absence of the second derivative (0) f .
Once this derivative has been correctly evaluated an analytical solution
of the boundary value problem may be readily found. Blasius [10] found
the following power series solution of the problem with
0
1
2
=
1
3 2
0
1
( ) ( ) (2.2)
2 (3 2)!
k
k k k
k
A
f
k

+
+
=
=
+

where
0 1
1 A A = = and
1
1,
0
3 1
2
3
k
k r k r
r
k
A A A k
r

=
| |
=
|
\

Where represents the unknown (0). f Howarth [17] solved (1.1), (with
0
1
2
= ), numerically and found
0.33206 (2.3) =
Several authors have devised numerical algorithms to find good
approximations to (0) f see, for example, Asaithambi [8] and references

therein . Asithambi [8] solved (2.1) , (with
0
1 = ), and found (0) f ,
denoted by , to be
0.46900 (2.4) =
Fang et al. [13] have shown that the substitution
0
0
1
( ) ( ) f F

=
transforms Eq. (2.1a) into
0 F FF + =
Therefore it is sufficient to consider the Blasius problem with
0
1 = .
Henceforth we shall treat the problem with
0
1. = Liao [18] applied his
homotopy analysis method to the Blasius problem and obtained the
solution to a high level of accuracy. J. H. He [16] has used an iterative
perturbation technique to find an approximate analytical solution of the
Blasius problem. Abbasbandy [3] has used a modified version of the
Adomian decomposition method to find a numerical solution while Cortell
[12] has studied the dependence of the solution on the parameter
0
.
Recently Wang [24] has introduced a fresh approach to the Blasius problem
. He used an ingenious idea to transform the Blasius problem into a simpler
problem governed by a second order differential equation. He used the
transformation
( ) , ( ) (2.5) x f y f = =
to transform (2.1) to
2
2
0, [0,1) (2.6 )
d y x
x a
dx y
+ =
with the boundary conditions
1
(0) (0), (0) 0, lim ( ) 0 (2.6 )
x
y f y y x b

= = =

Wang [24] used the Adomian decomposition method to solve (2.6), and
found
3 6 9 12
3 5 7
( ) ...... (2.7)
6 180 2160 19008
x x x x
y x

=
To find the equation y(1)=0 is solved for . He solved this equation
retaining six terms of the series (2.7) and found 0.453539. = Recently
Hashim [15] improved this value to 0.466799 = by finding terms of the
series (2.7) up to
24
x by the Adomian decomposition method (ADM) and
then approximating this function by the [12/12] diagonal Pade'
approximating .

2.2 The Adomian decomposition method
Adomian's decomposition method is used for solving operator
equations of the from
(2.8) u Nu = +
where N is a nonlinear differential operator and is known. Since the
method does not resort to linearization or assumptions of week
nonlinearity, the nonlinearities which can be handled are quite general and
the solutions generated may be more realistic than those achieved by
simplifying the model of the physical problem to achieve conditions
required for other techniques [14].
Adomian's method assumes that the solution u can be expanded as an
infinite series:
0
(2.9)
n
n
u u

=
=

With , .
n
u X n The image Nu has an expansion
0
(2.10)
n
n
Nu A

=
=

With , .
n
A X n Substituting (2.9) and (2.10) into (2.8) gives
0 0
(2.11)
n n
n n
u A

= =
= +


which is satisfied formally if we set
0
1
(2.12)
(2.13)
n n
u
u A

+
=
=

Thus, we need to determine the so-called Adomian polynomials
, 1, 2, 3,...
n
A n = To determine the
n
A s , a scalar parameter is introduced to
set
0
( ) (2.14)
n
n
n
u u

=
=

And
0
( ) (2.15)
n
n
n
Nu A

=
=


As in [4,5], we can deduce that the coefficients , 1, 2, 3,...
n
A n = are given by
0
1
( ) (2.16)
!
n
n n
d
A Nu
n d

=
=
Thus,
0
0
1
1 0 1 1
0
2 2
2 0 2 0 1
( ) ,
( ) ( )( ) ,
1
( )( ) ( )( ) ,
2!
A Nu
d
A Nu DN u u
d
A DN u u D N u u

=
=
=
= =
= +

And so on, where
[ ]
0
( )
r
D N u denotes the r-th derivative of N at
0
u X . We
should note that while
n
A is a polynomial in
0 1 2
, , ,..., u u u its dependence on
0
u is decided by the form of N and its derivatives. However generally,
n
A
depends only on
0 1
, ,...,
n
u u u so that (2.16) allows successive calculation of

, 0,1, 2,...
n
u n = at any stage the solution can be approximated by the partial
sum
1
0
N
N j
j
s u

=
=



2.3 Wang's transformation
Define ( ), ( ), x f y f = = then
( ) ( ) .
( ) ( )
,
( )
dy d d
f
dx d dx
f f
dx
f
d

=

= =

Using this in (2.1a) we get
0
( ) ( ) ( ) 0 (2.17)
dy
f f f
dx
+ =
Also assuming ( ) 0, f we can write (2.1a) as
0
( )
( ) (2.18)
( )
f
f
f

Integrate from 0 to .
[ ]
0
0
0
0
0
0
0
ln( ( )) ( )
( )
ln ( )
(0)
( ) (0) exp ( ) (2.19)
f f s ds
f
f s ds
f
f f f s ds

=
| |
=
|

\

=

Therefore ( ) 0 f > if (0) 0. f > Our assumption that ( ) 0 f is justified.
We can cancel ( ) f in Eq.(2.17). We get
0
( ) 0
dy
f
dx
+ =
Differentiate again with respect to x


2
0 2
2
0 2
2
0 2
( ) 0
0
0
d y d
f
dx dx
d y x
or
dx
dx
d
d y x
or
dx y

+ =
+ =
+ =

The above equation was obtained by Wang [24]. The new boundary
condition are as follows
x=0 corresponds to 0
df
d
= which occurs at 0 = . At this point
2
2
d f
y
d

= = . Thus (0) y = .
Also
0
( )
dy
f
dx
= . At 0, = the right side will be zero giving us (0) 0 y = .
The condition lim 1
df
d

= indicates that the domain over which Wang's

equation holds is [0,1). From (2.19) we conclude that ,
lim ( ) 0 f

=
This condition transforms to
1
lim ( ) 0
x
y x

=

2.4 S0lution of Wang's equation by Adomian
decomposition method
0, 1, (2.20)
x
y o x
y
+ = < <
Subject to the conditions
(0) 0, (1) 0, (2.21) y y = =
Along with the condition
(0) (2.22) y =

Following the standard procedure of the ADM [4], we write Eq. (2.20) in
the operator form
, (2.23) Ly Ny =
Where the linear operator
2
2
d
L
dx
= is easily invertible
x
Ny
y
= represents
the nonlinear term. Applying the inverse operator
0 0
(.)
x x
dxdx

to both sides of
Eq.(2.23) and incorporating the first condition of (2.21) and condition
(2.22) yield
1
( ) ( ) (2.24) y x L Ny

= +
The ADM defines the solution y(x) by the series:
0
n
n
y y

=
=

The nonlinear term Ny is defined by:
0
n
n
Ny A

=
=

where the Adomian polynomials
n
A can be generated for all types of
nonlinearities according to algorithm set by Adomian [4],
0
0
1
, 0,1, 2,... (2.25)
!
n
i
n i n
i
d
A N y n
n d

=
=
| |
= =
|
\

In the ADM, the components
n
y are determined by the recursive algorithm:
1
0 1
, ( ), 0
k k
y y L A k

+
= =
Some of the Adomian polynomials for
x
Ny
y

= calculated form (10) are

0
0
1
1 2
0
2
1 2
2 3 2
0 0
3
3 1 2 1
3 2 3 4
0 0 0
1
,
,
2
,
A x
y
y
A x
y
y y
A x
y y
y y y y
A x
y y y
| |
=
|
\
| |
=
|
\
| |
=
|
\
| |
= +
|
\

For numerical comparison purposes, we construct the solution
0 1 2 3
( ) ( ) ( ) ( ) ( ) ... as y x y x y x y x y x = + + + +
lim ( ) ( ), (2.26)
n
n
x y x

=
Where ( )
n
x is the n-term approximate solution. The theoretical treatment
of the convergence of the ADM has been considered and discussed in [1,2].
Below we present first few members of the set of Adomian polynomials
associated with this problem as well as the functions
0 1 2
, , ... y y y calculated
from them.
0 0
3 4
1 1 3
6 7
2 2 3 5
9 10
3 3 5 7
12 13
4 4 7 9
15 16
5 5 9 11
, ,
, ,
6 6
, ,
180 30
, ,
2160 144
2099
, ,
19008 1425600
2099 31453
, ,
2993766000 99792000
x
y A
x x
y A
x x
y A
x x
y A
x x
y A
x x
y A

= =
= =
= =
= =
= =
= =


The value of can be found by using the last condition in (2.21), i.e.
y(1)=0.

By solving (1) 0, 1, 2.., 6
k
k = = , we find
1
2
3
4
5
6
0.408248
0.441743
0.452576
0, 457674
0.460566
0.462404

=
=
=
=
=
=



2.5 Pade approximation
Hashim [15] improved the above values by approximating the function y
into a rational function with respect to the variable x using diagonal Pade
approximation. He employed the Maple built-in diagonal Pade
approximants together with the condition (1) 0
n
= to get the approximate
value for .
He presented the calculated values for using the
n
for both the ADM
and ADM-Pade approximation. The ADM with Pade approximation gives
more accurate results compared with the standard ADM without Pade
approximation. The accuracy of the solution will be further improved if
more terms in the series solution are used. Hashim's results are reproduced
in Table 3.
Level of Pade approximation Value of
[3/3] 0.447214
[6/6] 0.463257
[9/9] 0.466791
[12/12] 0.466799
Table 3: Evaluation of using Pade approximation

Chapter 3 Analytical Solution of Wang's Equation

3.1 Introduction
In this chapter we shall find an analytical solution of Wang's equation
by a method independent of Adomian decomposition method. This method
can easily extend the series (2.7) to as many terms as we wish. From the
results in [15,24] it seemed that one could determine to arbitrary
accuracy once a large number of terms of the series (2.7) was available.
Unfortunately this expectation turns out to be false. We shall derive two
sequences one of them increasing the other decreasing, both of them
converging to . The convergence is hopelessly slow. However the two
sequences enable us to check accuracy of the approximation at each step.
We shall establish that the integral
2
2
2
0
d f
d
d

| |
|
\

is bounded below by
3
4

and above by . If we use the value of 0.469606 = then we get the exact
result
2
2
2
0
0.37118
d f
d
d

| |
=
|
\

We use the above result to find an approximate asymptotic expression for
( ) f . The power series solution (2.2) is known to be valid for 0 5.69
[18]. This solution combined with an asymptotic solution will represent the
function ( ) f analytically on the entire domain 0 .

3.2 Blasius solution
First we obtain Blasius solution give by (2.2) of the last chapter.
We want to solve
0
( ) ( ) ( ) 0 (3.1)
(0) 0, (0) 0, (0)
f f f
f f f

+ =
= = =


Let
0
( ) ,
n
n
n
f a

=
=


Successivederivatives give us
1
1
2
2
3
3
( ) ,
( ) ( 1) ,
( ) ( 1)( 2) ,
n
n
n
n
n
n
n
n
n
f na
f n n a
f n n n a

=
=
=
=

Put them in (3.1)
3 2
0
3 2 0
( 1)( 2) ( 1) 0
n k n
n k n
n k n
n n n a k k a a


= = =
+ =


The above expression is equivalent to
3 0 2
0 0 0
( 3)( 2)( 1) ( 2)( 1) 0
n k n
n k n
n k n
n n n a k k a a

+ +
= = =
+ + + + + + =


Or
3 0 2
0 0 0
( 3)( 2)( 1) ( 2)( 1) 0 (3.2)
m k n
m k n
m k n
m m m a k k a a

+
+ +
= = =
+ + + + + + =


Define . Therefore . m k n n m k = + =
Since 0, this implies 0 . n k m . Eq. (3.2) becomes
3 0 2
0 0 0
3 0 2
0 0
3 0
( 3)( 2)( 1) ( 2)( 1) 0
( 3)( 2)( 1) ( 2)( 1) 0
Sinceeach coefficient in theaboveseries must vanish, we get
( 3)( 2)( 1) ( 2)
m
m m
m k m k
m m k
m
m
m k m k
m k
m
m m m a k k a a
m m m a k k a a
m m m a k

+ +
= = =

+ +
= =
+
+ + + + + + =

+ + + + + + =


+ + + + +


2
0
0
3 2
0
( 1) 0
or
( 2)( 1) (3.3)
( 3)( 2)( 1)
m
k m k
k
m
m k m k
k
k a a
a k k a a
m m m

+
=
+ +
=
+ =
= + +
+ + +

Since
1 2
0, 0, ,
2!
o
a a a

= = = we get from (3.3)

2
0
3 4 5
2 3
0
6 7 8
3 4
0
9 10 11
0, 0,
5!
11
0, 0,
8!
375
0, 0,
11!
a a a
a a a
a a a



= = =
= = =
= = =

It is clear that we shall get non zero
m
a for m=2,5,8,11,
Let 3 1, 1, 2, 3,... m = =
3 1
0
3 2 2 3 1
0
( 2)( 1) , 0, 3, 6,..., 3 3
(3 )(3 1)(3 2)
k k
k
a k k a a k

+ +
=
= + + =
+ +

We shall get non zero coefficients if
2 2, 5, 8,..., for this reason wedefine 2 3 2, 0,1, 2,...
or 3 , 0 1
k k r r
k r k
+ = + = + =
=

Now Eq. (3.3) transforms to
1
0
3 2 3 2 3 3 1
0
(3 2)(3 1) , (3.4)
(3 )(3 1)(3 2)
r r
r
a r r a a

+ +
=
= + +
+ +

In order to get the solution in the form (2.2) we define
1
3 2
3 2
(3.5)
(3 2)!
b
a

+
+
+
=
+

and replace
3 2 3 2
,
r
a a
+ +
etc. in (3.4). We get
1 1 1
3 2 0 3 2 3 3 1
0
(3 2)(3 1) .
(3 2)! 3 (3 1)(3 2) (3 2)! (3 3 1)!
r r
r r
r
b b b
r r
r r

+ +
+ +
=
= + +
+ + + +

Simplification gives
1
3 2 3 3 1
3 2 0
0
1
0 3 2 3 3 1
0
(3 1)!
(3 )!(3 3 1)!
3 1
(3.6)
3
r r
r
r r
r
b b
b
r r
b b
r

+
+
=

+
=
=

| |
=
|
\

From (3.5) we find
2
2
2!
b
a =

Since
2
,
2
a

= we get
2
1 b =
Formula (3.6) with
2
1 b = produces the solution given by Blasius [10] in
1908.

3.3 Solution of Wang's equation
Wang's equation can be made a useful tool to elucidate analytical
features of the Blasius problem. He solved this equation by Adomian
decomposition method. However this method requires a lot of computation
including double integration. In this section we shall solve Wang's equation
by the same approach as was used to find Blasius solution. Re-write Eq.
(2.6a) in the form
(3.7) yy x =
We assume the solution in the form
0
( ) (3.8)
n
n
n
y x a x

=
=

From the boundary conditions (2.6b), it is clear that
0 1
(0), 0 (3.9) a f a = =
We define (0) f = . We are following the same notation as used by Wang
[24] and Hashim [15], since in (2.1a) has now been put equal to unity.
Differentiate y twice and substitute y and y in (3.7). We get
2
0 2
( 1) (3.10)
n k
k n
k n
a a n n x x

+
= =
=

Writing first few terms of the series on the left we obtain
2 2
2 1 2 3 4 1 3 2
3
5 1 4 2 3
2 (2 6 ) (12 6 2 )
(20 12 8 ) ... (3.11)
a a a a x a a a a x
a a a a a x x

+ + + + + +
+ + + =




from which we get
2 3 4 5
1
0, , 0, 0 (3.12)
6
a a a a

= = = =
Define new summation index 2 m n k = + and change the order of
summation in (3.10). We get
2
0 0
( 2)( 1) (3.13)
m
m
k m k
m k
a a m k m k x x

+
= =
+ + =

On the left side of (3.13) coefficient of each power of x other than unity
must vanish. This gives for 2 m
2
0
( 2)( 1) 0 (3.14)
m
k m k
k
a a m k m k
+
=
+ + =

which can be re-written as
2 0 2
1
( 2)( 1) ( 2)( 1)
m
m k m k
k
m m a a a a m k m k
+ +
=
+ + = + +

We have obtained the recurrence relation
2 2
1
1
( 2)( 1) ( 2)( 1), 2 (3.15)
m
m k m k
k
m m a a a m k m k m

+ +
=
+ + = + +

If 3 is not a divisor of 2 m + it will not divide both k and 2 m k + . From
(3.9) and (3.12) we note that, with the exception of
0
a and
3
a , other
coefficients upto and including
5
a are all zero. Using this in (3.15) we find
that
6
0 a but
7 8
0 a a = = . Applying the principle of mathematical induction
we can easily establish that only those coefficients whose suffixes are
multiples of 3 fail to vanish. We can replace (3.15) by
1
3 3 3 3
1
1
3 (3 1) (3 3 )(3 3 1) , 2 (3.16)
n
n k n k
k
n n a n k n k a a n

=
=

Define
3
3 2 1
(3.17)
(3 )!
n
n n
b
a
n

=
then (3.16) becomes

1
3 3 3 3
2 1 2 1 2( ) 1
1
( ) ( ) ( ) 1
3 (3 1) (3 3 )(3 3 1) .
(3 )! (3 )! (3 3 )!
n
n k n k
n k n k
k
b b b
n n n k n k
n k n k

=

=

Or
1
3 3 3 3
1
(3 2)!
(3 )!(3 3 2)!
n
n k n k
k
n
b b b
k n k

Or
1
3 3 3 3
1
3 2
(3.18)
3
n
n k n k
k
n
b b b
k

=
| |
=
|
\

From (3.17) we get
0
0 0 1
3
3
6
b
a b
b
a

= =
=

therefore
0 3
1, 1 b b = = .
Formula (3.18) with
0 3
1, 1 b b = = gives a solution of Wang's equation
similar to Blasius solution found in the last section.
In (3.18) let n=2. We get
6 3 3
4
4
3
b b b
| |
= =
|
\

Also n=3 give,
2
9 3 9 3 3 6 6 3
1
7 7 7
168
3 3 6
k k
k
b b b b b b b
k

=
| | | | | |
= = + =
| | |
\ \ \

Then (3.17) gives,
6 3 3
9 5 5
4 1
6! 180
168 1
9! 2160
a
a


= =
= =

We calculate first few coefficients in the above manner and list them in the
following lines.

0 0
3 3
6 6 3
9 9 5
12 12 7
6
15 15 9
6
18 18 11
7
21 21 13
24
1,
1
1,
6
1
4,
180
1
168,
2160
1
25200,
19008
7.01125 10
9168432,
1.03002 10
6594561792,
1.61614 10
8257020125920,
16498091239838
b a
b a
b a
b a
b a
b a
b a
b a
b

= =
= =
= =
= =
= =

= =

= =

= =
=
8
24 15
9
27 27 17
10
30 30 19
2.65906 10
208,
4.53471 10
49377815204079398400,
7.95335 10
210964851697107227166720,
a
b a
b a

= =

= =


Hashim [15] found the first nine coefficients (up to
24
a ) of (3.8), in
complete agreement with members of the above list. Equation (3.16) can be
easily used in a computer algorithm to extend the series (2.7) to an arbitrary
number of terms with very little effort. As an example we give below
coefficient of the
1500
x term in the series (3.8).
335
1500 999
4.1809830 10
a

=






3.4 An increasing sequence
To find we have to solve the equation (1) 0, y = where
3
3
0
( ) (3.19)
n
n
n
y x a x

=
=

This gives
3
2 1
1
0 (3.20)
n
n
n
a

=
+ =

Hashim [15] solved (3.20) after truncating the series after 7 terms and
found 0.463662 . He improved this value to 0.466799 by approximating
the series for y(x) up to
21
x by the [12/12] Pade approximant. Since we can
find an arbitrary number of terms in the series (3.20), we can attempt to
solve (3.23)with a large number of terms. Let ( ) F denote the left side of
Eq. (3.20) and let
3
2 1
1
( ) , 1, 2, 3,.... (3.21)
k
n
k n
n
a
F k

=
= + =

and let
k
be a root of the equation
( ) 0 (3.22)
k
F =
We shall solve (3.8) by a slightly modified Newton's method. We first find
1 1
bysolving ( ) 0 F = . To find , 2,
k
k we use the iterative scheme
1 1
1
1
(3.23)
( )
, 1, 2,...
( )
k
k i
i i
k k
x
F x
x x i
F

=
= =

and we approximate the derivative at
1 k

by
1 1
1
( ) ( )
( ) (3.24)
k k k k
k k
F h F
F
h

+
=
and choose h a sufficiently small number. The above scheme works
efficiently and produces a sequence some of whose members are given in
the following table.

1 0.408248 100 0.469365
2 0.441743 200 0.469496
5 0.460566 300 0.469535
7 0.463662 1000 0.469583
15 0.467293 2000 0.469592
25 0.468366 3000 0.469595
50 0.468066 4000 0.469597
k k
k k

Table 4
Roots of Eq. (3.22)
The above sequence increases because for larger k more terms in the series
(3.23) are retained and a larger value of will be needed to satisfy the
equation. The sequence converges very slowly and no estimate about the
accuracy of any term is available. In the next section we shall obtain a
decreasing sequence also converging . The two sequences together will
determine precisely lower and upper bounds for at each level of
approximation.

3.5 A decreasing sequence for .
Hashim [15] and Wang [24] have found the following expression for
1
y

3 6 9 12
3 5 7 9
3
3 2 1
1
1 1 2099
... (3.25)
6 30 144 1425600
1
n
n n
n
x x x x
y
x
c

+
=
= + + + + +
= +

where
3n
c is defined to be the coefficient of
3
2 1
n
n
x

+
in (3.25). The solution
y(x) of (3.7) is related to (3.25) in the following manner
0 0
( ) (3.26)
x x
x
y x dx
y
=


If we use (3.25) in (3.26), we obtain

3 3 3
3 2 1
1
3 3
3
2 1
0
3
3 3
2 1
1
( ) (3.27. )
6 (3 3)(3 2)
(3 3)(3 2)
(3.27. )
3 (3 1)
n
n n
n
n
n
n
n
n
n
n
n
x x
y x c a
n n
c x
n n
c x
b
n n

+
+
=
+
+
=

=
=
+ +
=
+ +
=

Compare (3.27.b) with (3.19), we find
3 3 3
3 (3 1) , 1 (3.28)
n n
c n n a n

=
Let x=1 in (3.27.b) . We shall truncate the series after k terms and find an
approximate sum of the remainder. We have
3 3 3
2 1 2 1
1
0 (1) ...... (3.29)
3 (3 1) (3 3)(3 2)
k
n k
n k
n
c c
y
n n k k

+
=
= =
+ +

If k is large, we can approximately set
2 3 3
3
4 3 6
3
k
k
k
k
c
c
c
c

+
+
=
=

and so on. Eq. (3.29) becomes
3 3 3
2 1 2 1
1
1
0 (3.30)
3 (3 1) (3 3)(3 2)
k
n k
n k
n n k
c c
n n n n

+
= =
=
+ +


It is easy to show that
1 1
(3 3)(3 2) 3 2
n k
n n k

=
=
+ + +

and (3.30) becomes
3 3 3
2 1 2 1
1
0
3 (3 1) (3 2)
k
n k
n k
n
c c
n n k

+
=
=
+

In the notation of Section 3, the above equation may be written as
3 3 2 1
3 3
( ) 0 (3.31)
k k k
k
F a

+ +
+
+ =
Since the approximation used to obtain (3.30) replaces a larger number by a
smaller one, we have

( ) ( ) F G
Where ( ) and ( ) F G denote respectively the left sides of Eqs. (3.20) and
(3.31). The above inequality ensures that if
e
is a zero of ( ) F then a zero
of ( ) G greater than
e
exists. We solve (5.3) for k=1,2, by the modified
Newton's method.
Let
( ) k
denote a root of (5.3). We display a few members of this sequence
in Table 5.
( ) ( ) ( ) ( )
5 0.482463 100 0.470006 900 0.469633 2100 0.4696130
7 0.478413 200 0.469784 1000 0.4696298 3000 0.4696088
15 0.473268 300 0.469716 1500 0.4696190 3500 0.4696074
50 0.470505 500 0.469665 1800 0.4696155 4000 0.469
k k k k
k k k k
6064

Table 5
Roots of Eq. (3.31)
Using the last entries of Tables 4 and 5 we have the following inequality
for
0.469597 0.4696064 (3.32) < <

3.6 Pade approximation
In chapter 2 we discussed Hashim's results which were found by
approximating the power series solution of Wang by a Pade approximant.
letting x=1 and putting the numerator equal to zero. From the [12/12]
approximant he found 0.466799.
We have shown above that the exact value of lies between 0.4695975
and 0.4696064. Thus Hashim's result is correct to only two decimal
positions and, in order to achieve a more accurate result by this method,
Pade' approximants of very high order must be used. However the amount
of work involved in ADM becomes prohibitive after a few terms which

precludes the use of this approach to high orders. We can use our solution
to carry this method to higher orders to generate a sequence of numbers
which should converge to . We find that, up to a point, this sequence
increases and appears to be converging to the exact value of , each term
being a better approximation as compared with the corresponding number
given by a straightforward solution of the equation y(1)=0 after truncating
the series (2.7) and retaining an appropriate number of terms. However
after this stage the Pade' approximations start yielding numbers which,
instead of giving an improved approximation, produce errors of increasing
magnitude. Thus this method appears to be of limited utility.
We illustrate our method by an example. Let us truncate the series (2.7)
after five terms and let x=1. We get an approximate value of by solving
the equation
3 5 7
1 1 1 1
0 (3.33)
6 180 2160 19008

=
If we divide through with
2
1
andlet p

= , the equation becomes

2 3 4
1 0 (3.34)
6 180 2160 19008
p p p p
=
We replace the left side of (3.34) by its [2/2] Pade' approximant. A Pade'
approximant can be easily found by using algebraic softwares like
MATLAB or Mathematica. This gives
2
2
119 73
1
396 3960
0 (3.35)
53 1
1
396 594
p p
p p
+
=
+

By equating the numerator of the fraction on the left side of (3.35) to zero
we find p =4.65968. An approximate value of is given by
1
0.463257
p
= . This agrees with the entry which appears in the table of

Hashim[15] corresponding to his Pade [6/6] approximant. In Table 6 we
present approximate values of calculated by using successive Pade'
approximants starting with the [2/2] approximant and going up to the
[23/23] approximant. For comparison, we also include values of this
parameter obtained by solving an equation of the from(3.34) after steadily
increasing the number of terms. Note that our [23/23] approximant
corresponds to the [69/69] approximant of Hashim assuming that one could
go as far as that by adopting the ADM approach.
2 0.457674 0.463257 14 0.468525 0.469075
3 0.462404 0.466791 15 0.468611 0.469086
4 0.464572 0.468061 16 0.468686 0.469097
5 0.465791 0.468645 17 0.468750 0.469107
6 0.466561 0.468956 18 0.468807 0.469118
7 0.467088 0.4689
i i
i i
i i
80 19 0.468857 0.469123
8 0.467469 0.468997 20 0.468901 0.469124
9 0.467756 0.469011 21 0.468941 0.469053
10 0.467980 0.469025 22 0.468977 0.469977
11 0.468158 0.469038 23 0.469009 0.474672
12 0.468303 0.469051 24 0.469039 no value
13 0.468424 0.469063 25 0.469066 no value

Table 6. Values of
i
found by solving the equation y(1)=0, when we
retain 2i+1 terms in (2.7) and
i
given by the [i/i] Pade' approximant.

A glance at the Table indicates that successive terms of the sequence
i
give better approximation to the exact value of for i =2,3,,22 but
23
is much worse compared with
23
. Also the sequence ends abruptly at
23
because the Pade' [24/24] approximant equated to zero does not yield a
real root leading to a suitable entry in the Table. We are faced with a
dilemma. On the one hand the sequence
i
continues beyond i =23 and the

subsequent terms yield an increasing sequence which converges to , on
the other hand the corresponding Pade' approximant either produces
an
i
which is worthless or it fails to generate any value of
i
at all. This is
strange since it is well known that an [m/m] Pade' approximant gives a
better representation of a function than its Taylor series truncated after 2m
terms. The first 22 terms of Table 4 mirror this fact. The anomalous
behavior after i=23 can be explained in only one way that the Pade'
approximants from the stage [23/23] onwards no longer adequately
represent the corresponding Taylor polynomial on an interval containing
the unique zero of the polynomial. Probably this occurs due to ill-
conditioned matrices which are involved in the calculation of coefficients
in the Pade' approximants of sufficiently high orders.

3.7 The integral
2
0
[ ( )] . f d

Integrate (2.7) with respect to x on [0.1). We get
1
3 5
0
1 1 1
.... (3.36)
4.6 7.180 10.2160
(3.37)
ydx

=
<

Also we have from (3.36)
1
3 5
0
1 1 1 1
( ....)
4 6 180 2160
1
(3.38)
4
3
(3.39)
4
ydx

> + + +
=
=

Eq. (3.38) follows from the line preceding it because satisfies the
equation y(1)=0, which becomes, on letting x=1 in (2.7),
3
1 1
....
6 180

= + +

Now
2
2
and .
d f df
y x
d d
= = Hence
1 2
2
2
0 0
[ ] (3.40)
d f
ydx d
d

=


From (3.37), (3.39) and (3.40) we obtain the inequality,
2
2
2
0
3
(3.41)
4
d f
d
d

| |
< <
|
\

If we use 0.469606 = in (3.36) then we can replace (3.41) by the equality
2
2
2
0
0.37118 (3.42)
d f
d

| |
=
|
\

Write the Blasius equation 0 f ff + = in the form
f
f
f

An integration gives
0
( ) exp ( ) (3.43) f f u du

| |
=
|
|
\

It is clear from (3.43) that for 0 , ( ) 0. f < As a consequence ( ) f is
nonnegative and monotonically increasing and ( ) f is also increasing and
unbounded. Since ( ) f is bounded above by unity, the curve ( ) y f = does
not intersect the line , y = because otherwise ( ) f being a convex
function, its slope, at the point of intersection must exceed unity. This
means the graph of ( ) y f = is flanked on the left by the line . y = The
initial condition lim ( ) 1 f

= indicates that the function ( ) f has an

asymptote with slope unity but lying below the line . y = Let the equation
of this asymptote be
0
y =
Since
0
0
0
( ) ( ) , f d d

an exponentiation leads to the inequality

2
0
0
1
( ) exp ( ) exp ( ) (3.44)
2
f f u du

| |
| |
=
|
|
|
\
\

Square both sides of (3.44) and integrate on [0, ) . We find
2 2
0
0
0.37118 exp{ ( ) } (3.45)

where we have used (3.42). Now
2
0
0
2 2
0
2
2
0
0
0 0
0
exp{ ( ) }
(3.46)
2
u
u u
u
e du
e du e du
e du

=
= +
= +

Combining (3.45) and (3.46) we obtain
0
2
2
0
0.37118
2
0.79695 (3.47)
u
e du

The above inequality is satisfied if
0
1.16 (3.48)
The above result indicates that the line
1.16 (3.49) y =
is approximately an asymptote to the curve ( ). y f = Since this curve
eventually behaves like a straight line of slope unity, it follows that, for
large ,
( ) 1.16 f
A numerical solution of the Blasius problem indicates that
0
1.23. = The
above result gives this number with an error less than 6 percent.
We were able to apply a classical method to solve Wang' equation which
gave the solution to an arbitrary number of terms. In principle a modern

method such as the Adomian decomposition method also gives the solution
to as many terms as we wish but the computational effort involved to
calculate the 3000th term, for example, would be tremendous. However our
advantage in easy calculation of the two sequences
(k)
{ }and{ }
k
, is
somewhat neutralized by their extremely slow convergence to .































Chapter 4 Limitations of Adomian Decomposition Method

4.1 Introduction
In the last two chapters we solved Blasius equation and its transformed
version, Wang equation, by Adomian decomposition method as well as by
a direct method.
In both cases the solution comes out in the form of an infinite series.
However the series converges within a finite interval and it is impossible to
get any information about the solution for large values of . To evaluate
(0) f , correct to six decimal places, we need to find a series solution of the
Wang equation up to several thousand terms.
Even after using this value of (0) f to solve the Blasius problem, the series
solution converges only for 5.90 . To get a complete picture we have to
solve the Blasius problem numerically such as Runge-Kutta method.
Liao has developed a method, called the homotopy analysis method,
which attempts to accelerate the converge of the series solution. In chapter
1 we described his method of solution. However there are a number of
arbitrary parameters in the method and one has to depend on guess work to
estimate a suitable value of them.
In general it is useful to make an asymptotic analysis of a problem
before tackling it by the Adomian or the homotopy analysis methods. Such
methods produce series solutions which, for nonlinear problems, only
converge within a small interval.





4.2 A simple problem
To highlight the deficiencies of the Adomian and the homotopy
analysis method, we consider a very simple nonlinear problem
2
2
2
(4.50)
(0) 0, (0) 0 (4.51)
d y
y x
dx
y y
+ =
= =

First we find an approximate analytical solution for large x. Consider a
transformation
(4.52) y u x = +
Then
1
2
3
2
2 2
1
2
1
(4.53)
4
and 2 (4.54)
y u x
y u x
y u x u x

= +
=
= + +

For 1 x >> , the second term on the right of (4.53) may be dropped. Also we
assume u x << , an assumption justified later. Therefore we drop
2
u in
(4.54) and Eq. (4.50) is transformed to a linear equation
2 0 (4.55) u x u + =
Now it is well-known that the equation
2
2 2 2 2 2 2 2
2
(1 2 ) [( ) ] 0 (4.56)
r
d y dy
x s x s r a r x y
dx dx
+ + + =
has the general solution
1 2
[ ( ) ( )] (4.57)
s r r
y x c J ax c Y ax

= +
Multiply (4.55) with
2
x . We get
5
2
2
2 0 (4.58) x u x u + =



Compare (4.58) with (4.56). We get
2 2 2
2 2
1 2 0
0
5
2
2
2
s
s r
r
a r

=
=
=
=

Therefore
1 5 2 4 2
, , ,
2 4 5 5
s r a = = = = and the general solution of
(4.55) is
2
5
5 5
4 4
1 2 2
5
4 2 4 2
[ ( ) ( )]
5 5
u x c J x c Y x = +
Since the Bessel functions are oscillatory with decreasing amplitude such
that
( ) 0 as
and ( ) 0 as
J x x
Y x x

our assumption that u x << is justified for large x.
From (4.52) we see that the solution of Eq. (4.50), for large x, will be small
oscillations superimposed on the parabola y x = .
In Fig.4 we present the numerical solution of the problem which confirms
our qualitative analysis.


0 10 20 30 40 50
0
1
2
3
4
5
6
7

Fig.4 : Numerical solution of Eq. (4.50).


In Fig.5 we present the solution along with the curve y x = ,
shown dashed .
0 10 20 30 40 50
0
1
2
3
4
5
6
7

Fig. 5 : The solution of Eq. (4.50) together with y x = shown dashed.



4.3 Solution by Adomian method

If we solve the problem by Adomian decomposition method, we
get the following solution
3 8 13 18 23
28 33
38 43
x x x 95 31x
6 2016 943488 48502831104 9203412201984
74849x 189251x
13507369511417413632 21395673306085183193088
38700523x 1061168785x
2801579230592089310166515712 50011106613036187601787024
x
y = + +
+
+
48
53
58
310272
1527994261379x
47494608245909822416547577359008530432
986872159994333x
20444184729624559311214550180022959197913088
3659317869389202813517x
50960628520373191763798652378101228449208971561009

63
152
63328266525722085749107x
597156645001733061088192608566590194967830728751905243136
+

In Fig. 6 we plot the above solution. It is clear that the series diverges
when 3.5 x > and the oscillatory behavior depicted in Figs. 4 and 5 is
nowhere to be seen.
0 0.5 1 1.5 2 2.5 3 3.5
0
1
2
3
4
5
6

Fig. 6

If we add one more term to the solution it becomes

3 8 6 13
-9 18 -12 23 -15 28
-18 33 -20 38 -23 43
-26 48 -29 53 -32 58
y= 0.166667x 0.000496032x 1.0599 10
1.95865 10 3.36832 10 5.54135 10
8.84529 10 1.38138 10 2.12187 10
3.2172 10 4.82715 10 7.18068 10
1.0605
x
x x x
x x x
x x x

+
+
+ +
+
+
-34 63 -37 68
10 1.55666 10 . x x


The graph of shown in Fig.7. Again the divergence of the series is clear
beyond x =3.3



0 0.5 1 1.5 2 2.5 3 3.5
-3
-2
-1
0
1
2


Fig. 7







4.4 Homotopy Analysis Method

It has been shown by Liao [19] that the Adomian solution up to m
terms
0
m
n
n
n
y a x
=
=


Can be improved by the homotopy analysis method to
,
0
( ) (4.59)
m
n
m n n
n
y h a x
=
=


where
,
( )
m n
h is called an approaching function and is defined as
( )
,
0
1
( ) ( ) 1 (4.60)
m n
k
n
m n
k
n k
h h h
k

=
+ | |
= +
|
\

where h is arbitrary and should be chosen so that the solution (4.59) agrees
with the exact solution over as wide an interval as possible.
If we choose m =7 , the homotopy solution becomes

2 3 4 5 6 3
2 2 3 4 5 8 3 2 3 4 13
4 2 3 18 5 2 23
5 2
1
(1 )(1 )
6
(1 ) (1 2 3 4 5 6 ) (1 ) (1 3 6 10 15 )
2016 943488
95(1 ) (1 4 10 20 ) 31(1 ) (1 5 15 )
48502831104 9203412201984
74849(1 ) (1 5 15 )
y h h h h h h h x
h h h h h h x h h h h h x
h h h h x h h h x
h h h
= + + + + + +
+ + + + + + + + +
+
+ + + + +
+
+ +

28 6 33
189251(1-h) (1 6 )
13507369511417413632 21395673306085183193088
x h x +
+

We plot the above solution for various values of h and observe that the
homotopy solution also suffers from the convergence problem. In any case
it fails to produce the oscillatory behavior of the solution shown in Fig. 4
and 5 .





0 0.5 1 1.5 2 2.5 3 3.5
0
1
2
3
4
5
6
7

Fig. 8 : Homotopy solution with h = 0.1







0 0.5 1 1.5 2 2.5 3 3.5
0
1
2
3
4
5
6
7

Fig. 9 : Homotopy solution with h = -0.1


0 0.5 1 1.5 2 2.5 3 3.5
0
1
2
3
4
5
6

Fig. 10 : Homotopy solution with h = 0.2





0 0.5 1 1.5 2 2.5 3 3.5
-3
-2
-1
0
1
2

Fig. 11 : Homotopy solution with h = -0.2




0 1 2 3 4
-3
-2
-1
0
1
2

Fig. 12 : Homotopy solution with h = 0.5




0 0.5 1 1.5 2 2.5 3 3.5
-40
-30
-20
-10
0
10
20

Fig. 13 : Homotopy solution with h =- 0.5

We note that the homotopy solution in Figs. 11-13 produces a zero but the
exact solution is free of zeros. This shows that a bad choice of the
parameter h can lead to a solution which is totally wrong even for small
values of x.

4.5 Conclusion
By applying the Adomian and homotopy analysis method to our
simple nonlinear problem we have shown that these methods produce
series solutions which converge over a relatively short interval and it is
impossible to get any information for large x. For this one should turn to
asymptotic analysis of the problem.































References

1. K. Abbaoui, Y. Cherruault, New ideas for proving convergence of
decomposition methods, Comput. Math. Appl. 29 (1995) 103-108.
2. K. Abbaoui, Y. Cherruault, Convergence of Adomian's method
applied to differential equations, Comput. Math. Appl. 28 (1994)
103-109.
3. S. Abbasbandy, Anumerical solution of Blasius equation by
Adomian decomposition method and comparison with homotopy
perturbation method, Chaos, Solitons and Fractals 31 (2007) 257-
260.
4. G. Adomian, Solving Frontier Problem of Physics: The
Decomposition Method, Kluwer, Boston, MA, 1994.
5. G. Adomian, A review of the decomposition method in applied
mathematics, J. Math. Anal. Appl. 135 (1988) 501-544.
6. G. Adomian, Stochastic Systems, Academic Press, New York,
1986.
7. A. Asaithambi, A finite difference method for the solution of the
Falkner-Skan equation, Appl. Math. Comput. 156(2004), 779-786.
8. A. Asaithambi, Solution of the Falkner-Skan equation by recursive
evaluation of Taylor coefficients, J. Comput. Appl. Math.
176(2005), 203-214.
9. F. Ahmad, Application of Wang equation to the Blasius problem,
Electronic Journal Technical Acoustics (2007), 2.
10. H. Blasius, Grenschten in Flussigkeiten mit kleiner reibung, Z.
Math. Physik,56(1908), 1-37. English translation NACA-TM-1256.
11. B.J. Cantwell, Introduction to Symmetry Analysis, Cambridge
University Press, Cambridge, 2002.

12. R.Cortell, Numerical solutions of the classical Blasius flat-plate
problem, Appl. Math. Comput. 170 (2005) 706-710.
13. T. Fang, F. Guo, C. F. Lee, A note on the extended Blasius
equation, Applied Math. Lethers 19 (2006) 613-617.
14. K.L. Guo, Numerical Heat Transfer, University of Science and
Technology of China Press, Beijing, 1988.
15. I. Hashim, Comments on " A new algorithm for solving classical
Blasius equation," by L. Wang, Appl. Math. Comput. , (2006), in
the press.
16. Ji-Huan He, A simple perturbation approach to Blasius equation,
Appl. Math. Comput. 140 (2003) 217-222.
17. L. Howarth, On the solution of the laminar boundary layer
equations, Proc. London Math. Soc. A164(1938), 547-579.
18. S. J. Liao, An explicit, totally analytic approximate solution for
Blasius viscous flow problems, Int. J. Non-Linear Mech. 34 (1999),
759-778.
19. S. J. Liao, Beyond Perturbation : Introduction to homotopy analysis
method, Chapman and Hall/CRC, New York, 2004.
20. S. J. Liao, A uniformly valid analytic solution of 2-D viscous flow
over a semi-infinite flat plate, J. Fluid. Mech. 385 (1999) 101-128.
21. A. M. Lyapunov, General problem on Stability of Motion, Taylor
and Francis, London, (1992).
22. H. Schlichting and K. Gersten, Boundary Layer Theory, 8
th

revised edition, Springer, Berlin 2000.
23. H. Stephan, Differential Equations, Their Solutions Using
Symmetries, Cambridge University Press, Cambridge 1989.
24. L. Wang, A new algorithm for solving classical Blasius equation,
Appl. Math. Comput. 157 (2004), 1-9.