1 D Greens Functions
1 D Greens Functions
1 D Greens Functions
In these notes I just want to explain things on pages 663-667 in Arfken and give an
example for solving a Sturm-Liouville dierential equation using one-dimensional Greens
function method.
Denition 1. Let p(x) and q(x) be two real functions which are continuous on the
interval [a, b] and let p(x) be positive on this interval. Operator L dened as
L[ ] =
d
dx
p(x)
d
dx
+ q(x) , (1)
is called a Sturm-Liouville operator.
Any dierential equation
L[y] + f(x) = 0, (2)
where f(x) is function dened on [a, b] is called a dierential equation in the sturm-Liouville
form or simply a Sturm-Liouville dierential equation.
Example 1. Let p(x) = 1 + x
2
and q(x) = 1 x
2
. Determine L[sin x].
Solution.
L[sin x] =
d
dx
(1 + x
2
)
d
dx
sin x
+ (1 x
2
) sin x
=
d
dx
[(1 + x
2
) cos x] + (1 x
2
) sin x
= 2xcos x (1 + x
2
) sin x + (1 x
2
) sin x
= 2xcos x 2x
2
sin x.
Now suppose we want to nd the solution y(x) of the equation (1) on the interval [a, b]
with the boundary conditions
y(a) + y
(a) = 0
y(b) + y
(b) = 0
, (3)
where Greek letters are real numbers.
1
First we should construct the Green function G(x, t) corresponding to the operator L
and the boundary conditions (3). Here is the recipe for doing this.
Solve the homogeneous equation L[y] = 0 and nd two linearly independent solutions
u(x) and v(x) such that
u(a) + u
(a) = 0
v(b) + v
(b) = 0
. (4)
In other words u(x) should fullls the boundary condition at a and v(x) should fullls
the boundary condition at b
1
.
Determine the constant
2
A from the following equation
A = p(t)
u(t) u
(t)
v(t) v
(t)
. (5)
Using A, u(x), and v(x) construct the Green function G(x, t) as follows
G(x, t) =
1
A
u(t)v(x), t < x b
1
A
u(x)v(t), a x < t
, (6)
and since we dont need to stress the interval [a, b] explicitly, the above Greens func-
tion can be re-written as
G(x, t) =
1
A
u(t)v(x), t < x
1
A
u(x)v(t), x < t
. (7)
Then based on the theorem which is proved on pages 665 and 666, the solution y(x) can
be determined from the following relation
y(x) =
b
a
G(x, t) f(t) dt. (8)
Example 2. Using Greens function method, solve the following boundary value problem
+
2
y = x
y(0) = y(1/4) = 0
.
1
If u(x) and v(x) failed to exist, this method would not work.
2
According to Abels formula A would actually b be a constant.
2
Solution. First, comparing y
+
2
y = x to the operator L dened in (1) and the equation
(2), it is easy to see that p(x) = 1, q(x) =
2
, and f(x) = x. Second, taking a = 0, b = 1/4,
= 1, = 0, = 1, and = 0 reduces the boundary conditions (3) to y(0) = y(1/4) = 0.
Therefore, the given problem is exactly one of the problems that can be solved by Greens
function method.
We know that the general solution to the equation y
+
2
y = 0 is
y(x) = M sin x + N cos x. (9)
So sin x and sin x cos x can play the role of u(x) and v(x) respectively, since they are
linearly independent solutions of the corresponding homogeneous equation and moreover,
sin x fullls the boundary condition at 0 and sin xcos x fullls the boundary condition
at 1/4.
Now we can go to the second step and determine the constant A from the equation ().
A =
sin x cos x
sin x cos x cos x + sin x
1/4
0
t G(x, t) dt
=
1
(sin x cos x)
x
0
t sin t dt +
1
sin x
1/4
x
t (sin t cos t) dt.
(11)
Integration by parts yields
t sin t dt =
sin t t cos t
2
+ C,
and
t cos t dt =
t sin t + cos t
2
+ C.
3
Using these in equation (11) and manipulating the expressions a little bit, one ends up with
the following solution
y(x) =
x
2
4
2
sin x.
4