# Phys 7221 Homework #3

Gabriela Gonz´alez
September 27, 2006
1. Derivation 2-4: Geodesics on a spherical surface
Points on a sphere of radius R are determined by two angular coordinates, an az-
imuthal angle ψ and a polar angle θ:
r = x
ˆ
i +y
ˆ
j +z
ˆ
k = R(sin ψ cos θ
ˆ
i + sin ψ sin θ
ˆ
j + cos ψ
ˆ
k)
When moving on the sphere, the diﬀerential arc length ds is
ds
2
= dx
2
+dy
2
+dz
2
= R
2
((cos ψ cos θdψ −sin ψ sin θdθ)
2
+ (cos ψ sin θdψ + sin ψ cos θdθ)
2
+ (−sin ψdψ)
2
)
= R
2
(dψ
2
+ sin
2
ψdθ
2
)
The distance on the sphere between two points is then
l =
_
ds = R
_ _

2
+ sin
2
ψdθ
2
= R
_

¸
_

_
2
+ sin
2
ψ = R
_
dθf(ψ, ψ

)
We can use a variational principle for ﬁnding the path with minimum length between
two point. The path is described by a function ψ(θ), and the (diﬀerential) equation for
ψ can be obtained from the Euler-Lagrange equation using f(ψ, ψ

) =
_
sin
2
ψ +ψ
2
.
Back to the variational principle: the equation for ψ is
0 =
d

∂f
∂ψ

∂f
∂ψ

=
d

_
ψ

f
_

sin ψ cos ψ
f
=
ψ

f

ψ

f

f
2

sin ψ cos ψ
f
=
ψ

f

ψ

f
2
ψ

sin ψ cos ψ +ψ

ψ

f

sin ψ cos ψ
f
1
0 = (ψ

−sin ψ cos ψ)f
2
−ψ
2

+ sin ψ cos ψ)
0 = (ψ

−sin ψ cos ψ)(ψ
2
+ sin
2
ψ) −ψ
2

+ sin ψ cos ψ)
= ψ

sin
2
ψ −2ψ
2
sin ψ cos ψ −sin
3
ψ cos ψ
This looks like a complicated equation to solve! It’s always useful if we know the
solution before we obtain it, admittedly not the most common case, but true in this
case. We know that the shortest path between points in the sphere are great circles.
Great circles are the intersection between the sphere and a plane. If the unit vector
normal to the plane as ˆ n = a
ˆ
i +b
ˆ
j +c
ˆ
k, the points in the great circle are those points
in the sphere that satisfy ˆ n · r = 0 = R(sin φ(a cos θ + b sin θ) + c cos ψ), or those
points with coordinates ψ, θ satisfying
cos ψ
sin ψ
= Acos θ +Bsin θ
with A
2
+B
2
< 1. If we deﬁne a function q(θ) = cos ψ(θ)/ sin(ψ(θ)), we are looking
for an equation of the form d
2
q/d
2
θ = −q. If q = 1/ tan ψ, then q

= −ψ

/ sin
2
ψ,
and q

= −ψ

/ sin
2
ψ + 2ψ
2
cos ψ/ sin
3
ψ. Lagrange’s equation can then be written
as
0 = −q

sin
4
ψ −sin
3
ψ cos ψ
q

= −cos ψ/ sin ψ = −q
which is the equation we were looking for, with a general solution
q =
cos ψ
sin ψ
= Acos θ +Bsin θ
which we know describes points on a great circle.
2. Exercise 2-14: A hoop rolling on a cylinder
We can ﬁnd out the angle at which the hoop falls from the cylinder by obtaining an
expression for the normal force on the hoop as a function of the position of the hoop:
the hoop will fall oﬀ when the normal force vanishes.
We set up a coordinate system with the origin at the center of the cylinder, and
describe the center of mass of the hoop with polar coordinates r, θ, and an angular
coordinate φ for the rotation about the hoop’s axis, as shown in the ﬁgure.
The kinetic energy is
T =
1
2
mv
2
+
1
2

2
=
1
2
m( ˙ r
2
+r
2
˙
θ
2
) +
1
2
ma
2
˙
φ
2
2
and the potential energy is
V = −mg · r = mgr sin θ
The Lagrangian is
L = T −V =
1
2
m( ˙ r
2
+r
2
˙
θ
2
) +
1
2
ma
2
˙
φ
2
−mgr sin θ
There are two constraints while the hoop is rolling on the cylinder :
f
1
= r −(R +a) = 0 (1)
f
2
= (R +a)
˙
θ +a
˙
φ = 0 (2)
Note that if the hoop is rolling down,
˙
θ < 0 and
˙
φ > 0, if the angles are deﬁned like in
the ﬁgure. The rolling constraint is formulated setting up the velocity of the contact
point instantaneously equal to zero, and expressing it as the velocity of the center
of mass r
˙
θ, plus the velocity with respect to the center of mass, a
˙
φ. The equivalent
condition for the hoop rolling on a plane is ˙ x +a
˙
φ = 0.
The ﬁrst constraint f
1
is holonomic, and we’ll associate with it a Lagrange multiplier
λ (which will be related to the normal force of the cylinder on the hoop). The second
constraint is semi-holonomic, i.e., it depends on velocities but it could be integrated
into a holonomic constraint. We will associate a Lagrange multiplier µ with it, which
will be related to the friction force producing the rolling.
There are three Lagrange’s equations for the coordinates r, θ, φ:
d
dt
∂L
∂ ˙ q
i

∂L
∂q
i
= λ
∂f
1
∂q
j

∂f
2
∂ ˙ q
j
m¨ r −mr
˙
θ
2
+mg sin θ = λ (3)
2mr ˙ r
˙
θ +mr
2
¨
θ +mgr cos θ = µ(R +a) (4)
ma
2
¨
φ = µa (5)
3
We have then 5 equations (1)...(5) for ﬁve unknowns, r, θ, φ, λ, µ.
We use the ﬁrst constraint to solve for the coordinate r: r = R + a, ˙ r = ¨ r = 0. We
use this solution in Lagrange’s equations for r, θ:
−m(R +a)
˙
θ
2
+mg sin θ = λ (6)
m(R +a)
2
¨
θ +mg(R +a) cos θ = µ(R +a) (7)
We use the rolling constraint to ﬁnd an expression for φ as a function of θ:
φ = −
a +R
a
θ +φ
0
(8)
and use this in Lagrange’s equation (5) for φ to obtain
µ = ma
¨
φ = −m(R +a)
¨
θ (9)
We use this expression for µ in (7), and obtain an equation for
¨
θ:
¨
θ = −
g
2(R +a)
cos θ (10)
We can integrate this equation by multiplying by
˙
θ:
¨
θ +
g
2(R +a)
cos θ = 0
¨
θ
˙
θ +
g
2(R +a)
˙
θ cos θ = 0
d
dt
_
1
2
˙
θ
2
+
g
2(R +a)
sin θ
_
= 0
1
2
˙
θ
2
+
g
2(R +a)
sin θ = C
If the hoop starts from rest at the top, then
˙
θ = 0 when θ = π/2, which tells us the
value of the constant of integration C:
˙
θ
2
=
g
R +a
(1 −sin θ) (11)
We now use this in Eq.(6), to get an expression for the normal force as a function of
the angle θ:
λ = −m(R +a)
˙
θ
2
+mg sin θ
= −mg(1 −sin θ) +mg sin θ
λ = mg(2 sin θ −1) (12)
4
At the top, when θ = π/2, we obtain λ = mg, as expected for the normal force. If
we try to apply this equation at the bottom, when θ = 0, we obtain a negative value
for λ, which tells us that the formulation of the problem cannot apply at that point,
since the normal force cannot be negative. Equation (12) tells us that if sin θ < 1/2,
the multiplier λ becomes negative: this is the angle at which the hoop falls from the
cylinder, θ = 30

.
3. Exercise 2-18: A bead on a rotating hoop
A bead with mass m can slide without friction on a vertical hoop of radius a. The
hoop is rotating along a vertical diameter with constant angular velocity ω.
Take the origin of a coordinate system at the center of the hoop, with the z-axis
pointing down, along the rotation axis. If we use spherical coordinates r, ψ, θ to
describe the position the mass, we know that r = a and
˙
θ = ω, so the only generalized
coordinate needed to describe the mass’ postion is ψ. The kinetic energy is
T =
1
2
mv
2
=
1
2
m
_
˙ r
2
+r
2
˙
ψ
2
+r
2
sin
2
ψ
˙
θ
2
_
=
1
2
ma
2
_
˙
ψ
2

2
sin
2
ψ
_
The potential energy due to the gravitational acceleration g = g
ˆ
k (since the z-
axispoints down) is
V = −mg · r
= −mgz
= −mga cos ψ
The Lagrangian is
L = T −V =
1
2
ma
2
˙
ψ
2
+
1
2
ma
2
ω
2
sin
2
ψ +mga cos ψ
Lagrange’s equation of motion is
0 =
d
dt
∂L

˙
ψ

∂L
∂ψ
= ma
2
¨
ψ −ma
2
ω
2
sin ψ cos ψ +mga sin ψ
a
¨
ψ = −g sin ψ
_
1 −

2
g
cos ψ
_
(13)
The Lagrangian has ∂L/∂
˙
ψ = 0, so the canonical momentum conjugate to ψ, pro-
portional to the angular momentum component L
z
, is not conserved.
5
However, the Lagrangian does not depend explicitly on time, so there is an integral
of motion:
h
ψ
=
∂L

˙
ψ
˙
ψ −L
=
1
2
ma
2
˙
ψ
2

1
2
ma
2
ω
2
sin
2
ψ −mga cos ψ
˙
ψ
2
=
2h
ψ
ma
2

2
sin
2
ψ + 2
g
a
cos ψ
The integral of motion is not the total energy E = T + V , but it is related to the
energy by h
ψ
= E −ma
2
ω
2
sin
2
ψ.
For the mass to remain stationary on the hoop, we need ψ = ψ
0
and
˙
ψ =
¨
ψ = 0.
From Eq. ??, we see that is possible only if sin ψ
0
= 0 (top or bottom of the hoop),
or cos ψ
0
= g/aω
2
, which is possible only if the hoop’s velocity is high enough so that
g/aω
2
< 1.
If the mass starts near the bottom, where ψ 1, we can use a small angle approxi-
mation in the equation of motion, and
a
¨
ψ ≈ −gψ(1 −aω
2
/g).
If the angular velocity is not larger than ω
2
0
= g/a, this equation describes a harmonic
oscillator with frequency Ω
2
= (g/a)(1 − aω
2
/g) = g

/a. The mass oscillates as the
pendulum bob of a pendulum with length a, in a gravitational acceleration reduced
by the rotation, g

= g(1 −aω
2
/g)
1
.
In general, if aω
2
/g < 1, the acceleration given by Eq. ?? will be always negative
(since sin ψ > 0), and will drive the mass to the bottom, making it oscillate about
the lowest point (unless it starts there with zero velocity, when it will stay at the
bottom).
If the angular velocity is larger than ω
2
0
= g/a, there is angle α given by cos α =
g/aω
2
, where the acceleration given by Eq.?? is zero. If the mass starts near such a
point, we can deﬁne a small angle ψ

= ψ −α 1, so that
¨
ψ

=
¨
ψ, and use Eq.??:
¨
ψ

= −
g
a
sin ψ
_
1 −
ω
2
a
g
cos ψ
_
= −
g
a
sin(ψ

+α)
_
1 −
ω
2
a
g
cos(ψ

+α)
_
1
Strictly speaking, the angular coordinate ψ can only be positive, so it cannot oscillate about zero, which
is a singular point for the spherical coordinate system. However, we could deﬁne another coordinate system
where the point at the bottom of the hopp is not singular, and we would obtain the same SHO equation of
motion.
6
= −
g
a
_
sin ψ

cos α + cos ψ

sin α
_
_
1 −
ω
2
a
g
_
cos ψ

cos α −sin ψ

sin α
_
_
≈ −
g
a
sin α
_
1 −cos ψ

+
ω
2
a
g
sin αsin ψ

_
≈ −(ω
2
sin
2
α)ψ

The equation for ψ

is again that of a simple harmonic oscillator, with frequency

2
= ω
2
sin
2
α: a smaller frequency than the hoop’s angular frequency. For very high
rotation frequencies ω
2
a/g, we have cos α ∼ 0 (α ∼ π/2), sin α ∼ 1, and Ω ∼ ω:
the mass stays in the center of the hoop, in a constant position relative to the hoop’s
coordinate system.
4. Exercise 2-19: Symmetries and conserved quantities
We consider the gravitational forces created on particles by diﬀerent mass distri-
butions. If the mass distribution has a particular symmetry, so will the potential
associated with the force, and so will the Lagrangian. Since symmetries are associ-
ated with conserved quantities through Noether’s theorem, we can ﬁnd the conserved
quantities: translational symmetries are associated with components of the linear
momentum; rotational symmetries with components of the angular momentum, and
time independence with conservation of energy. Since the potential is ﬁxed in all
cases (i.e., independent of time), the energy is conserved in all systems.
(a) The mass is uniformly distributed in the plane z = 0 (an inﬁnite, ﬂat, Earth):
the forces do not depend on the coordinates x, y, and thus the components of
the linear momentum p
x
, p
y
will be conserved. Also, the force is invariant under
a rotation about the z axis, so L
z
is conserved.
(b) The mass is uniformly distributed in the half plane z = 0, y > 0 (a ﬁnite, ﬂat,
Earth, like Columbus feared): there’s only translational symmetry with respect
to x, and no rotational symmetries: only p
x
will be conserved.
(c) The mass is uniformly distributed in a circular cylinder of inﬁnite length, with
axis along the z-axis: the conﬁguration has translational symmetry along z, and
rotational symmetry about z, so p
z
and L
z
are conserved.
(d) The mass is uniformly distributed in a circular cylinder of ﬁnite length, with
axis along the z-axis: there is now no translational symmetry, but there is still
rotational symmetry about z: only L
z
is conserved.
(e) The mass is uniformly distributed in a right cylinder of elliptical cross section
and inﬁnite length, wit axis along the z axis: there is now no rotational symme-
try, but because the cylinder is inﬁnite along z, there is translational symmetry
along z: only p
z
is conserved.
7
(f) The mass is uniformly distributed in a dumbbell whose axis is oriented along
the z axis: no translational symmetries, but there is rotational symmetry about
the z axis, so L
z
is conserved.
(g) The mass is the form of a uniform wire wound in the geometry of an inﬁnite
helical solenoid, with axis along the z axis. There are no pure translational
or rotational symmetries, but there is a symmetry combining a z-translation of
distance h (the distance between coils), and a rotation about z of 2π. Thus,
although p
z
or L
z
are not individually conserved, hp
z
+L
z
will be conserved.
5. Exercise 2-20: A particle on a sliding wedge
A mass m is sliding down without friction along a wedge with angle α and mass M.
The wedge can move without friction on a smooth horizontal surface.
There are two objects in the problem, the mass m and the wedge M. The edge is a
rigid body, but since it cannot rotate, it is described by the coordinates of just one
point, say the top corner. We can treat the problem as a two dimensional problem
so we have two coordinates for each mass. Let us choose a coordinate system with
the origin at the initial position of the top of the wedge, with a horizontal x-axis and
a vertical y-axis pointing down, as shown in the ﬁgure.
Figure 1: Problem 2-20: a mass sliding down a sliding wedge.
The coordinates of the top corner of the wedge will be R = (X, 0). The coordinates
of the mass sliding down the wedge are r = (x, y). The constraint that the mass is
on the wedge is
r = R+l(cos α, sin α) , or
x = X +l cos α and
y = l sin α
8
where l is the distance the mss traveled down the wedge. This is one constraint,
which we can express as a function of x, y, X as
f = (x −X) sin α −y cos α = 0.
The kinetic energy of the system is
T =
1
2
m( ˙ x
2
+ ˙ y
2
) +
1
2
M
˙
X
2
.
The potential energy is just gravitational, with g = g
ˆ
j. The gravitational potential
energy of the wedge is constant, so we can ignore it. The gravitational potential
energy of the mass m is
V = −mg · r = −mgy
The Lagrangian is
L = T −V =
1
2
m( ˙ x
2
+ ˙ y
2
) +
1
2
M
˙
X
2
+mgy (14)
There are three Lagrange equations, for q
i
= X, x, y, which together with the con-
straint, form a system of four equations for the four variables X, x, y, λ (where λ is
the Lagrange multiplier associated with the constraint).
d
dt
∂L
∂ ˙ q
i

∂L
∂q
i
= λ
∂f
∂q
i
m¨ x = λsin α
m¨ y −mg = −λcos α
M
¨
X = −λsin α
The equations for x, X can be added and result in
m¨ x +M
¨
X = 0
This equation is saying that the horizontal position of the center of mass of the system
has a constant velocity: we know this, since the only external force is gravity, and it
is vertical (Notice that the ﬁgure does not represent the actual motion then!).
We can assume the initial position of the mass is at the top of the wedge, and the
initial velocity of the center of mass is zero, and then we have a solution for X in
terms of x:
X = −mx/M.
9
We can use the constraint to solve x in terms of y (or viceversa):
y = (x −X) tan α = x(1 +m/M) tan α.
We can also use Lagrange’s equation for y to solve for λ, and use these results in
Lagrange’s equation for x:
m¨ x = λsin α = m(g − ¨ y) tan α
= mtan α(g − ¨ x(1 +m/M) tan α)
(1 + (1 +m/M) tan
2
α)¨ x = g tan α
1 + (m/M) sin
2
α
cos
2
α
¨ x = g tan α
¨ x = g
sin αcos α
1 + (m/M) sin
2
α
¨ x = a
x
The acceleration of x is constant, and the general solution is x = x
0
+v
0x
t+(1/2)a
x
t
2
.
Since we assumed the mass started at the origin, x
0
=0. If the mass starts from rest,
v
0x
= 0. We can now use this result to obtain the equation for y:
¨ y = ¨ x(1 +m/M) tan α
= g
sin αcos α
1 + (m/M) sin
2
α
(1 +m/M) tan α
= g sin
2
α
1 +m/M
1 + (m/M) sin
2
α
= a
y
The acceleration of the mass along the direction tangent to the wedge is
a
s
= a
x
cos α +a
y
sin α
= g
sin αcos α
1 + (m/M) sin
2
α
cos α +g sin
2
α
1 +m/M
1 + (m/M) sin
2
α
sin α
= g sin α
The solutions for X, λ are:
¨
X = −m¨ x/M
= −g
m
M
sin αcos α
1 + (m/M) sin
2
α
= −a
M
10
λ =
m
sin α
¨ x
= mg
cos α
1 + (m/M) sin
2
α
The multiplier λ is the normal force from the wedge on the mass: it is mg if the
wedge is horizontal (α = 0), and it is smaller as the wedge is steeper α →π/2.
If the wedge has a large mass (M m), it does not move much (
¨
X ≈ 0), and the
system approximates that of a mass sliding down a ﬁxed incline. The tangential
acceleration approximates a
s
≈ g sin α, and the normal force approximates λ ≈
mg cos α, as it should be well known.
The constraint force, the normal force, is always perpendicular to the wedge: N =
λ(sin α, −cos α). If the wedge is ﬁxed, this is perpendicular to the mass’ motion,
but if the wedge is not ﬁxed, it will have a component along the mass’ velocity: the
constraint force works on the particle. The work done in time t by the force N on
the mass m is given by:
dW
m
dt
= N· v
= λ( ˙ xsin α − ˙ y cos α)
= λt(a
x
sin α −a
y
cos α)
= λgt sin
2
αcos α
m/M
1 + (m/M) sin
2
α
The work done by the normal force tends to zero as m/M as the wedge mass M gets
large. The constraint force also works on the wedge:
dW
M
dt
= N· V
= λ
˙
X sin α
= −λ(m˙ x/M) sin α
= −λ(m/M)a
x
t sin α
= −λgt sin
2
αcos α
m/M
1 + (m/M) sin
2
α
That is, the net work done by the constraint on the system is zero.
To ﬁnd out conserved quantities, we need to express the Lagrangian ?? in terms of
independent coordinates. If we use the constraint to solve for X, we get
X = x −
y
tan α
11
L =
1
2
m( ˙ x
2
+ ˙ y
2
) +
1
2
M
˙
X
2
+mgy
=
1
2
m( ˙ x
2
+ ˙ y
2
) +
1
2
M
_
˙ x −
˙ y
tan α
_
2
+mgy
The coordinate x is cyclical, so its canonical momentum is a constant of motion:
p
x
=
∂L
∂ ˙ x
= (m+M) ˙ x −M
˙ y
tan α
This can also be obtained from the solutions, since ˙ x = a
x
t, ˙ y = a
y
t and a
y
=
a
x
(1 +m/M)/ tan α.
The Lagrangian does not depend explicitly on time, so there is a Jacobi integral,
which is equal to the total energy E = T +V .
6. A carriage on rotating cross-rails
The position vector of the mass m in the inertial frame is r = x
ˆ
i+y
ˆ
j. The coordinates
x, y are related to the spring lengths R, r, where the length R of the spring with force
constant K and rest length R
0
, and the length r of the spring on the perpendicular
rail with force constant k (and zero rest length). We can express x, y in terms of R, r:
x = Rcos ωt −r sin ωt
y = Rsin ωt +r cos ωt
or R, r in terms of x, y:
R = xcos ωt +y sin ωt
r = −xsin ωt +y cos ωt
We can choose as generalized coordinates x, y (natural coordinates in the inertial lab
frame) or R, r (natural coordinates in the rotating frame).
The kinetic energy is
T =
1
2
( ˙ x
2
+ ˙ y
2
)
=
1
2
m((
˙
R −rω)
2
+ ( ˙ r +ωR)
2
)
=
1
2
m(
˙
R
2
+ ˙ r
2
+ 2ω(R˙ r −
˙
Rr) +ω
2
(r
2
+R
2
))
12
The potential energy is
V =
1
2
K(R −R
0
)
2
+
1
2
kr
2
=
1
2
K(xcos ωt +y sin ωt −R
0
)
2
+
1
2
k(−xsin ωt +y cos ωt)
2
In terms of x, y, the Lagrangian is
L = T −V
=
1
2
m( ˙ x
2
+ ˙ y
2
) −
1
2
K(xcos ωt +y sin ωt −R
0
)
2

1
2
k(−xsin ωt +y cos ωt)
2
Since the Lagrangian depends explicitly on time, we know the energy function is not
conserved. Because the potential does not depend on time derivatives ˙ x, ˙ y, and the
kinetic energy is homogeneous of second degree in ˙ x, ˙ y, we know that the energy
function is the mechanical energy, and is not conserved:
h =

˙ q
i
∂L
∂ ˙ q
i
−L
= ˙ x
∂L
∂ ˙ x
+ ˙ y
∂L
∂ ˙ y
−L
=
1
2
m( ˙ x
2
+ ˙ y
2
) +
1
2
K(xcos ωt +y sin ωt −R
0
)
2
+
1
2
k(−xsin ωt +y cos ωt)
2
= T +V = E
If we now instead choose R, r as our generalized coordinates, the Lagrangian is
L = T −V
=
1
2
m(
˙
R
2
+ ˙ r
2
+ 2ω(R˙ r −
˙
Rr) +ω
2
(r
2
+R
2
)) −
1
2
K(R −R
0
)
2

1
2
kr
2
=
1
2
m(
˙
R
2
+ ˙ r
2
) +mω(R˙ r −
˙
Rr) +
1
2

2
(r
2
+R
2
) −
1
2
K(R −R
0
)
2

1
2
kr
2
and is independent of time, so the “energy function” or “Jacobi integral” will be
conserved. However, since the kinetic energy is not homogenous of second degree in
˙
R, ˙ r, then the energy function is not equal to the mechanical energy.
The Jacobi integral is
h

=

˙ q
i
∂L
∂ ˙ q
i
−L
13
= ˙ r
∂T
∂ ˙ r
+
˙
R
∂T

˙
R
−(T −V )
= m˙ r( ˙ r +ωR) +m
˙
R(
˙
R −ωr) −T +V
= m˙ r
2
+m
˙
R
2
+mω(R˙ r −r
˙
R) −T +V
=
1
2
m(
˙
R
2
+ ˙ r
2
) −
1
2

2
(r
2
+R
2
) +
1
2
K(R −R
0
)
2
+
1
2
kr
2
= T +V −mω(R˙ r −
˙
Rr) −
1
2

2
(r
2
+R
2
)
We see that if the beams are not rotating, ω = 0 and h = T + V = E. With the
rotation on, the mechanical energy is not conserved, and the rate of change of the
energy is
dE
dt
=
d
dt
_
h +mω(R˙ r −
˙
Rr) +
1
2

2
(r
2
+R
2
)
_
= mω(R¨ r −r
¨
R) +mω
2
(r ˙ r +R
˙
R) (15)
(The following was not asked in the homework, but it has more interesting facts
We can ﬁnd expressions for the rate of change in energy from Lagrange’s equations
of motion for R, r, which we can ﬁnd from the Lagrangian:
L =
1
2
m(
˙
R
2
+ ˙ r
2
) +mω(R˙ r −
˙
Rr) +
1
2

2
(r
2
+R
2
) −
1
2
K(R −R
0
)
2

1
2
kr
2
=
1
2
m(
˙
R
2
+ ˙ r
2
) −U −V
We have written the Lagrangian as the kinetic energy in the (non-inertial) rotating
system, minus the potential energy in the inertial system, minus an extra potential
U due to the rotation of the frame:
U(r, R, ˙ r,
˙
R) = −mω(R˙ r −
˙
Rr) −
1
2

2
(r
2
+R
2
).
The forces derived from this potential are the “centrifugal forces”, and they are not
conservative (they depend on velocities).
Lagrange’s equations are
0 =
d
dt
∂L
∂ ˙ r

∂L
∂r
m¨ r =
d
dt
∂U
∂ ˙ r

∂(U +V )
∂r
m¨ r = −mω
˙
R +mω
2
r −kr
14
0 =
d
dt
∂L

˙
R

∂L
∂R
m
¨
R = −
d
dt
∂U

˙
R

∂(U +V )
∂r
m
¨
R = mω ˙ r +mω
2
R −K(R −R
0
)
From these equations, we can ﬁnd an expression for the combinations we need for
power in Eq(??):
dE
dt
= mω(R¨ r −r
¨
R) +mω
2
(R
˙
R +r ˙ r) = ωr(−kR +K(R −R
0
))
The equilibrium coordinates are the solutions to ˙ r = ¨ r =
˙
R =
¨
R = 0, which are r = 0
and R = R
0
/(1 −mω
2
/K). The equilibrium position for the K spring is not R
0
, but
a smaller length. If the system rotates too fast, and ω
2
≥ K/m, the cross rail will be
pushed against the rotation axis. Assuming small oscillations about the equilibrium
length, and deﬁning ξ = R −R
0
/(1 −mω
2
/K), the equations of motion are
m¨ r = −mω
˙
ξ −(k −mω
2
)r
m
¨
ξ = mω ˙ r +mω
2
_
ξ +
R
0
1 −mω
2
/K
_
−K
_
ξ +
R
0
1 −mω
2
/K
−R
0
_
= mω ˙ r −(K −mω
2

We see that the springs are “softened” by the rotation, and the oscillations in the
diﬀerent directions are coupled. We will learn how to ﬁnd solutions to the equations
of motion of these systems in Chapter 6.
15

θ. Great circles are the intersection between the sphere and a plane. The kinetic energy is 1 1 1 1 ˙ ˙ ˙ T = mv 2 + Iω 2 = m(r2 + r2 θ2 ) + ma2 φ2 2 2 2 2 2 . 2. and describe the center of mass of the hoop with polar coordinates r. then q = −ψ / sin2 ψ. If q = 1/ tan ψ. with a general solution q= cos ψ = A cos θ + B sin θ sin ψ which we know describes points on a great circle. Exercise 2-14: A hoop rolling on a cylinder We can ﬁnd out the angle at which the hoop falls from the cylinder by obtaining an expression for the normal force on the hoop as a function of the position of the hoop: the hoop will fall oﬀ when the normal force vanishes. as shown in the ﬁgure. We know that the shortest path between points in the sphere are great circles. admittedly not the most common case.0 = (ψ − sin ψ cos ψ)f 2 − ψ 2 (ψ + sin ψ cos ψ) 0 = (ψ − sin ψ cos ψ)(ψ 2 + sin2 ψ) − ψ 2 (ψ + sin ψ cos ψ) = ψ sin2 ψ − 2ψ 2 sin ψ cos ψ − sin3 ψ cos ψ This looks like a complicated equation to solve! It’s always useful if we know the solution before we obtain it. and q = −ψ / sin2 ψ + 2ψ 2 cos ψ/ sin3 ψ. but true in this case. or those points with coordinates ψ. we are looking for an equation of the form d2 q/d2 θ = −q. If we deﬁne a function q(θ) = cos ψ(θ)/ sin(ψ(θ)). Lagrange’s equation can then be written as 0 = −q sin4 ψ − sin3 ψ cos ψ q = − cos ψ/ sin ψ = −q which is the equation we were looking for. and an angular coordinate φ for the rotation about the hoop’s axis. We set up a coordinate system with the origin at the center of the cylinder. θ satisfying cos ψ = A cos θ + B sin θ sin ψ with A2 + B 2 < 1. If the unit vector ˆ ˆ normal to the plane as n = aˆ + bˆ + ck. the points in the great circle are those points i j ˆ in the sphere that satisfy n · r = 0 = R(sin φ(a cos θ + b sin θ) + c cos ψ).

. φ: d ∂L ∂L − dt ∂ qi ∂qi ˙ = λ ∂f1 ∂f2 +µ ∂qj ∂ qj ˙ ˙ m¨ − mrθ2 + mg sin θ = λ r ¨ 2mrrθ + mr2 θ + mgr cos θ = µ(R + a) ˙˙ ¨ ma2 φ = µa 3 (3) (4) (5) . it depends on velocities but it could be integrated into a holonomic constraint. aφ. which will be related to the friction force producing the rolling. and expressing it as the velocity of the center ˙ ˙ of mass rθ. The rolling constraint is formulated setting up the velocity of the contact point instantaneously equal to zero. plus the velocity with respect to the center of mass. θ.e. The equivalent ˙ condition for the hoop rolling on a plane is x + aφ = 0. if the angles are deﬁned like in the ﬁgure. The second constraint is semi-holonomic. We will associate a Lagrange multiplier µ with it. and we’ll associate with it a Lagrange multiplier λ (which will be related to the normal force of the cylinder on the hoop). θ < 0 and φ > 0. There are three Lagrange’s equations for the coordinates r.and the potential energy is V = −mg · r = mgr sin θ The Lagrangian is 1 1 ˙ ˙ L = T − V = m(r2 + r2 θ2 ) + ma2 φ2 − mgr sin θ ˙ 2 2 There are two constraints while the hoop is rolling on the cylinder : f1 = r − (R + a) = 0 ˙ ˙ f2 = (R + a)θ + aφ = 0 (1) (2) ˙ ˙ Note that if the hoop is rolling down. ˙ The ﬁrst constraint f1 is holonomic. i.

We ˙ ¨ use this solution in Lagrange’s equations for r. r. µ. λ. We use the ﬁrst constraint to solve for the coordinate r: r = R + a. θ. r = r = 0. then θ = 0 when θ = π/2. to get an expression for the normal force as a function of the angle θ: ˙ λ = −m(R + a)θ2 + mg sin θ = −mg(1 − sin θ) + mg sin θ λ = mg(2 sin θ − 1) 4 (12) . φ. θ: ˙ −m(R + a)θ2 + mg sin θ = λ ¨ m(R + a)2 θ + mg(R + a) cos θ = µ(R + a) We use the rolling constraint to ﬁnd an expression for φ as a function of θ: φ=− a+R θ + φ0 a (8) (6) (7) and use this in Lagrange’s equation (5) for φ to obtain ¨ ¨ µ = maφ = −m(R + a)θ ¨ We use this expression for µ in (7). which tells us the value of the constant of integration C: ˙ θ2 = g (1 − sin θ) R+a (11) We now use this in Eq..(6).We have then 5 equations (1). and obtain an equation for θ: ¨ θ=− g cos θ 2(R + a) (10) (9) ˙ We can integrate this equation by multiplying by θ: g cos θ 2(R + a) g ¨˙ ˙ θθ + θ cos θ 2(R + a) 1 ˙2 g θ + sin θ 2 2(R + a) 1 ˙2 g θ + sin θ 2 2(R + a) ¨ θ+ = 0 = 0 = 0 = C d dt ˙ If the hoop starts from rest at the top.(5) for ﬁve unknowns..

when θ = π/2. is not conserved. along the rotation axis. the multiplier λ becomes negative: this is the angle at which the hoop falls from the cylinder. we obtain a negative value for λ. since the normal force cannot be negative. If we try to apply this equation at the bottom. 3.At the top. Exercise 2-18: A bead on a rotating hoop A bead with mass m can slide without friction on a vertical hoop of radius a. The hoop is rotating along a vertical diameter with constant angular velocity ω. with the z-axis pointing down. 5 . ψ. The kinetic energy is T = = = 1 mv 2 2 1 ˙ ˙ m r2 + r2 ψ 2 + r2 sin2 ψ θ2 ˙ 2 1 ˙ ma2 ψ 2 + ω 2 sin2 ψ 2 ˆ The potential energy due to the gravitational acceleration g = g k (since the zaxispoints down) is V = −mg · r = −mgz = −mga cos ψ The Lagrangian is 1 1 ˙ L = T − V = ma2 ψ 2 + ma2 ω 2 sin2 ψ + mga cos ψ 2 2 Lagrange’s equation of motion is d ∂L ∂L − ˙ dt ∂ ψ ∂ψ ¨ = ma2 ψ − ma2 ω 2 sin ψ cos ψ + mga sin ψ aω 2 ¨ aψ = −g sin ψ 1 − cos ψ g 0 = (13) ˙ The Lagrangian has ∂L/∂ ψ = 0. we obtain λ = mg. θ to ˙ describe the position the mass. so the only generalized coordinate needed to describe the mass’ postion is ψ. Equation (12) tells us that if sin θ < 1/2. so the canonical momentum conjugate to ψ. as expected for the normal force. which tells us that the formulation of the problem cannot apply at that point. when θ = 0. If we use spherical coordinates r. we know that r = a and θ = ω. proportional to the angular momentum component Lz . Take the origin of a coordinate system at the center of the hoop. θ = 30◦ .

??. 6 .?? is zero. ?? will be always negative (since sin ψ > 0). the acceleration given by Eq. ˙ ¨ For the mass to remain stationary on the hoop. which is possible only if the hoop’s velocity is high enough so that g/aω 2 < 1. and will drive the mass to the bottom. and we would obtain the same SHO equation of motion. this equation describes a harmonic oscillator with frequency Ω2 = (g/a)(1 − aω 2 /g) = g /a. the Lagrangian does not depend explicitly on time.??: ¨ ψ g ω2a = − sin ψ 1 − cos ψ a g g ω2a = − sin(ψ + α) 1 − cos(ψ + α) a g 1 Strictly speaking. and use Eq. and 1.However. The mass oscillates as the pendulum bob of a pendulum with length a. the angular coordinate ψ can only be positive. where ψ mation in the equation of motion. From Eq. we need ψ = ψ0 and ψ = ψ = 0. there is angle α given by cos α = 2 . when it will stay at the bottom). in a gravitational acceleration reduced by the rotation. we see that is possible only if sin ψ0 = 0 (top or bottom of the hoop). we could deﬁne another coordinate system where the point at the bottom of the hopp is not singular. If the mass starts near such a g/aω ¨ ¨ point. which is a singular point for the spherical coordinate system. making it oscillate about the lowest point (unless it starts there with zero velocity. However. if aω 2 /g < 1. but it is related to the energy by hψ = E − ma2 ω 2 sin2 ψ. or cos ψ0 = g/aω 2 . g = g(1 − aω 2 /g) 1 . so it cannot oscillate about zero. 2 If the angular velocity is larger than ω0 = g/a. 2 If the angular velocity is not larger than ω0 = g/a. where the acceleration given by Eq. so that ψ = ψ. If the mass starts near the bottom. we can use a small angle approxi- ¨ aψ ≈ −gψ(1 − aω 2 /g). In general. so there is an integral of motion: hψ = = ˙ ψ2 = ∂L ˙ ψ−L ˙ ∂ψ 1 1 ˙ ma2 ψ 2 − ma2 ω 2 sin2 ψ − mga cos ψ 2 2 2hψ g + ω 2 sin2 ψ + 2 cos ψ 2 ma a The integral of motion is not the total energy E = T + V . we can deﬁne a small angle ψ = ψ − α 1.

and no rotational symmetries: only px will be conserved. so will the potential associated with the force. (e) The mass is uniformly distributed in a right cylinder of elliptical cross section and inﬁnite length. Earth): the forces do not depend on the coordinates x. Since the potential is ﬁxed in all cases (i. so Lz is conserved. Exercise 2-19: Symmetries and conserved quantities We consider the gravitational forces created on particles by diﬀerent mass distributions. (a) The mass is uniformly distributed in the plane z = 0 (an inﬁnite. ﬂat. and thus the components of the linear momentum px . ﬂat. there is translational symmetry along z: only pz is conserved. y > 0 (a ﬁnite. we can ﬁnd the conserved quantities: translational symmetries are associated with components of the linear momentum. Earth. 4. (b) The mass is uniformly distributed in the half plane z = 0. we have cos α ∼ 0 (α ∼ π/2).g ω2a sin ψ cos α + cos ψ sin α 1 − cos ψ cos α − sin ψ sin α a g g ω2a ≈ − sin α 1 − cos ψ + sin α sin ψ a g = − ≈ −(ω 2 sin2 α)ψ The equation for ψ is again that of a simple harmonic oscillator. and Ω ∼ ω: the mass stays in the center of the hoop. sin α ∼ 1. but because the cylinder is inﬁnite along z. and so will the Lagrangian. and time independence with conservation of energy. 7 . in a constant position relative to the hoop’s coordinate system. with axis along the z-axis: there is now no translational symmetry. but there is still rotational symmetry about z: only Lz is conserved. rotational symmetries with components of the angular momentum.e. wit axis along the z axis: there is now no rotational symmetry. (d) The mass is uniformly distributed in a circular cylinder of ﬁnite length. the force is invariant under a rotation about the z axis. py will be conserved. If the mass distribution has a particular symmetry. with frequency Ω2 = ω 2 sin2 α: a smaller frequency than the hoop’s angular frequency. like Columbus feared): there’s only translational symmetry with respect to x. independent of time). and rotational symmetry about z. the energy is conserved in all systems. y. with axis along the z-axis: the conﬁguration has translational symmetry along z. For very high rotation frequencies ω 2 a/g. so pz and Lz are conserved. (c) The mass is uniformly distributed in a circular cylinder of inﬁnite length.. Since symmetries are associated with conserved quantities through Noether’s theorem. Also.

There are two objects in the problem. hpz + Lz will be conserved. Figure 1: Problem 2-20: a mass sliding down a sliding wedge. although pz or Lz are not individually conserved. with axis along the z axis. but since it cannot rotate. The coordinates of the mass sliding down the wedge are r = (x. (g) The mass is the form of a uniform wire wound in the geometry of an inﬁnite helical solenoid. y). Exercise 2-20: A particle on a sliding wedge A mass m is sliding down without friction along a wedge with angle α and mass M . The edge is a rigid body. The wedge can move without friction on a smooth horizontal surface. with a horizontal x-axis and a vertical y-axis pointing down. so Lz is conserved. The coordinates of the top corner of the wedge will be R = (X. the mass m and the wedge M .(f) The mass is uniformly distributed in a dumbbell whose axis is oriented along the z axis: no translational symmetries. as shown in the ﬁgure. sin α) . say the top corner. 0). Thus. or x = X + l cos α and y = l sin α 8 . and a rotation about z of 2π. 5. There are no pure translational or rotational symmetries. but there is rotational symmetry about the z axis. but there is a symmetry combining a z-translation of distance h (the distance between coils). Let us choose a coordinate system with the origin at the initial position of the top of the wedge. it is described by the coordinates of just one point. The constraint that the mass is on the wedge is r = R + l(cos α. We can treat the problem as a two dimensional problem so we have two coordinates for each mass.

y. which together with the constraint. for qi = X. The kinetic energy of the system is 1 1 ˙ T = m(x2 + y 2 ) + M X 2 . The gravitational potential energy of the mass m is V = −mg · r = −mgy The Lagrangian is 1 1 ˙ L = T − V = m(x2 + y 2 ) + M X 2 + mgy ˙ ˙ 2 2 (14) There are three Lagrange equations. since the only external force is gravity. x. 9 . and then we have a solution for X in terms of x: X = −mx/M. We can assume the initial position of the mass is at the top of the wedge. form a system of four equations for the four variables X. so we can ignore it. energy of the wedge is constant. y. y. This is one constraint. X can be added and result in ¨ m¨ + M X = 0 x This equation is saying that the horizontal position of the center of mass of the system has a constant velocity: we know this. X as f = (x − X) sin α − y cos α = 0. ∂L ∂f d ∂L − = λ dt ∂ qi ∂qi ˙ ∂qi m¨ = λ sin α x m¨ − mg = −λ cos α y ¨ M X = −λ sin α The equations for x. and it is vertical (Notice that the ﬁgure does not represent the actual motion then!). which we can express as a function of x. λ (where λ is the Lagrange multiplier associated with the constraint).where l is the distance the mss traveled down the wedge. x. and the initial velocity of the center of mass is zero. ˙ ˙ 2 2 The potential energy is just gravitational. with g = gˆ The gravitational potential j.

We can also use Lagrange’s equation for y to solve for λ. and the general solution is x = x0 +v0x t+(1/2)ax t2 . If the mass starts from rest. x0 =0. We can now use this result to obtain the equation for y: y = x(1 + m/M ) tan α ¨ ¨ sin α cos α = g (1 + m/M ) tan α 1 + (m/M ) sin2 α 1 + m/M = g sin2 α 1 + (m/M ) sin2 α = ay The acceleration of the mass along the direction tangent to the wedge is as = ax cos α + ay sin α 1 + m/M sin α cos α cos α + g sin2 α sin α = g 2 1 + (m/M ) sin α 1 + (m/M ) sin2 α = g sin α The solutions for X. v0x = 0. λ are: 2 ¨ X = −m¨/M x m sin α cos α = −g M 1 + (m/M ) sin2 α = −aM 10 . Since we assumed the mass started at the origin. and use these results in Lagrange’s equation for x: m¨ = λ sin α = m(g − y ) tan α x ¨ = m tan α(g − x(1 + m/M ) tan α) ¨ (1 + (1 + m/M ) tan α)¨ = g tan α x 2 1 + (m/M ) sin α x = g tan α ¨ cos2 α sin α cos α x = g ¨ 1 + (m/M ) sin2 α x = ax ¨ The acceleration of x is constant.We can use the constraint to solve x in terms of y (or viceversa): y = (x − X) tan α = x(1 + m/M ) tan α.

To ﬁnd out conserved quantities. The constraint force. we get y X =x− tan α 11 . the net work done by the constraint on the system is zero. If the wedge is ﬁxed. as it should be well known. and it is smaller as the wedge is steeper α → π/2. it does not move much (X ≈ 0).λ = m x ¨ sin α cos α = mg 1 + (m/M ) sin2 α The multiplier λ is the normal force from the wedge on the mass: it is mg if the wedge is horizontal (α = 0). − cos α). we need to express the Lagrangian ?? in terms of independent coordinates. and the system approximates that of a mass sliding down a ﬁxed incline. it will have a component along the mass’ velocity: the constraint force works on the particle. this is perpendicular to the mass’ motion. but if the wedge is not ﬁxed. ¨ If the wedge has a large mass (M m). the normal force. If we use the constraint to solve for X. and the normal force approximates λ ≈ mg cos α. The work done in time t by the force N on the mass m is given by: dWm dt = N·v = λ(x sin α − y cos α) ˙ ˙ = λt(ax sin α − ay cos α) m/M = λgt sin2 α cos α 1 + (m/M ) sin2 α The work done by the normal force tends to zero as m/M as the wedge mass M gets large. The constraint force also works on the wedge: dWM dt = N·V ˙ = λX sin α = −λ(mx/M ) sin α ˙ = −λ(m/M )ax t sin α = −λgt sin2 α cos α m/M 1 + (m/M ) sin2 α That is. is always perpendicular to the wedge: N = λ(sin α. The tangential acceleration approximates as ≈ g sin α.

y in terms of R. The Lagrangian does not depend explicitly on time. so there is a Jacobi integral. since x = ax t. r (natural coordinates in the rotating frame). A carriage on rotating cross-rails The position vector of the mass m in the inertial frame is r = xˆ ˆ The coordinates i+y j. y = ay t and ay = ˙ ˙ ax (1 + m/M )/ tan α. so its canonical momentum is a constant of motion: px = ∂L y ˙ = (m + M )x − M ˙ ∂x ˙ tan α This can also be obtained from the solutions. The kinetic energy is T = = = 1 2 (x + y 2 ) ˙ ˙ 2 1 ˙ m((R − rω)2 + (r + ωR)2 ) ˙ 2 1 ˙ ˙ m(R2 + r2 + 2ω(Rr − Rr) + ω 2 (r2 + R2 )) ˙ ˙ 2 12 . y are related to the spring lengths R. y: R = x cos ωt + y sin ωt r = −x sin ωt + y cos ωt We can choose as generalized coordinates x. r: x = R cos ωt − r sin ωt y = R sin ωt + r cos ωt or R. We can express x. x. r in terms of x. which is equal to the total energy E = T + V .L = = 1 m(x2 + y 2 ) + ˙ ˙ 2 1 m(x2 + y 2 ) + ˙ ˙ 2 1 ˙ M X 2 + mgy 2 1 y ˙ M x− ˙ 2 tan α 2 + mgy The coordinate x is cyclical. r. y (natural coordinates in the inertial lab frame) or R. and the length r of the spring on the perpendicular rail with force constant k (and zero rest length). where the length R of the spring with force constant K and rest length R0 . 6.

we know the energy function is not conserved. r. and the ˙ ˙ kinetic energy is homogeneous of second degree in x. since the kinetic energy is not homogenous of second degree in ˙ ˙ R. y. the Lagrangian is L = T −V 1 1 1 ˙ ˙ = m(R2 + r2 + 2ω(Rr − Rr) + ω 2 (r2 + R2 )) − K(R − R0 )2 − kr2 ˙ ˙ 2 2 2 1 1 1 1 ˙ ˙ = m(R2 + r2 ) + mω(Rr − Rr) + mω 2 (r2 + R2 ) − K(R − R0 )2 − kr2 ˙ ˙ 2 2 2 2 and is independent of time. The Jacobi integral is h = qi ˙ ∂L −L ∂ qi ˙ 13 . the Lagrangian is L = T −V 1 1 1 = m(x2 + y 2 ) − K(x cos ωt + y sin ωt − R0 )2 − k(−x sin ωt + y cos ωt)2 ˙ ˙ 2 2 2 Since the Lagrangian depends explicitly on time. so the “energy function” or “Jacobi integral” will be conserved. we know that the energy ˙ ˙ function is the mechanical energy. Because the potential does not depend on time derivatives x. However. y. r as our generalized coordinates.The potential energy is V = = 1 1 K(R − R0 )2 + kr2 2 2 1 1 K(x cos ωt + y sin ωt − R0 )2 + k(−x sin ωt + y cos ωt)2 2 2 In terms of x. and is not conserved: ∂L −L ∂ qi ˙ ∂L ∂L = x ˙ +y ˙ −L ∂x ˙ ∂y ˙ 1 1 1 = m(x2 + y 2 ) + K(x cos ωt + y sin ωt − R0 )2 + k(−x sin ωt + y cos ωt)2 ˙ ˙ 2 2 2 = T +V =E qi ˙ h = If we now instead choose R. y. then the energy function is not equal to the mechanical energy.

minus an extra potential U due to the rotation of the frame: 1 ˙ U (r. r. ω = 0 and h = T + V = E. With the rotation on. the mechanical energy is not conserved. which we can ﬁnd from the Lagrangian: L = = 1 1 1 1 ˙ ˙ m(R2 + r2 ) + mω(Rr − Rr) + mω 2 (r2 + R2 ) − K(R − R0 )2 − kr2 ˙ ˙ 2 2 2 2 1 2 2 ˙ m(R + r ) − U − V ˙ 2 We have written the Lagrangian as the kinetic energy in the (non-inertial) rotating system. and the rate of change of the energy is dE dt = 1 ˙ h + mω(Rr − Rr) + mω 2 (r2 + R2 ) ˙ 2 2 ¨ ˙ = mω(R¨ − rR) + mω (rr + RR) r ˙ d dt (15) (The following was not asked in the homework. R. and they are not conservative (they depend on velocities). R) = −mω(Rr − Rr) − mω 2 (r2 + R2 ). r. minus the potential energy in the inertial system. Lagrange’s equations are d ∂L ∂L − dt ∂ r ˙ ∂r d ∂U ∂(U + V ) m¨ = r − dt ∂ r ˙ ∂r ˙ + mω 2 r − kr m¨ = −mω R r 0 = 14 . We can ﬁnd expressions for the rate of change in energy from Lagrange’s equations of motion for R. but it has more interesting facts about this system).∂T ˙ ∂T − (T − V ) +R ˙ ∂r ˙ ∂R ˙ ˙ = mr(r + ωR) + mR(R − ωr) − T + V ˙ ˙ ˙ ˙ = mr2 + mR2 + mω(Rr − rR) − T + V ˙ ˙ 1 1 1 1 ˙ m(R2 + r2 ) − mω 2 (r2 + R2 ) + K(R − R0 )2 + kr2 ˙ = 2 2 2 2 1 2 2 2 ˙ = T + V − mω(Rr − Rr) − mω (r + R ) ˙ 2 = r ˙ We see that if the beams are not rotating. ˙ ˙ ˙ 2 The forces derived from this potential are the “centrifugal forces”.

the equations of motion are ˙ m¨ = −mω ξ − (k − mω 2 )r r ¨ mξ = mω r + mω 2 ξ + ˙ R0 1 − mω 2 /K −K ξ+ R0 − R0 1 − mω 2 /K = mω r − (K − mω 2 )ξ ˙ We see that the springs are “softened” by the rotation. Assuming small oscillations about the equilibrium length. We will learn how to ﬁnd solutions to the equations of motion of these systems in Chapter 6. but and R = R0 /(1 − mω 0 a smaller length. The equilibrium position for the K spring is not R . which are r = 0 ˙ ¨ 2 /K). and ω 2 ≥ K/m. and the oscillations in the diﬀerent directions are coupled. 15 .d ∂L ∂L − ˙ dt ∂ R ∂R d ∂U ∂(U + V ) ¨ mR = − − ˙ dt ∂ R ∂r ¨ = mω r + mω 2 R − K(R − R0 ) mR ˙ 0 = From these equations. and deﬁning ξ = R − R0 /(1 − mω 2 /K). we can ﬁnd an expression for the combinations we need for power in Eq(??): dE ¨ ˙ = mω(R¨ − rR) + mω 2 (RR + rr) = ωr(−kR + K(R − R0 )) r ˙ dt ˙ ¨ The equilibrium coordinates are the solutions to r = r = R = R = 0. the cross rail will be pushed against the rotation axis. If the system rotates too fast.

Sign up to vote on this title