See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/264949489

Note on the integral function Ni(x) and its elementary properties

Article  in  International Journal of Pure and Applied Mathematics · January 2011

CITATIONS READS

0 23

2 authors:

M. H. Hamdan Merzik T. Kamel

University of New Brunswick, Saint John, New Brunswick… University of New Brunswick
136 PUBLICATIONS   496 CITATIONS    37 PUBLICATIONS   257 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Grid Distortion Errors in Streamfunction Coordinates View project

Sedimentation of suspensions View project

All content following this page was uploaded by M. H. Hamdan on 12 January 2016.

The user has requested enhancement of the downloaded file.

International Journal of Pure and Applied Mathematics
————————————————————————–
Volume 70 No. 4 2011, 549-557

A NOTE ON THE INTEGRAL FUNCTION N i(x) AND

ITS ELEMENTARY PROPERTIES
M.H. Hamdan1 § , M.T. Kamel2
1,2 Department
of Mathematical Sciences
University of New Brunswick
P.O. Box 5050, Saint John, N.B., E2L 4L5, CANADA

Abstract: In this note, we discuss the integral function Ni(x) which has
recently been introduced. Some features of this function make it attractive in
the solution to non-homogeneous Airy’s differential equation. We will examine
some of its elementary properties, its derivatives, and the differential equations
it satisfies.

AMS Subject Classification: 33E99

Key Words: Airy’s non-homogeneous equation, Ni(x) function

1. Introduction

In their analysis of flow over porous layers, Nield and Kuznetsov, [5], formulated
the physical problem by introducing a transition layer of variable thickness be-
tween a porous layer and a free-space channel. This resulted in a mathematical
problem involving the non-homogeneous Airy’s differential equation (see [1],
[8]), namely
y ′′ − xy = f (x). (1)
In their elegant analysis, Nield and Kuznetsov, [5], introduced an integral func-
tion, Ni(x), to facilitate obtaining the solution to their resulting governing
equation. Some interesting properties of Ni(x) were discussed by Nield and
Kuznetsov, [5], and point out to the need to look further into this function
and its promising features. The aim of the current work is to investigate some
elementary properties of Ni(x), its connection to the Airy and Scorer functions,
its nth derivative, and the ordinary differential equations this function satisfies.

Received: February 7, 2011 c 2011 Academic Publications, Ltd.

§ Correspondence author
550 M.H. Hamdan, M.T. Kamel

This work is organized as follows. In Section 2, we provide an overview

of solutions to equation (1). In Section 3 we discuss Ni(x) and some of its
properties. We devote Section 4 to establishing the relationships between Ni(x),
the Airy and Scorer functions. Sections 5 and 6 are devoted, respectively, to
establishing derivatives of Ni(x) and the values of the derivatives at x = 0.
In Section 7 we provide a few of the differential equations that Ni(x) satisfies.
Throughout this work, prime notation is used to denote ordinary differentiation.

2. Solutions to Airy’s Equation

As is well-known, (cf. [1], [2], [6], [7], [8]), general solution to equation (1) takes
the following forms. When f (x)= -1/π, solution is given by

y = c1 Ai (x) + c2 Bi (x) + Gi(x) (2)

and when f (x)= 1/π, solution is given by

y = c1 Ai (x) + c2 Bi (x) + Hi(x). (3)

The functions Ai(x) and Bi(x) are the linearly independent homogeneous Airy’s
functions defined by:

1 ∞
 
1 3
Z
Ai (x) = cos xt + t dt, (4)
π 0 3
1 ∞ 1 ∞
   
1 3 1 3
Z Z
Bi (x) = sin xt + t dt + exp xt − t dt, (5)
π 0 3 π 0 3

with their Wronskian given by, [1]:

1
W (Ai (x) , Bi (x)) = Ai (x) Bi′ (x) − Bi (x) Ai′ (x) = . (6)
π
The functions Gi(x) and Hi(x) are the Scorer functions, see [1], [2], [6], (or
the non-homogeneous Airy’s function, as they represent particular solutions to
equation (1)), which are defined by:

1 ∞
 
1 3
Z
Gi (x) = sin xt + t dt, (7)
π 0 3

1 ∞
 
1 3
Z
Hi (x) = exp xt − t dt. (8)
π 0 3
A NOTE ON THE INTEGRAL FUNCTION N i(x) AND... 551

It can be seen from (7), (8) and (5) that

Gi (x) + Hi (x) = Bi(x). (9)

Particular values of Airy’s and Scorer’s functions at x = 0, expressed in terms

of the Gamma function are as follows, [1], [2], [8]:

1 1 1 1
Gi (0) = Hi (0) = Bi (0) = √ Ai (0) = 7/6 2 , (10)
2 3 3 3 Γ( 3 )
1 1 1 1
Gi′ (0) = Hi′ (0) = Bi′ (0) = − √ Ai′ (0) = 5/6 1

. (11)
2 3 3 3 Γ 3

3. The Integral Function Ni(x)

Nield and Kuznetsov [5], defined N i (x) by

Z x Z x
N i (x) = Ai (x) Bi(t)dt − Bi(x) Ai(t)dt, (12)
0 0

with its first derivative given by

Z x Z x
′ ′ ′
N i (x) = Ai (x) Bi (t)dt − Bi (x) Ai (t)dt (13)
0 0

and obtained the following interesting zeroes of N i (x) and N i′ (x):

N i (0) = N i′ (0) = 0. (14)

We recognize here that another important derivative of N i (x) is its second,

which we express in the following form that involves the Wronskian, W (Ai (x),
Bi (x)):
Z x Z x
′′ ′′ ′′
N i(x) = Ai (x) Bi (t)dt − Bi (x) Ai(t)dt − W (Ai (x) , Bi (x)) . (15)
0 0

where, as given in equation (6), W (Ai (x) , Bi (x)) = π1 .

The value of the second derivative at x = 0 is given in terms of the said
Wronskian, namely
′′ 1
N i (0) = −W (Ai (0) , Bi (0)) = − . (16)
π
552 M.H. Hamdan, M.T. Kamel

Now, using the method of variation of parameters, and with the help of (16),
the following expression for a particular solution to (1) when f (x) = constant
= R, is obtained with great ease:
 x x 
1
Z Z
yp = ′′ Ai (x) f (t) Bi (t) dt − Bi(x) f (t) Ai (t) dt
N i (0) 0 0
R
= ′′ N i (x) (17)
N i (0)

For the special cases of f (x)= -1/π, and f (x)= 1/π, general solutions to (1)
take the following forms, respectively:

y = c1 Ai (x) + c2 Bi (x) + N i(x), (18)

y = c1 Ai (x) + c2 Bi (x) − N i(x). (19)

Clearly, solutions (18) and (19) offer some advantages over solutions (2) and
(3), which include the use of one function, Ni(x), in the particular solutions of
equation (1), as opposed to the two functions Gi(x) and Hi(x)

4. Relationships between Ni(x) and the Scorer Functions

There exist relationships between Ni(x) and the Scorer functions Gi(x) and
Hi(x) that we can establish with the knowledge of the following integrals of the
Airy’s homogeneous function, reported in [8]:
Z x
1 n ′ ′
o
Ai (t)dt = + π Ai (x) Gi (x) − Gi (x) Ai (x) , (20)
0 3
Z x
2
Ai (t)dt = − − π Ai′ (x) Hi (x) − Hi′ (x) Ai(x) ,

(21)
3
Z 0x n ′ o
′
Bi (t)dt = π Bi (x) Gi (x) − Gi (x) Bi (x) , (22)
Z0 x
Bi (t)dt = −π Bi′ (x) Hi (x) − Hi′ (x) Bi(x) .

(23)
0

Multiplying (22) by Ai(x) and (20) by Bi(x), then subtracting the latter prod-
uct from the former, and making use of the Wronskian of Ai (x) and Bi (x), we
obtain:
1
N i (x) = Gi (x) − Bi (x) . (24)
3
A NOTE ON THE INTEGRAL FUNCTION N i(x) AND... 553

In a similar fashion, multiplying (23) by Ai(x) and (21) by Bi(x), then sub-
tracting the latter product from the former, and making use of the Wronskian
of Ai (x) and Bi (x), we obtain:

2
N i (x) = Bi (x) − Hi(x). (25)
3
Now, from (24) and (25) we obtain:

2 1
N i (x) = Gi(x) − Hi(x). (26)
3 3
The following integral representation for Ni(x) can then be obtained by using
any one of equations (24), (25) or (26) together with the appropriate integral
representations of equations (4), (5), (7) and (8):
Z ∞   Z ∞  
2 1 3 1 1 3
N i (x) = sin xt + t dt − exp xt − t dt. (27)
3π 0 3 3π 0 3

5. Derivatives of Ni(x)

First derivative of N i (x) was obtained by Nield and Kuznetsov, [5], and is
given by equation (13). We have obtained the second derivative in equation
(15). In what follows we will develop an iterative formula for higher derivatives
of N i (x) that are of utility when dealing with nth order differential equations
that N i (x) satisfies.
We can write the second derivative of N i (x) in the following equivalent
forms:
 Z x Z x 
′′
N i (x) = x Ai (x) Bi (t)dt − Bi(x) Ai (t)dt
0 0
− W (Ai (x) , Bi (x)) , (28)
′′
N i (x) = xN i (x) − W (Ai (x) Bi (x)). (29)
Similarly, third-order derivative of N i (x) can be written in the following equiv-
alent forms:

 x
Z
′′′  ′
N i (x) = Ai (x) + xAi (x) Bi (t)dt − [Bi (x)
0
554 M.H. Hamdan, M.T. Kamel

Z x
′

+xBi (x) Ai (t)dt, (30)
0
′′′
N i (x) = N i (x) + xN i′ (x) . (31)
We can obtain higher derivatives by repeated differentiation of (30) or (31), as
follows. Repeated differentiation of (31) yields the following iterative formula:
N i(n) (x) = (n − 2)N i(n−3) (x) + xN i(n−2) (x) ; n¿2. . . . (32)
Alternatively, we can express second and higher derivatives in terms of Ni(x)
N i′ (x) and W (Ai (x) , Bi (x)) = π1 by assuming that the nth derivative is of
the form:
′
N (n) (x) = g (x) N i (x) + h (x) N i (x) − p (x) W (Ai (x) Bi (x)), (32)
where g(x), h(x), and p(x) are the coefficients of Ni(x), N i′ (x) and W(Ai(x),Bi(x)),
respectively, that appear in the nth derivative Then the n + 1st derivative is
given by:

N (n+1) (x) = g′ (x) + xh (x) N i (x) + g (x) + h′ (x) N i′ (x)

   

− [h (x) + p′ (x)]W (Ai (x) Bi (x)). (33)

Equation (33) takes the following form in terms of Ai (x) and Bi (x) :
n h ′
i oZ x
(n+1) ′ ′

N (x) = g (x) + xh (x) Ai (x) + g (x) + h (x) Ai (x) Bi (t)dt
0
n h ′
i ′ oZ x
− g′ (x) + xh (x) Bi (x) + g (x) + h (x) B i (x)

Ai (t)dt
0
h ′
i
− h (x) + P (x) W (Ai (x) , Bi (x)) (34)

6. Values of Ni(x) and its Derivatives at x = 0

In this section, we derive iterative formulae for evaluating the derivatives of

Ni(x) at x = 0 They may be of utility when dealing with initial value problems
of nth order differential equations that Ni(x) satisfies.
Nield and Kuznetsov, [5] provided the following values:
′
N i (0) = N i (0) = 0. (35)
In this work we have shown that:
′′ 1
N i (0) = −W (Ai (0) , Bi (0)) = − (36)
π
A NOTE ON THE INTEGRAL FUNCTION N i(x) AND... 555

Now, taking x = 0 in the n+1st derivative (equation (33) or (34)), we obtain:

N (n+1) (0) = − π1 [h (0) + p′ (0)] ; n = 2, 3, 4, . . . . . . (38)
′
In using (38) we need the values of h(0)and p (0), which must be deter-
mined from the nth derivative. Alternatively, we develop the following iterative
formula for more convenient computations:
N i(n+1) (0) = (n − 1) N i(n−2) (0) for n = 2,3, 4, . . . . . . (39)
As an illustration of the resulting pattern, using (39) gives the following
first eleven derivatives of N i (x) at x = 0:
′ ′′′
N i (0) = N i (0) = N i (0) = N iv (0) = N vi (0) = N vii (0) = N ix (0)
= N x (0) = 0. (37)

′′ 1 3 18 162
N i (0) = − ; N iv (0) = − ; N iviii (0) = − ; N ixi (0) = − . (38)
π π π π

7. Differential Equations that Ni(x) Satisfies

The function Ni(x) satisfies the following nth order ordinary differential equa-
tion, obtained from (33):
h ′
i ′ h ′ i 1 ′
y (n) − g (x) + h (x) y − g (x) + xh (x) y = − [h (x) + P (x)]. (39)
π
For example, Ni(x) satisfies the following equations of orders two to seven,
computed iteratively using (39):

′′ 1
y − xy = − , (40)
π

y ′′′ − xy ′ − y = 0, (41)
1
y iv − 2y ′ − x2 y = − x, (42)
π
′ 3
y v − x2 y − 4xy = − , (43)
π
1
y vi − 6xy ′ − 4 + x3 y = − x2 ,
 
(44)
π

′ 8
y vii − 10 + x3 y − 9x2 y = − x. (45)
π
556 M.H. Hamdan, M.T. Kamel

It is worth noting that equation (41) is a special case of the so-called comparison
differential equation that received interest in the literature (see [8], page 108,
and Langer, [3,4]), namely:
y ′′′ − xy ′ − µy = 0. (46)
When µ = 1 (46) reduces to (41). Solutions to (46) have been discussed in
detail in [8]. However, we see here that N i (x) is a solution to (46) for the case
of µ = 1. Furthermore, as discussed in [8], if y is a solution of (46) then z = y ′
is a solution of the following differential equation:
z ′′′ − xz ′ − (1 + µ)z = 0. (47)
Accordingly, since the function y = N i (x) is a solution to (41), then the func-
tion z = N i′ (x) is a solution of
z ′′′ − xz ′ − 2z = 0. (48)
In fact, this is easily verified by computing the fourth derivative of N i (x) and
substituting in equation (48).

Conclusion

The main theme of this work has been a discussion and development of
elementary properties the function Ni(x) possesses. This integral function, in-
troduced by Nield and Kuznetsov, [5], can offer a viable alternative to the
solution of the non-homogeneous Airy’s differential equation. We discussed the
various ordinary differential equations that Ni(x) satisfies. In particular, Ni(x)
satisfies a third-order equation that is a special case of Langer’s comparison
equation, [3,4]

References

[1] M. Abramowitz, I.A. Stegun, Handbook of Mathematical Functions, Dover,

New York (1984).
[2] A. Gil, J. Segura, N.M. Temme, On non-oscillating integrals for computing
inhomogeneous Airy functions, Mathematics of Computation, 70 (2001),
1183-1194.
[3] R.E. Langer, On the asymptotic forms of the solutions of ordinary differ-
ential equations of the third order in a region containing a turning point,
Trans. Am. Math. Soc., 80 (1995), 93-123.
A NOTE ON THE INTEGRAL FUNCTION N i(x) AND... 557

′′′ ′
[4] R.E. Langer, The solutions of the differential equation v +λ2 w +3µλ2 v =
0, Duke Math., 22 (1955), 525-541.

[5] D.A. Nield, A.V. Kuznetsov, The effect of a transition layer between a
fluid and a porous medium: Shear flow in a channel, Transport in Porous
Media, 78 (2009), 477-487.

[6] R.S. Scorer,

R x2 Numerical evaluation of integrals of the form
I R = iφ(x)
x1 f (x)
e dx and the tabulation of the function
1 ∞ 1 3
π 0 sin uz + 3 u du, Quarterly J. Mech. Appl. Math., 3 (1950),
107-112.

[7] N.M. Temme, Special Functions: An Introduction to the Classical Func-

tions of Mathematical Physics, John Wiley and Sons, New York (1996).

[8] O. Vallée, M. Soares, Airy Functions and Applications to Physics, World

Scientific, London (2004).
558

View publication stats