Greens Function Associated with one and two dimensional Problem

In mathematics a greens function is type of function used to solve

inhomogeneous differential equations subject to specific initial conditions or boundary
conditions. Greens functions provide an important tool when we study the boundary
value problem. They also have intrinsic value for a mathematician.
Also greens functions in general are distribution, not necessarily proper function. Green
functions are also useful for solving wave equation, diffusion equation and in quantum
mechanics, where the greens function of the Hamiltonian is a key concept, with
important links to the concept of density of states. The greens function as used in
physics is usually defined with the opposite sign that is

) ( ) , ( s x s x G
This definition does not change significantly any of the properties of the Greens function
in heat conduction we know that the Greens function represents that temperature at a
field point due to a unit heat source applied at source point. In electro static the greens
function stand for the displacement in the solid due to the application of unit point force.
In this project construction of greens function in one and two dimension has
shown. There are more then one way of constructing greens function (if it exist) but the
result is always same. Due to this we can say that greens function for a given linear
system is unique.
We start with the brief introduction of the Dirac delta or Diracs delta function
which is not strictly a function in real sense of functions.
Dirac Delta Function:-
i. (t-a) = 0 if t 0
ii. (t-a)dt = 1
iii. (t-a)f(t)dt = f(x)
iv. (t-a)= (a-t)
Greens Function Associated with one dimensional boundary value problem:
Consider the following boundary value problem.
M[y] = F(x)(1)
x
1
< x <
x
2
[ ] y
B1
[ ] y
B2
M is defined by
M = ( ) ( ) x A
dx
d
A
dx
d
x
A 0 1
2
2
2
+ +
From(1)
( ) ( ) ( ) ( ) x F y x A
dx
dy
x A
dx
y d
x
A
+ +
0 1
2
2
2
Divide by ( ) x A
2
we will get
( )
( )
( )
( )
( )
( )
+ +
x A
x F
y
x A
x A
dx
dy
x A
x A
dx
y d
2 2
0
2
1
2
2
(2)
Let
( )
( )
( )

dx
x A
x A
e x P
2
1
Multiply equation (2) by P(x)
( )
( )
( )
( )
( )
( )
( )
( ) ( )
( )
( ) ( ) ( ) x f y x q
dx
dy
x P
dx
d
x A
x F x P
y x P
x A
x A
dx
dy
x P
x A
x A
dx
y d
x P +
]
]
]

+ +
2 2
0
2
1
2
2
( ) ( ) ( )
[ ] ( ) x f y
x f y x q
dx
d
x p
dx
d

]
]
]

+
,
`

.
|

Where
( ) ( ) x q
dx
d
x P
dx
d
+
]
]
]

Consider the self-ad-joint boundary value problem
[ ] ( )
[ ] [ ]
( ) 3
,
2 1

'

y B y B
x f y
Now split the boundary value problem into the following boundary value problems.
[ ]
[ ] [ ]
( ) 4
,
0
2 1

'

y B y B
y
[ ] ( )
[ ] [ ]
( ) 5
0 , 0
2 1

'

y B y B
x f y
The solution of problem (3) is written as
P H
y y y +
H
y
is complementary function which satisfy the homogeneous differential equation
0
H
y
P
y
is the particular solution to the inhomogeneous differential Equation
( ) x f y
P
+
The general solution of (4) is written as
( ) 6
2 2 1 1
+ y c y c y
H
Here
1
y &
2
y are solution of homogeneous Equation [ ]
2 1
, 0 c c and y are arbitrary
constant and they can be determined by applying the boundary condition.
Now consider the problem (5). Suppose that the solution of problem 5 can be
expressed in the integral form as
( ) ( ) ( ) 7 ,
2
1

ds s f s x g y
x
x
P
Where g(x,s) is the greens function which is to be defined later. The negative sign in (7)
describe the physical interpretation. Apply the differential operator on both side of
7
[ ] ( ) ( )

,
`

.
|

ds s f s x g y
x
x
P
2
1
,
Since & the integral operator commute each other.
[ ] ( ) ( ) ) (
2
1
i ds s f g y
x
x
P

Also
( ) ( ) ( ) ) (
2
1
ii x f ds s f g x
x
x

And also we have

[ ] ( ) ) (iii x f y
P

From i, ii, and iii.
( ) ( )

ds s f g
x
x
2
1

( ) ( )ds s f s x
x
x

2
1
,
[ ] ( ) [ ] ( ) 0
2
1
+

ds s f s x g
x
x

Since
( ) s f
is arbitrary
( ) 0 s f
[ ] ( ) 0 + s x g
[ ] ( ) s x g
(8)
( ) s x
is dirac delta function.
To, determine the unique greens function condition (8) is not enough. We have to
determined other condition also from homogeneous boundary condition.
[ ] 0
1
y B , [ ] 0
2
y B
[ ] 0
1

P
y B , [ ] 0
2

P
y B
( ) ( ) 0
2
1
1

]
]
]
]

ds s f s x g B
x
x
( ) ( ) 0
2
1
2

]
]
]
]

ds s f s x g B
x
x
[ ] ( ) 0
2
1
1

ds s f g B
x
x
And
[ ] ( ) 0
2
1
2

ds s f g B
x
x
Since
( ) s f
Can be almost any function the above relations are satisfied only if.
[ ] 0
1
g B , [ ] 0
2
g B (9)
Hence the greens function we are looking for is solution of the following boundary
problem
[ ] ( ) s x g
(10)
Related with boundary condition
[ ] 0
1
g B , [ ] 0
2
g B
Where s is a fix value lies between x
1
and x
2.
and. The above problem is similar to that
given equation in (5) only the forcing function in (9 ) is delta function rather then
arbitrary function f(x). This means that solving the problem for g is simpler then solving
the corresponding problem y.
And once the greens function has been determined for particular operator and
set of boundary conditions it may be used for solving problem (5) for any umber of time
where only the function f(x) changes from problem to problem. It is this feature of
greens function that make it most useful I application.
The greens function g(x,s) associated with boundary value problem
[ ] ( ) x f y
[ ] y B
1
, [ ] y B
2
Where
( ) ( ) x q
dx
d
x P
dx
d
+
]
]
]

Satisfying
(a)
[ ] ( ) s x g
1
x <
x
<
2
x
(b) [ ] 0
1
g B , [ ] 0
2
g B
(c) ( ) ( ) s s g s s g , ,
+

(d)
( ) ( )
( ) s P x
s s g
x
s s g 1 , ,

+
The condition d is called jumped discontinuity of greens function at x=5.
Based on above condition an exploit formula for the greens function can be
designed. It is observed from the condition (a) if either x<s or x s then
[ ] 0 g
By definition of Dirac Delta function if are solution of
homogeneous differential equation
[ ] 0 g
Such that [ ] 0
1 1
Z B [ ] 0
2 2
Z B so, from
condition (a) and (b) the greens function has following form
( )
( ) ( )
( ) ( )

'

<

s x x Z s v
s x x Z s u
s x g
2
1
,
where u and v are function to be determined imposing the
condition c and d. the unknown function u & v must be chosen such that
( ) ( ) ( ) ( ) 0
1 2
s Z s u s Z s v
( ) ( ) ( ) ( )
( ) s P
s Z s u s Z s v
1
1 2


By solving the above two simultaneous equation we will get
( )
( )
( ) ( )( ) s Z Z W s P
s Z
s v
2 , 1
1

( )
1 2 2 1 2 , 1
Z Z Z Z Z Z W
Since Z
1
and Z
2
are such solution we can write
( ) ( )( ) ( ) ( )( )( ) c s x Z Z W s P s Z Z W s P
2 , 1 2 , 1
The greens function can be written as
( )
( ) ( )
( ) ( )

'

<

< <

2
2 1
1
2 1
,
x x s
c
s Z x Z
s x x
c
s Z x Z
s x g
Greens function is symmetric in x and s.
i.e g(x,s) = g(s,x)
Greens function associated with two dimensional problem:
To, introduce greens function in two dimensions, we consider the PDE (poissons
equation which here represents static deflection of a rectangular membrane).
( ) ) 11 ( ,
2
2
2
2

y x f
y
u
x
u
Here f(x,y) represent the external load per unit area, divided by T(tention in the
membrane, which here has the dimension of force per length)
The B.CS. in this case are
( ) ( ) 0 , , 0 y a u y u
,
( ) ( ) ) 12 ( 0 , 0 , b x u x u
A concentrated force acting at a point (x
/
,y
/
) may be simulated by the two dimension delta
function
( ) ( ) y y x x
F

Let
( ) 0 , y y x x G
be the greens function associated with the problem

2
2 2
y
G
x
G
( ) ( ) ) 13 ( y y x x
F

G(0,y) = G(a,y) = 0, G(x,0) = G(x,b) = 0 (14)

Then the solution of the problem(11) and (12) is given by
( ) ( ) ( )

a b
y d x d y x f y y x x G y x u
0 0
, , ,
We first find a complete set of Eigen functions of the associated homogeneous Eigen
value problem viz.
( ) ( ) y x u y x u
y x
, ,
2
2
2
2

,
`

.
|

with the same B.CS as on u(x,y).

We find
,... 3 , 2 , 1 , ,
2
2 2
2
2 2

,
`

.
|
+ n m
b
n
a
m
mn

and
b
y n
a
x m
ab
y x u y x u
mn

sin sin
2
) , ( ) , (
which have been normalized to unity. Now the required greens function can be
represented as
( )
b
y n
a
x m
y x A
ab
y y x x G
n m
mn

sin sin ) , (
2
;
' '
1 , 1
' '

Now the required greens function can be represented as

( )
b
y n
a
x m
y x A
ab
y y x x G
mn
n m

sin sin ) , (
2
;
' '
1 , 1
' '

Substituting this into (13)

b
y n
a
x m
y x A
b
n
a
m
n m
mn

sin sin ) , (
' '
,
2 2 2 2

]
]
]

) ( ) (
/ /
y y x x
F

Multiplying by
( ) ( )
b
y
a
x
n m
' '
s i n s i n
and integrating w.r.t x and y we have.
b
y n
a
x m F b a
A
b
n
a
m
ab
mn
' '
2
2 2
2
2 2
sin sin
2
.
2
2

,
`

.
|
+
Or
b
y n
a
x m F
ab
y x A
b
n
a
m
mn
' '
' '
2
2 2
2
2 2
sin sin
2
) , (

,
`

.
|
+
Hence on substitution
( )
b
y n
a
x m
b
y n
a
x m F
n a m b
b a
ab
y y x x G
n m


sin sin sin sin
4
;
/ '
,
2 2 2 2 2 2
2 2
' '

+

1 ,
/ '
2 2 2 2 2
sin sin sin sin
1 4
n m
b
y n
a
x m
b
y n
a
x m
n a m b
abF

References
Boas, M.L. (1983). Mathematical Method in physical Sciences. (2
nd
Edition).
John Wiley and sons, New York
Khalid Latif Mir. (1997). Problem and Methods in Mathematical Physics and
Applied Mathematics. Markazi Kutab Khana, Urdu Bazar, Lahore.
Barton, G (1989). Element of Greens function and Propagation. Oxford, UK