1.

Partial Differential Equation (PDE)

A pde is an equation containing at least one partial derivative. For
example if
) , ( t x u u
then
,
2
2
2
x
u
k
t
u

Diffusion / Conduction
n
q E
.. (i)
,
2
2
2
2
2
x
u
c
t
u

Wave
.
n
q E
.
(ii)
and
0

x
u
c
t
u
. (iii)
are examples of
c k s e d p ,
are constants. The order of a pde is
the order of the highest partial derivative in the
. .
n
q e
For example
) ( , ) ( ii i
is 2
nd
. Order, whilst (iii) is 1
st
order.
The solution to a pde is when a function
) , ( t x
satisfies the pde
identically. For example x a e t x
t a k
sin ) , (
2 2

is a solution to
pde
. ) (i
A is arbitrary constant. To see this, we observe that
, sin ) (
2 2
2 2
x a e a k
t
i of s h L
t a k

whilst
{ } x a e a
x
k
x
k i of s h R
t a k
cos ) ( .
2 2
2
2
2
2

{ } x a e a k
t a k
sin
2 2
2 2

s h L
Now let
) , ( y x u u
satisfies

. 0
2

y x
u
x y u
.... (iv)
Then
. 0

,
_

y
u
x
Integrating
, . . x t r w
we obtain
. ) ( , y f
y
u

Integrating again ,
y t r w

+
y
x g s d s f y x u ) ( ) ( , ) , (
. ) ( ) ( ) , ( x g y f y x u +
2. Classification of 2
nd
order linear pde.
In this course, were going to concentrate on two variables only. Let
. ) , ( y x u u
Then the most general 2
nd
. Order linear pde is
y
u
y x E
x
u
y x D
y
u
y x C
y x
u
y x B
x
u
y x A

) , ( ) , ( ) , ( ) , ( ) , (
2
2 2
2
2
. ) , ( ) , ( y x G u y x F +

. (vi)
Any pde which is not in the form (vi) is called non-linear pde. When
, ) ) , ( y x G
pde (vi) is called a homogenous pde.
When :
, 0 4
2
< C A B
PDE is Elliptic .
0 4
2
> C A B
, PDE is Hyperbolic .
, 0 4
2
C A B
PDE is Parabolic .
3. Examples of pde solvable by direct integration
Ex . 1 Solve
1
2

y
u
y x
u
. (i)
Subject to
. sin ) , 0 ( , 0 ) 0 , ( y y
y
u
x u

Form (i) ,
. 1

'

u
x
u
y
assume
.
2
2 2
x y
u
y x
u

Integrating
, y t r w
) (
~
x f y u
x
u
+

e i . { } { }. ) (
~
x f y e u e
x
x x
+

Integrating
, x t r w

+
x
s x x
e y e u e
) ( x f
. ) ( ) ( ) , ( y g e x f e y y x u
x x
+ + . (ii)
{ } ) ( ) ( ) ( 9 ) ( ) , ( o g x f o g x f e o o o x u C B
x
+
From
y d
g d
e
y
u
x
+

1
and
y y o
y
u
sin ) , (

y d
g d
y + 1 sin
i.e.
c y y y g + cos ) (
,
c
.
and
. 1 ) ( c o g +
{ } y c y y c e y x u ii
x
+ + cos 1 ) , ( ) (
. . e i
. ) cos 1 ( ) , ( y y y e y x u
x
+ .
(iii)
Ex 2 . Solve
0 4
2
2

u
y
u
subject to
. ) , ( , ) , (
3 x
e o x
y
u
x o x u

. (i)
From the hint given by the BC, let
n
l o S
to (i) be
x
e y g x y f y x u

+ ) ( ) ( ) , (
3
.
(ii)
Substituting (ii) into (i) ,
{ } 0 4
3
2
2
3
2
2

'

+
x x
e g x f e
y d
g d
x
y d
f d
i.e.
. 0 4
4
2
2
2
2
3

'

'

g
y d
g d
e
f y d
f d
x
x
x
e x

,
3

is linearly independent of each other ,

, 0 4
2
2
f
y d
f d
and
0 4
2
2
g
y d
g d
y h D y C y g y h B y A y f 2 sin 2 cosh ) ( , 2 sin 2 cosh ) ( + +
x
e y h D y h C x y h B y A y x u

+ + + ) 2 sin 2 cos ( ) 2 sin 2 cos ( ) , (
3
0 , 1 ) . 1 . ( ) . 1 . ( ) , (
3 3 3
+ + +

C A e o D c x o B A x x o x u C B
x
From
x
e y h D y h C x y h A Z
y
u

+ +

) 2 cos 2 2 sin 2 ( ) 2 sin (

3
and
, ) , (
x
e o x
y
u

we obtain
2 / 1 . 1 2 , 0 ) 1 . 2 ( ) 1 . 2 (
3
+

D e i D B e D x B e
x x
.
) (iii
now becomes
. . 2 sin
2
1
. 2 cos ) , (
3 x
e y h x y h y x u

+
.
(iv)
Ex 3. Find the general solution of
. 0 , 0 . ,
2 2
2
2
> > +

y x e A u y
x
u
x
. (i)
A
.
Complementary
n
l o S .
:
Consider
. 0
2
2
2
+

u y
x
u
. (ii)
n
l o S .
to (ii) is
, sin ) ( cos ) ( ) , ( x y y g x y y f y x u
c
+
. (iii)
Where
g f ,
are arbitrary function of
y
.
Particular
n
l o S .
:
From the hint given by the non . homogenous term in (i), let
x
e y h y x p U
2
) ( ) , ( . (iv)
Substituting (iv) into (i) ,
x x x
e A e y h y e y h
2 2 2 2
) ( ) ( 4 +
4
) (
2
+

y
A
y h
.
4
) , (
2
2
x
e
y
A
y x p U
+

. (v)
p c
u u y x u + ) , (
i.e
+
+ +
4
sin ) ( cos ) ( ) , (
2
2
y
e A
x y y g x y y f y x u
x
.
(vi)
1. Modelling and PDE
(a) Difffusion / Conduction
.
n
g e
Let
) , , , ( t z y x T T
be the
temp at a point
) , ( z y x r
at the
time
t
.
The energy outflux through
s d
is
.
~ ~
~
~
s d q s d n q
The energy outflux through S is

S
s d q
~
~
,
~
V d q
v

Gouss Theorem. . (i)
The amount of energy per unit
T c p s i l o v .
Where
,
~
r at density p
. heat specific c
The amount of enery in
, ) ( m d T c dV T c V d
and the total energy in

V
V d T c V .
. (ii)
Rate of increase of energy / heat in V is
. ) ( V d T c
t
V d T c
t
V V

.
(iii)
Let
) , , , ( t z y x f
be the internal heat source / sink per unit
. l o v

per unit time. Then the rate of energy input is

V
V d f .
. (iv)
Now, from Conservation of Energy

+

V V V
V d f V d q V d T c
t ~
) (
i.e

'

V
V d f q T c
t
0 ) (
~

.
(v)
. l o v
is non-zero, from (v) we obtain
f q T c
t
+

~
) (
.
(vi)
From Fouriers Law,
,
~
T K q
. (vii)
Where K is the heat conductivity. Inserting (vii) into (vi)
. ) (
2
f T K T c
t
+

If the medium is uniform, them

,
2
c
f
T K
t
T

. (viii)
Where
c
K
k

is the thermal diffusivity.

(b) For the derivation of wave
, .
n
q e
see standard ref / texts, e.g
Zill , Cullen .
. ) , , , (
2 2
2
2
t z y x p u c
t
u
+

.
(ix)
(c) when
t
the 2 above
n
q e
become
0
2
T
or
, 0
2
u
. (x)
Called the Laplace
n
q e
. Here
p f ,
are zero.
(d) For pde (viii) , (ix), we need initial condition (s) IC and Boundary
Conditions (BC), whilst pde (x) requires appropriate BCs.
Separation variables method (Fourier Method)
1. Heat conduction
n
q e
(Diffusion
n
q e
)
PDE :
,
2
2
2
x
u
k
k
u

. 0 , 0 > < < t x

.(i)
BC :
. 0 , 0 ) , ( , 0 ) , ( > t t u t o u
.
(ii)
IC :
. 0 , ) ( ) 0 , ( < < x x f x u
. (iii)
Seek
n
l o s
as
. ) ( ) ( ) , ( t T x X t x u
.
(iv)
Substitute (iv) into PDE (i) ,
T
x d
X d
k
t d
T d
X
2
2
2

i.e.
,
1 1
2
2
2

x d
X d
X t d
T d
T k

2
1
k
t d
T d
T

t k
e A t T
2
) (

.
(v)
and
. 0
2
2
+ X
x d
X d

. (vi)
From (iv) , BC (ii) , we obtain
. 0 ) , ( 0 ) (
. 0 ) ( 0 ) ( ) (
, 0 ) 0 ( 0 ) ( ) 0 (

t x u t T
X t T X
X t T X

. (vii)
n
l o S .
for X is defined by
.
(a) when
. 0 <
Let
. 0 ,
2
>

. 0 ) (
2
2
2
X
x d
X d
vi

. sin
~
cos
~
) ( x h B x h A x X l o S
n
+
n
b
a
l o S Trivial t x u e i
x X
B h B vii
A B A vii C B
. . 0 ) , ( .
0 ) (
0
~
sin
~
0 ) (
0
~
0 .
~
1
~
0 ) (

+

(b) when
. 0

. 0 ) (
2
2

x d
X d
vi . ) ( . B x A x X l o S
n
+

+
0 0 ) ( ,
0 0 0 ) (
A A vii
B B A vii C B
b
a

n
l o S Trivial
t x u e i x X
.
0 ) , ( . 0 ) (

(c) when
. 0 >
Let
. 0 ,
2
>
. 0 ) (
2
2
2
+ X
x d
X d
vi
n
l o S .

x B x A x X sin

cos

) ( +
0

0 ) ( + A B A vii C B
a
. sin

0 ) ( , B vii
b

For . 0

, 0 ) ( B x X 0 sin
,.... 3 , 2 , 1 , sin 0 sin . . n n e i
.

n
n
.
2

,
_

n
Eigen Value. . (viii)
and x
n
x x n X x X
n

sin sin ) ( ) ( Eigen Function

. (ix)
From (v),
t k
n
t k n
n n
n
e A e A t T
2 2 2
) (

. (x)


,
_

x n
t e C t T x X t x U
k n
n n n n
sin ) ( ) ( ) , (
2
) (i PDE
is linear, then linear combination of
) , ( t x U
n
is also a
n
l o S .
.

,
_

x n
e C t x u
n
k n
n
sin ) , (
1
2
.
(xi)
Applying IC (iii),

x n
C x f
n
n
sin ) (
1
..
(xii)
Now multiplying ) (xii q e
n
with

x m
x X
m
sin ) ( and integrating
x t r w .
from
0 x
to
x
we obtain
x d
x m
x f
o

sin ) ( x d
x n
C
x m
n
n
o
1
]
1

1
sin sin

x d
x n x m
C
o
n
n

sin sin
1
From

'

n m
n m o
x d
x n x m
o
,
2
,
sin sin

We obtain
2
sin ) (

m
o
c x d
x m
x f
. .e i

x d
x n
x f c
o
n

sin ) (
2
.
(xiii)
n
l o S to BVP (i) (iii) is given by (xi), with coeff.
n
c
given by (xiii) .
NOTE : (1)
) ( , ) ( ii BC i PDE
are both homogenous.
(2) ) (xii q E
n
is Fourier Sine Series for
) (x f
(3)
n
l o S
for
) (x X
is given by the eigen fct .
) (x X
n
(4) . ) ( ( . 0 . , 0 viii q e n
n
>
(5)
n
l o S
method thus far is formal.
(6)
n
l o S
. sin ) ( sin
2
) , (
1
2
s d
s n
s f
x n
e t x u
o
n
t
k n

,
_

2. Wave Equation .
PDE :
. 0 , 0 ,
2
2
2
2
2
> < <

t x
x
u
c
t
u

. (i)
BC :
. 0 , 0 ) , ( , 0 ) , 0 ( >

t t
x
u
t
x
u

.
(ii)
IC :
< <

x x g x
t
u
x f x u 0 , ) ( ) 0 , ( , ) ( ) 0 , (
.
(iii)
Lets
. ) ( ) ( ) , ( t T x X t x u
. (iv)
Sub. (iv) into (i) , we obtain
T
x d
X d
c
t d
T d
X
2
2
2
2
2

2
2
2
2
2
1 1
. .
x d
X d
X t d
T d
T c
e i
, 0
2
2
2
+ T c
t d
T d

. (v)
. 0 ,
2
2
+ X
x d
X d

.
(vi)
(iv) and
0
) 0 (
) (
x d
X d
ii C B
and
. 0
) (

x d
X d
.
(vii)
Now again there are 3 cases for
n
eq t r w . . . ,
(vi) .
(a) when
0 <
. Let
. 0 ,
2
>
. 0 ) (
2
2
2
X
x d
X d
vi

x h B x h A x X l o S
n
sin cos ) ( : +
n
b
a
l o S Trivial
t x u e i x X
A h A vii
B B A vii C B
.
0 ) , ( . 0 ) (
0 0 sin ) (
0 0 1 0 ) (

+


(b) when
0
0 ) (
2
2
x d
X d
vi
+ B x A x X l o S
n
~ ~
) (

0
~
) ( ,
0
~
) (
A vii
A vii C B
b
a
B x X
~
) (

when B x X
~
) ( , 0 , a constant.
(c) when
. 0 >
Let
. 0 ,
2
>

. 0 ) (
2
2
2
+ X
x d
X d
vi
. sin

cos

) ( x B x A x X l o S
n
+
. 0

0 1

) ( + B B A vii C B
a

, 0 sin . 0 sin

) ( A vii
b
. 0 , 0

A
, 3 , 2 , 1 , sin 0 sin n n

n
n
and
x A x n X x X
n n
cos

) ( ) (
From case (b) , (c) , we can redefine

n
n
, , 3 , 2 , 1 , 0 n
,
x X X
n n
cos
. (viii)
From (v)
. sin cos ) ( 0
2 2
2
2
t c D t c C t T T c
t d
T d
n n n n n n n
n
+ +
( ) . cos sin cos ) , ( x t c D t c C t x U
n n n n n n
+

PDE is linear ,
( )

+
0
. cos ) sin cos ) (
n
n n n n n
x t c D t c C t x u
. (ix)
From
a
iii C I ) (
. cos ) (
0
x C x f
n
n
n

Multiplying by
x x X
m m
cos ) (
and integrating
x t r w . .
from
0 x
to
, x

,
_

0
0
0
cos cos cos ) ( x d x c x x d x x f
n
n n m m

0
0
cos cos x d x x C
n m
n
n

2

m
c

0
. cos ) (
2
dx x x f C
n n
. (x)
From
, ) (
b
iii C I
and
( ) x t c c D t c c c t x
t
u
n
n n n n n n n

0
cos cos sin ) , (

0
. cos ) ( ) (
n
n n n
x c D x g
Multiplying by
x x X
m m
cos ) (
and then integrating
x t r w . .
from
0 x
to
x
we obtain
2
cos ) (
0

c D x d x x g
m m m
e i.

0
cos ) (
2
x d x x g
c
D
n
n
n
. (xi)
Thus the formal
n
l o S
of
) ( ) ( iii i P V B
is given by (ix) , with its cueff.
n n
D C ,
given by
) ( , ) ( xi x
respectively.
3. Laplace Equation.
: E D P , 0
2
2
2
2

y
u
x
u
. 0 , 0 b y a x < < < <
. (i)
: C B
. 0 , ) ( ) ( , 0 ) , (
. 0 , 0 ) , ( , 0 ) , (
b y y f a u y o u
a x b x u o x u

< <
) (
) (
iii
ii


Using Fourier Method, it can be shown that
b
y n
b
x n
h D y x u
n
n

sin sin ) , (
1
where
y d
b
y n
y f
b
a n
b
D
b
n

0
sin ) (
sin
2
y
b
a
x
0 ) , ( b x u
0 ) , ( y o u
0 ) , ( o x u
) ( ) , ( y f y a u
0
2
2
2
2
+
y a
u a
x a
u a
Superposition Principle.
Note :
0
2
2
2
2

y
u
x
u
is homogeneous
All 4 BCs are non-homogeneous
Now let
) , ( , ) , ( . . ) , ( ) , ( ) , ( y x w y x v t s y x w y x v y x u +
each has a pair of homogeneous BCs.
w v u Sub + .
into the Laplace
n
q e
and re-arranging
0
2
2
2
2
2
2
2
2

,
_

,
_

y
w
x a
w
y a
v
x a
v
. .e i 0
2
2
2
2

y
v
x
v
and

0
2
2
2
2
y
w
x
w
.
(i)
Now

+
+
+
+
) ( ) , ( ) , ( ) , (
) ( ) , ( ) , ( ) , (
) ( ) , ( ) , ( ) , (
) ( ) , ( ) , ( ) , (
y G y a w y a v y a u and
y F y o w y o v y o u
x g b x w b x v b x u
x f o x w o x v o x u
.
(ii)
y
b
a
x
) ( ) , ( x f o x u
) ( ) ( y G a u ) ( ) , ( y F y o u
) ( ) , ( x g b x u
0
2
2
2
2

y a
u
x a
u
Then we can choose

) ( ) , ( , ) ( ) , (
0 ) , ( 0 ) , (
y G y a v y F y o v
and
b x v and o x v

0 ) (
0 ) , (
) ( ) , (
) ( ) , (

u a w
y o w
x g b x w
x f o x w
Thus now we can solve for
w v ,
using Fourier Method And hence
w v u +
is obtained.
Alternatively we can decide to put

) ( ) , ( 0 ) , (
) ( ) , ( 0 ) , (
x g b x v and b x w
x f o x v and o x w
0 ) , ( , ) ( ) , (
0 ) , ( ) ( ) , (

y a v y G y a w
y o v y F y o w
And solve for
w v ,
with these BCs.
Ultimately, the
ns
l o s
are the same.
Non Homogenous BVP
Non homogenous PDE and / or non homogenous BCs.
For illustrative purpose only, we consider BVP consisting of ;
PDE :
0 , 0 , ) , (
2
2
2
2
2
> < < +

t x t x P
x
u
c
t
u

. (1)
BC : 0 , ) (
~
) , ( , ) ( ,
~
) , 0 (
2
> t t f t u t f t u . (2)
IC :

) ( ) 0 , (
) ( ) 0 , (
x g x
t
u
x f x u
< < x 0
. (3)
NOTE : The ideas and solution methods can be carried or to other linear non
homogenous BVP.
Here ) ( , ) (
~
, ) (
~
2 1
x f t f t f and
) (x g
are given.
) , ( t x P
is the generating term (source term) or annihilating term
(sink term) .
1. Homogenous PDE, non homogenous BCs.
PDE :
. 0 , 0 ,
2
2
2
2
2
> < <

t x
x
u
c
t
u

. (4)
BC :

b t u
a t u
) , (
) , 0 (

. 0 > t

b a ,
. (5)
I.C :
. ) ( ) 0 , ( , ) ( ) 0 , ( x g x
t
u
x f x u

. (6)
Lets
. ) , ( ) ( ) , ( t x w x v t x u +
. (7)
Sub. (7) into (4) , then
.
2
2
2
2
2
2
2

,
_

x
w
x d
v d
c
t
w
choose
0
2
2

x d
v d
.
From (5a) ;
a t w v + ) , 0 ( ) 0 (
. Select
. 0 ) , 0 ( ) 0 ( t w t s a v
(5b) ;
. ) , ( ) ( b t w v +
Take
. 0 ) , ( ) ( t w t s b v
From (6a);
) ( ) ( ) 0 , ( ) ( ) 0 , ( ) ( x v x f x w x f x w x v +
(6b) ;
. ) ( ) 0 , ( x g x
t
w

Then solve
0
2
2

x d
v d
subject to
b v a v ) ( , ) 0 (
a x
a b
x v e i +

) ( . .
and homogenous BVP
PDE :
0 , 0 ,
2
2
2
2
2
> < <

t x
x
w
c
t
w

BC :
, 0 ) , 0 ( t w , 0 ) , ( t w 0 > t
IC :
, ) ( ) ( ) 0 , ( x v x f x w
. ) ( ) 0 , ( x g x
t
w

From these known

, . ,
ns
l o s w v
we obtain
. ) , ( t x u
2. Non homogenous PDE , non homogenous BCs
PDE :
, ) (
2
2
2
2
2
x P
x
u
c
t
u
+

0 , 0 > < < t x

. (8)
BC :

b t u
a t u
) , (
) , 0 (

. 0 > t b a ,
. . (9)
IC :
, ) ( ) 0 , ( x f x u , ) ( ) 0 , ( x g x
t
u

< < x 0

. (10)
Again lets
. ) , ( ) ( ) , ( t x w x v t x u +

. (11)
From
, ) 8 ( , ) ( II . ) (
2
2
2
2
2
2
2
x P
x
w
x d
v d
c
t
w
+

,
_

Choose
0 ) (
2
2
2
+ x P
x d
v d
c

2
2
2
2
2
x
w
c
t
w
t s
From (9a) ;
. ) , 0 ( ) 0 ( a t w v +
Put
. 0 ) , 0 ( ) 0 ( t w t s a v

(9b) ;
. ) , ( ) ( b t w v +
Impose
. 0 ) , ( ) ( t w t s b v
From (10a) ;
) ( ) ( ) 0 , ( ) ( ) 0 , ( ) ( x v x f x w x f x w x v +
(10b) ;
. ) ( ) 0 , ( x g x
t
w

Then Solve
(i)
2 2
2
) (
c
x P
x d
v d

subject to
b v a v ) ( , ) 0 (
and
(ii) PDE :
0 , 0 ,
2
2
2
2
> < <

t x
x
w
t
w

BC :
0 ) , ( ) , 0 ( t w t w
IC :
) ( ) 0 , ( , ) ( ) ( ) 0 , ( x g x
t
w
x v x f x w

To obtain

n
l o S t x u ) , (
3. Non homogenous PDE, homogenous BCs (Standard Form)
Variation of parameters / Eigenfunction expansion .
PDE :
0 , 0 , ) , (
2
2
2
2
2
> < < +

t x t x P
x
u
c
t
u

.
(12)
BC : >

0
0 ) , (
0 ) , 0 (
t
t u
t u

. (13)
IC :
, ) ( ) 0 , ( x f x u < <

x x g x
t
u
0 , ) ( ) 0 , (
. (14)
When
, 0 ) , ( t x P
the corresponding BVP
.
n
l o S
is

1
, ) ( ) ( ) , (
n
n n
x X t T t x u
. (15)
with
t c B t c A t T
n n n n n
sin cos ) ( +
x x X
n n
sin ) (
. . . . , 3 , 2 , 1 ,

n
n
n

Now, with
, 0 ) , ( t x P
generalised
t s l o s
n
) 15 ( .

1
) ( ) ( ) , (
n
n n
x X t H t x u
. (16)
Satisfies (12) (14) . Substituting (16) into (12) , we obtain

'

+
1
2 2
2
2
. ) ( ) (
n
n n n
n
t x P x X H c
t d
H d

But
, ) ( ) ( ) , (
1
x X t P t x P
n
n n

. (17)
with

0
0
2
0
. ) ( ) , (
1
) (
) ( ) , (
) ( x d x X t x P
N
x d x X
x d x X t x P
t P
n
m
n
n
n
. (18)
Hence

'

+
1 1
2 2
2
2
) ( ) ( ) (
n n
n n n n n
n
x X t P x X H c
t d
H d

+ 1 , ) ( . .
2 2
2
2
n t P H c
t d
H d
e i
n n n
n
. (19)
For the IC of (19) ;

1
) ( ) ( ) 0 ( ) 16 ( , ) 14 (
n
n n
x f x X H a
. ) ( ) (
1
) 0 ( . .
0
x d x X x f
N
H e i
n
m
n

. (20)
) ( ) ( ) 0 ( ) 16 ( , ) 14 (
1
x g x X
t d
H d
b
n
n
n

0
. ) ( ) (
1
) 0 ( x d x X x g
N t d
H d
e i
n
m
n
.
(21)
) 21 ( ) 19 (
n
q E will give the unique
n
l o s
of
, ) (t H
n
with
. ) ( ) ( ) ( t H t H t H
p n c n n
+
. (22)
,
n
c n
l o s ary complement H
and
.
n
p n
l o s particular H
Thus the ) , ( t x u l o s
n
is now known , with
) (t H
n
given by (22) .
Note : (1) ) 17 ( , ) 16 (
n
q E examples of eigenfunction expansion
(2) Eigenfunction
) (x X
n
satisfies BCs (13)
(3)




0
2
2
) ( x d x x N
n m
(4)
. ) , ( t x P P
4. Non homogenous PDE, non homogenous BCs.
PDE :
, ) , (
2
2
2
2
2
t x P
x
u
c
t
u
+

. 0 , 0 > < < t x
.
(23)
BC :
, 0
) , (
) , 0 (
>

t
b t u
a t u

b a ,
. .
(24)
IC :
< <

x
x g x
t
u
x f x u
0
) ( ) 0 , (
) ( ) 0 , (
. (25)
Assume
n
l o s . . ) , ( ) ( ) , ( t x w x v t x u +
. (26)
Sub . (26) into (23)
. ) , (
2
2
2
2
2
2
2
t x P
x
w
x d
v d
c
t
w
+

,
_

Impose
) , ( 0
2
2
2
2
2
2
2
t x P
x a
w
c
t
w
t s
x d
v d
+

.
Then
( 24 a ) ;
. 0 ) , 0 ( ) 0 ( + t w v
Choose
a v ) 0 (
0 ) , 0 ( t w t s
( 24 b ) ;
b t w v + ) , ( ) (
Choose
0 ) , ( ) ( t w t s b v
( 25 a ) ;
) ( ) ( ) 0 , ( ) ( ) 0 , ( ) ( x v x f x w x f x w x v +
( 25 b ) ;
. ) ( ) 0 , ( x g x
t
w

The
n
l o s .
to
) , ( t x u
will be given by solving
(i)
0
2
2

x d
v d
subject to
, ) ( , ) 0 ( b v a v
and
(ii) PDE :
. ) , (
2
2
2
2
2
t x P
x
w
c
t
w
+

BC :
0 ) , ( ) , 0 ( t w t w
IC :
, ) ( ) ( ) 0 , ( x v x f x w

) ( ) 0 , ( x g t
t
w

.
5. Non homogenous PDE, non homogenous BCs.
PDE :
. ) , (
2
2
2
2
2
t x P
x
u
c
t
u
+

. 0 , 0 > < < t x

. (27)
BC :
0
) (
~
) , (
) (
~
) , 0 (
2
1
>

t
t f t u
t f t u

. (28)
IC :
< <

x
x g x
t
u
x f x u
0
) ( ) 0 , (
) ( ) 0 , (
.
(29)
Lets
. ) , ( ) , ( ) , ( t x w t x v t x u +
.
(30)
Sub. (30) into (27), and re-arranging
) , (
2
2
2
2
2
2
2
2
2
2
t x P
t x
v
c
x
w
c
t
w
+

,
_

. (31)
From . 0 ) , 0 ( ) (
~
) , 0 ( , ) 30 ( , ) 28 (
1
t w t s t f t v choose a .
(32)
0 ) , ( ) (
~
) , ( , ) 30 ( , ) 28 (
2
t w t s t f t v choose b .
(33)
Now put
. ) ( 1 ) ( ) , (
1 0
m n
x
t A
x
t A t x v

'

'

. (34)
with
m n ,
positive integer, chosen s.t. it has the smallest tve to make the
BCs for
) , ( t x w
homogenous (though this will possibly
n
eq
for
) , ( t x w
non homogenous ) .
From
) (
~
) ( ) 34 ( , ) 32 (
1 0
t f t A
) (
~
) ( ) 34 ( , ) 33 (
2 1
t f t A

,
_

+
,
_

x
t f
x
t f t x v ) (
~
1 ) (
~
) , (
2 1
, . (35)
with
1 n m
Hence solve, from (31) , (35) ,
PDE :

'

'

x
t f
x
t f t x P
x
w
c
t
w
) ( "
~
1 ) ( "
~
) , (
2 1
2
2
2
2
2

. (36)
Subject to
BC :
0 , 0 ) , ( ) , 0 ( > t t w t w

. (37)

IC :

'

'

x
f
x
f x f x v x f x w ) 0 (
~
1 ) 0 (
~
) ( ) ( ) ( ) 0 , (
2 1


) 0 , ( ) ( ) 0 , ( x
t
v
x g x
t
w

. (38)

'

'

x
f
x
f x g ) 0 (
~
1 ) 0 (
~
) (
2 1
Thus with
) , ( t x w
known from solving ( 36 ) ( 39 ) and
) , ( t x v
from
( 35 ) , the
n
l o s
to
) , ( t x u
will be given by
. ) , ( ) , ( ) , ( t x w t x v t x u +
Note: (1) the BC (28) here is specification on
) , ( t x u
at
. , 0 x x
If these
are replaced with
) , ( t x
x
u

at
0 x
and / or
. 1 , , n m x
Example 1 Non homogenous BC.
PDE :
,
2
2
2
x
u
k
t
u

0 , 0 > < < t x

. . (1)
BC :
>
0 0
, 0 , ) , ( , 0 ) , 0 ( t t u t u
. .
(2)
IC :
, 0 ) 0 , ( x u . 0 < < x
.
(3)
Let
) , ( ) ( ) , ( t x w x v t x u +
. . (4)
Sub. (4) into (1) and (2), we obtain
(i)
, ) ( , 0 ) 0 ( , 0
0
2
2
v v
x d
v d
and . (5)
(ii)
< <

x
x
w
k
t
w
0 ,
2
2
2
. (6)
0 ) , ( ) , 0 ( t w t w
.
(7)
) ( ) 0 , ( x v x w
.
(8)
The
n
l o s
to (5) is x x v

0
) (

.
(9)
The
n
l o s
to BVP (6) (8) :
. . . . , 3 , 2 , 1 , , sin ) (

n
n x n
x X
n n

x d
x n
x v c
n





0
sin ) (
2

'

n
n
x d
x n
x ) 1 (
2
sin
2
2
2
0
0
2
0

. ) 1 (
2
.
0 n
n
n
C e i

( )

x n
e
n
t x w
n
t
n
k n

sin
) 1 ( 2
) , (
1
0
2

. (10)

'

,
_

1
0
sin
) 1 ( 2
) , ( .
2
n
t
k n
x n
e
n
x t x u e i

. (11)
Example 2. Internally generated heat.
PDE :
. ) (
2
2
2
t P
x
u
k
t
u
+

. 0 , 0 > < < t x

. (1)
BC :
. 0 ) , ( , 0 ) , 0 ( t u t u . 0 > t
. (2)
IC :
) ( ) 0 , ( x f x u
. (3)
Seek the
.
n
l o s
of BVP (1) (3) as

1
, ) ( ) ( ) , (
n
n n
x X t A t x u
. (4)
where
x x X
n n
sin ) (
,

n
n
. (5)
is the corresponding homogenous BVP. Also expand
) (t P
as

1
) ( ) (
n
n n
X t P t P
. (6)

x d x X t P t P e i
n n
) ( ) (
2
) ( .
0

{ } { } . ) ( ) 1 ( 1
2
) 1 ( 1 ) (
2
t P
n
t P
n n

. (7)
Sub.
) 6 ( , ) 4 (
into
, ) 1 (
and using
,
2
2
2
n n
n
X
x d
X d

we obtain
[ ]

'

+
1
2 2
0 ) ( ) ( ) 1 ( 1
2
n
n
n
n n
n
x X t P
n
A k
t d
A d

. (8)
0

. . . . , 3 , 2 , 1 , ) ( n x X
n

is linearly independent of each other

(8)
[ ] . ) ( ) ( ) 1 ( 1
2
2 2
t P t P
n
A k
t d
A d
n
n
n n
n

+
{ } ) (
2 2 2 2
t P e A e
t d
d
t k
n n
t k
n n

+ +
t
k
n
t k t k t k
n n
d P e e e e c t A e i
n n n n
0
) ( ) (
2 2 2 2 2 2 2 2


. (9)
From (4) and IC (3), we obtain

1
) ( ) 0 ( ) (
n
n n
x X A x f

0
, ) ( ) (
2
) 0 ( y a s x d x X x f A e i
n n n

. (10)
. ) 10 ( , ) 9 (
n n
c
Hence (9) becomes
[ ]

+
t
t k n t k
n n
d P e
n
e t A
n n
0
) (
. ) ( ) 1 ( 1
2
) (
2 2 2 2



. (11)
Then sub. (11) into (4), we obtain the BVP
n
l o s .
as

'

1
0
) ( ) (
2
sin ) , (
2 2
n
n n
t k
s d s X s f x e t x u
n

+
t
t k
n
T d P e
n
x n
n
0
) (
1
. ) (
) 1 2 (
/ ) 1 2 ( sin 4
2 2

. (12)
Note : (1) When
0 ) ( t P
n
in (12), we recover the homogenous BVP.
(See
pg. 3/9 previous notes ).
(2) The 2
nd
term in (12) is the generating heat effect.
Fourier Series
1. Motivation
ODE :
< < + x y
x d
y d
, 0
2
2
2
. (1)
BC :
. ) ( ) ( y y
.
(2)

.
) (
) (
x d
y d
x d
y d

. (3)
n
l o S .
to (1) is
, sin cos ) ( x B x A x y + B A ,
. .
(4)
x B x A x
x d
y d
cos sin ) ( +
. .
(5)
Sub. (4) into BC (2),
sin cos ) ( sin ) ( cos B A B A + +
sin cos sin cos B A B A +
0 sin 2 B
.
(6)
Sub. (5) into BC (3) ,
cos sin cos sin B A B A + +

0 sin 2 A
.
(7)
. . . . , 2 , 1 , 0 , sin 0 sin ) 7 ( , ) 6 ( n n
. . . . , 2 , 1 , 0 ,

n
n
e i
n

.
(8)
and
x y y and x y y
n n n n
sin , cos
2 2 1 1

. .
(9)

+ +
1
0
. sin cos
2
) (
n
n n n n
x b x a
a
x y
.
(10)
) 10 ( .
n
q E Fourier Series .
Academy Paris
ob Conduction Heat
Fourier
Joseph
Pr
1811 , 1807

2. Basics
. sin cos
2
1
0

x m
b
x m
a
a
S F
m
m
m

+

.
(11)
Important properties of trigonometric functions :
(i) Periodic
period T x f T x f + , ) ( ) (
For (11) ,
.
2 2
m
m
T

Note : If
f
and
g
periodic fcts with period T then
g f g f , ) ( +

are also periodic fcts. With period T .
(ii) Orthogonality.

'

. ,
, , 0
cos cos
n m
n m
x d
x n x m

.
(12)

'

. 12 ,
, , 0
sin sin
m
n m
x d
x n x m

.
(13)
and , 0 sin cos

x d
x n x m

for all
. , n m
. (14)
Note :
. . . . , 4 , 3 , 2 , 1 , n m
In
, 0 , ) 12 ( n m when

2 ) 0 ( cos ) 0 ( cos x d
, 0 , ) 13 ( n m when

, 0 ) 0 ( sin ) 0 ( sin x d
, 0 , ) 14 ( n m when

. 0 ) 0 ( sin ) 0 ( cos x d
3. The Euler. Fourier Formulas .
Lets
) (x f
be FS, i.e.

,
_

+
1
0
sin cos
2
) (
m
m m
x m
b
x m
a
a
x f

.
(15)
Integrate (15) w.r.t
x
from
x
to
, x
we obtain

x d x f a ) (
1
0
. (16)
Multiply (15) with

x n
cos , and integrate the resulting , .
n
q e
taking into account that it is permissable to interchange the integral and
summation, and also using (12) (14) , we obtain
. cos ) (
1
x d
x n
x f a
n

.
(17)
Multiply (15) with , sin

x n
and repeat the above procedure, gives
x d
x n
x f b
n

sin ) (
1
.
(18)
(16) (18) are Euler Fourier Formulas .
Thus given an orbitrary fct.
) (x f
, and n
a
and b n can be calculated from
(16) (18) , then
) (x f
can be represented by FS (15) .
Note : If
) (x p
is a polinomial of degree
m
, and
) (x f
ctn. , then

+
+ + + + , ) 1 ( . . . . ) ( ) (
1
) (
3 2 1 m
m m
F p F p F p F p x d x f x p
where

etc x d F F x d f F . ,
1 2 1
4. Application.
Example 1
Find the FS of

'

< <
<

. 0 ,
0 , 0
) (

x x
x
x f
`
(16)
{ }


+
0
0
0
0
1
) (
1

x d x x d x d x f a
2 2
1
0
2

'

1
]
1

x
(17) x d
x n
x x d
x n
x f a
n

0
cos
1
cos ) (
1

0
2
cos
1
sin
1
1
1
1
1
1
]
1

,
_

,
_

,
_

n
x n
n
x n
x

y

o 2 3 4 x 5
( )
[ ] 1 ) 1 ( . .
2

n
n
n
a e i

(18) x d
x n
x x d
x n
x f b
n

0
sin
1
sin ) (
1

0
2
0
sin
. 1
cos
1
1
1
1
1
1
]
1

,
_

,
_

,
_

n
x n
n
x n
x
1
) 1 (
+

n
n
n
b e i

[ ] , sin ) 1 ( cos 1 ) 1 (
) (
4
) (
1
1
2

,
_

+
n
n n
x n
n
x n
n
x f

for
< < x
. The extension is also shown in the above figure.
Example 2
Find the FS of

'

<

x x
x x
x f
0 ,
0 ,
) (
) ( ) 2 ( x f x f +
(16)

{ }

'

1
]
1

+
1
]
1

0
2
0
2
0
0
0
2 2
1 1
) (
1 x x
x d x x d x x d x f a

'

+
2 2
1
2 2
4 3 2 2 3 4

y
(17)
x d
x n
x x d
x n
x a
n


+

0
0
cos
1
cos
1

'

1
1
1
1
1
]
1

,
_

,
_

,
_

+
1
1
1
1
1
]
1

,
_

,
_

,
_

0
2
0
2
cos
1
0
sin cos
1
0
sin
1
n
x n
n
x n
x
n
x n
n
x n
x
( ) ( ) ( ) [ ] 1 ) 1 (
) (
2
1 ) 1 ( 1 1
1
2
2 2

'

,
_

+ +

,
_

n n
n
n
n n

(18)

'

+

0
0
sin sin
1

x d
x n
x x d
x n
x b
n

'

1
1
1
1
1
]
1

,
_

,
_

,
_

+
1
1
1
1
1
]
1

,
_

,
_

,
_

2
0
2
0
sin
1
cos
0
sin
1
cos
1

n
x n
n
x n
x
n
x n
n
x n
x
0 ) 1 ( ( ) 1 (
1

'

,
_

+ +

n n
n n

, cos
1 ) 1 ( 2
2
) (
1
2 2

n
n
x n
n
x f

[ ]

< <

1
2 2
,
) 1 2 (
/ ) 1 2 ( cos 4
2
) (
n
x
n
x n
x f e i

Example 3
Given that
< < x x x f , ) (
and
, ) ( ) 2 ( x f x f +

find the FS of
. ) (x f
(16)

0
1
0
x d x a
(17) 0 cos
1

x d
x n
x a
n

(18)

'

1
1
1
1
1
]
1

,
_

,
_

,
_

2
0
sin
1
cos
1
sin
1
n
x n
n
x n
x x d
x n
x b
n
( )

'

n n
n
cos cos
1

1
) 1 (
2
+

n
n
n
b e i

( )
, sin
1 2
) (
1
1

n
n
x n
n
x f

< < x
Note : In example 2 , 3 , the extension out of
< < x
can be effected
using

+

T T
x d x f x d x f

0
, ) ( ) (

when
) ( ) ( x f T x f +
Example 4
Find the FS of

'

< <
< <

. 0 , 1
, 0 , 1
) (
x
x
x f
(16) , (17)
. . . . , 2 , 1 , 0 , 0 n a
n
(18)

0
sin
2
sin ) (
1
x d x n x d
x n
x f b
n
) 1 cos (
2 cos 2
0

1
]
1

n
n n
x n
[ ]
n
n
n
b ) 1 ( 1
2

, sin
) 1 ( 1 2
) (
1
x n
n
x f
n
n

1
]
1

y
x
-1
1

1
) 1 2 (
) 1 2 ( sin 4
) (
n
n
x n
x f e i
. ( * )
The graf of
11 2 1
, , f f f
and 55
f
are shown on page 8. As
n
increases
) (x f f
n

; at the pt. of discontinuities the overshoot remains
0
0
9 ( of
jump height, even as
n
). Called Gibbs phenomenon .
.
) 1 2 (
) 1 2 ( sin 4
) (
1

n
n
x n
x f
5. Conditions for the existence of FS of
) (x f
Definition : A
) (x f t c f
is said to be piecewise continous on an
interval
x
if the interval can be partitioned by a finite
number of points
< < < <
n
x x x x . . .
2 1 0 such that
(i)
) (x f
is continous on each
, ;
1
x x x
i
< <
and
(ii)
) (x f
approach a finite limits as the end points of each
, ;
1
x x x
i
< <
are approached from within.
Theorem Suppose
f
and
x d
f d
are piecewise ctn. On the interval
x
. Suppose further that
f
is defined outside
x
such that
) ( ) ( x f T x f +
. Then
f
has a FS
, sin cos
2
) (
1
0

,
_

+
n
n n
x n
b
x n
a
a
x f

with n
a
and n
b
given by Euler-Fourier Formulas (16) (18).
The FS converges to
) (x f
at all points where
f
is ctn and to
1
]
1

+ +
2
) ( ) ( x f x f
at all points where
f
is disctn.
6. Even and odd fcts.
(i)
) (x f
is even if
) ( ) ( x f x f
(ii)
) (x f
is odd if
) ( ) ( x f x f
x n x x g e
n
sin , ,
1 2 +

NOTES
(i) The sum (difference) and product (quotient) of two even fcts are even.
(ii) The sum (diff) of two odd fcts is odd ; the product (quotient) of two odd
fcts is even.
(iii) The sum (diff) of an odd tct and an even fct is neither even nor odd ;
the product (quotient) of odd and even fcts. is odd.
(iv) If
) (x f
is even fct , then for
< < x
`

x d x f x d x f ) ( 2 ) (
and
, cos
2
) (
1
0

+
n
n
x n
a
a
x f

with
. . . , 2 , 1 , 0 , cos ) (
2
0

n x d
x n
x f a
n


. . . , 3 , 2 , 1 , 0 n b
n
( Refer to Example 2 pg . 5 ) Fourier Cosine Series
(v) If
) (x f
is odd on the interval
, < < x
then

0 ) ( x d x f
and
, sin ) (
1

n
n
x n
b x f

with
. . . , 2 , 1 , 0 , 0 n a
n
. . . , 3 , 2 , 1 . sin ) (
2
0

n x d
x n
x f b
n


( Refer to Example 3 , pg. 6 ) Fourier Sine Series
y
x


7. Complex form of the FS .
Substituting
( )
x n i x n i
e e x n

+
2
1
cos
and
( )
x n i x n i
e e
i
x n


2
1
sin
into FS

,
_

+
1
0
sin cos
2
) (
n
n n
x n
b
x n
a
a
x f

we obtain

x n
n
n
x n i
i
n
e
n b a
e
n b a a
x f

,
_

+
+
,
_

+

1
1
0
2
1
2
1
2
) (
/ 1
1
/
0
x n
n
n
x n i
n
e c e c c

+ +

e i , ) (
/ 1

o b n
x n
n
e c x f

< < x
with
. ) (
2
1
/ 1
x d e x f c
x n
n

: b n
2
0
0
a
C
( )
n n n
b a C 1
2
1

( )
n n n
b a C 1
2
1
+

Boundary value problems in Cylindrical and Spherical

Coordinates
BVP of interest :
,
2
u k
t
u

Diffusion
. q E
. (i)
,
2 2
2
2
u c
t
u

Wave
. q E
. (ii)
. 0
2
u
Laplace
n
q E .
. (iii)
Plus the appropriate BCs and ICs.
(a) Cartesian Coordinate.

.
2
2
2
2
2
2
2
z
u
y
u
x
u
u

For example, wave
n
q e

in
[ ] 3
Is

,
_

2
2
2
2
2
2
2
2
z
u
y
u
x
u
c
t
u
.
(b) Cylindrical Coordinate .
z z y x , sin , cos
.
1 1
2
2
2
2
2
2
z
u u u
u

,
_

Z
y
0
x
) , , ( z y x P
z
z

y p

0
x
) , , ( z P
(c) Spherical Coordinate .
sin sin , cos sin r y r x
. cos r z

,
_

,
_

u
r r
u
r
r r
u sin
sin
1 1
2
2
2
.
sin
1
2
2
2 2

+
u
r
A. Gamma
) (x
and Beta
) , ( y x
functions

>
0
1
.
. 0 , ) ( : x t d t e x f e D
x t n
. (i)

> >

1
0
1 1
. 0 , 0 , ) 1 ( ) , ( y x t d t t y x
y x

.
(ii)
From (i) , and also by integration by ports, we can show that
, 1 ) 1 (
0 , ) ( ) 1 ( > + x x x x
and . . . . , 3 , 2 , 1 , 0 , ) 1 (
1
+ n n n
Putting
2
u t
in (i) , we obtain

0
1 2
. 2 ) (
2
u d u e x
x u
Then

,
_

2
1
.
) , , ( r P
y
p

z
x
0
Letting

2
cos t
in (ii)
.
) (
) ( ) (
) , (
y x
y x
y x
+


.
(iii)
Hence can use (iii) for the evaluation of
) , ( y x
B. Cauchy Euler
n
q E
.
0 ) (
2
+ + x y c y x b y x a .(i)
Here
c b a , ,
and
. 0 > x
If
, 0 < x
let
. 3 x
Letting


r x y ,
, . (ii)
and substituting into (i), we get
0 ) (
2
+ + c r a b r a Indicial
n
q e
.
. (iii)
. ,
2 1

) ( ) ( ) (
2 1
x y B x y A x y +
.
2 1

+ x B x A . (iv)
If
,
2 1

1
) (
1

x x y
0 , ) (
1
2
>

x x n I x x y ( reduction of order )
. ) (
1 1
x n I x B x A x y

+ . (v)
Note : (i) For any
x
that makes the coeff. Of
y
non-zero , is called a
regular pt.
(ii) If it is zero, it is called singular pt.
0 x
is a singular pt. of (i) .
Applying separation variables method to (i) (iii) of page 1/6 , one of the
resulting ODE is in the form of
0 ) ( ) ( ) ( ) ( + + x y x R y x Q y x P
. (vi)
In theoretical studies, (vi) is generally transformed to
, 0 ) ( ) ( ) (
2
+ + x y x r y x q x y x . (vii)
Where
) (x q
and
) (x r
are expanded in power series , i.e

0
) (
m
m
m
x q x q
.
(viii)
, ) (
0

m
m
m
x x
. (ix)

m m
q , . Here
0
0
x
is chosen.
Example
1.
0 2 + + y y y x
0
2
1
2
1
2
y x y x y x + +
x x x q
2
1
) ( ,
2
1
) (
. . . . , 3 2 , 1 , 0 ; ,
2
1
0
i q q e i
. . . , 4 , 3 , 2 , 0 ,
2
1
, 0
1 0
i
i
2.
0 ) 1 ( ) 1 ( + y y x y x x
0
1
2

+ y
x
x
y x y x

<
0
1
1 , ) 1 ( ) ( , 1 ) (
m
m
x x x x x x x q e i

+

0
1
m
m
x
Note :
1 , 0 x
singular pt. of
;
.
n
q e
the rest is regular pt.
3. . 0 ) 3 ( 3
2
+ y x y x y x
. 3 ) ( , 3 ) ( x x r x q
4. . 0 3 ) 1 (
2
+ + y x y x y x
1 2 3 1 2 2
) 1 ( ) ( , ) 1 ( 3 ) (

x x x r x x x q
) ( , ) ( x r x q
can be expanded as a power series ( use Binomial
Theorem ) , converge for
1 < x
.
5. . 0 ) (
2 2 2
+ + y n x y x y x Bessel
;
.
n
q e
. ) ( , 1 ) (
2 2
n x x r x q
Power Series Solution to ODE
Consider
) ( ) ( ) ( ) ( x G y x R y x Q y x P + +
.
(i)
In
< < x
, where
R Q P , ,
and G ctn. For any
0
x
in
< < x

such that
(a)
, 0 ) (
0
x P
then
0
x
is defined as an ordinary pt. of (i) , and
(b)
0 ) (
0
x P
then
0
x
is called singular pt. of (i) .
Example :
El Bessel
0 ) (
2 2 2
.
+ + y D x y x y x q E
n
. . (ii)
0 x
is a singular pt.
0 x
is an ordinary pt.
E2 Legendre
. 0 ) 1 ( 2 ) 1 (
2
.
+ + y y x y x q E
n

.
(iii)
1 + x
singular pt.
1 + x
ordinary pt.
E3 Airy
. ,
.
< < x y x y q E
n
. (iv)
Every pt. in
< < x
is ordinary pt.
Theorem 1
If 0
x
is an ordinary pt. of
, 0 ) ( ) ( ) ( + + y x R y x Q y x P
. (V)
that is
R Q p /
and
p R q /
analytic at
,
0
x
then the general
n
l o s
.
to (v) is given by

+
0
2 1 1 0 0
, ) ( ) ( ) ( ) (
n
n
n
x y a x y a x x a x y
. (vi)
Where
1 0
, a a
are arbitrary constant ,
2 1
, y y
are independent
power series
ns
l o s
.
. Which are analytic at
.
0
x
The radius of
convergence of
2 1
, y y
is at least as big as the smaller bet the radius
of convergence of
) (x p
and
. ) (x q
Now lets say
0
x
is a singular pt f (i) ,
0 ) (
0
x P e i
. Lets find a
n
l o s
.
of (i) close to
.
0
x

n
Def :
0
x
is a regular singular pt of ODE (i) if
) (
) (
) (
0
x P
x Q
x x
and
) (
) (
) (
2
0
x P
x R
x x
analytic at
e i x ,
0
if both terms can be expanded as a
Taylor series. If P , Q , R are polinomials, these conditions
can be replaced by the existence and finiteness of
.
) (
) (
) (
lim
) (
) (
) (
lim
2
0
0
0
0
x P
x R
x x
x x
and
x P
x Q
x x
x x

Otherwise
0
x
is irregular singular pt. of (i)
Example
. 0 ) 1 ( 2 ) 1 ( ) 1 (
2 2 3 2
+ + + y x y x y x x . (vii)
Singular pts
. 1 , 0
0
+ x
0
0
x

. 0
) 1 (
) 1 ( 2
) 0 (
0
lim
, 1
) 1 (
) 1 (
) 0 (
0
lim
3 2
2
3 2
2 2

'

'

x x
x
x
x x x
x
x
x
0 x
is a regular singular pt. of (vii)
1
0
x

2
1
) 1 ( ) 1 (
1
1
lim
) 1 (
) 1 (
) 1 (
1
lim
3 2
2 2

+
+

'

+
x x x
x
x x x
x
x
x

1
0
x
is a regular singular pt. of ODE (vii)
1
0
x

2
1
) 1 ( ) 1 (
) 1 (
1
lim
) 1 (
) 1 (
) 1 (
1
lim
3 2
2 2

'

x x x
x
x x x
x
x
x


+
+

'

3 3
2
3 2
2
) 1 ( ) 1 (
) 1 ( ) 1 ( 2
1
lim
) 1 (
) 1 ( 2
) 1 (
1
lim
x x x
x x
x x x
x
x
x

1
0
x
is an irregular singular pt. of ODE (vii) .
From (v)
0 + + y
P
R
y
P
Q
y
e i

0 ) ( ) ( + + y x q y x p y
,
. ) ( / ) ( ) ( , ) ( / ) ( ) ( x P x R x q x P x Q x p
Hence { } { } . 0 ) ( ) (
2 2
+ + y x q x y x p x x y x . (viii)
Theorem 2
Let
0
0
x
be the regular singular pt. of
[ ] [ ] , 0 ) ( ) (
2 2
+ + y x q x y x p x x y x . (viii)
e i
at
0
0
x

0 0
2
) ( , ) (
n n
n
n
x
x q x q x n P x p x
n
Converge for
, < x
where

is the smaller rad. Of convergent between

) (x p x
and . ) (
2
x q x Let
1

and
2

be the roots of the indicial

n
q e
.

0
) (
n
r n
n
x a x y
, 0 ) 1 ( ) (
0 0
+ + q r p r r r F
.
(viii)
and
2 1
r r
if

2 1
, r r
. Then in the interval
0 < < x
or
, 0 < < x
the
n
l o s
.
of ODE (viii) is given by ;
(a) If
0
2 1
r r
or integer
, N

'

1
1 1
) ( 1 ) (
1
n
n
n
r
x r a x x y
. (ix)

'

1
2 2
) ( 1 ) (
2
n
n
n
r
x r a x x y
(b) If
,
2 1
r r

'

1
1 1
) ( 1 ) (
1
n
n
n
r
x r a x x y
. (x)

+
1
1 1 2
) ( ) ( ) (
2
n
n
n
r
x r b x x In x y x y
(c) If
,
2 1
N r r
positive integer

'

n
n
n
r
x r a x x y ) ( 1 ) (
1
1
1
1
. (xi)

'

+ +

1
2 1 2
) ( 1 ) ( ) (
2
n
n
n
r
x r c x x In x y a x y
Where
) ( , ) ( , ) ( , ) (
2 1 2 1
r c r b r a r a
n n n n
and constant a can be
determined by sub.
) ( / ) ( / ) (
.
i x x x i l o s
n
into ODE (viii) . Infact
. . . , 3 , 2 , 1 , ) ( ) ( ) ( , ) (
2
2 2 1 1

1
]
1

1
]
1

n r a r r
r d
d
r c r r
r d
a d
r b
r r
n n
n
n
and
. ) ( ) (
lim
2
2
r a r r
r r
a
N

Bessels equation, Bessel functions and BVP in Cylindrical Coordinate.

(A). . 0 0 ) (
2 2 2
+ + v y x y x y x .
(1)
Bessel
n
q e
.
order

Seek
n
l o s near regular singular pt.
e i x . , 0
0

let

0
) (
n
r n
n
r
x a x x y
. . (2)
Substituting (2) into (1) results
{ } { }
1
1
2
0
2
) 1 ( ) 1 ( ) 1 (
+
+ + + + +
r r
x a r r r x a r r r
[ ] { }

+ + + +
2
2
2
. 0 ) ( ) (
n
r n
n n
x a a n r n r
.
(3)
From coeff. Of
,
r
x
we obtain
0 ) (
0
2 2
a r
. 0 , 2 ,
0 1
+ a r .
(4)
Note : when
, 0 > r , 0 0
0
x as x a
r
and

, 0 < r . 0
0
x as x a
r
(a) Regular
n
l o s
.

) 0 ( > r
From coeff. Of
1 + r
x in (3) , we see that
0 ) 1 2 (
1
+ a
. 0 ) 1 2 ( , 0
1
> + a
. (5)
From coeff. Of
r n
x
+
in (3) ,
.
) 2 (
2
n n
a
a
n
n
+

. (6)
. . . 3 , 2 , 1 , 0 . 0 . . . . . . . . ) 6 ( , ) 5 (
1 2 7 5 3 1

+
k a a a a a
k
and (6)
. . . , 3 , 2 , 1 ,
) 1 ( ) ( 2
) 1 (
1 2
0
2

+ +

k
k k k
a
a
k
k
k


) 1 ( ! 2
) 1 ( 2 ) 1 (
2
0
+
+

+
k v k
a v
v k
v k
v k
v k
v k
k
x
k v k
v
a x y
+
+

+ +
+


2
2
0
0 1
) 1 ( ! 2
) 1 ( 2 ) 1 (
) (
Now let
) 1 ( 2
1
0
+

a
such that

+
+
+ +

0
2
2
1
.
) 1 ( ! 2
) 1 (
) ( ) (
k
k
k k
v
k k
x
x J x y

.
(7)
Bessel fct. Of the 1
st
. kind of order

.
(b) Irregular
n
l o S
.

) 0 ( < v r
From coeff. Of 0 ) 1 2 ( ,
1
1
+
+
a x
r
in (3)
. 0 1 2 , 0
1
+ a
For
,
2
1
v
see below ,
pg 4
. . . . , 3 , 2 , 1 , 0 . 0 . . . . . . .
1 2 7 5 3 1

+
k a a a a a
k
From coeff. Of
r n
x
+
. 2 ,
) 2 (

n
v n n
a
a
z n
n
Using similar argument to derive (7), we obtain

+ +

0
2
2
2
.
) 1 ( ! 2
) 1 (
) ( ) (
k
v k
v k k
v
k v k
x
x J x y
.
(8)
Bessel fct. Of the 1
st
kind of order
v
.

General
n
l o S
.
to Bessel
n
q e
.
(1) is given by
) (x J
v
and
, ) (x J
v
whenever
e i n eger v . , int
) ( ) ( ) (
2 1
x J c x J c x y
v v
+
. (9)
Note :
v
J
and
v
J

converge for all

x
, and is linearly independent
from each other.
When
n
, an integer , then

+ +

0
2
2
,
) 1 ( ! 2
) 1 (
) (
k
n k
n k k
n
k n k
x
x J

+
+

+ +

0
2
2
,
) 1 ( ! 2
) 1 (
) 1 (
k
n k
n k k
n
k n k letting
k n k
x
) ( ) 1 ( x J
n
n

. (10)
) ( , ) ( x J x J
n n

linearly dependent. Therefore when

a n v ,
second
) (
2 .
x y l o S
n
is sought, using reduction order method / variation of
parameter or other appropriate technique. The most common second
n
l o S
.

used is
[ ]
.
sin
) ( ) ( cos
) (

v
x J x J v
x Y
v v
v
.
(11)
Bessel fct. Of the second kind of order
v
.
) (x J
v
and
) (x Y
v
linearly independent.
Hence the general
n
l o S
.
to Bessel
n
q e
.
(1) is :
(i)
) ( ) (
2 1
x J c x J c y
v v
+
if
v
not an integer ,
or (ii)
) ( ) (
2 1
x Y c x J c y
n n
+
if
v
an integer
n
,
or (iii)
) ( ) (
2 1
x Y c x J c y
v v
+
for all
v
.
Note : When
,
2
1
v
, sin
2
) (
2
1
2
1
x
x
x J

,
_

. cos
2
) (
2
1
2
1
x
x
x J

,
_

.
(12)
Note : For
, n v
integer ,
n
l o s
.
(11) can also be given by
) (
1
2
1
2
2
) (
1
x J
r
x
In x Y
n
n
r
n

'

,
_

+
+

,
_

0 1
2
1 1
2 ! ) ( !
1
) 1 (
1
s
s
r
s n
s
n r r
x
s n s
,
2 !
! ) 1 ( 1
2
1
0
s n
n
s
x
s
s n
+

,
_

.
(13)
where =
. 5772 . 0
1
lim
1

,
_

n
r
n In
r n
.
(14)
Euler-Moscheroni constant.
B. Some important properties of Bessel fct.
(1) When
n v
, integer then
(a)
) ( ) 1 ( ) ( x x J
n
n

(b)
) ( ) 1 ( ) ( x Y x Y
n
n
n

(c)
[ ] d x n x J
n

0
sin cos
1
) (
(d)

'

,
_

n
n
n
t x J
t
t x ) (
1
2
1
exp
(2)
{ } ) ( ) (
1
x J x x J x
dx
d
n
n
n
n

(3)
{ } ) ( ) (
1
x J x x J x
x d
d
n
n
n
n
+


(4)
) ( ) ( ) (
1
x J
x
n
x J x J
n n n

(5)
) ( ) ( ) (
1
x J x J
x
n
x J
n n n +

(6)
{ } ) ( ) (
2
1
) (
1 1
x J x J x J
n n n +

(7) ) (
2
) ( ) (
1 1
x J
x
n
x J x J
n n n
+
+
(8)
) (x Y
n
also satisfies
) 7 ( ) 2 (
.

n
l e r
(9) when
x

'

,
_

,
_

'

,
_

,
_

2 2
1
sin
2
~ ) (
2 2
1
cos
2
~ ) (
2
1
2
1
n x
x
x Y
n x
x
x J
n
n
(10) when
0 x
n
n
x
n
x J
,
_

+ 2 ) 1 (
1
~ ) (
^

'

,
_

. 0 ,
2
0 ,
2
) (
1
~ ) (
n x In
n
x
n
x Y
n
n
C. Orthogonality of Bessel fct.
(1) If i

and j

are roots of
( ) 0 a J
n
then
{ }
j i
a
i n j n i n
a J
z
a
x d x J x J x
2
0
1
2
) ( ) ( ) (

+
Where

'

.
, 0
, 1
j i
j i
j i

(2) Lets
) (x f
be an arbitrary fct. Defined on
a x 0
and be
expanded as

1
, ) (
i
i n i
x J c
where i

is root of
, 0 ) ( a J
n

then
{ }
.
) (
) ( ) ( 2
2
1
2
0
a J a
x d x J x f x
c
i n
a
i n
i

.
(*)
e i .

1
, ) ( ~ ) (
i
i n i
x J c x f
Fourier Bessel series .
(3) Lets
) (x f
be defined on
a x 0
and
x d x f x
a
) (
0

finite.
Then

1
) (
i
i n i
x J c
.(* *)
converges on
[ ] . (*) 0 by given c with a x
i

If
) (x f
ctn. at
0
x

then
( )
equals
) (x f
, and if
) (x f
disctn. Then
( )
equals to
{ }. ) ( ) (
2
1
0 0
+ + x f x f
f
) (
0
r x f
0
x
) (
0
x f
x
D. Some important facts for application
(1)
( )

1
0
1 1 2
, ) ( ) 1 (
) (
2
2
) ( t d t x J t t
m n
x
x J
m
m m n
m n
n
where
. 1 > >m n
(2)

>
+

0 2 2
0
0 ,
1
) ( a
b a
x d x b J e
x a
(3)
. . . . , 3 , 2 , 1 , 0 ,
1
) (
0

n
b
x d x b J
n
(4)
2
1
) (
.
2
1
2
) (
2 2
0
+

,
_

n
n
n
x a n
n
b a
b
n
x d e x x b J
(5)
0 ,
) (
.
2
3
2
) (
2
3
2 2
0
1
1
2
>
+

,
_

+
+

a
b a
b a
n
x d e x x b J
n
n
n
x a n
(6)

+
+
>
0
1
1
0 ,
) 2 (
) (
4
2
2
a e
a
b
x d e x x b J
a
b
n
n
x a n
n
(7)

+ +
+
>

'

+
0
2
2 1
3
0 ,
4
1
2
) (
4
2
2
a e
a
b
n
a
b
x d e x x b J
a
b
n n
n
x a n
n
(8)

'

>

>

a b
b a
a b
x d x b J x a
,
1
, 0
) ( . sin
2 2
0
0
(9)

'

>
>

a b
a b
a b
x d x b J x a
, 0
,
1
) ( . cos
2 2
0
0
(10)

0
1 ) (
n
x d
x
x J
n
(11) If
i

root of
, 0 ) (
0
J
then
{ }
. 1 0 ,
2
1
) (
) (
1
2
1
0
< <

x x In
J
x J
i
i i
i
(12)
x
x Y x J x Y x J
n n n n

2
) ( ) ( ) ( ) (

(13) { } { }

+

x
n n n n
x J x J x J
x
t d t J t
0
1 1
2
2
2
) ( ) ( ) (
2
) (
For further results, please consult Abrahamovit3 , Stegun pg. 355 494 .
E. Temperature in an infinite cylinder
(a) Let
) , , , ( t Z r u
be the temperature at
) , , ( z r P
in an infinite
cylinder . The governing heat conduction
n
eq.
in a solid cylinder is
u k
t
u
2 2

.
1 1
2
2
2
2
2
2

'

,
_

Z
u
o
u
r r
u
r
r r
k
.
(1)
Let the surface temperature at all times be
c
0
0
, and initial condition
. ) (r f
From these BC and IC, we infer that
. ) , ( t r u u
. (2)
Then (1) becomes
,
1
2
2
2

'

r
u
r r
u
k
t
u
. (3)
with

) ( ) , (
) , (
r f o r u C I
C O t a u C B
. (4)
For
, 0 > t
we will assume that
(i)
) , ( t r u
ctn. in the cylinder , including at
0 r
, and
(ii)
) (r f
is at least piecewise continous in cylinder
Letting
) ( ) ( ) , ( t T r R t r u
. (5)
and substituting into the homogenous , ) 3 ( .
n
eq we obtain
,
1 1 1
2
2
2
2

'

+
r d
R d
r r d
R d
R t d
T d
T k
say
Then
0
1
2
2
2
+ + R
r d
R d
r r d
R d

e i.
0
2 2
2
2
2
+ + R r
r d
R d
r
r d
R d
r
. (6)
and
t k
e A t T e i k
t d
T d
T
2 2
) ( .
1
2 2

.
(7)
From (4) , , the condition on R is
) ( , 0 ) ( r R a R
ctn. in cylinder
Now putting ) 6 ( . ,
n
eq r s becomes
0
2
2
2
2
+ + R s
s d
R d
s
s d
R d
s
(8)
and
) ( , 0 ) ( s R a R C B
ctn on
. 0 a s
n
l o S .
to ) ( . g eq
n
is
) ( ) (
1
s J s R

and
. ) ( ) (
2
s Y s R

) (s R
is finite
) 0 ( a s
, then ) ( ) ( .
2
s Y s R l o s
n

must be
rejected.
[ ] . 0 ) ( : s as s Y note

Hence
) ( ) ( ) ( r J s J r R


.
(9)
From
, 0 ) ( a R C B
we see that
) ( 0 ) (
0
s J a J


. (10)
. . . . , 3 , 2 , 1 , i a s s
i i i

where

, 8 4 0 4 . 2
1 1
s

, 1 0 2 5 . 5
2 2
s

, 7 3 5 6 . 8
3 3
s

, 5 1 9 7 . 11
4 4
s . . . . , 9 0 3 9 . 4 1
5 5
s
Thus ,
a
i
i

eigenvalues
} . (11)
and
, ) ( ) (
0 0

,
_

a
r
J r J r R
i
i i

eigenfunction .
) 7 ( , ) 5 (
t k
i i i i
i
e r J t T r R t r u
2 2
) ( ) ( ) ( ) , (
0

) 3 (
is linear and homogenous ,

1
) , ( ) , (
i
i i
t r U A t r u
e i .
t k i
i
i
e r i J A t r u
2 2
) ( ) , (
1

. .
(12)
From
i C I ) 4 (
and
) 12 (
,
. ) ( ) (
1
r J A r f
i
i
i

.
(13)
(13) is Fourier-Bessel series. Therefore
r d r J A r J r r d r J r f r
a
i
i i j j
a

'

0
1
0 0
0
) ( ) ( ) ( ) (

1
0
0
) ( ) (
i
i
a
j i
r d r J r J r A

{ }
2
(
2 2
a J a
A
j i
j

{ }
. . . . , 3 , 2 , 1 , ) ( ) (
) (
2
0
2
1
2

i r d r J r F r
a J a
A
i
a
i
i

. (14)
Hence
n
l o s .
to (3) (4) is given by (12) , where
i
A
is obtained from (14).
Example 2
PDE :
0 , 0 , 9
1
0
2
2
2
> < < +

'

t a r
r
u
r r
u
k
t
u
. (1)
BC :
u t a u , 0 ) , (
ctn. on
a r o
. (2)
IC :
. 0 ) 0 , ( r u

9
is constant . (1) is nonhomogenous , we let
. ) ( ) ( ) , (
1
r J t A t r u
i
i
i

.
(3)
From

1 1
) (
) ( 2
1
i i i
i
a J
r J
a

and sub. (3) into (1) , we obtain

'

+
1
0
2
2
2
1
) (
1
) ( ) ( ) (
i
i i i i
i
i
r J
r d
d
r
r
r d
J d
t A k r J
t d
A d

.
) (
) ( 2
1 1
a J
r J
a
q
i i
i
i

+
. (4)
And from
, ) ( ) (
1
) (
0
2
2
2
r J r J
r d
d
r
r J
r d
d
i i i i
+

(4) becomes
) (
) (
2
) (
1 1 1
2 2
r J
a J a
q
r J A k
t d
A d
i
i i i i
i i
i

'

+
e i . .
) (
2
1
0 2 2
a J a
q
A k
t d
A d
i i
i
i

+
.
(5)
From IC
0 ) 0 , ( r u
we know that, from (3) ,
. 0 ) 0 ( 0 ) ( ) 0 (
1

i i
i
i
A r J A
.
(6)
Now (5) is
{ }
t k i
i i
i
t k i
e
a J a
q
t A e
t d
d 2 2 2 2
) (
2
) (
1

Integrating from
0
to
, t
and using (6), we get
) (
) ( exp 1 2
) (
1
3
2 2
2
a i J i
t k i
k a
q
t A
i

. (7)

) (
) (
) ( exp 1 2
) , (
1
1
3
2 3
2
r i J
a i J i
t k i
k a
q
t r u
i

where
. ) 0 ( . 0 ) ( >
i i
a J

Legendre equation, Legendre polinomial and BVP in spherical

coordinate.
, 0 ) ( 2 ) 1 (
2
+ x y y x y x .
(1)
a
or
[ ] , 0 ) 1 (
2
+

y y x
.
(1)
b
Where .
1 + x
singular pts, the rest being ordinary pt. seek
n
l o s .

to (1) on the interval
1 1 < < x
about
. 0 x
Then
. ) (
0

i
i
i
x a x y
.
(2)
Substituting (2) into (1) will result in
[ ]

+
2 0
2
0 2 ) 1 1 ( ) 1 (
i i
i
i
i
i
x a i i x a i i
Rearranging
[ ] { }

+
+ + +
0
2
. 0 ) 1 ( ) 1 ( ) 2 (
i
i
i i
x a i i a i i
.
(3)
i
x
is linearly independent of each other for
. . . , 2 , 1 , 0 i
(3) implies
. . . . , 3 , 2 , 1 , 0 ,
) 2 ( ) 1 (
) 1 (
2

+ +
+

+
i a
i i
i i
a
i i

.
(4)
If
. 0 . . . . ) 4 ( , 0
6 4 2 0
a a a a
Also,
If
. 0 . . . . ) 4 ( , 0
7 5 3 1
a a a a
Hence

+
1
0
, 2
2 0 1
. 0 , ) (
k
k
k
a x a a x y
. (5)
and
. 0 , ) (
1
1 2
1
1 2 1 2
+
+

+
a x a x a x y
k
k
k .(6)
Where
. . . . , , ,
4 3 2
a a a
will be determined by (4) . Note that
n
l o s . (5) ,
(6) is independent of each other, and also note that for
n
l o s .
(5)
2 2
2
2
) 1 ( 2
) 1 ( 2
) 1 2 ( ) 2 2 (
) 1 2 ( 2 lim lim
x x
k k
k k
k x a
x a
k
k
k
k
k

+ +
+

+
+

n
l o s .
(5) converges for
, 1
2
< x
and diverges when
. 1
2
x
The same
conclusion is also true for
n
l o s .
(6).
n
l o s .
(5) or (6) is discontinous at
, 1 + x
but if (5) / (6) becomes a
polinomial, then at
1 + x

n
l o s .
(5) / (6) will be ctn. This is so if
. . . , 3 , 2 , 1 , 0 , ) 1 ( + n n n
In this instance , i.e when
, ) 1 ( + n n
reccurence relation (4) becomes
. . . . , 2 , 1 , 0 ,
) 2 ( ) 1 (
) 1 ( ) 1 (
2

+ +
+ +

+
i a
i i
n n i i
a
i i .
(7)
, 0 ) 7 (
2

+ n
a
and therefore
. 0 . . . .
8 6 4

+ + + n n n
a a a
1
y
or
2
y
can be a polinomial.
when
) 1 ( + n n
and
n
is even , (7) implies
, . . . .
4
4
2
2 0 1
n
n
x a x a x a a y + + + +
and when is odd,
, . . . . ) (
5
5
3
3 1 2
n
n
x a x a x a x a x y + + + +
Where
. 0 ,
1
a a

Note that when

n
is even, series
n
l o s .

) (
2
x y
is an
infinite power series, and when
n
is odd
n
l o s .

) (
1
x y
becomes infinite
power series.
For this polinomial
n
l o s .
, normally we choose the function value to be unity
) 1 (
at
1 x
therefore when
n
is even we select
.
) ! ( 2
! ) 2 (
2
n
n
a
n
n

. (8)
When
n
is odd, we chose the same
n
a
as given in (8) . Using this
n
a
, then
.
! ) ( ! ) 2 (
! 2 2 (
!
) 1 (
2
1
2
k n k n
k n
k
a
k
n
k n

.
(9)
Hence the polinomial
n
l o s .
of
) (
1
x y
or
) (
2
x y
can be written as

m
k
k n
k
n
n
x
k n k n
k n
k
x P
0
2
! ) ( ! ) 2 (
! ) 2 2 (
!
) 1 (
2
1
) (
.
(10)
For
. . . . , 2 , 1 , 0 n

2
n
m
when
n
is even and
2
1

n
m when
n

is odd. Polinomial (10) is known as Legendre polinomial of order
n
.
Note that
. . . . , 2 , 1 , 0 , ) ( ) 1 ( ) ( n x P x P
n
n
n
) (x P
n

is even function when

n
is even, and

) (x P
n
is odd function when
n
is odd.
(10)
( ) 1 3
2
1
) ( , ) ( , 1 ) (
2
2 1 0
x x P x x P x P

( ) ( ) 3 30 35
8
1
) ( , 3 5
2
1
) (
2 4
4
3
3
+ x x x P x x x P

( ) . , 15 70 63
8
1
) (
3 5
5
etc x x x x P +
In general then, the general
n
l o s .
to Legendre
n
eq.
(1) is
) ( ) ( ) (
2 1
x Q c x P c x y
n n
+
. (11)
Where
) (x Q
n
is given by either
n
l o s .
(5) or (6) .
Orthogonality of
) (x P
n
Theorem : The eigenvalues and eigenfunctions of Legendre
n
eq.
0 ) ( 2 ) 1 (
2
+ x X X x X x . (1)
are
. ) ( , ) 1 ( x P X n n
n n
+
. (2)
Note that if
(i)
, 1 0 0 ) 0 ( < < x on X
then
, ) ( , ) 1 2 ( 2
2
x P X n n
n n n
+
and
(ii)
, 1 0 0 ) 0 ( < < x on X
then
. ) ( , ) 2 2 ( ) 1 2 (
1 2
x P X n n
n n n +
+ +
Note also that
(i)


'
0
2 2
, 0 ) ( ) ( n m x d x P x P
n m
(ii)


+ +
'
0
1 2 1 2
, 0 ) ( ) ( n m x d x P x P
n m
and
(iii)

'
1
1 2
2
) ( ) (
n m n m
m
x d x P x P
where

'

. 0
, 1
n m if
n m if
n m

Example : . . . , 3 , 2 , 1 , 0 , ) (
2
) 1 2 (
) (
+
n x P
n
x
n n

Is orthonormal on
. 1 1 < < x
Hence, on
1 1 < < x
it can
represent any arbitrary fct.
. ) (x f
Theorem : Lets
) (x f
be piecewise ctn. on
, 1 1 < < x
and at any pt.
that it is disctn. Its value be given as
{ }. ) ( ) (
2
1
+ + x f x f

Then on
1 1 < < x

0
, ) ( ) (
n
n n
x P A x f
Fourier-Legendre Series
where

1
1
. . . , 2 , 1 , 0 , ) ( ) (
2
1 2
n x d x P x f
n
A
n n
Example
(i)
. 1 0 , 1 ) ( < < x x f
Then

+
< <

,
_

+
+

0
1 2
2 2
. 1 0 , ) (
) ! ( 2
! ) 2 (
2 2
3 4
) 1 ( 1
n
n
n
n
x x P
n
n
n
n
(ii)
) (
3
2
) (
3
1
2 0
2
x P x P x +
(iii)
) (
5
2
) (
5
3
3 1
3
x P x P x +
(iv)
. . . . ) (
16
3
) (
8
5
) (
2
1
4 2
+ + x P x P x P x

1 0 < < x
(v)

'

< <
< <

1 0 ,
2
1
0 1 ,
2
1
) (
x
x
x f
then

0
) ( ) (
n
n n
x P A x f
where
( )
[ ] [ ]

'

,
_

even n when
odd n when
n n
n n
A
n
n
n
, 0
,
! 2 / ) 1 ( ! 2 / ) 1 ( 2
! 1
2
1
) 1 (
2 / ) (
Example
(i)
. 1 0 , 1 ) ( < < x x f
Then

+
< <

,
_

+
+

0
1 2
2 2
. 1 0 , ) (
) ! ( 2
! ) 2 (
2 2
3 4
) 1 ( 1
n
n
n
n
x x P
n
n
n
n
(ii)
) (
3
2
) (
3
1
2
2
x P x P x +

(iii)
) (
5
2
) (
5
3
3 1
3
x P x P x +
(iv)
. . . . ) (
6
3
) (
8
5
) (
2
1
4 2
+ + x P x P x P x

1 0 < < x
(v)

'

< <
< <

1 0 ,
0 1 ,
) (
2
1
2
1
x
x
x f
then

0
) ( ) (
n
n n
x P A x f
where
( )
[ ] [ ]

'

,
_

even n when
odd n when
n n
n n
A
n
n
n
, 0
,
! 2 / ) 1 ( ! 2 / ) 1 ( 2
! 1
2
1
) 1 (
2 / ) (
Heat Conduction in a solid sphere
(i) Let the temperature of a solid sphere of radius
a r
be
. ) , ( r u

Here were considering the steady state condition, and
. 0 , ) (
~
) , ( < < f a u For the steady state condition, the
governing
n
eq.
is
, 0 sin
sin
1
) (
2
2

,
_

u
u r
r
r
. (1)
where
. 0 , < < < a r
Were also to assume
) , ( r u
and its first and second derivatives ctn. in
the sphere
, 0 a r <
0 .
Note : The above
n
q e .
and its BCs, can also solve the electrostatic
potential problem.
Let the
n
l o s
.
of (1) be given as
. ) ( ) ( ) , ( H r R r u
. (2)
Sub. (2) into (1), we obtain

,
_

, ) ( sin
. sin
1
2
2
R r
r d
d
R
r
d
H d
d
d
H

e i . , , 0 ) (
2
2
a r R R r
r d
d
r <
.
(3)
and
. 0 , 0 sin
sin
1
< < +

,
_

H
d
H d
d
d
. (4)
Put
cos x
into (4) , then
( ) , 0 1
2
+

'

H
x d
H d
x
x d
d
, 1 1 < < x
. (5)
n
q E .
(5) is Legendres
, .
n
q e
and the
n
l o s .
exist when
. . . . 2 , 1 , 0 , ) 1 ( + n n n
n

.
(6)
and
. . . . 3 , 2 , 1 , 0 , ) cos ( ) ( ) (
2
n P x P x H H
n n

.
(7)
Sub.
n

into (3), we obtain

0 ) 1 ( 2
2
2
2
+ +
n
n n
R n n
r d
R d
r
r d
R d
r
.
(8)
e i
, ) (
) ( +
+
n
n
n
n n
r B r A r R
n n
B A ,
. . (9)
{ } ) cos ( ) , (
) 1 (

n
n
n
n
n n
P r B r A r u
+
+
Since (1) is linear, we built the
n
l o s .
as
{ }

+
+
0
) 1 (
) cos ( ) , (
n
n
n
n
n
n
P r B r A r u
.
(10)
Now our
n
l o s .

) , ( r u
must be finite / ctn as
0 r
, we infer that
. 0
n
B
Therefore

0
. ) cos ( ) , (
n
n
n
n
P r A r u
.
(11)
From BC ) 11 ( , ) (
~
) , ( f a u gives

0
. ) cos ( ) 0 (
~
n
n
n
n
P a A f
. (12)

+

1
1
) ( ) (
2
1 2
x d x P x f
n
a A
n
n
n

1
1
, ) ( ) (
2
1 2
. x d x P x f
a
n
A e i
n
n
n .
(13)
Where . 1 1 , ) cos (
~
) (
1
< <

x x f x f
Else we can let

,
_

0
, , ) cos ( ) , (
n
n
n
n
a r P
a
r
c r u
Where
. . . . , 2 , 1 , 0 , sin ) cos ( ) (
~
2
1 2
0

+

n d P f
n
c
n n

as
n
l o s .
to (1).
(ii) Now lets say we are interested in finding the temperature outside the
sphere,
, a r >
with the same BC at the surface.
For finite temperature as
r
, we must impose
0
n
A
in (10).
Then

0
) 1 (
. ) cos ( ) , (
n
n
n
n
P r B r u
.
(14)
From the surface BC ,

0
) 1 (
. ) cos ( ) (
~
n
n
n
n
P a B f

+

+

0
1
. . . , 3 , 2 , 1 , 0 , sin ) cos ( ) (
~
2
1 2
n d P f
n
a
B
n
n
n

Letting (14) be
, ) cos ( ) , (
1
0

n
n
n
n
P
r
a
c r u
+

,
_

.
(15)
Where

0
sin ) cos ( ) (
~
2
1 2
d P f
n
c
n n
. (16)
n
l o s . ) 16 ( , ) 15 ( to our external problem.
Laplace Transform (LT)
(1) Basics
The idea is to use integral as a transformation. If
) (t f
is given, then
{ }

, ) ( ) , ( ) ( ) ( t d t f t s K s F t f T
where
) , ( t s K
is the kernel of the transformation.
,
.
(a) Definition

0 0
, ) (
lim
) (
A
t d t f
A
t d t f
when the rhs integral exist. For example 1 , ) (

t t t f
m
( )
1
]
1

1
1
1
lim lim
1
1
m
A
m
A
m A
t d t
A
(i) If
, 0 , 1
1
>

A as A m
m
(ii) If
, , 1
1
<

A as A m
m
and
(iii) If
, 1 m
the integral
A as

>

1
. 1 ,
1
1
m when
m
t d t
m
PDE /
ODE
) (t f
Simpler
PDE / ODE
) (s F
1
) (

T L
T L
(b) The Laplace transform of
f
, denoted by
{ } ) (t f L
or by
, ) (s F
is
defined by
{ }

0
. ) ( ) ( ) ( t d t f e s F t f L
t s
.
(1)
The Laplace inverse of
, ) (s F
denoted by { } , ) (
1
s F L

is defined
by
{ }

+

i c
i c
t s
s d s F e
i
t f s F L . ) (
2
1
) ( ) (
1
. (2)
From (1) , (2), we can see that given
{ } ) ( ) ( s F t f L
and
{ } ) ( ) ( s G t g L
then
{ } ) ( ) ( ) ( ) ( s G s F t g t f L + +
. (3)
and
{ } , ) ( ) ( ) ( ) (
1
t g t f s G s F L + +

.
(4)
1
, .

L L e i linear operators . , .
2. Examples , building a table.
E 1 Find the LT of
. 1 ) ( t f
By
n
f e d .
{ }

0
1 1 t d e L
t s
. 0 ,
1
0
>
1
]
1

s
s s
e
t s
{ } . 0 ,
1
1 . > s
s
L e i
Also
{ } . 0 , > s
s
k
k L
E2. Find the
T L
of
t
.
By
n
f e d .
{ }

0
. t d t e t L
s

+
1
]
1

0
0
1
t d e
s
t
s
e
st
t s

0
s
1

{ } 0 ,
1
2
> s
s
t L
Also,
{ } . 0 ,
!
1
>
+
s
s
n
t L
n
n
. . . , 3 , 2 , 1 n
E3. What is the LT of
t a
e ?
Using , ) 1 ( .
n
f e d
{ }

0
) (
0
t d e t d e e e L
a s t a t s t a

1
]
1

0
) (
) ( a s
e
t a s
{ } . ,
1
. a s
a s
e L e i
t a
>

E4. If
, sin ) ( t w t f
find the
T L
of
. f
Using ) 1 ( .
n
f e d and
( )
t w i t w i
e e
i
t w


2
1
sin
{ }

,
_

0 0
2
sin sin t d
i
e e
e t d t w e t w L
t w i t w i
t s t s

( )

+

0
) ( ) (
2
1
t d e e
i
t w i s t w i s

'

'

2 2
2
2
1 1 1
2
1
w s
w i
i w i s w i s i
{ } . 0 , sin .
2 2
>
+
s
w s
w
t w L e i
E5. Also note that
{ } { } { } { } t w L i t w L t w i t w L e L
t w i
sin cos sin cos + +
{ }




0 0
) (
t d e t d e e e L
t w i s t w i t s t w i

,
1
2 2
w s
w i s
w i s +
+

{ } , cos
2 2
w s
s
t w L
+

and
{ } . sin
2 2
w s
w
t w L
+

Similarly using result in E3 we can see that

{ }

'

+

) (
2
1
cos
t w t w
e e L t w h L
{ } { } { }
t w t w
e L e L

+
2
1

,
_

'

+
+

2 2
2
2
1 1 1
2
1
w s
s
w s w s
{ } . cosh .
2 2
w s
s
t w L e i

and,
{ } . sin
2 2
w s
w
t w h L

From ) 2 ( .
n
f e d and E1 E2 we obtain ;
(i)
0
1
, 1
1
1 1 1
>

'

'

'

s k
s
L k
s
k
L
s
L
(ii)
0 ,
!
,
1
1
1
2
1
>

'

'

+

s t
s
n
L t
s
L
n
n
(iii)
a s e
a s
L
t a
>

'

,
1
1
(iv)
0 , cos , sin
2 2
1
2 2
1
>

'

'

+

s t w
w s
s
L t w
w s
w
L
(v)
w s t w h
w s
s
L t w h
w s
w
L >

'

'

, cos , sin
2 2
1
2 2
1
3. Theorem Sufficient Condition for LT.
Suppose that
(a)
f
is piecewise ctn. on the interval
A A t , 0
,
ve t
(b)
t
e k t f

) (
when
a K M t , .
and M real const.
. , ve M +
Then
T L
of
f
exist for
> s
as defined by (1) .
4. Operational propeties of LT .
P1. LT of derivatives
From
, ) 1 ( .
n
f e d

'

0
t d
t d
y d
e
t d
y d
L
t s

[ ]

+
0
0
) ( t d t y e s e y
t s t s

) ( ) 0 ( s Y s y +
, ) 0 ( ) ( . y s Y s
t d
y d
L e i

'

.
(5)
where
{ } . ) ( ) ( s Y t y L
Applying ) 1 ( .
n
f e d and
, ) 5 (
. ) 0 ( ) 0 ( ) (
2
2
2
y y s s Y s
t d
y d
L

'

. (6)
Application to ODE
E1. Solve . 6 ) 0 ( , 6 ) 0 ( , 12 2 3
2
+

y y e y y y
t
. (i)
Classical Method
) (i ODE
is constant coeff.
n
eq.
try
. ) (
t m
c
e t y
.
(ii)
C.F : Sub. (ii) into the homogenous
n
q e .
of (i), we obtain
0 ) 2 3 (
2
+
t m
e m m
. ) 2 ( ) 1 ( 0 ) 2 3 (
2
+ m m m m
2 , 1
2 1
m m

t t
c
t t
e B e A t y e y e y
2 2
2 1
) ( , +
.(iii)
P.I : For inhomogenous , .
n
q e put
.
~
) (
2 t
p
e A t y

Sub p
y
into (i)
t t
e e A A A
2 2
12 )
~
2
~
6
~
4 (

+ +
1
~
A
t
p
e t y
2
) (


. ) ( ) ( ) (
2 2 t t t
p c
e e B e A t y t y t y

+ + +
. (iv)
IC
. 2 , 3
6 2 2 6 ) 0 (
6 1 6 ) 0 (

+
+ +
A B
B A y
B A y
# . 3 2 ) (
2 2 t t t
e e e t y

+ +
. (v)
LT Method :
Taking LT of (i)
{ } { }
t
e L y y y L
2
12 2 3

+
{ } { } { } { }
t
e L y L y L y L
2
12 2 3

+
{ } { }
2
1
. 12 2 ) ( 3 ) (
2
+
+
s
Y o y Y s o y s Y s
{ } { }
2
12
18 6 6 2 3
2
+
+ + +
s
s s s Y
e i . .
2
1
2
3
1
2
) (
+
+

s s s
s Y
{ } ) ( ) (
1
s Y L t y

'

+
+

2
1
2
3
1
2
1
s s s
L

'

+
+

'

'

2
1
2
1
3
1
1
2
1 1 1
s
L
s
L
s
L

# . 3 2 ) (
2 2 t t t e
e e t y

+ +
. (vi)
E2 Analyse constant coeff. IVP :
ODE :
. 0 , ) ( A t t f y c y b y a < < + +
.
(i)
IC : . ) 0 ( , ) 0 (



y y y y
Taking LT towards (i) ,
{ } { } ) ( ) 0 ( ) 0 ( ) 0 (
2
s F Y c y Y s b y y s Y s a + +
,
) ( ) (
) ( .
2 2
c s b s a
s F
c s b s a
y a y b s a
s Y e i
+ +
+
+ +

+ +

, ) ( ) ( ) ( ) ( ) ( s F s Q y a y b s a s Q s Y +
1
]
1

+ +
.
(ii)
where , ) ( ) (
1 2
+ + c s b s a s Q called the Transfer function.
{ } ) ( ) (
1
s Y L t y


) ( ) ( ) ( .
2 1
t y t y t y e i +
.
(iii)
where
) (
1
t y
is homogenous
n
l o s .
satisfying the initial conditions
whilst
) (
2
t y
nonhomogenous
n
l o s .
when . 0

y y
When , 0

y y (ii) implies
{ }
{ } ) (
) (
) (
) (
) (
t f L
t y L
s F
s Y
s Q
. .
(iv)
E3 Solve IVP
ODE :
, 0 4 2 + y y t y
. (i)
IC :
. 0 ) 0 ( ) 0 ( y y
. (ii)
Define
{ }

0
. ) ( ) ( ) ( t d t y e s Y t y L
t s
.
(iii)
Inorder to solve (i), we reqd.
{ }

0
t d y t e y t L
t s

[ ]

0
0
) ( ) ( t d e t s e y e t y
t s t s t s

0

+
0
) ( ) ( t d y t e s s Y
t s

{ } . ) ( y t L s s Y +
. (iv)
But from diff (iii) w.r.t
s
( )




0 0
) ( ) ( ) ( t d y t e t d t y e t s Y
s d
d
t s t s
{ } , .
s d
y d
y t L e i
. (v)
{ } . ) (
s d
y d
s s Y y t L
. (vi)
Therefore taking LT of (i) gives
{ } 0 4 2 ) 0 ( ) 0 (
2

'

+ Y Y
s d
y d
s y y s Y s
Y
s
s
s d
y d
e i
2
) 6 (
.
2

s
s
s d
y d
y
3
2
1 1

Integrate w.r.t
s
, we obtain
In
3
4
2

s
Y
In
s
In
, A
A
4 /
2 3
s Y S A n I

. ) (
4
2
3
s
e
s
c
s Y
. (vii)
Note that at this stage we dont know the IC/BC of (vii). But from sufficient cdt.
theorem for the existance of LT we can see that
{ }

0
) ( ) ( ) ( t d t f e t y L s Y
t s

( )

0
t d e k e
t t s

0
) (
t d e K
t s

1
]
1

0
) (
) (

s
e
K
t s

>

'

s
s
K ,
1

. 0 s as
e i vii n i c . , ) ( 0 0 ) ( s Y and { } . 0 ) ( ) (
1


s Y L t y
E4 Now consider solving IVP
ODE :
, 1 4 2 + y y t y
. (i)
IC :
. 0 ) 0 ( , 0 ) 0 ( y y
. (ii)
Taking
T L
onto (i), and also using IC (ii) , we obtain
,
2
1
2
3
2
s
y
s
s s d
y d

,
_

+
{ }
4
2
4
2
2
3
s s
e
s
Y e s
s d
d


.
1
) (
3 3
4
2
s
c C
s
s Y
s
+
. (iii)
For the same reason as in E3,
0 c
and therefore
.
2
1
) ( . ,
1
) (
2
3
1
3
t
s
L t y e i
s
s Y

'

From (i) ,

,
_

+
2
4 ) ( 2 1 . .
2
t
t t s h l

1

. . . s h r
NOTE
1/. Application of LT onto constant coeff. ODE transforms the ODE into
purely algebraic
. .
n
q e
See EI .
2/. Application of LT onto polinomial coeff. ODE transforms the ODE into
ODE ; see E#. Also note
{ } . 2
2
2
2
s d
y d
s
s d
y d
y t L +
3/. Patial Fraction.
) ( / ) ( s Q s p
.
If the degree of polinomial
) (s P
is less than
) (s Q
then
(a)
,
) (
. . . .
) ( ) ( ) ( . . . ) ( ) (
) (
2
2
1
1
2 1 n
n
n
s s
A
s s
A
s s
A
s s s s s s
s P

+ +

(b)
) (
. . . .
) ( ) ( ) ( . . . . ) ( . . . . ) (
) (
1
2
2
1
1
1 m n
p
m
s s
B
s s
A
s s
A
s s s s s s
s P

+ +

,
) (
. . . .
) (
. . . .
) (
2
2
n
n
p
m
p
m
s s
A
s s
B
s s
B

+ +

+ +

+
(c)
[ ]
[ ]
[ ]
. . . .
) (
) (
) (
) (
2
2 2
2 2
2 2
1 1
2 2
+
+
+
+
+
+

+

s
D s c
s
D s c
s
s P
m

[ ]
.
) (
2 2
m
m m
s
D s c
+
+
+
P2. Translation Theorems.
First Translation Theorem : Let
{ } , ) ( ) ( s F t f L
and for any
a
, then
{ } . ) ( ) ( a s F t f e L
t a
.
(i)
e i .
{ } . ) ( ) (
1
t f e a s F L
t a

. (ii)
Proof : By , ) 1 ( .
n
f e d
{ } t d t f e e t f e L
t a t s t a
) ) ( ( ) (
0

>
0
) (
. , ) ( a s t d t f e
t a s
. ) ( a s F
#
E1 Find the LT of
3 2
t e
t
and
. 2 sin
3
t e
t
(a) From
{ } ,
! 3
4
3
s
t L
then using (i)
{ } .
) 2 (
! 3
4
3 2

s
t e L
t
(b) From
{ }
2 2
2
2
2 sin
+

s
t L
and using (i) ,
{ } .
2 ) 3 (
2
2 sin
2 2
3
+ +

s
t e L
t
Note : With help of this theorem, we can built the LT of combination between
exponential fct. and other polinomial, trigonometric, hyperbolic etc. fcts.
: .
n
f e D
The unit step function
, ) ( a t H
or
, ) (t H
a
is defined by

'

<

. , 1
0 , 0
) (
a t
a t
a t H
Sometimes it is common in the litrature to use
) ( a t U
instead of
. ) ( a t H
) ( ) ( t U or t H
a a

Second Translation Theorem : Let
{ } , ) ( ) ( s F t f L
and a any real
number. Then
{ } , ) ( ) ( ) ( s F e a t f a t H L
s a
.
(iii)
. . e i
{ } . ) ( ) ( ) (
1
a t F a t H s F e L
s a


.
(iv)
We can also put (iii) into this form :
{ } { } ) ( ) ( ) ( a t f L e t f a t H L
as
+

.
(v)
E2 Find the LT of

'

<

. , sin 2
0 , 0
) (
t t
t
t f
From
, ) ( sin 2 ) ( t H t t f
and (iii), we obtain
{ } .
1
1
) (
2
+

s
e t f L
s
P3. Scaling . Let
{ } ) ( ) ( s F t f L
and
a
. Then
{ } ( ) .
1
) (
a
s
F
a
t a f L .
(vi)
. . e i
( ) { } ) (
1
t a f a
a
s
F L

.
(vii)
E1 Find { } ,
t a
e L given
{ } .
1
1

s
e L
t
From (vi) ,
{ }
1
1 1

a
s a
e L
t a
{ } .
1
.
a s
e L e i
t a

E2 Find
{ } , cos t w L
if
{ } .
1
1
cos
2
+

s
t L
{ } .
1
1 1
cos ) (
2
+
,
_

w
s
w
t w L i v
{ } . cos
2 2
w s
w
t w L
+

P4 Derivative of the transform. Let
{ } . ) ( ) ( s F t f L
For
0 n

then
{ } .
) (
) 1 ( ) (
n
n
n n
s d
s F d
t f t L
.
(viii)
. . e i . ) ( ) 1 (
1
t f t
s d
F d
L
n n
n
n

'

. (ix)
(viii)
LT
is not a good method to solve non-constant coeff ODE / PDE.
P5. Integration. Let
{ } . ) ( ) ( s F t f L
Then
{ } .
) (
) (
0
S
s F
d f L
t

.
(x)
e i .

'

t
d f s F
s
L
0
1
. ) ( ) (
1

.
(xi)
P6. Let
{ } ) ( ) ( s F t f L
for 0
s s >
and
t t f / ) (
bdd. for
. 0 > t
Then

'

0
. ) ( ) (
1
p d p F t f
t
L
. (xii)
. . e i { } . ) (
1
) (
0
1
t f
t
p d p F L

.
(xiii)
E1 From
{ } w s
w s
w
t w L >
+
, sin
2 2
then

'

s
p d
w p
w
t
t w
L
2 2
sin
. tan
1
s
w

P7. Convolution
n
Def .
The convolution between
) (t f
and
) (t g
is defined by

t
d t g f t g t f
0
. ) ( ) ( ) ( * ) (

0
, ) ( ) (
t
d g t f t

t
d t F g
0
) ( ) (
. ) ( * ) ( ) ( * ) ( t f t g t g t f
Note :
h g f h g f * ) * ( ) * ( *
. * * ) ( * h f g f h g f + +
Let
{ } { } . ) ( ) ( , ) ( ) ( s G t g L s F t f L
Then for
.
0
s s >
{ } { } ) ( ) ( , ) ( ) ( s G t g L s F t f L
{ } g f G F L . .
1

0
) ( ) ( d g t f
{ } , ) ( ) ( * s G s F g f L
.
(xiv)
e i .
{ } . * ) ( ) (
1
g f s G s F L

. (xv)
E1 Find the inverse of
4 4
1
) (
2
+ +

s s
s F
2
1
.
2
1
4 4
1
2
+ +

+ + s s s s
t t
e e
s
L
2 2
2
1
*
) 2 (
1

'

t
t
d e e
0
) ( 2 2
.

t
t t
t e d e
0
2 2
.
# .
4 4
1
2
2
1 t
e t
s s
L

'

+ +

E2 What is th inverse of
.
) ( ) (
2 2
b s a s
s

t b
e hat
b s a s
s
L

'

cos
1
2 2
1

t
t b
d e a
0
) (
cosh


( )

+
t
b a a t b
d e e e e
0
2
1

[ ]

+
+
t
b a b a
t b
d e e
e
0
) ( ) (
2

'

+
2 2
) ( ) (
2
) ( 2 b a
b
b a
e
b a
e e
t b a t b a t b
.
2
2
1
) ( ) (
2 2 2 2
1

'

'

b a
e b
b a
e
b a
e
b s a s
s
L
t b t a t a
Now lets . 1 , ) ( < < n t t f
n
Then

{ }


+

0 0
1
, ) (
1
t s d t s
s
t d t e t L
n n
n
n t s n

. ,
1
0
1
t s z z d z e
s
n z
n

+
{ } 1 ,
) 1 (
1
>
+

+
n
s
n
t L
n
n
.
Example.
{ } ( )
{ } ( ) .
2
2
3
.
2
1
2
3
2
3
2
1
2
1
2
1

s s t L
s
s t L
E3. Find inverse of
.
) 1 (
1
s s
Let
.
1
1
) ( ,
1
) (

s
s G
s
s F
Then
. ) ( ,
1
) (
t
e t g
t
t f

t
e
t s s
L

'

1
) 1 (
1
1

d e
t
t

0
1 1

t
t
d
T
e e
0

t
y
t
y y d e
e
0
, 2
2

( ) ( ) # .
) 1 (
1
1
t f r e e t f r e
e
s s
L
t
t

'

By using 1
st
. Translation / Shifting theorem P2, we obtain
( ) { } ( ) # .
1
1
1
t f r e t f r e e e
s s
L
t t

'

+

P8. Periodic Function. Let
) (t f
be a periodic function with period
e i T , ,
.
, 0 , ) ( ) ( < + t t f T t f
and piecewise ctn in
[ ]. , 0 T
Then
{ }
S T
e
s F
t f L

1
) (
) (
1
. (xvi)
where

T
t s
t d t f e s F
0
1
. ) ( ) (

. ) (
1
) (
.
1 1
t f
e
s F
L e i
S T

'

.
(xvii)
E1. Find the LT of
) (t f
if

'

< <
< <

2 ,
0 ,
) (
t h
t h
t f
and
. ) ( ) 2 ( t f t f +

+

2
0
2
0
1
) ( ) ( ) ( d h e d h e d f e s F
s q s s
. ) 1 (
2 s
e
s
h


{ }
s
s
e
e
s
h
t f L

2
2
1
) 1 (
) (

.
) (
) (
) 1 (
) 1 (
2 2
2 2
s s
s s
e e
e e
s
h
e
e
s
h
s
s

{ } .
2
tan ) ( .
s
h
s
h
t f L e i

#
Application of LT to selected ODEs and PDEs
1. Step Function
(a) Define
) (
) (
0
, , 1
, , 0
) ( ) (
c t H
t H
c
c t
c t
c t U t U
c
c

'

<

Graph of
) (t U y
c

(b)
Graph of
) ( 1 t U y
c

(c)
Graph of
0 , ) ( ) ( ) (
2
t t U t U t h y

2
2
0
, 0
, 1
, 0
) (

<

'

t
t
t
t h
y
1
c t
y
1
c t
y
1

2 3
(d)

'

<

c t
c t
c t f
t g y
), (
, 0
) (
) ( ) ( ) ( c t f t U t g
c

2. ODE
(a) Solve
ODE :
) (
4
5
t g y y y + +
. (i)
IC :
0 ) 0 ( , 0 ) 0 ( y y
. (ii)
where
) ( 1 ) ( t U t g


.
0
, 0
, 1

>
<

'

t
t
Taking LT to (i), we obtain
. / ) 1 (
)} ( { ) 1 {
4
5
) 0 ( ) 0 ( ) 0 (
2
s e
t U L L Y y sy y sy Y s
s

+ +

i.e

,
_

+ +

4
5
1
) (
2
s s s
e
s Y
s
), ( ) 1 ( s H e
s
say . (iii)
where

,
_

+ +

4
5
1
) (
2
s s s
s H

( )
( ) 1
5
4 1
5
4
1
5
4 1
5
4
2
2
1
2
1
2
1
4
5
2
+ +
+ +

+
+

s
s
s
s s
s
s
y
) 0 ( f
t 0
) (t f y
y
) 0 ( f
t 0
c
) ( ) ( c t f t U y
c

) (t g
1

t
0
)} ( { )} ( {
} { ) (
1 1
1
s H e L s H L
Y L t y
s

) ( ) ( ) (

t h t U t h
. (iv)
where
) sin
2
1
cos (
5
4
5
4
)} ( { { ) (
2 2 1 1 1
t e t e s H L L t h
t
+
{ }
( ) { }

'

+ +

< +

t t e t t e e
t t e t e
t y
t t
t t
, sin 2 cos 4
5
1
, sin 2 cos 4 (
5
1
5
4
) (
2 / 2 / 2 /
2 / 2 /
Graph of sol
n
(iv) against
t
(b) Solve the following IVP:
ODE:
, ) ( 4 t g y y +
. (i)
where
t
t
t
t g

<

'

2
2
0
,
,
2
) (

IC:
0 ) 0 ( , 0 ) 0 ( y y
. (ii)
1
1
2 4
6 8 10
t
y
0
2

t
2

Taking LT to
n
eq. (i), and using (ii), we obtain
) 4 (
1
) 1 ( ) (
2 2
+


s s
e s Y
z
s

), ( ) 1 ( s H e
z
s


say.


)} ( { ) ( ,
2
) ( ) (
)} ( { ) (
1
1
2
s H L t h t h t U t h
s Y L t y

,
_

From
,
4
1
4
1 1
4
1
) (
2 2
+

s s
s H
t t t H 2 sin
8
1
4
1
) (

'

+ + ) ( )] ( 1 [ 2 sin )] ( 1 [ 2
8
1
) (
2 2 2
t U t U t t U t t y


. (iii)
Graph of sol.
n
(iii) against
t
.
3. PDE
(a) Heat conduction in a semi-infinite rod
PDE : 0 , 0 ,
2
2
2
> < <

t x
x
u
t
u
. (i)
BC :
0 ) , ( Limit , ) , 0 (

t x u K t
x
u
x
. (ii)
IC :
< < x o x u . 0 ) 0 , (
. (iii)
y
1
2 4
6
8 10
t
Define

0
) , ( ) , (
~
)} , ( { dt t x u e s x U t x u L
st
. (iv)
Applying LT to (i),
, ) 0 , ( ) , (
~
0
2
2
2
2
2
2
dt
x
u
e
x
u
L x u s x U s
st

,
_

'

,
~
, ) , (
2
2
2
0
2
2
2
x
U
dt t x u e
x
st

i.e. 0
~
~
2 2
2

U
s
x
U


+
x x
s s
e s B e s A s x U ) ( ) ( ) , (
~
. . (v)
Now LT towards BC (ii) we see that
s
K
s
x
U
L L t
x
u
L

'

) , 0 (
~
} { ) , 0 (
, . (vi)
and
0 ) , (
~
Limit } 0 { ) , ( Limit

'

s x U L t x u L
x
x
. . (vii)
(vii) & (v)
0 ) ( s A
, whilst (vi) & (v)

2
3
) (
s
k
s B

s s
e
s
k
s x U
2
3
) , (
~
. .(viii)
Inverting (viii)

'

x
s
s
e
s
k
L t x u
2
1
1
) , (

d e k
x
t
2 2
4
0
1

. (ix)
Integrating by parts,

'

,
_

1
]
1

d
x
e e
k
t x u
x
t
t
x
2
2 2
4
0 0
4
2 2 ) , (
2
2
2
1
2
4
2
2
1

d e k
x
t
2
2
4
0
1

Now let
.
4
2
1 2
2

T x

d
x
d
2
2
2
4
2

,
_

d e
x
d
e
x
t
x
x
4
4 2
2 2
2
2
1
2 2
2
2
3
2
2

d e
x
2
4
2

nb:
, 0

d e
x
t
x
2 ) 2 (
2

t
x
t

2
,

'

d t
x
e
x
e
t
K t x u
t
x
2 ) 2 ( 2
2
2
2 ) , (
) 4 (

'

,
_

t 2
x
x e

t
2 K t) u(x,
t) (4
x
2
2
erfc
. (x)
where
.
2
) ( erfc
2
ds e
s

From sol ) ( . x
n
,
.

t
K 2 t o, u ) ( . (xi)
Note From
2
2
0

d e
i.e
1
2
2
0

d e

1
2
2 2
0 0

'

d e d e
1 ) ( erfc ) ( erf + t t

where

d e t
t 2
0
2
: ) ( erf

,
and
.
2
: ) ( erfc
2

t
d e t

Given any
t
, the value of
) ( ert t
, and
) ( erfc t
, can be obtained from Matlab,
Mapple, Mathematica et. al. It can be also glanced from stegun, Abramowitz
(pg. 310) You can also use the probability integral relationship (in your Prob.,
statistic course).
(b) Vibration of semi-infinite string (Travelling Wave)
PDE : 0 , 0 ,
2
2
2
2
2
> >

t x
x
u
c
t
u
. . (i)
BC :
0 ) , ( Limit ), ( ) , 0 (

t x u t g t u
x
. (ii)
IC :
0 , 0 ) 0 , ( , 0 ) 0 , ( >

x x
t
u
x u
.
(iii)
Define
{ } . ) , ( ) , (
~
) , (
0
dt t x u e s x U t x u L
st

.
(iv)
From LT of (i),
2
2
2 2
~
) 0 , ( ) 0 , ( ) , (
~
x
U
c x
t
u
x su s x U s

,
0 0
0
~
~
2
2
2
2

U
c
s
x
U
. (v)
c
sx
c
sx
e s B e s A s x U

+ ) ( ) ( ) , (
~
. (vi)
From LT of BC (ii)
{ } , ) ( ) ( ) , 0 (
~
s G t g L s U say . (vii)
and
0 ) , (
~
Limit

s x U
x
. (viii)
Then (viii), (vi)
0 ) ( s A
, whilst (vii), (vi)
) ( ) ( s G s B
c
x
s
e s G s x U

) ( ) , (
~
. .
(ix)
Inverting,

'

s
c
x
e s G L t x u ) ( ) , (
1

,
_

,
_

c
x
t g
c
x
t H t x, u ) (
. . (x)
For example

>
< <

'

t
t t h
t g
0
, 0
, sin
) (
0
Then
t
c
x
c
x
t
c
x
c
x
t
c
x
t h
t x u
< +
+ < <

'

,
_

, 0
, sin
, 0
) , (
0
2

t
1
h
) (t g
At
0
u
c ct ct
x
(c) Heat conduction in a infinite slab
PDE : 0 , 0 ,
2
2
> < <

t x
x
u
k
t
u
. (i)
BC :
0 , ) , ( , ) , ( >

t u t u u t o u

. (ii)
IC :
. 0 ) 0 , ( x u
< < x 0 . (iii)
Define
{ } dt t x u e s x U t x u L
st
) , ( ) , (
~
) , (
0

. (iv)
Taking LT to (i),
2
2
~
) 0 , ( ) , (
~
x
U
k x u s x U s

0
i.e. 0
~
~
2
2

U
k
s
x
U
. . (v)
Sol
n
.
to (v) is x
k
s
s B x
k
s
s A s x U sinh ) ( cosh ) ( ) , (
~
+ .
(vi)
Applying Lt to BC (ii) we obtain
s u s U / ) , 0 (
~

and
s u s U / ) , (
~

.
(vii)
Then
(vi) , (vii)
a
0 . 1 . B A
s
u
+

i.e
s
u
A

(vi) , (vii)
b


k
s
B
k
s
s
u
s
u
sinh cosh +

'

k
s
x
s
s
u
B cosh 1
sinh
1
x
k
s
k
s
k
s
s
u
x
k
s
k
s
s
u
x
k
s
s
u
s x U sinh
sinh
cosh
sinh
sinh
1
cosh ) , (
~

+

.(viii)
But

u t u ) , (
(a)

'

2
1
cosh
1
x
x
k
s
k
s
e e
s
x
k
s
s
,
) , (
~
1
s x U , say
(b)

k
s
k
s
k
s
k
s
k
s
k
s
k
s
e
e e
s
e e
e e
s
k
s
x
k
s
s
x x x x
2
) ( ) (
1
{ 1
} {
} { 1
sinh
sinh
1

'

0
2 ) ( ) ( 1
n
n x x
k
s
k
s
k
s
e e e
s

'

+ +

+
0
) 2 (
0
) 2 ( 1
n
n x
n
n x
k
s
k
s
e e
s

) , (
~
2
s x U , say
and
(c)

'

'

+ +
0 0
2 (
) 2 (
2
1
sinh
sinh
cosh
1
n n
n x
n x
k
s
k
s
k
s
k
s
e e
e e
s
k
s
x
k
s
k
s
s

+ +

'

'

'

'

0
) 2 2 (
0
) 2 2 (
0
) 2 (
0
) 2 (
1
2
1 1
2
1
2
1
2
1
2
1
2
1
n
x n
k
s
n
x n
k
s
n
x n
n
x n
e
s
e
s
e e
k
s
k
s


), , (
~
3
s x U
say
Therefore inverting
) , (
~
s x U
in (viii) and using above results in (a), (b), (c),
we get
) , (
~
{ ) , (
1
s x U L t x u

)} , (
~
{ )} , (
~
{ )} , (
~
{
3
1
2
1
1
1
s x U L U s x U L U s x U L u

+

i.e

'

,
_

+ +

,
_

+
+

'

,
_

,
_

n
kt
n x
kt
n x
U
kt
x

kt
x u
t x u
2
2
erfc
2
2
erfc
2
erfc
2
erfc
2
) , (

'

,
_

+ +

,
_

+
+

,
_

,
_

0
2
2 2
erfc
2
2 2
erfc
2
2
erfc
2
2
erfc
2
n
kt
x n
kt
x n
kt
x n
kt
x n u

.(ix).
Note: We can also solve BVP (i) . (iii) by using Fourier Method by putting
) , ( ) ( ) , ( t x w x v t x u +
.
(x)
This requires us to solve

u x v ) ( u v u v
dx
v d
) ( , ) 0 ( , 0
2
2
.
(xi)
and
0 , 0 ,
2
2
> < <

k x
x
w
k
t
w
.
(xii)
0 ) , ( , 0 ) , 0 ( t w t w
.
(xiii)

u
x v x w

) ( ) 0 , (
.
(xiv)
2
The eigenvalues, eigenfct. of (xii) (xiii) are

x n
X
n
n n n

sin ,
2
2

,
_

Sol
n
.
to
) , ( t x w
is
t
n
k
n
n
e
x n
A t x w
2
1
sin ) , (

,
_

. (xv)
From IC (xiv), (xv),

1
sin
n
n
x n
A u

dx
x n
u A
n

0
sin ) (
2
} 1 ) 1 {(
2

n
n
u

i.e
t
n
k
n
e
x n
n
u
t x w

2
] ) 1 2 [(
1
) 1 2 (
sin
) 1 2 (
1 4
) , (

Hence

'

1 n
t
1 2n
k
2
e
x 1 2n
sin
1 2n
1

4
1 u k x, u

] ) [(
) (
) (
) (

.
(xvi)
Although Sol
ns
.
(ix) and (xvi) are equivalent, computationally it is easier to do
Sol
n
.
(ix). Sol
n
.
(xvi) converges very slowly, and many terms are reqd. for the
desired accuracy.
D. Vibrating Water Surface With Source
) (t f
PDE : ) (
2
2
2
2
2
t f
x
u
c
t
u
+

,
0 , 0 > < < t x
. (i)
BC :
0 , 0 ) , ( Limit , 0 ) , 0 (
x
>

t t x
x
u
t u
. (ii)
IC :
< <

x x
t
u
x u 0 , 0 ) 0 , ( , 0 ) 0 , (
.
(iii)
Define
{ }

0
. ) , ( ) , ( ) , ( dt t x u e s x u t x u L
st
. (iv)
LT ) 1 (
.
n
eq will give
0
u
)} ( { ) 0 , ( ) 0 , (
2
2
2 2
t f L
x
u
c x
x
u
x su u s +

i.e
2 2
2
2
2
) (
c
s F
u
c
s
x
u

where
)} ( { ) ( t f L s F
. (v)
Sol
n
.
to (v) is
2
) (
) ( ) ( ) , (
s
s f
e s B e s A s x u
x x
c
s
c
s
+ +
. .
(vi)
From LT BC (ii) we obtain
0 ) , 0 ( s u
and
0 ) , ( Limit

s x
x
u
x
. (vii)
Using BC (vii) b towards (vi)

, 0 ) ( s A
and (vii) a
2
) (
) (
s
s F
s B

} 1 {
) (
) , (
2
x
c
s
e
s
s F
s x u


.
(viii)
Special case (a)

f t f ) (
, constant
Then
s f f L s F

} { ) (
.

) 1 ( ) , (
3
x
c
s
e
s
f
s x u



. (ix)
From
1
!
} {
+

n
n
s
n
t L
and shifting Theorem ) ( } ( ) ( { s F e a t f a t U L
as

results

'

,
_

,
_

'

,
_

1
]
1

'

'

2
2
2
2
3
1
3
1
1
2
2
) (
2
1 1
)} , ( { ) , (
c
x
t
c
x
t U t
f
t
c
x
t U
t
f
e
s
L
s
L f
s x u L t x u
c
x
s
c
x

0 0
i.e

'

'

,
_

c
x
t
c
x
t t
f
c
x
t
f
t x u
when
when
,
2
,
2
) , (
2
2

. (x)
Special case (b)
t w A t f cos ) (
, where
w A ,
const.
Then
} cos { ) ( wt A L s F

2 2
s w
As
+

Sol
n
.
(viii) ) 1 (
) (
) , (
2 2
s
c
x
e
w s s
A
s x u

+

) 1 (
1
2 2 2
s
c
x
e
w s
s
s w
A

,
_

+

i.e

'

,
_

,
_

+

s
c
x
e
w s
s
s w s
s
s w
A
s x u
2 2 2 2 2
1 1
) , (
From
) ( 2
2
sin 2 cos 1
1
2
2 2
1
t
wt
wt
w s
s
s
L

'

, say, then
)} , ( { ) , (
1
s x u L t x u

'

,
_

,
_

c
x
t
c
x
t U t
w
A
t x u ) (
2
) , (
2
. (xi)
That is

'

'

,
_

<

c
x
t
c
x
t
w wt
w
A
c
x
t
wt
w
A
t x u
when ,
when
2
sin
2
sin
2
,
2
sin
2
) , (
2 2
2
2
2
. (xii)
E. Heat conduction in a semi-infinite rod with heat sink
PDE : 0 , 0 ,
2
2
2
> < <

t x hu
x
u
k
t
u
. . (i)
BC :

u t u ) , 0 (
, and
0 , 0 ) , ( Limit >

t t x u
x
. (ii)
IC :
< < x x u 0 , 0 ) 0 , (
. (iii)

u h ,
constants.
Define

0
) , ( ) , ( )} , ( { dt t x u e s x U t x u L
st
. . (iv)
From LT of eq
n
.
(i), we obtain
U h
x
U
k x u s x U s
~
~
) 0 , ( ) , (
~
2
2
2

i.e 0
~
~
2 2
2

U
k
s h
x
U
. (v)
Sol
n
.
of (v) is
. ) ( ) , (
~
k
x
k
x
s h s h
Be e s A s x U
+ +
+
.
(vi)
From LT of BC (ii)
s U s U

) , 0 (
~
, . (vii)
and
0 ) , (
~
Limit

s x U
x
. (viii)
Applying BC (viii) to (vi)
, 0 ) ( s A
(vii) to (vi)
s u s B

) (
k
x
h s
e
s
u
s x U
+


) , (
~
. (ix)
Using
s
s F
d f L
t
) (
) (

'

and
{ } ,
2
3 2
) 4 (
1
2 2
t k
e x
e L
t k x ht
h s
k
x

'

k x h s
e
s
L U
s x U L t x u
.
1
)} , (
~
{ ) , (
1
1

d
k
e x
U t x u
t
k x h

3 2
) 4 (
2
) , (
2 2
. (x)
Now let

k
x
2

. Then
) (x
becomes
d e

u
t x, u
t k
x
2 2
2
2
4k
hx

,
_

) (2
2
) (

.
(xi)
when 0 h ,

) 2 (
2
2
) , (
t k x
d e
U
t x u

,
_

t 2k
x
u t x, u erfc

) (
F. Vibration of finite string initiated by one of its end.
0
PDE : 0 , 0 ,
2
2
2
2
2
> < <

t x
x
u
c
t
U
. .
(i)
BC :
0 ,
) (
) , ( , 0 ) , 0 ( >

t
t f
t
x
u
t u

. (ii)
IC :
< <

x x
t
u
x u 0 , 0 ) 0 , ( , 0 ) 0 , (
. (iii)
where
, c
constants and

'

<
>

. 0 , 0
, 0 ,
) ( ) (
t
t f
f t U t f

Define
dt t x u e s x U t x u L
st

) , ( ) , (
~
)} , ( {
. (iv)
From LT of (i) and re-arranging, we get
0
~
~
2
2
2
2

U
c
s
x
U
.
(vi)
From Lt of BC (ii),
0 ) , 0 (
~
s U
.
(vii)
and

'

f
t U L s
x
U
) ( ) , (
~

)} ( { t U L
f

i.e
s
f
s
x
U

) , (

.
(viii)
(vi) , (vii)
0 + B A
i.e . B A
(vi) , (viii)
{ }
s
f
e c A
c
s
c c
s s

c
s
e e s
cf
B A
c
s

+

(
2

,
_

s s
s s
c c
c
x
c
x
e e
e e
s
cf
s x U

2
) , (
~

,
_

,
_

s s
s s
s s
s s
c c
c c
c c
c
x
c
x
e e
e e
e e
e e
s
cf

3
3
2

{ }
c
s
c
s
c
s
c
s
c
s
e
e e e e
s
cf
x x x x

4
) 3 ( ) 3 ( ) ( ) (
2
1
1

+ +

.
(ix)
From
{ }
ks
e
s
k t k t U L


2
1
) ( ) (
and
Ts
e
F
t f L

1
)} ( {
1
where
) ( ) ( t f T t f +
and dt t f e F
T
st
) (
0
1


, we see that
{ } ) , (
~
) , (
1
s x U L t x u

'

,
_

+

,
_

+

,
_

,
_

c
x
t
c
x
t U
c
x
t
c
t U
cf

,
_

+

,
_

+
+
,
_

,
_

c
x
t
c
x
t U
c
x
t
c
x
t U
3 3 3 3
.
(x)
i.e

'

< < +
+ < <

,
_

+
+
< < +

,
_

+ < <

,
_

< <

. 4 3 , 0
, 3 3 ,
3
, 3 ,
2
, ,
, 1 0 , 0
) , (

ct x
x ct x
c
x
t
cf
x ct x
c
x cf
x ct x
c
x
t
cf
x ct
t x u

.
(xi)
Graph of
) , ( t x u
against time
t
Note : In
) (x
we have used
), (
1
) ( s F
s
d f L
t

'

P5 pg. 13/17.
Now
s
s F s F L dt t f e L
T
st
1
) ( )} ( { ) (
1 1

'

) ( ) (
1
s F s F
, . (xii)
Hence
) (t f
is obtained; by taking the inverse of (xii)
G. Impulse function; its LT and application
c
x
c

c
x +
c
2
c
x 3
c
3
c
x + 3
4
) , ( t x u

,
_

c
x cf 2

0
t
2
a
+ a
a
0 t
h
) (t P
Def
n
.
: Impulse function
) (t p
,

'

+
+ < <

a t a t
a t a h
t p
, , 0
,
) (
. (i)
Here 1 >> h and 1 << . Impulse
) ( I
will be defined by
h
hdt
dt t p I
a
a

2
) ( ) (

Choose
2
1
h
such that
1 ) ( I
when 0 .
1 ) ( I
.
(ii)
Let
) ( ) ( t p t p

whenever
2 1 h
. From (i), (ii)

. 1 ) ( Limit
, 0 ) ( Limit
0
0

I
a t t p
. (iii)
From (iii) define Dirac delta function by

. 1 ) 9
, 0 ) (
dt a t
a t a t

.
(iv)
Note that
) (
) ( )} ( {
s s
as
a
a
st
a
a
st
st
e e
s
e h
s
e
h
hdt e
dt t p e t p L

1
]
1

i.e s
s
zhe
t p L
as
sinh )} ( {

.
(v)
whilst
s
s e
t p L a t L
as

sinh
Limit
)} ( { Limit )} ( {
0
0

i.e
as
e a t L )} ( { .
(vi)
and
1 )} ( { t L
.
Note also that
dt
t f
dt t f t p dt t f a t
a
a

2
) (
Limit
) ( ) ( Limit ) ( ) (
0
0

+ < <

a t a t f
* *
0
, ) ( 2 .
2
1
Limit
(Mean Value Theorem)

) ( ) ( ) ( a f dt t f a t
Consider the following wave motion with impulsive input
PDE :
0 , 0 ,
2
2
2
2
2
> < <

,
_

t x
y
x
t f
x
u
c
t
u

.
(i)
BC :
) , ( , 0 ) , 0 ( t x u t u
bounded as
x
,
. (ii)
IC :
0 ) 0 , ( , 0 ) 0 , (

x
t
u
x u
.
. (iii)
Define dt t x u e s x u t x U L
st
) , ( ) , ( )} , ( {

.
. (iv)
Applying LT towards eq
n
.
(i)
. ) 0 , ( ) 0 , ( ) , (
2
2
2 2
s
y
x
e f x
u
c x
t
u
x su s x u s

i.e .
2 2
2
2
2
x
y
s
e
c
f
u
c
s
x
u

.
(v)
Sol
n
.
to (v) is

'

+ +

c v
cs
xe f
c v
s v c C
e v f
e s B e s A s x u
x
x
x x
y
s
y
s
c
s
c
s
when ,
2
when ,
)
) ( ) ( ) , (
2 2 2
2

.
(vi)
Applying LT towards eq
n
.
(i),
s
y
x
e f
x
u
c x
t
u
x u s s x u s

2
2
2 2
) 0 , ( ) 0 , ( ) , (
i.e
x
y
s
e
c
f
u
c
s
x
u

2 2
2
2
2

. .
(v)
Sol
n
.
to (v) is

'

+ +

c v
cs
xe f
c v
s v c
e v f
e s B e s A s x u
x
x
x s
v
s
v
s
c
s
c
s
when
when
,
2
,
) (
) ( ) ( ) , (
2 2 2
2

.
(vi)
From LT BC (ii) we obtain
0 ) , 0 ( s u
and
) , ( t x u
bounded as
x
.
(vii)
(vii)
0 ) ( s A
b
.
(vii)

'

c v
c v
s v c
v f
s B
a
when
when
, 0
,
) (
) (
2 2 2
2

'

c v
cs
xe f
c v
s v c
e e v f
s x u
x
x x
c
s
c
s
v
s
when
when
,
2
,
) (
) (
) , (
2 2 2
2

.
(viii)
Inverting,

'

,
_

'

,
_

,
_

,
_

,
_

c v
c
x
t H
c
x f
c v
c
x
t H
c
x
t
y
y
t H
v
x
t
v c
v f
t x u
,
2
, . .
) (
) , (
2 2
2

Note:
as
as as
as as
e a t L
s e L e a t H L
s e t L e a t a t H L

)} ( {
. / } 1 { } 1 . ) ( {
. / } { )} ( . ) ( {
2

For other examples, please refer to Zill, Cullen; Example 1 4, page 490
495.
PDE : 0 , 0 ,
2
2
> > <

t x
x
u
K
L
u
. (1)
BC : 0 ) , ( , ) , 0 (

t x u Ac t u
t
as
x
, 0 > t .
(2)
IC :
x x u 0 , 0 ) 0 , (
. (3)
Define
{ }

) , ( ) , ( ) , ( t x u L dt t x u e s x u
st
. (4)
Applying LT(4) toward PDE (1),

,
_

2
2
) 0 , (
x
u
k x u u s
0
2
2

u
k
s
x
u
i.e
x x
k
s
k
s
De Be s x u +

) , (
. (5)
Applying LT(4) towards Be (2),
0 ) , ( lim , ) , 0 (
+

s u u
s
A
s u
x
. (6)
From BC (6)
2
, we obtain 0 D . And from (6)
1
,
+

s
A
B
.
x
k
s
e
s
A
s x u

+
) , ( . (7)
0
From
{ }

,
_

'

+
t
k
t
k
e L e
s
L
s k t
4
2
3
1 1
2
,
1

, and convolution
Theorem, we get

d e e
k
Axe
t x u
kt
x t
t
2
3
) 4 (
2
0
2
) , (

.
hb:

)

'

) 4 (
2
2
3
t
x
e t L
.
2
x s
e
x

KERJA KURSUS
INSTITUT SAINSMATEMATIK
UNIVERSITI MALAYA
KXEX3244 PERSAMAAN PEMBEZAAN SEPARA
SEMESTER 2 2007-08
1. Pertimbangkan seutas tali panjang dari 0 x hingga ke
x
. Bila 0 t
tali ini duduk diam secara mendatar. Bila 0 > t titik 0 x serta merta
diberikan pergerakan mencancang
). (t h
Jika setiap pergerakan
mencancang titik dalam tali ini mematuhi pergerakan gelombang, dengan
menggunakan Jelmaan Lapalace atau sebarang cara lain, carikan kedudukan
mencancang sebarang titik
x
pada masa t .
Jika
) sin( ) ( t t h
jika 2 0 t dan
0 ) ( t h
untuk 2 > t , carikan
penyelesaian masalah ini.
Sila hantar penyelesaian anda sebelum pukul 12 tengah hari Khamis
27hb. Mac 2008.
1.
PPS :
2
2
2
2
2
x
u
c
k
u

.
(i)
S.S :

'

>

2 , 0
2 0 , sin
) ( ) , 0 (
t
t t
t h t u
.
(ii)
0 ) , ( k x u
apabila
x
. (iii)
S.A :
, 0 ) 0 , ( x u 0 ) 0 , (

x
t
u
. (iv)
Takrifkan

0
) , ( )} , ( { ) , ( dt t x u e t x u L s x U
st
. (v)
Ambil JL. Thp. (i)
2
2
2 2
) 0 , ( ) 0 , ( ) , (
x
U
c x
t
u
x u s s x U s

i.i 0
2
2
2
2

U
c
s
x
U
x x
c
s
c
s
e s B e s A s x U

+ ) ( ) ( ) , (
.
(v)
Dr.
0 ) , ( had

t x u
x
, i.i.
0 ) , ( had

s r U
x
dapat
0 ) ( s A
.
Dr. (ii),
{ } { } { } ) ( ) ( ) ( ) ( ) ( ) , 0 ( s H s B s H t h L t h L t u L
x
c
s
e s H s x U

) ( ) , (
. (vi)
Dr. Teorem Anjakan { } ) (

) ( ) (
1
a t U a t f s F e L
as


dapat
u
x
2 t
0
0

'

+ < <

,
_

,
_

,
_

,
_

,
_

lain lain untuk

bila siri
, 0
2
sin

} ) ( { ) , (
1

c
x
t
c
x
c
x
t
c
x
t u
c
x
t
c
x
t u
c
x
t h
e s H L t x u
x
c
s
iaitu

'

< <

,
_

lain lain untuk

bila siri
,
ct x c 2 t
c
x
t
t x, u
0
) (
) (
.
) , (
) , (
2
2
2
2
2
2
2
2
2
2
2
2
x
s x U
c
dt t x u e
x
c
dt
x
u
e c
x
u
c L
st
st

'

2 t
ct
u
c
x
2 +
c
x
u
x
KXEX3244 PERSAMAAN PEMBEZAAN SEPARA
ISM, UNIVERSITI MALAYA
KERJA KURSUS INI MENYUMBANG 10% KEPADA KESELURUHAN
MARKAH KURSUS KXEX3244.
SOALAN
Dalam suatu sungai sedalam h meter, air mengalir secara seragam dalam
arah
x
. Endapannya menjadi terampai dalam aliran air ini. Akibat dari
resapan bergelora aliran air tersebut. Oleh kerana endapan ini ada jusim,
maka ia akan mengenap ke dasar sungai. Katakan halaju enapan partikel
endapan ini ialah
w
meter sesaat dan pekali resapan bergelora ialah k
(anggap
k w h , ,
malar). Tunjukkan bahawa konsentrasi endapan
) , ( t z C

memuaskan persamaan
. 0 ,
2
2
h z
z
C
k
z
C
w
t
C
< <

. (1)
Andaikan bahawa kadar fluks mencancang endapan dipermukaan air
z
C
k wC

adalah sifar, sementara di dasar sungai pula ia bertindak
sebagai punca endapan tak terhingga banyaknya iaitu
0
) , 0 ( C t C
. Di sini
0
C
malar.
(a) Carikan
) , ( t z C
, jika diberitahu bahwa syarat awal
0 ) 0 , ( z C
.
(b) Bila
t
, penyelesaian
) , ( t z C
bahagaian (a) anda menuju kepada
nilai apa?
(c) Untuk kes mantap, per (1) menjadi
. 0 ,
2
2
h z
z
C
k
z
C
w < <

. (2)
Selesaikan per (2) ini tertakluk kepada syarat sempadan yang diberikan
Adakah penyelesaian per (2) ini sama dengan penyelesaian di dalam
bahagian (b)?
Sila hantar penyelesaian anda sebelum pukul 10.05 pagi khamis 8 hb Feb
2007.
(Soalan ini adalah dari buku Mathematical Analysis in Engineering karangan
Chiang C. Mei, terbitan Cambridge University. Press, ms 87)
Khamis 25hb Jan 2007
P.P.S : h z
z
C
k
z
C
w
t
C
< <

0 ,
2
2
.
(1)
S. S :
0 ) , ( ) , ( , ) , 0 (
0

t h
z
c
k t h C w C t C
. (2)
S. A :
0 ) 0 , ( Z C
.
(3)
(a) PPS (1) homogen, SS (2) , tidak homogen. Letak
) , ( ) ( ) , ( t z U z V t z C +
.
(4)
Masukkan (4) ke dalam (1),

'

'

2
2
2
2
z
U
dx
v d
k
z
U
dz
dv
w
t
U
Pilih
2
2
dz
v d
k
dz
dv
w , justeru itu
(5)
2
2
z
U
k
z
U
w
t
U

. (6)
s.s (2)
1
, (4)
) , 0 ( ) 0 (
0
t U V c +
i.i.
0 ) , 0 ( ), 0 (
0
t U V c
s.s. (2)
2
, (4)
0 ) , ( ) ( )} , ( ) ( {

'

+ + + t h
z
U
h
dz
dV
k t h U h V w
akan memberi
) ( ) ( h
dz
dV
k h wV +
.
(7)
0 ) , ( ) , (

t h
z
C
k t h C w
2
2
z
C
k
z
C
k
z
C
w
T
C

x
h
z
0

U
,
0 ) , ( ) , (

+ t h
z
U
k t h U w
. (8)
SA (3)
0 ) 0 , ( ) ( + z U z V
i.i
) ( ) 0 , ( z v z U
.
(9)
Dr. (5), 0
2
2
+
dz
dv
k
w
dz
v d
Letak
0 ) (
2

,
_

+
k
w
m m e e e z V
mz mz mz
0
mz
e , dapat
0

,
_

+
k
w
m m
i.i
k
w
m m
2 1
, 0
.
z
k
w
Be A z V

+ ) (
dengan B A,
Dr. s.s.

C V ) 0 (
dapat
B A C +
.
Dr. s.s. (7),
{ } 0 0

'

+ +

A Be
k
w
k Be A w
h h
k
w
k
w
z
k
w
e C z V


0
) ( .
(10)
Untuk menyelesaikan Pr, (6) tertakluk kepada SS
, 0 ) , 0 ( t U
(8) serta SA
(9), letak
). ( ) ( ) , ( t T z H t z U
. (11)
Masukkan (11) ke dalam 6
T
dz
H d
k T
dz
dH
w
dt
dT
H
2
2

i.i

'

+
dt
dH
k
w
dz
H
H dt
dT
kT
2
2
1 1

pemalar pemisah
0
2
2
+ H
dz
dH
k
w
dz
H d
.
(12)
,
0 T k
dt
dT

.
(13)
s.s
0 ) 0 ( 0 ) , 0 ( H t u
.
(14)
, ss. (8)
0 ) ( ) ( + h
dz
dH
k h H w
Dr. (12), kita boleh letak
mz
e z H py ) ( dan hujahkan bahawa untuk
mendapatkan
py
. dalam sebutan
cos , sin
, maka
.
4
2
2

,
_

<
k
w

Tetapi di sini
ditunjukkan jalan panjang, tetapi lebih jelas.
Katakan
) ( ) ( ) ( z P z S z H
.
(15)
dengan
2
2
2
2
2 ,
dz
P d
S
dz
dP
dz
dS
P
dz
H d
dz
dP
S P
dz
dS
dz
dH
+ + + .
(16)
Masukkan (15), (16) ke dalam (12) kita peroleh
. 0 2
2
2
2
2

'

,
_

+ +

'

,
_

+ + P S
dz
dS
k
w
dz
s d
dz
dP
S
k
w
dz
dS
dz
P d
S
. (17)
Dalam Pr (17) pilih pekali
dz
dP
sifar i.i
0 2

,
_

+ s
k
w
dz
dS
i.i
z
k
w
e z s
2
) (

.
(18)
Supaya
py
. Pr. (17) ada cos, siri , yaitu jadikan pr. (17) sbg. Pr. gerakan
harmonic ringkas. Dari (18) dan
,
2
,
2
2 2
2
2
2
z z
k
w
k
w
e
k
w
dz
s d
e
k
w
dz
dS

,
_

Pr. (17) menjadi

0
4
2
2
2
2

,
_

+ P
k
w
dz
p d

.
(19)
Dr. s.s (14)
1
, (15);
0 ) 0 ( P
.
(20)
s.s (14)
2
, (15) ;
0 ) ( 2 ) ( + h
dz
dP
k h P w
Kes 1. 0
4
0
2
2
2
2
P
k
w
dz
P d

z
k
w
z
k
w
Be Ae z P Py
2 2
) ( :

+
) (
2
) (
2 2
z
k
w
z
k
w
Be Ae
k
w
z
dz
dP

s.s.
0 0 ) 0 ( + B A P
, i.i A B .
s.s. 0 ) (
2
2 ) ( ) 20 (
2 2 2 2
2
+

h
k
w
h
k
w
z
k
w
h
k
w
e e
k
w
A k e e A w
0 0 B A
0 ) ( z P
.
Py
. remeh
0
z siri B z P
n n n
) (
Dr. (15), z e B z H
n
z
n n
k
w
sin ) (
2

.
Dari (13), iaitu
kT T k
dt
dT
2

i.i n n
n
T k
t d
T d
2

kita dapat
kt
n n
n
e A t T
2
) (

.
, 3 , 2 , 1 , sin ) , (
2
2


n z e e D t z u
n
kt z
n n
n k
w

Oleh kerana Pr. untuk

u
, iaitu Pr. (6) linear, bentuk
py
sebagai

w
k
f
2
) (
2

) ( f
1

. sin ) , (
2
2
1
z e e D t z u
n
kt
n
z
n
n k
w

z e e D e c t z C
n
kt
n
z
n
z
n k
w
k
w

sin ) , (
2
2
1

+
.
(23)
Dari SA
, 0 ) 0 , ( z C
(23) memberi
z e D e C
n
n
z
n
z
k
w
k
w
sin
1
0
2

1
0
sin
2
n
n n
z
z D e C
k
w

.
Maka dz z e
h
c
D
n
h
z
n
k
w
sin
2
0
0 2

1
1
1
1
1
]
1

,
_

2
2
2
0
2
2
n
n
k
w
h
c

.
2
2
2
2
2

'

,
_

n
n
k
w
h
C

(b) Bila
, t
dari (23) ,
. ) , (
z
k
w
e C t z C

(c) Pr. mantap -
. 0 ,
2
2
h z
z
c
k
z
c
w
s s
< <

A z
k
w
z
c
n I
k
w
z
c
z
c
s s s
~
2
2
+

z
s
k
w
e A
z
c

Z
k
w
s
z
w
k
s
e A
Z
C
e
w
k
A B t z C

) , (

C
w
k
A B C C S S
s

0
) 0 ( . ..
0 . .

'

h
k
w
h
k
w
e A k e
w
k
A B w h Z i d S S

C
k
w
A B , 0
. ) (
z
k
w
s
e C z C



) ( ) , ( z C t z C
s

apabila
t
Nota :
.
4
2
2
2 2
k
w
n n

0
2
>
n

maka
. . . . 3 , 2 , 1 ,
4
2
2
2
> n
k
w
n

PDE :
x
c
h
x
c
k
t
c

2
2
. (1)
BC :
0 , 0 ) , ( , 0 ) , 0 ( > t t c t c
.
(2)
IC :
< < x c x c 0 , ) 0 , (

.
(3)
Let
) ( ) ( ) , ( t T x X t x c
. (4)
T
x d
X d
h T
x d
X d
k
t d
T d
X
2
2
) 1 ( , ) 4 (
.
1
2
2
2

x d
X d
x
h
x d
X d
x
k
t d
T d
T
t
e A t T
t d
T d
T
2
) (
1
2

.
(5)
0
2
2
2
+ X
x d
X d
h
x d
X d
k
. (6)
Let . ) (
x m
e x X .
(7)
0 ) 7 ( , ) 6 (
2 2
+ m h m k
k
k h h
m
2
4
2 2
+

Case 1
) 0 (
4
0 4
2
2 2 2
>
k
h
k h
.
2 ,
) ( 1
x
k
h
x m
x
e e X
x
k
h
e x x X
2
2
) (
x
h
x
h
e x B e A x X
2 2
~ ~
) ( +
. (8)
From
0 ) ( , 0 ) 0 ( , ) 2 ( , ) 4 ( X X
.
(9)
. . 0 ) , ( 0 ) (
0
~
0 0 ) (
, 0 0 ) 0 (
2
n
h
Sol Trivial t x C x X
B e B X
A X

>



Case2
3
2
4
. 0 4 , 0 4
2
2 2 2 2 2 2

,
_

< > >

k
h
k h k h
let

k
h
m
k
h
m
k
h
m
2
,
2
,
2
2 1 2 , 1

+


0 >

0 >
x m x m
e B e A x X
, ,
) ( +
B A B A X + 0 0 ) 0 (
( ) B A e e A X
m m
0 0 0 ) (
2 1

n
l o S trivial t x C x X . , 0 ) , ( 0 ) (
Case 3
0 , 4
4
, 0 4
2 2 2 2
2 2 2
2 2
> > +

<

h h k
k h
k h
k
h
4
2
2
>
k
h
m
k
h
m
k
h
m
2
1
,
2
1
,
2
1
2 1

x
k
h
e
k
x
B
k
x
A x X
2
2
sin

2
cos

) (

,
_

+

A X

0 0 ) 0 (
. 0

, 0
2
sin .
2
sin

0 0 ) (
2
B
k
e
k
B X
k
h

n
k
sin
2
sin

. . . , 4 , 3 , 2 , 1 , 2

n
n
k
n

. 4 .
2 2 2
h k e i
n n

. (10)
.
2
sin ) ( ) (
2
k
x
e x X x X
n
x
k
h
n

. (11)
From
) 10 ( , ) 5 (
t
k
h
n
n
e t T t T
4
2 2
) ( ) (
+

t
k
h
x
k
h
n
n n n
n
e e
k
x
T X t x c
4 2
2 2
. .
2
sin ) , (
+

,
_

1
4
2
2 2
.
2
sin ) , (
n
t
k
h
n
n
x
k
h n
e
k
x
D e t x c

. (12)
From IC (3) ,

1
2
.
2
sin
n
n
n
x
k
h
k
x
D e C

. (13)
e i .

1
2
.
2
sin
n
n
n
x
k
h
k
x
D e C

,
_

x d x
m
x
n
D
m
e c
n
n
x
k
h
sin sin sin
1
2

1
0
sin sin
n
n
x d
x m x n
D



,
_

m
D
x d
x n
e c D
x
k
h
n

sin
2
0
2

Now let

+
1
1
1
1
]
1

0
2
0
2 2
cos
2
cos
sin x d
x n
e
k
h
n
n
x n
e x d
x n
e I
k
h
x
k
h
x
k
h
n

'

1
1
1
1
]
1

'

0
2
0
2 2
sin
2
sin
2
cos 1 x d
x n
e
k
h
n
n
x n
e
k n
h
n e
n
x
k
h
x
T
h
k
h

n I
k n
h
n e
n
k
h 2
2
2
cos 1

,
_

'

,
_

,
_

,
_

+

n e
n k n
h
n I
k
h
cos 1
2
1
2
2

,
_

,
_

+

n e
n k n
h k n
n I
k
h
cos 1
4
4
2
2 2 2
2 2 2 2 2

,
_

n e
h k n
k n
n I
k
h
cos 1
4
4
2
2 2 2 2 2
2

1
1
]
1

k
h
n
n
e
h k n
C k n
D
2
2 2 2 2 2
2
) 1 ( 1
4
8

. (14)
( )

+
1
1
]
1

+
1
4
4
2 2 2 2 2
2
2 2
2
2 2 2 2 2
4
) 1 ( 1
8 ) , (
n
t
k
h k n
k
h
n
x
k
h
e
h k n
e n
e C k x C

. (15)