DIFFERENTIAL EQUATIONS

January 4, 1980

Objectives
At the end of the course, the student should be able to produce analytic
solutions of standard rst and second order linear and non-linear dierential
equations including power series solutions, non linear dierential equations in-
cluding power series solutions, and to handle analyitcal solutions of some ele-
mentary functions

Course Outline
Partial dierentiation including rst and second order partial derivatives,
total derivatives, change of variables for two independent variables. First order
equations; their origin, solutions and applications. Second order linear equa-
tions; homogeneous with constants and variable coecients and their applica-
tions. Systems of Linear dierential equations and their applications. Non-
lineat dierential equations. Linear equations of nth order. Introduction to
partial dierential equations and their applications.

References
1. R Bronson: Theory and Problems of Dierential equations, 2nd edition
Schaum Outline Series. Mc Graw Hill, 1994.
2. RK. Nagle and EB sta: Fundamentals of Dierential equations, 4th
edition, Addison-Wesley 1996.
3. SJ Sales and E Hille Calculus One and Several variables, 7th edition,
Wesley 1995.

1
Contents

TABLE OF CONTENTS 2

1 PARTIAL DIFFERENTIATION 4
1.1 First Order Partial derivatives . . . . . . . . . . . . . . . . . . . . 4
1.2 Second Order Partial dierential equations . . . . . . . . . . . . . 9
1.3 Total Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.4 The Chain Rule for functions of functions . . . . . . . . . . . . . 19
1.5 Implicit Functions . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2 ORDINARY DIFFERENTIAL EQUATION 28

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.1.1 Solution of a Dierential equation . . . . . . . . . . . . . 30
2.2 Methods of Solving Dierential equations . . . . . . . . . . . . . 30
2.2.1 Change of Variables for two independent variables . . . . 30
2.2.2 Homogeneous Dierential Equations . . . . . . . . . . . . 34
2.2.3 Equations Reducible to Homogeneous Forms . . . . . . . 40

3 LINEAR EQUATIONS 48
3.1 Leibnitz equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.2 Bernoulli equation . . . . . . . . . . . . . . . . . . . . . . . . . . 51

4 EXACT DIFFERENTIAL EQUATIONS 56

4.1 Equations Reducible to Exact Dierential Equations . . . . . . . 59

5 APPLICATION OF FIRST ORDER DIFFERENTIAL EQUA-

TIONS 62
5.1 In Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.2 Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

6 SECOND ORDER DIFFERENTIAL EQUATIONS 70

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

2
 6.2 The D-Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
6.2.1 The Inverse of D-Operator f (D)
1
. . . . . . . . . . . . . . . 79

7 CAUCHY HOMOGENEOUS LINEAR EQUATIONS 97

8 LEGENDER LINEAR DIFFERENTIAL EQUATION 105

9 SIMULTANEOUS DIFFERENTIAL EQUATIONS 112

10 VARIATION OF PARAMETERS 119

11 METHOD OF UNDETERMINED COEFFICIENTS 130

12 LAPLACE TRANSFORMATIONS 134

13 SOLUTIONS OF DIFFERENTIAL EQUATIONS IN SERIES138

13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
13.1.1 Analytic Function . . . . . . . . . . . . . . . . . . . . . . 139
13.1.2 Ordinary Point . . . . . . . . . . . . . . . . . . . . . . . . 139
13.1.3 Singular Point . . . . . . . . . . . . . . . . . . . . . . . . 140
13.1.4 Regular Singular Point . . . . . . . . . . . . . . . . . . . . 140
13.1.5 Irregular Singular Point . . . . . . . . . . . . . . . . . . . 140
13.2 Solution About an Ordinary Point . . . . . . . . . . . . . . . . . 141
13.3 Solution about a regular Singular Point . . . . . . . . . . . . . . 147

14 TAYLOR SERIES 155

15 PARTIAL DIFFERENTIAL EQUATIONS 157

15.1 Solution of pde's by direct integration . . . . . . . . . . . . . . . 159

3
1 PARTIAL DIFFERENTIATION

1.1 First Order Partial derivatives

A variable depends upon more than one variable. For example production de-
pends upon variables 'labour' and 'capital'. A variable Z is said to be a function
of two variables x and y, if for each given ordered pair of real numbers(x, y),
there exists one and only one value of z. We use the notation to denote the value
of the function(x, y)and write z = f (x, y). If z is a function of variables x and
y, then z is called the dependent variable and (x, y)are called the independent
variables. The set of ordered pairs of real numbers (x, y)for which the function
z = f (x, y)is dened is called the domain of the function z = f (x, y).

Let z = f (x, y)be a function of two independent variables x and y. By holding

y a constant, z = f (x, y)essentially becomes a function of x and y only and so
we can talk of its derivative with respect to. The derivative of z with respect
to x, keeping y constant is called the partial derivative of z with respect to x,
keeping y constant and is written as ∂z δf
δx , δx , fx , zx , fx (x, y). Mathematically,

∂z f (x + ∂x) − f (x)
= lim
δx δx→0 ∂x

∂z
δx measures the instantaneous rate of change of z with respect to x while y
is held constant. Similarly, the derivative of z with respect to y keeping x
constant is called the partial derivative of z with respect to y and is written as
∂z δf
δy , δy , fy , zy , fy (x, y). Mathematically,

∂z f (y + ∂y) − f (y)
= lim
δy δy→0 ∂y

∂z
δy measures the instantaneous rate of change of z with respect to y while x is
held constant.
Partial derivatives rules include scalar multiplication rule, addition
Example:
1. Find ∂z
δx and ∂z
δy of z = 5x6 y 4

4
Solution
∂z δ
(1)
4
5x6 y 4 = 30x5 y


=
δx δx
δz δ
5x6 y 4 = 20x6 y 3 (2)


=
δy δy

2. Find ∂z
δx and ∂z
δy of
(i)z = 3x2 + 6xy + 9y 4
(ii)z = 7x2 y 3 + 8x5 y 8
(iii)z = (4x − 6y)7
Solution
(i)z = 3x2 + 6xy + 9y 4
∂z δ
3x2 + 6xy + 9y 4 = 6x + 6y (3)


=
δx δx
δz δ
3x2 + 6xy + 9y 4 = 6x + 36y 3 (4)


=
δy δy

(ii)z = 7x2 y 3 + 8x5 y 8

∂z δ
7x2 y 3 + 8x5 y 8 = 14xy 3 + 40x4 y 8 (5)


=
δx δx
δz δ
7x2 y 3 + 8x5 y 8 = 21x2 y 2 + 64y 7 x5 (6)


=
δy δy

(iii)z = (4x − 6y)7

∂z δ
(4x − 6y)7 = 7(4x − 6y)6 · 4 = 28(4x − 6y)6 (7)


=
δx δx
δz δ
(4x − 6y)7 = 7(4x − 6y)6 · −6 = −42(4x − 6y)6 (8)


=
δy δy

2. Determine ∂z
δx and ∂z
δy of
(i) z = xy + y x
(ii)z = (x + 2y)10 (4x − y)12
(iii)z = 4x−9y
5x+2y

Solution
(i) z = xy + y x
∂z δ
= (xy + y x ) = yxy−1 + y x lny (9)
δx δx
δz δ
= (xy + y x ) = xy lnx + xy x−1 (10)
δy δy

5
(ii)z = (x + 2y)10 (4x − y)12
∂z δ  10 12

10 δ 12 12 δ 10
= (x + 2y) (4x − y) = (x + 2y) (4x − y) +(4x − y) (x + 2y)
δx δx δx δx
(11)
= (x + 2y)
10
· 12 (4x − y)
11
· 4 + (4x − y)
12
· 10 (x + 2y) (12)

= 48 (x + 2y)
10
(4x − y)
11
+ 10 (4x − y)
12
(x + 2y) (13)

= (x + 2y) (4x − y)
9 11
[48(x + 2y) + 10(4x − y)] (14)

= (x + 2y) (4x − y)
9 11
[88x + 86y] (15)

= 2 (x + 2y) (4x − y)
9 11
[44x + 43y] (16)
δz δ  10 12

10 δ 12 12 δ 10
= (x + 2y) (4x − y) = (x + 2y) (4x − y) +(4x − y) (x + 2y)
δy δy δy δy
(17)
= (x + 2y)
10
· 12 (4x − y)
11
· −1 + (4x − y)
12
· 10 (x + 2y) · 2
9
(18)

= −12 (x + 2y)
10
(4x − y)
11
+ 20 (4x − y)
12
(x + 2y)
9
(19)

9
= (x + 2y) (4x − y)
11
[−12(x + 2y) + 20(4x − y)] (20)

= (x + 2y) (4x − y)
9 11
[68x − 44y] (21)

= 4 (x + 2y) (4x − y)
9 11
[17x − 11y] (22)

(iii)z = 4x−9y
5x+2y 
∂z δ 4x − 9y 4(5x + 2y) − 5(4x − 9y) 53y
= = 2
= (23)
δx δx 5x + 2y (5x + 2y) (5x + 2y)2
 
δz δ 4x − 9y −9(5x + 2y) − 2(4x − 9y) −53x
= = = (24)
δy δy 5x + 2y (5x + 2y)2 (5x + 2y)2

3. Determine ∂z
δx and ∂z
δy of
(i) z = (7x3 + 5xy + 9x2 y 3 + 10y 5 )11
Let u = 7x3 + 5xy + 9x2 y 3 + 10y 5 ⇒ z = u11 =⇒ δz = 11u10
δz δz δu
= 11(7x3 + 5xy + 9x2 y 3 + 10y 5 )10 · 21x2 + 5y + 18xy 3 (25)


= ·
δx δu δx
δz δz δu
= 11(7x3 + 5xy + 9x2 y 3 + 10y 5 )10 · 5x + 27x2 y 2 + 50y 4 (26)


= ·
δy δu δy

6
(ii)z = e2x
2
+4xy 3 +y 7

Let u = 2x2 + 4xy 3 + y 7 ⇒ z = eu =⇒ δz = eu δu

δz δz δu
(27)
2 3 7
= e2x +4xy +y · 8x + 4y 3


= ·
δx δu δx
δz δz δu
(28)
2 3 7
= e2x +4xy +y · 12xy 2 + 7y 6


= ·
δy δu δy

(ii)z = ln 8x5 + 3xy 3 + y 6 + 4

Let u = 8x5 + 3xy 3 + y 6 + 4 ⇒ z = lnu =⇒ δz = u1 δu

δz δz δu 1 4 3

40x4 + 3y 3
= · = 5 · 40x + 3y =
δx δu δx 8x + 3xy 3 + y 6 + 4 8x5 + 3xy 3 + y 6 + 4
(29)
δz δz δu 1 2 5

9xy 2 + 6y 5
= · = 5 · 9xy + 6y =
δy δu δy 8x + 3xy 3 + y 6 + 4 8x5 + 3xy 3 + y 6 + 4
(30)
4. If u = ln x3 + y 3 + z 3 − 3xyz , show that δu δu δu 3


δx + δy + δz = x+y+z

Solution
δu 3x2 − 3yz
= 3
δx x + y 3 + z 3 − 3xyz

δu 3y 2 − 3xz
= 3
δx x + y 3 + z 3 − 3xyz

δu 3z 2 − 3yx
= 3
δx x + y 3 + z 3 − 3xyz

δu δu δu 3x2 − 3yz + 3y 2 − 3xz + 3z 2 − 3yx

+ + = (31)
δx δy δz x3 + y 3 + z 3 − 3xyz


3 x2 − yz + y 2 − xz + z 2 − yx 3
= = (32)
(x + y + z) (x2 − yz + y 2 − xz + z 2 − yx) x+y+z

5. If z is a function of x and y, given by z x2 + y 2 = x + y . Show that

δz δz
x δx + y δy = −z

Solution
x+y
z=
x2 + y 2


δz x2 + y 2 − 2x (x + y) x2 + y 2 − 2x2 − 2xy −x2 + y 2 − 2xy
= 2 = 2 = 2
δx (x2 + y 2 ) (x2 + y 2 ) (x2 + y 2 )


δz x2 + y 2 − 2y (x + y) x2 + y 2 − 2xy − 2y 2 −y 2 + x2 − 2xy
= 2 = 2 = 2
δy (x2 + y 2 ) (x2 + y 2 ) (x2 + y 2 )

7
 ! !
δz δz −x2 + y 2 − 2xy −y 2 + x2 − 2xy
x +y =x 2 +y 2 (33)
δx δy (x2 + y 2 ) (x2 + y 2 )

1  3
−x + xy 2 − 2x2 y + yx2 − y 3 − 2xy 2

= 2
(x2 + y2 )


−1 3 2 2 3
 − (x + y) x2 + y 2 x+y
= 2 x + x y + xy + y = 2 =− 2 = −z
2 2
(x + y ) 2 2
(x + y ) x + y2
(34)
6. If u = f y
, show that x δu δu


x δx + y δy = 0

Solution
δu y
−y  y 
= f0 y · (−1)x−2 = 2 f 0
δx x x x
 
δu  y  1 1  y 
= f0 = f0
δy x x x x
   
δu δu −y 0  y  1 0 y y 0 y y 0 y
x +y =x f + y f = − f + f =0
δx δy x2 x x x x x x x
(35)
 2
δu 2 δu 2
7. If u = x2 + y 2 + z 2 , show that
p
δu


δx + δy + δz =1

Solution
δu 1 2
− 1 x
= x + y 2 + z 2 2 · 2x = p
δx 2 x + y2 + z2
2

δu 1 2
− 1 y
= x + y 2 + z 2 2 · 2y = p
δy 2 x + y2 + z2
2

δu 1 2
− 1 z
= x + y 2 + z 2 2 · 2z = p
δz 2 x + y2 + z2
2

 2  2  2 !2 !2 !2
δu δu δu x y z
+ + = p + p + p
δx δy δz x2 + y 2 + z 2 x2 + y 2 + z 2 x2 + y 2 + z 2
(36)
x2 y2 z2 x2 + y 2 + z 2
= 2 + + = =1 (37)
x + y2 + z2 x2 + y 2 + z 2 x2 + y 2 + z 2 x2 + y 2 + z 2

Exercise

− 12
1. If u = 1 − 2xy + y 2 , prove that x δu δu 2 3
δx − y δy = y u

2. The production formula of a rm is given by Q = 4L 4 K 4 , L > 0,K > 0.

3 1

Show that
L δQ δQ
δL + K δK = Q

8
3. If f (x, y, z) = x3 + y 3 + z 3 − 3xyz , show that x δf δf δf
δx + y δy + z δz = 3f (x, y, z)

1.2 Second Order Partial dierential equations

Let z = f (x, y)be a function of x and y. The rst order partial derivatives are
given by δz
δx and δz
δy which are also functions of x and y.
The partial derivative of δz
with respect to x, i.e. δ δz
is written as δ2 z
,zxx or


δx δx δx δx2

fxx .
 
The partial derivative of δz
δy with respect to y, i.e. δ
δy
δz
δy is written as δ2 z
δy 2 ,zyy or
fyy .

The partial derivative of δz

with respect to y, i.e. δ δz
is written as δ2 z
,zyx or


δx δy δx δyδx

fyx .
 
The partial derivative of δz
δy with respect to x, i.e. δ
δx
δz
δy is written as δ2 z
δxδy ,zxy or
fxy .
δ2 z
and δy 2 are called second order direct partial derivatives whereas, and δxδy are
2
δ z δ2 z δ z 2

δx2 δyδx

called second order mixed partial derivatives. By Young's theorem if δyδx and δxδy are
2 2
δ z δ z

continuous then δ2 z
= δxδy .
2
δ z
δyδx

Examples
1. Find , ,
δ2 z δ2 z δ2 z
δx2 δy 2 δyδx and δ2 z
δxδy of the function z = 8x2 y − 11xy 3
Solution
∂z δ
8x2 y − 11xy 3 = 16xy − 11y 3


=
δx δx
δz δ
8x2 y − 11xy 3 = 8x2 − 33xy 2


=
δy δy

δ2 z
 
δ δ δ
16xy − 11y 3 = 16y (38)


2
= =
δx δx δx δx

δ2 z
 
δ δ δ
8x2 y − 11xy 3 = 8x2 − 33xy 2 (39)


2
= =
δy δy δy δx

δ2 z
 
δ δ δ
8x2 y − 11xy 3 = 16xy − 11y 3 (40)


= =
δxδy δx δy δx

δ2 z
 
δ δ δ
16xy − 11y 3 = 16xy − 11y 3 (41)


= =
δyδx δy δx δy

9
2. Find , , and of the function z = x2 + 2y
δ2 z δ2 z δ2 z δ2 z

4
δx2 δy 2 δyδx δxδy

Solution
∂z δ  2
4 
3
3
= x + 2y = 4 x2 + 2y · 2x = 8x x2 + 2y
δx δx
δz δ  2
4 
3
3
= x + 2y = 4 x2 + 2y · 2 = 8 x2 + 2y
δy δy

δ2 z
 
δ δ δ 
3 
(42)
2

2
= 8 x2 + 2y 7x2 + 2y


2
= = 8x x + 2y
δx δx δx δx

δ2 z
 
δ δ δ 
3 
(43)
2

2
2
= = 8 x + 2y = 48 x2 + 2y
δy δy δy δy

δ2 z
 
δ δ δ 
3 
2
2
= = 8 x2 + 2y = 24 x2 + 2y · 2x = 48x x2 + 2y
δxδy δx δy δx
(44)
δ2 z
 
δ δ δ 
3 
2
2
= = 8x x2 + 2y = 8x.3 x2 + 2y · 2 = 48x x2 + 2y
δyδx δy δx δy
(45)
3. If u = x2 + 2y 2 , then verify that = δxδy .
p 2 2
δ u δ u
δyδx

Solution
∂z δ  2
1  1 2
− 1 x
= x + 2y 2 2 = x + 2y 2 2 · 2x = 1
δx δx 2 (x + 2y 2 ) 2
2

δz δ  2
1  1 2
− 1 2y
= x + 2y 2 2 = x + 2y 2 2 · 4y = 1
δy δy 2 (x2 + 2y 2 ) 2

δ2 z
 
δ δ δ 
− 1  1
− 3 −2xy
= = 2y x2 + 2y 2 2 = 2y·− x2 + 2y 2 2 ·2x = 3
δxδy δx δy δx 2 (x + 2y 2 ) 2
2
(46)
2
 
δ z δ δ δ 
− 1  1
− 3 −2xy
= = x x2 + 2y 2 2 = x.− x2 + 2y 2 2 ·4y = 3
δyδx δy δx δy 2 (x + 2y 2 ) 2
2
(47)
 2
4. If r2 = x2 + y 2 , show that δ2 r
δx2 · δ2 r
δy 2 = δ2 r
δxδy

Solution
2rδr = 2xδx

δr x
=
δx r
δr
δ2 r δ  x  r − x. δx r − x. xr r 2 − x2 y2
2
= = 2
= 2
= 3
= 3 (48)
δx δx r r r r r

2rδr = 2yδy

10
 δr y
=
δy r
δr
δ2 r δ  y  r − y. δy r − y. yr r2 − y2 x2
= = = = = (49)
δy 2 δy r r2 r2 r3 r3

δ2 r
 
δ δr δ y δr −y x −xy
= = = y(−1)r−2 = 2 · = 3 (50)
δxδy δx δy δx r δx r r r

δ2 r δ2 r y 2 x2 x2 y 2
2
· 2 = 3 · 3 = 6 (51)
δx δy r r r
2 2
δ2 r x2 y 2
 
−xy
= = (52)
δxδy r3 r6

Hence proven.
5.If u = f (x + 2y) + g (x − 2y), show that 4 δδxu2 =
2
δ2 u
δy 2

Solution
δu
= f 0 (x + 2y) · 1 + g 0 (x − 2y) · 1 = f 0 (x + 2y) + g 0 (x − 2y)
δx
δu
= f 0 (x + 2y) · 2 + g 0 (x − 2y) · −2 = 2f 0 (x + 2y) − 2g 0 (x − 2y)
δy

δ2 u δ
= (f 0 (x + 2y) + g 0 (x − 2y)) = f 00 (x + 2y)·1+g 00 (x − 2y)·1 = f 00 (x + 2y)+g 00 (x − 2y)
δx2 δx
(53)
δ2 u δ
2
= (2f 0 (x + 2y) − 2g 0 (x − 2y)) = 2f 00 (x + 2y)·2+(−2 · −2) g 00 (x − 2y) = 4f 00 (x + 2y)+4g 00 (x − 2y)
δy δy
(54)
δ2 u δ2 u
4 2 = 2
δx δy

Hence proven.
6. If u = , then show that x2 δδxu2 + 2xy δxδy
2 2
xy δ u 2
δ2 u
x+y +y δy 2 =0

Solution
δu y(x + y) − (x − y) y2
= 2 = 2
δx (x + y) (x + y)

δu x(x + y) − (x − y) x2
= 2 = 2
δy (x + y) (x + y)
!
δ2 u δ y2 2 −3 −2y 2
= 2 = y · −2 (x + y) · 1 = 3
δx2 δx (x + y) (x + y)

11
 !
δ2 u δ x2 −3 −2x2
= 2 = x2 · −2 (x + y) ·1= 3
δy 2 δy (x + y) (x + y)
!
2
δ2 u δ x2 (x + y) · 2x − 2x2 (x + y) 2xy
= 2 = 4 = 3
δxδy δx (x + y) (x + y) (x + y)
! ! !
δ2 u δ2 u 2
2 δ u −2y 2 2xy −2x2
x2 2 +2xy +y = x2 3 +2xy 3 +y 2 3
δx δxδy δy 2 (x + y) (x + y) (x + y)
(55)
1 1
−2x2 y 2 + 4x2 y 2 − 2x2 y 2
(56)


= 3 = 3 (0) =0
(x + y) (x + y)

Hence proven
7. If x2 + y 2 + z 2 = 1
u2 , show that δ2 u
δx2 + δ2 u
δy 2 + δ2 u
δz 2 =0

Solution
1
u= p
x2 + y 2 + z 2

δu 1
− 3 −x
= − x2 + y 2 + z 2 2 · 2x = 3
δx 2 (x2 + y 2 + z 2 ) 2
!
23 3

21
δ2 u δ −x −1 x2 + y 2 + z 2 +x· 2 x2 + y 2 + z 2 · 2x
= 3 = 3
δx2 δx (x2 + y2 + z2) 2 (x2 + y2 + z2)

23
21
12  2

− x2 + y 2 + z 2 + 3x2 x2 + y 2 + z 2 x2 + y 2 + z 2 3x − x2 + y 2 + z 2
= 3 = 3
(x2 + y 2 + z 2 ) (x2 + y 2 + z 2 )


3x2 − x2 + y 2 + z 2
= 5
(x2 + y 2 + z 2 ) 2


δ2 u 3y 2 − x2 + y 2 + z 2
= 5
δy 2 (x2 + y 2 + z 2 ) 2


δ2 u 3z 2 − x2 + y 2 + z 2
= 5
δz 2 (x2 + y 2 + z 2 ) 2




δ2 u δ2 u δ2 u 3x2 − x2 + y 2 + z 2 3y 2 − x2 + y 2 + z 2 3z 2 − x2 + y 2 + z 2
+ + = 5 + 5 + 5 =0
δx2 δy 2 δz 2 (x2 + y 2 + z 2 ) 2 (x2 + y 2 + z 2 ) 2 (x2 + y 2 + z 2 ) 2
(57)
8. If R = ln x2 + y 2 + z , then prove that x δyδz
2 2 2
δ R δ R δ R
2


= y δzδx = z δxδy

Solution
δR 2x
= 2
δx x + y2 + z2

12
 δR 2y
= 2
δy x + y2 + z2
δR 2z
= 2
δz x + y2 + z2

δ2 R
 
δ 2z
−2 −4yz
= = 2z · −1 x2 + y 2 + z 2 · 2y = 2
δyδz δy x2 + y 2 + z 2 (x2 + y 2 + z 2 )

δ2 R
 
δ 2x
−2 −4xz
= = 2x · −1 x2 + y 2 + z 2 · 2z = 2
δzδx δz x2 + y 2 + z 2 (x2 + y2 + z2 )

δ2 R
 
δ 2y
−2 −4xy
= = 2y · −1 x2 + y 2 + z 2 · 2x = 2
δxδy δx x2 + y 2 + z 2 (x2 + y2 + z2 )
!
δ2 R −4yz −4xyz
x =x 2 = 2 (58)
δyδz (x2 + y2 + z2) (x2 + y2 + z2 )
!
δ2 R −4xz −4xyz
y =y 2 = 2 (59)
δzδx (x2 + y2 + z2) (x2 + y2 + z2 )
!
δ2 R −4xy −4xyz
z =z 2 = 2 (60)
δxδy (x2 + y2 + z2) (x2 + y2 + z2 )

δ2 R δ2 R δ2 R
x =y =z (61)
δyδz δzδx δxδy

Hence proven
9. If u = ln x3 + y 3 − x2 y − xy 2 . Show that δδxu2 + 2 δxδy
2 2 2
δ u −4
+ δδyu2 =


(x+y)2

Solution
 
 
x3 +y 3 −x2 y−xy 2 = x + y x2 − xy + y 2 − xy (x + y) = (x + y) x2 − xy + y 2 − xy =
2
(x + y) (x − y)
h i
2
u = ln (x + y) (x − y) = ln (x + y) + 2ln (x − y)
δu 1 2 3x + y
= + = 2
δx x+y x−y x − y2
δu 1 2 −x − 3y
= − = 2
δy x+y x−y x − y2


δ2 u 3 x2 − y 2 − 2x (3x + y) 3x2 − 3y 2 − 6x2 − 2xy −3x2 − 2xy − 3y 2
 
δ 3x + y
2
= 2 2
= 2 = 2 = 2
δx δx x − y (x2 − y 2 ) (x2 − y 2 ) (x2 − y 2 )


δ2 u −3 x2 − y 2 + 2y (−x − 3y) −3x2 + 3y 2 − 2xy − 6y 2 −3x2 − 2xy − 3y 2
 
δ −x − 3y
= = 2 = 2 = 2
δy δx x2 − y 2 (x2 − y 2 ) (x2 − y 2 ) (x2 − y 2 )

13


δ2 u −1 x2 − y 2 − 2x (−x − 3y) −x2 + y 2 + 2x2 + 6xy x2 + 6xy + y 2
 
δ −x − 3y
= = 2 = 2 = 2
δxδy δx x2 − y 2 (x2 − y 2 ) (x2 − y 2 ) (x2 − y 2 )
!
δ2 u δ2 u δ2 u −3x2 − 2xy − 3y 2 x2 + 6xy + y 2 −3x2 − 2xy − 3y 2
+2 + = 2 +2 2 + 2
δx2 δxδy δy 2 (x2 − y 2 ) (x2 − y 2 ) (x2 − y 2 )
(62)
−3x2 − 2xy − 3y 2 + 2x2 + 12xy + 2y 2 − 3x2 − 2xy − 3y 2 −4x2 + 8xy − 4y 2
= 2 = 2
(x2 − y 2 ) (x2 − y 2 )
(63)
−4(x − y)2 −4
= 2 = 2 (64)
(x2 − y 2 ) (x + y)

10. A utility function is given by u = 2q12 q2 + q1 q23 , Show that the rate of change
of marginal utility q1 with respect to q2 is equal to the rate of change of marginal
utility of q2 with respect to q1 .
Solution
Marginal utility of q1 is given by δu


δq1 = (2q2 ) 2q1 + q23 · 1 = 4q1 q2 + q23
 
Rate of change of marginal utility of q1 with respect to q2 is given by δ
δq2
δu
δq1 =
δ


δq2
3
4q1 q2 + q2 = 4q1 + 3q22

Marginal utility of q2 is given by δu

δq2 = 2q12 + 3q1 q22
 
Rate of change of marginal utility of q2 with respect to q1 is given by δ
δq1
δu
δq2 =
δ


δq1 2q12 + 3q1 q2 = 4q1 + 3q22
2

Therefore the rate of change of marginal utility of q1 with respect to q2 is equal

to the rate of change of marginal utility of q2 with respect to q1 .

1.3 Total Derivatives

Let z = f (x, y)be a function of independent variables x and y . Let δxand δy be
small increments of the independent variables x and y respectively and δz , the
corresponding increment of the dependent variable z.
z + δz = f (x + δx, y + δy)

δz = f (x + δx, y + δy) − f (x, y)

The quantities δxand δy , are called the dierentials of the independent variables

14
x,y respectively and are denoted by dx, dy respectively. δx = dx,δy = dy
 
The quantity δz δz
dy is called the total dierential of the dependent


δx dx + δy

variable z and is written as dz .The increment δz and total dierential dz of z

are nearly equal when the changes in δxand δy in the independent variables
are numerically suciently small. The total dierential dz is much easier to
nd than the incrementδz because the total dierential dz is often taken as an
approximation of the increment δz of the variable z.
Example:
1. If z = 3x2 + xy − 2y 2 . Determine the total dierential dz of z.
Solution
   
δz δz
dz = dx + dy
δx δy

dz = (6x + y) dx + (x − 4y) dy (65)

2. If z = x2 y 3 + exy . Determine the total dierential dz of z.

Solution
   
δz δz
dz = dx + dy
δx δy

dz = 2xy 3 + yexy dx + 3x2 y 2 − xexy dy (66)

3. Find the appropriate change of volume of a right circular cylinder if the

radius is changed from 3 to 3.01 and the altitude from 5 to 4.98.
Solution
V = πr2 h
   
δV δV
dV = dr + dr
δr δr

dV = ((πh)2r) dr + (πr2 )1 dr = (2πrh) dr + πr2 dr

r = 3,dr = 3.01 − 3.00 = 0.01

h = 5,dh = 4.98 − 5.00 = 0.02

dV = 2π(3)(5)(0.01) + π(3)2 (−0.02) (67)

22
= (0.3)π − (0.18)π = 0.12 × = 0.377 (68)
7

15
4. In a right angled triangle, one side changes from 3cm to 3.1cm and the other
changes from 4cm to 3.9cm.
(i) Find the approximate change in area
(ii) Find the approximate change in the hypotenuse
Solution
Let A be the area and z be the hypotenuse
A = 12 xy and z =
p
x2 + y 2
   
1 1
dA = y dx + x dy
2 2
! !
x y
dz = p dx + p dy
x2 + y 2 x2 + y 2

x = 3,dx = 3.1 − 3.0 = 0.1

y = 4,dy = 3.9 − 4.0 = −0.1

Change in area dA,

   
1 1
dA = (4) (0.1) + (3) (−0.1) = (69)
2 2

Change in hypotenusedz ,
   
3 4
dz = √ (0.1) + √ (−0.1) = (70)
3 + 42
2 3 + 42
2

4. Approximate the area of a rectangle of dimensions 35.02cm by 24.97cm.

Solution
A = xy
   
δA δA
dA = dx + dy
δx δy

dA = (y) dx + (x) dy

x = 35,dx = 35.02 − 35 = 0.02

y = 25,dh = 24.97 − 25 = −0.03

dA = (25) (0.02) + (35) (−0.03) = −0.55 (71)

Approximate area is Area = A + dA

= (30 × 35) + (−0.55) = 874.45cm2 (72)

16
5. Approximate the change in the hypotenuse of a right triangle of legs 15 and
20cm when the shorter leg is lengthened by 85 cmand the longer is shortened by
5
16 cm .
Solution
Let x,y.z be the shorter leg, longer leg and the hypotenuse of the triangle re-
spectively.
p
z= x2 + y 2
   
δz δz
dz = dx + dy
δx δy
! !
x y xdx + ydy
dz = p dx + p dy = p (73)
x2 + y 2 x2 + y 2 x2 + y 2

x = 15,dx = 5
8

y = 20,dy = − 16
5

15 85 + 20 − 165



xdx + ydy 1
dz = p = √ = cm (74)
2
x +y 2 2
15 + 20 2 8

Therefore the hypotenuse is lengthened by 18 cm.

6. The power consumed in an electrical resistor is given by P = E2
R watts.
If E=200 volts and R=8 ohms, by how much does the power change if E is
decreased by 5 volts and R is decreased by 0.2 ohms
δP 2E δP −E 2
= and =
δE R δR R2
   
δP δP
dP = dE + dR
δE δR
   2
2E E
dP = dE − dR (75)
R R2

E = 200,dE = −5

R = 8,dR = −0.2
2002
   
2(200)
dP = (−5) − (−0.2) = −250 + 125 = −125volts (76)
8 82

The power is reduced by 125 volts.

7. In measuring a rectangular block of wood, the dimensions were found to
be 25, 30 and 50cm with a possible error of 0.125cm in each of the measure-
ments. Find approximately the greatest error in the surface area of the block

17
and the percentage of error in the area caused by the errors in the individual
measurements.
Solution
The surface area is given by s = 2 (xy + yz + zx) = 2 ((25 × 30) + (30 × 50) + (25 × 50)) =
7000
     
δs δs δs
ds = dx + dy + dz
δx δy δz

ds = 2(y + z)dx + 2(x + z)dy + 2(y + x)dz (77)

The greatest error in s will occur when the errors in the lengths are of the same
sign, say positive. Then
ds = 2(30 + 50)(0.125) + 2(25 + 50)(0.125) + 2(25 + 30)(0.125) = 52.5cm2 (78)

The percentage error is

Error 52.5
× 100% = × 100% = 0.75% (79)
Area 7000

8. In using the formula R = E

C , nd the maximum error and the percentage
error if C = 20with a possible error of 0.1 and E=120 with a possible error of
0.05.
Solution
  
δR δR
dR = dE + dC
δE δC
   
1 E
dR = dE + dC
C C2

The maximum error will only occur when

   
1 120
dR = (0.05) + (−0.1) = 0.0325 (80)
20 202

The percentage error is

dR 0.0325
× 100% = × 100% = 0.41%
R 8

9. The sides of a triangle were measured as 50cm by 65cm and the included
angle as 600 . If the possible errors are 0.06m in measuring the sides and 10 in

18
the angle, determine the greatest error in the computed area.
Solution
A = 12 xysinθ
     
δA δA δA
dA = dx + dy + dθ
δx δy δθ
     
1 1 1
dA = ysinθ dx + xsinθ dy + xycosθ dθ (81)
2 2 2

x = 50,dx = 0.06

y = 65,dy = 0.06

θ = 600 ,dx = π
180

then
     
1 0 1 0 1 0 π 
dA = (65)sin60 (0.06)+ (50)sin60 (0.06)+ (60)(65)cos60 = 17.169m2
2 2 2 180
(82)

1.4 The Chain Rule for functions of functions

If z = f (x, y)is a continuous function of the variable x,y with continuous partial
derivatives δz
δx and δz
δy , and if x and y are dierentiable functions of variable t,
x = g(t)andy = h(t), then z is a function of t and dz
dt is called the total derivative
of z with respect to t and is given by:
dz δz dx δz dy
= +
dt δx dt δy dt

Similarly, if w = f (x, y, z, ...)is a continuous function of the variable x,y,z,...with

continuous partial derivatives s, and if x,y,z,...are dierentiable functions of a
variable t, the total derivative of w with respect to t is given by:
dw δw dx δw dy δw dz
= + + + ...
dt δx dt δy dt δz dt

Example
1. Find dz
dt , given that z = x2 + 3xy + 5y 2 ,x = sintand y = cost
Solution
dz δz dx δz dy
= +
dt δx dt δy dt

19
 = (2x + 3y) cost − (3x + 10y) sint (83)

2. Find dz
, given that z = ln x2 + y 2 ,x = e−t and y = et


dt

Solution
dz δz dx δz dy
= +
dt δx dt δy dt

2 (yet − xe−t )
   
2x 2y
= 2 2
· −e −t
+ 2 2
et = (84)
x +y x +y x2 + y 2

Let z = f (x, y)is a continuous function of the variable x,y with continuous
partial derivatives δz
δx and δz
δy , and let y be a dierentiable function of x. Then z
is a dierentiable function of x and y above.
dz δf dx δf dy
= +
dx δx dx δy dx
dz δf δf dy
= +
dx δx δy dx

The shift notation from z to f is made here in order to avoid possible confusion
of using δz
δx and dz
dx in the same expression
Example
1. Find dz
dx given that z = f (x, y) = x2 + 2xy + 4y 2 and y = eax
Solution
dz δf δf dy
= +
dx δx δy dx
dz
= (2x + 2y) + (2x + 8y) aeax (85)
dx

= 2 (x + y) + 2a (x + 4y) eax (86)

2. Determine δx
δz
and dz
dx given that z = f (x, y) = xy 2 + x2 y and y = lnx
Solution
dz δf δf dy
= +
dx δx δy dx
dz
1
= y 2 + 2xy + 2xy + x2 = y 2 + 2xy + 2y + x (87)


dx x
dz δf dx δf
= +
dy δy dy δy

20
 dz
= y 2 + 2xy x + 2xy + x2 = xy 2 + 2x2 y + 2xy + x2 (88)



dy

3. The altitude of a right circular cone is 15cm and is increasing at 0.2cms−1 .

The radius of the base is 10cm and is decreasing at 0.3cms−1 .How fast is the
volume changing?
Solution
Let x be the radius and y the altitude of the cone. From v = 31 πx2 y , considering
x and y as functions of t.
dv δv dx δv dy
= +
dt δx dt δy dt
 
1 dx dy
= π 2xy + x2 (89)
3 dt dt

1 70π 3 −1
π 2 · 10 · 15 · (−0.3) + 102 (0.2) = − (90)


= cm s
3 3

7. A point P is moving along the curve of intersection of the parabola

2
x2
16 − y9 =

z and the cylinder x2 + y 2 = 5, with x,y and z expressed in cm. If x is increasing

at 0.2cms−1 , how fast is z changing when x=2?

dz δz dx δz dy
= +
dt δx dt δy dt
 
dz  x  dx 2y dy
= − (91)
dt 8 dt 9 dt

Since x2 + y 2 = 5,y = ±1when x = 2, also x dx dy

dt + y dt = 0

When y = 1, dy
dt =
−x dx
y dt = −2
1 (0.2) = −0.4and dz
dt = 2
8 (0.2) − 29 (−0.4) =
5 −1
36 cms

If z = f (x, y)is a continuous function of the variable x,y with continuous partial
derivatives δz
δx and δz
δy ,and if x and y are continuous functions of x = g(r, s)and
y = h(r, s)of the independent variables r and s, then z is a function of r and s

with
dz δz dx δz dy
= +
dr δx dr δy dr

and
dz δz dx δz dy
= +
ds δx ds δy ds

21
Similarly, if w = f (x, y, z, ...)is a continuous function of the variable x,y,z,...with
continuous partial derivatives ,
δw δw
δx δy and δw
δz , and if x,y,z,...are continuous func-
tions of the n independent variables r,s,t,...then
dw δw dx δw dy δw dz
= + +
dr δx dr δy dr δz dr

and
dw δw dx δw dy δw dz
= + +
ds δx ds δy ds δz ds

Examples
1. Find dz
dr and dz
ds , given z = x2 + xy + y 2 , x = 2r + sand y = r − 2s
Solution
dz
= (2x + y) (2) + (x + 2y) = −3y
dr

2. Find δu
δρ , δβ , δθ , given that u = x +2y +2z ,x = ρsinβcosθ , y = ρsinβsinθ and
δu δu 2 2 2

y = ρcosβ

Solution
du δu dx δu dy δu dz
= + +
dρ δx dρ δy dρ δz dρ

= 2xsinβcosθ + 4ysinβsinθ + 4zcosβ (92)

du δu dx δu dy δu dz
= + +
dβ δx dβ δy dβ δz dβ

= 2xρcosβcosθ + 4yρcosβsinθ − 4zρsinβ (93)

du δu dx δu dy δu dz
= + +
dθ δx dθ δy dθ δz dθ

= −2xρsinβsinθ + 4yρsinβsinθ (94)

3. Determine du
dx , given that u = f (x, y, z) = xy + yz + zx, y = x1 , z = x2
Solution
du δf δf dy δf dz
= + +
dx δx δy dx δz dx
   
1 x+z
= (y + z) + (x + z) − 2 + (y + x) 2x = y + z + 2x x + y − (95)
x x2

22
1.5 Implicit Functions
The dierentiation of a function of one variable dened implicitly by a relation
f (x, y) = 0. was earlier treated. For this case , we state without proof.

Theorem 1: If f (x, y)is a continuous function including a point (x0 , y0 ), then

there is a neighborhood of (x0 , y0 ) in which f (x, y) = 0 can be solved for y as a
δf
continuous dierentiable function of x, y = φ(x), with y0 = φ(x0 )and dy
dx = − δf
δx

δy

Example
1. Use the above theorem to show that x2 + y 2 − 13 = 0denes y as a continuous
dierentiable function of x in any neighborhood of the point (2, 3)that does not
include a point of the x-axis. Determine the derivative at that point.
Solution
f (x, y) = x2 + y 2 − 13
δf
dy −x −2
δx
= − δf = = (96)
dx δy
y 3

2. Find dy
dx givenf (x, y) = y 3 + xy − 12 = 0
Solution
δf
dy −y
δx
= − δf = 2 (97)
dx δy
3y + x

3. Find dy
dx givenf (x, y) = ex siny + ey sinx − 1 = 0
Solution
δf
dy ex siny + ex cosx
δx
= − δf =− x (98)
dx δy
e cosy + ey sinx

Extending this theorem,

Theorem 2. If F (x, y, z)is continuous in a region including a point (x0 , y0 , z0 )for
which F (x0 , y0 , z0 ) = 0, if ,
δF δF
δx δy ,and δF
δz are continuous throughout on the re-
gion, and if δF
δz 6= 0at (x0 , y0 , z0 ), then there is a neighborhood of (x0 , y0 , z0 )in

which F (x0 , y0 , z0 ) = 0can be solved for z as a continuous dierentiable function

of x and y, z = φ(x, y)with z0 = φ(x0 , y0 )and

δz − δF
= δFδx
δx δz

23
 δz − δF
δy
= δF
δy δz

Example
1. Find δz
δx and δz
δy , given that sinxy + sinyz + sinzx = 1
Solution
F (x, y, z) = sinxy + sinyz + sinxz − 1
δz − δF
= δFδx
δx δz

ycosxy + zcoszx
=− (99)
ycosyz + xcoszx

δz − δF
δy
= δF
δy δz

xcosxy + zcosyz
=− (100)
ycosyz + xcosxz

Theorem 3: If f (x, y, u, v)and g (x, y, u, v)are continuous in a region including

the point (x0 , y0 , z0 )for which f (x0 , y0 , u0 , v0 ) = 0, if the rst partial derivatives
of f and g are continuous through out
 the region,
 and if at (x0 , y0 , u0 , v0 ),the
 δf δf 
   
determinant is given by J = 
 δu δv 
f,g
u,v 6 0then there is a neighbor-
 =
 δg δg 
 δu δv 
hood of (x0 , y0 , z0 )in which f (x, y, u, v) = 0and g (x, y, u, v) = 0can be solved
simultaneously for u and v, as continuous dierentiable functions of x and y,
 
u = φ(x, y)and v = ϕ(x, y). If at (x0 , y0 , u0 , v0 ), the determinant J f,g
x,y 6=

0then there is a neighborhood of (x0 , y0 , u0 , v0 )in which f (x, y, u, v) = 0and

g (x, y, u, v) = 0can be solved for x and y as continuous dierentiable functions

of u and v, u = φ(x, y)and v = ϕ(x, y).

Examples
1. If u and v are dened as functions of x and y by the equations
f (x, y, u, v) = x + y 2 + 2uv

g(x, y, u, v) = x2 − xy + y 2 + u2 + v 2

Determine (a) δu
δx and δv
δx (b) δu
δy and δv
δy

(a)Dierentiating f and g partially with respect to x, we obtain:

24
 δu δv
1 + 2v + 2u =0
δx δx

and
δu δv
2x − y + 2u + 2v =0
δx δx

Solving these equations simultaneously, for δu

δx and δv
δx , we obtain
δu v + u(y − 2x)
=
δx 2 (u2 − v 2 )

and
δv v (2x − y) − u
=
δx 2 (u2 − v 2 )

(b) Dierentiating f and g partially with respect to y, we obtain:

δu δv
2y + 2v + 2u =0
δy δy

and
δu δv
−x + 2y + 2u + 2v =0
δy δy

Solving these equations simultaneously, for δu

δy and
δv
δy , we obtain
δv v(2y − x) − 2uv
=
δy 2 (u2 − v 2 )

2. Given that u2 − v 2 + 2x + 3y = 0and uv + x − y = 0, determine

(a) , , and
δu δv δu
δx δx δy
δv
δy

(b) δu , δv , δu and
δx δx δy δy
δv

Solution
(a) Here x and y are considered as independent variables . Dierentiating the
given equations with respect to x.
δu δv
2u − 2v +2=0
δx δx

and
δu δv
v +u +1=0
δx δx

Solving these equations simultaneously, for δu

δx and δv
δx , we obtain
δu u+v
=− 2
δx u + v2

and

25
 δv v−u
= 2
δx u + v2

Dierentiate partially with respect to y;

δu δv
2u − 2v +3=0
δy δy

and
δu δv
v +u −1=0
δy δy

Solving these equations simultaneously, for δu

δy and
δv
δy , we obtain
δu 2v − 3u
=
δy 2 (u2 + v 2 )

δv 2u + 3v
=
δy 2 (u2 + v 2 )

(b) Here u and v are to be considered as independent variables. Dierentiate

the given equations partially with respect to u obtaining:
δx δy
2u + 2 +3 =0
δu δu

and
δx δy
v+ − =0
δu δu

Solving these equations simultaneously, for δx

δu and δy
δu , we obtain
δx 2u + 3v
=−
δu 5

and
δy 2 (v − u)
=
δu 5

Dierentiating with respect to v , we obtain

δx δy
−2v + 2 +3 =0
δv δv

and
δx δy
u+ − =0
δv δv

Solving these equations simultaneously, for δx

δv and δy
δv , we obtain
δx 2v − 3u
=
δv 5

and

26
 δy 2u (u + v)
=
δv 5

Exercise:
1. Find dy
dx , given
(a) x3 − x2 y + xy 2 − y 3 = 1and (b) ln x2 + y 2 − tan−1 y



x =0

Ans:(a) (b)
2 2
3x −2xy+y 2x+y
x2 −2xy+3y 2 x−2y

2. Determine δz
δx and δz
δy ,given
(a) 3x2 + 4y 2 − 5z 2 = 60and (b) x + 3y + 2z = lnz
Ans:(a) δz
δx = 3x
5z and δz
δy = 4y
5z

(b) δz
δx = z
1−2z and δz
δy = 3z
1−2z

3. Find all the rst and second partial derivatives of z, given that x2 + 2yz +
2zx = 1

Ans: δx
δz
x+y , δy =
= − (x+z) δz
,
−z δ 2 z
x+y δx2 = ,
x−y−2z δ 2 z
(x+y)2 δxδy
= x+2z
(x+y)2
and δ2 z
δy 2 = 2z
(x+y)2

4. Find the rst partial derivatives of u and v with respect to x and y and the rst
partial derivatives of x and y with respect to u and v given:2u−v+x2 +xy = 0and
u + 2v + xy − y 2 = 0.

Ans: δu
δx = − 5 (4x + 3y); δx =
1 δv 1
5 (2x − y); δu
δy =
1
5 (2y − 3x); δy
δv
= 4y−x δx
5 δu ; =
4y−x δy
2(x2 −2xy−y 2 ) δu ; = y−2x
2(x2 −2xy−y 2 ) ; δx
δv = 3x−2y δy
2(x2 −2xy−y 2 ) δv; = −4x−3y
2(x2 −2xy−y 2 )

5. If u = x + y + z ; v = x2 + y 2 + z 2 and w = x3 + y 3 + z 3 , show that

δx
δu = yz δy
(x−y)(x−z) δv ; = x+z δz
2(x−y)(y−z) δw ; = 1
3(x−z)(y−z)

27
2 ORDINARY DIFFERENTIAL EQUATION

2.1 Introduction
A dierential equation is an equation that involves dierentials or dierential
coecients. An ordinary dierential equation is a dierential equation in which
all the dierential coecients have reference to a single independent variable.
A partial dierential equation is a dierential equation in which there are two
or more independent variables and partial dierentials coecients with respect
to them.
The order of a dierential equation is the order of the highest order derivative
in the equation. For example:
dn y dy
n
+ =0
dx dx

is of the nth order if n>1. The degree of a dierential equation is the degree of
the highest derivative in the equation. For example:
 2
dy dy
y =x +x
dx dx

is of the second degree but rst order.

A dierential equation can be formed from its given solution by elimination of
arbitrary functions of elimination of arbitrary constants. For example:
1. Consider the equation of all circles of radius 1, (x − h)2 + (y − k)2 = 1where
(h, k)are coordinates of the circle. Form a dierential equation from it.

Solution
Dierentiating with respect to x, then
dy
2(x − h) + 2(y − k) =0 (101)
dx

or
dy
(x − h) + (y − k) =0 (102)
dx

Dierentiating again with respect to x

2
d2 y

dy
1 + (y − k) 2 + =0 (103)
dx dx

28
 "  2 #
d2 y dy
(y − k) 2 = − 1 + (104)
dx dx
  2 
dy
− 1 + dx
(y − k) = d2 y
(105)
dx2

From above,
dy
(x − h) = −(y − k) (106)
dx
  2 
dy
dy  1 + dx
(x − h) = d2 y

(107)
dx
 
2 dx

Substituting (x − h)and (y − h)in the circle equation to eliminate the constants

we have:   2 2   2 2
dy dy
 2 1 + dx 1 + dx
dy
 2 2 +1  2 2 =1 (108)
dx d y d y
dx2 dx2

  2 2
dy
1 + dx "  2 #
dy
 2 2 + 1+ =1 (109)
d y dx
dx2

" 2 # 3 2
d2 y
 
dy
1+ = (110)
dx dx2

2. Form a dierential equation of the simple harmonic notation given by:

x = Acos (nt − α)

Solution
Because these are two constants; A and α, the dierential equation is dieren-
tiated twice
dx
= −nAsin (nt − α) (111)
dt
d2 x
= −n2 Acos (nt − α) = −n2 x (112)
dt2

or
d2 x
+ n2 x = 0 (113)
dt2

29
2.1.1 Solution of a Dierential equation
A relation between the variables that satisfy the given dierential equation is
called the solution of the dierential equation. There are two types of solution:
The general solution or complete solution which is one that the number of
arbitrary constants is the order of the dierential equation
The particular solution is that which can be obtained from the general solution
by giving solutions to the arbitrary constants.

2.2 Methods of Solving Dierential equations

2.2.1 Change of Variables for two independent variables

A dierential equation is said to be separable if the variables can be separated.
That is:
F (y)dy = G(x)dx

Once this is done all that is needed is to apply integration on both sides. The
method of solving separable dierential equations can therefore be summarized
as follows:
separate the varables
Integrate
Examples
1. Solve the dierenctial equation2ydy = x2 + 1 dx

Solution
ˆ ˆ
x2 + 1 dx (114)


2ydy =

2y 2 x3
= +x+c (115)
2 3

30
 x3
y2 = +x+c (116)
3

2. Solve the dierential equation dy

= (1 + e−x ) y 2 − 1


dx

Solution
dy
= 1 + e−x dx (117)


y2
−1
ˆ ˆ
dy
= 1 + e−x dx (118)
y2 − 1
ˆ ˆ ˆ
− 21 1
dy + 2
dy = 1 + e−x dx (119)
y+1 y−1
1 1
− ln |y + 1| + ln |y − 1| = x − e−x + c (120)
2 2
 
y − 1
 = 2 x − e−x + c (121)


ln 
y + 1

Let 2c = A  
y − 1
ln 
  = 2x − 2e−x + A (122)
y + 1

3. Solve the dierenctial equationxydx − x2 + 1 dy = 0

Solution
x dy
dx = (123)
x2 + 1 y
ˆ ˆ
x dy
2
dx = (124)
x +1 y
1  2
(125)

ln x + 1 = lny + c
2

4. Solve the dierential equation dy

dx = e3x−2y + x2 e−2y

Solution
dy
e2y = e2y e3x−2y + x2 e−2y (126)


dx

e2y dy = e3x + x2 dx (127)

ˆ ˆ
e2y dy = e3x + x2 dx (128)

e2y 1 x3
= e3x + +c (129)
2 3 3

31
5. Solve the dierential equation 1 − x2 (1 − y) dx = xy (1 + y) dy

Solution
1 − x2 y (1 + y)
dx = dy (130)
x 1−y
ˆ  ˆ 
y2
 
1 y
− x dx = + dy (131)
x 1−y 1−y
ˆ 
x2
  
1 1
lnx − = −1 + + −y − 1 + dy (132)
2 1−y 1−y

x2 y2
lnx − = −y − ln |1 − y| − − y − ln |1 − y| + c (133)
2 2

Eliminating the denominator by multipling R.H.S and L.H.S by 2,

2lnx − x2 = −2y − 2ln |1 − y| − y 2 − 2y − 2ln |1 − y| + 2c (134)

Let 2c = A
2lnx − x2 + 4y + 4ln |1 − y| + y 2 = A (135)

EXERCISE
Solve the dierential equations by variable separable method
1. xdx +secxsinydy = 0Ans: cosy = xsinx + cosx + c
2

x ex +2
2. dx ,y(0) = 1Ans:4y 3 = ex + 2x2 + 3
dy 2
= 6y 2

If we have a dierential equation of the form dy

dx = f (ax + by + c), we put

ax + by + c = tand then separate the variables.

Examples
Solve the following dierential equations
1. dy
dx = (4x + y + 1)
2

Solution
Let 4x + y + 1 = t
Dierentiating with respect to x,
dy dt
4+ = (136)
dx dx
dy dt
= −4 (137)
dx dx

32
Replacing back to the given dierential equation
dt
− 4 = t2 (138)
dx
dt
= t2 + 4 (139)
dx

Separating the variables,

dt
= dx (140)
t2 + 4
ˆ ˆ
dt
= dx (141)
t2 + 4
1 t
tan−1 = x + c (142)
2 2

where c is a constant of integration

Replacing t,
 
1 4x + y + 1
tan−1 =x+c (143)
2 2

2. cos(x + y)dy = dx
Solution
Let x + y = t
dx dt
+1= (144)
dy dy
dx dt
= −1 (145)
dy dy

But cos(x + y) = dx
dy
dt
cost = −1 (146)
dy
dt
cost + 1 = (147)
dy

Separating the variables,

dt
= dy (148)
cost + 1
ˆ ˆ
dt
= dy (149)
cost + 1
ˆ ˆ
dt
= dy (150)
2cos2 2t

33
 ˆ ˆ
1 t
sec2 dt = dy (151)
2 2
 
t
tan =y+c (152)
2
 
y+x
tan =y+c (153)
2

2.2.2 Homogeneous Dierential Equations

A function f (x, y)of variables x and y is called a homogeneous function for any
constant k if
f (kx, ky) = k n f (x, y)

The nmber n is called the degree of the homogeneous function f (x, y)

Example
1. Which of the following functions is homogeneous?
(i) x2 + 7xy
(ii)3x + y 2
(iii)x5 − 4x3 y 2 + 2xy 4 + 3y 5
Solution
(i) x2 + 7xy
f (x, y) = x2 + 7xy (154)

2
f (kx, ky) = (kx) + 7 (kx) (ky) (155)

f (kx, ky) = k 2 x2 + 7xy (156)

This function is homogeneous and is of degree 2.

(ii)3x + y 2
f (x, y) = 3x + y 2 (157)

f (kx, ky) = 3 (kx) + (ky)

2
(158)

f (kx, ky) = k 3x + ky 2 (159)

34
This function is not homogeneous since there still remains a k in the brackets.
(iii)x5 − 4x3 y 2 + 2xy 4 + 3y 5
f (x, y) = x5 − 4x3 y 2 + 2xy 4 + 3y 5 (160)

f (kx, ky) = (kx)5 − 4(kx)3 (ky)2 + 2(kx)(ky)4 + 3(ky)5 (161)

f (kx, ky) = k 5 x5 − 4x3 y 2 + 2xy 4 + 3y 5 (162)

This function is homogeneous and is of degree 5.

2. Determine the degree of the following homogeneous functions
(i) x3 +3xy 2
2xy−y 2

Solution
x3 + 3xy 2
f (x, y) = (163)
2xy − y 2

(kx)3 + 3(kx)(ky)2
f (kx, ky) = (164)
2(kx)(ky) − (ky)2


k 3 x3 + 3xy 2
 3
x + 3xy 2

f (x, y) = 2 = k (165)
k (2xy − y 2 ) 2xy − y 2

This function is homogeneous and is of degree 1.

√ √
(ii) x+ y
x+y

Solution √ √
x+ y
f (x, y) = (166)
x+y
√ √
kx + ky
f (kx, ky) = (167)
kx + ky
1 √ √
√ √ 
k2 x+ y x+ y
f (kx, ky) = =k − 21
(168)
k (x + y) x+y

This function is homogeneous and is of degree − 12 .

In general an expression of the form a0 xn + a1 xn−1 y + a2 xn−2 y 2 + a3 xn−3 y 3 +
... + an y n in which each term is of degree n is called a homogeneous function of

degree n. We can rewrite the expression as:

h i
y y 2 y 3 y n
. This function can be ex-





xn a0 + a1 x + a2 x + a3 x + ... + an x

pressed as xn η y
.To solve a homogeneous dierential equation we put y =


x

35
vxand dy
dx
dv
= v + x dx . We then substitute the values in the given dierential
equation and separate the variables for v and x, integrate them, substitute the
values of v and then get the solution for the dierential equation.
Examples:
1. Solve x2 − y 2 dx = 2xydy

dy x2 − y 2
= (169)
dx 2xy

Let y = vxand dy dv
= v + x dx
dx

dv x − (vx)2
2
x2 − v 2 x2 x2 1 − v 2 1 − v2
v+x = = = = (170)
dx 2x(vx) 2vx2 2vx2 2v

dv 1 − v2 1 − v 2 − 2v 2 1 − 3v 2
x = −v = = (171)
dx 2v 2v 2v

Separating the variables,

2v dx
2
dv = (172)
1 − 3v x
ˆ ˆ
2v dx
2
dv = (173)
1 − 3v x
ˆ ˆ
1 −6v dx
− dv = (174)
3 1 − 3v 2 x
1 
− ln 1 − 3v 2  = ln |x| + c (175)

3

3ln |x| + ln 1 − 3v 2  + 3c = 0 (176)

 

3ln |x| + ln 1 − 3v 2  = −3c (177)

 

ln x3 1 − 3v 2  = −3c (178)



Replacing v = y
x    y 2 
ln x3 1 − 3 (179)

 = −3c
x 

x3 − 3xy 2 = e−3c (180)

Let e−3c = A
x3 − 3xy 2 = A (181)

2. Solve x (x + y) dy − y 2 dx = 0

36
Solution
dy y2
= 2 (182)
dx x + xy

Let y = vxand dy dv
= v + x dx
dx
dv (vx)2 v 2 x2 x2 v 2 v2
v+x = 2 = 2 = = (183)
dx x + x(vx) x + x2 v x2 (1 + v) 1+v

dv v2 v 2 − v (1 + v) v2 − v − v2 −v
x = −v = = = (184)
dx 1+v 1+v 1+v 1+v

Separating the variables

(1 + v) dx
− = (185)
v x
ˆ ˆ
(1 + v) dx
− = (186)
v x
ˆ   ˆ
1 dx
− + 1 dv = (187)
v x

− (ln |v| + v) − ln |x| + c (188)

−ln |v| − v − ln |x| = c (189)

ln |v| + v + ln |x| = c (190)

v + ln |vx| = c (191)

Replacing v = y
x
y y 
+ ln  · x = c
 
(192)
x x
y
+ ln |y| = c (193)
x

y + xln |y| = cx (194)

3. Solve the dierential equation xdy − ydx =

p
x2 + y 2 dx

Solution p
dy x2 + y 2 + y
= (195)
dx x

Let y = vxand dy
dx
dv
= v + x dx

37
 p p √

dv x2 + (vx)2 + (vx) x (1 + v 2 ) + (vx) x 1 + v2 + v p
v+x = = = = 1 + v 2 +v
dx x x x
(196)
dv
(197)
p p
x = 1 + v2 + v − v = 1 + v2
dx

Separating the variables

dv dx
√ = (198)
1 + v2 x
ˆ ˆ
dv dx
√ = (199)
1+v 2 x

put v = tanθand dv = sec2 θdθ

ˆ ˆ ˆ ˆ ˆ
sec2 θdθ sec2 θdθ
 
dv secθ + tanθ
√ = √ = = secθdθ = secθ· dθ = ln |secθ + tanθ|
1 + v2 1 + tan2 θ secθ secθ + tanθ
(200)
 
(201)
p
= ln v + 1 + v 2 
 

 
(202)
p
ln v + 1 + v 2  = ln |x| + c
 

 v + √1 + v 2 
 
ln 

=c

(203)
 x 
√
v+ 1 + v2
= ec (204)
x

Let ec = A √
v+ 1 + v2
=A (205)
x

Replacing v = y
x q
y y 2



+ 1+
x
=A
x
(206)
x
p
y + x2 + y 2
=A (207)
x2

(208)
p
y + x2 + y 2 = Ax2

4.Solve x3 dx − y 3 dy = 3xy (ydx − xdy)

Solution
dy x3 − 3xy 2
= 3 (209)
dx y − 3x2 y

38
Let y = vxand dy dv
= v + x dx
dx

dv x3 − 3x(vx)2 x3 1 − 3v 2 1 − 3v 2
v+x = 3 2
= 3 3
= 3 (210)
dx (vx) − 3x (vx) x (v − 3v) v − 3v


dv 1 − 3v 2 1 − 3v 2 − v v 3 − 3v 1 − 3v 2 − v 4 + 3v 2 1 − v4
x = 3 −v = 3
= 3
= 3
dx v − 3v v − 3v v − 3v v − 3v
(211)
When a dierential equation contains xy a number of times. We can solve this
as an homogeneous equation by putting y
x = v.

Examples
1. Solve xsin xy − ycos xy dx + xcos xy dy = 0

Solution
dy −xsin xy + ycos xy
= (212)
dx xcos xy

Let y = vxand dy dv
= v + x dx
dx
dv −xsinv + vxcosv −sinv + vcosv
v+x = = (213)
dx xcosv cosv
dv −sinv + vcosv −sinv + vcosv − vcosv −sinv
x = −v = = (214)
dx cosv cosv cosv

Separating the variables,

−cosv dx
dv = (215)
sinv x
ˆ ˆ
−cosv dx
dv = (216)
sinv x

−ln |sinv| = ln |x| + c (217)

ln |xsinv| = −c (218)

Replacing y
x = v.
 y 
ln xsin  = −c

(219)
x
y
xsin = e−c (220)
x

Let e−c = A
y
xsin =A (221)
x
   
2. Solve 1 + e y dx + e y 1 − xy dy = 0
x x

39
Solution x
 
dx −e y 1 − xy
= x (222)
dy 1 + ey
x
y = v and dx
dy =
dv
v + y dy

dv −e 1 − xy
v
−ev (1 − v) − v (1 + ev ) −ev + vev − v − vev − (ev + v)
y = −v = = =
dy 1 + ev 1 + ev 1 + ev 1 + ev
(223)
Separating the variables
−dy 1 + ev
= (224)
y v + ev
ˆ ˆ
−dy 1 + ev
= (225)
y v + ev

−ln |y| = ln |v + ev | + lnc (226)

ln |y (v + ev )| = −lnc (227)

y (v + ev ) = −c (228)

Replacing x
y =v
 
x
(229)
x
y + ey = −c
y

Let −c = A
(230)
x
x + ye y = A

2.2.3 Equations Reducible to Homogeneous Forms

An equation of the form dy
dx = ax+by+c
a0 x+b0 y+c0 can be reduced to homogeneous form:
CASE 1:
Where a
a0 6= b
b0 , we put x = X + hand y = Y + kwhere h and k are arbitrary
contants. Then dx = dX and dy = dY .
Therefore the equation becomes:
dy a (X + h) + b (Y + k) + c (aX + bY ) + (ah + hk + c)
= 0 = 0
dx a (X + h) + b0 (Y + k) + c0 (a X + b0 Y ) + (a0 h + b0 k + c0 )

40
We choose h and k such that ah + hk + c = 0and a0 h + b0 k + c0 = 0. From
these, we calculate the values of h and k. From these the equation now becomes
homogeneous i.e:
dy aX + bY
= 0
dx a X + b0 Y

CASE 2:
Where a
a0 = b
b0 = 1
m , then a0 = amand b0 = bmand the equation becomes
dy ax + by + c
=
dx m (ax + by) + c0

ax + by = t,a + b dx
dy
= dt
dx
dy 1 dt


dx = b dx − a

and the equation becomes

 
1 dt t+c
−a =
b dx mt + c0

We can now separate the variables abd integrate accordingly to get thae solution
in t and x, then subtitute back the value of t to obtain the reqiured solution.
Examples:
1. Solve dy
dx = y+x−2
y−x−4

Solution
a
a0 = 1
−1 = −1and b
b0 = 1
1 = 1. Therefore a
a0 6= b
b0 hence we apply case 1.
Let x = X + hand y = Y + k.Then dx = dX and dy = dY .
dY Y + X + (k + h − 2)
= (231)
dX Y − X + (k − h − 4)

Choosing h and k such that k + h − 2 = 0and k − h − 4 = 0and solving for k

and h simultaneously, we obtain: k = 3and h = −1
Therefore the equation now becomes

dY Y +X
= (232)
dX Y −X

which is homogeneous.
Let Y = V X , dX
dY dV
= V + X dX
dV (V X) + X V +1
V +X = = (233)
dX (V X) − X V −1

41
 dV V +1 V + 1 − V (V − 1) V +1−V2+V 1 − V 2 + 2V
X = −V = = =
dX V −1 V −1 V −1 V −1
(234)
Separating the variables,
V −1 dX
dV = (235)
1 − V 2 + 2V X
ˆ ˆ
V −1 dX
dV = (236)
1 − V 2 + 2V X
ˆ ˆ
1 2 − 2V dX
− 2
dV = (237)
2 1 + 2V − V X
1 
− ln 1 + 2V − V 2  = ln |X| + c (238)

2

−ln 1 + 2V − V 2  = 2ln |X| + 2c (239)

 

ln  1 + 2V − V 2 · X 2  = −2c (240)



X 2 1 + 2V − V 2 = e−2c (241)

Let e−2c = A
 2 !
Y Y
X 2
1+2 − =A (242)
X X

X 2 + 2XY − Y 2 = A (243)

x + 1 = X and y − 3 = Y .

Therefore,
2
(x + 1) + 2 (x + 1) (y − 3) − (y − 3) = A
2
(244)

x2 + 2xy − y 2 − 4x + 8y − 14 = A (245)

2. Solve (3y + 2x + 4) dx − (4x + 6y + 5) dy = 0

Solution
dy 3y + 2x + 4
= (246)
dx 4x + 6y + 5
a
a0 = 12 and b
b0 = 12 . Therefore a
a0 = b
b0 hence we apply case 2.
Let t = 2x + 3y . Thus, dt dy
and dx
dy
= 31 dxdt


dx = 2 + 3 dx −2
 
1 dt t+4
−2 = (247)
3 dx 2t + 5

42
 dt 3t + 12
−2= (248)
dx 2t + 5
dt 3t + 12 3t + 12 + 2 (2t + 5) 7t + 22
= +2= = (249)
dx 2t + 5 2t + 5 2t + 5

Separating the variables,

2t + 5
dt = dx (250)
7t + 22
ˆ ˆ
2t + 5
dt = dx (251)
7t + 22
ˆ  9  ˆ
2
− 7
dt = dx (252)
7 7t + 22

2 9
t − ln |7t + 22| = x + c (253)
7 49
2 9
(2x + 3y) − ln |7 (2x + 3y) + 22| = x + c (254)
7 49
2 9
(2x + 3y) − ln |14x + 21y + 22| = x + c (255)
7 49

14 (2x + 3y) − 9ln |14x + 21y + 22| = 49x + 49c (256)

Let 49c = A
14 (2x + 3y) − 9ln |14x + 21y + 22| − 49x = A (257)

3. Solve dy
dx = x+2y−3
2x+y−3

Solution
a
a0 = 12 and b
b0 = 2
1 = 2. Therefore a
a0 6= b
b0 hence we apply case 1.
Let x = X + hand y = Y + k.Then dx = dX and dy = dY .
dY X + 2Y + (h + 2k − 3)
= (258)
dX 2X + Y + (2h + k − 3)

Choosing h and k such that h + 2k − 3 = 0and 2h + k − 3 = 0and solving for k

and h simultaneously, we obtain: k = 1and h = 1
Therefore the equation now becomes

dY X + 2Y
= (259)
dX 2X + Y

which is homogeneous.

43
Let Y = V X , dX
dY dV
= V + X dX
dV X + 2(V X) 1 + 2V
V +X = = (260)
dX 2X + (V X) 2+V

dV 1 + 2V 1 + 2V − V (2 + V ) 1 + 2V − 2V − V 2 1−V2
X = −V = = =
dX 2+V 2+V 2+V 2+V
(261)
Separating the variables, we have:
2+V dX
2
dV = (262)
1−V X
ˆ ˆ
2+V dX
dV = (263)
1−V2 X
ˆ   ˆ
2 V dX
+ dV = (264)
1−V2 1−V2 X
 
1 + V  1 
 − ln 1 − V 2  = ln |X| + C (265)

ln 
1−V  2
 
1 + V 
 − ln 1 − V 2  = 2ln |X| + 2C (266)
 
2ln 
1−V 
 
 1 + V 2 1 
ln 

×  = 2ln |X| + 2C

(267)
 1−V (1 − V ) (1 + V ) 
 
 1+V 
ln 

 = 2ln |X| + A

(268)
 (1 − V )3 
 
 1 1 + V 
ln  2 ·

=A (269)
 X (1 − V )3 

1 1+V
· = eA = B (270)
X 2 (1 − V )3
Y
1 1+ X
·
=B (271)
X2 1 − Y 3
X

X +Y
3 =B (272)
(X − Y )

x − 1 = X and y − 1 = Y .

Therefore,
(x − 1) + (y − 1)
3 =B (273)
((x − 1) − (y − 1))

44
 x+y−2
3 =B (274)
(x − y)

x + y − 2 = B (x − y)
3
(275)

4. Solve (3x − y − 2) dx
dy
= x + 3y − 4

Solution
dy x + 3y − 4
= (276)
dx 3x − y − 2

Solution
a
a0 = 13 and b
b0 = 3
−1 = −3. Therefore a
a0 6= b
b0 hence we apply case 1.
Let x = X + hand y = Y + k.Then dx = dX and dy = dY .
dY X + 3Y + (h + 3k − 4)
= (277)
dX 3X − Y + (3h − k − 2)

Choosing h and k such that h + 3k − 4 = 0and 3h − k − 2 = 0and solving for k

and h simultaneously, we obtain: k = 1and h = 1
Therefore the equation now becomes
dY X + 3Y
= (278)
dX 3X − Y

Let Y = V X , dX
dY dV
= V + X dX
dV X + 3(V X) 1 + 3V
V +X = = (279)
dX 3X − (V X) 3−V

dV 1 + 3V 1 + 3V − V (3 − V ) 1 + 3V − 3V + V 2 1+V2
X = −V = = =
dX 3−V 3−V 3−V 3−V
(280)
Separating the variables,
3−V dX
dV = (281)
1+V2 X
ˆ ˆ
3−V dX
dV = (282)
1+V2 X
ˆ ˆ ˆ
3 V dX
2
dV − 2
dV = (283)
1+V 1+V X
1 
3tan−1 V − ln 1 + V 2  = ln |X| + C (284)

2

6tan−1 V − ln 1 + V 2  = 2ln |X| + 2C (285)

 

45
Let 2C = A
6tan−1 V − ln 1 + V 2  = 2ln |X| + A (286)
 

6tan−1 V = ln X 2 1 + V 2  + A (287)



ButV = Y
X
Y2 
  
Y
6tan−1 = ln X 2 1 + 2  + A (288)

X X

Y
6tan−1 = ln X 2 + Y 2  + A (289)
 
X

x − 1 = X and y − 1 = Y .

Therefore,
y−1  
6tan−1
2 2
= ln (x − 1) + (y − 1)  + A

(290)
x−1

5. Solve (x + y) (dy − dx) = dx + dy

Solution
dy x+y+1
= (291)
dx x+y−1
a
a0 = 11 and b
b0 = 11 . Therefore a
a0 = b
b0 hence we apply case 2.
Let t = x + y . Thus, dt
dx =1+ dy
dx and dt
= dxdy
−1
dx
dt t+1
−1= (292)
dx t−1
dt t+1 t+1+t−1 2t
= +1= = (293)
dx t−1 t−1 t−1

Separating the variables,

t−1
dt = dx (294)
2t
ˆ ˆ
t−1
dt = dx (295)
2t
ˆ ˆ ˆ
1 1
dt − dt = dx (296)
2 2t
1 1
t − lnt = x + c (297)
2 2

t − lnt = 2x + 2c (298)

Let 2c = A

46
 t − lnt = 2x + A (299)

But t = x + y ,
(x + y) − ln |x + y| = 2x + A (300)

Exercise:
Solve the following dierential equations
1. dy
dx = x+y+1
2x+2y+3 Ans:3 (2y − x) + ln |3x + 2y + 4| = C

47
3 LINEAR EQUATIONS

3.1 Leibnitz equation

This is a dierential equation in which the dependent variable and its dierential
coecient occur only in the rst degree and not multiplied together. The stan-
dard form of the linear dierential equation of rst order dx
dy
+p(x)y = Q(x)where

p(x) and Q(x) are functions of x. This equation is called the leibnitz equation.
´
To solve this kind of equation, we multiply both sides by e pdx
to get:
´ ´ ´
pdx dy pdx pdx
e + p(x)e y = Q(x)e
dx
d h ´ pdx i ´
ye = Q(x)e pdx
dx

Integraing both sides,

´
ˆ ´
pdx pdx
ye = Q(x)e

´
The factor e pdx
is called the integrating factor of the dierential equation.
Exarcise:
Solve the following dierential equations by nding a suitable integrating factor.
a) (x + 1) dx
dy
− y = ex (x + 1)
2

Solution
dy y
− = ex (x + 1) (301)
dx x + 1
´ ´
Let p = −1
x+1then pdx = −1
x+1 dx = −ln |x + 1| = ln |x + 1|
−1

´ −1
pdx 1
I.F. = e = eln|x+1| = x+1

Multiplying all through by the integrating factor

1 dy y
· − = ex (302)
x + 1 dx (x + 1)2
 
d y
= ex (303)
dx x + 1
ˆ
y
= ex (304)
x+1
y
= ex + c (305)
x+1

48
 y = (ex + c) (x + 1) (306)

b) 1 + y 2 dx = tan−1 y − x dy

Solution
dy tan−1 y − x
= (307)
dx 1 + y2

dx tan−1 y x
= − (308)
dy 1 + y2 1 + y2

dx x tan−1 y
+ 2
= (309)
dy 1+y 1 + y2
´ ´
Let p = 1
1+y 2 then pdy = 1
1+y 2 dy = tan−1 y
´ −1
pdy
I.F. = e = etan y

Multiplying all through by the integrating factor

−1
tan−1 y dx xetan y tan−1 y tan−1 y
e + = ·e (310)
dy 1 + y2 1 + y2

d h tan−1 y i tan−1 y tan−1 y

xe = ·e (311)
dy 1 + y2
ˆ
tan−1 y tan−1 y
xetan−1 y
= ·e dy (312)
1 + y2

Let tan−1 y = tthen 1

= dtand dy = 1 + y 2 dt


dy
1+yˆ2 ˆ
t t 2
tet dt (313)


= · e · 1 + y dt =
1 + y2

u = t, du = 1and dv = et , v = et
ˆ ˆ
(314)
−1 −1
te dt = te − et dt = tet − et + c = tan−1 yetan y − etan y + c
t t

(315)
−1 −1
xetan y
= tan−1 y − 1 etan y + c

(316)
−1
x = tan−1 y − 1 + cetan y

c) xlnx dx
dy
+ y = 2lnx
dy y 2
+ = (317)
dx xlnx x
´ ´
Let p = 1
xlnx then pdx = 1
xlnx dx ,
Let t = lnxthen x1 dx = dt ⇒ dx = xdt
´ ´ 1
´ 1
pdx = xt · xdt = t dt = lnt = ln (lnx)

49
 ´
pdx
I.F. = e = eln(lnx) = lnx

Multiplying all through by the integrating factor

dy y 2
lnx · + = · lnx (318)
dx x x
d 2
[ylnx] = · lnx (319)
dx x
ˆ
2
ylnx = · lnx (320)
x
ˆ
1 t2
ylnx = 2
2
· txdt = 2 · + c = (lnx) + c (321)
x 2
c
y = lnx + (322)
lnx

d) + ycotx = 5ecosx . If y = 4when x =

dy
dx
π
2
´ ´
pdx = cotxdx = lnsinx
´
pdx
I.F. = e = elnsinx = sinx

Multiplying all through by the integrating factor

dy
sinx + ycotx · sinx = 5ecosx · sinx (323)
dx
dy
sinx + ycosx = 5sinxecosx (324)
dx
d
[ysinx] = 5sinxecosx (325)
dx
ˆ
ysinx = 5sinxecosx dx (326)

Let cosx = t ⇒ −sinxdx = dtand dx = −dt

ˆ sinx
dt
=5 sinxet · (327)
−sinx
ˆ
= −5 et dt (328)

= −5ecosx + c (329)

ysinx = −5ecosx + c (330)

Using the initial conditions

4 = −5 + c ⇒ c = 9 (331)

50
 ysinx = −5ecosx + 9 (332)

3.2 Bernoulli equation

The equation dy
dx + py = Qy n is known as bernoulli equation. To solve this

equation we divide both sides by y n to get

dy
y −n + py 1−n = Q
dx

We then put y 1−n = z , so that

dy dz
(1 − n) y −n =
dx dx

Substituting these in the original equation

1 dz
+ pz = Q
1 − n dx

Thus
dz
+ p (1 − n) z = (1 − n) Q
dx

To solve this we require a suitable integrating factor and multiply it by the

equation, we then integrate and get the solution in terms of x and z. On
substituting the values of z we obtain the required solution.
Exercise
Solve the following bernoulli equations
a) x dx
dy
+ y = x3 y 6

Solution
dy y
+ = x2 y 6 (333)
dx x
dy y −5
y −6 + = x2 (334)
dx x

Let y −5 = z , −5y −6 dx
dy
= dz
dx and dy
= − 51 y 6 dx
dx
dz

1 dz z
− + = x2 (335)
5 dx x
dz 5z
− = −5x2 (336)
dx x

51
 ´ 5
−x
I.F. = e dx
= e−5lnx = x−5

Multiplying all through by the integrating factor

dz 5z
x−5 · − = −5x−3 (337)
dx x6
d
zx−5 = −5x−3 (338)


dx
ˆ
zx−5 = −5x−3 dx (339)

5 −2
zx−5 = x +c (340)
2

But y −5 = z
5 −2
y −5 x−5 = x +c (341)
2
5 3 5
1= x y + cx5 y 6 (342)
2

An equation of the form f 0 (y) dx

dy
+pf (y) = Qcan be reduced to leibnitz equation

by putting f (y) = z .
a) Solve dy
dx + xsin2y = x3 cos2 y

Solution
1 dy 2xsinycosy
· + = x2 (343)
cos2 y dx cosy
dy
sec2 y + 2xtany = x3 (344)
dx

Let z = tany
dy dz
sec2 y dx = dx
dy 1 dz
dx = sec2 y dy

Thus,
dz
+ 2xz = x3 (345)
dx
´ 2
2xdx
I.F. = e = ex

Multiplying all through by the integrating factor

dz
(346)
2 2 2
ex + 2xex z = x3 ex
dx

52
 d h x2 i
(347)
2
ze = x3 ex
dx
ˆ
(348)
2 2
zex = x3 ex dx

Let x2 = t ⇒ 2xdx = dtand dx = dt

ˆ ˆ 2x ˆ ˆ
dt 1 1
te x2
= 3 t
x e · = 2 t
x e = tet dt (349)
2x 2 2

u = t, du = 1and dv = et , v = et
ˆ  ˆ 
1 1 1
tet dt = tet − et dt = (t − 1) et + c (350)
2 2 2

1 2
(351)
2
2
zex = x − 1 ex + c
2

But z = tany
1 2
(352)
2
2
tanyex = x − 1 ex + c
2
1 2
(353)
2
x − 1 + ce−x


tany =
2

b) Solve dy
dx + x (x + y) = x3 (x + y) − 1
3

Solution
Let x + y = u
dy du
1+ dx = dx
dy du
dx = dx −1

Replacing,
du
− 1 + xu = x3 u3 − 1 (354)
dx
du
+ xu = x3 u3 (355)
dx
du
u−3 + xu−2 = x3 (356)
dx

u−2 = z

−2u−3 du
dx =
dz
dx and du
dx = − 12 u3 dx
dz

1 dz
− + xz = x3 (357)
2 dx
dz
− 2xz = −2x3 (358)
dx

53
 ´ 2
−2xdx
I.F. = e = e−x

Multiplying all through by the integrating factor

dz
(359)
2 2 2
e−x − 2xe−x z = −2x3 e−x
dx
d h −x2 i
(360)
2
ze = −2x3 e−x
dx
ˆ
(361)
2 2
ze−x = −2x3 e−x

Let x2 = tthen dx = dt
ˆ 2x ˆ ˆ  ˆ 
−x2 3 −t dt 2 −t
te dt = − −te + e dt = − −te−t − e−t +c
−t −t −t
 
ze = −2x e =− x e dt = −
2x
(362)
= (t + 1) e −t
+c (363)

(364)
2 2
ze−x = x2 + 1 e−x + c

(365)
2 2
u−2 e−x = x2 + 1 e−x + c

(366)
2
u−2 = x2 + 1 + cex

(367)
−2 2
= x2 + 1 + cex


(x + y)
 
c) Solve ey dy
dx + 1 = ex
dy
ey + ey = ex (368)
dx

Let ey = z
dy dz
ey dx = dx
dy
dx = e−y dx
dz

dz
+ z = ex (369)
dx
´
dx
I.F. = e = ex

Multiplying all through by the integrating factor

dz
ex + zex = e2x (370)
dx
d
(zex ) = e2x (371)
dx

54
 ˆ
zex = e2x dx (372)

1 2x
zex = e +c (373)
2

But ey = z
1 2x
ey ex = e +c (374)
2
1 2x
e(y+x) = e +c (375)
2

55
4 EXACT DIFFERENTIAL EQUATIONS

A dierential equation of the form M (x, y)dx + N (x, y)dy = 0is said to be exact
if the left hand side is an exact dierentil of some function, say M (x, y). That
is M (x, y)dx + N (x, y)dy = du(x, y) = 0.
Theorem: The necessary and sucient condition for an equation M (x, y)dx +
N (x, y)dy = 0to be exact is δM
δy = δN
δx .
Proof: The Necessary Part,
Given that M (x, y)dx + N (x, y)dy = 0is an eact dierential equation then
M (x, y)dx + N (x, y)dy = du(x, y) = 0.......................................(i) Therefore,

du = δu
δx du + δu
δy dy ..................................................................(ii)
From (i) and (ii) then M = δu
δx ,N = δu δM
δy δy , = ,
δ 2 u δN
δyδx δx = δ2 u
δxδy

But, from young's theorem δ2 u

δyδx = δ2 u
δxδy then δM
δy = δN
δx

Proof: The Sucient Part,

Given that M (x, y)dx + N (x, y)dy = du(x, y) = 0.is an exact dierential equa-
tion and δM
δy = δN
δx , then
δu
M= δx
δM
δy = δ2 u
δyδx .................................(i)
..................................(ii)
2
δN δ u
δx = δxδy

Therefore,
δM
δy = δN
δx ,
Integrating (ii) with respect to x,
δu
N= δy + f (y)

Therefore the equationM (x, y)dx + N (x, y)dy = 0becomes δu δu

δx du + δy dy = 0
h i
δu δu
δx du + + f (y) dy = 0
δy
h i
δu δu
δx du + δy dy + f (y)dy = 0
´

d u + f (y)dy = 0

Thus the equation M (x, y)dx + N (x, y)dy = 0is an exact dierential equation.
On integration then
´
u+ f (y)dy = c

56
Exercise
1. Solve dy
dx + ycosx+siny+y
sinx+xcosy+x =0

Solution
(ycosx + siny + y) dx = − (sinx + xcosy + x) dy (376)

(ycosx + siny + y) dx + (sinx + xcosy + x) dy = 0 (377)

M = ycosx + siny + y

N = sinx + xcosy + x
δM
δy = cosx + cosy + 1
δN
δx = cosx + cosy + 1
δM δN
= (378)
δy δx

and therefore the equation is exact.

δu
M= δx = ycosx + siny + y
δu
N= δy = sinx + xcosy + x

Thus,
ˆ
u= (ycosx + siny + y) dx = c (379)

ysinx + xsiny + xy = c + f (y) (380)

and
ˆ
u= (sinx + xcosy + x) dy = c (381)

ysinx + xsiny + xy = c + f (x) (382)

In comparing these then;

ysinx + xsiny + xy = c (383)

2. Solve x2 + y dx + y 3 + x dy = 0

Solution
M = x2 + y and N = y 3 + x
δM δN
δy = δx =1

Therefore the equation is exact

57
 ˆ
y 3 + x dy (384)


u=

y4
+ xy + f (x) = c (385)
4

and
ˆ
x2 + y dx (386)


u=

y4
+ xy + f (y) = c (387)
4

In comparing these then;

y4
+ xy = c (388)
4

3. Solve y 1 + x1 + cosy dx + [x + lnx − xsiny] dy = 0




Solution
1
+ cosy and N = x + lnx − xsiny


M =y 1+ x
δM δN 1
δy = δx =1+ x − siny

Therefore the equation is exact

ˆ    
1
u= y 1+ + cosy dy (389)
x

xy + ylnx + xcosy + f (x) = c (390)

and
ˆ
u= (x + lnx − xsiny) dx (391)

xy + ylnx + xcosy + f (y) = c (392)

In comparing these then;

xy + ylnx + xcosy = c (393)

58
4.1 Equations Reducible to Exact Dierential Equations
Group of terms Integrating factor Exact dierential
1
d xy


xdy − ydx x2
 
xdy − ydx 1
y2 d −xy
1
d ln xy


xdy − ydx xy
1
d tan−1 xy


xdy − ydx x2 +y 2
 
1 x+y
xdy − ydx x2 −y 2 d 12 ln x−y
4. Solve xdy − ydx = y lnxdx
2

Solution
Multiplying both sides by 1
y2 ,

 
x 1 xdy − ydx
dy − dx = lnxdx = lnxdx (394)
y2 y y2

 
−x
d = lnxdx (395)
y

 
−x
d − lnxdx = 0 (396)
y

ˆ   ˆ
−x
d = lnxdx (397)
y

ˆ
−x 1
= xlnx − x· dx (398)
y x

−x
= xlnx − x + c (399)
y
 p 
5. Solve xdy − ydx = x x2 + y 2 dx
Solution

 p 
xdy − ydx = x x2 + y 2 dx (400)

59
 r  y 2
1 y
dy − 2 dx = 1+ (401)
x x x

y r  y 2
d = 1+ (402)
x x

ˆ y

ˆ
d
q x

= dx (403)
y 2
1+ x
´ √
√ 1
 
a2 +x2
dx = ln x + a2 + x2

" r #
y  y 2
ln + 1+ =x+c (404)
x x

6. Solve (1 + xy) ydx + (1 − xy) xdy = 0

Solution

(ydx + xdy) + xydx − x2 ydy = 0 (405)

Dividing all through by x2 y 2

ydx + xdy xy 2 x2 y
2 2
+ 2 2 dx − 2 2 dy = 0 (406)
x y x y x y

 
d(xy) 1 1
+ dx − dy =0 (407)
x2 y 2 x y

Integrating,

ˆ ˆ  
d(xy) 1 1
+ dx − dy =0 (408)
x2 y 2 x y

1
− + lnx − lny = c (409)
xy

 
1 x
− + ln =c (410)
xy y

60
7. Solve x4 ex − 2mxy 2 dx + 2mx2 ydy = 0

Solution

x4 ex dx − 2m xy 2 dx − x2 ydy = 0 (411)

Dividing by x4

xy 2 dx − x2 ydy
 
x
e dx − 2m =0 (412)
x4

(ydx − xdy)
ex dx − 2my =0 (413)
x3

 y  (ydx − xdy)
ex dx − 2m =0 (414)
x x2

y y
ex dx − 2m d =0 (415)
x x

Integrating,

 y 2
ex dx − m =c (416)
x

61
5 APPLICATION OF FIRST ORDER DIFFER-

ENTIAL EQUATIONS

5.1 In Mechanics
1. A moving body is opposed by a force per unit mass of value Cx and the
resistance per unit mass of radius bv 2 whose x and v are the displacement nand
velocity of the particle rspectively at that instant. Determine the velocity of
the particle in terms of x if it starts from rest.
Solution
Let m be the mass of the particle , then the equation of motion from Newton's
second law of motion is:
F = −mcx − mbv 2

ma = −mcx − mbv 2

dv
m = −mcx − mbv 2 (417)
dt

dv dx
m · = −mcx − mbv 2 (418)
dx dt

dv
= − cx + bv 2 (419)


v
dx

dv
v + bv 2 = −cx (420)
dx

Put v 2 = z

dv dz
2v = (421)
dx dx

dv 1 dz
= (422)
dx 2v dx

62
 The above equation becomes:

1 dz
+ bz = −cx (423)
2 dx

dz
+ 2bz = −2cx (424)
dx
´
2bdx
I.F = e = e2bx

Multiplying all through by the integrating factor

dz
e2bx + 2bze2bx = −2cxe2bx (425)
dx

d  2bx 
ze = −2cxe2bx (426)
dx

Integrating both sides, then

ˆ
ze2bx = −2c xe2bx dx (427)

 ˆ 
x 2bx 1 2bx
ze 2bx
= −2c e − e dx (428)
2b 2b

" #
x 2bx 1 2bx
ze2bx
= −2c e − 2e +k (429)
2b (2b)

 
−c 1
z= x− + ke−2bx (430)
b 2b

 
−c 1
v = 2
x− + ke−2bx (431)
b 2b

Whenv = 0,x = 0

 
−c 1
0= − +k (432)
b 2b

63
 −c
k= (433)
2b2

Thus,
c
cx
v2 = 2
1 − e−2bx − (434)
2b b

2. A body of mass m falling from rest is subject to the force of gravity

and air resistance proportional to the square of the velocity i.e.kv 2 . If it falls
to a distance x and processes a velocity v at that instant, prove that: 2kx
m =

ln a2a−r2 where mg = ka2

2

Solution
If the velocity of the body is r after falling a distance r then the equation of
motion is

dv
mv = mg − kv 2 (435)
dx

dv
mv = ka2 − kv 2 (436)
dx

dv
= k a2 − v 2 (437)


mv
dx

ˆ ˆ
dv k
v 2 = dx (438)
a − v2 m

ˆ ˆ
1 −2vdv k
− 2 2
= dx (439)
2 a −v m

1  k
− ln a2 − r2  = x + c (440)

2 m

Whenv = 0,x = 0

1
− lna2 = c (441)
2

64
 1  k 1
− ln a2 − r2  = x − lna2 (442)

2 m 2

1  a2 
 
k
ln  2 2
= x (443)
2 a −r  m

a2  2kx
 
(444)

ln  =
a2 − r2  m

Chemistry
1. Uranium disintegrates at a rate proportional to the amount present at any
instant. If m1 and m2 are the masses present at times T1 and T2 respectively,
show that the half of uranium is T2 −T1
m
ln m1
ln2
2

Solution
Let m be the mass of uranium present at time t. Then the equation of
disentegration will be

dm
= −µm (445)
dt

where µis a constant of proportionality

ˆ ˆ
dm
= −µdt (446)
m

lnm = −µt + k (447)

where k is a constant of integration. Let the original mass of uranium be

m0 . Thus when t = 0,m = m0 . and lnm0 = k , therefore

m0
ln = µt (448)
m

Now at
t = T1 ,m = m1

65
 t = T2 ,m = m2

Thus from eqn(i)

m0
ln = µT1 (449)
m1

m0
ln = µT2 (450)
m2

m0 m0
µ (T2 − T1 ) = ln − ln (451)
m1 m2

   
m0 m2 m1
µ (T2 − T1 ) = ln · = ln (452)
m1 m0 m2

 
1 m1
µ= ln (453)
T2 − T1 m2

Thus eqn(i) beomes

T2 − T1  m0 
t=   ln (454)
ln m 1 m
m2

Let the mass be half the original mass = m0

after t = T , half time, then:


2

 
T2 − T1 m0
T =   ln m0 (455)
m1
ln m 2
2

T2 − T1
T =   ln2 (456)
m1
ln m 2

5.2 Mixtures
1. A tank initially contains 50 gallows of fresh water. Brine containing 2lb of
salt per gallow ows inrto the tank at a rate of 2lb/min and the mixture, kept
uniform by stirring, runs out at the same rate. How long will it take for the
quantity of salt in the tank to increase from 40lb to 80lb?

66
 Solution
Let the salt content at time t be ulb in the tank. Therefore the rate of change
of the content of salt in the tank is du
dt .The rate of increase of salt content in
the tank will be 2 × 2 = 4lb/min. If C is the concentration of brine in the tank
at time t then the rate of decrease of salt content in the tank is

u
2clb/min = 2 lb/min
50

since the concentration is c = u

50

Thus
du u 100 − u
=4− = (457)
dt 25 25

ˆ ˆ
du 1
= dt (458)
100 − u 25

t
−ln |100 − u| = +k (459)
25

Initally, u = 0,t = 0. Therefore k = −ln100

 
100 1
ln = t (460)
100 − u 25

100
t = 25ln (461)
100 − u

Let t = t1 ,u = 40
t = t2 ,u = 80

100 100 5
t1 = 25ln = 25ln = 25ln (462)
100 − 40 60 3

100 100
t2 = 25ln = 25ln = 25ln5 (463)
100 − 80 20

67
 t2 − t1 = 25ln3mins (464)

2. Under certain conditions sugar cane in wtaer is converted to dextrose

which is proportional to the amount that is unconverted at any time. If of 75g
at time t = 0, 8g are converted during the rst 30 minutes, nd the amount
that is converted in one and a half hours.
Solution
Let m0 be the original mass of sugar cane and m be the mass converted after
time t.

dm
= µ (75 − m) (465)
dt

ˆ ˆ
dm
= µdt (466)
75 − m

−ln (75 − m) = µt + k (467)

When, m = 0,t = 0. Therefore k = −ln75

75
ln = µt (468)
75 − m

When t = 30min,m = 8g

75
ln = 30µ (469)
67

1 75
µ= ln (470)
30 67

75 t 75
ln = ln (471)
75 − m 30 67

68
 30 75
t= 75 ln 75 − m (472)
ln 67

When t = 1 21 hrs = 90min

75 75
ln = 3ln (473)
75 − m 67

 −3
75 75
(474)
75
= e−3ln 67 = = 0.7129
75 − m 67

(475)
75
75 − m = 75e−3ln 67 = 53.469

m = 75 − 53.469 = 21.531 (476)

69
6 SECOND ORDER DIFFERENTIAL EQUA-

TIONS

6.1 Introduction
The general form of the the nth order linear dierential equation is given by

dn y dn−1 y dn−2 y
+ P 1 + P 2 + ... + Pn y = Q(x)
dxn dxn−1 dxn−2

where P1 , P2 , ...Pn are functions of x. A linear dierential equation with

constant coecients will have the form

dn y dn−1 y dn−2 y
n
+ k1 n−1 + k2 n−2 + ... + kn y = 0
dx dx dx

where k1 , k2 , ...kn are constants.

Theorem: If y1 and y2 are the onlty solutions of the above equation then
c1 y1 + c2 y2 is also a solution where c1 and c2 are arbitrary constants.

Proof: Since y = y1 is a solution of equation(i) above then it satises it i.e.

dn y1 dn−1 y1 dn−2 y1
+ k 1 + k2 + ... + ky1 = 0
dxn dxn−1 dxn−2

and for y = y2

dn y2 dn−1 y2 dn−2 y2
+ k 1 + k2 + ... + ky2 = 0
dxn dxn−1 dxn−2

Multiplying eqn(2) by c1 and eqn(3) by c2

dn dn−1 dn−2
(c1 y1 + c 2 y2 )+k 1 (c1 y1 + c2 y2 )+k2 (c1 y1 + c2 y2 )+...+k (c1 y1 + c2 y2 ) = 0
dxn dxn−1 dxn−2

which shows that if y1 and y2 are the onlty solutions of the above equation
then c1 y1 + c2 y2 is also a solution where c1 and c2 are arbitrary constants.

70
6.2 The D-Operator
The following D-Operator method will be adopted as indicated:
= D, dx 2 = D , dx3 = D , dx4 = D ,..., dxn = D
2 3 4 n
d d 2 d 3 d 4 d n
dx

In this case the linear dierential equation with constant coecients can be
written as:

Dn + k1 Dn−1 + k2 Dn−2 + ... + kn y = 0

f (D)y = 0

f (D)y = 0is also called the auxillary form of the dierential equation f (D) =

Case 1:
If all the roots of f (D) = 0are real and dierent then f (D)y = 0can be
written as

[(D − m1 )(D − m2 )(D − m3 )......(D − mn )] y = 0

Let (D − mn ) y = 0, from the nth root

 
d
− mn y = 0
dx

dy
− mn y = 0
dx

Which is a leibnitz linear equation

´
I.F. = e− mn dx
= e−mn x

Multiplying all through by the integrating factor

dy
e−mn x − mn e−mn x y = 0
dx

71
 d  −mn x 
ye =0
dx

ye−mn x = cn

y = cn emn x

From above we see that

y = c1 e m 1 x

y = c2 e m 2 x

y = c3 e m 3 x

e.t.c.
Therefore the complte solution will be the source of these independent solu-
tions i.e.

y = c1 em1 x + c2 em2 x + c3 em3 x + ..... + cn emn x

Case 2:
If all the roots are real but two of them are equal and the rest dierent, say,
m1 = m2 , m3 , m4 , ..., mn dierent then

[(D − m1 )(D − m1 )(D − m3 )(D − m4 )......(D − mn )] y = 0

so that the solution will be

y = (c1 + c2 x) em1 x + c3 em3 x + c4 em4 x + ..... + cn emn x

72
 Therefore there will be (n − 1)solutions and we say that the solution is not
complete. An equation of order n have n independent solutions. Considering
the rst two equal roots in this case,

[(D − m1 )(D − m1 )] y = 0

and

(D − m1 )y = z

then

(D − m1 )z = 0

 
d
− m1 z = 0
dx

dz
− m1 z = 0
dx
´
I.F. = e− m1 dx
= e−m1 x

Multiplying all through by the integrating factor

dz
e−m1 x − m1 e−m1 x z = 0
dx

d  −m1 x 
ze =0
dx

Integrating

ze−m1 x = k1

z = k1 em1 x

73
Now
(D − m1 )y = z

(D − m1 )y = k1 em1 x

dy
− m1 y = k1 em1 x
dx
´
I.F. = e− m1 dx
= e−m1 x

Multiplying all through by the integrating factor

dy
e−m1 x − m1 e−m1 x y = k1 em1 x · e−m1 x
dx

dy
e−m1 x − m1 e−m1 x y = k1
dx

d  −m1 x 
ye = k1
dx

Integrating,

ˆ
ye−m1 x = k1 dx

ye−m1 x = k1 x + k2

y = (k1 x + k2 ) em1 x

Thus the complete solution will be

y = (c1 + c2 x) em1 x + c3 em3 x + c4 em4 x + ..... + cn emn x

If m1 = m2 = m3 , m4 , ..., mn dierent then the complete solution will be

74
 y = c1 + c2 x + c3 x2 em1 x + c4 em4 x + ..... + cn emn x

Case 3:
If one pair of the roots is comples and the other rfeal and dierent saym1 =
α + iβ, m2 = α − iβ, m3 , m4 , ..., mn are the roots then the solution is given by

y = c1 e(α+iβ)x + c2 e(α−iβ)x + c3 em3 x + ..... + cn emn x

Considering the rst two roots,

c1 e(α+iβ)x + c2 e(α−iβ)x = eαx c1 eiβx + c2 e−iβx

 

= eαx [c1 (cosβx + isinβx) + c2 (cosβx − isinβx)]

= eαx [(c1 + c2 ) cosβx + i (c1 − c2 ) sinβx]

Let k1 = c1 + c2 and k2 = i (c1 − c2 )

Therefore the complete solution is given by

y = eαx (c1 cosβx + c2 sinβx) + c3 em3 x + c4 em4 x + ..... + cn emn x

Case 4: If two of the complex roots are equal and the others real and dierent
i.e. m1 = m2 = α + iβ, m3 = m4 = α − iβ, m5 , m6 , ..., mn then the complete
solution will be given as:

y = eαx ((c1 + c2 x) cosβx + (c3 + c4 x) sinβx) + c5 em3 x + c6 em4 x + ..... + cn emn x

Exercise

75
 1. Solve d2 x
dt2 + 6 dx
dt + 9x = 0

Solution

D2 − 6D + 9 x = 0 (477)

2
(D + 3) x = 0 (478)

D = −3, −3

Hence the roots are real but equal. Thus the complete solution will be given
by

x = (c1 t + c2 ) e−3t

2. Solve D3 + D2 + 4D + 4 x = 0

Solution
By inspection, D = −1satises the equation. Therefore (D + 1)is a factor

D2 (D + 1) + 4 (D + 1) = 0 (479)

D2 + 4 (D + 1) = 0 (480)

Thus D = −1, 2i, −2i

Therefore the complete solution is

y = c1 e−x + c2 cos2x + c3 sin2x (481)

3. Solve d4 y
dx4 + 4y = 0

Solution

D4 + 4 y = 0 (482)

76
 D4 + 4D2 + 4 − 4D2 = 0 (483)

(484)

2 2
D2 + 2 − (2D) = 0

D2 − 2D + 2 D2 + 2D + 2 = 0 (485)

When D2 − 2D + 2
√ √
2± 4−8 2± −4 2±2I
D= 2 = 2 = 2 =1±i

When D2 + 2D + 2
√ √
−2± 4−8 −2± −4 −2±2I
D= 2 = 2 = 2 = −1 ± i

The complete solution is therefore,

y = ex (c1 cosx + c2 sinx) + e−x (c3 cosx + c4 sinx) (486)

4. Solve d3 y 2
d y dy
dx3 − 3 dx 2 + 3 dx − y = 0

Solution

D3 − 3D2 + 3D + 1 y = 0 (487)

D3 − 3D2 + 3D + 1 = 0 (488)

(D − 1) = 0
3
(489)

D = 1, 1, 1

Therefore the complete solution is given by

y = c1 x2 + c2 x + c3 ex (490)

6. Solve d4 x
dt4 + m4 x = 0

77
Solution

D4 + m4 x = 0 (491)

D4 + 2D2 m2 + m4 − 2D2 m2 = 0 (492)

√ 2
(493)

2
D2 + m2 − 2Dm = 0

7. Solve d4 y 2
d y
dx4 + 8 dx 2 + 16y = 0

Solution

D4 + 8D2 + 16 y = 0 (494)

D4 + 8D2 + 16y = 0 (495)

(496)

2
D2 + 4 =0

D = 2i, 2i, −2i, −2i

The complete solution is therefore

y = (c1 + c2 x) cos2x + (c3 + c4 x) sin2x (497)

Consider the following dierential equation

dn y dn−1 y dn−2 y
n
+ k1 n−1 + k2 n−2 + ... + kn y = Q(x)
dx dx dx

 n
D + k1 Dn−1 + k2 Dn−2 + ..... + kn y = Q(x)


This kind of equation will have two types of solutions; We solve the L.H.S=0

78
to get the complimetary function(C.F) of the equation. Then we nd the par-
ticular integral as a result of Q(x) given by

1
P.I. = Q(x)
Dn + k1 Dn−1 + k2 Dn−2 + ..... + kn

Thus the complete solution will be

y = C.F + P.I

6.2.1 The Inverse of D-Operator f (D)

1

1
f (D) x is that function of x containing arbitrary constants which when operated
on by f (D)gives x.i.e.

 
1
f (D) x =x
f (D)

Thus f (D)and 1
f (D) are inverse operators of each other. Equation(2) can be
written as f (D)y = Q(x). Operating both sides by 1
f (D) then

1 1
f (D)y = Q(x)
f (D) f (D)

1
y= Q(x)
f (D)

which contains the particular integral and contains no arbitrary constants.

Exercise:
1. Evaluate 1
D Q(x) where Q(x)is some function of x and D = d
dx

Solution
Let y = 1
D Q(x)

Operating D on both sides

 
1
Dy = D Q(x)
D

79
 Dy = Q(x)

dy
= Q(x)
dx

Integrating both sides with respect to x

ˆ
y= Q(x)dx

2. Evaluate 1
D−a Q(x)

Solution
Let y = 1
D−a Q(x)

Operating D − a on both sides

1
(D − a) y = D − a Q(x)
D−a

(D − a) y = Q(x)

dy
− ay = Q(x)
dx
´
−adx
I.F. = e = e−ax

Multiplying all through by the integrating factor

dy
e−ax − ae−ax y = e−ax Q(x)
dx

d  −ax 
ye = Q(x)e−ax
dx

Integrating,

ˆ
ye−ax = Q(x)e−ax

80
 ˆ
y = eax Q(x)e−ax dx

Rules of Finding Particular Integral

Case 1:When Q(x) = eax then

1 ax 1
P.I. = e = n eax
f (D) D + k1 Dn−1 + ... + kn

Deax = aeax

D2 eax = a2 eax

Dn eax = an eax

Dn + k1 Dn−1 + ... + kn eax = an + k1 an−1 + ... + kn eax = f (a)eax

\
f (D)eax = f (a)eax

1 1
[f (D)eax ] = [f (a)eax ]
f (D) f (D)

1 ax
eax = f (a) e
f (D)

1 ax 1 ax
e = e
f (D) f (a)

provided that f (a) 6= 0

If f (a) = 0, then D − a is a factor of the equation f (D) = 0. Thusf (D) =

81
(D − a) φ(D).

Let φ(a) 6= 0then

1 ax 1
e = eax
f (D) (D − a) φ(D)

1
= eax
(D − a) φ(a)

1 1
= eax
φ(a) D − a

ˆ
1
Q(x) = eax Q(x)e−ax dx
D−a

ˆ
1
eax = eax eax e−ax dx
D−a

ˆ
ax
=e dx

= xeax

Therefore, when f (a) = 0with f (D) = (D − a) φ(D) then P.I. = 1

f (D) e
ax
=
x ax
φ(a) e

Now f (D) = (D − a) φ(D)

Dierentiating both sides with respect to D then,

f 0 (D) = φ(D) + (D − a) φ0 (D)

When D = a

f 0 (a) = φ(a) + 0.φ0 (a)

82
 f 0 (a) = φ(a)

Thus

1 ax 1 ax
e = e if f (a) 6= 0
f (D) f (a)

1 ax x
e = 0 eax if f 0 (a) 6= 0
f (D) f (a)

1 ax x2
e = 00 eax if f 00 (a) 6= 0
f (D) f (a)

Exercise:
1. Determine the particular integral of D2 + 5D + 6 y = ex

Solution

1
yp = ex (498)
D2 + 5D + 6

Put D = 1

1 1 x
yp = ex = e (499)
12 + 5(1) + 6 12

2. Determine the particular integral of (D + 2)(D − 1)2 y = e−2x + 2sinhx

Solution

1
(500)
 −2x
+ ex − e−x

yp = 2
e
(D + 2)(D − 1)

" #
1 1 1
e −2x
= e−2x
(501)
(D + 2)(D − 1)2 D + 2 (−2 − 1)2

 
1 1 −2x
= e (502)
D+2 9

83
 1 x −2x
= · e (503)
9 1

x −2x
= e (504)
9

" #
1 1 1
e−2x = 2 2e
x
(505)
(D + 2)(D − 1)2 (D − 1) (D + 2)

 
1 1 x
= 2 e (506)
(D − 1) 3

1 1
= · · ex (507)
3 (D − 1)2

x2
 
1
= ex (508)
3 f 00 (D)

1 2 x
= x e (509)
6

1 1 −e−x
e −x
= e −x
= (510)
(D + 2)(D − 1)2 (−1 + 2)(−1 − 1)2 4

x −2x 1 2 x e−x
yp = e + x e − (511)
9 6 4

Case 2:
When Q(x) = sin(ax + b)or Q(x) = cos(ax + b)
In this case:
Dsin(ax + b) = asin(ax + b)

D2 sin(ax + b) = −a2 sin(ax + b)

D3 sin(ax + b) = −a3 sin(ax + b)

D4 sin(ax + b) = a4 sin(ax + b)

84

2
2
D2 sin(ax + b) = −a2
sin(ax + b)
2 3 3



D sin(ax + b) = −a2 sin(ax + b)

r
r
D2 sin(ax + b) = −a2 sin(ax + b)

operating, f (D1 2 ) on both sides then

1 2 1
f (D 2 ) f (D )sin(ax + b) = f (−a2 ) f (D 2 ) sin(ax + b)

1
f (−a2 ) f (D 2 ) sin(ax + b) = sin(ax + b)

(a) f (D1 2 ) sin(ax + b) = 1

f (−a2 ) sin(ax + b) f (−a2 ) 6= 0
1
f (D 2 ) sin(ax + b) = x
f 0 (−a2 ) sin(ax + b) f 0 (−a2 ) 6= 0

(b) f (D1 2 ) cos(ax + b) = 1

f (−a2 ) cos(ax + b) f (−a2 ) 6= 0
1
f (D 2 ) cos(ax + b) = x
f 0 (−a2 ) cos(ax + b) f 0 (−a2 ) 6= 0

Exercise
1. Determine the particular integral of D3 + 1 y = cos (2x − 1)

Solution

1
yp = cos (2x − 1) (512)
D3 + 1

1
yp = cos (2x − 1) (513)
D(D2 ) + 1

put D2 = −a2 = −22 = −4

1
yp = cos (2x − 1) (514)
1 − 4D

1 + 4D
= cos (2x − 1) (515)
(1 − 4D) (1 + 4D)

1 + 4D
= cos (2x − 1) (516)
1 − 16D2

1 + 4D
= cos (2x − 1) (517)
1 − 16(−4)

85
 1
= [cos (2x − 1) + 4Dcos (2x − 1)] (518)
65

1
= [cos (2x − 1) − 8sin (2x − 1)] (519)
65

2. Determine the particular integral of d3 y

dx3
dy
+ 4 dx = sin2x

Solution

D3 + 4D y = sin2x (520)

1
yp = sin2x (521)
D3 + 4D

put D2 = −a2 = −22 = −4

f −22 = 0

1
yp = sin2x (522)
−4D + 4D

Now f (D) = D3 + 4D
f 0 (D) = 3D2 + 4

f 0 (−a2 ) = 3(−4) + 4 = −8 6= 0

−x
yp = sin2x (523)
8

Case 3:
When Q(x) = eax V , where V is a function of x then we have
D (eax V ) = eax DV + aeax V

D (eax V ) = eax (D + a) V

D2 (eax V ) = eax D2 V + aeax DV + aeax DV + a2 eax V

D2 (eax V ) = eax D2 V + aeax DV + aeax DV + a2 eax V

2
D2 (eax V ) = eax (D + a) V
n
Dn (eax V ) = eax (D + a) V

86
Thus 1
f (D)
1
(eax V ) = eax f (D+a) V

Exercise
1. D2 − 2D + 4 y = ex cosx

Solution

1
yp = ex cosx (524)
D2 − 2D + 4

1
yp = ex 2 cosx (525)
(D + 1) − 2 (D + 1) + 4

1
yp = ex · cosx (526)
D2 + 2D + 1 − 2D − 2 + 4

1
yp = ex · cosx (527)
D2 + 3

f (D) = D2 + 3

f (D) = −12 + 3 = 2 6= 0

1
yp = ex · cosx (528)
2

1 x
yp = e cosx (529)
2

2, Solve (D − 2)2 y = 8 e2x + sin2x

Solution

(D − 2) = 0
2
(530)

D = 2, 2

yc = (c1 x + c2 ) e2x (531)

87
 1
(532)
 2x 
yp = 2 8e + 8sin2x
(D − 2)

1 1
yp = 2 8e
2x
+ 2 8sin2x (533)
(D − 2) (D − 2)

f (D) = D2 − 4D + 4for the rst partf (D) = 0

f 0 (D) = 2D − 4,f 0 (2) = 0

f 00 (D) = 2 6= 0

1 8x2 2x
2 8e
2x
= e = 4x2 e2x (534)
(D − 2) 2

For the second part,

1 8
2 8sin2x = sin2x (535)
(D − 2) D2 − 4D + 4

ˆ
8 −2
= sin2x = sin2x = −2 sin2xdx = cos2x (536)
−4 − 4D + 4 D

or

8 −2 −2D −2D 1
= sin2x = sin2x = sin2x = sin2x = ·2cos2x = cos2x
−4 − 4D + 4 D D2 4 2

yp = 4x2 e2x + cos2x (537)

3. Solve d2 y
dx2 + a2 y = secax

Solution

D2 + a2 y = secax (538)

88
 D2 + a2 = 0 (539)

D2 = −a2 (540)

D = ±ai

yc = c1 cosax + c2 sinax (541)

1
yp = secax (542)
D2 + a2

1
= secax (543)
(D + ai) (D − ai)

 
1 1 1
= − secax (544)
2ia D − ai D + ai

 
1 1 1
= secax − secax (545)
2ia D − ai D + ai
´
But 1
D−a Q(x) = eax Q(x)e−ax dx

ˆ
1
secax = eaix secaxe−iax dx (546)
D − ai

ˆ
cosax − isinax
= eiax dx (547)
cosax

ˆ
= eiax (1 − itanax) dx (548)

 
i
=e iax
x + ln |cosax| (549)
a

89
 ˆ
1
secax = e−aix secaxeiax dx (550)
D + ai

ˆ
cosax + isinax
=e −iax
dx (551)
cosax

ˆ
= e−iax (1 + −itanax) dx (552)

 
i
=e −iax
x + ln |cosax| (553)
a

    
1 i i
yp = eiax x + ln |cosax| − e−iax x − ln |cosax| (554)
2ia a a

x eiax − e−iax e + e−iax

   iax 
1
= + 2 ln |cosax| (555)
a 2i a 2i

x 1
= sinax + 2 cosaxln |cosax| (556)
a a

y = yc + yp (557)

x 1
y = c1 cosax + c2 sinax + sinax + 2 cosaxln |cosax| (558)
a a

4. Solve D2 + 4 y = ex sin2 x

Solution

D2 + 4 = 0 (559)

D = ±2i

90
 yc = c1 cos2x + c2 sin2x (560)

1
yp = ex sin2 x (561)
D2 +4

1
= ex 2 sin2 x (562)
(D + 1) + 4

1 1
= ex [1 − cos2x] (563)
D2 + 2D + 5 2

 
1 x 1 1
= e e 0x
− cos2x (564)
2 D2 + 2D + 5 D2 + 2D + 5

 
1 x 1 1
= e − cos2x (565)
2 5 −4 + 2D + 5

 
1 x 1 1
= e − cos2x (566)
2 5 2D + 1

 
1 x 1 2D − 1
= e − cos2x (567)
2 5 (2D + 1) (2D − 1)

 
1 x 1 2D − 1
= e − cos2x (568)
2 5 4D2 + 1

 
1 x 1 2D − 1
= e − cos2x (569)
2 5 −17

 
1 x 1 1
yp = e − (−4sin2x − cos2x) (570)
2 5 17

 
1 x 1 1
yp = e − (4sin2x + cos2x) (571)
2 5 17

91
 y = yc + yp (572)

 
1 x 1 1
y = c1 cos2x + c2 sin2x + e − (4sin2x + cos2x) (573)
2 5 17

5. Solve d2 y
dx2
dy
− 2 dx + 4y = ex sin x2

Solution

x
D2 − 2D + 4 y = ex sin (574)


2

D2 − 2D + 4 = 0 (575)

D2 − 2D + 1 = −3 (576)

√
(D − 1) = 1 ± 3i (577)

 √ √ 
yc = ex c1 cos 3x + c2 sin 3x (578)

1 x
yp = ex sin (579)
D2 − 2D + 4 2

1 x
= ex 2 sin (580)
(D + 1) − 2 (D + 1) + 4 2

1 x
= ex sin (581)
D2 + 2D − 1 − 2D − 2 + 4 2

1 x
= ex sin (582)
D2 + 3 2

92
 1 x
= ex sin (583)
− 14 + 3 2

4 x x
yp = e sin (584)
11 2

y = yc + yp (585)

 √ √  4 x
yp = ex c1 cos 3x + c2 sin 3x + ex sin (586)
11 2

Casa 4:
When Q(x) = X m
Particular integral will be yp = 1
f (D) x
m −1
= [f (D)] xm . We expand this

using the binomial theorem:

Exercise
1. Solve 2D2 + 2D + 3 y = x2 + 2x − 1

Solution

2D2 + 2D + 3 = 0 (587)
√ √
−2± 4−24 5
D= 4 = − 12 ± 2 i

√ √ !
5 5
yc = e − 12 x
c1 cos x + c2 sin x (588)
2 2

1
yp = · x2 + 2x − 1 (589)
2D2 + 2D + 3

1
= 2


· x2 + 2x − 1 (590)
3 1+ 3D + 23 D2

  −1
1 2 2 2
x2 + 2x − 1 (591)


= 1+ D+ D
3 3 3

93
  
1 2 2
1 − D − D2 − ... x2 + 2x − 1 (592)


=
3 3 9

 
1
2 2
= 2
x + 2x − 1 − (2x + 2) − · 2 (593)
3 3 9

 
1 2 2 25
= x + x− (594)
3 3 9

1 2 2 57
= x + x− (595)
3 9 27

y = yc + yp (596)

√ √ !
5 5 1 2 57
yp = e − 21 x
c1 cos x + c2 sin x + x2 + x − (597)
2 2 3 9 27

2. Solve D3 − 2D + 4 y = x4 + 3x2 − 5x + 2

Solution

D3 − 2D + 4 y = 0 (598)

(D + 2) D2 − 2D + 2 y = 0 (599)

D = −2, 1 ± i

yc = c1 e−2x + ex (c2 cosx + c3 sinx) (600)

1
x4 + 3x2 − 5x + 2 (601)


yp =
D3 − 2D + 4

94
 1

 x4 + 3x2 − 5x + 2 (602)


=  1 1 3
4 1− 2D − 4D

  −1
1 1 1 3
x4 + 3x2 − 5x + 2 (603)


= 1− D− D
4 2 4

"    2  3  4 #
1 1 1 3 1 1 3 1 1 3 1 1 3
+ ... x4 + 3x2 − 5x + 2


= 1+ D− D + D− D + D− D + D− D
4 2 4 2 4 2 4 2 4
(604)

 
1 1 1 2 1 3 3 4
1 + D + D − D − D + ... x4 + 3x2 − 5x + 2 (605)


=
4 2 4 8 16

 
1 4 2 1
3

1 2

1 3
= x + 3x − 5x + 2 + 4x + 6x − 5 + 12x + 6 − (24x) − (24) + ...
4 2 4 8 16
(606)

1 4 1 3 3 2 5 7
= x + x + x − x− (607)
4 2 2 4 8

y = yc + yp (608)

1 1 3 5 7
y = c1 e−2x + ex (c2 cosx + c3 sinx) + x4 + x3 + x2 − x − (609)
4 2 2 4 8

Exercise:
Determine the particular integrals of the following dierential equations
a) D2 + D + 1 y = e2x + 6ex − 3e−2x + 5Ans: yp = 1 3x
+ 2ex − e−2x + 5


13 e

b) D4 − 1 y = sin2xAns:yp = 1


15 sin2x

c) D3 + 1 y = cosxAns:yp = 12 (cosx − sinx)

95
d) D4 + 4 y = sin2xAns:yp = − 14 xcos2x

e) D2 − 1 y = x2 Ans:yp = −x2 − 2

f) D4 D2 − 1 y = x2 Ans:yp = − 360
1



x6 + 30x4

g) D2 + 2 y = x3 + x2 Ans: 1



2 x3 + x2 − 3x − 1

96
7 CAUCHY HOMOGENEOUS LINEAR EQUA-

TIONS

An equation of the form(i)

dn y dn−1 y dn−2 y dn−1 y

xn n
+ k1 xn−1 n−1 + k2 xn−2 n−2 + ... + kn−1 x n−1 + kn y = Q(x)
dx dx dx dx

where k's ae contants and Q(x) is some function of x is called a cauchy homo-
geneous Linear equation, Such an equation can be redced to a linear dierential
equation with constant coecients by puttingx = et ort = lnx.

dy dy dt
= ·
dx dt dx

1 dt
= ·
x dx

1
= Dy
x

where D = d
dt

d2 y
 
d 1
2
= Dy
dx dx x

 
1 dy 1 d dy
=− +
x2 dt x dt dt

 
1 dy 1 dt d dy
=− 2 + · ·
x dt x dx dt dt

1 dy 1 d2 y
=− 2
+ 2· 2
x dt x dt

97
 1 1 1
Dy + 2 D2 y = 2 D2 − D y


=− 2
x x x

1
== D (D − 1) y
x2

Thus,
dy
x = Dy
dx

d2 y
x2 = D (D − 1) y
dx2

d3 y
x3 = D (D − 1) (D − 2) y
dx3

By subtituting this back to the equation (i), we get a linear dierential

equation with constant coecients
Exercise:
1. Solve x2 dx 2 − x dx + y = lnx...........................(i)
2
d y dy

Solution
Let x = et ort = lnx.
dy
x dx = Dy
2
d y
x2 dx 2 = D (D − 1) y

3
d y
x3 dx 3 = D (D − 1) (D − 2) y

Replacing back to equation (i)

[D (D − 1) − D + 1] y = t (610)

D2 − D − D + 1 y = t (611)

D2 − 2D + 1 y = t (612)

98
 2
(D − 1) y = t (613)

(D − 1) = 0
2
(614)

D = 1, 1

yc = (c1 + c2 t) et (615)

1
yp = 2t (616)
(D − 1)

1
= 2t (617)
(1 − D)

= (1 − D)
−2
t (618)

= (1 + 2D + ....) t (619)

=t+2 (620)

y = yc + yp (621)

y = (c1 + c2 t) et + t + 2 (622)

But t = lnx.

y = (c1 + c2 lnx) x + lnx + 2 (623)

99
2. Solve x2 dx 2 − 4x dx + 6y = x ................................................(i)
2
d y dy 5

Let x = et ort = lnx.

dy
x dx = Dy
2
d y
x2 dx 2 = D (D − 1) y

3
d y
x3 dx 3 = D (D − 1) (D − 2) y

Replacing back to equation (i)

[D (D − 1) − 4D + 6] y = e5t (624)

D2 − D − 4D + 6 y = e5t (625)

D2 − 5D + 6 y = e5t (626)

(D − 3) (D − 2) y = e5t (627)

(D − 3) (D − 2) = 0 (628)

D = 3, 2

yc = c1 e3t + c2 e2t (629)

1 1 1
yp = e5t = e5t = e5t (630)
(D − 3) (D − 2) (5 − 3) (5 − 2) 6

y = yc + yp (631)

1
y = c1 e3t + c2 e2t + e5t (632)
6

100
 But t = lnxand x = et

1
y = c1 x3 + c2 x2 + x5 (633)
6

3. The radial displacement in a rotating disc is given by r2 ddru2 + r du

2

dr −

u + kr3 = 0where k is a constant. Solve the equation under the conditions

u = 0when r = 0and u = 0when r = a

Solution

d2 u du
r2 +r − u = −kr3 (634)
dr2 dr

Let t = lnrand r = et
r du
dr = Du
2
r2 ddru2 = D (D − 1) u

[D (D − 1) + D − 1] u = −ke3t (635)

(636)
 2
D − D + D − 1 u = −ke3t


D2 − 1 u = −ke3t (637)
 

(D − 1) (D + 1) u = −ke3t (638)

(D − 1) (D + 1) = 0 (639)

uc = c1 e−t + c2 et (640)

101
 1 1 k
up = − ke3t = − ke3t = − e3t (641)
(D − 1) (D + 1) (3 − 1) (3 + 1) 8

u = uc + up (642)

k 3t
= c1 e−t + c2 et − e (643)
8

c1 k
u= + c2 r − r3 (644)
r 8

k 4
ur = c1 + c2 r2 − r (645)
8

under the conditions u = 0when r = 0

c1 = 0

k 4
ur = c2 r2 − r (646)
8

under the conditions u = 0when r = a

k 4
0 = c2 a2 − a (647)
8

k 4
c2 = a
8

k 4 2 k 4
ur = a r − r (648)
8 8

k
r a2 − r2 (649)


ur =
8

102
4. Solve x2 D2 − 2xD + 2 y = x3 +sin (5lnx)...................................................(i)

Solution
Let x = et ort = lnx.
dy
x dx = Dy
2
d y
x2 dx 2 = D (D − 1) y

3
d y
x3 dx 3 = D (D − 1) (D − 2) y

Replacing back to equation (i)

[D (D − 1) − 2D + 2] y = e3t + sin5t (650)

D2 − D − 2D + 2 y = e3t + sin5t (651)

D2 − 3D + 2 y = e3t + sin5t (652)

(D − 1) (D − 2) y = e3t + sin5t (653)

(D − 1) (D − 2) = 0 (654)

yc = c1 et + c2 e2t (655)

1
(656)
 3t 
yp = e + sin5t
(D − 1) (D − 2)

1 1
yp = e3t + 2 sin5t (657)
(D − 1) (D − 2) D − 3D + 2

1 1
yp = e3t + sin5t (658)
(3 − 1) (3 − 2) −25 − 3D + 2

103
 1 3t 1
= e + sin5t (659)
2 −23 − 3D

1 3t 23 − 3D
= e − 2 sin5t (660)
2 23 − 9D2

1 3t 23 − 3D
= e − sin5t (661)
2 329 + 225

1 3t 1
= e − (23sin5t − 15cos5t) (662)
2 754

y = yc + yp (663)

1 1
y = c1 et + c2 e2t + e3t − (23sin5t − 15cos5t) (664)
2 754

But x = et ort = lnx.

1 1
y = c1 x + c2 x2 + x3 − (23sin (5lnx) − 15cos (5lnx)) (665)
2 754

104
8 LEGENDER LINEAR DIFFERENTIAL EQUA-

TION

An equation of the form

n dn y n−1 d
n−1
y n−2 d
n−2
y n−1 d
n−1
y
(ax + b) n
+k1 (ax + b) n−1
+k2 (ax + b) n−2
+...+kn−1 (ax + b) +kn y = Q(x)
dx dx dx dxn−1

where k's ae contants and Q(x) is some function of x is called a legender Lin-
ear dierential equation, Such an equation can be redced to a linear dierential
equation with constant coecients by putting(ax + b) = et ort = ln (ax + b).

dy dy dt
= ·
dx dt dx

a dt
= ·
(ax + b) dx

a
= Dy
(ax + b)

where D = d
dt

d2 y
 
d a
2
= Dy
dx dx (ax + b)

a2
 
dy a d dy
=− 2 +
(ax + b) dt (ax + b) dt dt

a2
 
dy a dt d dy
=− 2 dt + (ax + b) · dx · dt
(ax + b) dt

a2 dy a2 d2 y
=− + 2 2 · 2
(ax + b) dt (ax + b) dt

105
 d2 y
 2 
2 2 d y dy
(ax + b) = a −
dx2 dt2 dt

= a2 D2 − D y

Thus,
dy
(ax + b) = aDy
dx

2 d2 y
(ax + b) = a2 D (D − 1) y
dx2

3 d3 y
(ax + b) = a3 D (D − 1) (D − 2) y
dx3

Exercise
1. Solve (1 + x)2 dx 2 +(1 + x) dx +y = 2sin [ln (1 + x)]................................(i)
2
d y dy

Solution
Put(1 + x) = et ort = ln (1 + x)
then

dy
(1 + x) = Dy
dx

2 d2 y
(1 + x) = D (D − 1) y
dx2

equation (i) becomes:

[D (D − 1) + D + 1] y = 2sint (666)

D2 − D + D + 1 y = 2sint (667)

106
 D2 − 3D + 2 y = e3t + sin5t (668)

D2 + 1 y = 2sint (669)

nding the complimentary solution

D2 + 1 y = 0 (670)

D = ±i (671)

yc = c1 cost + c2 sint (672)

1 2t t D
yp = 2sint = sint = sint = t 2 sint = −tcost (673)
D2 +1 2D D D

y = yc + yp (674)

y = c1 cost + c2 sint − tcost (675)

But t = ln (1 + x)

y = c1 cost [ln (1 + x)] + c2 sin [ln (1 + x)] − [ln (1 + x)] cos [ln (1 + x)] (676)

2. Solve (1 + x)2 dx 2 + (1 + x) dx = (2x + 3) (2x + 4)................................(i)

2
d y dy

Solution
Put(1 + x) = et ort = ln (1 + x)

107
then

dy
(1 + x) = Dy
dx

2 d2 y
(1 + x) = D (D − 1) y
dx2

equation (i) becomes:

[D (D − 1) + D] y = 2et + 1 2et + 2 (677)

D2 y = 4e2t + 6et + 2 (678)

nding the complimentary solution

D2 y = 0 (679)

D = 0, 0 (680)

yc = c1 t + c2 (681)

nding the particular integral

1
4e2t + 6et + 2 (682)


yp =
D2

= 4e2t + 6et + t2 (683)

nding the general solution

y = yc + yp (684)

108
 y = c1 t + c2 + 4e2t + 6et + t2 (685)

But t = ln (1 + x)

y = c1 ln (1 + x) + c2 + 4e2[ln(1+x)] + 6e[ln(1+x)] + [ln (1 + x)]

2
(686)

2 2
y = c1 ln (1 + x) + c2 + (1 + x) + 6 (1 + x) + [ln (1 + x)]
2
(687)

3. Solve (x + 3)2 dx 2 +(x + 3) dx +6y = 2sin [ln (x + 3)]................................(i)

2
d y dy

Put(x + 3) = et ort = ln (x + 3)
then

dy
(x + 3) = Dy
dx

2 d2 y
(x + 3) = D (D − 1) y
dx2

equation (i) becomes:

[D (D − 1) − 4D + 6] y = t (688)

D2 − 5D + 6 y = t (689)

nding the complimentary solution

D2 − 5D + 6 y = 0 (690)

(D − 2) (D − 3) = 0 (691)

109
 D = 2, 3 (692)

yc = c1 e2t + c2 e3t (693)

nding the particular integral

1
yp = (t) (694)
D2 − 5D + 6

  −1
1 5 1 2
= 1− D− D t (695)
6 6 6

 
1 5 1
= 1 + D − D2 t (696)
6 6 6

t 5
= + (697)
6 36

OR

 
1 1
yp = − t
D−3 D−2

1 1
= t− t (698)
D−3 D−2

ˆ
1
Q(x) = eax Q(x)e−ax dx
D−a

ˆ
1
t = e3t te−3t dt (699)
D−3

 ˆ 
t −3t 1 −3t
3t
=e − e − − e dt (700)
3 3

110
  
t 1 t 1
= e3t − e−3t − e−3t = − − (701)
3 9 3 9

and

ˆ
1
t = e2t te−2t dt (702)
D−2

 ˆ 
t 1
= e2t − e−2t − − e−2t dt (703)
2 2

 
t −2t 1 −2t t 1
=e 2t
− e − e =− − (704)
2 4 2 4

 
t 1 t 1 t 5
yp = − − − − − = + (705)
3 9 2 4 6 36

nding the general solution

y = yc + yp (706)

t 5
y = c1 e2t + c2 e3t + + (707)
6 36

But t = ln (1 + x)

ln (x + 3) 5
y = c1 2ln (x + 3) + c2 3ln (x + 3) + + (708)
6 36

1 5
2 3
y = c1 (x + 3) + c2 (x + 3) + ln (x + 3) + (709)
6 36

111
9 SIMULTANEOUS DIFFERENTIAL EQUATIONS

If two or more independent variables are functions of a single independent vari-

able, the equations involving their derivatives are called simultaneous dierential
equations.e.g.

dx
+ 4y = t
dt

dy
+ 2x = et
dt

The method of solving these equations is based on the process of elimination,

the same way we solve algebraic simultaneous equations .
Example:
1. Solve the equations:
dx
dt = −ωy
dy
dt = ωx

Solution
On using the D-Operator method

Dx + ωy = 0...........................................................................(i)

−ωx + Dy = 0.....................................................................(ii)

On muliplyng (i) with ω and (ii) with D, we get

ωDx + ω 2 y = 0...........................................................................(iii)

112
 −ωDx + D2 y = 0.....................................................................(iv)

On adding (iii) and (iv)

ω 2 y + D2 y = 0...........................................................................(v)

or

ω 2 + D2 y = 0..................................................................(v)

Now we have to solve (v) to get the value for y

Auxillary equation is

ω2 + D2 = 0

D2 = −ω 2 or D = ±iω

y = Acosωt + Bsinωt....................................................................(vi)

Dy = −Aωsinωt + Bωcosωt....................................................................(vii)

On substituting (vii) in (ii), we get;

−ωx − Aωsinωt + Bωcosωt = 0

113
 ωx = −Aωsinωt + Bωcosωt

x = −Asinωt + Bcosωt

2. Solve the following

dx dy
dt + dt + 2x + y = 0
dy
dt + 5x + 3y = 0

Solution

Dx + Dy + 2x + y = 0...................................(i)

Dy + 5x + 3y = 0...................................(ii)

(D + 2) x + (D + 1) y = 0.............................(iii)

5x + (D + 3) y = 0.....................................(iv)

On multiplying (iii) by 5 and (iv) by D + 1, to eliminate y;

5 (D + 2) x + 5 (D + 1) y = 0.............................(v)

5 (D + 2) x + (D + 3) (D + 2) y = 0.....................................(vi)

Subtracting (v) from (vi)

[(D + 3) (D + 2) − 5 (D + 1)] y = 0.....................................(vi)

114
 D2 + 5D + 6 − 5D − 5 y = 0 or D2 + 1 y = 0

D = ±i

y = C1 cost + C2 sint

Dy = −C1 sint + C2 cost

Putting value of Dy and y in (ii)

5x − C1 sint + C2 cost + 3C1 cost + 3C2 sint

5x = C1 sint − C2 cost − 3C1 cost − 3C2 sint

1 1
x= (C1 − 3C2 ) sint − (3C1 + C2 ) cost
5 5

3. Solve
dx dy
dt − dt + 2y = cos2t
dx dy
dt + dt − 2x = sin2t

Solution

Dx − Dy + 2y = cos2t...................................(i)

Dx + Dy − 2x = sin2t...................................(ii)

Dx − (D − 2) y = cos2t.............................(iii)

115
 (D − 2) x + Dy = sin2t.....................................(iv)

On multiplying (iii) by D and (iv) by D − 2, to eliminate y;

D2 x − D (D − 2) y = −2sin2t.............................(v)

2
(D − 2) x + D (D − 2) y = 2cos2t − 2sin2t.....................................(vi)

Adding (v) and (vi)

h i
2
D2 + (D − 2) x = 2cos2t − 4sin2t.....................................(vi)

D2 + D2 − 4D + 4 x = 0 or 2D2 − 4D + 4 x = 2cos2t − 4sin2t

D2 − 2D + 2 x = cos2t − 2sin2t

For the complimentary solution

D2 − 2D + 2 x = 0

√
2± 4−8
D= =1±i
2

xc = et [C1 cost + C2 sint]

For the particular integral

116
 1 2
xp = cos2t − 2 sin2t
D2 − 2D + 2 D − 2D + 2

1 2
= cos2t − sin2t
−4 − 2D + 2 −4 − 2D + 2

−1 2
= cos2t − sin2t
2D + 2 D+1

−2D − 2 D−1
= cos2t − 2 sin2t
4D2 − 4 D −1

2D − 2 D−1
= cos2t − sin2t
−16 − 4 −4 − 1

1 1
= (D − 1) cos2t − (D − 1) sin2t
10 5

1 1
= (−2sin2t − cos2t) − (2cos2t − sin2t)
10 5

1
= − cos2t
2

x = xc + xp

1
x = et [C1 cost + C2 sint] − cos2t.........................................(vii)
2

dx
= et [−C1 cost + C2 sint] + et [C1 cost + C2 sint] + sin2t..............(viii)
dt

117
 Adding equations (i) and (ii), we get

dx
2 + 2y − 2x = cos2t + sin2t.........................................(ix)
dt

Putting the values of x and dx

dt from (viii)

 
t t

t 1
2 e [−C1 cost + C2 sint] + e [C1 cost + C2 sint] + sin2t +2y−2 e [C1 cost + C2 sint] − cos2t
2

= cos2t + sin2t

1
y = et [C1 cost + C2 sint] − sin2t
2

Exercise:
Solve the following system of equations:
(a)(D − 3) x + 2 (D + 2) y = e2t
2 (D + 1) x + (D − 1) y = 0

118
10 VARIATION OF PARAMETERS

Consider the general second-order linear dierenmtial equation with variable

coecients.

a0 (x)y 00 + a1 (x)y 0 + a2 (x)y = Q(x)..............................(i)

Suppose that y1 and y2 are linearly independent solutions of the corresponding

homogeneous solution

a0 (x)y 00 + a1 (x)y 0 + a2 (x)y = 0..............................(ii)

Then the complimentary function of equation(i) is

C1 y1 (x) + C2 y2 (x)

where C1 and C2 are arbitrary constants. The procedure in the method of

variation of parameters is to rplace the arbitrary constants in the complimentary
function by respective functions, V1 and V2 which will nbe determined so that the
resulting function will be dened by:

V1 y1 (x) + V2 y2 (x)..................................................(iii)

will be a particular integral of equation (i). Hence the name variation of

parameters.
We have at our disposal the two functions V1 and V2 with which to satisfy the
one condition that (iii) is a solution of (i). Since we have two functions but ony
one condiion on them, we are thus free to impose a second condition, provided
that theis second condition does not violate the rst one. We shall see when
and how to impose this additional condition as we proceed.
We thus assume the solution is of the form equation (iii)

119
 yp (x) = V1 (x)y1 (x) + V2 (x)y2 (x)..................................................(iv)

Dierentiating (iv), we have:

yp0 (x) = V1 (x)y10 (x)+V2 (x)y20 +V10 (x)y1 (x)+V20 (x)y2 (x)..................................................(v)

At this point we impose the second condition, we simplify yp0 by demanding

that;

V10 (x)y1 (x) + V20 (x)y2 (x) = 0.............................................(vi)

with this condition imposed, (v) reduces to:

yp0 (x) = V1 (x)y10 (x) + V2 (x)y20 ...................................(vii)

Now dierentiating (vii), we obtain:

yp00 (x) = V1 (x)y100 (x)+V2 (x)y200 +V 0 (x)y10 (x)+V20 (x)y20 ...................................(viii)

We now impose the basic condition that (iv) be a solution of equation (i).
Thus we substitute (iv), (vii)and (viii) for y, y 0 and y 00 , repectively in equation
(i) and obtain the following identity;

a0 (x) [V1 (x)y100 (x) + V2 (x)y200 + V 0 (x)y10 (x) + V20 (x)y20 ]+a1 (x) [V1 (x)y10 (x) + V2 (x)y20 ] +

a2 (x) [V1 (x)y1 (x) + V2 (x)y2 (x)] = Q(x)..............................(ix)

120
 Since y1 and y2 are solutions of the corresponding homogeneous dierential
equation (ii), the expression in the rst two brackets in (ix) are identically zero.
This leaves merely;

Q(x)
[V10 (x)y10 (x) + V20 (x)y20 (x)] = ..............................(x)
a0 (x)

This is actually what the basic condition demand. Thus the imposed con-
ditions require that the fnctions V1 and V2 be chosen such that the system of
equations

V1 (x)y10 (x) + V2 (x)y20 (x) = 0

Q(x)
V20 (x)y10 (x) + V20 (x)y20 (x) = ..............................(xi)
a0 (x)

is satised. The determinant of coecients of this system is precisely;

 
 
y1 (x) y2 (x)
ω (y1 (x), y2 (x)) = 
 

 0
y1 (x) y20 (x)


Since y1 and y2 are linearly independent solutions of the corresponding homo-

geneous dierential equation(ii), we know that ω (y1 (x), y2 (x)) 6= 0. Hence the
system (ii) has a unique solution. Actually, solving the system, we obtain:
 
 
 0 y2 (x)

 
 Q(x)
y20 (x)

 a0 (x) Q(x)y2 (x)
V10 (x) =  =
y1 (x)
 a0 (x)ω (y1 (x), y2 (x))
 y2 (x)
 
 0
y1 (x) y20 (x)


121
  
 
y1 (x) 0 
 
 
 0 Q(x) 
y1 (x) a0 (x)  Q(x)y1 (x)
V20 (x) =  =
y1 (x) y2 (x)
 a0 (x)ω (y1 (x), y2 (x))
 
 
 0
y1 (x) y20 (x)


Thus we obtain,

ˆx
Q(x)y2 (x)
V1 (x) = dx..............................(xii)
a0 (x)ω (y1 (x), y2 (x))

ˆx
Q(x)y1 (x)
V2 (x) = dx..............................(xiii)
a0 (x)ω (y1 (x), y2 (x))

Therefore the particular integral is given by

yp (x) = V1 (x)y1 (x) + V2 (x)y2 (x)

where V1 (x)and V2 (x)are dened by (xii)and(xiii).

Example:
1. Consider the dierential equation D2 + 1 y = tanx....................................(i)

Solution

yc (x) = C1 sinx + C2 cosx.......................................(ii)

Assume

yp (x) = V1 (x)sinx + V2 (x)cosx.......................................(iii)

where the functions V1 (x)and V2 (x) will be determined such that this is a
particular integral of the dierential equation (i). Then

yp0 (x) = V1 (x)cos(x)+V2 (x)sin(x)+V10 (x)sin(x)+V20 (x)cos(x)..................................................(iv)

122
 We impose the condition;

V10 (x)sin(x) + V20 (x)cos(x) = 0.............................................(v)

leaving

yp00 (x) = −V1 (x)sin(x)−V2 (x)cos(x)+V10 (x)cos(x)−V20 (x)sin(x)...................................(vi)

Substituting (vi) and (iii), in to (i)

V10 (x)cos(x) − V20 (x)sin(x) = tanx.............................................(vii)

Thus we have two equations,

V10 (x)sin(x) + V20 (x)cos(x) = 0

V10 (x)cos(x) − V20 (x)sin(x) = tanx

 
 
 0 cos(x) 

 
tanx −sin(x)
 
−cos(x)tan(x)
V10 (x) =  = = sin(x)

sin(x) cos(x)  −1
 
 
cos(x) −sin(x)
 

 
 
sin(x) 0 
 
 
 cosx −tan(x)
 
0 −sin(x)tan(x) −sin2 (x) cos2 (x) − 1
V2 (x) = 
  = = = = cos(x)−sec(x)

sin(x) cos(x)  −1 cos(x) cos(x)
 
 
cos(x) −sin(x)
 

123
 Integrating we nd that:

V1 (x) = −cos(x) + C3 ....................................(viii)

V2 (x) = sin(x) − ln |sec(x) + tan(x)| + C4 ....................(ix)

Substituting V1 (x)and V2 (x) into (iii),

yp = (−cos(x) + C3 ) sin(x) + (sin(x) − ln |sec(x) + tan(x)| + C4 ) cos(x)

= −cos(x)sin(x)+C3 sin(x)+sin(x)cos(x)−ln |sec(x) + tan(x)| cos(x)+C4 cos(x)

= C3 sin(x) + C4 cos(x) − (ln |sec(x) + tan(x)|) cos(x)

y = yc + yp

y = C1 sinx + C2 cosx + C3 sin(x) + C4 cos(x) − (ln |sec(x) + tan(x)|) cos(x)

= (C1 + C3 ) sinx + (C2 + C4 ) cosx − (ln |sec(x) + tan(x)|) cos(x)

Let C1 + C3 = Aand C2 + C4 = B

y = Asinx + Bcosx − (ln |sec(x) + tan(x)|) cos(x)

124
 2. Solve the dierential; equatin using method of variation of parameters
y 000 − 6y 00 + 11y 0 − 6y = ex

Solution

D3 − 6D2 + 11D − 6 y = ex ...........................(i)

yc = C1 ex + C2 e2x + C3 e3x ..........................(ii)

Assume,

yp = V1 (x)ex + V2 (x)e2x + V3 (x)e3x .............................(iii)

Since we have three functions V1 (x), V2 (x)and V3 (x)

yp0 = V10 (x)ex +V20 (x)e2x +V30 (x)e3x +V1 (x)ex +2V2 (x)e2x +3V3 (x)e3x .............................(iv)

Proceeding in a manner analogous to the second order case, we impose the

condition,

V10 (x)ex + V20 (x)e2x + V30 (x)e3x = 0..................................(v)

Equation(iv) reduces to:

yp0 = V1 (x)ex + 2V2 (x)e2x + 3V3 (x)e3x .............................(vi)

Then

yp00 = V10 (x)ex +2V20 (x)e2x +3V30 (x)e3x +V1 (x)ex +4V2 (x)e2x +9V3 (x)e3x .............................(vii)

We now impose the condition,

125
 V10 (x)ex + 2V20 (x)e2x + 3V30 (x)e3x = 0.............................(viii)

Then equation (vii) reduces to:

yp00 = V1 (x)ex + 4V2 (x)e2x + 9V3 (x)e3x .............................(ix)

yp000 = V10 (x)ex +4V20 (x)e2x +9V30 (x)e3x +V1 (x)ex +8V2 (x)e2x +27V3 (x)e3x .............................(x)

Substituting equations (iii),(vi),(ix),(x) in equation (i),

V10 (x)ex +4V20 (x)e2x +9V30 (x)e3x +V1 (x)ex +8V2 (x)e2x +27V3 (x)e3x −6 V1 (x)ex + 4V2 (x)e2x + 9V3 (x)e3x

+11 V1 (x)ex + 2V2 (x)e2x + 3V3 (x)e3x −6 V1 (x)ex + V2 (x)e2x + V3 (x)e3x = ex

V10 (x)ex +4V20 (x)e2x +9V30 (x)e3x +V1 (x)ex +8V2 (x)e2x +27V3 (x)e3x −6V1 (x)ex −24V2 (x)e2x −54V3 (x)e3x

+11V1 (x)ex + 22V2 (x)e2x + 33V3 (x)e3x − 6V1 (x)ex − 6V2 (x)e2x − 6V3 (x)e3x = ex

or

V10 (x)ex + 4V20 (x)e2x + 9V30 (x)e3x = ex ............................(xi)

Thus we have three equations (v),(viii) and (xi) from which to determine

126
V1 (x), V2 (x)and V3 (x),

V10 (x)ex + V20 (x)e2x + V30 (x)e3x = 0

V10 (x)ex + 2V20 (x)e2x + 3V30 (x)e3x = 0

V10 (x)ex + 4V20 (x)e2x + 9V30 (x)e3x = ex

Solving we nd,
 
2x 3x
 
0 e e 

 
0
 2e2x 3e3x 
 
 x
e 4e2x 9e3x 

0 1
V1 (x) =  =
ex 2
e2x e3x 


 
ex 2e2x 3e3x 

 
 x
e 4e2x 9e3x 


 
 x
e3x 

e 0

 
ex 0 3e3x 

 
 x
e ex 9e3x 

V20 (x) =   = −e−x
ex e2x e3x 


 
ex 2e2x 3e3x 

 
 x
e 4e2x 9e3x 


 
 x 2x

e e 0 

 
ex 2e2x 0 

 
 x
e 4e2x ex 

1
V30 (x) =   = e−2x
ex 2
e2x e3x 


 
ex 2e2x 3e3x 

 
 x
e 4e2x 9e3x 


127
 Integrating,

1
V1 (x) = x + C4 ....................................(xii)
2

V2 (x) = e−x + C5 .................................(xiii)

1
V3 (x) = − e−2x + C6 .................................(xiv)
4

Replacing (xii),(xiii) and (xiv) in (iii)

   
1 x −x

2x 1 −2x
yp = x + C4 e + e + C5 e + − e + C6 e3x .............................(xv)
2 4

1 x 1
yp = xe + C4 ex + ex + C5 e2x − ex + C6 e3x
2 4

1 x 3 x
= xe + e + C4 ex + C6 e3x
2 4

y = yc + yp

1 3
= C1 ex + C2 e2x + C3 e3x + xex + ex + C4 ex + C6 e3x
2 4

 
3 1
= C1 + + C4 ex + C2 e2x + (C3 + C6 ) e3x + xex
4 2

Let C1 + 43 + C4 = Aand C3 + C6 = B

1
y = Aex + C2 e2x + Be3x + xex
2

Exercise:

128
 Evaluate the folowing dierential equations using method of variation of
parameters:
a) y 00 + 4y = sec(x)
b) y 00 + 9y = 12sec(3x)
c) y 00 + y = sec(x)
d) d2 y
dx2 + y = sec(x)tan(x)

e)
2
d y
+ y = x − cot(x)
dx2

f) D2 − 6D + 9 y = ex2

3x

g) y 00 + 4y = sec(x)
h) y 00 + 9y = 12sec(3x)

129
11 METHOD OF UNDETERMINED COEFFI-

CIENTS

The method of undetermined coecients is based on expressing the form of

yp upto arbitrary multiplicative constants and these constants are then evaluated

by substituting tthen proposed solution into the given dierential equations and
equating coecients of the like terms.
Case 1: If φ(x) = Pn (x)is an nth order polynomial in x, then we assume
that the solution of yp is given by:

yp = An xn + An−1 xn−1 + ... + A1 x + A0

where the Aj are the multiplicative constants

Case 2: If the φ(x) = eαx Pn (x)where αis a known constant and Pn is as
above. Therefore,

yp = eαx An xn + An−1 xn−1 + ... + A1 x + A0

Case 3: If the φ(x) = eαx Pn (x)sinβxwhere αis a known constant and Pn is

as above. Therefore,

yp = eαx sinβx An xn + An−1 xn−1 + ... + A1 x + A0 +eαx sinβx An xn + An−1 xn−1 + ... + A1 x + A0

Case 4: If the φ(x) = eαx Pn (x)cosβxwhere αis a known constant and Pn is

as above. Therefore,

yp = eαx sinβx An xn + An−1 xn−1 + ... + A1 x + A0 +eαx sinβx An xn + An−1 xn−1 + ... + A1 x + A0

The modication of this method is that if any term of the assumed solution
regardless of the constant is a term of yc then the assumed solution must be

130
modied by multiplying it by xm where m is the lowest positive integer such that
the product xm with the assumed solution has no common term with yc .
The limitation of this method is that it cannot be used if p(x) does not
fall in any of the above four cases discussed. If the equation does not contain
coecients, in such a situation another method can be used.

• If yc = c1 x + c0 is the solution of P (x) = 9x2 + 2x − 1. Since Pn (x)is

of second order degree polynomial(case 1) we assume a colution of the

form yp = A2 x2 + A1 x + A0 . However this assumed solution has terms in
common with yc , particularly c1 xand A1 x. Therefore, we multiply yp by
x2 so that the new yp becomes yp = A2 x4 + A1 x3 + A0 x2 .Hence, no ters is

now common with yc .

• If p(x) = (x − 1) sinx + (x + 1) cosx. Assumed solution for (x − 1) sinx

will be that yp = (A1 x + A0 ) sinx + (B1 x + B0 ) cosx. Whereas that

of (x + 1) cosx, will be yp = (A1 x + A0 ) sinx + (B1 x + B0 ) cosx. The
total solution is therefore, yp = (A1 x + A0 ) sinx + (B1 x + B0 ) cosx +
(A1 x + A0 ) sinx + (B1 x + B0 ) cosx

Example:

1. Solve by method of undetermined coecients d2 y

dx2
dy
− 2 dx = ex sinx.

Solution

D2 − 2D y = 0

[D (D − 2)] y = 0

D = 0, D = 2

yc = c1 + c2 e2x

131
Assume the solution is of the form;

yp = A1 ex sinβx + A2 ex cosβx...............................(i)

yp0 = A1 (ex sinβx + ex cosβx) + A2 (ex cosβx − ex sinβx)

yp00 = A1 (ex sinβx + ex cosβx + ex cosβx − ex sinβx)+A2 (ex cosβx − ex sinβx − ex sinβx − ex cosβx)

yp00 = 2A1 ex cosβx − 2A2 ex sinβx

Replacing,

2A1 ex cosβx−2A2 ex sinβx−2A1 ex sinβx−2A1 ex cosβx−2A2 ex cosβx+2A2 ex sinβx = ex sinx

−2A1 ex sinβx − 2A2 ex cosβx = ex sinx

equation the like terms;

−2A1 ex sinβx = ex sinx

1
A1 = −
2

2A2 ex cosβx+ = 0

132
 A2 = 0

Replacing A1 and A2 back to equation (i), we have:

1
yp = − ex sinβx
2

The coplete solution is given by

y = yc + yp

1
= c1 + c2 e2x − ex sinβx
2

2. Exercise: By the method of undetermined coecients, solve the dier-

ential equation d2 y
dx2
dy
− 2 dx + 3y = x3 + cosx.

133
12 LAPLACE TRANSFORMATIONS

´∞
By denition L {y(t)} = 0
e−st y(t)dt = y(s).
´∞
L {y 0 (t)} = 0
e−st y 0 (t)dt.

By integration by parts;

du
u = e−st =⇒ = −se−st
dt

dv
= y 0 (t) =⇒ v = y(t)
dt

ˆ ∞
0 −st
L {y (t)} = e y(t) |∞
0 + se−st y(t)dt
0

ˆ ∞
= 0 − y(0) + s e−st y(t)dt
0

= sy(s) − y(0)

´∞
L {y 00 (t)} = 0
e−st y 00 (t)dt.

du
u = e−st =⇒ = −se−st
dt

dv
= y 00 (t) =⇒ v = y 0 (t)
dt

ˆ ∞
L {y 00 (t)} = e−st y 0 (t) |∞
0 + se−st y(t)dt
0

ˆ ∞
= 0 − y 0 (0) + s e−st y 0 (t)dt
0

= s2 y(s) − sy(0) − y 0 (0)

134
 Similarly,
´∞
L {y 000 (t)} = 0
e−st y 000 (t)dt = s3 y(s) − s2 y(0) − sy 0 (0) − y 00 (0).

In general for a diernetial equation a0 y n +a1 y n−1 +...+an y = y(t).......................(i).

We require the solution y(t) with initial conditions y(0) = y0 , y 0 (0) = y1 ,
y 00 (0) = y2 ............y n−1 (0) = yn−1 to take the laplace transform,

L {y n (t)} = sn (y(s)) − sn−1 y(0) − sn−2 y(0)..... − y n−1 (0)

Example:

1. Solve the Initial Value Problem y 00 − 3y 0 + 2y = 2e3t , given y(0) = 2and

y 0 (0) = 3 by laplace transforms method.

Solution

2
s2 y(s) − sy(0) − y 0 (0) − 3 [sy(s) − y(0)] + 2y(s) =
s−3

2
s2 y(s) − 2s − 3 − 3 [sy(s) − 2] + 2y(s) =
s−3

2 2 + (2s − 3) (s − 3) 2s2 − 9s + 11
s2 − 3s + 2 y(s) =


+2s−3 = =
s−3 s−3 s−3

2s2 − 9s + 11
y(s) =
(s − 3) (s2 − 3s + 2)

2s2 − 9s + 11
=
(s − 3) (s − 1) (s − 2)

In partial fractions,

2s2 − 9s + 11 A B C
= + +
(s − 3) (s − 1) (s − 2) s−3 s−1 s−2

135
 A = 1 B = 2 C = −1

     
1 2 1
y(t) = L−1 {y(s)} = L−1 + L−1 − L−1
s−3 s−1 s−2

= e3t + 2et − e2t

2. Solve d2 y
dt2 dt − 8y = 0 using laplace transforms with initial values as
− 2 dy

y(0) = 3and y 0 (0) = 6

Solution

s2 y(s) − sy(0) − y 0 (0) − 2 [sy(s) − y(0)] − 8y(s) = 0

s2 y(s) − sy(0) − y 0 (0) − 2sy(s) + 2y(0) − 8y(s) = 0

s2 y(s) − 3s − 6 − 2sy(s) + 6 − 8y(s) = 0

s2 − 2s − 8 y(s) = 3s

3s 3s
y(s) = =
s2 − 2s − 8 (s − 4) (s + 2)

Applying partial fractions

3s A B
= +
(s − 4) (s + 2) s−4 s+2

136
 A=2B=1

   
−1 2 −1 1
y(t) = L +L
s−4 s+2

y(t) = 2e4t + e−2t

The following represent the standard inverse Laplace transforms:

1. L−1 1


s =1
 
2. L−1 1
s−a = eat

3. L−1 1 tn−1


sn = (n−1)!
 
4. L−1 1
(s−a)n = eat tn−1
(n−1)!
 
5. L−1 a
s2 +a2 = sinat
 
6. L−1 s
s2 +a2 = cosat
 
7. L−1 a
s2 −a2 = sinhat
 
8. L−1 s
s2 −a2 = coshat
 
9. L−1 b
(s−a)2 +b
= eat sinbt
 
10. L−1 s−a
(s−a)2 +b2
= eat cosbt
 
11. L−1 s
(s2 +a2 )2
= 1
2a tsinat
 
12. L−1 s
(s2 +a2 )2
= 1
2a3 (sinat − atcosat)

Exercise: Using method of Laplace transforms solve the following ordinary

dierential equations.

1. d2 y
dt2 + y = e−2t sint where, y(0) = 0 and y 0 (0) = 0

137
 2. d2 y
dt2 − dy
dt − 2y = 18e−t sin3t where, y(0) = 0 and y 0 (0) = 3

3. y 000 + 4y 00 + 5y 0 + 2y = 10cost where, y(0) = 0, y 0 (0) = 0 and y 00 (0) = 3

13 SOLUTIONS OF DIFFERENTIAL EQUATIONS

IN SERIES

13.1 Introduction
Consider the second order dierential equation

a0 (x)y 00 + a1 (x)y 0 + a2 (x)y = Q(x)..............................(i)

and suppose that the equation has no solution that is expressible as a nite
linear combination of known elementary functions due to the nature of the
variable coecients a0 (x), a1 (x) and a2 (x).If the coecients were constants then
the equation (1) could be easily solved by means of complimentary functions.
Let us assume, however, that it does have a solution that could be expressed in
the form of an innite series. Specically, we assume that it has a solution of
the form:

2
X n
y = C0 +C1 (x − x0 )+C2 (x − x0 ) +........ = Cn (x − x0 ) .......................(ii)

where C0 , C1 , C2 , ...., Cn are constants. An expression of the form (ii) is called

power series in (x − x0 ). We have thus assumes that the dierential equation (i)
has a so called power series solution of the form (ii). Assuming that this assump-
tion is valid, we can proceed to determine the constants;C0 , C1 , C2 , ...., Cn in (ii),
in such a manner that the expression (ii) does indeed satisfy equation (i). Now
we can write equation (i) into the equivalent normalized form as:

138
 d2 y a1 (x) dy a2 (x)
+ + y=0
dx2 a0 (x) dx a0 (x)

where a0 (x) 6= 0.
or

d2 y dy
2
+ P1 (x) + P2 (x)y = 0..................................(iii)
dx dx

where P1 (x) = a1 (x)

a0 (x) ,P2 (x) = a2 (x)
a0 (x)

13.1.1 Analytic Function

A function f(x) is called an analytic function at a point x = x0 , if its Taylor
series about x0 given by

X f n (x0 ) n
f (x) = (x − x0 )
n!

exists and converges to f(x) for all x in an open interval of x0 . This is equivalent
to saying that the function is single valued and possesses derivatives of all orders
of that point. All polynomial functions are analytic everywhere and so also are
the functions et , cos(x), sin(x). A rational function is analytic except at these
points at which the denominator is zero. For example, the rational function;

1 1
= 2
x2 − 2x + 1 (x − 1)

is analytic everywhere except at x=1.

13.1.2 Ordinary Point

A point x = x0 is said to be ordinary if both P1 (x)and P2 (x)in the equivalent
normalized form (iii) are analytic at x = x0 . If the Taylor series expansion of
P1 (x)and P2 (x)are valid in |x − x0 | < R, then the Taylor series expansion of

the solution is also valid for the same range.

139
13.1.3 Singular Point
If either one or both P1 (x)and P2 (x)in the equivalent normalized form (iii) is
not analytic at the point x = x0 , then we say that the point x0 is a singular
point of the dierential equation(i0.
Exercise:
Determine the singular points of the equation
(x − 1) y 00 + xy 0 + x1 y = 0

Solution

x 1
y 00 + y0 + y=0
x−1 x (x − 1)

P1 (x) = x
x−1 ,P2 (x) = 1
x(x−1)

At x=0 and x=1, the dierential equation becomes innite and so these are
singular points. All the other points are ordinary points.

13.1.4 Regular Singular Point

If in the normalized form,

d2 y dy
+ P1 (x) + P2 (x)y = 0
dx2 dx

at least one of P1 (x)and P2 (x)is not analytic at x = x0 . (x0 is a singular

point ) but both the products (x − x0 ) P1 (x)and (x − x0 ) P2 (x)are analytic at
x = x0 then x0 is said to be a regular singular point of the dierential equation(i).

13.1.5 Irregular Singular Point

If x0 is a singular point of the normalized form of equation (i) and either
(x − x0 ) P1 (x)or (x − x0 ) P2 (x)is not analytic at x = x0 , then the point x = x0 is

said to be an irregular singular point of the dierential equation(i)

Exercise 1:
Find the singular, regular singular and irregular singular points of the dif-
ferential equation:

140
 (a)2x2 dx
2
d y dy
2 − x dx + (x − 5)y = 0

(b)x2 (x − 2)2 dx
2
d y dy
2 + 2(x − 2) dx + (x + 1)y = 0

Solution
(a) Writing the equation in the normalized form

d2 y x dy (x − 5)
− 2 + y=0
dx2 2x dx 2x2

P1 (x) = x
2x2 ,P2 (x) = (x−5)
2x2

We see that at x = 0, both P1 (x)and P2 (x)are innite. Therefore x = 0 is a

singular point. Forming the product then xP1 (x) = −1
2 ,x2 P2 (x) = (x−5)
2

which are both analytic at x = 0 . Therefore x = 0 is a regular singular

point of the dierential equation.
(b) The normalized form is:

d2 y 2(x − 2) dy (x + 1)
2
+ 2 + 2 y=0
dx x (x − 2) dx x (x − 2)2
2

P1 (x) = 2
x2 (x−2) ,P2 (x) = (x+1)
x2 (x−2)2

at x = 0 and x = 2both P1 (x)and P2 (x)are innite. Therefore x = 0and

x = 2are singular points of the dierential equation. At x = 0we dene the

products as:
xP1 (x) = 2
x(x−2) ,x2 P2 (x) = (x+1)
(x−2)2

Now xP1 (x)is not analytic but x2 P2 (x)is analytic at x = 0is an irregular
singular point..
At x = 2we dene the products,
(x − 2)P1 (x) = 2
x2 ,(x − 2)2 P2 (x) = (x+1)
x2

which are both analytic at x = 2. Therefore x = 2is a regular singular point

of the dierential equation.

13.2 Solution About an Ordinary Point

Method of Solution
Let the equation be

141
 a0 (x)y 00 + a1 (x)y 0 + a2 (x)y = Q(x)..............................(i)

then we assume it has a solution of the form

∞
X n
y= Cn (x − x0 ) .......................(ii)
n=0

where x0 is the ordinary point about which we are nding the solution. From
equation (ii) then,

∞
X n−1
y0 = nCn (x − x0 )
n=1

∞
X n−2
y 00 = n(n − 1)Cn (x − x0 )
n=2

On substituting these in equation (i) then

hX i hX i hX i
n−2 n−1 n
a0 (x) n(n − 1)Cn (x − x0 ) +a1 (x) nCn (x − x0 ) +a2 (x) Cn (x − x0 ) = Q(x)...........(ii

From equation (iii) we have to nd the radius of C0 , C1 , C2 , ...., Cn by equat-

ing constant terms, coecients x, x2 , x3 , ....to zero and nally substitute the
values of C0 , C1 , C2 , ...., Cn in (ii) to get the solution:
Exercise:
1. Solve the equation y 00 + 4y = 0near the ordinary point x = 0.
Solution
y 00 + 4y = 0.................................(i)

Here we let the solution be a power series of the form

∞
X
y= Cn xn .......................(ii)
n=0

142
 ∞
X
y0 = nCn xn−1 ....................(iii)
n=1

∞
X
y 00 = n(n − 1)Cn xn−2 .................(iv)
n=2

Substituting these in equation (i) then,

∞
X ∞
X
n(n − 1)Cn xn−2 + 4 Cn xn = 0.....................................(v)
n=2 n=0

We now change the indexing so that the series has the general form xn−2 by
putting n = n − 2in the second sum

∞
X ∞
X
n(n − 1)Cn xn−2 + 4 Cn−2 xn−2 = 0.....................................(vi)
n=2 n=2

Adding the two because they have the same index then

∞
X
[n(n − 1)Cn + 4Cn−2 ] xn−2 = 0.....................................(vii)
n=2

For the power series to vanish identically over interval then each coecient
in the series must be zero. Thus

n(n − 1)Cn + 4Cn−2 = 0 , n ≥ 2

−4Cn−2
Cn = , n ≥ 2..................................(viii)
n (n − 1)

Relation (viii) is called a recurrent relation and is used to nd Cn for n ≥ 2in
terms of C0 and C1 , which are left arbitrary we therefore have
C2 = −4C0
2.1 ,C4 = −4C2
4.3 ,.......C2k = −4C2k−2
2k(2k−1)

143
 C3 = −4C1
3.2 ,C5 = −4C3
5.4 ,......,C2k+1 = −4C2k−1
(2k+1)(2k)

We have taken both even and odd constants separately. Multiplying the
corresponding numbers of the even and odd;
−4C0 −4C2 −4C2k−2
C2 · C4 · ..... · C2k = 2.1 · 4.3 · ..... · 2k(2k−1)
(−1)k 4k
C2 · C4 · ..... · C2k = (2k)! C0 · C2 · C4 · ..... · C2k−2
k k
(−1) 4
C2k = (2k)! C0 , k ≥ 1...................................................(ix)

For the odd numbers,

−4C1 −4C3 −4C2k−1
C3 · C5 · ........ · C2k+1 = 3.2 · 5.4 · ..... · (2k+1)(2k)
(−1)k 4k
C3 · C5 · ........ · C2k+1 = (2k+1)! C1 · C3 · C5 · ..... · C2k−1
k k
(−1) 4
C2k+1 = (2k+1)! C1 , k ≥ 1................................................(x)

Therefore the solution becomes:

∞
X
y= C n xn
n=0

∞
X
= C0 + C1 x + Cn xn
n=2

∞
X ∞
X
= C0 + C1 x + C2k x2k + C2k+1 x2k+1
k=1 k=1

∞ ∞
X (−1)k 22k x2k X (−1)k x2k+1
= C0 + + C1 x + C1
(2k)! (2k + 1)!
k=0 k=0

∞ ∞
X (−1)k 22k x2k C1 X (−1)k x2k+1
= C0 +
(2k)! 2 (2k + 1)!
k=1 k=0

C1
= C0 cos(2x) + sin(2x)
2

2. Use the power series method to solve the dierential equation

1 − x2 y 00 − 6xy 0 − 4y = 0............................(i)

near the ordinary pointx = 0.

Solution

144
Here we let the solution be a power series of the form

∞
X
y= Cn xn .......................(ii)
n=0

∞
X
y0 = nCn xn−1 ....................(iii)
n=1

∞
X
y 00 = n(n − 1)Cn xn−2 .................(iv)
n=2

Substituting in equation (i) yields;

"∞ # " ∞
# " ∞
#
2

X n−2
X
n−1
X
n
1−x n(n − 1)Cn x − 6x nCn x −4 Cn x =0
n=2 n=1 n=0

∞
X ∞
X ∞
X ∞
X
n(n − 1)Cn xn−2 − n(n − 1)Cn xn − 6 nCn xn − 4 C n xn = 0
n=2 n=2 n=1 n=0

∞
X ∞
X
n−2
n2 + 5n + 4 Cn xn = 0


n(n − 1)Cn x −
n=2 n=0

Shifting the index in the second summation by replacing n with n-2,

∞
X ∞ 
X 
2
n(n − 1)Cn xn−2 − (n − 2) + 5 (n − 2) + 4 Cn−2 xn−2 = 0
n=2 n=0

∞
X ∞
X
n(n − 1)Cn xn−2 − (n − 1) (n + 2) Cn−2 xn−2 = 0
n=2 n=2

∞
X
[n(n − 1)Cn − (n − 1) (n + 2) Cn−2 ] xn−2 = 0................(ii)
n=2

Since the coecients of various powers of x must be identically zero then,

145
 n(n − 1)Cn − (n − 1) (n + 2) Cn−2 = 0

Thus for n ≥ 2,

n+2
Cn = Cn−2 .....................(iii)
n

The equation (iii) is called a recurrence relation. It gives Cn in terms of the

proceed C 0 s.
C2 = 24 C0 ,C4 = 64 C2 ,C6 = 68 C4 ,.....,C2k = 2k+2
2k C2k−2

C3 = 53 C1 ,C5 = 75 C3 ,C6 = 79 C5 ,.....,C2k+1 = 2k+3

2k+1 C2k−1

Getting the product of the even and odd rows above,

C2 · C4 · C6 · ... · C2k = 42 C0 · 46 C2 · 86 C4 · ... · 2k+2
2k C2k−2
2k+2
C2k = 2 C0

C2k = (k + 1) C0

Also,
C3 · C5 · C7 · ... · C2k+1 = 53 C1 · 75 C3 · 97 C5 · ... · 2k+3
2k+1 C2k−1
1
C2k+1 = 3 (2k + 3) C1

Since the nature of the expressions for C 0 s depends on where the subscript
is odd or then the expression of y becomes:

∞
X
y= C n xn
n=0

∞
X ∞
X
2k
= C2k x + C2k+1 x2k+1
k=0 k=0

∞ ∞
X X 1
= (k + 1) C0 x2k + (2k + 3) C1 x2k+1
3
k=0 k=0

∞ ∞
X C1 X
= C0 (k + 1) x2k + (2k + 3) x2k+1
3
k=0 k=0

146
13.3 Solution about a regular Singular Point
Rules of nding the Solution
a) Put y = xm ,y 0 = mxm−1 ,y 00 = m (m − 1) xm−2 and nd the common
dierence of the powers of x. If it is s(say) then we take the assumed solution
as

∞
X
y= ar xm+rs
r=0

b) Find the derivatives of the above series and substitute in the given dier-
ential equation.
c) Find the initial equation from the identity by equating the coecients of
either the lowest powers of x(or the highest power of x) to zero.
d) From the initial equation nd the roots.
e) From the roots, a recurrence solution can be obtained which will be a
relation in ar+1 and ar .
f) Find the constants a1 , a2 , ...in terms of a0 from the recurrence relation.

• If the roots of the initial equation are determined and not dened by an

integer then the completed solution will be

y = C1 (y) m1 + C2 (y) m2

where m1 and m2 are the roots of the initial equation.

• If the roots of the initial equation are equal, say m1 and m1 then the com-

plete solution will be

 
δy
y = C1 (y) m1 + C2 m1
δm

g) If the roots of the initial equation are distinct and dierent integers and if
some of the coecients of the y-series become innite when m = m1 , we modify
the form of y by replacing a0 by b (m − m1 ). We then obtain the complete
solution as

147
  
δy
y = C1 (y) m1 + C2 m1
δm

Exercise:
1. Solve the dierential equation 9x (1 − x) dx
2
d y dy
2 −12 dx +4y = 0..................(i)

about the point x = 0.

Solution
Let y = xm ,y 0 = mxm−1 ,y 00 = m (m − 1) xm−2 in equation (i) then,

9x (1 − x) m (m − 1) xm−2 − 12 mxm−1 + 4xm = 0

   

9m (m − 1) xm−1 + m (m − 1) xm − 12mxm−1 + 4xm = 0

So here the common dierence is

s = m − (m − 1) = 1

Therefore the solution assumed is of the form:

∞
X
y= ar xm+r
r=0

∞
X
0
y = ar (m + r) xm+r−1
r=0

∞
X
y 00 = ar (m + r − 1) (m + r) xm+r−2
r=0

Substituting these in equation (i)

∞
X ∞
X ∞
X
9x (1 − x) ar (m + r − 1) (m + r) xm+r−2 −12 ar (m + r) xm+r−1 +4 ar xm+r = 0
r=0 r=0 r=0

148
∞
X ∞
X ∞
X ∞
X
m+r−1 m+r m+r−1
9ar (m + r − 1) (m + r) x − 9ar (m + r − 1) (m + r) x − 12ar (m + r) x + 4ar xm+
r=0 r=0 r=0 r=0

∞
X
9 (m + r − 1) (m + r) xm+r−1 − 12 (m + r) xm+r−1 − [9 (m + r − 1) (m + r) − 4] xm+r = 0
 
ar
r=0

∞
X
ar (m + r) [9 (m + r − 1) − 12] xm+r−1 − [9 (m + r − 1) (m + r) − 4] xm+r = 0..........(iii)

r=0

Equation (iii) is an identity. We nd the indicial equation by equating the

coecients of the lowest power of x to 0 at dierent values of r. The lowest
power of x to 0 at dierent values of x is from xm+r−1 when r=0, i.e. xm−1 so at
r=0 the coecient of the lowest power of x is:

a0 m (9m − 12) = 0

which is the indicial equation

Thus, m = 0or m = 7
3

We see that the dierence between these roots is not an integer. On equating
the coecient of the general term xm+r to zero, replace r by r+1 in the rst
summation, then

ar+1 (m + r + 1) [9 (m + r + 1) − 12] xm+r−1 − [9 (m + r − 1) (m + r) − 4] xm+r ar = 0

 

so

9 (m + r − 1) (m + r) − 4
ar+1 = ar
(m + r + 1) [9 (m + r + 1) − 21]

149
which is the recurrence relationship for a0r s. So at r=0,

9 (m − 1) (m) − 4
a1 = a0
(m + 1) [9 (m + 1) − 12]

9m2 − 9m − 4
= a0
(m + 1) (9m − 12)

(3m + 1) (3m − 4)
= a0
3 (m + 1) (3m − 4)

(3m + 1)
= a0
3 (m + 1)

At r=1,

9 (m + 1) (m) − 4
a2 = a1
(m + 2) [9 (m + 2) − 12]

9m2 − 9m − 4
= a1
(m + 2) (9m − 3)

(3m − 1) (3m + 4)
= a1
3 (m + 2) (3m − 1)

(3m + 4)
= a1
3 (m + 2)

(3m + 4) (3m + 1)
= · a0
3 (m + 2) 3 (m + 1)

Therefore

(3m + 1) (3m + 4) (3m + 7)

a3 = · · a0
33 (m + 1) (m + 2) (m + 3)

(3m + 1) (3m + 4) (3m + 7) (3m + 10)

a4 = · · · a0
34 (m + 1) (m + 2) (m + 3) (m + 4)

150
 Therefore the solution is

∞
X
y= ar xm+r
r=0

∞
X
= xm ar xr
r=0

= xm a0 + a1 x + a2 x2 + a3 x3 + .......
 

at m = 0

 
1 1·4 2 1·4·7 3
y1 = 1 a0 + a0 x + 2 a0 x + 3 a0 x + .....
3 3 (2 · 1) 3 (3 · 2 · 1)

 
1 1·4 1·4·7
y1 = a0 1 + x + 2 x2 + 3 x3 + .....
3 3 (2 · 1) 3 (3 · 2 · 1)

at m = 73 then,

 
7 8 8 11 2 8 11 14 3
y2 = x a0 + a0 x +
3 · a0 x + · · a0 x + .....
10 10 13 10 13 16

 
7 8 8 11 2 8 11 14 3
y2 = a 0 x 3 1 + x + · x + · · x + .....
10 10 13 10 13 16

The complete solution will be a linear combination of y1 and y2

  
1 1·4 2 1·4·7 3 7 8 8 11 2 8 11 14 3
y = C1 a0 1 + x + 2 x + 3 x + ..... +C2 a0 x 1 + x +
3 · x + · · x + ..
3 3 (2 · 1) 3 (3 · 2 · 1) 10 10 13 10 13 16

2. Solve the equation x dx + xy = 0...............................(i)about the

2
d y dy
2 + dx

singular point x=0.

Solution
Let y = xm ,y 0 = mxm−1 ,y 00 = m (m − 1) xm−2 in equation (i) then,

151
 x m (m − 1) xm−2 + mxm−1 + x · xm = 0
 

m (m − 1) xm−1 + mxm−1 + xm+1 = 0

Therefore the common dierence is s = (m + 1) − (m − 1) = 2

Let
∞
X
y= ar xm+2r
r=0

∞
X
y0 = ar (m + 2r) xm+2r−1
r=0

∞
X
y 00 = ar (m + 2r − 1) (m + 2r) xm+2r−2
r=0

Substituting these in equation (i) then,

∞
X ∞
X ∞
X
x ar (m + 2r − 1) (m + 2r) xm+2r−2 + ar (m + 2r) xm+2r−1 +x ar xm+2r = 0
r=0 r=0 r=0

∞
X ∞
X ∞
X
ar (m + 2r − 1) (m + 2r) xm+2r−1 + ar (m + 2r) xm+2r−1 + ar xm+2r+1 = 0
r=0 r=0 r=0

∞
X
ar [(m + 2r − 1) (m + 2r) + (m + 2r)] xm+2r−1 + xm+2r+1 = 0

r=0

∞
X
ar (m + 2r) (m + 2r) xm+2r−1 + xm+2r+1 = 0

r=0

152
 ∞
X n o
2
ar (m + 2r) xm+2r−1 + xm+2r+1 = 0
r=0

To nd the indicial equation from (2), we equate the coecient of the lowest
power of x into zero to get (at r = 0, xm−1 )

a0 m2 = 0, a0 6= 0

m = 0, 0which are two equal roots

Now equating the coecient of the general term xm+2r+1 to zero, we get

2
ar+1 (m + 2r + 2) + ar = 0

−1
ar+1 = 2 ar
(m + 2r + 2)

at r=0

−1
a1 = 2 a0
(m + 2)

at r=1

−1 1
a2 = 2 a1 = 2 2 a0
(m + 4) (m + 2) (m + 4)

at r=2

−1 −1
a3 = 2 a2 = 2 2 2 a0
(m + 6) (m + 2) (m + 4) (m + 6)

Therefore,

∞
X
y= ar xm+r
r=0

153
 ∞
X
= xm ar xr
r=0

= xm a0 + a1 x2 + a2 x4 + a3 x6 + .......
 

" #
m −1 2 1 4 −1 6
=x a0 + 2 a0 x + 2 2 a0 x + 2 2 2 a0 x + .......
(m + 2) (m + 2) (m + 4) (m + 2) (m + 4) (m + 6)

" #
−1 1 −1
= a0 xm 1 + 2x
2
+ 2 2x
4
+ 2 2 2x
6
+ .......
(m + 2) (m + 2) (m + 4) (m + 2) (m + 4) (m + 6)

At m=0
" #
−1 2 1 4 −1 6
y1 = a0 1 + 2x + 2 2x + 2 2 2x + .......
(2) (2) (4) (2) (4) (6)

" #
δy m 1 2 1 4 1 6
= x lnxa0 1 − 2x + 2 2x − 2 2 2 x + .......
δm (m + 2) (m + 2) (m + 4) (m + 2) (m + 4) (m + 6)

" ( ) #
2x2 2 2
+xm a0 3 − x4 3 2 + 2 3 + ......
(m + 2) (m + 2) (m + 4) (m + 2) (m + 4)

Therefore,

 
δy
m=0
δm

   2   
1 1 1 x 1 1
= a0 lnx 1 − 2 x2 + 2 2 x4 − 2 2 2 x6 + ...... +a0 2 − 2 2 1 + x4 + .......
2 2 ·4 2 ·4 ·6 2 2 ·4 2

154
 The general solution is:

 
δy
y = C1 (y) m1 + C2 m1
δm

" #
−1 2 1 4 −1 6
y = C1 a0 1 + 2x + 2 2x + 2 2 2x + .......
(2) (2) (4) (2) (4) (6)

   2   
1 1 1 x 1 1
+a0 lnx 1 − 2 x2 + 2 2 x4 − 2 2 2 x6 + ...... +a0 2 − 2 2 1 + x4 + .......
2 2 ·4 2 ·4 ·6 2 2 ·4 2

14 TAYLOR SERIES

1. Use the Taylor series to nd the power series solution of the initial value
problem:
dy
dx = x2 + y 2 .....................(i)where y(0) = 1in powers of x.

Solution
The Taylor series is dened by

x2 00 x3
p(x) = f (0) + xf 0 (0) + f (0) + f 000 (0) + ....
2! 3!

Let

x2 00 x3
y(x) = y(0) + xy 0 (0) + y (0) + y 000 (0) + ....
2! 3!

be the solution from (i)

y 0 = x2 + y 2

y 0 (0) = 0 + y 2 (0) = 0 + 1 = 1

Dierentiating (i) with respect to x,

155
 d2 y dy
= y 00 = 2x + 2y = 2x + 2yy 0 = 0 + 2y(0)y 0 (0) = 0 + 2.1.1 = 2
dx2 dx

2
y 000 = 2 + 2yy 00 + 2 (y 0 ) = 2 + 2.1.2 + 2.1 = 8

Therefore the solution will be,

x2 x3
y =1+x+ ·2+ · 8 + ......
2! 3!

4
= 1 + x + x2 + x3 + ....
3

156
15 PARTIAL DIFFERENTIAL EQUATIONS

Partial dierential equations are those equations which contain partial dieren-
tial coecients, independent and dependent variables. The independent vari-
ables will be denoted by x and y and the dependent variables by z. The partial
dierential coecients are denoted as follows:
= p, = q, δ2 z
= r, δxδy = s,
2
δz δz δ z δ2 z
δx δy δx2 δy 2 =t

Order of a partial dierential equation is the same as that of the order of

the highest dierential coecient in it. Methods of forming partial dierential
coecients are:
a) B y eliminating of arbitrary constants
b) By eliminating arbitrary functions
(A) Method of eliminating arbitrary constants
Exercise
1. Form a p.d.e from: x2 + y 2 + (z − c)2 = a2
Solution
This equation contains arbitrary constants ; and c. Dierentiating partially
with respect to x, we get;

δz
2x + 2 (z − c) =0
δx

2x + 2 (z − c) p = 0................(i)

Dierentiating partially with respect to y

δz
2y + 2 (z − c) =0
δy

2y + 2 (z − c) q = 0................(ii)

Eliminating c from (i) and (ii),

157
 x y
(z − c) = − =−
p q

x
y− q=0
p

yp − xq = 0

(B) Method of eliminating arbitrary functions

2. Form a p.d.e from z = f x2 − y 2

Solution
Dierentiate with respect to x and y, to get

δz
= f 0 x2 − y 2 · 2x


p=
δx

δz
= f 0 x2 − y 2 · −2y


q=
δy

Dividing yields;

p x
=−
q y

py = −qx or yp + xq = 0

Exercise:
a) z = (x + a) (y + b) Ans: pq = z
b) (x − h)2 + (y − k)2 + z 2 = a2 Ans: z 2 p2 + q 2 + 1 = a2

c) 2z = (ax + y)2 + b Ans: px + qy = q 2

d) ax2 + by 2 + z 2 = 1 Ans: z (px + qy) = z 2 − 1
e)x2 + y 2 = (z − c)2 tan2 α Ans: yp − xq = 0

158
15.1 Solution of pde's by direct integration
1. Exercise: Solve δ3 z
δx2 δy = cos (2x + 3y)

Solution
Integrate with respect to x, we get

δ3 z 1
= sin (2x + 3y) + f (y)
δxδy 2

Integrating with respect to x

δz 1
= − cos (2x + 3y) + xf (y) + g (y)
δy 4

Integrating with respect to y

ˆ ˆ
1
z = − sin (2x + 3y) + x f (y)dy + g(y)dy
12

1
=− sin (2x + 3y) + xΦ1 (y) + Φ2 (y)
12

2. Solve δ2 z
δxδy = x2 y subject to the condition z (x, 0) = x2 and z = (1, y) =

cosy

Solution
On integrating with respect to x, we obtain

δz x3
= y + f (y)
δy 3

Integrating with respect to y we obtain

ˆ
x3 y 2
z= + f (y)dy + g(x)
3 2

or

x3 y 2
z= + F (y) + g(x)........................(i)
6

159
 ´
where F (y) = f (y)dy
Condition 1: Putting z = x2 and y = 0in (i) we get,

x2 = 0 + F (0) + g(x)

g(x) = x2 − F (0)

Substitute in equation (i), we get

x3 y 2
z= + F (y) + x2 − F (0)........................(ii)
6

Condition 2: z (1, y) = cos(y)

Putting x = 1, and z = cos(y)in equation (ii), we get

y2
cos(y) = + F (y) + 1 − F (0)
6

y2
cos(y) − − 1 + F (0) = F (y)
6

Putting the values of F(y) in (ii), we get

x3 y 2 y2
z= + cos(y) − − 1 + F (0) + x2 − F (0)
6 6

x3 y 2 y2
z= + cos(y) − − 1 + x2
6 6

3. Solve δ2 z
δy 2 = z , if y = 0,z = ex and δz
δy = e−x

Solution
If z was a function of y above, then

z = Asinh(y) + Bcosh(y)

where A and B are constants. Since z is a function of x and y, A and B are

160
functions of x. Then

z = sinh(y)f (x) + cosh(y) · Φ(x)..................................(i)

On putting y = 0,z = ex in (i) above, we obtain;

ex = Φ(x)

Substitute:

z = sinh(y)f (x) + cosh(y) · ex ..................................(ii)

On dierentiating with respect to y we get;

δz
= cosh(y)f (x) + sinh(y) · ex ..................................(iii)
δy

On putting y = 0, and δz
δy = e−x , we get

e−x = f (x)

and equation (ii) becomes:

z = e−x sinh(y) + ex cosh(y)

Exercise:
Solve the following
1. δ2 z
δyδx = xy 2

2. δ2 z
δxδy = ey cos(x)

3. δ2 z
δxδy = y
x +2

4. = a2 z , when x = 0, = asin(y)and
2
δ z δz δz
δx2 δx δy =0

5. = sin(x)sin(y)if = −2sin(y)and z = 0whenx = 0.

2
δ z δz
δxδy δy

Answers
1. z = x2 y 3
6 + f (y) + Φ(x)

161
2. z = ey sin(x) + f (y) + Φ(x)
3. z = y2
2 ln(x) + 2xy + f (y) + Φ(x)

4. z = sin(x) + ey cos(x)

162