Professional Documents
Culture Documents
Method of Green's Functions: 1 Preliminary Ideas and Motivation
Method of Green's Functions: 1 Preliminary Ideas and Motivation
Matthew J. Hancock
Fall 2006
where f is continuous at x = a. The last is called the sifting property of the -function.
To make proofs with the -function more rigorous, we consider a -sequence, that
is, a sequence of functions that converge to the -function, at least in a pointwise
sense. Consider the sequence
n 2
n (x) = e(nx)
Note that
2n 2
Z Z Z
(nx)2 2
n (x) dx = e dx = ez dz = erf () = 1
0 0
for any scalar function G and vector valued function F. Setting F = u gives what
is called Greens First Identity,
Z Z Z
2
G u + u G dA = G (u n) dS (2)
D C
2 u = F in D, (4)
u = f on C.
2 G = ( x, y) = (r) in D, (5)
G = 0 on C.
2
The notation (r) is short for ( x, y). Substituting (4) and (5) into Greens
second identity (3) gives
Z Z Z
u (x, y) GF dA = f G ndS
D C
Rearranging gives Z Z Z
u (x, y) = GF dA + f G ndS (6)
D C
Therefore, if we can nd a G that satises (5), we can use (6) to nd the solution
u (x, y) of the BVP (4). The advantage is that nding the Greens function G depends
only on the area D and curve C, not on F and f .
Note: this method can be generalized to 3D domains.
Hence
1
v (r) = ln r
2
3
This is called the fundamental solution for the Greens function of the Laplacian on
2D domains. For 3D domains, the fundamental
q solution for the Greens function of
the Laplacian is 1/(4r), where r = (x )2 + (y )2 + (z )2 .
The Greens function for the Laplacian on 2D domains is dened in terms of the
corresponding fundamental solution,
1
G (x, y; , ) = ln r + h,
2
h is regular,
2 h = 0, (, ) D,
G = 0 (, ) C.
2 h = 0 (, ) D,
1
h = ln r (, ) C.
2
The denition of G in terms of h gives the BVP (5) for G. Thus, for 2D regions D,
nding the Greens function for the Laplacian reduces to nding h.
2.2 Examples
Ref: Myint-U & Debnath 10.6
(i) Full plane D = R2 . There are no boundaries so h = 0 will do, and
1 1
ln ( x)2 + ( y)2
G= ln r =
2 4
(ii) Half plane D = {(x, y) : y > 0}. We nd G by introducing what is called an
image point (x, y) corresponding to (x, y). Let r be the distance from (, ) to
(x, y) and r the distance from (, ) to the image point (x, y),
q q
2 2
r = ( x) + ( y) , r = ( x)2 + ( + y)2
We add
1 1
q
h= ln r = ln ( x)2 + ( + y)2
2 2
to G to make G = 0 on the boundary. Since the image point (x, y) is NOT in D,
then h is regular for all points (, ) D, and satises Laplaces equation,
2 2h 2h
h= 2 + 2 =0
4
0
0.1
0.2
G(21/2,21/2;,)
0.3
0.4
0.5
0.6 5
0.7
0
0
2
4
6
8
10 5
Figure 1: Plot of the Greens function G (x, y; , ) for the Laplacian operator in the
upper half plane, for (x, y) = ( 2, 2).
1 1 1 1 r 1 ( x)2 + ( y)2
G= ln r + h = ln r ln r = ln = ln (7)
2 2 2 2 r 4 ( x)2 + ( + y)2
G (x, y; , ) is plotted in the upper half plane in Figure 1 for (x, y) = 2, 2 .
Note that G as (, ) (x, y). Also, notice that G < 0 everywhere and
G = 0 on the boundary = 0. These are, in fact, general properties of the Greens
function. The Greens function G (x, y; , ) acts like a weighting function for (x, y)
and neighboring points in the plane. The solution u at (x, y) involves integrals of
the weighting G (x, y; , ) times the boundary condition f (, ) and forcing function
F (, ).
On the boundary C, = 0, so that G = 0 and
G 1 y
G n = =
=0 ( x)2 + y 2
The solution of the BVP (6) with F = 0 on the upper half plane D can now be
written as, from (6),
y f ()
Z Z
u (x, y) = f G ndS = d,
C ( x)2 + y 2
which is the same as we found from the Fourier Transform, on page 13 of fourtran.pdf.
(iii) Upper right quarter plane D = {(x, y) : x > 0, y > 0}. We use the image
points (x, y), (x, y) and (x, y),
1 1
q q
2 2
G = ln ( x) + ( y) ln ( x)2 + ( + y)2 (8)
2 2
1 1
q q
ln ( + x)2 + ( y)2 + ln ( + x)2 + ( + y)2
2 2
For (, ) C = D (the boundary), either = 0 or = 0, and in either case, G = 0.
Thus G = 0 on the boundary of D. Also, the second, third and fourth terms on the
r.h.s. are regular for (, ) D, and hence the Laplacian 2 = 2 / 2 + 2 / 2 of each
of these terms is zero. The Laplacian of the rst term is (r). Hence 2 G = (r).
Thus (8) is the Greens function in the upper half plane D.
For (, ) C = D (the boundary),
Z 0 ! Z !
G G
Z
f G ndS = f (0, ) d + f (, 0) d
C =0 0 =0
Z Z
G G
= f (0, ) d f (, 0) d
0
=0 0
=0
Note that
G 4yx
=
=0 x2 + (y + )2 x2 + (y )2
G 4yx
=
=0 (x ) + y 2 (x + )2 + y 2
2
The solution of the BVP (6) with F = 0 on the upper right quarter plane D and
boundary condition u = f can now be written as, from (6),
Z
u (x, y) = f G ndS
C
4yx f (0, )
Z
= d
0 x2 + (y + )2 x2 + (y )2
4yx f (, 0)
Z
+
0 (x ) + y 2 (x + )2 + y 2
2 d
(iv) Unit disc D = {(x, y) : x2 + y 2 1}. By some simple geometry, for each
point (x, y) D, choosing the image point (x , y ) along the same ray as (x, y) and a
p
distance 1/ x2 + y 2 away from the origin guarantees that r/r is constant along the
circumference of the circle, where
q q
2 2
r = ( x) + ( y) , r = ( x )2 + ( y )2 .
6
[DRAW] Using the law of cosines, we obtain
r2 = 2 + 2 2 cos
2 2 1
r = + 2 2 cos
p p
where = x2 + y 2 , = 2 + 2 and , are the angles the rays (x, y) and (, )
make with the horizontal. Note that for (, ) on the circumference ( 2 + 2 = 2 = 1),
we have
2
r2 1 + 2 cos
= = 2 , = 1.
r 2
1 + 12 2 1 cos
Thus the Greens function for the Laplacian on the 2D disc is
2 2
1 r 1 + 2 cos
G (, ; x, y) = ln = ln
2 r 4 2 2 + 1 2 cos
Note that
1 2
G 1
G n = =
=1 2 1 + 2 2 cos
Thus, the solution to the BVP (5) on the unit circle is (in polar coordinates),
Z 2
1 1 2
u (, ) = f d
2 0 1 + 2 2 cos
2 2
1
Z 2 Z 1 + 2 cos
+ ln F ,
d
d
4 0 0 2 2 + 1 2 cos
This is called the Poisson integral formula for the unit disk.
then by the Riemann Mapping Theorem, there is a conformal map w (z) (analytic
and one-to-one) from D into the unit disk, which maps to the origin, w () = 0
and the boundary of D to the unit circle, |w (z)| = 1 for z D and 0 |w (z)| < 1
for z D/D. The Greens function G is then given by
1
G= ln |w (z)|
2
To see this, we need a few results from complex analysis. First, note that for z D,
|w (z)| = 0 so that G = 0. Also, since w (z) is 1-1, |w (z)| > 0 for z =
. Thus, we
n
can write w (z) = (z ) H (z) where H (z) is analytic and nonzero in D. Since
w (z) is 1-1, |w (z)| > 0 on D. Thus n = 1. Hence
w (z) = (z ) H (z)
and
1
G= ln r + h
2
where
q
r = |z | = ( x)2 + ( y)2
1
h = ln |H (z)|
2
Since H (z) is analytic and nonzero in D, then (1/2) ln H (z) is analytic in D and
hence its real part is harmonic, i.e. h = ((1/2) ln H (z)) satises 2 h = 0 in D.
Thus by our denition above, G is the Greens function of the Laplacian on D.
Example 1. The half plane D = {(x, y) : y > 0}. The analytic function
z
w (z) =
z
maps the upper half plane D onto the unit disc, where asterisks denote the complex
conjugate. Note that w () = 0 and along the boundary of D, z = x, which is
equidistant from and , so that |w (z)| = 1. Points in the upper half plane (y > 0)
are closer to = x + iy, also in the upper half plane, than to = x iy, in the
lower half plane. Thus for z D/D, |w (z)| = |z | / |z | < 1. The Greens
function is
1 1 |z | 1 r
G= ln |w (z)| = ln = ln
2 2 |z | 2 r
which is the same as we derived before, Eq. (7).
2 u + u = F in D,
u = f on D,
2 u u = F in D,
u = f on D,