Physics 210 Homework #3

Hanwen Qin

November 21, 2016

1 Cyclotron Motion
Part (a) The Hamiltonian I don’t need the speed of light c if I have magnetic field B in teslas and
electric charge e in coulombs. In Cartesian coordinates, the Lagrangian is
m 2 eB
L(x, y, ẋ, ẏ) = (ẋ + ẏ 2 ) − (xẏ − ẋy) (1.1)
2 2
Planar polar coordinates (r, θ) are defined such that x = r cos θ, y = r sin θ, so the velocities are
(
ẋ = ṙ cos θ − r sin θ θ̇
ẏ = ṙ sin θ + r cos θ θ̇

Substituting into (1.1),

m 2 eB 2
(ṙ + r2 θ̇2 ) −
L(r, ṙ, θ̇) = r θ̇ (1.2)
2 2
To move over to the Hamiltonian formalism, I first compute the conjugate momenta,
∂L
pr = = mṙ
∂ ṙ
∂L eB 2
pθ = = mr2 θ̇ − r
∂ θ̇ 2
Then I take the Legendre transform of the Lagrangian (1.2),

m 2 mr2 2
H = pr ṙ + pθ θ̇ − L = ṙ + θ̇
2 2
The last step is to eliminate the velocities in favor of the momenta, and write the Hamiltonian as an
explicit function of the phase space variables,

p2r p2θ eBpθ e2 B 2 r 2

H(r, pr , pθ ) = + + +
2m 2mr2 2m 8m

Part (b) Circular Orbits For circular orbits the easiest thing to use is Newton’s Second Law. Because
the Lorentz force F = evB = eωr0 B provides the centripetal acceleration a = ω 2 r0 , I immediately know
the orbital frequency
eB
ω=
m
As for the orbital radius, in terms of the kinetic energy E = (1/2)mv 2 = (1/2)mω 2 r02 , I have
r √
2E 2mE
r0 = =
mω 2 eB

1
Part (c) Perturbations on Circular Orbits When the orbit is not circular is when the advantage
of the Hamiltonian approach shows. For θ is ignorable and therefore pθ is conserved. It helps then to
consider an effective potential,

p2θ eBpθ e2 B 2 r2
V eff (r) = + +
2mr2 2m 8m
in which the radial coordinate r moves. As long as pθ > 0, V eff always has a global minimum r0 that
corresponds to a stable circular orbit.
r
0 p2θ e2 B 2 r 2pθ
V (r0 ) = − 3 + = 0 → r0 =
mr 4m eB
If the perturbation ρ is small, I may Taylor-expand the effective potential about the minimum,
1
V eff (r0 + ρ) = V (r0 ) + V 0 (r0 ) ρ + V 00 (r0 ) ρ2 + O(ρ3 )
2
The first term is constant. The second term is zero. The third term is a harmonic oscillator potential
with the “force constant”
3p2θ e2 B 2 e2 B 2
V 00 (r0 ) = + =
mr04 4m m
So the frequency of small oscillations is
r
V 00 eB
ωρ = =
m m

This is identical to the orbital frequency. So a small oscillation in the radial direction coupled with
uniform circular motion causes a small overall translation of the circular orbit.

2 Spinning Disk
Part (a) Inertia Tensor I am given the inertia tensor with respect to the center of mass. To shift
the reference point to the edge of the disk, I invoke the tensor form of the parallel-axis theorem,
   2   
2 1 0 0 R 0 0 2 5 0 0
MR  MR 
Idisk = 0 1 0 + M  0 0 0  = 0 1 0
4 2 4
0 0 2 0 0 R 0 0 6

Then I separately compute the bead’s inertia tensor about the same reference point,
 2
−R2
  
R 0 2 1 −1 0
5M R 
Ibead = m −R2 R2 0 = −1 1 0
2 4
0 0 2R 0 0 2

Knowing the moment of inertia is linear in mass distribution, I add up the bead to the disk.
 
5/2 −5/4 0
I = Idisk + Ibead = M R2 −5/4 3/2 0
0 0 4

2
Part (b) Principal Axes and Principal Moments Lamentably, the bead breaks enough symmetries
that the x, y axes are no longer principal axes. The z axis, however, is still a principal axis, with principal
moment
I3 = I33 = 4M R2
To find the other two axes, I only need to diagonalize the top-left 2 × 2 block of the inertia tensor.
 √ ! √
 29 2 0 (−2 + 29)x̂ + 5ŷ
I1 = 2 −


 MR x̂ = p √


 4 58 − 4 29
√! √
(−2 − 29)x̂ + 5ŷ

 29 0
M R2

 I2 = 2+ ŷ = √

 p
4

58 + 4 29

Part (c) Force Needed for Uniform Rotation Given a constant angular velocity ω
~ = ω ŷ, I can
decompose it in the body system,
5ω

ω10 = ω
 ~ · x̂0 = p √
58 + 4 29




5ω
ω20 = ω
~ · ŷ 0 = p √



 58 − 4 29
 0
ω3 = ω~ · ẑ 0 = 0

The inertia tensor is constant and diagonal in the body frame, the diagonal elements being the principal
moments I just computed. So the angular velocity in the body frame is

~ = I1 ω10 x̂0 + I2 ω20 ŷ 0

~ = Iω
L

Although it appears constant, the body frame is rotating with angular velocity ω
~ , so the rate of change
of angular momentum in the inertial frame is

~
dL ~ = (I2 − I1 )ω10 ω20 ẑ 0
~ ×L
=ω
dt
However, ẑ 0 is not constant either—it rotates about the y axis at angular velocity ω
~ . Simple trigonometry
tells me ẑ 0 = cos ωt ẑ + sin ωt x̂. On the other hand, the torque exerted on the disk about point A is
~ = ~rB × F~ = 2RFz x̂ − 2RFx ẑ
N
~ = dL/dt,
The law of angular momentum says N ~ so I can equate

(I − I1 )ω10 ω20

Fx = − 2
 cos ωt
2R
0 0
F = (I2 − I1 )ω1 ω2 sin ωt

z
2R
while Fy is undetermined. Plugging in the numbers now, I obtain

5
F~ = M ω 2 R(− cos ωt x̂ + sin ωt ẑ)
8

3
3 Bead on a Hoop
Part (a) Equations of Motion The stable equilibrium of the system is when the hoop hangs straight
down, with the bead at the bottom of the hoop (Figure 1). Define angular coordinates θ1 , θ2 that
measure small deviations from equilibrium (Figure 2). The hoop’s center of mass is at
(
x1 = a sin θ1
y1 = −a cos θ1

and the bead is at (

x2 = x1 + a sin θ2
(3.1)
y2 = y1 − a cos θ2

Parallel-axis theorem tells me the hoop’s rotational inertia about the pivot, I = M a2 + M a2 = 2M a2 .
Therefore the kinetic energy of the hoop is
I 2
T hoop = θ̇ = M a2 θ̇12
2 1
The kinetic energy of the bead is more complicated. I need to first differentiate (3.1),
(
ẋ2 = a cos θ1 θ̇1 + a cos θ2 θ̇2
ẏ2 = a sin θ1 θ̇1 + a sin θ2 θ̇2

and then compute the quadratic form

m ma2 h 2 i
T bead = (x˙2 2 + y˙2 2 ) = θ̇1 + θ̇22 + 2 cos(θ1 − θ2 ) θ̇1 θ̇2
2 2
As for potential energy, since only gravity is acting,

V hoop = M gy1 = −M ga cos θ1

V bead = mgy2 = −mga(cos θ1 + cos θ2 )

Finally, the Lagrangian of the hoop-bead system is

L(θ1 , θ2 , θ̇1 , θ̇2 ) = T hoop + T bead − V hoop − V bead

 m  2 2 ma2 ˙ 2
= M+ a θ̇1 + θ2 + ma2 cos(θ1 − θ2 ) θ̇1 θ̇2
2 2
+ (M + m)g cos θ1 + mga cos θ2

The Euler-Lagrange equations are


m h
2
i M +m g
 θ̈1 = − 2M + m sin(θ1 − θ2 ) θ̇2 + cos(θ1 − θ2 ) θ̈2 − 2M + m a sin θ1




g
 θ̈2 = sin(θ1 − θ2 ) θ̇12 + cos(θ1 − θ2 ) θ̈1 − sin θ2



a

4
 y y

x x
a
θ1
M M
a
θ2

m
m

Figure 1: Stable Equilibrium Figure 2: Generalized Coordinates

Part (b) Small Oscillations For θ1 , θ2  1, the equations of motion are linear.

m M +m g

θ̈1 = −
 θ̈2 − θ1 + O(θ2 )
2M + m 2M + m a
θ̈ = θ̈ − g θ + O(θ2 )

2 1 2
a
To discover normal modes, I take a Fourier transform, and put the linear, algebraic equations in matrix
form,
M +m g
" m
#" #
ω 2 − 2M +m a 2M +m ω 2 θ̂1
2 2 g
=0 (3.2)
ω ω −a θ̂2

For non-trivial solutions θ̂1 , θ̂2 to exist, the matrix of coefficients A must be singular:
2M 3M + 2m g 2 M + m  g 2
det A = ω4 − ω + =0
2M + m 2M + m a 2M + m a

This equation is quadratic in ω 2 , so the solutions are given by the quadratic formula, and I discover the
normal-mode frequencies
r r
m g g
ω= 1+ or
M a 2a
p
For the higher frequency ω = (1 + (m/M ))(g/a), equation (3.2) becomes

(M +m)2 g
" (M +m)m g
#" #
M (2M +m) a M (2M +m) a
θ̂1
=0
M +m g m g θ̂2
M a M a

By inspection θ̂1 = m, θ̂2 = −(M + m) is a solution. In this mode the hoop andpthe bead oscillate out
of phase, and the bead always has larger amplitude. For the lower frequency ω = (g/2a), the equation
to solve is
− 2(2Mm+m) ag 2(2Mm+m) ag θ̂1
" #" #

g g
=0
2a − 2a θ̂2

Obviously θ̂1 = 1, θ̂2 = 1 is a solution. In this mode the hoop and the bead oscillate in phase with equal
amplitude.