Lesson 15 Potential Theory Using Complex Analysis
Lesson 15 Potential Theory Using Complex Analysis
Introduction
“Potentials” in physics can simplify the derivation of forces. They are typically described by
solutions to Laplace’s equation ∇2Φ=0. The solutions are called “harmonic functions” if they
The real and imaginary parts of a complex analytic function F(z=x+iy)=Φ(x,y)+iΨ(x,y) are
satisfied with 2-D Laplace’s equation: ∇2Φ=∇2Ψ=0 (proved by CR-conditions, Lesson 9). If
Φ(x,y) represents the potential function, by working with F(z), we can: (1) handle both
equipotential lines (Φ=constant) and lines of force (Ψ=constant) simultaneously; (2) solve
Complex Potential
Examples of real potentials
1) Parallel plates: The electrostatic potential Φ between two parallel conducting plates is
d d
governed by: Laplace’s equation: ∇2Φ=Φ"(x)=0, and BCs: Φ − =Φ1, Φ =Φ2.
2 2
Φ(x)=ax+b (15.1)
Φ1 − Φ 2 Φ + Φ2
where a= − , b= 1 . The equipotential line Φ=Φ0, is a vertical line x=x0
d 2
v
parallel to the plates. The E-field is E = − ∇Φ , ⇒
v v
E = − ax (15.2)
Note: Infinite dimension (along y-axis) causes constant potential (Φ is independent of y).
where a, b are determined by BCs [Φ(r1)=Φ1, Φ(r2)=Φ2]. The equipotential line Φ=Φ0 is
v
a circle r=r0. The E-field is E = − ∇Φ , ⇒
v av
E =− r (15.4)
r
which is in radial direction, perpendicular to the equipotential lines.
3) Angular region: If the region of interest is confined by two plates in the radial directions
and with an included angle α, it is difficult to directly solve the potential Φ by traditional
equation for it is the imaginary part of an analytic function: F(z)= Ln z =ln |z|+i⋅Arg(z) [eq.
(9.7)]. By (1-2), ⇒
Φ=a+bθ (15.5)
Φ1 + Φ 2 Φ − Φ1
where a= , b= 2 . The equipotential line Φ=Φ0 is a ray θ =θ 0. The E-field
2 α
v
is E = − ∇Φ , ⇒
v bv
E =− θ (15.6)
r
For a given real potential Φ(x,y), we can uniquely (except for an additive constant) determine
is analytic. As a result, F(z) maps curves in the xy-plane onto curves in the ΦΨ-plane
“conformally” or vise versa, i.e. included angle is preserved during mapping (Appendix 9A).
Since Φ=Φ0 (vertical line) is always perpendicular to Ψ=Ψ0 (horizontal line) in the ΦΨ-plane,
⇒ the corresponding curves in the xy-plane: Φ(x,y)=Φ0 (equipotential line), and Ψ(x,y)=Ψ0
Since gradient defines the steepest ascent/descent direction, which is always perpendicular
v
with the equipotential lines (zero-variation direction), ⇒ E-field E = −∇Φ is perpendicular
Ψ=ay+c (15.7)
2) Coaxial cylinders: By eq. (15.3), Φ=aln r+b =aln |z|+b. By CR-equations, its conjugate is:
Ψ=aArg(z)+c (15.8)
0 and points on the negative real axis. The E-field lines are Ψ=constant, ⇒
3) Angular region: By eq. (15.5), Φ=a+bθ =a+bArg(z), which is the imaginary part of
The E-field lines are Ψ= constant, ⇒ |z|=r =constant, same as eq. (15.6).
Find an analytic transformation function f(z) to map a complicated domain D (in the z-plane)
onto a simpler domain D* (in the w-plane), where the complex potential F*(w) can be easily
solved [Re{F*(w)}=Φ*(w) is satisfied with 2-D Laplace’s equation and boundary conditions
in the w-plane]. Then the complex potential in the z-plane is derived by inverse transform:
The strategy works because harmonic functions remain harmonic under conformal mapping.
Proof: f(z) and F*(w) are analytic (s.t. Re{F*(w)} is harmonic). By the chain rule, F(z) is also
dF *
analytic: F ′(z ) = ⋅ f ′( z ) w=f(z) exists. ⇒ Φ(x,y)=Re{F(z)} is harmonic.
dw
2 2
E.g. Non-coaxial cylinders: C1: |z|=1, C2: z − = ; U1=0, U2=110.
5 5
2z − 1
The equipotential lines Φ(x,y)=Φ0, ⇒ =constant, are circles (with different centers) in
z−2
the z-plane (see plot); corresponding to concentric circles in the w-plane. The lines of force
are circular arcs (see plot), corresponding to rays Arg(w)=constant in the w-plane.