CONFORMAL MAPPING

In a limited group of problems one can use a short cut to the solution of the
Laplace's equation, conformal mapping. The problem has to be 2D (two
dimensional), i.e., the boundaries, boundary conditions and sources contain only
two variables, x and y say.
Some examples where this method can be used is: the electrostatic potential;
heat conduction; flow of fluids.
We will of course focus on the electrostatic potential. The idea is that the
problem at hand with a certain geometry may be mapped into a problem with
simpler geometry or with a geometry that we have already solved.
For polygonal boundaries a general method of constructing the mapping is
offered by Schwartz-Cristoffel transformations. For non-polygonal boundaries
one must rely on "dictionaries" of mappings. One such dictionary with 30
different mappings is found in a student text book by R. V. Churchill
1
.
Another dictionary is found in a book by K. J. Binns and P. J. Lawrenson
2
.
For further reading about conformal mapping see W. R. Smythe
3
.
1
R. V. Churchill: Complex variables and applications, 2nd Edition, McGraw-
Hill, New York 1960.
2
K. J. Binns and P. J. Lawrenson, Analysis and Computation of Electric and
Magnetic Field Problems, 2nd Edition, Pergamon Press, New York 1973.
3
W. R. Smythe, Static and Dynamic Electricity, 3rd Edition, Taylor&Frances
1989.
Bo E. Sernelius 5:2
The conformal mapping relies on the properties of analytic functions. Thus we
need to refresh our knowledge of these.
The derivative of a complex valued function of complex variables
f z
f z f z
'
0
0 0
( )
=
+
( )

( )

lim
z 0
z
z

Let z x iy = + , and f z u z iv z u x y iv x y ( ) = ( ) + ( ) = ( ) + ( ) , , .
x and y real valued. u and v are real valued functions of the real valued
variables x and y.
One can show that:
1) If the derivative of f(z) exists at a point z then the partial derivatives of u
and v exist at that point and obey the following conditions:

u
x
v
y
u
y
v
x
= = and The Cauchy-Riemann conditions
2) Let u and v be real and single valued functions of x and y which,
together with their partial derivatives of the first order, are continuous at a
point. If those partial derivatives satisfy the Cauchy-Riemann conditions at that
point, then the derivative of f exists at that point.
Analytic Functions
A function f(z) is analytic at a point z
0
if its derivative f'(z) exists not only at
z
0
but at every point z in a neighborhood of z
0
.
One can show that if f(z) is analytic the partial derivatives of u and v of all
orders exist and are continuous functions of x and y.
Thus we have
Bo E. Sernelius 5:3

2
2
2 2 2
2
u
x
v
x y
v
y x y
v
x y
u
y
u
y
= = =

_
,
=

_
,

=
or

2
2
2
2
0
u
x
u
y
+ =
and in an analogous way we get

2
2
2
2
0
v
x
v
y
+ =
Thus both u(x,y) and v(x,y) satisfy Laplace's equation.
Harmonic functions
Any function that has continuous partial derivatives of the second order and
that satisfies the Laplace's equation is called a Harmonic function.
Thus both the real part, u, and imaginary part, v, of f are harmonic functions.
If the function f = u+iv is analytic, u and v are conjugate harmonic
functions. Given one of two harmonic functions, the Cauchy-Riemann
equations can be used to find the other.
That u and v are harmonic functions can be used to find the potential. Both
u and v are candidates for the potential since they satisfy Laplace's equation.
Alternatively the function f can be viewed as a change of variables, a
transformation from the complex z-plane to the complex -plane.
z x iy
u iv
= +
= +
;

Bo E. Sernelius 5:4
iy
x
z-plane iv
u
-plane
One can show that if the function f is analytic at a point z=z
0
, where
f'(z
0
)0, there exists a neighborhood of the point
0
in the -plane in
which the function
= ( ) f z
has a unique inverse
z F = ( )
The functions f and F defines a change of variables from (x,y) to (u,v) and
from (u,v) to (x,y), respectively.
y
x
xy-plane v
u
uv-plane
When this change of basis is defined through an analytic function all curves in
the xy-plane that cross each other at an angle are mapped into curves in the
uv-plane that cross each other at exactly the same angle. This is why the
mapping is called conformal. The curves can be the boundaries or equi-
potential curves, electric field lines or any curves.
In particular since the set of curves that are horizontal in the xy-plane are
Bo E. Sernelius 5:5
perpendicular to the set of curves that are vertical the two sets will be mapped
into perpendicular sets in the uv-plane. Of course the opposite is also true. To
be noted is that in points where f or its inverse is not analytical the
transformation is not conformal.
Transformation of boundary conditions.
Two types of boundary conditions remain unchanged by the transformation
and problems with these boundary conditions are best suited for conformal
mapping. The types are:
1) The value of a harmonic function is constant on the boundary.
(one type of Dirichlet problem)
2) The normal derivative of a harmonic function is zero on the boundary.
(one type of Neumann problem)
(The conditions are that f and its inverse are analytical)
Bo E. Sernelius 5:6
Further properties
Let the real part of the function f(x,y) be the actual potential of our problem.
Then the imaginary part is called the stream function. Its level curves are
called stream lines. The name comes from the analogy in fluid flow.
f x y u x y iv x y x y i x y , , , , , ( ) = ( ) + ( ) = ( ) + ( )
In electrostatic problems these stream lines are parallel to the electric field. We
furthermore have
E z E x y iE x y
df z
dz
x y
( ) = ( ) + ( ) =
( )
, ,
*
or
E x y
u x y
x
E x y
u x y
y
x
y
,
,
,
,
( ) =
( )
( ) =
( )

Example 1: The potential from a line charge, of charge density

l
, along the
z-axis is r r z
l l
( ) = ( ) = ( )
[ ]
2 2 ln Re ln . The electric field calculated in the
two different ways is:
( )
[ ]

= = =
+
+
+

1
]
1
d
dz
z
z
z
z
x
x y
i
y
x y
l l l l
2 2
1
2 2
2 2 2 2 2
ln
*
*
E
x
x y
E
y
x y
x l y l
=
+
=
+
2 2
2 2 2 2
;
and
E x y
u x y
x
x y x
x
x y
E x y
u x y
y
x y y
y
x y
x l l
y l l
,
,
ln
,
,
ln
( ) =
( )
= +

_
,

=
+
( ) =
( )
= +

_
,

=
+

2 2
2 2
2 2
2 2
2 2
2 2
Bo E. Sernelius 5:7
Example 2: The capacitance between two conductors
If Q is the electrical charge on either conductor and V is the difference
in potential between one conductor and the other, then the capacitance C is
defined to be
C
Q
V
=

In two-dimensional problems we calculate the capacitance per unit length, c,

whose cross section is typically displayed in the complex plane. Let in this case
Q
L
be the charge per unit length of one of the conductors. Then we have
c
Q
V
L
=

Theorem: The electrical charge per unit length on a conductor that belongs to
a charged two-dimensional configuration of conductors is
Q z
L
= ( ) 4
where is the permittivity, or dielectric constant, of the surrounding material
and z ( ) is the decrement (initial value minus final value) of the stream
function as we proceed in the positive direction once around the boundary of
the cross section of the conductor in the complex plane.
Usually (z) will be a multivalued function defined by means of a branch cut.
Thus (z) does not return to its original value when we encircle the
conductor, and thus 0. We get:
c
V
=

4

Theorem: The capacitance of a two-dimensional system of conductors is

unaffected by a conformal transformation of its cross section.
Bo E. Sernelius 5:8
EXAMPLES
We will just give a few examples of conformal mapping.
The Linear Fractional Transformation (Mbius transformation)

=
+
+
( )
=
+

az b
cz d
ad bc
z
d b
c a
, ; 0
where a, b, c and d are complex constants. Each point in the z-plane,
except z = -d/c, has a unique image point in the -plane. And each point in
the -plane, except = a/c, has a unique image point in the z-plane. We
may include the points z = and = and then have a one to one
correspondence between all points in the z- and -planes.
This transformation always transforms circles and lines into circles and lines.
There is just one bilinear transformation that maps three given distinct points
z
1
, z
2
, z
3
into three specified distinct points
1
,
2
,
3
, respectively.
Find one transformation that maps the unit circle and its interior on the real
axis and the whole upper half plane.
iy
x
iv
u
We let
Bo E. Sernelius 5:9
z = i be mapped into = , z = - i be mapped into = 0, and
z = 0 be mapped into a point with v > 0
The first point guarantees that the circle transforms into a straight line, the
second that the line passes through the origin and the third that it maps the
interior of the circle to the upper half-plane and not the lower.
The following transformation has the wanted properties:
=
+ ( )
( )
i
z i
z i
The inverse transformation is
z i
i
i
=

+

It is useful to introduce the cross ratio of four points:

z z z z
z z z z
z z z z
1 2 3 4
1 2 3 4
1 4 3 2
, , ,
( )
=
( )
( )
( )
( )
Theorem:
The cross ratio is invariant under the mbius transformation :


1 2 3 4
1 4 3 2
1 2 3 4
1 4 3 2
( )
( )
( )
( )
=
( )
( )
( )
( )
z z z z
z z z z
This is very useful when we want to find the transformation that maps three
specific points, z
1
, z
2
, z
3
onto three specific images,
1
,
2
,
3
. Let the
forth point be taken as the general point:


1 2 3
1 3 2
1 2 3
1 3 2
( )
( )
( )
( )
=
( )
( )
( )
( )
z z z z
z z z z