Problem 1.

A half submerged metal sphere (UIC comprehensive

exam)
A very light neutral hollow metal spherical shell of mass m and radius a is slightly submerged
by a distance b  a below the surface of a dielectric liquid. The liquid has mass density ρ
and electrical permitivity . The liquid sits in air which has negligble density ρo  ρ, and
the permitivity of air is approximately unity, air ' 1. The pressure at the air liquid interface
is p0 . Recall that stress tensor of an ideal fluid at rest is T ij = p(z)δ ij where p(z) is the
pressure as a function of z.

Air
b
Liquid
(a) Use the formalism of stress tensor to show that p(z) increases as p = p0 + ρgh, where
h = −z is the depth below the surface, z < 0. Here p0 is the pressure at the surface.
Hint: what is the net force per volume for a static fluid?

(b) Use the formalism of stress tensor to prove that the boyancy force (for any shape) equals
the difference in weight of the displaced fluid volume ∆V and the corresponding weight
of the air:
F = (ρ − ρo )g∆V ' ρg∆V.

Now a charge Q is added to the sphere, and the sphere becomes half submerged.

(c) Determine the potential, and the electrostatic fields E and D, in the top and lower
halves of the sphere. Verify that all the appropriate boundary conditions are satisfied.

(d) What is the surface charge density on the top and lower halves of the sphere?

(e) Determine the electrostatic attractive force as a function of Q, a, and . What must Q
be for the sphere to be half submerged? Make all reasonable approximations. Express
your approximate result in terms of ρ, g, a, . Use dimensional reasoning to show that
for a light sphere, p
Q = ρga5 × function of  . (1)

(f) (Optional) Estimate Q numerically for typical liquids.

1
Problem 2. A cylinder in a magnetic field (Jackson)
A very long hollow cylinder of inner radius a and outer radius b of permeability µ is placed
in an initially uniform magnetic field Bo at right angles to the field.

(a) For a constant field Bo in the x direction show that Az = Bo y is the vector potential.
This should give you an idea of a convenient set of coordinates to use.
Remark: See Wikipedia for a list of the vector Laplacian in all coordinates. Most often
the vector Laplacian is used if the current is azimuthal and solutions may be looked
for with Aφ 6= 0 and Ar = Aθ = 0 (or Aρ = Az = 0 in cylindrical coordinates). This
could be used for example in Problem 3. Similarly if the current runs up and down,
with Az 6= 0 and Aρ = Aφ = 0, so that B = (Bx (x, y, z), By (x, y, z), 0) is independent
of z, then the vector Laplacian in cylindrical coordinates −∇2 Az is a good way to go.

(b) Show that the magnetic field in the cylinder is constant ρ < a and determine its
magnitude.

(c) Sketch |B|/|Bo | at the center of the as function of µ for a2 /b2 = 0.9, 0.5, 0.1 for µ > 1.

2
Problem 3. Helmholtz coils (Jackson)
Consider a compact circular coil of radius a carrying current I, which lies in the x y plane
with its center at the origin.

(a) By elementary means compute the magnetic field along the z axis.

(b) Show by direct analysis of the Maxwell equations r · B = 0 and r ⇥ B = 0 that

slightly o↵ axis near z = 0 the magnetic field takes the form

Bz ' 0 + 2 z2 1 2
2
⇢ , B⇢ ' 2 z⇢ , (2)
⇣ ⌘
1 @ 2 Bzo
where 0 = (Bzo )and 2 = 2 @z 2
are the field and its z derivatives evaluated at the
origin. For later use give 0 and 2 explictly in terms of the current and the radius of
the loop.
Remark: The magnetic field near the origin satisfies r ⇥ B = 0, so r · B = 0. We
say it is harmonic function1 . Because the function is harmonic, the taylor series of B
on the z axis, is sufficient to determine the taylor series close to the z axis.

(c) Now consider a second identical coil (co-axial with the first), having the same magni-
tude and direction of the current, at a height b above the first coil, where a is the radii
of the rings. With the coordinate origin relocated at the point midway between the
two centers of the coils, determine the magnetic field on the z-axis near the origin as
an expansion in powers of z to z 4 . Use mathematica if you like. You should find that
the coefficient of z 2 vanishes when b = a
Remark For b = a the coils are known as Helmholtz coils. For this choice of b the z 2
terms in part (c) are absent. (Also if the o↵-axis fields are computed along the lines
of part (b), they also vanish.) The field near the origin is then constant to 0.1% for
z < 0.17 a.

1
This means that B can be written B = r where r2 =0

3
Problem 4. A magnetized sphere and a circular hoop
A uniformly magnetized sphere of radius a centered at origin has a permanent total magnetic
moment m = m ẑ pointed along the z-axis (see below). A circular hoop of wire of radius b
lies in the xz plane and is also centered at the origin. The hoop circles the sphere as shown
below, and carries a small current Io (which does not appreciably change the magnetic field).
The direction of the current Io is indicated in the figure.

Io

(a) Determine the bound surface current on the surface of the sphere.

(b) Write down (no long derivations please) the magnetic field B inside and outside the
magnetized sphere by analogy with the spinning charged sphere disucssed in class.

(c) Show that your solution satisfies the boundary conditions of magnetostatics on the
surface of the sphere.

(d) Compute the net-torque on the circular hoop. Indicate the direction and interpret.

4
 Solution

(a) The surface current is given by the discontinuity in the magnetization

Kb
= n ⇥ (Mout Min ) (8)
c
Then since Mout = 0 we have
Kb
= M (n ⇥ ẑ) (9)
c
where M = m/( 43 ⇡a3 ) is the magnetic dipole moment per volume. With n = r̂ and
ˆ + cos ✓r̂, we find
ẑ = sin ✓✓
Kb 3
= m sin ✓ ˆ (10)
c 4⇡a3

(b) A formal analogy with the charged spinning sphere gives the solution. In the spinning
sphere case the surface current was also / sin ✓ ˆ . In the spinning sphere case we
found that the magnetic field outside is one of a magnetic dipole, where all of magnetic
moment is placed at the origin
1
B= [3(m · r̂)r̂ m] (11)
4⇡r3
Or more explicitly
1 1 ˆ.
B= 3
2m cos ✓ r̂ + m sin ✓ ✓ (12)
4⇡r 4⇡r3
Inside sphere, the magnetic field was constant

B = Bo ẑ (13)

The constant Bo can be picked o↵ from the boundary conditions as we will do in the
next item.

(c) The boundary conditions read

n ⇥ (H2 H1 ) =0 (14)
n · (B2 B1 ) =0 (15)

Then from the second boundary condition at r = a

Br |out = Br |in . (16)

With the magnetic field outside the sphere

1
Br |out = 2m cos ✓ , (17)
4⇡r3

5
 and inside the sphere
r̂ · B|in = Bo r̂ · ẑ = Bo cos ✓ , (18)
comparison at r = a gives
1
Bo = 2m . (19)
4⇡a3
Thus we have shown that n · (B2 B1 ) = 0 is satisfied, provided Bo takes a specific
value.
Note that inside the sphere with M = m/(4⇡a3 /3) and the inductance is
m
Hz,in = Ho = Bo M= . (20)
4⇡a3
Now one can verify the parallel boundary condition n ⇥ (H2 H1 ) = 0 using eq. (20)
✓ ◆
1
H✓,out H✓,in |r=a = m sin ✓ + Ho sin ✓ =0 (21)
4⇡r3 r=a

(d) To compute the torque we first compute the lorentz force on a element of length
d` = bd✓.
Io
dF = d`B? (22)
c
Io
= bd✓ Br (23)
c
Io 2m cos ✓
= bd✓ (24)
c 4⇡b3
The right hand rule indicates that the force is in the ŷ direction in the upper hemi-
sphere, and in the positive ŷ direction in the lower hemisphere. This implies that the
net torque points along the x-axis. This can be intuited by noting that the magnetic
moment of the hoop tends to align with the magnetic field from the sphere

(e) The torque around the x-axis

Z Z
⌧= d⌧ = b cos ✓ dF (25)
Z ⇡
Io 2m cos ✓
=2 bd✓
b cos ✓ (26)
0 c 4⇡b3
Z
4m(Io /c)b2 ⇡
= d✓ cos2 ✓ (27)
4⇡b3 0
4m(Io /c)b2 ⇡
= (28)
4⇡b3 2
2m Io 2
= ⇡b (29)
4⇡b3 c

6
Problem 5. Electrodes in an ohmic material filling half of space
Two small spherical electrodes of radius a are embedded in a semi-infinite medium of con-
ductivity σ, each at a distance d  a from the plane face of the medium and at a distance
b  a from each other.

positive lead

negative lead

(a) State the boundary conditions on all surfaces. Assume that the electrodes emit and
absorb a total current I which is spread uniformly over the surface of the sphere. Do
not assume that a is small for this part (but of course a < b and a < d).

(b) Argue for small a the potential between the two electrodes satisfies

I 3 I
∇2 ϕ = δ (r − r1 ) − δ 3 (r − r2 ) (3)
σ σ
where r1 is the position of the emitting electrode while r2 is the position of the ab-
sorbing electrode.

(c) Find the resistance between the electrodes. Sketch the flow lines of current if the two
electrodes are held at a potential difference ∆V . (Hint: use images to solve Eq. (3)
with the right boundary conditions.)