EP 222: Classical Mechanics

Tutorial Sheet 2: Solution

This tutorial sheet contains problems related to the calculus of variations, and Hamilton’s
principle.

1. Show that the geodesics of a spherical surface are great circles, i.e., circles whose centers
lie at the center of the sphere.

Soln: Length element on the surface of a sphere of radius a is

q
ds = a dθ2 + sin2 θdφ2 ,

So the distance between two points 1 and 2

ˆ2 ˆθ2 q
S= ds = a 1 + sin2 θφ0 2 dθ (1)
1 θ1

0
where φ = dφdθ
. The fact that the integral of equation (1) is a minimum, implies the
Euler-Lagrange equations :  
d ∂I ∂I
− = 0,
dθ ∂ φ̇ ∂φ
( )
d φ̇ sin2 θ
p = 0,
dθ 1 + sin2 θφ0 2
( )
φ̇ sin2 θ
p = b = constant ,
1 + sin2 θφ0 2
0
φ̇2 sin4 θ = b2 + b2 sin2 θφ 2 ,
dφ b
= p ,
dθ sin θ sin2 θ − b2
ˆ
csc2 θ
φ=b √ dθ + C,
1 − b2 csc2 θ
using, 1 + cot2 θ = csc2 θ, we get,
ˆ
csc2 θ
φ=b √ dθ + C,
1 − b2 csc2 θ
ˆ
csc2 θ
φ= q dθ + C,
(1−b2 ) 2
b2
− cot θ
(1−b2 )
Let, d2 = b2
, and t = cot θ, dt = − csc2 θdθ,
ˆ
dt
φ=− √ + C,
d2− t2
 
−1 t
φ = − sin + C,
d

1
 −t = d sin (φ − C) ,
cos θ
− = d (sin φ cos C − cos φ sin C) ,
sin θ
− cos θ = A sin θ sin φ − B sin θ cos φ, (2)

above, A = d cos C,B = d sin C. Multiply both sides of equation (2) by radius a, and
recognize,
a sin θ sin φ = y,

a sin θ cos φ = x,

a cos θ = z,

we get above,
Bx − Ay − z = 0 (3)

So, the shortest length curve on the surface of a sphere lies on its intersection with a
plane which passes through the origin of the sphere (Eq. (3)). This is the definition of a
great circle.

2. 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 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.

Soln: For the most general motion, we need three generalized coordinate,

Figure 1:

2
a = Distance between the center of the hoop and the center of the cylinder= OO’. θ =
Angle made by OO’ from OA direction. This describes the motion of the c.m. of the
hoop along the surface of the cylinder. Angle φ, as shown above, defines the rotation of
the hoop about its center. Clearly,

1 1 1
T = mȧ2 + ma2 θ̇2 + Ihoop φ̇2 ,
2 2 2

Since, Ihoop = mr2 ,

1 1 1
T = mȧ2 + ma2 θ̇2 + mr2 φ̇2 ,
2 2 2
and, V = mga cos θ, so,

1 1 1
L = mȧ2 + ma2 θ̇2 + mr2 φ̇2 − mga cos θ
2 2 2

However, the motion of the hoop has constraints: (1) It moves along the surface of
cylinder,
a = r + R,

da = 0,

(2) Hoop rolls without slipping,

adθ = rdφ,

(r + R) dθ − rdφ = 0,

We will have to introduce two Lagrange multipliers λ1 (for a) and λ2 (for θ and φ).
 
d ∂L ∂L
− = λ1 ,
dt ∂ ȧ ∂a

mä + mg cos θ − maθ̇2 = λ1 , (4)

 
d ∂L ∂L
− = (r + R) λ2 ,
dt ∂ θ̇ ∂θ
d  2 
ma θ̇ − mga sin θ = (r + R) λ2 , (5)
dt
 
d ∂L ∂L
− = −rλ2 ,
dt ∂ φ̇ ∂φ

mr2 φ̈ = −rλ2 , (6)

If in (4), (5) and (6) we substitute the constraint equations,

a = r + R,

ȧ = ä = 0,
 
r+R
φ̈ = θ̈,
r
we get,
λ1 = mg cos θ − m (r + R) θ2 ,

3
  
m (r + R)2 θ̈ − mg (r + R) sin θ = (r + R) λ2 ,

mr (r + R) θ̈ = −rλ2 ,

The point where hoops leaves the cylinder, reaction force λ1 should vanish

mg cos θ − m (r + R) θ̇2 = 0. (7)

However, θ̇ at that point can be found by conservation of energy (gain in K.E.= loss in
P.E.),
1 1
m (r + R)2 θ̇2 + mr2 φ̇2 = mg (r + R) (1 − cos θ) ,
2 2
with no slipping constraint,
 2
1 1 r+R
m (r + R)2 θ̇2 + mr2 θ̇2 = mg (r + R) (1 − cos θ) ,
2 2 r

m (r + R)2 θ̇2 = mg (r + R) (1 − cos θ) ,

m (r + R) θ̇2 = mg (1 − cos θ) , (8)

putting equation (8) in (7) to get,

mg cos θ − mg (1 − cos θ) = 0,

1
cos θ = ,
2
◦
θ = 60 .

So, hoop will fall out at θ = 60◦ .

3. A point mass is constrained to move on a massless hoop of radius a fixed in a vertical plane
that rotates about its vertical symmetry axis with constant angular speed ω. Obtain the
Lagrange equations of motion assuming the only external forces arise from gravity. What
are the constants of motion? Show that if ω is greater than a critical value ω0 , there can
be a solution in which the particle remains stationary on the hoop at a point other than
at the bottom, but that if ω < ω0 , the only stationary point for the particle is at the
bottom of the hoop. What is the value of ω0 ?

Soln: Here, we have two generalized coordinates θ and φ. Thus the Lagrangian is,

4
 Figure 2:

1  
L = ma2 θ̇2 + sin2 θφ̇2 + mga cos θ,
2
But the motion in φ is trivial i.e.

φ̇ = ω = constant.

So, the effective Lagrangian has only one generalized coordinate θ

1  
L = ma2 θ̇2 + sin2 θω 2 + mga cos θ,
2

and the equation of motion is,

 
d ∂L ∂L
− = 0,
dt ∂ θ̇ ∂θ

ma2 θ̈2 − ma2 sin θ cos θω 2 + mga sin θ = 0, (9)

Since, the Lagrangian is time independent energy function will be a constant, which can
be obtained by multiplying equation (9) by θ̇ in both sides,

ma2 θ̇θ̈2 − ma2 sin θ cos θω 2 θ̇ + mga sin θθ̇ = 0,

 
d 1 2 2 1 2
ma θ̇ + ma cos 2θω 2 − mga cos θ = 0,
dt 2 4
1 2 2 1 2
ma θ̇ + ma cos 2θω 2 − mga cos θ = c,
2 4
at t = 0, θ = θ̇ = 0 (assume), we obtain

1
c = ma2 ω 2 − mga.
4

With this,
1 2 2 1 2
ma θ̇ = ma (1 − cos 2θ) ω 2 − mga (1 − cos θ) ,
2 4

5
 We are looking for solution θ̇ = 0,

1 2
ma (1 − cos 2θ) ω 2 − mga (1 − cos θ) = 0,
4

1 2 2 θ
ma ω 2 sin2 θ − 2mga sin2 = 0,
4 2
θ θ θ
2ma2 ω 2 sin2 cos2 − 2mga sin2 = 0,
2 2 2
 
θ θ g
sin2 cos2 − 2 = 0,
2 2 aω
θ
=⇒ sin2 = 0,
2
=⇒ θ = 0,

or,
θ g
cos2 = .
2 aω 2
we know that
θ
cos2 ≤ 1,
2
g
≤ 1,
aω 2
g
ω 2 ≥ = ω02 .
a
So, if ω < ω0 , the only stationary point is at the bottom. But for ω ≥ ω0 , we will have
stationary point for θ > 0.

4. A particle of mass m slides without friction on a wedge of angle α and mass M that can
move without friction on a smooth horizontal surface, as shown in the figure. Treating
the constraint of the particle on the wedge by the method of Lagrange multipliers, find
the equation of motion for particle and wedge. Also obtain an expression for the forces
of constraint. Calculate the work done in time t by the forces of constraint acting on the
particle and on the wedge. What are the constants of motion for the system?

Soln: We can describe the rotation of the wedge w.r.t a coordinate system fixed in
the ground (X − Y ) and that of the particle w.r.t a coordinate system fixed on the wedge
(x − y). The kinetic energy of the wedge can be written in terms of the (X − Y ) coordi-
nated of an point, say O’, on the wedge.

6
 Figure 3:

1  
Twedge = M Ẋ 2 + Ẏ 2 ,
2
where (X, Y ) are the coordinate of O’ w.r.t the fixed frame. If Xp and Yp are the coordi-
nates of m w.r.t the fixed frame we get the kinetic energy of the mass m,

1  
Tm = m Ẋp2 + Ẏp2 ,
2

so, the Lagrangian is,

1  2 2
 1 
2 2

L = M Ẋ + Ẏ + m Ẋp + Ẏp − mgYp − M g (Y − h) , (10)
2 2

here, (Y − h) is the Y i c.m coordinate of wedge. But from the geometry it is clear,

Xp = X + x cos α,

Yp = Y − x sin α,

Substituting Xp in equation (10)

1   1   
L = M Ẋ 2 + Ẏ 2 + m Ẋ + ẋ cos α 2 + Ẏp2 − mgYp − M g (Y − h) ,
2 2
1   1 1 1
L = M Ẋ 2 + ẋ2 + mẋ2 cos α + mẊ ẋ cos α + mẎp2 + M Ẏ 2 − mgYp − M g (Y − h) ,
2 2 2 2
This Lagrangian is expressed in four generalized coordinates X, x, Y and Yp . But, Here
we have two constraints: (1) wedge is moving only along the X−direction,

dY = 0, (11)

(2) particle is moving along the wedge in the x−direction. So the constraint equation is,

Yp = Y − x sin α,

dYp = dY − x sin αdx,

7
 dYp − dY + x sin αdx = 0. (12)

So, we introduce Lagrange multipliers corresponding to these constraints. Let λ1 and λ2

corresponds to equation (11) and (12), respectively and we have,
 
d ∂L ∂L
− = 0,
dt ∂ Ẋ ∂X
 
d ∂L ∂L
− = λ1 − λ2 ,
dt ∂ Ẏ ∂Y
 
d ∂L ∂L
− = sin αλ2 ,
dt ∂ ẋ ∂x
!
d ∂L ∂L
− = λ2 ,
dt ∂ Ẏp ∂Yp
so, we get,
(m + M ) Ẍ + mẍ cos α = 0,

M Ÿ + M g = λ1 − λ2 , (13)

mẍ cos2 α + mẌ cos α = sin αλ2 ,

mŸp + mg = λ2 , (14)

the constraints of motion imply,

Ÿ = 0,

and
Ÿp = −ẍ sin α,

So, we get from (13) and (14),

λ1 − λ2 = M g, (15)

mg − mẍ sin α = λ2 , (16)

and from (15) and (16),

λ1 = λ2 + M g = (m + M ) g − mẍ sin α, (17)

with these constraints we have only two equations of motion left,

(m + M ) Ẍ + mẍ cos α = 0,

mẍ cos2 α + mẌ cos α = sin α (mg − mẍ sin α) ,

or we get,
(m + M ) Ẍ + mẍ cos α = 0, (18)

mẌ cos α + mẍ = mg sin α, (19)

8
 from equations (18) and (19) we get separate equations of motions for Ẍ and ẍ,

m sin2 α + M Ẍ = −mg sin α cos α,

and
mg sin α (m + M )
mẍ = ,
M + m sin2 α
leading to,
−mg sin α cos α
Ẍ =
, (20)
m sin2 α + M
g sin α (m + M )
ẍ = . (21)
M + m sin2 α
Let us substitute equation (20) and (21) in equation (16) and (17) to get λ1 and λ2 ,

mg sin2 α (m + M )
λ1 = (m + M ) g − ,
M + m sin2 α

or
M (m + M ) g
λ1 = ,
M + m sin2 α
and
M mg cos2 α
λ2 = .
M + m sin2 α
Constants of Motion : (1) Since Lagrangian is not an explicit function of time, so the
energy function h will be a constant of motion.
(2) From equation (18), we get

d  
(m + M ) Ẋ + mẋ cos α = 0 ,
dt

(m + M ) Ẋ + mẋ cos α = constant,

X−component of the momentum of the whole system (particle + wedge) is constant. This
is obvious because there is no force in the x−direction on the system, hence momentum
in that direction should be conserved.

5. The one-dimensional harmonic oscillator has the Lagrangian L = 12 mẋ2 − 12 kx2 . Suppose
you did not know the solution to the motion but realized that the motion must be periodic
and therefore could be described by a Fourier series of the form
X
x(t) = aj cos jωt,
j=0

(taking t = 0 at a turning point) where ω is the unknown angular frequency of the mo-
tion. This representation for x(t) defines a many-parameter path for the system point
in configuration space. Consider the action integral I for two points t1 and t2 separated
by the period T = 2πω
. Show that with this form for the system path, I is extremum for
nonvanishing x only if aj = 0, for all j 6= 1, and only if ω 2 = k/m.

9
Soln: The Lagrangian of one-dimensional harmonic oscillator is,

1 1
L = mx2˙− kx2 ,
2 2

and it is given that motion can be describe by Fourier series,

∞
X
x (t) = aj cos jωt,
j=0

∞
X
ẋ (t) = − jωaj sin jωt,
j=0

and ∞
2
X
(ẋ (t)) = ijω 2 ai aj sin iωt sin jωt,
i,j=0

∞
X
2
(x (t)) = ai aj cos iωt cos jωt.
i,j=0

2π


So, the action integral between t1 and t1 + ω
, will be,

(t1ˆ+ 2π
ω )

I= Ldt,
t1

(t1ˆ+ 2π
ω )(
∞ ˙ ∞
)
1 X 1 X
= m ijω 2 ai aj sin iωt sin jωt− k ai aj cos iωt cos jωt dt,
2 i,j=0 2 i,j=0
t1

∞
(t1ˆ+ 2π
ω )
∞
(t1ˆ+ 2π
ω )
1 X 1 X
= m ijω 2 ai aj sin iωt sin˙ jωt dt− k ai aj cos iωt cos jωt dt, (22)
2 i,j=0 2 i,j=0
t1 t1

But we can show,

(t1ˆ+ 2π
ω ) (t1ˆ+ 2π
ω )

˙ jωt dt = π
sin iωt sin cos iωt cos jωt dt = δi,j . (23)
ω
t1 t1

So we get after substituting equation (23) in (22),

∞ ∞
1 X 2 π 1 X π
I= m ijω ai aj δij − k ai aj δij ,
2 i,j=0 ω 2 i,j=0 ω

∞
π X 2 2
aj j mω 2 − k .


I= (24)
2ω j=0

10
If x(t) is the correct solution, variation of (24) with respect to a0j s should be satisfactory,

∞
π X
2aj j 2 mω 2 − k δaj = 0,


δI =
2ω j=0

or ∞
X
2aj j 2 mω 2 − k δaj = 0.


a0 kδa0 +
j=1

Since all the variation of δaj are independent of each other, their coefficients should vanish,

a0 k = 0, (25)

aj j 2 mω 2 − k = 0, for j = 1, 2, 3, ...


(26)

Clearly from equation (25), we conclude that

a0 = 0,

because
k 6= 0.

Clearly, all but one aj (for j > 0) above will vanish because, otherwise (j 2 mω 2 − k) = 0
for non-vanishing aj ’s leading to more than one value of frequency. If we choose aj = 0,
for j = 1, we get from equation (26)

mω 2 − k = 0,
r
k
ω= .
m
Actually the book leads to a false impression. You can have any one aj 6= 0, and the final
solution will be the same as above, as shown in the tutorial. Try it !!!!!

11