File

PHY 5246: Theoretical Dynamics, Fall 2015
September 2nd , 2015

Assignment # 2
(Graded problems are due Wednesday September 9th , 2015)
1 Graded problems
1. A bead of mass m slides without friction in a uniform gravitational field on a vertical circular
hoop of radius R. The hoop is constrained to rotate at a fixed angular velocity about its
vertical diameter. Let be the position of the bead on the hoop measured from the lowest
point.

(1.a) Write down the Lagrangian L(, ).
(1.b) Find how the equilibrium values of depends on . Which are stable, which unstable?
(1.c) Find the frequencies of small oscillations about the stable equilibrium positions. Say
something concerning the motion about the only stable equilibrium point when =
q
g/R.
2. Consider a stretchable plane pendulum, that is, a mass m suspended from a spring of spring
constant k and unstretched length l, constrained to move in a vertical plane. Write down the
Lagrangian and obtain the Euler-Lagrange equations. Interpret the form of the equations
in terms of the forces acting on the system.
3. Two points of mass m are joined by a rigid weightless rod of length l, the center of which is
constrained to move on a circle of radius a. Express the kinetic energy in terms of generalized
coordinates.
4. Two mass points of mass m1 and m2 are connected by a string passing through a hole in a
smooth table so that m1 rests on the table surface and m2 hangs suspended. Assuming m2
moves only in a vertical line, what are the generalized coordinates for the system? Write
the Lagrange equations for the system and, if possible, discuss the physical significance each
one of them might have. Reduce the problem to a single second-order differential equation
and obtain a first integral of the equation. What is its physical significance? (Consider the
motion only until m1 reaches the hole.)
2 Non-graded suggested problems

Any problem in Chapter 1 of Goldsteins book will be a very good exercise. In particular solutions
will be posted for Problems 15, 18, and 22.
Assignment # 2, Solutions
1 Graded Problems
Problem 1
(1.a)
The coordinates can be written using spherical coordi-
nates as:
z x = R sin cos(t) , (1)

t y = R sin sin(t) ,
z = R cos ,
where we have used the constraints: r = R and = t

(i.e. = ), and reduced the number of generalized coor-
The problem
dinates to just one () and its velocity ().
is therefore equivalent to a one-dimensional problem. No-
y tice that the constraints are time dependent, which hint
R to the fact that the mechanical energy of the system may
not be conserved. The kinetic and potential energy of the
system are

1
x T = m(R2 2 + R2 sin2 2 ) , (2)
2
V = mgR cos ,
and the Lagrangian of the system is

1
L = T V = m(R2 2 + R2 sin2 2 ) + mgR cos . (3)
2
The Euler-Lagrange equation of motion is
d L L g
= 0 = 2 sin cos sin . (4)

dt R
Using the equation of motion one can indeed verify that dE/dt is not zero, as expected.
(1.b)
The equilibrium points are defined as the points where the bead, if placed there (i.e. with zero
velocity), does not move. They are given therefore by the condition

2 g
= 0 sin cos =0 , (5)
R
and are g g
0 = 0, , and 0 = arccos if <1 . (6)
R 2 R 2
Let us notice that we could get to the same result by interpreting the r.h.s of Eq. (4) as a force,
F () and defining an effective potential, Veff () such that
Veff ()
F () =

1
Veff () = mR2 sin2 2 mgR cos .
2
Then, the equilibrium condition is equivalent to the condition that gives the extrema of Veff (),
i.e.
Veff ()
= 0 F () = 0 =0 . (7)

The nature of the equilibrium points (stable vs unstable) can then be determined by looking at
the second derivative of Veff (),
2 Veff ()
2
= mR2 2 cos2 + mR2 2 sin2 + mgR cos , (8)

at each equilibrium point. We find that 0 = is always an unstable equilibrium point, since:
2 Veff ()

= mR2 2 mgR < 0 , (9)
2 0 =
while the nature of 0 = 0 and 0 = arccos Rg 2 depends on the ratio g/(R 2). Indeed,

2 Veff ()
g
2 2
= mR 1 (10)
2 0 =0 R 2
and 0 = 0 is a stable equilibrium point if g/(R 2) > 1 while it is unstable if g/(R 2) < 1.
Viceversa,
2 Veff ()
g 2
2 2
= mR 1 , (11)
2 0 =arccos( g ) R 2
R 2
and 0 = arccos(g/(R 2)) is a stable equilibrium point if g/(R 2) < 1, while it is unstable if
g/(R 2) > 1. The shape of Veff () therefore changes in going from g/(R 2) > 1 to g/(R 2) < 1.
In the first case Veff () has a minimum at 0 = 0 and no other extrema except a maximum at
0 = , while in the second case, 0 = 0 becomes a maximum (as well as 0 = ) and a new
minimum develops at 0 = arccos(g/(R 2)). Of course the potential is symmetric with respect
to , so an analogous discussion holds for between and 2. We can actually say that for
g/(R 2) < 1 there are two stable equilibrium points, one to the right and one to the left of the
= 0 position, both at an angle 0 = arccos(g/(R 2)) from the vertical.
(1.c)
To find the frequency of small oscillations about the stable equilibrium positions (0 ) we consider
a small displacement from 0 (lets call it ), i.e. we write
= 0 + , (12)
and plug it into the equation of motion. We then expand the equation of motion in and its
derivatives, keeping only up to linear terms, i.e. we write
sin(0 + ) = sin 0 cos + cos 0 sin = sin 0 (1 + O( 2 )) + cos 0 ( + O( 3 ) , (13)

cos(0 + ) = cos 0 cos sin 0 sin = cos 0 (1 + O( 2)) sin 0 ( + O( 3 ) ,
and we get the equation of motion for small oscillations about 0 in the form
g h g i
= 2 sin 0 cos 0 sin 0 + 2 cos2 0 sin2 0 cos 0 . (14)
R R
The terms in the first parenthesis cancel because 0 is a solution of = 0 (or equivalently a
minimum of Veff and therefore satisfies the extreme condition in Eq. (7)), and the equation for
reduces to
+ 2 = 0 , (15)
which is the equation of a one-dimensional harmonic oscillator with frequency
g
2 = cos 0 2 cos2 0 sin2 0 .

(16)
R
We then have two cases:
g/(R 2) > 1, the stable equilibrium point is 0 = 0 and the frequency of small oscillation
about 0 is
g 1/2
= 2 , (17)
R
g/(R 2) < 1, the stable equilibrium point is 0 = arccos(g/(R 2)) and the frequency of
small oscillation about 0 is
1/2
g2

= sin 0 = 1 2 4 . (18)
R
At the critical value of 2 = g/R, the system is transitioning between the two cases, and the
critical values all occur at = 0.
Problem 2
The system has two degrees of freedom and given the symmetry of the prob-
lem, it is natural to choose polar coordinates, r and . Indeed, we will denote
by r the elongation of the spring from its unscratched length. They are the z
generalized coordinates of the problem, in terms of which we can write the
kinetic and potential energy of m as x
1
T = m(r 2 + (l + r)2 2 ) , (19) l
2
1
V = mg(l + r) cos + kr 2 ,
2 r
m
and the Lagrangian of the system as
1 1
L = T V = m(r 2 + (l + r)2 2 ) + mg(l + r) cos kr 2 . (20)
2 2
The equation of motion for r is then
r (r + l)2 ) = mg cos kr ,
m( (21)
while the equation of motion for is
m[(l + r) + 2r ]
= mg sin . (22)
Notice that Eq. (21) and (22) are of the form mar = Fr and ma = F , where (ar , Fr ) and (a , F )
are the components of the acceleration and of the total applied force in the radial and tangent
direction respectively.
Problem 3 (1.14 of Goldsteins book)

Notice that the problem is in 3d. So, imagine the center of mass of the system to move in a circle
in the (x, y) plane, while the rod+masses can also move in z.
The kinetic energy (K.E.) of the system is the K.E. of the CM (with mass 2m) plus the K.E.
of the two masses rotating with respect to the CM.
T = TCM + T1+2 wrt CM
1
TCM = (2m)a2 2
2
2
1 l
T1+2 wrt CM = 2 m (2 + sin2 2 ).
2 2
Here the coordinates of 1 and 2 with respect to the CM are
l l
x1 = cos sin x2 = cos sin (23)
2 2
l l
y1 = sin sin y2 = sin sin (24)
2 2
l l
z1 = cos z2 = cos (25)
2 2

Here we use generalized coordinates {r, } and set the length of the string to be fixed, l.
1 1
T = m1 (r 2 + r 2 2 ) + m2 r 2
2 2
V = m2 g(l r)
1 1
L = T V = (m1 + m2 )r 2 + m1 r 2 2 + m2 g(l + r).
2 2
The equation of motion for r from the Euler-Lagrange equations is
r m1 r 2 m2 g = 0.
(m1 + m2 )
z
y CM

a
y
x
Figure 1: Figures for Problem 3
For the equation we note that since the Lagrangian

is not a function of we get
m1
d L L d L r
=0 = 0.
dt dt
This implies that the angular momentum of m1 about
lr
the origin is conserved,
L p m2
p = = m1 r 2 = constant = .

m1 r 2
Thus we can rewrite the r equation of motion as
p2
(m1 + m2 )
r + m2 g = 0,
m1 r 3
which is now just a 1d problem.
2 Non-graded Problems
U(r, v) = V (r) + L
(a)
U(r, v) = V (r) + (r mv)

= V (r) + m [x (yvz zvy ) + y (zvx xvz ) + z (xvy yvx )]
V d V
Fx = +
x dt x
V x
= m(y vz + z vy ) + m(y vz z vy )
r r
V x
= + 2m(y vz z vy )
r r
V y
Fy = + 2m(z vx x vz )
r r
V z
Fz = + 2m(x vy y vz )
r r
V
F = r + 2m( v)
r
Now putting these equations into spherical coordinates and choosing =
z, we can plug U(r, v)
into the general Lagrange equations

dU d U
Qi = i + .
dx dt x i
We find the following:
U = V (r) + mr 2 sin2
dV
Qr = 2mr sin2
dr
Q = 2mr 2 sin cos
Q = 2mr 2 sin cos + 2mr r sin2 .
(b)
Putting in the
z direction it is easy (but rather lengthy) to show the result
X ri
Qj = Fi .
i
qj
(c)
1
T = m(r 2 + r 2 2 + r 2 sin2 2 ),
2
where the equations of motion are
d T T
= Qj .
dt qj qj
m k
L= (ax 2 + 2bx y + cy 2) (ax2 + 2bxy + cy 2 ).
2 2
Equations of motion:
m(a
x + b
y ) + k(ax + by) = 0
m(c
y + b
x) + k(cy + bx) = 0
Case 1, a = c = 0.
y + 2 y = 0
x + 2 x = 0
p
where = k/m. So we have two decoupled 2d harmonic oscillators with the same fre-
quency.
Case 2, b = 0, c = a.
x + 2x = 0
y + 2 y = 0.
and the result is the same as in the first case.

If we make a change of variables

u a b x u = ax + by
=
v b c y v = bx + cy
the system decouples at the level of the Lagrangian. The b2 ac 6= 0 condition is the condition
for the transformation matrix to not be singular.

x1 = l1 sin 1 x 1 = l1 1 cos 1
n n
y1 = l1 cos 1 y 1 = l1 1 sin 1
x2 = l1 sin 1 + l2 sin 2 x 2 = l1 1 cos 1 + l2 2 cos 2
n n
y2 = l1 cos 1 l2 cos 2 y 2 = l1 1 sin 1 + l2 2 sin 2
The Lagrangian is
1 1
L = T1 + T2 V1 V2 = m1 (x 21 + y 12) + m2 (x 22 + y 22) V1 V2
2 2
1 1
= m1 l12 12 + m2 [l12 12 + l22 22 + 2l1 l2 1 2 cos(1 2 )] + m1 gl1 cos 1 + m2 g(l1 cos 1 + l2 cos 2 ).
2 2
The equations of motion (via Euler-Lagrange) are
(m1 + m2 )l1 1 = (m1 + m2 )g sin 1 m2 l2 cos(1 2 )2 22 m2 l2 sin(1 2 ).
m2 l2 2 = m2 g sin 2 m2 l1 1 cos(1 2 ) + 12 m2 l1 sin(1 2 ).

y
m1
m2
September 9th , 2015

Assignment # 3
1 Graded problems
1. A particle slides on the inside surface of a frictionless cone subject to the action of the
gravitational force. The cone is fixed with its tip on the ground and its axis vertical. Let
the half-angle at the tip be , let r be the distance from the particle to the axis of the cone,
and let be the angle around the cone.
(1.a) Find the equations of motion.

(1.b) If the particle moves in a circle of radius r0 , what is the frequency, , of the motion?
If the particle is perturbed slightly from this circular motion, what is the frequency, ,
of the oscillations about the radius r0 ? Under what conditions does = ?
2. A block of mass m is held motionless on a frictionless plane of mass M and angle of inclination
. The inclined plane rests on a frictionless horizontal surface. The block is released and
start sliding under the action of the force of gravity. What is the horizontal acceleration
of the inclined plane? (Try solving this problem using Newtonian mechanics, i.e. F = ma.
You will have a greater appreciation for the Lagrangian method!)
3. A particle is subject to a potential V (x) = F x where F is a constant. The particle travels

from x = 0 to x = a in a time interval t0 . Assume the motion of the particle can be expressed
in the form x(t) = A + Bt + Ct2 . Find the values of A, B, and C such that the action is a
minimum.
4. Suppose a particle moves in space subject to a conservative potential V (r) but is constrained
to always move on a surface whose equation is (r, t) = 0 (where the explicit dependence on t
indicates that the surface may be moving). The instantaneous force of constraint is taken as
always perpendicular to the surface. Show analytically that the energy of the particle is not
conserved if the surface moves in time. What physically is the reason for non-conservation
of the energy under this circumstance?

5. Goldsteins book, Chapter 2, Problem 19.
6. Goldsteins book, Chapter 2, Problem 6.

1 Graded Problems
Problem 1
(1.a)
We use cylindrical coordinates and notice that z = r cot . We use {r, } as generalized coordinates
and write the kinetic energy of the bead as
1 2 1 1
2 2 2 2 2 2
T = m r + r + z = m r + r , (1)
2 2 sin2
and its potential energy as
V = mgz = mgr cot , (2)
where we have assumed V = 0 at z = 0.
The Euler-Lagrange equation of motion for the r coordinate is
d L
= 0 r r sin2 2 + g sin cos = 0 , (3)
dt r
while the Euler-Lagrange equation of motion for is
d L d = 0 l = mr 2 = constant ,
= 0 (mr 2 ) (4)
dt dt
which expresses the conservation of angular momentum about the z axis.
~r

(1.b)
If the bead is in equilibrium at r = r0 then
r=r = = l ,
| (5)
0
mr02
since l is constant (see Eq. (4)). The equilibrium condition is that
l2 sin2
r = 0 = g sin cos , (6)
m2 r03
and the frequency of small oscillations about the (stable) equilibrium position is found by inserting
r = r0 + (with an infinitesimal displacement) in the r equation of motion and expanding linearly
in and its derivatives, which gives
l2 sin2 l2 sin2

2 + g sin cos = 13 + g sin cos = 0 , (7)
m (r0 + )3 m2 r03 r0
which, using Eq. (6) gives

3l2 sin2
+ =0 , (8)
m2 r04
and shows that the bead is performing small oscillations of frequency about r = r0 , where
2 3l2 sin2 3g
= 4
= sin cos . (9)
m r0
2 r0
Problem 2
Since there is no rotational motion, the motion
of the wedge and of the block is completely de-
scribed by the motion of their center(s) of mass. yb
y yb
The problem is planar, so we have two coordi-
nates for each center of mass. Lets call them
(xw , yw ) for the center of mass of the wedge and xb
(xb , yb ) for the center of mass of the block. We
also have two constraints: xb m
yw = constant and (10)

yb M
= tan , xw
xb

so we will need only two generalized coordi- x
nates. We pick xw for the wedge and xb for the

block, where xb is neither xb nor yb but is the

coordinate along the incline (i.e. it is the x co-
ordinate of a rotate system of Cartesian coordinates that has the x axis along the incline and the
y axis orthogonal to it; we notice that in this coordinate system the second constraint becomes
simply yb =constant as well). The relation between xb and yb and the chosen set of generalized
coordinates is
xb = xw + xb cos + constant , (11)

yb = xb sin + constant
where the constant terms are irrelevant either in the definition of the Lagrangian or in the form
of the equations of motion, so we do not specify them any further.
The kinetic energy of the system can then be written as
1 1 2 1 1 2
T = M x 2w + x b + y b2 = M x 2w +

x w + x 2
b + 2 x
w x

b cos , (12)
2 2 2 2
while the potential energy is
V = mgyb = mgxb sin + constant , (13)
where the constant term can be dropped because it does not affect the equations of motion. The
Lagrangian is therefore
1 1 2
L = T V = M x 2w +

x w + x 2
b + 2 x
w x

b cos + mgxb sin . (14)
2 2
The equation of motion for xw is
d
[M x w + m(x w + x b cos )] = 0 (15)
dt
which expresses the conservation of the x component of the linear momentum of the system. It
provides a relation between xw and xb of the form
m cos
xw = x . (16)
M +m b
Furthermore, the equation of motion for xb is
d
(x w + x b cos ) mg sin = 0 , (17)
dt
which gives
xb = g sin cos
xw . (18)
Solving the system of Eqs. (16) and (18) we get
mg sin cos
xw = , (19)
M + m sin2
g sin
xb = 2 .
1 mMcos
+m
Problem 3 (Goldstein 2.5)
The action is t0 t0
1 2
Z Z
S= Ldt = mx + F x dt.
0 0 2
From the problem statement we know x(t)
= B + 2Ct, and we can set the initial conditions
x(0) = A = 0 A = 0
1
x(t0 ) = Bt0 + Ct20 = a B = (a Ct20 ).
t0
Now we can explicitly calculate the action:
Z t0
1 2 2
S = m(B + 2Ct) + F (Bt + CT ) dt
0 2
Z t0
1 2 2 2 2
= mB + 2mBCt + 2mC t + F Bt + F Ct dt
0 2
1 2 1 1
= mB 2 t0 + mBCt20 + mC 2 t30 + F Bt20 + F Ct30
2 3 2 3
1 1 1
= mB t0 + (mBC + F B)t0 + (2mC + F C)t30
2 2 2
2 2 3
1 1 1 1 1
= m 2 (a Ct0 ) t0 + (mC + F ) (a Ct20 )t20 + (2mC 2 + F C)t30
2 2
2 t0 2 t0 3
1 1 2 1 1
= m (a + C 2 t40 2aCt20 ) + t0 (mC + F )(a Ct20 ) + (2mC 2 + F C)t30
2 t0 2 3
2
1 a 1 1 1 2 1
= m + mC 2 t30 maCt0 + maCt0 mC 2 t30 + aF t0 CF t30 + mC 2 t30 + F Ct30
2 t0 2 2 2 3 3
1 21 1 1
= ma + aF t0 + (mC 2 F C)t30 .
2 t0 2 6
Notice that since we have already set our initial conditions, the only unknown in our equation of
motion is C. Thus, this is what we want to find the minimum with respect to, S/C = 0.
S 1
= (2mC F )t30 = 0.
C 6
Thus we have
F
C =
2m
1 F 2
B = a t .
t0 2m 0
Note that we can check this is a minimum by looking at the second derivative;
2S 1 3
= mt > 0.
C 2 3 0
In general, the conservation of the total mechanical energy of the system (defined as E = T + U)
needs to be established by looking explicitly at the total differential of the energy with respect
to time, i.e. dE dt
. Under special circumstances, i.e. when the relation between cartesian and
generalized coordinates does not depend on time and the active forces are conservative, it is true
that dE
dt
= L
t
and the conservation of energy can be established by simply looking at the explicit
time dependence/independence of the Lagrangian.
In the case of the problem, since the equation of the constraint is time-dependent ((r, t) = 0),
the relation between cartesian and generalized coordinates is in general time dependent and the
conservation of energy cannot be deduced from the Lagrangian. In general we can then write that,
dE dT dU
= + (20)
dt dt dt
T T T U U
= qi + qi + + qi + ,
qi qi t qi t
where repeated indeces indicate summation. Using that for a system subject to both potential
U
(Qi = q i ) forces the Euler-Lagrange equations read,
) and non-potential (Q
i
d T T U i ,
= +Q (21)
dt qi qi qi
we can recast Eq. (20) in the following form,

dE d T U T T U U
= + Qi qi + qi + + qi + (22)
dt dt qi dq qi t qi t
i
d T T U U i qi
= qi + +2 qi + Q
dt qi dt qi dt
dT d(T1 + 2T0 ) T dU U i qi ,
= 2 + +2 Q (23)
dt dt dt dt t
or equivalently,
dE d(T1 + 2T0 ) T U i qi ,
= + +Q (24)
dt dt t t
where we have written the kinetic energy T as a polinomial in the qi ,
T = Aij qi qj + Bi qi + C T2 + T1 + T0 , (25)
with,
r r r r r r
Aij = , Bi = , and C = , (26)
qi qj qi t t t
and we have used that,
dT2 dT1 dT0
= 2T2 , = T1 , and =0 . (27)
dt dt dT
If the constraints are time dependent (or reonomic), the relations between cartesian and gen-
eralized coordinates will in general be also time dependent ( r = r(qi , t)) and one or more terms on
the right hand side of Eq. (24) will be non zero causing the total mechanical energy of the system
not to be conserved. Physically, we can interpret this by observing that, in the case or reonomic
or time-dependent constraints, the work of the forces of constraints is non zero since they do not
remain orthogonal to the trajectory during the time evolution of the system (given that the con-
straints change with time). The work of the forces of constraints per unit time corresponds indeed
to the last term in Eq. (24), and as we see it is just one of the reasons why the total mechanical
energy of the system is not conserved. Occasionally the explicit time-dependence of either kinetic
or potential energies, or both, can contribute to it as well.
If the mass distribution has a given symmetry, so will the potential and therefore so will the
Lagrangian. From the symmetry, we deduce the conserved quantity.
(a) The force does not depend on (x, y) (px , py ) conserved. It also does not depend on
the angle of rotation about z, so lz is conserved as well.
(b) The force does not depend on x, so px is conserved.
(c) The force does not depend on z so pz is conserved. It is also independent of the angle of
rotating about z, so lz is also conserved.
(d) The force does not depend on the angle of rotation about z (although it is now a function
of z since it is finite), so lz is conserved while pz is not.
(e) The force does not depend on z so pz is conserved.
(f) The force does not depend on the angle of rotation about z, so lz is conserved.
(g) For h the distance between coils, the combination hpz + lz is conserved.

First assume the Earth has a uniform mass density , so that the mass distribution is
Z r
2 4
M(r) = 4 r 2 dr = r 3 ,
0 3
M(R) ME r3
= 4 3 = 4 3 M(r) = ME 3 ,
3
R 3
R R
Where R is the radius of the Earth and ME is the mass. Using this we can find the gravitational
force and corresponding potential on a mass m,
GM(r)m GME m
F(r) = r = rr,
Z rr R3
2
GME m 2
V (r) = (F(r ) r )dr = r .
0 2R3
So at a distance r from the center of the Earth we can find the energy to be
1 GME m 2
E = T + V = mv 2 + r .
2 2R3
Since the gravitational force is conservative we can use energy conservation to find the velocity.
If the velocity at the surface is zero we have
r
GME m 2 1 2 GME m 2 GME 2
E(R) = E(r) R = mv + r v(r) = (R r 2 ).
2R3 2 2R3 R3
We want to find the curve that minimizes the time between points a and b,
Z b
ds
t= .
a v(r)
If we consider coordinates {x, y} which define the plane passing through the center of the Earth
and the two points at the ends of the curve, we can expressed the arc length and the time traveled
as
s
2 2
p dx dy
ds = dx + dy =
2 2 + d
d d
p
= x2 + y 2 d,
v h
u
dx 2
i
dy 2
Z u 3
t R d
+ d
t = d
0 GM E R 2 r2
Z s
x2 + y 2
= k2 2 d .
0 R (x + y )
2 2
p
We have defined k = R3 /GME . To minimize the function
s
x2 + y 2
f (x, y, x , y ) = k (28)
R2 (x2 + y 2)
we consider the Euler-Lagrange equations:

d f f
= 0
d x x
d f f
= 0.
d y y
Since f is not explicitly a function of , the Euler-Lagrange equations are equivalent to
f
f x = const. = C1
x
f
f y = const. = C2
y
Using our function from eq (28) we can write this as:
s
y 2 x2 + y 2 C1
=
x2 + y 2 R2 (x2 + y 2) k
s
x2 x2 + y 2 C2
=
x2 + y 2 R2 (x2 + y 2) k

s
x2 + y 2 C1 C2
= +
R (x + y )
2 2 2 k k
x2 + y 2 C2
=
R2 (x2 + y 2) k2
C2 2
R (x2 + y 2) ,

x2 + y 2 = (29)
k 2
where C = C1 + C2 . The solution to this equation is a hypocycloid with parametric equations

Rb
x() = (R b) cos + b cos
b

Rb
y() = (R b) sin b sin .
b
These give the coordinates of a point on a circle of radius b rolling with no slipping inside a circle
of radius R. The constant b is determined by finding

Rb
x () = (R b) sin (R b) sin
b

Rb
y () = (R b) cos (R b) cos .
b
and substituting into eq. (29). We find

r
C2 Rb Rb
= C=k .
k 2 b b
The time is takes to travel between two points on the Earths surface is
s
B B
ds x2 + y 2
Z Z
B
tAB = = k d
A v(r) A R2 (x2 + y 2)
C B
Z R
s=
= k d = CdAB
2 b
k A
r AB
2b R b 2b
= C =k
sR b R
R 2s s A
= k s
2
R
r
2R s s
= k
r s R
sR
= (2R s) 1640s 27.4min.
GME
We have used the fact that the distance will simply be an arc length of a circle connecting the
two points A and B (see figure). We have also used the constants s = 4800 km, R 6371 km,
ME 5976 1024 kg, G = 6.6726 1011 Nm2 /kg2 . At the deepest point the tunnel would be
2b = s/ 1528 km!
Using another property of the hypocycloid we can determine the length of the tunnel:

8(R b)b 2 R
l() = sin
R 4b
s
s
R 2

2b 2 2 R 2b
l = 8 sin
R R 4b R
s s
= 8 R 5378 km.
2 2R

Assignment # 4
(Graded problems are due Wednesday September 23rd , 2015)
1 Graded problems
1. A spherical pendulum consists of a bob of mass m attached to a weightless, extensionless
rod of length l. The end of the rod opposed to the bob pivots freely (in all directions) about
some fixed point.
(1.a) Set up the Lagrangian function in spherical coordinates and derive the equations of
motion. How can you interpret the equation for , the angle of rotation about the
vertical axis?
(1.b) Discuss the limiting case = 0 , i.e. =constant.
(1.c) Discuss the case = 0 , i.e. the case of a conical pendulum ( is the angle with respect
to the vertical). Which condition needs to be verified for the pendulum to move at
= 0 ? Find the frequency of small oscillations about = 0 .
2. A particle of mass m starts at rest on top of a smooth hemisphere of radius a. Find the
force of constraint, and determine the angle at which the particle leaves the hemisphere.
3. A uniform hoop of mass m and radius r rolls without slipping on a fixed cylinder of radius
R. The only external force is that of gravity. If the smaller cylinder (hoop) starts rolling
from rest on top of the bigger cylinder, use the method of Lagrange multipliers to find the
point at which the hoop falls off the cylinder.
4. Problem 2.21 of Goldsteins book.


1 Graded Problems
Problem 1
(1.a)
The problem has spherical symmetry and is therefore naturally solved using spherical coordinates
(see figure). The fixed length of the pendulum gives the constraint: r = l and reduces the number
of generalized coordinates to two: {, }.
The kinetic and potential energies of the pendulum are
1 2 2
m l + l2 sin2 2 ,

T = (1)
2
V = mgl(1 cos ) , (2)
and the Lagrangian is:

1
L = T V = m l2 2 + l2 sin2 2 mgl(1 cos ) .

(3)
2
Since the Lagrangian does not depend on , the Euler-Lagrange equation of motion for states
the conservation of the angular momentum with respect to the vertical (or z) axis:
d 2 2 p
ml sin = 0 ml2 sin2 = constant = p = . (4)
dt ml2 sin2
On the other hand, the Euler-Lagrange equation of motion for reads
l2 l2 2 sin cos + gl sin = 0 . (5)
Substituting in Eq. (5) in terms of p , as derived in Eq. (4), the equation for reduces to a one
dimensional equation in the only variable (and its derivatives) of the form
p2 g
3 cos + sin = 0 . (6)
m2 l4 sin l
(1.b)
The case = 0 , i.e. = 0 is trivial and reduced to the plane simple pendulum. Indeed, Eq. (5)
reduces to
g
+ sin = 0 . (7)
l
(1.c)
The case = 0 corresponds to the case of a conical pendulum, since the pendulum describes a
conical surface during its motion. = 0 is an equilibrium point, and therefore needs to verify
the condition = = 0. Imposing this condition in Eq. (5) we get that, in order to realize the
condition of conical pendulum we need to have = 0 such that
g g 1 g
20 sin 0 cos 0 = sin 0 20 = = sec 0 . (8)
l l cos 0 l
= 0 is indeed a stable equilibrium position, as we can prove by studying the motion of the
pendulum for small displacements (in ) from the equilibrium angle = 0 . In order to do
that, lets keep = 0 and lets consider a small displacement from the equilibrium position, i.e.
= 0 + , for infinitesimal. Plugging this into Eq. (5), we get
g g sin4 0 cos(0 + )
+ sin(0 + ) =0 . (9)
l l cos 0 sin3 (0 + )
Expanding in and keeping only up to the linear terms in , we find

sin o
g sin 0 cos 0 1 cos 0
4
+ sin 0 + cos 0 =0 , (10)
l cos 0 sin3 0 1 + cos 0 3

sin 0
" !#
g sin 0 cos 0
+ sin 0 + cos 0 sin 0 1 3 =0 ,
l cos 0 sin 0
g
+ (3 cos 0 + sec 0 ) = 0 ,
l
which is the equation of an harmonic oscillator with frequency
g
2 = (3 cos 0 + sec 0 ) . (11)
l
So, a conical pendulum rotates about the vertical axis with constant velocity 0 (given in Eq. (8))
while it performs small oscillations about the = 0 position with frequency (given in Eq. (11)).
The period of small oscillations in is
s
2 l
T = = 2 . (12)
g(3 cos 0 + sec 0 )
During one complete oscillation in the pendulum sweeps an angle 1 > , since
s
g l 2
r
1 = 0 T = sec 0 2 = > . (13)
l g(3 cos 0 + sec 0 ) 1 + 3 cos2 0
If we imagine to project the motion of the pendulum on the (x, y) plane, it will look like a curve
that could close after m turns if the condition 2n = mT is verified. On the other hand, if we
imagine to trace the motion of the pendulum on a sphere, we would see it oscillating between two
circles (one lower and one higher of the = 0 circle) describing a sort of sinusoidal curve on the
surface of the sphere.
Problem 2
The problem is planar (given the symmetry in the azimuthal angle), and can be treated using
polar coordinates {r, }. Normally, we would impose the constraint r = a and work with just one
generalized coordinate, . However, the problem asks for the angle at which the particle leaves
the hemisphere, which corresponds to the angle at which the force of the constraint vanishes. We
then need to determine the force of the constraint, keeping the full set of coordinates, {r, }, and
using the method of Lagrange undetermined multipliers.
We have one constraint, r a = 0, and therefore we introduce one undetermined multiplier,
. Comparing the equation of the constraint, r a = 0, to the generic constraint equation in
differential form
ar dr + a d = 0 we deduce: ar = 1 , a = 0 .
The kinetic and potential energies of the particle are
1
T = m(r 2 + r 2 2 ) , (14)
2
V = mgr cos ,
and the Lagrangian is
1
L = T V = m(r 2 + r 2 2 ) mgr cos . (15)
2
The equation of motions will then be, for r,
d L L
r mr 2 + mg cos = ,
= ar m (16)
dt r r
and for ,
d L L
= a mr 2 + 2mr r mgr sin = 0 . (17)
dt
We will solve them for r(t), (t) and , together with the constraint equation r = a (i.e. r = r = 0).
After using the equation of the constraint, Eqs. (16) and (17) become:
ma2 mg cos + = 0 (18)
mga sin ma2 = 0 . (19)
Noticing that
d d d d
= = = , (20)
dt d dt d
we can derive, from Eq. (19), that
g d
= sin = , (21)
a d
which, upon integration, yields
1 2 g g
= cos + , (22)
2 a a
where we have assumed that = 0 at t = 0, when = 0. Substituting 2 into Eq. (18) we get
= mg(3 cos 2) , (23)
and therefore the particle leaves the hemisphere when
2

= 0 = arccos . (24)
3
There are two constraints in this problem: (1) the distance from
the center of the cylinder to the center of the hoop is R+a, and (2)
a there is no slipping, i.e. the velocity of the contact point between
the cylinder and the hoop is zero. Describe this motion using the
coordinates of the center of mass of the hoop and the rotation angle
of the hoop about the center of mass. The constraints are thus
R
r =R+a (25)
(R + a) a = 0 (26)
We will use the method of Lagrange multipliers to find the normal force on the hoop. The value
of for which the force is zero will correspond to the point at which the hoop falls of the cylinder.
The generalized coordinates will be {r, , }.
The equations of constraints are
r = R + a dr = 0 a1,r = 1, a1, = a1, = 0

= 0 (R+a)dad = 0 a2,r = 0, a2, = R+a, a2, = a.
(R+a)a
The Lagrangian is
~r
1 1
L = m(r 2 + r 2 2 ) + (ma2 ) 2 mgr cos ,
2 2
and the equations of motion (via Euler-Lagrange) are
d L L
= 1 a1,r m r mr 2 + mg cos = 1 ,
dt r r
d L L
= 2 a2, 2mr r + mr 2 mgr sin = (R + a)2 ,
dt
d L L
= 2 a2, ma = a2 ,
dt
where the Lagrange multipliers are 1 , 2 . These need to be solved together with the equations of
constraint (25) (which implies r = r = 0) and (26). We find
m(R + a) 2 + mg cos = 1 , (27)
m(R + a)2 mg(R + a) sin = (R + a)2 , (28)

ma2 = a2 . (29)
Note that
1 corresponds to the r-coordinate and is the normal force on the hoop. Thus, the condition
we are looking for is 1 = 0.
2 corresponds to the rotation of the hoop on the cylinder (in either or ), and is thus the
tangent friction that make the no-skipping condition possible.
Solving the system of equations:
R+a
(29): ma2 = a2 2 m(R + a),
a
(28): m(R + a)2 mg(R + a) sin = m(R + a)2 2(R + a) g sin = 0

g sin
= .
2(R + a)
Multiply this last equation by and rewrite it as
d 1 2 g g d

= sin = (cos )
dt 2 2(R + a) 2(R + a) dt
1 g
2 = cos + const.
2 2(R + a)
Setting = 0 and = 0 as the initial conditions we find that constant term is
g
const. =
2(R + a)
which gives us the equation of motion

g
2 = (cos 1).
R+a
Now plug this into (27):
mg(cos 1) + mg cos = 1 .
This gives us the Lagrange multiplier (and force of constraint)
1 = mg(2 cos 1),
and setting this to zero we can find the angle at which the constraint is violated:
1
1 = 0 cos = = = 60 .
2 3

(a)+(b)
The laboratory frame (x, y) and rotating frame (R, r) coordinates are defined as,
y n x = R cos r sin
(x, y) (30)
y = R sin + r cos
r
n R = x cos + y sin
(31)
r = x sin + y cos
R
(32)
with = t .
= t x
Using lab. frame coordinates (x, y), we can write the kinetic and potential energies as,
1
T = m(x 2 + y 2 ) , (33)
2
1 2 1 1 1
V = kr + K (R R0 )2 = k (x sin(t) + y cos(t))2 + K (x cos(t) + y sin(t) R0 )2 ,
2 2 2 2
such that the Lagrangian is,
1 1 1
L = T V = m(x 2 + y 2 ) k (x sin(t) + y cos(t))2 K (x cos(t) + y sin(t) R0 )2 , (34)
2 2 2
The energy function or Jacobi integral is,
L L
h= x + y L = 2T L = T + V = E , (35)
x y
as expected since the potential energy does not depend on the velocities and the kinetic energy is
a homogeneous function of second degree in the velocities. We can then use that,
dh L
= , (36)
dt t
to prove that the energy of the system is not conserved, since,
dE dh L
= = 6= 0 , (37)
dt dt t
given the explicit time dependence in the Lagrangian (via = t). This was expected since, in
order to keep the steady rotational motion of the system some work is done on the system and
this goes into the balance of the mechanical energy of the system, which, by itself (i.e. without
including the work done on the system) is not conserved. Stated differently, the mechanical system
of the two springs is not isolated.
For completeness, we will also derive here the equations of motion in the lab. frame coordinates:
d L L
= 0 m
x = k(y cos x sin ) sin K(x cos + y sin R0 ) cos (38)
dt x x
m
x = kr sin(t) K(R R0 ) cos(t)
d L L
= 0 m
y = k(y cos + x sin ) cos K(x cos + y sin R0 ) sin
dt y y
m
y = kr cos(t) K(R R0 ) sin(t)
(c)
In the (non-inertial) system of the rotating carriage, the two springs oscillate in the x and y
direction respectively, and we chose to call the corresponding coordinates R and r. The kinetic
energy is them simply given by,
1 1
T = mr 2 + mR 2 . (39)
2 2
The force acting on m in the rotating frame (F ~ rot ) is,
~ rot = F
F ~ lab m~ (~ ~r) 2m~ ~v , (40)
where F~ lab is the force acting on m in the laboratory frame, ~ is the vector angular velocity
r)
(orthogonal to the plane of motion), while ~r = (R, r) and ~v = (R, are the position and velocity
vectors of m in the rotating frame. The last two terms in Eq. (40) are often referred to as
centrifugal and Coriolis forces. As explained in Sec. 1.5 of Goldsteins book, for forces Qj that
also contain terms dependent on the velocities, a generalized potential function U(qj , qj ) can be
defined such that it satisfies,
d U L
= Qj . (41)
dt qj qj
In our case, it is easy to show that the generalized potential function is of the form,
1 1 1 .
U = kr 2 + K(R R0 )2 m 2 (r 2 + R2 ) m(Rr r R) (42)
2 2 2
In the rotating system the Lagrangian is then,
1 1 1 1 1
L = mr 2 + mR 2 kr 2 K(R R0 )2 + m 2 (r 2 + R2 ) + m(Rr r R)
. (43)
2 2 2 2 2
From this Lagrangian one can derive the equation of motions corresponding to the (R, r) system
of coordinates, i.e.
d L L = K(R R0 ) + m 2 R + 2m r ,
= 0 mR (44)

dt R R
d L L
= 0 m r = kr + m 2 R 2m R ,
dt r r
and verify that they indeed corresponds to the equations of motions generated by the non-inertial
force in Eq. (40). Moreover one can verify that they correspond to the equation of motion in the
laboratory frame, when the equations in Eq. (38) are transformed to the rotating frame (using
the formalism explained in the Note at the end of this problem).
Finally, in the rotating frame the Jacobi integral is,
L L
h = R+ r L (45)
R r
= m(r + R)r + m(R r)R L
= E mR(r + R) + mr(R r) .
Since L is independent of time,
dh L
= = 0, (46)
dt t
but since T is not a homogeneous function of the velocities (of second degree) then h 6= E and we
cannot infer anything from the time (in)dependence of h. Indeed, as seen in (a)+(b), the energy
is not conserved and this holds always, independently of the choice of frame or coordinates. It is
possible to check this explicitly (this is not needed but it can be an extra check of the consistency
of the formalism derived in this part of the problem),
dE dh d h
mR(r + R) mr(R r)
i
= +
dt dt dt
d h
+ m 2 (r 2 + R2 )
i
= m(Rr Rr)
dt
= m(R + 2m 2 (RR + r r).
r r R) (47)
This expression can be simplified further by using the two equations of motion:
d h
m(R r) m(r + R) + K(R R0 ) = 0
i
dt
mR = 2mr
+ m 2 R K(R R0 )
d
[m(r + R)] + m(R r) + kr = 0
dt
m r = 2m R + m 2 r kr . (48)
Plugging these two equations into (47) we find that many things cancel and we are left with
dE
= krR Kr(R R0 ) 6= 0 . (49)
dt
Note
The laboratory-frame and rotating-frame coordinates are related by a time-dependent rotation
( = t), i.e.
! ! ! ! !
x cos sin R R t x
= =M =M (50)
y sin cos r r y
and consequently,
R R
! ! ! ! !
x R x
= M +M =A +M , (51)
y r r y r
where, !
t= 0
A = MM . (52)
0
In the same way we can show that,
R
! ! ! ! !
x x x R
= A +A + M +M (53)
y y y r r
R R
! ! ! ! !
x x R
= A +A 2
+ AM + M +M (54)
y y r r r
R
! ! ! !
x R R
= A +A M2
+ 2M +M (55)
y r r r
R
( ! ! !)
2 R R
= M A + 2A + , (56)
r r r
where we have used that A = 0, A = M t AM, AM = MA, and A2 M = AMA = MA2 . From the
previous relation we deduce that,

! ! ! ! !
2
x R R R R 2 r + R
Mt = A2 + 2A + = , (57)
y r r r 2 r + 2 R + r
and therefore we show that the equations of motion in the laboratory frame (see Eq. (38)) transform
into the equations of motion on the rotating frame (see Eq. (44)), since,

! ! !
t x K(R R0 ) 2 R 2 r + R
mM = =m . (58)
y kr 2 r + 2 R + r
Consider a rotation about an arbitrary axis that we identify with

z. This corresponds to the rotation angle . Let us first consider
the one-particle case; if V (x,
y,
z)
then
z
V V x V y V z
= + +
x y z
~r V V
= (r sin sin ) + (r sin sin )
x y
= (~r ~ v V (~r ))
z = ~n (~r
~ v V (~r, ~r).

In the above we have defined ~n := z, and we have used the coor-
dinates
y
( x = r sin cos
y = r sin sin
z = r cos
( x = r sin cos + r cos cos + r sin sin
x y = r sin sin + r cos sin + r sin cos
z = r cos r sin .
Therefore we find
L T V X
~ v V ).
p = = = l ~n (~ri

i
i
The new term V / only exists for V = V (~r), and we have also extended the result to include
many particles. For electromagnetic forces the potential is
q~
V = q A ~v
c
so the gradient is
~ v V = q A.
~
c
Plugging this in gives us the final result;
qi ~
X
p = l ~n ~ri A .
i c
On a spheroidal surface of radius R:
( x = R sin cos
y = R sin sin
z = R cos
The arc length is given by
q A
ds2 = R2 d2 + R2 sin2 d2 ds = R d2 + sin2 d2 . A
B B
Thus the distance between two points (A,B) is given by
Z B Z B q
lAB = ds = R d2 + sin2 d2 A
A
v A B
u !2
B d
Z u
= R d sin2 +
t
.
A d
Let us define the integrand as

q
f (, ) = sin2 + 2 .
The variational principle tells us that the minimal dis-

tance is obtained when , satisfy the equation
d f f
=0
d
!
d sin cos
2 =0
d sin + 2
2
sin + 2
sin cos + sin cos
2 = 0.
f f f f
( sin cos f2 2 (sin cos + ) = 0
( sin cos )(2 + sin2 ) 2 (sin cos + ) = 0
sin2 22 sin cos sin3 cos = 0
h i
sin sin 22 cos sin2 cos = 0
sin2 22 cos sin2 cos = 0. (59)
Now lets work a little backwards. Great circles (which is what we are trying to prove being the
shortest path between two points on a sphere) are the intersections between a sphere and a plane
going through these two points. If
n
= nx x
+ ny y
+ nz
z
is the vector perpendicular to the plane, the points in the great circle are those such that
~r = 0
n
nx R cos cos + ny R sin cos + nz R cos = 0
R [sin (nx cos + ny sin ) + nz cos ] = 0
Thus we find the condition satisfied for points on a circle is
cos
= A cos + B sin .
sin
Now define a function
cos ()
g() := .
sin ()
Find the second derivative of this quantity:
! !
d2 g d dg d d sin cos2
= =
d2 d d d d sin2
d 1

= 2
d sin
2 cos 2 1
= 2
sin sin2
1 2 cos 2
= 2 +
sin sin2
1 h i
= 3 sin + 2 cos 2 .
sin
Our minimization condition (59) tell us that the quantity in brackets must be equal to sin2 cos ,
so we have
d2 g 1 2 cos d2 g
= sin cos = = g +g =0
d2 sin3 sin d2
g() = A cos + B sin .
This is exactly the equation for great circles on a sphere.
September 23rd , 2015

Assignment # 5
1 Graded problems
1. Consider a particle that moves in a logarithmic spiral orbit given by r = ke , where k and
are constants.
(1.a) Find the force law that allows the particle to move in this orbit.
(1.b) Determine r(t) and (t).
(1.c) What is the total energy of the orbit?
2. A particle of mass m moves in a potential given by V (r) = rk , where and k are constants.
Let the angular momentum be l.
(2.a) Find the radius r0 of the circular orbit.
(2.b) If the particle is given a tiny kick so that the radius oscillates around r0 , find the
frequency, r , of these small oscillations in r.
(2.c) What is the ratio of the frequency r to the frequency of the (nearly) circular motion,
= ? Describe the cases: k = 1, 2, 7, 7 , for which the ratio r / is rational,
4
that is, for which the path of the nearly circular motion closes back on itself. Can you
roughly plot the orbits for these four cases?
3. Two particles move about each other in circular orbits under the influence of gravitational
forces, with a period . Their motion is suddenly stopped at a given instant of time, and
they are released and allowed to fall into each other. Prove that they collide after a time

4 2
.

4. A particle moves in a force field described by
k r

F (r) = 2 exp ,
r a
where k and a are positive.
(4.a) Write the equations of motion and reduce them to the equivalent one-dimensional
problem. Use the effective potential to discuss the qualitative nature of the orbits for
different values of the energy and the angular momentum.
(4.b) Show that if the orbit is nearly circular, the apsides will advance approximately by
/a per revolution, where is the radius of the circular orbit.
1 Graded Problems
Problem 1
(1.a)
Using the equation of the orbit or force law
d2 1 1 mr 2

+ = 2 F (r) , (1)
d2 r r l
with r() = ke one finds
2 1 mr 2
+ = 2 F (r) , (2)
r r l
from which
(1 + 2 )l2 1
F (r) = . (3)
m r3
(1.b)
For a central force motion we have that
l l 2
= 2
= e , (4)
mr mk 2
where l is the magnitude of the conserved angular momentum. We can easily integrate this
equation by separation of variables, i.e.
l 1 2 l
e2 d = dt e + C = t , (5)
mk 2 2 mk 2
where C a constant of integration. Isolating the exponential term and taking the logarithm of
both l.h.s and r.h.s. one gets
" #
1 2l
(t) 0 = ln 2
t + C , (6)
2 mk
where C = 2C is determined by the initial conditions on 0 .

Substituting (t) into the expression of r() one gets
" #1/2 " #1/2
2l 2l
r(t) = K 2
t + C = t + k2C , (7)
mk m
where K = ke0 .
(1.c)
The total energy of the orbit is
1
E = m(r 2 + r 2 2 ) + V (r) , (8)
2
where, using Eqs. (6)-(7), we can calculate the kinetic energy as
1 (1 + 2 )l2 1
T = m(r 2 + r 2 2 ) = , (9)
2 2m r2
while the potential energy (modulus a constant of integration) is
(1 + 2 )l2 1 (1 + 2 )l2 1
Z Z " #
V (r) = F (r)dr = = , (10)
m r3 2m r2
and E = T + V = 0.
Problem 2
(2.a)
The problem is easily discussed in terms of the effective potential
1 l2 1 l2
V (r) = + V (r) = + r k . (11)
2 mr 2 2 mr 2
In order for a circular orbit to exist the effective potential has to have a minimum for some finite
value of r. The minimum condition is
V (r) l2
= 0 3 + kr k1 = 0 , (12)
r mr
which admits a real solution only if and k are either both positive or both negative. In which
case the radius of the circular orbit is
1
l2
!
k+2
r0 = . (13)
mk
(2.b)
Since about the equilibrium position r = r0 the system behaves as a linear harmonic oscillator
subject to a restoring force F (r) = (r r0 ), with potential energy V (r) = V (r0 ) + 21 (r r0 )2 ,
we can find by simply expanding V (r) about r = r0 and taking the coefficient of the quadratic
term in the expansion. The frequency of small oscillations will then be r = (/m)1/2 (where the
index r indicates that the oscillation are in the radial direction). The expansion of the potential
is
1 2 V (r)

V (r) = V (r0 ) + 2
(r r0 )2 + O((r r0 )3 ) , (14)
2 r

r=r0
such that
2 V (r)

= (15)
r 2 r=r0

3l2 1
= + k(k 1)r0k2
m r04
! 4 !k+2
l2 k+2
3l2 l2
= + k(k 1)
mk m mk
l2
= r04 (k + 2) ,
m
and the frequency of small oscillations r is
l
1/2

r = = k+2 . (16)
m mr02
(2.c)
The ratio of the frequency of small (radial) oscillation, r , to the frequency = of the (nearly)
circular motion is
l
r mr 2 k+2
= 0 l = k+2 . (17)
mr 2
0
The four given cases are:

r
k = 1 =1 (18)

r
k=2 =2

r
k=7 =3

7 r 1
k= =
4 2
1
which correspond to r making 1,2,3, or respectively 2
oscillation(s) for each complete revolution
in .

The reduced system also moves in a circular orbit with some radius r = a (and therefore r = 0).
The corresponding equation of motion is
l2 k
r = 0 = 3
2.
ma a

We solve this, using l = mr 2 :
l2 k k
3
= 2 2 = .
ma a ma3
s s
k 2 2 ma3
= = = = = 2 . (19)
ma3 k
Here we can note that is constant and must be the period of both the reduced system and
the original circular motion.
When the two masses are stopped and then released from rest, they have zero angular mo-
mentum l = 0, so they just satisfy a radial motion equation of the form
k 1 k
= mr 2 ,

a 2 r
which is easily found using conservation of energy. Therefore
2k 1 1

2
r =
ms r a s
2k 1 1 1/2 2k a r 1/2

r = = .
m r a m ra
Integrating the previous relation between t = 0 and t, we get,
s
Z 0 dr 2k
q = t, (20)
a ar m
ra
where t is the time it takes for the two masses to move from r = a to r = 0. Performing this
integration:
x2
s
Z 0 dr Z0 r Z0
q = a dr = 2 a dx
a ar a ar a a x2
ar
!#0
x
"
a x
= 2 a a x2 + sin1 = a a.
2 2 a
a
2
In the first line we have changed integration variables with r = x2 , and to get to the second line
we have used a standard integration table. Thus, from (20) we have
s
2k m
r
a a= tt= a a
2 m 2k 2
m 3 2 2
t2 =
a = t= .
2k 4 32 4 2
To get this final result we have used the period we found in (19).
(Note that the Yukawa potential is a kind of screened Coulomb potential, and can be used to
describe some common particle interactions - pion exchange between nucleons, for instance.)
The force corresponding to the Yukawa potential (for k, a > 0) is
k r/a
F (r) = e .
r2
(4.a)
The Lagrangian corresponding to a particle in the Yukawa potential is
1 1 k
L = m(r 2 r 2 2 ) V (r) = m(r 2 + r 2 2 ) + er/a .
2 2 r
The equation of motion for simply gives us conservation of angular momentum:
mr 2 = constant := l.
The equation of motion for r is

k k
r mr 2 + 2 er/a + a er/a = 0
m
r r
l2
!
k ak r/a
m
r 3
+ 2+ e = 0.
mr r r
Using this we can write the energy as:
1
E = m(r 2 + r 2 2 ) + V (r)
2
1 2 1 l2 k 1
= mr + 2
er/a = mr 2 + V (r),
2 2 mr r 2
where V (r) is the effective potential (see figure). Asymptotically, this potential has the feature
that for both large (r ) and small (r 0) it is dominated by the 1/r 2 term. In the middle
regions it will depend on the value of l.
(4.b)
The circular orbit condition is verified (for those values of l when V (r) has a minimum) if:
l2
!
V (r) k k
=0 2 + 2 + er/a = 0
r mr r ra
l2 r0

= r0 er0 /a a + . (21)
mk a
In this case we explain what happens when we examine small deviations from r = r0 . Take
r() = r0 [1 + ()]
and insert this into the equation for the orbit
d2 1 1 mr 2 mk

+ = 2 F (r) = 2 er/a .
d2 r r l l
Using the standard change of variables
1 1
u := = (1 ),
r r0
50
l=0.03*K
l=K
40
30
V(r)
20
10
-10
0 0.2 0.4 0.6 0.8 1
r/a
Figure 1: A graph of the effective Yukawa

potential for two different vales of angular momentum.
Here we have set k/a = 1 and K = 2mk.
we find that
d2 u mk 1/au
+ u = e
d2 l2

2
du mk r0 (1+) r0

+ (1 ) = r 0 e a a
d2 l2 a

d2
!
mk 2 r0 /a mk
2
+ 1 2 r0 e = 1 2 r0 er0 /a .
d l a l
This is the equation for a simple harmonic oscillator (with a constant shift) and frequency
mk 2 r0 /a r0 1 1
2 = 1 r 0 e = 1 r = ,
l2 a 1 + a0 1 + ra0
where we have used the definition of r0 from (21). Now choose to be at maximum when = 0,
then the next maximum will occur when
2 !
2 r0 r0

= 2 = = 2 1 + +o .
2a a
Therefore the apsides advance by
r0
=
a
each revolution.

Assignment # 6
(Graded problems are due Friday October 7th , 2015)
1 Graded problems
1. Consider an attractive 1/r potential and show that:
1.a) for circular and parabolic orbits having the same angular momentum, the perihelion
distance of the parabola is one-half the radius of the circle;

1.b) the speed of a particle at any point in a parabolic orbit is 2 times the speed in a
circular orbit passing through the same point.
2. Discuss the motion of a particle in a central inverse-square law force field for a super-imposed
force whose magnitude is inversely proportional to the cube of the distance from the particle
to the center of force, that is
k
F (r) = k, > 0 .
r2 r3
Show that the motion is described by a precessing ellipse. Consider the cases < l2 /m,
= l2 /m, and > l2 /m.
3. Consider a particle describing a circular orbit under the influence of an attractive central
force directed toward a point in the circle.
3.a) Show that the force varies as the inverse-fifth power of the distance.
3.b) Show that for the orbit described the total energy is zero.
3.c) Find the period of the motion.
3.d) Find x,
y, and v as a function of the angle around the circle and show that all three
quantities are infinite as the particle goes through the center of force.
4. A particle is moving in a potential

C
V (r) = (C > 0) .
3r3
4.a) Given l (angular momentum), find the maximum value of the effective potential.
4.b) Let the particle come in from infinity with speed v0 and impact parameter b. In terms
of C, m, and v0 , what is the largest value of b (call it bmax ) for which the particle is
captured by the potential? In other words, what is the cross section for capture, b2max ,
for this potential?
5. Chapter 3, Problem 10 of Goldsteins book.
1 Graded Problems
1.a)
For a circular orbit, by conservation of energy we have
k l2 k
E= = 2

2r0 2mr0 r0
l2 k l2 (c)
2
= r 0 = = rmin .
2mr0 2r0 mk
For a parabolic orbit with the same angular momentum l we have
l2 k (p) l2
E=0= 2
r min = .
2mrmin rmin 2mk
Thus we see
(p) 1 (c)
rmin = rmin .
2
1.b)
For the circular orbit, with v = r we have
1 2 2 k k
mr = E = ,
2 r 2r
s
1 2 k (c) 1k
mv = v(r) = .
2 2r mr
For a parabolic orbit with the same r (using v = r 2 + r 2 2 ) we can show
1 k
m(r 2 + r 2 2 ) = 0
2 r
s
1 2 k (p) 2k
mv = v(r) = .
2 r mr
This gives us the required result,
(p) (c)
v(r) = 2v(r) .
Problem 2
The equation of the orbit for the force given in this problem is
2 1 1 mr 2 mk m

+ = 2 F (r) = 2 + 2 , (1)
2 r r l l l r
that can be written as
2
!
1 m 1 mk

+ 1 2 = 2 . (2)
2 r l r l
This equation is of the form
2u mk
2
+ 2u = 2 , (3)
l
2 2
with = (1 m/l ) and gives different solutions according to the different values of .
2.a) 2 = (1 m/l2 ) > 0 < l2 /m,

the equation is the same we encountered for the case of gravitational interaction, with the
only difference that now 2 6= 1. The solution of the equation is of the form:
mk
u() = r 1 () = A cos() + , (4)
l2 2
where we have chosen the arbitrary phase = 0. Depending on the value of the energy
the orbit is either an ellipse, or a parabola or an hyperbola. In the case of bounded orbits
(ellipse) we notice that the main difference with respect to the pure gravitational interaction
is that the apsidal points are now reached at = 0 and = / > , which indicates a
precession of the orbit.
2.b) 2 = (1 m/l2 ) = 0 = l2 /m,

the solution of the equation is simply
mk 2
u() = r 1 () = + c1 + c2 , (5)
2l2
showing that the radial distance decreases for increasing angles with an inverse quadratic
law. The particle spirals in towards the center of force.
2.c) 2 = (1 m/l2 ) < 0 > l2 /m,

the solution is as for case 2.a), but expressed in terms of hyperbolic functions instead of
harmonic functions, i.e.
mk
u() = r 1 () = A cosh() + , (6)
l2 2
showing that the radial distance decreases for increasing angles with a negative exponential
law. In this case as well, the particle spirals in towards the center of force.
r

r
cos = 2a 2a

3.a)
We have for the equation of the orbit,
d2 1 1 mr 2

+ = 2 F (r). (7)
d2 r r l
Since r = 2a cos (see figure), we can write
d 1 d 1 1 d 1

= =
d r d 2a cos 2a d cos
1 sin
=
2a cos2
d2 1 cos3 + sin2 2 cos
! !
1 1 d sin

= =
d2 r 2a d cos2 2a cos4
sin2
" #
1 1
= +2 3
2a cos cos
1 1 2

= 1+ 2
2a cos cos2
1 2 4a2
" #
1 2

= 1 = 1
2a cos cos2 r r2
1 8a2
" #
= 1
r r2
Now plugging this into (7), we find
1 8a2 mr 2
" #
1
1 + = F (r),
r r2 r l2
l2 1 8a2 8a2 l2 1
F (r) = = .
m r2 r3 m r5
3.b)
Finding the potential and kinetic energies:
2 2
r
8a l 1 8a2 l2 11
Z
U(r) = dr = 4
m r 5 m 4r
2a2 l2 1
=
m r4
1
T = m(r 2 + r 2 2 )
2
1 h 2 + 4a2 cos2 2
i
= m (2a sin )
2
1 h 2 2 i
= m 4a (sin2 + cos2 ) = 2ma2 2 , now using l = mr 2
2
l2 2a2 l2 1
= 2ma2 2 4 =
mr m r4
2 2
2a l 1 2a2 l2 1
E = T +U = = 0.
m r4 m r4
3.c)
From the figure we know that
dr
= 2a sin ,
d
and so using the chain rule
dr dt d
= 2a sin , with = we have
dt d dt
dr l
= 2a sin = 2a sin 2
dt mr s
l r2 l
= 2a 1 cos2 2 = 2a 1 2
mr 4a mr 2
l
= 4a2 r 2 2 .
mr
Now we integrate this equation over one orbit to find the period. To avoid any possible problems
with multiple-valued functions, we will integrate from rmax = 2a to rmin = 0 and multiply the
result by two.
m r 2 dr
dt =
l 4a2 r 2
m r 2 dr
Z 0 !
T = 2 2
2a l 4a r 2
2m r 2 dr
Z 0
=
l 2a 4a2 r 2
#0
r 2 4a2
"
2m 2 1 r
= 4a r + sin
l 2 2 2a 2a
2m 4a2 ma2
" #
2m 2
= = a = 2 .
l 2 2 l l
3.d)
From the definition of in the figure we see that = 2, so we can write our Cartesian coordinates
as
n x = a + a cos
y = a sin

n x = a + a cos 2 = a(2 cos2 1 + 1) = 2a cos2
y = a sin 2 = 2a sin cos
Take a derivative with respect to time of these to find x and y:

l
x = 2a2 cos sin = 4a 2 cos sin
s mrs
2l r 2 2l 1 r2
= 2r 1 2 = 1 2
mr 4a mr 4a
l
y = 2a cos 2 = 2a 2 (2 cos2 1)
mr
2a2
" #
l 1 2 2al l
= 2a 2 r = 1 2 .
mr a mr 2 ma r
Here we have used r = 2a cos and 4a2 r 2 = 2a sin . Now we can see that as r 0,
n x q
v = x 2 + y 2 .
y
Problem 4
The effective potential is
l2 C
2
V (r) =
3 , (8)
2mr 3r
it becomes infinitely negative for r 0, it has a positive maximum for r = r0 (see (1.a)), and
vanishes for r , as illustrated in the plot.
1/r^2
-1/3r^3
4 V(r)
2
V(r)
-2
-4
0 0.5 1 1.5 2 2.5 3

r
4.a)
The extrema of the effective potential are found by searching the zeros of the derivative of V (r)
with respect to r,
l2 l2
" #
dV (r) C 1 C
= 3 + 4 = 3 + , (9)
dr mr r r m r
i.e.
dV (r) Cm
= 0 r = r0 = 2 . (10)
dr l
Both from the plot of V (r) and from the fact that

d2 V (r)

3C 4C C
2
= 5 5 = 5 <0 , (11)
dr r=r0 r0 r0 r0

we see that r = r0 is a maximum of the effective potential. The maximum value of the effective
potential is
l2 C C 1 C 1 C 1
V (r)max = V (r0 ) = 2
3 = 3
3
= . (12)
2mr0 3r0 2 r0 3 r0 6 r03
4.b)
If a particle comes in with energy E > V (r)max it can continue all the way to r = 0, i.e. it is
captured by the center of force. The condition on the energy gives automatically a condition on
the impact parameter since
l6 (mv0 b)6 m3 v06 b6

Emin = V (r)max = = = , (13)
6C 2 m3 6C 2 m3 6C 2
and then !1/6
1 m3 v06 b6 3C 2
E > Emin mv02 > b < bmax = . (14)
2 6C 2 m2 v04
Consequently the cross section for capture is
!1/3
3C 2
capture = b2max = . (15)
m2 v04
The planet with mass M is moving on a highly eccentric elliptic orbit, so its energy is
k
E0 = ,
2a
where note that 1 E 0, but not quite. In other words you need just a small change E
to get to E = 0. At the aphelion (where it has velocity v0 , all in the tangent direction) it is hit by
a comet (with mass m << M) moving with velocity vc in the same tangent direction. Thus the
collision is head-on with both objects moving in the same direction. The collision is assumed to
be completely inelastic, i.e. the two objects stick together forming a new object of mass (M + m).
Linear momentum is conserved while kinetic energy is not.
Conservation of linear momentum gives
Mv0 + mvc = (M + m)(v0 + v)

1
vc = [(M + m)(v0 + v) Mv0 ]
m
1
= [Mv + mv0 + o(mv)] (16)
m
M
v + v0 .
m
In the second to last line we have kept o(m/M) but are going to discard o(m/Mv) since v is a
small quantity because v0 is almost small enough to make E = 0 (but not quite!). The energy is
not conserved, and in fact the collision must increase the energy from the initial value E0 (E0 < 0)
to E = 0 in order for the resulting orbit to be parabolic. So, the comet has to have a kinetic
energy
1
Tc(min) = mvc2 (17)
2
so that the final energy is 0:
1 2 k 1 k
E = vf = (M + m)(v0 + v)2
2 r 2 r
1 1
= Mv02 + Mv0 v + mv02 + o(v 2 ) + o(mv)
2 2
1 2
= E0 + Mv0 v + mv0 ,
2
where E0 = 12 Mv02 kr . Setting E = 0 we get
E0 1m
v = v0 .
Mv0 2 M
Plugging this into (16) we find
M E0 1m

vc = v0 + v0
m Mv0 2 M
E0 1 E0 1
= v0 + v0 = + v0 .
mv0 2 mv0 2
Plugging this into (17) we finally find
2
1 E0 1

Tc(min) = m + v0
2 mr0 2
2
1 E0 1 E0 1 1

= m 2 2+ m + m v02
2 m v0 2 m 2 4
2
1 E0 E0 1 2
= 2
+ mv0 .
2 mv0 2 8
Now we can finally derive v0 in terms of the constant parameters of the orbit, the mass of planet
and k. From angular momentum conservation we have that (at the aphelion, and using = 1 )
l = Mrmax v0 = Ma(1 + )v0 = Ma(2 )v0

l2 aMk(1 2 )
v02 = =
M 2 a2 (2 )2 M 2 a2 (2 )2
aMk(1 (1 )2 ) k
= 2 2 2
.
M a (2 ) 2Ma
So now we have
1 E02 E0 1 k
Tc(min) = k + m
2 m 2M a 2 8 2Ma
M k k 1 m k
= + +
m 4a 4a 16 M a
k 1M m k

= +1 + .
4a m M 16a
Note that the second term is small compared to the first since << 1 and m/M << 1. Therefore
1M k
Tc(min) .
m 4a
October 14th , 2015

Assignment # 8
(Graded problems are due Friday October 21st , 2015)
1 Graded problems
1. Determine the eigenfrequencies and describe the normal mode motion for two pendula of
equal lengths b and equal masses m connected by a spring of force constant . The spring
is unstreched in the equilibrium position.
2. An idealized linear classical water molecule consists of three particles in a line connected by
equal springs and constrained to move along the line joining them. The outer two particles
have mass , the central one has mass , and the spring constant is k.
(2.a) Find the normal modes (describe them) and the normal frequencies.
(2.b) Write down the general solution.
(2.c) Write down the solution with initial conditions x1 (0) = A, x2 (0) = A/, x3 (0) = 0,
and x i (0) = 0 for i = 1, 2, 3.
1 Graded problems
1. In coordinates (1 , 2 ) (see figure), we have
1 1
T = m(b1 )2 + m(b2 )2
2 2
1
V = mgb(1 cos 1 ) + mgb(1 cos 2 ) + k(b sin 1 b sin 2 )2 .
2
Note the equilibrium length of the spring does not appear because of the manner in which
the problem is presented (unstreched in the equilibrium position). Now we expand these
coordinates about the equilibrium 0,1 = 0,2 = 0, and using the small displacements (1 , 2 )
we can make the following approximations:
i 1 1 2
sin i = , cos i 1 i2 = 1 2i .
b 2 2b
Now our energies are
1 2 1 2
T = m + m
2 1 2 2
1 2 1 2

1
V = mgb 1 1 21 + mgb 1 1 22 + k(1 2 )2 ,
2b 2b 2
and our Lagrangian is
1 mg 2 1
L = m( 12 + 22 ) (1 + 22 ) k(1 2 )2 .
2 2b 2
Our equations of motion (via Euler-Lagrange) are
mg
m1 + k + 1 k2 = 0
b
mg
m2 + k + 2 k1 = 0.
b
Using an ansatz i = ai eit these equations are
mg
m 2 a1 + k + a1 ka2 = 0 (1)
b
mg
m2 a2 + k + a2 ka1 = 0. (2)
b
These equations will have a nontrivial solution only if the determinant vanishes, which gives
us a condition on the eigenfrequencies .
k + mg m 2

k mg 2
2
b
mg
k + m k 2 = 0.
k k+ m 2
b b
mg
r
mg 2 2 k + b
k k + mg
b
k
k+ m = k = = .
b m m
Thus, as expected we have two different eigenfrequencies
r r
g g k
1 = , 2 = +2 ,
b b m
where the would be our positive and negative solutions if we wanted to do a full solution
with initial conditions. For our purposes we only want the normal modes, so we define
normal coordinates as X
i (t) = aij j (t), j (t) = eij t .
j
To find the normal coordinates we first plug 1 into equation (1)(setting ai = ai1 to match
our normal coordinates), and find
mg g
k+ m a11 ka21 = 0 a11 = a21 .
b b
Now plugging 2 into (2) (and with ai = ai2 ) we find

2k g mg
m + a12 + k + a12 ka22 = 0 a12 = a22 .
m b b
We could normalize these eigenvectors by simply making their length equal to 1:
1
a211 + a221 = 1 a11 = a21 =
2
1
a212 + a222 = 1 a12 = a22 =
2
The transformation between our small displacements and normal coordinates are then

X 1 1 1
i (t) = ~ (t) =
aij j (t) ~ (t).

j
2 1 1
It turns out that this matrix is its own inverse, so the normal coordinates are given by
1 = 12 (1 + 2 )

~ 1 1 1
(t) = ~ (t)
.
2 1 1 2 = 12 (1 2 )
Now to see what the physical meaning of these modes are, simply set the opposite one
to zero. When 1 = 0, we have that 1 = 2 , and we have that 2 corresponds to the
antisymmetric mode. When 2 = 0, we have 1 = 2 and 1 is therefore the symmetric phase
(see figure).
2. Choose coordinates as in the figure. Note here that xi are the small displacements. If the
equilibrium length of the springs is l then we can put our coordinate system with x = 0
on the mass , so that the equilibrium positions of the three masses are (l, 0, l). The
Lagrangian for this system is simply
1 1 1
L = (x 21 + x 23 ) + x 22 k[(x1 x2 )2 + (x2 x3 )2 ].
2 2 2
Antisymmetric M ode
Symmetric M ode
(2.a)
In these coordinates, the equations of motion are

X
(Tij xj + Vij xj ) = 0,
j
where the matrices are

1 1 0
T= , V = k 1 2 1 .
0 1 1
Using the ansatz
xj (t) = aj eit
we get the eigenequation
(V T 2 )~
a = 0,
which has nontrivial solutions if and only if the determinant vanishes:

k 2 k 0
2
2

|V T | = k 2k k = 0
0 k k 2
= 2 (k 2 )( 2 2k k) = 0.
The solutions to this characteristic polynomial are
s s
k k
1 = , 2 = 1+2 , 3 = 0.

These are the normal frequencies. To find the P normal modes, generalize our solution to
include a sum of these normal modes xj (t) = k ajk eik t , and so we solve

X (k k )a1k ka2k 0
(Vij Tij k )ajk = 0 ka1k (2k k2 )a2k ka3k = 0 (3)
2
j 0 ka2k (k k )a3k
for k = 1, 2, 3. First, setting k = 1 we see from the first line of (3) that a21 = 0, and from
the second line that a11 = a31 . Thus our eigenvector is

1
a1 = a11 0
(4)
1
The constant a11 is free, and could be fixed with normalization conditions. For k = 2 we
first see from the second line of (3) that a22 = 2 a32 , and plugging this into the first line
we also see that a12 = a32 . Thus our second eigenvector is

1
a2 = a32 2 (5)
1
With k = 3 in (3) we see immediately that a13 = a23 and a23 = a33 so we find

1
a3 = a13 1 (6)
1
Applying the normalization condition
X
Tij air ajs = rs ,
i,j
we find
1
a11 = = , (7)
2
1
a32 = = q
2

2 1 +
1
a13 = . (8)
2 +
Now we can simply invert the equation
X
xj = aij j
i
to find the normal modes j . The solution is

r

1 = (x1 x3 )
2

r
2 = [x1 2x2 + x3 ]
2(2 + )
1
3 = (x1 + x2 + x3 ).
2 +
In the first normal mode (see figure (a)), x1 = x3 , so the two outer masses vibrate out
of phase by 180 and with equal amplitudes. The central mass remains fixed (with zero
amplitude). In the second mode (figure (b)), x1 = x3 , so the two outer masses vibrate in
phase at frequency 2 and with equal amplitudes. Because x2 = 2x1 /, the central mass
vibrates out of phase by 180 at the same frequency and with 2/ times the amplitude. In
the first and second modes the center of mass remains stationary. In the third normal mode
(figure (c)), x1 = x2 = x3 , so the system moves as a whole. The center of mass moves at
some fixed velocity v. Clearly there is no force. This is only due to translation of the CM.
(2.b)
The full solution to this problem is
a1 (C1 ei1 t + C1 ei1 t ) + ~

x(t) = ~
~ a2 (C2 ei2 t + C2 ei2 t ) + ~
a3 (C + v0 t),
where we have required that the solution be real and the motion associated with the CM
has the form C + vo t, ie constant translations in time.
(2.c)
The initial conditions along with the general solution above lead to a set of linear equations:
x1 (0) = C1 + C1 + C2 + C2 + C = A

x2 (0) = 2 (C2 + C2 ) + C = A

x3 (0) = (C1 C1 ) + C2 + C2 + C = 0
x 1 (0) = i1 (C1 C1 ) + i2 (C2 C2 ) + v0 = 0

x 2 (0) = 2 i2 (C2 C2 ) + v0 = 0

x 3 (0) = i1 (C1 C1 ) + i(C2 C2 ) + v0 = 0.
There are most easily solved with a matrix method,

1 1 1 1 1 0 C1 A
0
0 2 2 1 0
C1 A
1 1 1 1 1 0 C2 = 0

1 1 2 2 0 i C2 0

0
0 2 2 2 2 0 i C 0
1 1 2 2 0 i v0 0
Using standard reduction techniques one can see that the solution to this equation is
C = v0 = 0,
A
C1 = C2 = .
4

k k
a)
b)
c)
d)
Thus the solution is

cos 1 t + cos 2 t
A
x(t) = 2 cos 2 t
~
2
cos 1 t + cos 2 t
This situation is shown in figure (d). The CM remains fixed, the central mass performs simple
harmonic motion at 2 and the other two masses move in a combination of the frequencies 1
and 2 . If these two frequencies are close to each other, we have the phenomena of beating,
although it is out of phase. In other words, when the amplitude of one of them is small, the
amplitude of the other is large.
October 21st , 2015
Assignment # 9
(Graded problems are due Wednesday October 28th , 2015)
1 Graded problems
1. Find the frequency of small oscillations for a thin homogeneous plate if the motion takes
place in the plane of the plate and if the plate has the shape of an equilateral triangle and
is suspended (a) from the midpoint of one side and (b) from the apex.
2. A homogeneous cube, each edge of which has a length l, is initially in a position of unstable
equilibrium with one edge in contact with a horizontal plane. The cube is then given a small
displacement and allowed to fall. Show that the angular velocity of the cube when one face
strikes the plane is given by
g
2 = A ( 2 1) ,
l
where A = 3/2 is the edge cannot slide on the plane and where A = 12/5 if sliding can occur
without friction.
3. A uniform right circular cone of height h, half-angle , and density rolls on its side without
slipping on a uniform horizontal plane in such a manner that it returns to its original position
in a time . Find expressions for the kinetic energy and the components of the angular
momentum of the cone.
4. A square sheet is constrained to rotate with an angular velocity about an axis passing
through its center and making an angle with the axis through the center of mass and
normal to the sheet (i.e. its axis of symmetry). At the instant the axis of rotation lies in
the plane determined by the axis of symmetry and a diagonal, the body is released. Find
the rate at which the axis of symmetry precesses about the constant direction of the angular
momentum.

5. Show that none of the principal moments of inertia can exceed the sum of the other two.
6. Calculate the moments of inertia I1 , I2 , and I3 for a homogeneous cone of mass M whose
height is h and whose base has a radius R. Choose the x3 -axis along the axis of the cone.
Choose the origin at the apex of the cone, and calculate the elements of the inertia tensor.
Then make a transformation such that the center of mass of the cone becomes the origin,
and find the principal moments of inertia.
7. Consider a thin disk composed of two homogeneous halves connected along a diameter of
the disk. If one half has density and the other has density 2, find the expression for
the Lagrangian when the disk rolls without slipping along a horizontal surface (the rotation
takes place in the plane of the disk).
1 Graded Problems
Problem 1
(1.a)
In order to find the equation of motion of the triangle,
we need to write the Lagrangian, with generalized coor-
y dinate {}. The potential energy is going to be based on
the location of the center of mass, and we find that via
l/2 l/2
Z l/2 Z 0
x yCM ( = 0) = 2 dx dyy
m 0 3(l/2x)
!2
2 m Z l/2 1 l
CM = 1 3
m l l 0
dx 3
2 2
x
2 2
!2
8 31 l l/2
= 2 x
3l 2 3 2 0
4 l3 l
23 l = 2 = ,
3l 8 2 3
where the density is . This is as expected by symmetry
considerations and by the fact that the CM has to be at the geometrical center of the triangle,
i.e. at one third of its height,
1 3 l
l= .
3 2 2 3
Note the negative sign is because the triangle lies below the x-axis. So, for generalized ,

yCM = cos ,
2 3
and the potential energy is
l
U = mgyCM = mg cos .
2 3
To calculate the kinetic energy,
1
T = Trot (about 0) = I3 2 ,
2
we need first to find the principal moment of inertia about the axis of rotation, which is an axis
perpendicular to the plane of the triangle, through 0. This is the only moment of inertia we need,
since
e3 and therefore,
~ =
1 T 1 1
T = ~ I ~ = 2 I33 = 2 I3 .
2 2 2
Z l/2 Z 0
I3 = 2 dy(x2 + y 2)
0 3(l/2x)
Z l/2 Z 0 Z l/2 Z 0
= 2 dxx2 dy + 2 dx dyy 2
0 3(l/2x) 0 3(l/2x)
!3
1 l
!
l/2 l l/2
Z Z
2
= 2 dxx 3 x + 2 3 3 x
0 2 0 3 2
1 l l3 1 l4 1 l4
!
= 2 3 +
32 8 3 16 4 16
m 1 l4 ml2
= 2 3 1 3 = .
l2 3 16
2 2
6
Therefore
ml2 2
!
1
T = ,
2 6
and the Lagrangian is
ml2 2
!
1 l
L= T V = + mg cos .
2 6 2 3
The equation of motion for our generalized coordinate is
d L L ml2 l
=0 + mg sin = 0

dt 6 2 3
g
+
3 sin = 0.
l
Under small oscillations about = 0 this is the equation for a simple harmonic oscillator with
frequency r
g
= 3 .
l
(1.b)
For this situation, the = 0 position of the CM is
yCM ( = 0) = l3 . Therefore
y
l l
yCM = cos U = mg cos .
3 3
We need to calculate the moment of inertia I3 with re- l/2 l/2
spect to the new center of rotation (and axis through x
it).

Z l/2 Z 3x
I3 = 2 dx dx(x2 + y 2)
0 3/2l
Z l/2 Z 3x Z l/2 Z
3x
= 2 dxx 2
dy + 2 dx dyy 2 CM
3 3
0 2
l 0 2
l
Z l3
" ! !#
l/2 l l/2 1
Z
2
= 2 3 dxx x + + dx 3 x3 + 23 l
0 2 0 3 8
1 l4 1 l4 l4
!
1
= 2 3 + l4 +
4 16 3 4 16 16
m l4 3 + 2 + 6 5
= 2 3 1 3 = ml2 .
l2 16
2 2
6 12
Thus the kinetic energy and Lagrangian are

1 5

T = ml2 2
2 12
1 5 l

L=T V = ml2 2 + mg cos ,
2 12 3
and the equation of motion is
d L L 5 l
= 0 ml2 + mg sin = 0

dt 12 3
g 12 1
+ sin = 0.
l 5 3
Thus the frequency of the small oscillation is
s
12 g
= .
5 3l
Problem 2
(a)
y
y
yCM = l cos
2
The cube rotates about the side perpendicular to the plane of the figure, through 0. We are
therefore referring to the case in which the frame has axes parallel to the side of the cube and
origin at one corner. In figure (1), the cube is rotating about the back side, along the y-axis. The
corresponding moment of inertia is I22 = 32 ml2 (also see p. 409-410 in Goldstein). We can find the
angular velocity of the cube when one face strikes the plane, by imposing conservation of energy:
l
Ei = Ui = mg
2
l 1 2 2 2

Ef = Uf + Tf = mg + ml f
2 2 3
l l 1
Ei = Ef mg = mg + ml2 f2 .
2 2 3
s
3
!
1 1 1 g
g = lf2 f = ( 2 1) .
2 2 3 2 l
z
O y
Figure 1:
(b)
If the cube can slide, there is no more fixed point. So, we have to calculate the kinetic energy as
CM CM
T = Ttrans + Trot
1 2 1 1 2 2

= mv + ml f ,
2 CM 2 6
where:
l l
vCM = y CM = sin = ,
2 2
and when = 45circ , the cube hits the plane. With ( = 45 ) = f , we have
1 l2 1 5
T = m f2 + ml2 f2 = lf2 ,
2 4 12 24
Or s
12
!
1 1 5 g
g = lf2 f = ( 2 1) .
2 2 24 5 l
We could also have calculated the kinetic energy as
CM CM
T = Ttrans + Trot
1 2 1 CM 2 1 1 1 2 2

= mv + I = + l f
2 CM 2 22 f 2 2 6
1 2 2 origin
= l f = Trot .
3
Now use v = R to find vCM = l2 f , and on p. 422 of Goldstein we can find I22
CM
= ml2
6
. We
thus obtain the same answer as above.
Take M as the mass of the cone, (x1 , x2 , x3 ) as the principal axes going through the CM, which
is at (0, 0, a = 34 h) in the body system. x3 is the cone axis, x2 is perpendicular to the plane going
through x3 and the segment OA, and x1 is orthogonal to the plane of x2 and x3 . As stated in
the problem, R is the radius, h is the height and is the half-angle of the cone. Due to the
condition of no-slipping, the angular velocity of the cone, ~ , is directed along the instantaneous
axis of rotation which is the line of contact between the surface of the cone and the surface on
which the cone is rolling (i.e. along the segment OA, see figure).
x3
x1
O
a x2

CM R x3
x1
x2 A
~
(a)
The kinetic energy (T ) can be calculated either using the center of mass as origin and taking as
reference frame the system of principal axes described above, or using a system of axes parallel to
that one but with origin at the apex of the cone O, which is a fixed point for the body.
In the first case, the kinetic energy is given by the sum of two terms: the kinetic energy of the
center of mass (i.e. the kinetic energy of a pointlike object of mass M located at the CM of the
cone and moving as CM moves) and the kinetic energy of the body as rotating about the center
of mass (expressed in terms of the principal moments of inertia and the component of the angular
velocity in the principal axes frame or body frame),
CM CM
T = Ttrans + Trot
CM
kinetic energy 1 1
= Mv 2 = Ma2 cos2 2 .
Ttrans = CM
of the CM 2 2
In the above equation we have used that vCM = a cos as it rotates about x
3 . This implies that
(angular velocity about OA) and are related. Thus we also have
vCM = a sin .
Combining these two relations we find
vCM
= = cot . (1)
a sin
The components of ~ in the body frame are
cos2
~ = ( sin , 0, cos ) = ( cos , 0, ).
sin
This follows from our choice of x2 and because of the relation (1). Therefore:
CM 1 1
Trot = I1 12 + I3 32
2 2
n I = I = 3 M R 2 + h2
1 2 20 4
= 3 2
I3 = 10
MR .
h2 4
!
1 3 2 2 2 1 3 2 cos 2
= M R + cos + MR
2 20 4 2 10 sin2
h2 cos4
!
1 1 3 1 3
CM CM 2 2
T = Ttrans + Trot = Ma cos + 2
M R + 2
cos2 2 + MR2 2 2
2 2 20 4 2 10 sin
9 3 1 3

= Mh2 cos2 2 + Mh2 tan2 + cos2 2 + Mh2 cos2 2 .
32 40 4 20
To get this last line we used a = 43 h, R = h tan , and the next line uses the trigonometric identity,
tan2 = cos12 1.
9 3 3 3 3
T = Mh2 cos2 2 Mh2 2 Mh2 cos2 2 + Mh2 cos2 2
32 40 40 4 20
3
= Mh2 2 (1 + 5 cos2 ).
40
In the second case, the kinetic energy is purely rotational, since O is a fixed point, and is given
by:
O 1 1
T = Trot = J1 12 + J3 32 ,
2 2
where J1 and J3 are the principal moments of inertia with respect to a system of axes parallel to
the ones drawn in the picture and with origin in O. They can be found using Steiners theorem
and they are simply given by:
J1 = I1 + M(a2 a1 a1 ) = I1 + Ma2 ,
J3 = I3 + M(a2 a3 a3 ) = I3 ,
where we have used that ~a = (0, 0, a) is the position vector of O in the CM frame. It is then
easy to see that,
1 1 1 1 1 1
O
T = Trot = I1 12 + Ma2 12 + I3 32 = Ma2 cos2 2 + I1 12 + I3 32 = Ttrans
CM CM
+ Trot .
2 2 2 2 2 2
(b)
The components of the angular momentum are given by
~ = L1
L e1 + L2 e2 + L3 e 3
e1 + I2 2
= I1 1 e2 + I3 3
e3 .
Therefore we know that L2 = 0, and
h2
!
3
L1 = I1 1 = M R 2 + cos 2
20 4
3 1 3

= Mh 2
cos ,
20 cos 4
3 cos2 3 cos2
L3 = I3 3 = MR2 = Mh2 tan2
10 sin 10 sin
3
= Mh2 sin .
10
Problem 4
In this problem the body axes are the principal axes, and ~ can
move in the the body fixed frame. Its easy to see that the plane is
x3 ~ ~ will be constant
a symmetric top. Therefore, in absence of forces L
and ~ will precess around it.
Let us calculate the moments of inertia explicitly:
l/2 l/2 1 2l3 l
Z Z
l I1 = I2 = dx dy x2 = 2
x2 l/2 l/2 3 8 2
O
2
ml
=
x1 12
l/2 l/2 ml2
Z Z
I3 = dx dy (x2 + y 2 ) = .
l/2 l/2 6
Now at t = 0,
!
sin sin
~ = , , cos ,
2 2
and the angular momentum is

2
!
~ = (I1 1 , I2 2 , I3 3 ) = ml
L
sin sin
, , 2 sin .
12 2 2
~ is (see discussion in class and in the text):

The velocity with which ~ precesses about L
L
pr = ,
I1
where
#1/2
ml2 sin2 sin2
"
L = (I1 12 + I2 22 + I3 32 )1/2 = + + cos2
6 8 8
ml2
= (1 + 3 cos2 )1/2 .
12
And so the frequency of precession is
(ml2 /12)(1 + 3 cos2 )1/2

pr = = (1 + 3 cos2 )1/2 .
ml2 /12
Problem 5
Take a generic point P and the principal axis going through it (see
figure). The principal moments of inertia are
Z h i
I1 = d3 rf (~r) ~r2 x21 x3
ZV h i
I2 = d3 rf (~r) ~r2 x22
ZV h i
I3 = d3 rf (~r) ~r2 x23 ,
V V
where f (~r) is the mass density. Therefore
Z h i
Ii + Ij = d3 rf (~r) 2~r2 x2i x2j
ZV
x2
h i
= d3 rf (~r) x2i + x2j + 2x2k
P
ZV
~ x2 + x2
h i
d3 rf (R) i j
ZV h i
d3 rf (~r) ~r2 x2k = Ik . x1
V
The equality in the above statement occurs when xk = 0. Here we
have assumed that i 6= j 6= k for i, j, k = 1, 2, 3.
Problem 6
Using cylindrical coordinates,
( x = r cos
z y = r sin
z = z,
and we have that r = (R/h)z. First determine the inertia tensor
with respect to 0:
CM Z h Z 2 Z Rz/h
I11 = dz d dr r(y 2 + z 2 )
0 0 0
3h
4 M Z h
Z Z Rz/h
2
= 1 2
dz d drR2 (r 2 sin2 + z 2 )
3
R h 0 0 0
z " Z
4 2
#
M h 1R 4 21 R
Z h
2
= 1 z dz + 2 z z dz
y
3
R2 h 0 4 h4 0 2 h2
r
M 1 2 1 2 3

2
= 1 2
R h R + h = M(R2 + 4h2 )
3
R h 20 5 20
x
I22 = I11 by symmetry
M
Z h Z 2 Z Rz/h
I33 = 1 2h 0
dz d dr r(x2 + y 2)
3
R 0 0
M R4 h5 3
= 1 2
2 4
= MR2 .
2
R h 4h 5 10
By symmetry we can also say that the non diagonal terms are zero (each of the chosen axes is a
symmetry axis for the cone). These can also be shown explicitly. For instance:
M h 2 Rz/h
Z Z Z
I12 = 1 dz d dr rxy = 0.
3
R2 h 0 0 0
This vanishes because xy = r 2 sin cos and

Z 2
d sin cos = 0.
0
Now in a similar way the other two vanish
M h 2 Rz/h
Z Z Z
I13 = 1 2 dz d dr rxz = 0,
3
R h 0 0 0
M
Z h Z 2 Z Rz/h
I23 = 1 2 dz d dr ryz = 0
3
R h 0 0 0
where we have used Z 2

xz = r cos z d cos = 0,
0
Z 2
yz = r sin z d sin = 0.
0
So, the chosen Cartesian axes are also principal axes with respect to the origin, and
3
(
20
MR2 0 0 )
I0 = 0 3
20
MR2 0
3 2
0 0 10
MR
To obtain ICM , we need first to determine (xCM , yCM , zCM ). By symmetry, we know xCM =
yCM = 0, while:
1 M h 2 Rz/h Z Z Z
ICM = 1 dz dr rz
M 3 R2 h 0 0 0
3 Z h
1 R2 2
= 2 dz z z
R2 h 0 2 h2
3 1 R2 h4 3
= 2 = h.
R2 h 2 h2 4 4
Finally, we can apply the generalized form of Steiners parallel axis theorem, according to which
h i
(ICM )ij = (I0 )ij M a2 ij ai aj ,
which in our case ~a = (0, 0, 43 h), so we have
3 2 3 9

(ICM )11 = (I0 )11 h M = M(R2 + 4h2 ) h2 M
4 ! 20 16
2
3 h
= M R2
20 4
(ICM )22 = (ICM )11 by symmetry
3
(ICM )33 = (I0 )33 = MR2 ,
10
and the non diagonal elements are still all zero.
Problem 7
Defining as the mass density, we use polar coordinates
(
x = r cos
y
y = r sin (x, y)
Given a coordinate system (x , y ) which rotates with the disk (see
figure (3)), the location of the CM is xCM = 0 and

Z Z R r
yCM = dxdyy + 2 dxdyy
M top bottom x
(Z ) 2
R Z Z R Z 2
= drr dr sin + 2 drr dr sin
M 0 0 0
R3 R3 2 R3
( )
4R
= 2 4 = = .
M 3 3 3 M 9
Where in the last line we have used Figure 2:
2 2
R R 3
M = + 2 = R2 .
2 2 2
Again referencing Figures (2) and (3) we see that the relationship
between the lab coordinates and the coordinates of the center of
mass are
n x
CM = R | yCM | sin = R 4R 9
sin
4R
yCM = R | yCM | cos = R 9 cos .

n x CM = R 4R
9
cos
y CM = 4R
sin .
9
The kinetic energy of the disk is made of 2 terms:
T = Ttrans + Trot ,
where Ttrans is the translational kinetic energy of the mass M with coordinates (xCM , yCM ) and
velocity (x CM , y CM ), and Trot is the rotational kinetic energy about the center of mass. Thus
1 1
T = M(x 2CM + y CM
2
) + I3 2 .
2 2
Here I3 is the moment of inertia about the CM, with respect to the z-axis, perpendicular to the
plane of the figure. To find I3 we will calculate it with respect to the center of disk (since it is
easier!) and use Steiners theorem to obtain it with respect to the CM.
Z R Z Z R Z 2
2 2
I3 = drr d(x + y ) + 2 drr d(x2 + y 2)

0 0 0 0
R4 R4 3 1
= + 2 = R4 = MR2
4 4 4 2

2 1 2 16 R2
I3 = I3 |0 M yCM = MR M

CM 2 81 2
1 32

= MR2 1 .
2 81 2
Then we can calculate the kinetic energy
1 1
T = Ttrans + Trot = M(x 2CM + y CM 2
) + I3 2
2 2 CM
" 2 2 #
1 4 4 1 32

2 2 2 2
= MR 1 cos + sin + MR 1
2 9 9 4 81 2
1 16 8 1 16

= MR2 2 1 + 2
cos +
2 81 9 2 81 2
1 3 8

= MR2 2 cos .
2 2 9
The potential energy must also be defined with respect to the center of mass,
4

U = MgyCM = MgR 1 cos ,
9
and the Lagrangian is the sum of these two terms,
1 3 8 4

L = T V = MR2 2 cos mgR 1 cos .
2 2 9 9
y
y

x
2
yCM
(xCM , yCM )
x
x
Figure 3:
November 16th , 2015

Assignment # 11
(Graded problems are due Monday November 23rd , 2015)
1 Graded problems
1. Consider a particle in a central force field.
1.a) Obtain the Hamiltonian and the canonical equations of motion.

1.b) Take two of the initial conditions to be p (0) = 0 and (0) = 0. Discuss the resulting
simplification of the canonical equations.
1.c) Consider now an attractive force of magnitude k/r2 : use plane polar coordinates and
find Hamilton equations of motion.
2. A particle of mass m moves in one dimension under the influence of a force

k t/
F (x, t) = e ,
x2
where k and are positive constants.
2.a) Compute the Lagrangian and Hamiltonian functions.

2.b) Compare the Hamiltonian and the total energy, and discuss the conservation of energy
for the system.
3. The Lagrangian for a particle of mass m and electric charge q moving under the influence
of a magnetic (but not electric) field is given by:
1 q
L(r, r ) = mr2 + r A(r) ,
2 c
where A(r) is the vector potential. Assume that the magnetic field is constant and given by
B(r) = B0 z.
3.a) Show that for such a constant magnetic field the vector potential can be written in the
form A(r) = 21 (B r). That is, show that such a vector potential satisfies: A = B.
3.b) Construct the Hamiltonian of the system in terms of the Cartesian coordinates and the
corresponding canonical momenta of the particle.
3.c) Denoting by = mr = mv the mechanical momentum of the particle, evaluate the
following Poisson brackets:
{ x , y } , { y , z } , { z , x } .
3.d) By re-expressing the Hamiltonian of part 3.b) in terms of the mechanical momentum of
the particle, and using the results derived in part 3.c), obtain the most general solution
for (t) by using the Poissons equation:
d
= {H,} ,
dt
where (using Landaus convention),
!
X f g f g
{f, g} = .
k pk qk qk pk
Interpret your results on the basis of a conventional (Newtons second Law plus Lorentz
force) approach.

8. Take any problem solved using Euler-Lagrange equations, derive the corresponding set of
Hamilton equations, and show that you can put them in a form equivalent to the Euler-
Lagrange ones. Ask yourself questions like: does the system admit constants of motion,
how do you use them in finding the dynamical evolution of a system, which formalism
(Lagrangian vs. Hamiltonian) better suits which problem, etc.
November 16th, 2015

1 Graded problems
Problem 1
1.a)
The Lagrangian is
1
L = m(r 2 + r 2 2 + r 2 sin2 2 ) V (r), (1)
2
and the conjugate momenta are
L
pr = = mr , (2)
r
L
p = = mr 2 ,

L
p = = mr 2 sin2 . (3)

By integrating these relations we get
pr
r = , (4)
m
p
= ,
mr 2
p
= .
mr sin2
2
Thus the Hamiltonian is

p2 p2

1 2
H = pi qi L = p + + + V (r). (5)
2m r r 2 r 2 sin2
Using our class discussion and Goldstein 8.1 we can find
a = 0,
~ (6)

1 1

T = m r2 T1 = 1 1 ,
r2
m
r 2 sin2 1
r 2 sin2

pr
1 T 1
H = p
~ T p ~ + V with p ~ = p .
2
p
Thus we see that
H =T +V =E ,
as expected. Hamiltons equations of motion are
H
= pmr
(

r r = pr h p2
i (7)
pr pr = H
r
= mr1 3 p2 + sin2
V (r)

( p
H
= = mr

p 2
p2 cos
p p = H =
mr 2 sin3
( p
= H
= mr2 sin

p 3
p H
p = = 0
1.b)
If p (0) = 0 then p (t) = 0 for all times (since p = 0 from the equations of motion). Then the
equations of motion become
( ( (
r = pmr = mr p
= 0
p2
2
(8)
p r = mr3 V (r) p = 0 p = 0
Thus if (0) = 0 (t) = 0 at all times, and the motion is planar (in the = 0 plane) as we
would expect. Given the initial conditions, it will be in the = 0 plane. The (r, pr ) and (, p )
sets of equations reduce to the usual equations for central-force motion:
(
= mr p 2
2 p = mr , conserved.
(9)
p = 0 p = 0 mr(r + 2r )
= 0 = ma = 0.
So we have that p = l is the magnitude of the angular momentum, and from p = 0 we get
Newtons second law in the direction. For the radial part:
(
r = pmr
p2 l2 (10)
p r = mr3 V (r) = mr 3 V (r) .

Taking a derivative of the first equation and plugging it into the second equation we find
p r l2 V (r)
r = = 2 3 . (11)
m mr m
Thus we can write
l2
p r = m
r= V (r) = mr 2 V (r) , (12)
mr 3
r r 2 ) = mar = V (r) = F (r) .
= m( (13)
So we find this set of equations gives us Newtons 2nd law in the radial direction.
1.c)
~ (r) = k2 r, since this is a
For the specific case of F r
conservative force with no constraint we already know
y that
H =T +U =E ,
m for constant E. From the diagram we see the coordinates
are
~r x = r cos
F~ = v 2 = r 2 + r 2 2 . (14)
y = r sin
The kinetic and potential energy, and the Lagrangian are
x 1
O T = m(r 2 + r 2 2 ) , (15)
2Z
k k
U = 2 dr = ,
r r
1 k
L = T U = m(r 2 + r 2 2 ) + .
2 r
For the conjugate momenta we have
(
L pr
pr = r
= mr r = m
(16)
p = L

= mr =
2 p
mr 2
Our Hamiltonian is
2
H = rp L = pr + p k = T + U ,
r + p (17)
2m 2mr 2 r
and E is conserved since
H
= 0 E = constant . (18)
t
The Hamiltonian (canonical) equations of motion are
H
pr
= r H
p
=
H H (19)
r
= p r
= p
From the equations of motion for (, p ) we get;
(
H
= = mr p
p
H
2
p = mr 2 = constant, (20)

= p = 0
which shows the angular momentum is conserved. From the set of equations for (r, pr ), we find
H pr
= r = , (21)
pr m
H p2 k
= p r = 3
+ 2
r mr r
p2 k k
p r = m
r= 3
2 = mr 2 . (22)
mr r r
Problem 2
2.a)
Our particle moves in 1D, so we use one generalized coordinate x. The potential is given by
integrating the force:
Z
k t/ k
F (x, t) = 2 e U(x, t) = F (x, t)dx = et/ + C , (23)
x x
where we assume as x that U 0 so we take C = 0. The kinetic energy and the Lagrangian
are therefore
1
T = mx 2 , (24)
2
1 k t/
= L = T U = mx 2 e . (25)
2 x
The conjugate momentum of x is
L px
px = = mx x = . (26)
x m
so the Hamiltonian is
1 p2x k t/
H = xp
xL= + e =T +U =E . (27)
2m x
2.b)
From above we see H = E, since there are no constraints and U is not a function of x. However,
since U = U(x, t) (explicitly depends on time!), the energy of the system is not conserved:
dE dH H
= = 6= 0 . (28)
dt dt t
Problem 3
3.a)
~ r ) = B0 z, and we can verify that the vector potential
The magnetic field is given to us as B(~
~ 1 ~
r) = 2 B ~
A(~ ~ ~ ~
r satisfies B = A in the following way:

~ ~ 1
( A)i = ijk j Ak = ijk j klm Bl xm (29)
2
1
= ijk klm Bl lm
2
1 1
= ijk klj Bl = 2il Bl = Bi .
2 2
This implies exactly that
~ = 1 B0 z ~
A
1
r = B0 (x
y yx
) . (30)
2 2
3.b)
The Lagrangian is
1 2 q ~
L(~r,~r) = m~r + ~r A(~ r) (31)
2 c
1 q
= m(x 2 + y 2 + z 2 ) + (xA
x + yA y + zA
z)
2 c
1 qB0
= m(x 2 + y 2 + z 2 ) + (y x + xy)
.
2 2c
The conjugate momenta are

L qB0 1 qB0
px = = mx y x = y + px (32)
x 2c m 2c

L qB0 1 qB0
py = = my + x y = + py
y 2c m 2c
L pz
pz = = mz z = .
z m
Thus the Hamiltonian is
H = px x + py y + pz z L (33)
2 2
p2 py py

qB0 qB0 1 2 qB0 qB0 2 qB0 2
= px y + x py x + + (px + p2y + p2z ) y x
2mc m 2mc m m 2m 2mc 4c 4c

1 2 2 2 qB0 qB0 2 qB0 2
= (p + py + pz ) y px y x + py x
2m x 2mc 4c 4c
2 2
1 qB0 1 qB0 1
= px + y + py x + p2z .
2m 2c 2m 2c m
3.c)
The mechanical momenta are ( = mx = p + qB0 y
x x 2c
y = my = py qB 2c
0
x (34)
z = mz
So we have (using Landaus definition of Poissons bracket, as also given in this problem),

qB0 qB0
{x , y } = px + y, py x (35)
2c 2c
2
qB0 qB0 qB0
= {px , py } + {y, py } {px , x} {y, x}
2c 2c 2c

qB0 qB0
= 2= ,
2c c
where we have used that {px , py } = 0 and {y, x} = 0. Similarly we have

qB0
{y , z } = py , pz = 0 , (36)
2cx

qB0
{z , x } = pz , px + y =0.
2c
(3.d)
In terms of the mechanical momenta:
x2 y2 2
H= + + z . (37)
2m 2m 2m
Now using

d~
= {H, ~
} , (38)
dt
we get
( = {H, } = 1 { 2 , } = qB0 ,
x x 2m y x mc y
y = {H, y } = 2m {x , y } = qB
1 2 0
mc x
, (39)
z = {H, z } = 0 .
The last expression implies z (t) = constant = z (0). Taking a derivative of the first expression
and plugging in the second gives
2
qB0 qB0
x = y = x = 2 x =
x + 2 x . (40)
mc mc
The solution to this is

x (t) = A cos(t) + B sin(t), (41)
while for y we have
1
y (t) = (A sin(t) + B cos(t)) = A sin(t) + B cos(t). (42)

With the initial conditions
x (0) = A, y (0) = B , (43)
we can write ( (t) = (0) cos(t) + (0) sin(t)
x x y
y (t) = x (0) sin(t) + y (0) cos(t) (44)
z (t) = z (0).
From Newtons 2nd law we would get
v
d~
d~ q ~ = qB0 ~
a=m
m~ = = ~vB z . (45)
dt dt c mc
qB0
This means that ~ moves precessing about the z-axis with frequency = mc
, as we have found
in the explicit expression for x , y , and z above. Also note that, since
qB0
{H, ~
} = z ,
~ (46)
mc
we have also found (38).
November 23rd , 2015

Assignment # 12
(Graded problems are due Wednesday December 2nd , 2015)
1 Graded problems
1. 1.a) For a one-dimensional system with Hamiltonian
p2 1
H= 2,
2 2q
show that there is a constant of motion given by
pq
D= Ht .
2
1.b) As a generalization of part 1.a), show that for a motion in a plane with the Hamiltonian
H = |p|n arn ,
where p is the vector of the momenta conjugate to the Cartesian cordinates, there is a
constant of motion given by
pr
D= Ht .
n
1.c) The transformation Q = q, p = P is obviously canonical. However, the same
transformation with t time dilatation, Q = q, p = P , t0 = 2 t, is not. Show that,
however, the equations of motion for q and p for the Hamiltonian in part 1.a) are
invariant under this transformation. The constant of motion D is said to be associated
with this invariance.
2. Given a system with Hamiltonian
!
1 1
H= + p2 q 4 ,
2 q2
2.a) find the equation of motion for q;
2.b) find a canonical transformation that reduces H to the Hamiltonian of a harmonic
oscillator. Show that the solution for the transformed variables is such that the equation
of motion found in part 2.a) is satisfied.
3. 3.a) Show that the transformation
p iaq
Q = p + iaq , P = ,
2ia
with a a real constant, is canonical and find a generating function.
3.b) Use the transformation to solve the linear harmonic oscillator problem.

Assignment # 12
(Graded problems are due Wednesday December 2nd , 2015)
1 Graded problems
1. 1.a) For a one-dimensional system with Hamiltonian
p2 1
H= 2,
2 2q
show that there is a constant of motion given by
pq
D= Ht .
2
1.b) As a generalization of part 1.a), show that for a motion in a plane with the Hamiltonian
H = |p|n arn ,
where p is the vector of the momenta conjugate to the Cartesian cordinates, there is a
constant of motion given by
pr
D= Ht .
n
1.c) The transformation Q = q, p = P is obviously canonical. However, the same
transformation with t time dilatation, Q = q, p = P , t0 = 2 t, is not. Show that,
however, the equations of motion for q and p for the Hamiltonian in part 1.a) are
invariant under this transformation. The constant of motion D is said to be associated
with this invariance.
2. Given a system with Hamiltonian
!
1 1
H= + p2 q 4 ,
2 q2
2.a) find the equation of motion for q;
2.b) find a canonical transformation that reduces H to the Hamiltonian of a harmonic
oscillator. Show that the solution for the transformed variables is such that the equation
of motion found in part 2.a) is satisfied.
3. 3.a) Show that the transformation
p iaq
Q = p + iaq , P = ,
2ia
with a a real constant, is canonical and find a generating function.
3.b) Use the transformation to solve the linear harmonic oscillator problem.

1 Graded problems
Problem 1
1.a)
Given the 1-dimensional system
p2 1
H= 2 , (1)
2 2q
we want to show that
pq
D= Ht (2)
2
is a constant of the motion. Indeed,
dD D h pq i D
= [D, H] + = Ht, H + (3)
dt t 2 t
h pq i 1 1
= , H H = [p, H]q + p[q, H] H
2 2 2
p2

1 1 1
= p, 2 q + p q, H
2 2q 2 2
2
1 1 1 p 1
= 3 q+ pp 2
2 2q 2 2 2q
2
1 1 p 1
= 2 + p2 + 2 =0.
2q 2 2 2q
1.b)
Now we consider a plane motion with
p|n arn ,
H = |~ (4)
and we want to show there is a constant of the motion

~~
p r
D= Ht , (5)
n
r (x, y, z) and p
where ~ ~ is the vector of the conjugate momenta to (x, y, z).

dD H ~~
p r
= [D, H] + = Ht, H H (6)
dt t n

~~
p r xpx + ypy + zpz n a
= ,H H = , |~
p| n H
n n r

~~
p r ~~
p r 1
= r , |~p|n a , H
n m rn
" # " #
1 X X 1 X 1
= p i xi , ( p2j )n/2 a pi xi , P 2 n/2 H
n i j
n i
( j xj )
1X n X
2 n/21 1X n2 2xk
= pk 2pk ( pj ) +a H
n k ( j x2j )n/2+1
P
n k 2 j
p|)n + arn (|~

= (|~ p|)n arn .
1.c)
Consider the canonical transformation Q = q, p = P but with the time dilation t0 = 2 t.
Clearly the canonical form of Hamilton equations is not preserved if t scale as in t0 = 2 t, but we
can check that the analytic form of the equations (in terms of (q, p) in one case and (Q, P ) in the
other) is the same. Indeed,
p2 p2

1
q = [q, H] = q, 2 = q, =p,
2 2q 2
p2

1 1 1 1
p = [p, H] = p, 2 = p, 2 = 3 = 3 .
2 2q 2q q q
Implement the transformation we can then derive that
dq 1 dQ 1 dQ dt0 1 dQ 2 dQ
q = = = = = (7)
dt dt dt0 dt dt0 dt0
dQ 1 1 1
0
= q = p = P = P Q = P ,
dt
dp dP dP dt0 dP dP
p = = = 0 = 0 2 = 3 0
dt dt dt dt dt dt
dP 1 1 1 1 1
0
= 3 p = 3 3 = 3 P = 3 ,
dt q Q Q
and prove that the form of Hamiltons equations is preserved.
Problem 2
2.a)
Given the Hamiltonian
1 1
H = ( 2 + p2 q 4 ) , (8)
2 q
the canonical equations of motion are

1 1 2 4
q = [q, H] = q, +p q (9)
2 q2
1 2 2 4 1
= [q , p q ] = 2p[q, p]q 4 = q 4 p ,
2 2
1 1 2 4
p = [p, H] = p, +p q
2 q2

1 1 1
= p, 2 + [p, p2 q 4 ]
2 q 2
1 1 1 1
= (2) 3 p2 4q 3 = 3 2p2 q 3 .
2 q 2 q
2.b)
The canonical transformation that will make H look like the the Hamiltonian of a harmonic
oscillator of position Q and conjugate momentum P is
(
m
( 1
P = q q = mP
1 (10)
1
Q = m pq 2 p = m QP 2
This transformation takes

1
H = H 0 = (P 2 + m2 Q2 )] . (11)
2m
The generating function is
1
F1 = m Q , (12)
q
such that (
F1
Q
= P = mq P = mq
2 . (13)
F1
P
= p = q12 mQ = pm mQ = 1m QP 2
The canonical equations in the (Q, P ) variables are
( 0
Q = H = P ,
P 0 m . (14)
P = Q = mQ
H
which, using Q = Q(q, p) and P = P (q, p) as given in Eq. (10), become
P 1 1 1
Q = (pq 2 + 2pq q) = (15)
m m mq

1 1
p = 2 2pq q
q q
q 1
P = mQ m 2 = m pq 2
q m
4
q = pq as before

1 1 4
p = 2 2pq pq
q q
1
2p2 q 3 as before.
q3
Problem 3
3.a)
We can prove that the transformation
p iaq
Q = p + iaq , P = , (16)
2ia
is canonical either by showing that it preserves the form of Hamiltons equations of motion, or
by verifying that the Jacobian matrix M of the change of variables (q, p) (Q, P ) satisfy the
symplectic condition J = MJMT , or by verifying that the fundamental Poisson brackets are
invariant under such transformation.
The last two proofs are very simple. Indeed, the matrix M is given by,

ia 1
M= , (17)
1
21 2ia
and one can easily verify that

ia 12

T ia 1 0 1 0 1
MJM = = =J . (18)
12 1
2ia
1 0 1
1 2ia 1 0
On the other hand, it is also easy to show that

p iaq 1 1
{Q, P } = p + iaq, = {p, q} + {q, p} = {q, p} (19)
2ia 2 2
such that the form of the fundamental Poisson brackets is preserved.
In order to prove that the canonical form of the equations of motion is preserved we need to
specify the Hamiltonian. From part 1.b) we know that the system is a one-dimensional harmonic
oscillator. Therefore, in terms of (q, p) variables the Hamiltonian is
p2 k
H= + q2 , (20)
2m 2
and the equations of motion are
H
= p = kq , (21)
q
H p
= q = ,
p m
which, upon further derivation with respect to time, can be cast in the form,
q + 2 q = 0 q(t) = D cos(t + ) , (22)

Dk
p = kq p(t) = sin(t + ) ,

where we can recognize the familiar solution for the one-dimensional harmonic oscillator in terms of
two arbitrary constants (D and ) that can be determined from the initial conditions q(t = 0) = q0
and p(t = 0) = p0 .
In order to find the form of Hamiltons equations in terms of the new variables (Q, P ), we
need to find the transformed hamiltonian, H 0 (Q, P ), which is obtained from H(q, p) by replacing
q = q(Q, P ) and p = p(Q, P ), where, through some simple algebra, one see that,
Q 2iaP Q + 2iaP
q= and p = . (23)
2ia 2
q
k
Choosing a = m in Eq. (23), where = m , we can write
p2 + m2 2 q 2
H= = iQP H 0 , (24)
2m
In terms of Q and P variables, Hamiltons equations are now,
H 0
= P = iP , (25)
Q
H 0
= Q = iQ ,
P
which, upon further derivation with respect to time, can be cast in the form
P + 2 P = 0 , (26)
Q + 2Q = 0 ,
of the same form of Eq. (22).
3.b)
The form of the Hamiltonian in terms of Q and P and the corresponding equations of motions have
been found in 1.a). Here we want to find the solution of Eqs. (26) and show that it corresponds
to the solution of Eqs. (22), i.e. to the motion of a one-dimensional harmonic oscillator. It is
indeed obvious from the form of the equations, but, to be pedantic, let us write the solution of
Eqs. (26) (in terms of two arbitrary constants A and B which could be thought as A = Q(t = 0)
and B = P (t = 0)) as
Q(t) = Aeit , (27)

P (t) = Beit .
Substituting it into q(t) of Eq. (23) we get

A it it A A
q(t) = e Be = B cos(t) + i + B sin(t) (28)
2ia 2ia 2ia
= D cos cos(t) D sin sin(t) = D cos(t + ) ,
which corresponds to q(t) in Eq. (22) if we identify

A A
D cos = B and D sin = i +B . (29)
2ia 2ia
Inverting these relations to get A and B,
A = iaD(cos + i sin ) = iaDei , (30)

D D
B = (cos i sin ) = ei ,
2 2
and substituting them in Eq. (23) to obtain p(t) gives
A it 1 D
p(t) = e + iaBeit = iaDei eit ia ei eit (31)
2 2 2
1
iaD ei(t+) ei(t+) = aD sin(t + ) ,

=
2
which corresponds to p(t) of Eq. (22) for a = m = k .
PHY6426/Fall 2007 CLASSICAL MECHANICS
HOMEWORK ASSIGNMENT #8: solutions
due by 9:35 a.m. Mon 10/22
Instructor: D. L. Maslov
maslov@phys.ufl.edu 392-0513 Rm. 2114
Please help your instructor by doing your work neatly.
1. Goldstein, Problem 6.12

Solution: 1.2 are displacement of the masses from their equilibrium positions. Lagrangian
1 1 1 2 2

L = m 12 + m 22 k1 + k22 + 3k (1 2 ) .
2 2 2
Equations of motion
d L L
=
dt 1,2 1,2
or
m
1 = k1 3k (1 2 )
m
2 = k2 + 3k (1 2 )
Looking for solutions in a form of 1,2 = A1.2 eit gives
k k
2 A1 = A1 3 (A1 A2 )
m m
k k
2 A2 = A2 + 3 (A1 A2 )
m m
which is equivalent to finding the roots of the equation
2
4k/m 3k/m
= 0.
3k/m 2 4k/m
2
2 4k/m (3k/m)2 = 0
p
12 = 7k/m; 2 = k/m.
2. Find the period of oscillations for a quartic one-dimensional oscillator

1 4
U (x) = ax .
4
Solution: energy conservation
1
mx 2 + U (x) = E.
2
Turning points, where x = 0, are at x = A. At these points, U (x) = E, so that E = 41 aA4 . Period
Z A
dx
T = 4 p
0 (2/m) (E U (x))
A
dx
Z
= 2 2m q
1 4 1 4
4 aA 4 ax
0
r Z A
2m 1 1
= 4 dx q
a A2 0 4
1 (x/A)
r
2m 1
= 4C ,
a A
2
where
1
1 1 1 1
Z
C= dy p = B , 1.31 . . .
0 1 y4 4 4 2
with B (x, y) = (x) (y) / (x + y) being a beta-function. Finally,

r
m1
T = 7.42 .
a A
Note that the period decreases with increasing amplitude.
3. Using the perturbation theory, find the eigenfrequencies and eigenmodes of a one-dimensional oscillator with
potential energy
1 2 1 4
U (x) = kx + ax
2 4
in the limit a 0. Which dimensionless parameter controls the convergence of the perturbation theory?
Solution: Equation of motion
x = kx ax3
m
or
x + 02 x = x3
02 02

2 3
x + 0 x = x 1 2 x
2
where 02 = k/m and = a/m. Seek for a solution in a form
x = A cos t + x1 . . .
= 0 + 1 + . . .
To leading order in , this gives
1 + 02 x1 = A3 cos3 t + 20 1 A cos t.
x
Transform cos3 t as
1
cos3 t = cos t cos2 t = cos t (1 + cos 2t)
2
1 1 1 1
= cos t + cos t cos 2t = cos t + (cos t + cos 3t)
2 2 2 4
3 1
= cos t + cos 3t.
4 4
The eq-n of motion now take the form
3 1
x1 + 02 x1 = A3 cos t A3 cos t + 20 1 A cos t.
4 4
The term oscillating with frequency in the RHS looks like a force almost at resonance with the original
frequency. This resonance is unphysical as our anharmonic oscillator is an isolated system. To avoid the
resonance, we require that the coefficient of the cos t term vanishes. This gives
3 A2
1 = .
8 0
Finally, the equation for x1 reduces to
A3
x1 + 02 x1 = cos 3t.
4
3
The solution is
A3
x1 = cos 3t.
3202
Collecting everything together, the lowest order solution is
A3
x = A cos t + cos 3t
3202
3 A2 3 A2

= 0 + = 0 1 + 2 .
8 0 8 0
3 A2
The dimensionless parameter controlling the convergence of the perturbative expansion is 8 02 .
PHY6426: CLASSICAL MECHANICS
HOMEWORK ASSIGNMENT #10: Solutions
Canonical Trasformations and Hamilton-Jacobi Equation
due by 9:35 a.m. Mon 11/05
1. Goldstein, Problem 9.10 (note that x is used for q) (33 points)

Solution
a)
p
Q =
x
P = x2
is this a canonical transformation? A criterion for canonicity is
P Q P Q
[P, Q]p,x = =1
p x x p

= 2x = 2 = 1
x
= 1/2
b) Re-write the transformation as
Qx
p =

P = x2 .
Then it is a transformation from (x, Q) to (p, P ) . Hence the generating function is a function of x and Q (type
I)
F Qx
p = =
x
F
P = = x2 .
Q
Integrating the first equation, we get
Qx2
F = .
2
Using the canonicity condition = 1/2, the equivalent form of F is
F = Qx2
Then the result for P is obtained as well.
c) Harmonic oscillator
p2 1
H= + kx2 .
2m 2
Hamiltons equations
H p
x = =
p m
H
p = = kx
x
2
or
p k
x = = x = 2 x
m m
Solution
x 0
x (t) = x0 cos t + sin t,

where x0 and x 0 are the initial coordinate and velocity, correspondingly. Choose x 0 = 0. Then
x (t) = x0 cos t
p (t) = mx = mx0 sin t
New coord. and momentum

p
Q (t) = = m0 tan t
x
P (t) = x2 = x20 cos2 t.
2. Goldstein, Problem 9.32 (33 points)

Solution:
H = q1 p1 q2 p2 aq12 + bq22
p1 aq1
F1 = , F2 = q 1 q 2 .
q2
Calculate the Poisson bracket for F1
H F1 H F1 H F1 H F1
[H, F1 ] = + .
p1 q1 p2 q2 q1 p1 q2 p2
Calculate the derivatives
H
= q1
p1
H
= q2
p2
H
= p1 2aq1
q1
H
= p2 + 2bq2
q2
F1 a
=
q1 q2
F1 p1 aq1
=
q2 q22
F1 1
=
p1 q2
F1
= 0.
p2
Substituting the derivatives into the Poisson brackets, we find

a p1 aq1 1
q1 + (q2 ) 2 (p1 2aq1 ) =0
q2 q2 q2
Jacobi identity (Goldstein, eq. 9.75e) states
[u, [v, w]] + [v, [w, u]] + [w, [u, v]] = 0.

3
If u = H and v and w are the two constants of motion, so that [H, v] = [H, w] = 0 then the Poisson bracket of
v and w is also a constant of motion. Check the Poisson bracket of F1 and F2
F1 F2 F1 F2 F1 F2 F1 F2
[F1 , F2 ] = +
p1 q1 p2 q2 q1 p1 q2 p2
1
= q2 = 1
q2
This bracket is another constant of motion but it is a trivial constant, which is not an algebraic function of q, p.
PHY6426/Fall 2007: CLASSICAL MECHANICS
Homework set #11: solutions
Due: Monday, Nov. 19.
1. Consider a particle of mass m which is constrained to move on the surface of a sphere of radius R. There are
no external forces of any kind on the particle.
a) Derive the Hamiltonian of the particle. Is it conserved? [15 ponts]
b) Using the Hamiltonian equations of motion, prove that the motion of the particle is along a great circle of
the sphere. [15 points]
NB: A great circle on a sphere is a circle on the spheres surface whose center is the same as the center of the
sphere
Solution:
x = R sin cos
y = R sin sin
z = R cos

x = R cos cos sin sin

y = R cos sin + sin cos
z = R sin
The Larangian
m 2 mR2 2
L=T = x + y 2 + z 2 = + 2 sin2 .
2 2
Momenta
L
p = = mR2

L
p = = mR2 sin2

The Hamiltonian
p2 p2
+ p
H = p L= +
2mR2 mR2 sin2
H does not depend on p is conserved
H
p = = 0.

By a suitable choice of the initial conditions, p can always be made equal to zero. e.g., starting the motion at
= 0. Velocity
H p
= = =0
p mR2 sin2
mR2 sin2 = 0.
As cannot be equal to zero for any instant of time, we conclude that = 0 or = const. This equation defines
a great circle.
2
2. Goldstein, Problem 10.5. [30 points]
m 2
S = q + 2 cot t mq csc t
2
m 2 cos t 1
= q + 2 mq
2 sin t sin t
Hamilton-Jacobi equation
" 2 #
S 1 S 2 2 2 S
H p= ,q = +m q =
q 2m q t
S m 2
= q + 2 csc2 t mq cot t csc t
t 2
S
= mq cot t m csc t
q
2
S 2 2
= (mq cot t) + (m csc t) 2m2 2 q cot t csc t
q
" #
2
1 S 2 2 2 1 h 2 2 2 2
i
+m q = m q cot2 t + 1 + (m csc t) 2m2 2 q csc t
2m q 2m
1 2 2 2
= m q + 2 csc2 t 2m2 2 q csc t
2m
Comparing the first and last lines, we see that they are the same.
Show that S generates a correct solution to the equations of motion:
S
p = = m (q cot t csc t)
q
S
= = m ( cot t q csc t)

Solve the last equation for q = q (, , t)

q = cos t sin t
m
q0 = q (t = 0) =

q0 = q(t
= 0) =
m

2
p = m cos t sin t cot t csc t
m

= m sin t + cos t = mq
m
3. A tennis ball of mass m is bouncing off the floor. The total energy of the ball is E. The ball is moving strictly
along the vertical. The collision between the ball and the floor is perfectly elastic.
a. Derive the action-angle variables for the tennis ball and determine the period of motion. [30 points]
p2
E = + mgz
2m
p
p = 2m (E mgz)
3
The action variable

I Z 0 p Z z0 p
J = pdz = () 2m (E mgz)dz + 2m (E mgz)dz
z 0
| 0 {z } | {z }
topbottom bottomtop
Z z0 Z z0
3/2
p p 4 2 E
= 2 2m (E mgz) = 2 2m2 g (z0 z)1/2 dz = m g CE 3/2 ,
0 0 3 mg
where

4 2 1
z0 E/mg; C .
3 mg
Expressing the Hamiltonian in terms of J
J 2/3
H =E=
C 2/3
Frequency
H 2
= =
J 3C 2/3 J 1/3
Period

1 3C 2/3 J 1/3 3 2/3 1/3 1/2 3 1/2 34 2 1
T = = = C C E = CE = E 1/2
2 2 2 2 3 mg
E 1/2
= 2 2 .
mg
b. Suppose that the tennis ball is a quantum-mechanical object. Using the results of part [a], surmise the
dependence of the quantized energy levels on the level number. [10 points]
Assuming the naive quantization rule J = h n, we obtain for the energy of quantized levels
!2/3
J 2/3 3 2 0
En = 2/3 = 2/3 n2 3 m1/3 g 2/3 .
h
C 8
The exact quantum-mechanical formula (see, e.g., S. Flugge, Practical Quantum Mechanics I, Problem 40)
gives in the limit of large n
!2/3
3 2 0
En = 2/3 n2 3 m1/3 g 2/3 .
h
8
By comparing the two results, we see that the correct quantization rule must have been J =
hn. The
wave-function of the problem is known as the Airy function.
Theoretical Dynamics October 01, 2010
Homework 4
Instructor: Dr. Thomas Cohen Submitted by: Vivek Saxena
Goldstein 9.7
Part (a)
F1 (q, Q, t) F2 (q, P, t)
F1
Pi = (1)
Qi
F2 (q, P, t) = F1 (q, Q, t) + Pi Qi (2)
F1 (q, Q, t) F3 (p, Q, t)
F1
pi = (3)
qi
F3 (p, Q, t) = F1 (q, Q, t) pi qi (4)
F1 (q, Q, t) F4 (p, P, t)
F1
pi = (5)
qi
F1
Pi = (6)
Qi
F4 (p, P, t) = F1 (q, Q, t) pi qi + Pi Qi (7)
F2 (q, P, t) F3 (p, Q, t)
F2
pi = (8)
qi
F2
Qi = (9)
Pi
F3 (p, Q, t) = F2 (q, P, t) pi qi Pi Qi (10)
F2 (q, P, t) F4 (p, P, t)
F2
pi = (11)
qi
F4 (p, P, t) = F2 (q, P, t) pi qi (12)
4-1
F3 (p, Q, t) F4 (p, P, t)
F3
Pi = (13)
Qi
F4 (p, P, t) = F3 (p, Q, t) + Pi Qi (14)
Part (b)
For an identity transformation, F2 = qi Pi and by equation (7), the type 4 generating
function is
F4 (p, P, t) = F2 (q, P, t) pi qi (15)

= qi P i p i qi
F2
= 0 as pi = = Pi
qi
For an exchange transformation, F1 = qi Qi and by equation (4), the type 3 generating
function is
F3 (p, Q, t) = F1 (q, Q, t) pi qi (16)

= qi Q i p i qi
F1
= 0 as pi = = Qi (17)
qi
Part (c)
Consider a type 2 generating function F2 (q, P, t) of the old coordinates and the new mo-
menta, of the form
F2 (q, P, t) = fi (q1 , . . . , qn ; t)Pi g(q1 , . . . , qn ; t) (18)
where fi s are a set of independent functions, and gi s are differentiable functions of the old
coordinates and time. The new coordinates Qi are given by
F2
Qi = = fi (q1 , . . . , qn ; t) (19)
Pi
In particular, the function
fi (q1 , . . . , qn ; t) = Rij qj (20)
where Rij is the (i, j)-th element of a N N orthogonal matrix, generates an orthogonal
transformation of the coordinates. Now,
F2 fi g g
pj = = Pi = Rij Pi (21)
qj qj qj qj
This equation can be written in matrix form, as
f g
p= P (22)
q q
4-2
where p denotes the N 1 column vector (p1 , . . . , pN )T , g/q denotes the N 1 column
vector (g/q1 , . . . , g/qn )T , and f
q denotes the N N matrix with entries

f fi
= = Rij (23)
q ij qj
From (22), the new momenta are given by

1

f g
P = p+ (24)
q q

g
= R1 p + (25)
q
= R1 (p + q g) (26)
As R is an orthogonal matrix, RRT = RT R = I, so R1 = RT is also an orthogonal

transformation.
This gives the required result: the new momenta are given by the orthogonal transformation
(R1 ) of an n-dimensional vector (p + q g), whose components are the old momenta (p)
plus a gradient in configuration space (q g).
Goldstein 9.25
Part (a)
The given Hamiltonian is
1 1
H= + p2 q 4 (27)
2 q2
The equation of motion for q is
H
q = = pq 4 (28)
p
Part (b)
Suppose we let Q2 = 1/q 2 and P 2 = p2 q 4 . Then, Q = 1/q and P = pq 2 . Now,
{Q, P } = {1/q, pq 2 }
= {q 1 , pq 2 }
= {q 1 , p}q 2 + p{q 1 , q 2 }
1
q p q 1 p 2

= q +p0
q p p q

1
= 2 q2
q
= 1
4-3
So, the signs on both Q and P cannot be identical. We take
1
Q = (29)
q
P = pq 2 (30)
which is a valid canonical transformation. This gives the Hamiltonian,

1
H(Q, P ) = (P 2 + Q2 ) (31)
2
The equations of motion are
H
Q = =P (32)
P
H
P = = Q (33)
Q
+ Q = 0, the solution to which is of the form Q = A cos t + B sin t. This gives
So, Q
P = Q = B cos t A sin t. Now,
1
q = = (A cos t + B sin t)1 (34)
Q
p = P Q2 = (B cos t A sin t)(A cos t + B sin t)2 (35)
so,
q = (A cos t + B sin t)2 (A sin t + B cos t) (36)
and hence
pq 4 = (B cos tA sin t)(A cos t+B sin t)2 (A cos t+B sin t)4 = (B cos tA sin t)(A cos t+B sin t)2 = q
(37)
So, the solution to the transformed equation for Q satisfies the original equation of motion
for q.
Problem 1
Part (a)
{X, Px } = {x + , px }
= {x, px }
= 1 (38)
{Y, Py } = {y, px }
= 1 (39)
4-4
{Z, Pz } = {z, pz }
= 1 (40)
{X, Py } = {X, Pz } = {Y, Px } = {Y, Pz } = {Z, Px } = {Z, Py } = 0 (41)

{X, X} = {Y, Y } = {Z, Z} = {Px , Px } = {Py , Pz } = {Pz , Pz } = {Px , Py } = {Py , Pz } = {Pz , Px } = 0
(42)
So, this is a canonical transformation. It corresponds to a translated canonical coordinate
system (translation along the x-direction in phase space).
dX
= [X, Px ] = 1 (43)
d
So Px is the generator of the canonical transformation.
Part (b)
{X, Px } = {x cos + y sin , px cos + py sin }

= cos2 {x, px } + sin2 {y, py }
= 1 (44)
{Y, Py } = {x sin + y cos , px sin + py cos }

= sin2 {x, px } + cos2 {y, py }
= 1 (45)
{Z, Pz } = {z, pz }
= 1 (46)
Using properties of the Poisson Bracket, we also have

(48)
So, this is a canonical transformation. It corresponds to a rotation about the z-axis in phase
space.
dX
= x sin + y cos (49)
d
whereas
{X, Lz } = {x cos + y sin , xpy ypx } = x sin y cos (50)
So,
dX
= {X, Lz } (51)
d
Therefore, Lz is the generator of the canonical transformation.
4-5
Part (c)
{X, Px } = {x, px + }
= 1 (52)
{Y, Py } = {y, py }
= 1 (53)
{Z, Pz } = {z, pz }
= 1 (54)

(56)
So, this is a canonical transformation. It corresponds to a translation along the px direction
in phase space. Now,

Px X Px X dPx
{Px , X} = =1= (57)
qi pi pi qi d
Therefore, X is the generator of the canonical transformation.
Part (d)
{X, Px } = {(1 + )x, (1 + )1 px }

= 1 (58)
{Y, Py } = {(1 + )y, (1 + )1 py }

= 1 (59)
{Z, Pz } = {(1 + )z, (1 + )1 pz }

= 1 (60)

(62)
4-6
So, this is a canonical transformation. It is a scaling transformation, which preserves the
volume element in phase space. Suppose g is the generator of the scaling transformation.
Then,
X
= x = [X, g] = [(1 + )x, g] (63)

which implies
x x g x g
= [x, g] = (64)
1+ x px px x
that is,
x g
= (65)
1+ px
the solution to which is
xpx
g= + f (y, z, py , pz ) (66)
1+
As dY /d = y = [(1 + )y, g] and dZ/d = z = [(1 + )z, g], following a similar argument for
Y and Z (or by symmetry) we get
xpx ypy zpz
g= + + + constant (67)
1+ 1+ 1+
as the generator of the scaling transformation.
Problem 2
As is a canonical transformation, we have
i
= {i , g} (68)

So,
H H i
=
i
H
= {i , g}
i
H i g
= Jjk (as is a canonical transformation)
i j k
H i g
= Jjk (as Poisson Brackets are invariant under canonical transformations)
i j k
H g
= ij Jjk
i k
H g
= Jij
i j
= {H, g}
= g (69)
H
But since H is conserved, = 0 and hence g = 0. Therefore, g is conserved.
4-7
Problem 3
The quantity , which was found to be an invariant of the system, can be expressed in
terms of the canonical coordinates x, y, px , py as
1 2 1
(x, y, px , py ) = (px p2y ) + m 2 (x2 y 2 ) (70)
2m 2
As is the conserved generator of a family of canonical transformations parametrized by
an infinitesimal parameter , we must have
x = {x, } (71)
y = {y, } (72)
px = {px , } (73)
py = {px , } (74)
(75)
We consider each condition separately below.
x = {x, }
1 2 1
= {x, (px p2y ) + m 2 (x2 y 2 )}
2m 2
p2x
= {x, }
2m
px
= (76)
m
y = {y, }
1 2 1
= {y, (p p2y ) + m 2 (x2 y 2 )}
2m x 2
p2y
= {y, }
2m
py
= (77)
m
px = {px , }
1 2 1
= {px , (px p2y ) + m 2 (x2 y 2 )}
2m 2
1 2 2
= {px , m x }
2
= m 2 x (78)
py = {py , }
1 2 1
= {py , (px p2y ) + m 2 (x2 y 2 )}
2m 2
1 2 2
= {py , m y }
2
2
= m y (79)
4-8
Now, let = where is a parameter. Then, the above equations become
dx px
= (80)
d m
dy py
= (81)
d m
dpx
= m 2 x (82)
d
dpy
= m 2 y (83)
d
So,
d2 x
+ 2x = 0 (84)
d2
d2 y
+ 2 y = 0 (85)
d2
The solutions to which are
x = A cos() + B sin() (86)

y = C cos( ) + D sin( ) (87)
and correspondingly
px = mA sin() + mB cos() (88)

py = m C sin( ) m D cos( ) (89)
Using the subscript 0 to denote the initial coordinates and momenta, we have
x0 = A (90)
y0 = C (91)
px 0 = mB (92)

py 0 = m D (93)
So,
px
x = x0 cos() + 0 sin()
m
py 0
y = y0 cos( ) sin( )
m
px = mx0 sin() + px 0 cos()

py = m y0 sin( ) + py 0 cos( )
Reverting to the notation in which x0 , px 0 , y0 , py 0 denote the original coordinates and
X, Y, Px , Py denote the canonically transformed coordinates, we get the form of the canon-
ical transformation as
px
X = x cos() + sin() (94)
m
Px = px cos() mx sin() (95)
py
Y = y cos( ) sin( ) (96)
m

Py = py cos( ) + m y sin( ) (97)
4-9
where is an arbitrary parameter, such that = 0 corresponds to the untransformed coor-
dinates. This canonical transformation is composed of two rotations in the 4-dimensional
phase space (one involving X and Px and the other involving Y and Py ), and its generator
is the conserved quantity .
Problem 4
Part (a)
P 2t

1 2
F2 (q, P, t) = q + gt (P mgt) (98)
2 2m
Now,
F2
p = = P mgt (99)
q
F2 1 Pt 1 pt
Q = = q + gt2 = q + gt2 gt2 (100)
P 2 m 2 m
So, the canonical transformation is
P = p + mgt (101)
pt 1 2
Q = q gt (102)
m 2
Part (b)
pt 1 2 pt 1 2
{Q, Q} = {q gt , q gt } = 0 (103)
m 2 m 2
{P, P } = {p + mgt, p + mgt} = 0 (104)
pt 1 2
{Q, P } = {q gt , p + mgt}
m 2
Q P Q P
=
q p p q

t
= (1)(1) (0)
m
= 1 (105)
So, the transformation satisfies the canonical Poisson Bracket relations.
Part (c)
The Lagrangian is
1 2
L(q, q)
= mq mgq (106)
2
4 - 10
The canonical momentum is
L
p =
q
= mq (107)
So the Hamiltonian is
H = pq L
p2
= + mgq (108)
2m
pt
Now, Q = q m 12 gt2 , so
pt 1 2 p2

{Q, H} = q gt , + mgq
m 2 2m
p2

pt
= q, , mgq
2m m
1
= {q, p2 } gt{p, q}
2m
p
= + gt (109)
m
Also
Q p
= gt (110)
t m
So,
dQ Q
= + {Q, H} = 0 (111)
dt t
Also, P = p + mgt, so
p2
{P, H} = {p + mgt, + mgq}
2m
= mg{p, q}
= mg (112)
and
P
= mg (113)
t
So,
dP P
= + {P, H} = 0 (114)
dt t
4 - 11
Part (d)
P2

F2 1 2
= gt(P mgt) + q + gt (mg)
t 2 2m
3 P 2
= P gt mg 2 t2 mgq
2 2m
3 (p + mgt)2
= (p + mgt)gt mg 2 t2 mgq
2 2m
p 2
= mgq mg 2 t2 (115)
2m
So, the Hamiltonian associated with Q, P is
F2
K = H+
t
p2 p2
= + mgq mgq mg 2 t2
2m 2m
= mg 2 t2 (116)
So, the Hamiltonian K is zero up to time-dependent constant term mg 2 t2 , but it is not a

function of P and Q (which are constant with time, since {Q, H} = {P, H} = 0 as shown
above).
Part (e)
Q and P are conserved quantities, that equal the initial position and the initial momentum
respectively. They are constant with time, as q and p vary:
q(t = 0) = Q
p(t = 0) = P
Part (f )
F2
= P mgt = p (117)
q
F2 mg 2 t2 P2
= P gt mgq (118)
t 2 2m
p2
H = + mgq
2m
F2 2

1
= + mgq (119)
2m q
So,
F2
K = H+ = mg 2 t2 (as shown in part d)
t
4 - 12
implies

F2 F2
H q, + = mg 2 t2 (120)
q t
So, the Hamilton-Jacobi equation is satisfied, except for a time-dependent constant term
appearing on the right hand side.
Part (g)
f (Q, P, t) = F2 (q(Q, P, t), P, t)

P 2t

1 2
= q + gt (P mgt)
2 2m
P 2t

pt
= Q+ + gt2 (P mgt)
m 2m
P 2t

(P mgt)t
= Q+ + gt2 (P mgt)
m 2m
2

Pt P t
= Q+ (P mgt)
m 2m
2
P t
= QP + Qmgt gP t2 (121)
2m
So,
f P2
= mgQ 2P gt (122)
t 2m
Also, p = mq = P mgt. So,
p2
L(q, q)
= mgq
2m
1 2 Pt 1 2
= (P mgt) mg Q + gt
2m m 2
1 1
= (P 2 + m2 g 2 t2 2P mgt) mgQ P gt + mg 2 t2
2m 2
P2
= 2P gt mgQ + mg 2 t2
2m
f (Q, P, t)
= + mg 2 t2 (123)
t
f (Q,P,t)
So, L(q(Q, P, t), q(Q,
P, t)) = t up to a time-dependent term mg 2 t2 .
Problem 5
The Hamilton-Jacobi equation, as expressed in the form
S(q, P )
H(q, S(q, P )) + =0 (124)
t
4 - 13
was obtained by constructing a generating function of the form
F = F2 (q, P, t) Qi Pi
where F2 denotes a generic type-2 generating function. For such a choice of F , the Hamil-
tonian K = H + Ft is zero.
Now, consider a type-3 generating function F3 of the old momenta and the new coordi-
nates, such that the Hamiltonian K is zero. Therefore,
K
Q i = =0 (125)
Pi
K
Pi = =0 (126)
Qi
Now,
F3
qi = = (p F3 )i (127)
pi
so,
F3
H(q(Q, p), p, t) + (Q, p, t) = 0 (128)
t
where the old coordinates q have been expressed in terms of the old momenta and the new
coordinates using equation (127). This is a PDE in (n + 1) variables p1 , . . . , pn , t. Let S
denote the solution of this PDE. Then, a solution of the form,
F3 S = S(p
1 , . . . , pn ; 1 , . . . , n+1 ; t) (129)
where Qi = i are the constants of motion (for i = 1, . . . , n), is consistent with equa-
tion (125). Here the constant n+1 must be a constant of integration, so the physically
meaningful solution is of the form
S = S(p
1 , . . . , pn ; 1 , . . . n ; t) (130)
equation (128) can be written as
So, in terms of S,

p, t) + S (Q, p, t) = 0
H(p S, (131)
t
which is of the desired form.
4 - 14
Exercises in Classical Mechanics
CUNY GC, Prof. D. Garanin No.4 Solution

1 Hamiltonian formalism for the double pendulum

(10 points) Consider a double pendulum that consists of two massless rods of length l1 and l2 with masses
m1 and m2 attached to their ends. The first pendulum is attached to a fixed point and can freely swing
about it. The second pendulum is attached to the end of the first one and can freely swing, too. The motion
of both pendulums is confined to a plane, so that it can be described in terms of their angles with respect
to the vertical, 1 and 2 .
a) Write down the Lagrange function for this system.
b) Introduce generalized momenta p1 and p2 and change to the Hamiltonian description. Find the
transformation matrix that yields the velocities 1 and 2 in terms of the momenta p1 and p2 . Write down
the Hamilton function H(1 , p1 , 2 , p2 ) using the transformation matrix.
c) Obtain the Hamilton equations.
Solution: a) Both kinetic and potential energy of the system are the sums of the contributions of the first
and second masses:
L = T U, T = T1 + T2 , U = U1 + U2 . (1)
For the coordinates of the masses 1 and 2 one has
x1 = l1 sin 1 , y1 = l1 cos 1 (2)
and
x2 = l1 sin 1 + l2 sin 2 , y2 = l1 cos 1 + l2 cos 2 , (3)
with the y-axis directed downwards. For the potential energies one has
U1 = m1 gy1 = m1 gl1 cos 1 , U2 = m2 gy2 = m2 g (l1 cos 1 + l2 cos 2 ) . (4)
Kinetic energies are given by

m1 2 2
T1 = l (5)
2 1 1
and
m2 2 m2 h 2 2 i
T2 = x 2 + y 22 = l1 1 + l22 22 + 2l1 l2 cos (1 2 ) 1 2 . (6)
2 2
All together yields
m1 + m2 2 2 m2 2 2
L = l1 1 + l + m2 l1 l2 cos (1 2 ) 1 2
2 2 2 2
+ (m1 + m2 ) gl1 cos 1 + m2 gl2 cos 2 . (7)
b) The generalized momenta are given by
L
p1 = = (m1 + m2 ) l12 1 + m2 l1 l2 cos (1 2 ) 2
1
L
p2 = = m2 l22 2 + m2 l1 l2 cos (1 2 ) 1 . (8)
2
This can be written in the matrix form as

! ! !
p1 1 (m1 + m2 ) l12 m2 l1 l2 cos (1 2 )
=K , K= . (9)
p2 2 m2 l1 l2 cos (1 2 ) m2 l22
Note that the kinetic energy in Eq. (7) can be written as

! !
1 1 1 p1 1
T = 1 2 K = 1 2 = ( p1 + 2 p2 ) (10)
2 2 2 p2 2 1
and that K T = K. The inverse transformation reads
! !
1 1 p1
=K , (11)
2 p2
where
!
1 1 m2 l22 m2 l1 l2 cos (1 2 )
K = , (12)
m2 l12 l22 [m1 + m2 m2 cos2 (1 2 )] m2 l1 l2 cos (1 2 ) (m1 + m2 ) l12
T
again K 1 = K 1 . Using the transposed relation
T
1 2 = p1 p2 K 1 = p1 p2 K 1 , (13)
one can write the kinetic energy, Eq. (10), in the form
!
1 p1
T = p1 p2 K 1 . (14)
2 p2
Now the Hamilton function becomes,

!
1 p1
H=T +U = p1 p2 K 1 (m1 + m2 ) gl1 cos 1 m2 gl2 cos 2 . (15)
2 p2
In the standard form this reads

m2 l22 p21 + (m1 + m2 ) l12 p22 2m2 l1 l2 cos (1 2 ) p1 p2
H= (m1 + m2 ) gl1 cos 1 m2 gl2 cos 2 . (16)
2m2 l12 l22 [m1 + m2 m2 cos2 (1 2 )]
c) The Hamilton equations read
H H
1 = , p1 =
p1 1
H H
2 = , p2 = . (17)
p2 2
The task to work out these equations is left to the reader. The equations for p1 and p2 are pretty cumbersome
since one has to differentiate the denominator. It is best to do with a mathematical software. The whole
system of Hamiltonian equations for the double pendulum is much more cumbersome than the system
of Lagrange equations. The only purpose to consider the Hamilton equations here is to show that the
Hamiltonian formalism is not well suited for engineering-type problems with constraints.
2 Canonical transformations
(10 points)
a) The canonical transformations between two sets of variables are

Q = ln (1 + q cos p) , P = 2 (1 + q cos p) q sin p. (18)
Show directly that this transformation is canonical. Show that

2
FpQ (p, Q) = eQ 1 tan p
is the generating function of this transformation.

b) For what values of and do the equations
Q = q cos (p) , P = q sin (p)
represent a canonical transformation? What is the form of the generating function FpQ (p, Q) in this case?
Solution: a) To check whether a transformation is canonical, one can show that the fundamental Poisson
brackets are invariant:
{Q, P }q,p = {Q, P }Q,P = 1. (19)
Explicit calculation is below:
Q P Q P
{Q, P }q,p =
q p p q
1 1
= cos p 2 [ q sin p q sin p + (1 + q cos p) q cos p]
1 + q cos p 2 q
" #
1 1
+ q sin p sin p + 2 cos p sin p
1 + q cos p q
1 h i
= q cos p sin2 p + cos2 p + q cos3 p + sin2 p + 2 q cos p sin2 p
1 + q cos p
1 h i
= 1 + q cos p sin2 p + q cos3 p
1 + q cos p
1 h i
= 1 + q cos p sin2 p + cos2 p = 1. (20)
1 + q cos p
Now check that 2

FpQ (p, Q) = eQ 1 tan p
is the generating function of this transformation. that is, one has to check the relations
FpQ FpQ
q= , P = . (21)
p Q
One obtains
FpQ Q 2 1
= e 1 . (22)
p cos2 p
On the other hand, from Eq. (18) follows

eQ 1 = q cos p, (23)
so that indeed
FpQ
= q. (24)
p
Further with the help of Eq. (23) one obtains
FpQ
= 2 eQ 1 eQ tan p = (2 q cos p) (1 + 2 q cos p) tan p
Q

= 2 (1 + 2 q cos p) q sin p = P, (25)
as it should be.
b) Let us calculate the Poisson brackets
Q P Q P
{Q, P }q,p =
q p p q
1
= q cos (p) q cos (p) + q sin (p) q 1 sin (p)
h i
= q 21 cos2 (p) + sin2 (p) = q 21 . (26)
For the transformation to be canonical, this Poisson bracket should be identically equal to 1 that requires
= 1/2, = 2, (27)
i.e.,
Q = q 1/2 cos (2p) , P = q 1/2 sin (2p) . (28)
The generating function FpQ (p, Q) should satisfy Eq. (21). To use Eq. (21) to find FpQ (p, Q), one should
first express q and P via the arguments p, Q. From Eq. (28) one obtains
FpQ FpQ
q= , P = . (29)
p Q
Q2
q= , P = Q tan (2p) .
cos2 (2p)
Now integrating the equation
FpQ
P = Q tan (2p) = (30)
Q
on Q one obtains
Q2
FpQ = tan (2p) + f (p). (31)
2
Here the integration Q-constant f (p) can be obtained from another relation
Q2 FpQ Q2 df (p)
q= = = + . (32)
cos2 (2p) p cos2 (2p) dp
This yields
df (p)
=0 f (p) = const, (33)
dp
an irrelevant constant that can be dropped. Thus FpQ is given by Eq. (31) with f (p) = 0.
3 Vortex dynamics
(10 points) Consider the equations of motions describing vortices of strength i with positions ri = (xi , yi )
in the plane
X yi yj X xi xj
x i = j 2
, y i = + j . (34)
j6=i
|ri rj | j6=i
|ri rj |2
Consider the Hamiltonian H and the following Poisson brackets:

n
1X X 1 f g f g
H= ln |ri rj |, {f, g} = . (35)
2 j6=i i j i
i xi yi yi xi
(a) Check that Hamilton equations
x i = {xi , H}, y i = {yi , H} (36)
reproduce the equations of motions. (Check that the standard Hamilton equations can be written in this
form, too)
(b) Show that the following quantities are conserved:
X X
Px = i yi , Py = i xi . (37)
i
(c) Find the Poisson brackets {Px , H}, {Py , H}, and {Px , Py }, as defined above.
(d) Find the solution of the equations of motion for a system of two vortices.
Solution: a) The Poisson brackets yield the Hamiltonian equations

!
X 1 xk H xk H 1 H
x k = {xk , H} = =
j
j xj yj yj xj k yk
!
X 1 yk H yk H 1 H
y k = {yk , H} = = . (38)
j
j xj yj yj xj k xk
Using Eq. (35) one obtains
1 H 1 X i j ln |ri rj | 1 X ij 1 |ri rj |
x k = = = (39)
k yk 2 j6=i k yk 2 j6=i k |ri rj | yk
and
1 H 1 X i j ln |ri rj | 1 X ij 1 |ri rj |
y k = = = (40)
k xk 2 j6=i k xk 2 j6=i k |ri rj | xk
In these expressions
q
|ri rj | (xi xj )2 + (yi yj )2 yi yj (yi yj ) yi yj
= =q = ( ik jk ) (41)
yk yk (xi xj )2 + (yi yj )2 yk |ri rj |
and
|ri rj | xi xj
= ( ik jk ) . (42)
xk |ri rj |
Substituting these results in the equations above one obtains
1 X i j yi yj 1X yk yj 1X yi yk
x k = 2
( ik jk ) = j 2
+ i (43)
2 j6=i k |ri rj | 2 j6=k |rk rj | 2 k6=i |ri rk |2
that is,
X yk yi
x k = i (44)
i6=k
|rk ri |2
and, similarly,
X xk xi
y k = i . (45)
i6=k
|rk ri |2
These equations indeed coincide with those given in the formulation of the problem. Note that i is the
strenght or vorticity of the ith vortex.
b) One has
Px = {Px , H}, Py = {Py , H}, (46)
so that calculating Poisson brackets with H amounts to calculating time derivatives. This can be proven
using Eq. (38). It is more direct to calculate these derivatives using the equations of motion:
X X xk xi
Px = k y k = ik = 0. (47)
k i6=k
|rk ri |2
It is easy to see that the result is zero since the summand is antisymmetric in i and k. Similarly one can
prove that Py = 0, too. Thus Px and Py are integrals of motion.
c) As argued above, {Px , H} = {Py , H} = 0. It remains thus to calculate
X 1 Px Py Px Py
X ij yi xj yi xj

{Px , Py } = = +
k
k xk yk yk xk ijk
k xk yk yk xk
X ij X
= ik jk = k. (48)
ijk
k k
From the Jacoby identity for the Poisson brackets follows that if Px and Py are integrals of motion, then
{Px , Py } is also an integral of motion. In our case, however, this new integral of motion is trivial.
d) For the system of two vortices the equations of motion read
y1 y2 y2 y1
x 1 = 2 , x 2 = 1
|r1 r2 |2 |r2 r1 |2
x1 x2 x2 x1
y 1 = 2 , y 2 = 1 . (49)
|r1 r2 |2 |r2 r1 |2
The integrals of motion of Eq. (37) for two vortices take the form
Px = 1 y1 + 2 y2 , Py = 1 x1 2 x2 . (50)
They can be interpreted as coordinates of the center of mass of the two vortices. The relative motion of the
vortices is described by the variables
X x1 x2 , Y y1 y2 . (51)
For X and Y one obtains the equations of motion

Y X
X = ( 1 + 2 ) 2 , Y = ( 1 + 2 ) 2 (52)
R R
with
R2 = X 2 + Y 2 . (53)
Note that the distance between the vortices R is a constant of motion:
dR2 XY YX
= 2X X + 2Y Y = 2 ( 1 + 2 ) 2 + 2 ( 1 + 2 ) 2 = 0. (54)
dt R R
Thus Eq. (52) can be written in the form
X = 0 Y, Y = 0 X, (55)
where
1 + 2
0 = (56)
R2
is an angular velocity. This is the angular velocity with which the two vortices are rotating around each
other. Indeed, the solution of Eq. (55) can be represented in the form
X = Re Z, Y = Im Z, (57)
where
Z = X + iY (58)
satisfies the equation
Z = i 0 Z (59)
and has the form
Z = Rei0 t+0 . (60)
It would be interesting to investigate the behavior of systems of more than two vortices.

File

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

File

Uploaded by

Copyright:

Available Formats

PHY 5246: Theoretical Dynamics, Fall 2015

September 2nd , 2015

2 Non-graded suggested problems

z x = R sin cos(t) , (1)

where we have used the constraints: r = R and = t

and the Lagrangian of the system is

sin(0 + ) = sin 0 cos + cos 0 sin = sin 0 (1 + O( 2 )) + cos 0 ( + O( 3 ) , (13)

Problem 3 (1.14 of Goldsteins book)

Problem 4 (1.21 of Goldsteins book)

Figure 1: Figures for Problem 3

For the equation we note that since the Lagrangian

U(r, v) = V (r) + (r mv)

We find the following:

and the result is the same as in the first case.

Problem 7 (1.22 of Goldsteins book)

(m1 + m2 )l1 1 = (m1 + m2 )g sin 1 m2 l2 cos(1 2 )2 22 m2 l2 sin(1 2 ).

m2 l2 2 = m2 g sin 2 m2 l1 1 cos(1 2 ) + 12 m2 l1 sin(1 2 ).

September 9th , 2015

(1.a) Find the equations of motion.

3. A particle is subject to a potential V (x) = F x where F is a constant. The particle travels

2 Non-graded suggested problems

6. Goldsteins book, Chapter 2, Problem 6.

since l is constant (see Eq. (4)). The equilibrium condition is that

which, using Eq. (6) gives

yw = constant and (10)

nates. We pick xw for the wedge and xb for the

block, where xb is neither xb nor yb but is the

xb = xw + xb cos + constant , (11)

V = mgyb = mgxb sin + constant , (13)

we can recast Eq. (20) in the following form,

(b) The force does not depend on x, so px is conserved.

(e) The force does not depend on z so pz is conserved.

Problem 6 (Goldstein 2.6)

we consider the Euler-Lagrange equations:

where C = C1 + C2 . The solution to this equation is a hypocycloid with parametric equations

and substituting into eq. (29). We find

September 16th , 2015

4. Problem 2.21 of Goldsteins book.

2 Non-graded suggested problems

6. Problem 2.4 of Goldsteins book.

and the Lagrangian is:

l2 l2 2 sin cos + gl sin = 0 . (5)

Expanding in and keeping only up to the linear terms in , we find

r = R + a dr = 0 a1,r = 1, a1, = a1, = 0

m(R + a) 2 + mg cos = 1 , (27)

m(R + a)2 mg(R + a) sin = (R + a)2 , (28)

(28): m(R + a)2 mg(R + a) sin = m(R + a)2 2(R + a) g sin = 0

which gives us the equation of motion

1 = mg(2 cos 1),

Problem 4 (Goldstein 2.21)

Let us define the integrand as

The variational principle tells us that the minimal dis-

September 23rd , 2015

2 Non-graded suggested problems

where C = 2C is determined by the initial conditions on 0 .

The four given cases are:

Problem 3 (Goldstein 3.11)

The equation of motion for r is

and insert this into the equation for the orbit

Figure 1: A graph of the effective Yukawa

September 30th , 2015

4. A particle is moving in a potential