CP1: Mechanics and Special Relativity

Toby Adkins

December 31, 2015

Contents

1 Mechanics I 3
1.1 Newton's Laws and Energy Conservation 4
1.1.1 Newton's Laws 4
1.1.2 Energy Conservation 4
1.2 Collisions 5
1.2.1 Centre of Mass frame 5
1.2.2 Oblique Scattering 7
1.2.3 Max Deection Angle 7
1.3 Resisted Motion 9
1.3.1 Resisted Projectile Motion 9
1.4 Variable Mass Problems 11
1.4.1 Rocket Motion 11
1.4.2 Relativistic Rocket 12
2 Mechanics II 14
2.1 Dimensional Analysis 15
2.2 Dynamical Coordinate Systems 16
2.2.1 Rotating Reference Frame 16
2.3 Rotational Dynamics 18
2.3.1 Angular Momentum and Torque 18
2.3.2 Moment of Inertia 18
2.3.3 Axis Theorems 20
2.3.4 Example Problems 21
2.4 Central Forces and Orbits 25
2.4.1 Kepler's Laws 25
2.4.2 Energy of Orbital Motion 26
2.4.3 The Eective Potential 27
2.4.4 The Reduced Mass System 28
2.4.5 Central Force Scattering 29
2.4.6 Example Problems 30
2.5 Lagrangian Mechanics 35
2.5.1 Denitions 35
2.5.2 The Lagrangian 36
2.5.3 The Hamiltonian 36
2.5.4 Example Problems 37
3 Special Relativity 45
3.1 The Michaelson-Morley Experiment and Einstein's Postulates 46
3.1.1 The Aether and The Michaelson-Morley Experiment 46
3.1.2 Einstein's Postulates 47
3.2 The Lorentz Transformations 49

1
Toby Adkins CP2

3.2.1 The Interval 49

3.2.2 Time Dilation 50
3.2.3 Length Contraction 50
3.2.4 Relativistic Velocity Addition 51
3.3 Space-Time Diagrams 53
3.3.1 An Example 53
3.4 Relativistic Dynamics 56
3.4.1 Photons 57
3.4.2 Approximations 57
3.4.3 Example Problems 58
3.5 The Relativistic Doppler Eect 63
3.5.1 Derivation 63
3.5.2 Superluminal Motion 63

2
1. Mechanics I

This chapter aims to cover the basics of mechanics, including:

• Newton's Laws and Energy Conservation

• Collisions

• Resisted Motion

• Variable Mass problems

Most of the material covered here will not be unfamiliar to students who have studied
physics in high-school, though it may be expressed in a slightly dierent way, such as using
vector notation in collisions. The mathematics required is quite simple, only requiring
basic knowledge of integration and linear algebra.

3
Toby Adkins CP2

1.1 Newton's Laws and Energy Conservation

Most of the physics dealt with in this chapter relies on these two concepts, and their
consequences. It is essential that any physics student is very much at ease with these.

1.1.1 Newton's Laws

All physics students should already be very familiar with Newton's Laws of Motion. They
are as follows:
1. Every body moves in a state of uniform motion unless acted upon by a net force
2. The rate of change of momentum is equal to the applied force:
dp dm dv
(1.1)
X
F = =v +m
dt dt dt

3. When a body exerts a force on a second body, the second body simultaneously exerts
a force equal and opposite to the second body.
Most students may not have seen (1.1) in the form shown above; generally, this will reduce
to F = m dv
dt for problems that do not involved variable mass.. However, (1.1) does lead to
an interesting result: in an isolated system, total momentum is conserved.
X
F =0
dp
→ =0
dt
This means that pr (the linear momentum) is a conserved quantity.

In general, we only use Newton's Laws for simple systems, as it is much more eective to
use other methods when the system becomes more complicated.

1.1.2 Energy Conservation

Most forces are what we term non-conservative forces. Energy is lost from the system as
a result of the force, usually dissipated as heat or light energy. However, there is a special
class of forces called (you guessed it!) conservative forces. In this case, the work done to
move an object between two points in space is independent of the path taken. An everyday
example of this is gravity. You may take any path from point a to b, but the work done
will always be given by:
Ua→b = U (b) − U (a) (1.2)
As a result, for a conservative eld, we can write the force as a gradient of some scalar
potential:
F = −∇U (1.3)
This result is proven in the notes on CP4 in the section on conservative forces.

The law of the conservation of energy states that in any closed system, subject to no external
non-conservative forces, energy is conserved. This can be used as a powerful analytic tool,
even when the system is subject to external forces. For example, one can nd the about
of energy dissipated by a resistive force by equating the energy before and after; we know
that the dierence must be equal to the work done by the non-conservative force.

4
Toby Adkins CP2

1.2 Collisions
These should be the types of problems that most people will already be familiar with at
this stage, but we would recommend reading this section in any case as there may be some
elements not previously encountered. There are a couple of crucial points to remember for
collisions:
• In all collisions subject to no external forces, momentum is conserved

• In elastic collisions, energy is conserved

• In in-elastic collisions, we have:

Energy After =  · (Energy Before) (1.4)

v2 − v1
= (1.5)
u1 − u2
The quantity  is known as the coecient of restitution. It is a measure of the amount
of energy that is conserved through the collision, and takes values 0 <  < 1. For
elastic collisions,  = 1, and for totally in-elastic collisions,  = 0.
• For an oblique collision, the motion is only changed along the common normal be-
tween two objects
We can use these results to derive a useful result for elastic collisions. Consider two objects
m1 and m2 that undergo a head-on collision. By the conservation of energy and momentum:

m1 u1 + m2 u2 = m1 v1 + m2 v2
m1 (v1 − u1 ) = m2 (u2 − v2 )

m1 u1 2 + m2 u2 2 = m1 v1 2 + m2 v2 2
m1 (v1 2 − u1 2 ) = m2 (u2 2 − v2 2 )
m1 (v1 − u1 ) · (v1 + u1 ) = m2 (u2 − v2 ) · (u2 + v2 )

Hence, we obtain:
v1 − v2 = u2 − u1 (1.6)
This means that in a elastic collision, relative velocity is maintained throughout.

1.2.1 Centre of Mass frame

This frame, referred to in short as CM, is a very useful tool for solving collision problems.
In this case, we dene all quantities relative to the centre of mass, which is dened as:
P
mi ri
rcm = Pi (1.7)
i mi

We then use the Galilean transformations to move into the CM frame.

u0i = ui − ṙcm
ui = u0i + ṙcm

Generally, one notates CM frame quantities using a prime.

The CM frame is useful for three main reasons:

5
Toby Adkins CP2

• There is net zero momentum in the CM

• The speeds of the particles before and after the collision are the same.

m1 u01 + m2 u02 = 0
m1 v 01 + m2 v 02 = 0
u02 − u01 = v 01 − v 02
m1 0 m1 0
− u − u01 = v 01 + v
m2 1 m 1
   2 
0 m1 0 m1
−u1 1 + = v1 1 +
m2 m2

→ |ui | = |v i | (1.8)
• The particles leave an oblique collision at the same angle in the centre of mass frame.
This is because angles are transformed in the CM frame; we need to use trigonometry
to convert back to the lab frame.
Consider the following example. A particle of mass 2m travelling at an initial speed u col-
lides inelastically with a stationary particle of mass m. What is the velocity of the second
particle after the collision?

First, we want to transform to the centre of mass frame:

 
2u 1
v cm =
3 0

The initial velocities in the CM are thus:

 
u 1
u01 =
3 0
 
2u −1
u02 =
3 0

Consider Figure (1.1).

Figure 1.1: Transforming between the CM and lab frames

6
Toby Adkins CP2

We can use (1.5) to write that:

 
u cos β
v 01
=·
3 sin β
 
0 2u − cos β
v2 =  ·
3 − sin β
Transforming back to the lab frame:
v 1 = v 01 + v cm
 
u  cos β + 2
=
3  sin β
v 2 = v 02 + v cm
 
2u − cos β + 1
=
3 − sin β

Using the cosine rule with reference to Figure (1.1):

2
v 02 = v 2 2 + v cm 2 − 2v 2 v cm cos θ
 s 
0
2
v2
v 2 = v cm cos θ ± − sin2 θ
v cm
 p 
2
→ v 02 = v cm cos θ ± 2 − sin2 θ

This is the nal velocity of the particle of mass m in the lab frame.

1.2.2 Oblique Scattering

Suppose that a particle of mass m undergoes an elastic collision with an identical stationary
particle. Let the angle separating the nal velocities of the two particles be θ. By the
conservation of energy and momentum:
u1 2 = v 1 2 + v 2 2
u1 = v 1 + v 2

Taking the dot product of both sides:

u1 · u1 = v 1 · v 1 + 2v 1 · v 2 + v 2 · v 2
u1 2 = v 1 2 + v 2 2 + 2v 1 v 2 · cos(θ)

→ 2v 1 v 2 · cos(θ) = 0 (1.9)
This leads to the interesting result that either v 1 = 0 or θ = π2 after the collision. In the
second case, u1 · v 1 > 0, and hence the rst particle is scattered through an angle of less
than π2 .

1.2.3 Max Deection Angle

Consider an elastic collision between a particle of mass m1 moving with a velocity u1 and
a stationary particle of mass m2 . By the conservation of momentum:
m1 u1 = m1 v 1 + m2 v 2
m1
v2 = · (u1 − v 2 )
m2

7
Toby Adkins CP2

By the conservation of energy:

m1 u1 2 = m1 v 1 2 + m2 v 2 2
m1 u1 2 − m1 v 1 2 = m2 v 2 2
m1 u1 2 − v 1 2 = m2 v 2 2

Eliminating v2 :

2
m1 m2 u1 2 − v 1 2 = m1 2 u1 2 − v 1 2



2
= m1 2 u1 2 + v1 2 − 2u1 v1 cos θ
   
u1 m2 v1 m2
cos(θ) = 1− + 1+
2v1 m1 2u1 m1

Dierentiating with respect to v1 :

   
u1 m2 1 m2
− 1− + 1+ =0
2v1 2 m1 2u1 m1
s
m2
1− m1
→ v1 = u1 m2
1+ m1

Hence, the maximum deection angle for the larger mass is given by:
 
m2
θ = sin −1
(1.10)
m1

8
Toby Adkins CP2

1.3 Resisted Motion

Most of the scenarios we deal with will not include resistive forces, as this evidently makes
the situation much more complicated. However, a particularly important case to take ac-
count of such restive forces is with motion under gravity, such as for falling objects. This
means that we can introduce the concept of the terminal velocity of an object; the object
will reach a stage where the resistive force is equal an opposite to the downwards gravi-
tational force, and so the object will stop accelerating. In general, terminal velocity (vT )
occurs as t → ∞ where dv dt = 0.

There are two main types of resistive forces in this case:

• Laminar ow - This is where the resistive force is ∝ v . It mainly acts when the body
is at lower speeds
• Turbulent ow - This is where the resistive force is ∝ v 2 . It begins to act when the
body achieves higher speeds
In most cases however, we shall only deal with one of these forces at a time. Recall that:
dv dx dv dv
= · =v (1.11)
dt dt dx dx
Depending on whether we want a time or distance dependant expression, we can solve the
equations of motion by writing the acceleration as one of these expressions.

1.3.1 Resisted Projectile Motion

This is a slightly more mathematically involved example. Suppose that the resistance
experiences by a projectile is launched at a speed u at some angle θ to the horizontal,
and that is experiences a resistive force F res = −mkv . Then, by (1.1), we can write the
equations of motion as:
dv
m = −F res + mg
dt
= −mkv + mg
Separating the two equations into components:
     
ẍ ẋ 0
= −k +
ÿ ẏ −g
We can then solve these two separable equations to obtain the solutions of:
u cos(θ)  
x= 1 − e−kt
k
1 g  g
y= u sin(θ) + 1 − e−kt − t
k k k
It has been left an an exercise to the reader to check these. For simplicity, let uo = u cos(θ)
and vo = u sin(θ). What is the max range of the projectile? First, let us nd the equation
of path:
uo  
x= 1 − e−kt
k  
1 kx
t = − ln 1 −
k uo
 
 g x
 g kx
→ y = vo + + ln 1 −
k uo k 2 uo

9
Toby Adkins CP2

Now, for max range we want y = 0. We can use the Taylor expansion of the logarithm and
ignore terms of O(k3 ) and above for small resistive forces:
1 k2 2 1 k3 3
 
vo g g k
0= x+ x+ 2 − x− x − x
uo kuo k uo 2 uo 2 2 uo 2
vo 1 g 1 gk 2
= − 2
x− x
uo 2 uo 3 uo 2
 s 
2
3uo  1 g 1 g 2 4 vo gk 
→x= − 2
± 4
+
2gk 2 uo 4 uo 3 uo 4

Then, using a second Taylor expansion, except keeping terms only in k:

2  2 !!
vo 2 k 2

3 uo 8 vo k 1 16
x=− 1± 1+ −
4 k 3 g 8 3 g2
8 vo k 32 vo 2 k 2
 
3 uo
=− 1−1− +
4 k 3 g 9 g2
2uo vo 8 vo 2 uo k
→ xmax = −
g 3 g2
Evidently, this is just an approximation, but there is no way to analytically solve the
problem.

10
Toby Adkins CP2

1.4 Variable Mass Problems

The form that (1.1) is written in becomes more relevant when the body that is being
considered has a variable mass. This is because the conservation of momentum has to be
applied both to the body and the mass that is lost.

For variable mass problems where the mass is being acquired or lost vertically at zero
velocity, the net force on the system is just equal to the weight of the mass. For a rate of
mass ejection α:
X dv
F =m
dt
= mo g − (mo − αt)g
= αgt

This just represents the loss of momentum from the system that results from the body
losing mass.

1.4.1 Rocket Motion

One very common type of problem that is encountered when considering variable mass
is that of a rocket, as it evidently it decreases in mass as fuel is forced out of the back
of the rocket. We can derive the equation of motion of the rocket my considering the
conservation of momentum from rst principles. Consider Figure (1.2), and suppose that
the fuel is ejected from the rocket at a rate α with relative speed u.

Figure 1.2: Deriving the Rocket Equation

By (1.1):
δp = (m − δm)(v + δv) + δm(v − u) − mv
= mδv − δmu − δmδv
= mδv − δmu

as δmδv is very small. Now, dm = −δm, so dividing through by δt and letting δt → 0, we

arrive at the Rocket Equation :
dp dv dm
=m +u (1.12)
dt dt dt
We can then use the fact that dm
dt = −α to nd the nal velocity of the rocket upon burnout

11
Toby Adkins CP2

(i.e when all the fuel has been used up).

m f uel
t burnout =
α
dv dm
−mg = m + u
dt dt
dv
= m − uα
dt
dv uα
= −g +
dt mo − αt
This has solution:
 
mo
→ v burnout = u · ln − gtb
mo − αtb
where tb = tburnout . A couple of points to bear in mind when solving rocket problems:
• To nd the height, integrate the expression of the velocity with respect to time up
until the burnout
• To nd the max height, use conservation laws to nd the extra height gained due to
the velocity that remains upon burnout
• For multi-stage rockets, transform to the frame moving at burnout velocity, and
solve the problem of the next stage in this frame. Then, transform back to the initial
frame by adding the initial velocity; the nal velocity is just the sum of the burnout
velocities of each stage assuming that the jettisoning of each stage is instantaneous

1.4.2 Relativistic Rocket

This is an interesting problem that involved relativistic variable mass. It is recommended
that the reader takes a look at Chapter 3 before reading through this problem.

Suppose we have a rocket that converts mass directly into energy and ejects it from the
back of the rocket in the form of photons. Initially, let us consider the problem in the rest
frame of the rocket. By the conservation of momentum:
mdv 0 + vγdmf = 0
By the conservation of energy:
dm − γdmf = 0
mdv 0 = −vdm
dv 0
m +v =0
dm
We now need to transform to the rest frame of the stationary observer. Using (3.9):
v + dv 0
v + dv = 0
1 + v dv
c2
 v 
≈ (v + dv 0 ) 1 − 2 dv 0
c
v 2 v 2
= v + dv 0 − 2 dv 0 − 2 dv 0
c c
v2
= v + dv 0 − 2 dv 0
c
2

v
→ dv = dv 0 1 − 2
c

12
Toby Adkins CP2

Now, as photons travel at v = c, we arrive at the equation of motion:

v2
 
dv
m +c 1− 2 =0
dm c

This separable equation can then be solved using partial fractions to nd the subsequent
evolution of the system.

13
2. Mechanics II

This section aims to cover some of the more advanced concepts in Mechanics, including:
• Dimensional Analysis

• Dynamical Coordinate Systems

• Rotational Dynamics

• Central Forces and Orbits

• Lagrangian Mechanics

In all likelihood, most of the concepts covered in this section will be new to most readers.
Quite a few examples have been included to help facilitate further familiarity with said
concepts. This chapter assumes that readers are already very competent with using the
concepts elucidated in Chapter (1).

14
Toby Adkins CP2

2.1 Dimensional Analysis

Dimensional analysis is a useful tool for physics in general; it can be used to nd the
form of physical formulae, or even check results obtained as a solution to a problem. A
dimension is a fundamental quality of a physical quantity, and can include things such as
length, time, mass and electric charge. Every other physical unit can be derived from some
combination of these dimensions. This means that, say, we expect an answer to have units
of m, then if the result does not have dimensions of length, then we know that something
has gone awry. Note that when we input numbers into an expression the dimensions do not
simply 'disappear'; numbers still have dimensions, unless of course that number is meant
to be dimensionless.

We denote the dimensions of some quantity Q by the notation [Q]. The 'fundamental'
dimensions are:
[T ime] = T
[M etres] = L
[M ass] = M
[Charge] = Q

The dimensions of some common physical units include:

[F orce] = M LT −2
[Energy] = M L2 T −2
[P ower] = M L2 T −3
[Current] = QT −1
[EM F ] = M L2 T −2 Q−1
We can also use dimensional analysis to derive formulae to within some dimensionless
constant. The general method is to write the desired quantity as a product of the quantities
it depends on to a set of powers, and then equate the powers of the resultant dimensional
quantities. For example, suppose we want to obtain the formula for the cyclotron frequency.
We know that Ω ∝ B, m , q . Thus:
Ω = k · B a · mb · q c
[Ω] = [k] · [B]a · [m]b · [q]c
T −1 = (M T −1 Q−1 )a · (M )b · (Q)c
= M a+b · T −a · Qc−a

Equating powers on either side of the equation, it becomes clear that:

a+b=0
a=1
c−a=0
→ a = 1 , b = −1 , c = 1
The cyclotron frequency is thus given by:
qB
Ω=k·
m

15
Toby Adkins CP2

2.2 Dynamical Coordinate Systems

Generally, we are used to coordinate systems being time-independent; the x-axis remains
in the same direction throughout all time, as is the case with all of the Cartesian coordinate
axes. However, some coordinate systems are dynamical, meaning that they change with
time. One such coordinate system is the polar coordinate system with unit vectors rb and
θb. Recall that rb · rb = 1.
d
(br · rb) = 0
dt
db r
rb · =0
dt
db r
→ rb ⊥
dt
This means that the time derivative of rb is not in the same direction as itself. In fact,
rb˙ = θ̇ · θb (2.1)
˙
θb = −θ̇ · rb (2.2)
This means that the acceleration in polar coordinates is written as:
   
r̈ = r̈ − rθ̇2 rb + 2ṙθ̇ + rθ̈ θb (2.3)

This follows from the dierentiation of the position vector r.

2.2.1 Rotating Reference Frame

For some problems, it helps to work in a rotating reference from. For F = mr̈, we can
write F 0 in the rotating frame as:
mr¨0 = F − mω̇ × r0 − 2mω × r˙0 + mω × (ω × r0 ) (2.4)
Note that in most cases, r0 k r , but the magnitude is not necessarily the same. The second
and third terms on the right-hand side of the equation represent the Coriolis force, and
the fourth represents the Centripetal force.

Consider the example of ring a projectile from a tower of height h on the equator of the
Earth. Let the Earth have angular velocity ω . The distance taken for the projectile to
reach the ground is most easily found by considering (2.4). We know that ω̇ = 0 as the
speed of rotation of the earth is constant, and we can ignore ω × (ω × r0 ) as ω is small.
mr¨0 = F − 2mω × r˙0
       
ẍ 0 0 ẋ
ÿ  = − −g  − 2  0  × ẏ 
z̈ 0 ω ż
   
0 ẏ
= − −g − 2ω −ẋ
  
0 0

16
Toby Adkins CP2

In comparison to ẏ , ẋ is very small.

ÿ = −g + 2ω ẋ
≈ −g
ẏ = −gt
1
y = − gt2
s 2
2h
→ t f all =
g

However, the other equation of motion gives:

ẍ = −2ω ẏ
= 2ωgt
1
x = ωgt3
3
Hence, the distance that the projectile ends up from the base of the tower is:
s
2 2h
x = ωh ·
3 g

This is assuming that the radius of curvature of the earth is suciently large that the
surface of the earth can be considered to be locally at.

17
Toby Adkins CP2

2.3 Rotational Dynamics

This section focusses on the concept of rotation. Generally, with think of the motion of
an object in mechanics as simple linear motion of the object. However, during it's motion,
the object may have some rotation about the centre of mass. In fact, as we will see, any
motion of the system can be split into the translation of the centre of mass, and a rotation
about the centre of mass.

2.3.1 Angular Momentum and Torque

Angular momentum is dened as:
L=r×p (2.5)
Under the condition that r ⊥ p, this can also be written as:
L = Iω
= mvr
= mr2 θ̇

From the cross product, it becomes clear that the direction of angular momentum is per-
pendicular to the plane of the instantaneous motion.
dL
= r × F + ṙ × p
dt
= r × F + ṙ × mṙ
=r×F

We dened the torque on an object as:

dL
τ =r×F = (2.6)
dt
Thus, angular momentum is conserved in the absence of external torques. This is the
'rotational analogue' (if you like) of (1.1). Let us consider the total angular momentum of
the system using the centre of mass quantities:
X
L= Li
i
X
= mi · (r0i + r cm ) × (v i + v cm )
i
X X X X
= mi · (r0i × v 0i ) + mi · (r0i × v cm ) + mi · (r cm × v 0i ) + mi · (r cm × v cm )
i i i i

The rst and third terms will disappear by the denition of the centre of mass. Hence, we
obtain:
L = L0 + r cm × M v cm (2.7)
As previously stated, this means we can characterise motion as the translation of the CM
and a rotation around it.

2.3.2 Moment of Inertia

There is a quantity used a lot in rotational mechanics called the moment of inertia of a
body. It is dened as: Z
I = r2 · dm (2.8)

18
Toby Adkins CP2

where dm = ρ · dV , and r is the perpendicular distance to the axis of rotation. If we wish

to work out rotational quantities about a particular axis, we rst need to calculate the
moment of inertia about said axis.

The moments of inertia of some common objects are as follows:

• Thin rectangular plate of side lengths a and b:
M
dm = · dxdy
ab
1
Ix = M b2
12
1
Iy = M a2
12
1
Iz = M a2 + b2


12

• Thin circular disk of radius R:

M
dm = · (2πr · dr)
πR2
1
→ Iz = M R 2
2

• A sphere of radius R. We want to calculate this by summing up the moments of

inertia of the disks that make up the sphere:
1
dI = x2 dm
2
dm = (πx2 )ρdz
3M
ρ=
4πR3
x = R2 − z 2
2

2
→ I = M R2
5

• Long, narrow rod of length l:

M
dm =
l
1
→ Iz = M l 2
12

• Solid Cylinder of radius R and length l:

Figure 2.1: Coordinate system for deriving the Moment of Inertia of a Cylinder

19
Toby Adkins CP2

 Ix :
M
dm = · (2πr · drdx)
πR2 l
1
→ Ix = M R 2
2
 Iy :
1
dIx = dmR2
2
= 2dIy
1 1
dIz = dmR2 + dmx2
4 4
1 1
→ Iz = M R2 + M L2
4 12
Some integration steps have not been included; it is recommended that the reader works
through each of these examples on their own to get a handle on deriving moments of inertia.

2.3.3 Axis Theorems

There are two 'axis' theorems that can be used to simplify or manipulate moment of inertia
calculations. These are as follows:
• Parallel Axis Theorem:

Figure 2.2: Deriving the Parallel Axis Theorem

Let d be a constant vector that gives the position of the parallel axis relative to the
axis through the centre of mass.
r0 = r + d
Z
2
I = r0 · dm
Z Z Z
= r2 · dm + d2 · dm + 2d · rdm

However, the third term will again disappear by the denition of the centre of mass.
Hence we obtain the Parallel Axis Theorem :
I = I cm + M d2 (2.9)

20
Toby Adkins CP2

• Perpendicular Axis theorem:

Z
Iz = d2 · dm
Z
= (x2 + y 2 ) · dm

→ Iz = Ix + Iy (2.10)
Note that this only holds true for a planar body. One consequence of this theorem
is if the body is uniform (e.g a regular pentagonal laminar), the moment of inertia
through any two orthogonal axes must be the same.

2.3.4 Example Problems

Below are a range of problems covering dierent aspects of the material in this section.
They should serve as examples that can be referenced when working on harder or more
involved problems.

1. Imagine that a ball of radius a rolls without slipping down a ramp inclined at an
angle θ to the horizontal. Let x be the distance moved by the ball down the ramp,
and h be the vertical distance of the ball below the top of the ramp. What is the
acceleration of the ball down the ramp?

Firstly, the condition for a ball to roll without slipping is given by v = ωr, as the edge
has to move at the same velocity as the CM. Furthermore, if the ball rolls without
slipping, we know that no energy is dissipated by frictional forces. This means that
energy is conserved. We can thus write:
1
mgh = m v 2 + Iω 2


2  
1 2 I 2
= m v + v
2 ma2
1
= mv 2 (k + 1)
2
I
f or k =
ma2
Re-arranging and substituting for h:
2gh
v2 =
k+1
2gx sin(θ)
=
k+1
Dierentiating with respect to time:
dv 2g sin(θ)
2v · = ·v
dt k+1
dv g sin(θ)
→ =
dt k+1
Evidently, it is clear that does not only hold for a sphere sphere, but in fact holds
for any object able to roll without slipping.

21
Toby Adkins CP2

2. Imagine that a snooker ball of radius a sits on a rough, horizontal table. At what
height does one have to hit the ball with the cue such that it moves o without slip-
ping?

Figure 2.3: Hitting a snooker ball

Equating torques around the ball's CM:

F (h − a) = I cm · ω̇

For rolling without slipping:

v̇
ω̇ =
a
v̇
F (h − a) = I cm ·
a
However, by (1.1):
F
v̇ =
m
I cm
h−a=
ma
I cm
→h= +a
ma
For a sphere, this becomes:
7a
h=
5
Now consider what happens if we do not hit the ball in this "sweet spot". How
far does it travel before rolling without slipping begins? Again, let us equate torques
around the centre of mass:
F · r = I · ω̇
µmgr
F = µmg → ω̇ =
I
µmgrt
ω=
I
µmgr2 t
v=
g

22
Toby Adkins CP2

We have imposed the rolling without slipping condition. The centre of mass acceler-
ation is given by r̈ = −µg .
µmgr2 t
= vo − µgt
I 
µmgr2

t µg + = vo
I
−1
µmgr2

→ to = vo · µg +
I
Thus, the distance travelled by the CM is:
1
d = vo · t − · r̈ · t2
2
1
= vo · to − · (µmg) · to 2
2
And then solve the equation numerically for d.
3. Imagine that a rod of mass m and length l is released horizontally under gravity
without resistive forces. An force of magnitude F is delivered instantaneously to
one end of the rod that causes the rod to rotate with angular frequency ωo . What
magnitude of force is required such that the rod returns exactly to its starting position?

The subsequent motion of the rod involve the motion of the centre of mass upwards
and a rotation around the centre of mass, as shown in Figure (2.4).

Figure 2.4: Decomposing the vertical impulse on a rod

We want to nd the time to max height by considering the change in angular mo-
mentum on the rod:
l
F δt = Iωo
2
F δt = mvo
l
mvo = Iωo
2
ωo l
vo =
6
ωo l
→ t max height =
6g
Using the fact that I = 12 1
ml2 about the CM for a rod. Now, for it to return exactly
to it's starting position, we require it to execute an integer number of full revolutions

23
Toby Adkins CP2

in the time it takes to rise and then fall again. Mathematically, we can express this
as:
 
ωo l
2· ωo = 2πn
6g
6πng
ωo 2 =
l
2Iωo
F =
l
1
= mlωo
6
1 p
→ F = m 6πngl
6
Thus, there are actually an innite number of forces that can be applied as n can
take any integer value; the only thing that changes is the maximum height and
subsequently the number of rotations.
4. Consider a rod of length l and mass m lying on a at horizontal table. Suppose that
an impulse is delivered at a distance x from one end of the rod. What is the value
of x that is required such that the end of the rod experiences no initial acceleration?
This is a typical 'centre of percussion' problem.

Figure 2.5: The centre of percussion for a rod

Finding the angular momentum about the end of the rod:

F · x = I end · ω
The linear acceleration of the CM is given by:
F = mv̇
F
v̇ =
m
The linear acceleration of the end of the rod is given by:
Fx l
v̇ = ·
I end 2
For the end of the rod to experience a net zero acceleration, this must be equal to
the linear acceleration of the CM.
F Fx l
= ·
m I end 2
2
→x= l
3
using the fact that the moment of inertia of a rod about its end is I end = 13 ml2 .

24
Toby Adkins CP2

2.4 Central Forces and Orbits

This section is all about tackling problems involving the orbital motion of bodies. Note
that in all cases, we will only consider the motion of a body under a single central force,
or at least a system that can be modelled as equivalent to a single central force.

What is a central force? Well, a central force is one that only acts in the radial direction,
i.e F = F (r)br. Readers will already be familiar with a number of central forces, such as
the force due to Coulomb's Law or Newton's Law of Universal Gravitation :
Gm1 m2
Fg = · rb (2.11)
r2
By it's denition, a central force will always pull an object inwards toward the origin of
the central force, and the motion of objects under the inuence of these central forces is
what we typically think of as orbits. For a very simple circular orbit, the central force is
analogous to the centripetal force that keeps the object in circular orbit.

Central forces have two main properties when applied to orbital mechanics:
• Under a central force, angular momentum is conserved

L=r×p
dL
= r × ṗ + ṙ × p
dt
= r × ṗ + ṙ × mṙ
=r×F

Using the fact that for a central force F = F (r)br :

dL
→ = r × F (r)b
r=0 (2.12)
dt
• Under a central force, the motion is conned to a plane. We know that the momentum
of a body is given by L = r × p = m(r × ṙ). It is clear that r · L = 0. As L is a
constant vector from (2.12), this is an equation of a plane through the origin. Hence
the motion is conned to a plane.

2.4.1 Kepler's Laws

Kepler's Laws govern the motion of orbital bodies, though they are less useful tools for
solving a lot of orbital problems. They are as follows:
1. The orbit of every plane is an ellipse with the sun at one of the foci
2. A line joining the planet and the sun sweeps out equal areas in equal times
1
dA = r(rdθ)
2
1 2
= r dθ
2
dA 1 2 dθ
= r
dt 2 dt
L
=
2m
dA
→ = constant
dt

25
Toby Adkins CP2

3. For a body orbiting a larger body of mass M with radius of orbit r and period T :
r3 GM
2
= (2.13)
T 4π 2
This can be derived by equating the gravitational (2.11) and centripetal forces for
an object performing circular motion.
mv 2 GmM
=
r r2
For a circle,
2πr
v=
T
Substituting this result in:
4π 2 r2 GM
= 2
rT 2 r
r 3 GM
→ 2 =
T 4π 2
Despite deriving this equation assuming a circular orbit, it does actually hold for any
periodic orbit (not hyperbolic orbits), though the proof for this is signicantly more
complicated.

2.4.2 Energy of Orbital Motion

Most problems in orbital mechanics are most easily solved by considering the total energy
of the object. As angular momentum is conserved under central forces, energy is also
conserved assuming no impulse is given to the body. We know that:
E total = KE + U
Let us rst consider the kinetic energy:
 
ṙ
v=
rθ̇
v 2 = ṙ2 + r2 θ̇2
1
KE = mv 2
2
1 2 1 2 2
= mṙ + mr θ̇
2 2
By the denition of angular momentum for the case where r ⊥ p:
L = mvr
= mr2 θ̇
1 L2
→ KE = mṙ2 +
2 2mr2
Now, the potential energy is just the work done against the central force to bring the object
in from the zero of potential, that we will dene as being at ∞.
Z
U =− F · dr
Z ∞
GmM
=− · dr
r r2
GmM
→U =−
r

26
Toby Adkins CP2

Note that this assumes that M  m, as we have assumed the centre of mass of the two
bodies is located at the centre of the body of mass M . If not, we have to use the Reduced
Mass System, as in Section (2.4.4).

The total energy of the system is thus given by the Orbit Equation:
1 L2 GmM
E total = mṙ2 + 2
− (2.14)
2 2mr r

2.4.3 The Eective Potential

The quantity:
L2 GmM
Uef f = 2
−
2mr r
is known as the Eective Potential of the system; it is the potential 'felt' by the object
when the system has been reduced to a single variable equation (in this case r). It is an
analytical tool that can be used to examine the eect that changes in L and ṙ have on the
system, as well as for classifying the type of orbit.

Graphically, we represent the eective potential as:

Figure 2.6: The Eective Potential

We have three distinct cases for the type of orbit based on the value of the energy:
• Eo - Circular motion with r = ro

• E1 - Elliptical motion with r1 < r < r2

27
Toby Adkins CP2

• E2 - Unbounded or Hyperbolic motion with r > r min

For an impulse delivered to an orbiting body, this can have an eect on the orbit in one of
two ways, depending on the direction in which the impulse is applied:
• Tangentially, L −→ L + ∆L - This causes a shift in the eective potential curve on
the graph (though the total energy remains the same), causing the radii to change.
• Radially, ṙ −→ ṙ + ∆ṙ - This causes a shift in the total energy, but the eective
potential curve remains the same. For example, the energy may shift from Eo to E1 .
When in doubt, always sketch the eective potential; seeing the energy represented graph-
ically can sometimes be the key to solving a tricky problem!

2.4.4 The Reduced Mass System

As previously stated, the Orbit Equation (2.14) is not valid when the two bodies are of
comparable size. In this case, we can treat it using the Reduced Mass System. Consider
Figure (2.7).

Figure 2.7: The Reduced Mass System

By (1.1),
m1 · r̈1 = F 12
m2 · r̈2 = F 21 = −F 12
F
r̈1 = 12
m1
F
r̈2 = − 12
m2

28
Toby Adkins CP2

Let M = m1 + m2 and µ = m1 m2
m1 +m2 (the reduced mass). Then:
r̈ = r̈2 − r̈1
1
= − · F 12
µ
GµM
→ µ · r̈ = − 2 · rb
r
We can use this equation of motion to nd the period of the orbit of the system.
GM
r̈ + ·r =0
r3 s
r3
→ T = 2π
G(m1 + m2 )

Note that this reduces to circular motion if m1 = m2 , as this implies that r1 = r2 .

2.4.5 Central Force Scattering

If an object located (initially) at the zero of potential energy is given a small velocity vo , it
will be attracted towards the body providing the central force. In this case, it will undergo
a hyperbolic orbit, which has two main characteristics:
• Distance of minimum approach - Initially, the body only has energy equal to its
angular momentum Eo = 12 mvo 2 . At the distance of closest approach, the energy is
no longer radial, meaning that ṙ = 0. Assuming energy is conserved:
1 L2 GmM
Eo = mṙ2 + 2
−
2 2mr r
1 L2 GmM
mvo 2 = −
2 2mr2 r
As angular momentum is also conserved, we know that L = mbvo for some charac-
teristic impact parameter b. Thus, solving for r, we obtain:
s
GM G2 M 2
r min = − + + b2 (2.15)
vo 2 vo 4

• Scattering Angle - Consider Figure (2.8). Evidently, the x b components of ∆p will

cancel one another, and so we just have to calculate the yb components.
GmM
Fy = · cos θ
r2
dθ L
L = mr2 → · dt = dθ
dt mr2
We can use this result to simplify the integral signicantly.
Z
∆p = Fy · dt
Z
GmM
= · dt
r2
Z π+ϕ
Gm2 M 2 2
= · cos θ · dθ
L − π2 − ϕ
2

Gm2 M ϕ
=2· · cos
L 2

29
Toby Adkins CP2

Figure 2.8: The Scattering Angle for a hyperbolic orbit

We now need another expression for ∆p. Again looking at Figure (2.8), it becomes
clear that this is equal to:
ϕ
∆p = 2 · mvo · sin
2
Equating these two expressions, we arrive at the nal expression for the scattering
angle of:
ϕ Lvo bvo 2
cot = = (2.16)
2 GmM GM
Note that this is independent of the mass of the body.

2.4.6 Example Problems

Below are a range of problems covering dierent aspects of the material in this section.
They should serve as examples that can be referenced when working on harder or more
involved problems.

1. Suppose that NASA scientists wanted to put a satellite into a circular orbit of radius
ro with a velocity vo , but instead accidentally directed the impulse at an angle θ to
the tangent of the desired orbit. What are the maximum and minimum radii of the
new orbit in terms of ro ?

In this case, the total energy of the orbit is the same, but the angular momentum is
changed. Initially, let us equate the gravitational and centripetal forces for circular
motion:
mvo 2 GmM
=
ro ro 2
GM
→ vo 2 =
ro
This is quite a useful result for simplifying expressions involving circular orbits, so
try to remember it!

30
Toby Adkins CP2

Figure 2.9: The Misdirected Launch! NASA is at it again....

The angular momentum is given by:

L = mro v perp
= mro vo cos θ

As we know the energy in each orbit is the same, we need to nd expressions for the
energy at points A and B .
1 GmM
EA = mvo 2 −
2 ro
1 GmM
=−
2 ro
1 2 L2 GmM
EB = mṙ + 2
−
2 2mr r
Equating these two expressions and letting ṙ = 0 :
1 GmM L2 GmM
− = −
2 ro 2mr2 r
1 GmM 2 L2
0= r − (GmM )r +
2 ro 2m
2 2 2
= r − (2ro )r + ro cos θ

using the results for L and vo 2 . Thus, we arrive at the quite simple expression of:
→ r = ro (1 ± sin θ)

2. A comet in a circular orbit around the Sun has speed vo and radius r0 = αR, where
R is the radius of the Earth's orbit, and α is a constant satisfying 0 < α < 1. The
comet has its velocity reduced by ∆v in a collision that does not change its initial
direction. What is the minimum ∆v required to move the comet into an orbit which
intersects that of the Earth?

31
Toby Adkins CP2

The initial energy of the comet is:

L2 GmM
Eo = 2
−
2mro ro
L = mro (vo − ∆v)

At the instant where it's velocity is reduced, it is still in circular motion. We require
that Eo = E R .
L2 GmM 1 2 L2 GmM
2
− = mṙ + 2
−
2mro ro 2 2mR R
 
1 1 GmM GmM
L2 2
− = −
2mro 2mR2 ro R
 
2 2 ro R
L = 2Gm M
ro + R
s  
2
αR
L = 2Gm M
1+α

Substituting L = mro (vo − ∆v):

s  
αR
mro (vo − ∆v) = 2Gm2 M
1+α
s  
2GM αR
vo − ∆v = ·
α2 R 2 1+α
r r
GM 2
= ·
ro 1+α
r !
2
→ ∆v = vo 1 −
1+α

How does the expression change if the impulse ∆v is delivered in the radial direction
towards the sun?

In this case, the initial energy of the comet is:

1 L2 GmM
Eo = m∆v 2 + −
2 2mro 2 ro
L = mro vo

Again, we require that Eo = E R . Using the fact that R = ro

α :
1 L2 GmM L2 GmM
m∆v 2 + 2
− = 2
−
2 2mro ro 2mR R
1 1 2
L α GmM α
m∆v 2 − mvo 2 = 2
−
2 2 2mro ro
1 1
m∆v = mvo · (α + 1) − mαvo 2
2 2 2
2 2
1
= mvo 2 (α2 − 2α + 1)
2
→ ∆v = vo (α − 1)

32
Toby Adkins CP2

Notice how the results are markedly dierent depending on the way in which the
impulse is applied. In general, it is much more 'energy-ecient' to transfer between
orbits by changing the angular momentum, rather than giving the body some radial
velocity.
3. A probe of mass m is initially in circular orbit with radius r1 . Show that with two
appropriately placed impulses, the probe is able to reach a second circular orbit of
radius r2 > r1 .

This question introduced the concept of a Hohmann transfer orbit; it is one of the
lowest energy methods for transferring an object between two circular orbits.

Before we tackle this question, we need to derive a useful result. Consider the energy
of a circular orbit ro :
1 GmM
|Ec | =
2 ro
1 GmM
= mv 2 −
2 r
Rearranging for v , we arrive at the vis-viva equation:
s  
2 1
v= GM − (2.17)
r a

for a semi-major axis a. Note that despite deriving it by assuming circular motion,
the result does hold true for any closed orbit. Now we are ready to tackle the problem
at hand.

We can use the impulses ∆v1 and ∆v2 , as shown in Figure (2.10), to achieve the
desired result.

Figure 2.10: The two impulses for a Hohmann Transfer Orbit

33
Toby Adkins CP2

Let us calculate each of the successive impulses. The initial orbital velocity is given
by:
r
GM
v1 =
r1
At perigee, the elliptical transfer orbit has velocity:
s  
2 2
v10 = GM −
r1 r1 + r2
s  
2r2
= GM
r1 (r1 + r2 )

Thus, the rst impulse must have value:

r r 
GM 2r1
→ ∆v1 = · −1
r1 r1 + r2

The orbital velocity in the second circular orbit is given by:

r
GM
v2 =
r2
At apogee, the elliptical transfer orbit has velocity:
s  
2 2
v20 = GM −
r2 r1 + r2
s  
2r1
= GM
r2 (r1 + r2 )

Thus, the second impulse must have value:

r  r 
GM 2r1
→ ∆v2 = · 1−
r2 r1 + r2

34
Toby Adkins CP2

2.5 Lagrangian Mechanics

As stated in Section (1.1.1), Newton's Laws are a very cumbersome way to solve more
complicated systems due to both the sheer computation required and the fact that it starts
to becomes horrendously complicated to resolve forces. Lagrangian Mechanics allows us
to solve such problems much more easily by considering the principle of stationary action;
that the dynamics of a system are the minimization of a function that contains all the
information in a system. Consider a function l:
Z x1
l= f (y, y 0 , x) · dx
xo

Writing the Taylor Expansion of the function f :

∂f ∂f
δf = δy + 0 δy 0 + O(δy 0 )
∂y ∂y
d
δy 0 = (δy)
dx
Z x1  
∂f ∂f ∂
δl = δy + 0 δy · dx
x ∂y ∂y ∂x
Z ox1  
∂f d ∂f
= δy − δy · dx
xo ∂y dx ∂y 0
using integration by parts. We want δl = 0 for the functional to be minimised:
Z x1   
∂f d ∂f
0= − δy · dx
xo ∂y dx ∂y 0
 
d ∂f ∂f
→ =
dx ∂y 0 ∂y
This principle of path minimisation is what the Lagrangian Method. In fact, we have just
derived the functional form of the Euler-Lagrange equations, but more on this later. It is
often quite helpful to default to using Lagrangian Mechanics if unsure of how to solve a
problem, as this will often give way to a solution.

2.5.1 Denitions
First, we need to dene a few key terms and concepts that are going to be referred to
throughout the course of this section. They are as follows:
• Generalised Coordinates - A set of parameters qk that specify the conguration of
the system. These do not have to be orthogonal
• Degrees of Freedom - The number of independent coordinates that is sucient to
uniquely describe a system
• Holonomic Constraints - Constraints on a system are described as holonomic assum-
ing that:
1. The constraints are independent
2. The system can be described by relations between general coordinate variables
and time
3. The number of generalised coordinates is reduced to the degrees of freedom of
the system
• Noether's Theorem - Every dierentiable symmetry of the action of a physical system
has a corresponding conservation law

35
Toby Adkins CP2

2.5.2 The Lagrangian

The Lagrangian (L) is not unique to mechanics; in fact, it crops up in many dierent areas
of physics. Generally, however, the Lagrangian is a function of the generalised coordinates
(qk ), their time derivatives (q˙k ) and time.
L = L(qk , q˙k , t)

In mechanics, the Lagrangian is simply dened as:

L = KE − U (2.18)
For example, we can dene L some of the coordinate systems that we are familiar with:
• Cartesian:
1
L = m ẋ2 + ẏ 2 + ż 2 − U (x, y, z)


2

• Cylindrical:
1  
L = m ṙ2 + r2 θ̇2 + ż 2 − U (r, θ, z)
2

• Spherical:
1  
L = m ṙ2 + r2 θ̇2 + r2 sin2 θ · φ̇2 − U (r, θ, φ)
2

Now, how do we use this? The hardest part about using the Lagrangian is usually nding
the Lagrangian itself; the rest is simply handle-turning. Once we have managed to nd L,
we can nd the equations of motion of the system using the Euler-Lagrange Equations :
 
d ∂L ∂L
= (2.19)
dt ∂ q˙k ∂qk

Note that for each coordinate and it's time derivative, we have one Euler-Lagrange (EL)
equation. Note that the time derivative on the left-hand side is total; that is, which dif-
ferentiate the coordinates contained in the expression with respect to time, rather than
dierentiate explicitly with respect to time. Don't worry if this seems a little confusing
right now; have a read through some of the examples to becomes familiar with the process.

The EL equations actually have some interesting consequences due to Noether's Theorem
(2.5.1). Essentially, if ∂q
∂L
= 0, then the quantity ∂∂L
q˙k is conserved. This is know as
the conjugate momentum, denoted by pk , and can be things such as linear or angular
k

momentum that are constants of the motion.

2.5.3 The Hamiltonian

The Hamiltonian is a quantity derived from the Lagrangian, and is often useful for exam-
ining perturbations around some minimum of a system.
!
∂L
(2.20)
X
H= q˙k −L
∂ q˙k
k

Usually, H is written as a function of p˙k , q˙k and time.

36
Toby Adkins CP2

Like with the Lagrangian, the Hamiltonian has a set of equations that we use to derive the
equations of motion, know as the Hamiltonian Equations :
∂H
p˙k = − (2.21)
∂qk
∂H
q˙k = (2.22)
∂pk
Note that the Hamiltonian itself is a characteristic of the motion of the system. Consider
its total derivative:
∂H ∂H ∂pk ∂H ∂qk dH
= · + · +
∂t ∂pk ∂t ∂qk ∂t dt
Using (2.20):
∂H dH
=
∂t dt
This means that if H has no explicit time dependence, it is a conserved quantity for the
system.

For small variations about a given minimum, we can either substitute qk = qo + δq and
expand (recalling that δq
qo  1 ) or we can use the concept of an eective potential:

∂Uo ∂ 2 U o qk 2
U ef f = Uo + · qk + · + ···
∂qk ∂qk 2 2!
∂ 2 Uo
q¨k = − · (qk − qo )
∂qk 2
around a minimum qo with a value of potential Uo at that point.

2.5.4 Example Problems

This material might seem complicated when reading over it (or it might not, who knows?),
so the best way to become better acquainted with it is to read over and work through some
examples.
1. A common problem is that of the compound pendulum; a non-uniform body that
rotates about some axis that may not be through the centre of mass. What is its
period for small angle oscillations?

Let the pendulum have total mass m and moment of inertia I about the axis of
rotation. Let us nd the Lagrangian:
U = −mgl cos θ
1
KE = I θ̇2
2
1 2
→ L = I θ̇ + mgl cos θ
2
Applying the EL equation:
d  
I θ̇ = −mgl sin θ
dt
I θ̈ = −mgl sin θ

37
Toby Adkins CP2

Figure 2.11: A typical compound pendulum

Using the small angle approximation that sin θ ≈ θ:

mgl
θ̈ + ·θ =0
I s
I
→ T = 2π
mgl

We can then substitute the moment of inertia of an appropriate body into this for-
mula.
2. Suppose that two equal masses are joined by a string that runs over two pulleys. One
of the masses is constrained to move in the vertical direction, while the other is free
to swing freely, as shown in Figure (2.12). Describe the evolution of the system for
small angle oscillations of the right-hand mass of magnitude .

Figure 2.12: Two constrained masses in a pulley system

Let the distance from the wheel of the pulley to the right hand mass be r. Dene
the zero of potential of the masses as the equilibrium position when both masses are

38
Toby Adkins CP2

at rest and the system is balanced.

1 1 1
KE = mṙ2 + mṙ2 + mr2 θ̇2
2 2 2
U = mgr − mgr cos θ
1
→ L = mṙ2 + mr2 θ̇2 − mgr + mgr cos θ
2
Applying the EL equations:
d
(2mṙ) = mrθ̇2 − mg + mg cos θ
dt
2r̈ = rθ̇2 − g(1 − cos θ)
mθ̈ = −mgr sin θ
g
θ̈ + sin θ = 0
r
We are told that the amplitude of the oscillations is . Thus, the small angle solution
to the θ coordinate equation is:
r
g
θ(t) =  cos( t + φ)
r
=  cos(ωt + φ)

For small oscillations, cos θ ≈ 1− θ2 . Substituting this into the r coordinate equation
2

of motion:
θ2
  
2
2r̈ = rθ̇ − g 1 − 1 −
2
θ 2
= rθ̇2 − g
2
g
= r (ω sin(ωt + φ))2 − ( cos(ωt + φ))2
 2 
2 2 1 2
=  g sin (ωt + φ) − cos (ωt + φ)
2
Averaging over a couple of periods:
2 g
2r̈ =
4
2
 g
r̈ =
8
As  > 0, g > 0, this is a positive quantity, and so the left hand mass gradually
climbs while the right hand mass gradually falls. This is a second order eect.
3. Consider a ball of mass m rolling without slipping on the surface of a hemisphere
that is able to deform slightly. Find the normal force. At what angle will the ball lose
contact with the surface?

As the surface is able to deform, we need to take account of the potential energy of
this by a 'step' potential function V (r).
1 1
KE = mr2 θ̇2 + mṙ2
2 2
U = mgr cos θ + V (r)
1 1
→ L = mr2 θ̇2 + mṙ2 − mgr cos θ − V (r)
2 2

39
Toby Adkins CP2

Figure 2.13: A ball rolling on a 'springy' hemisphere

Applying the EL equations, we obtain the equations of motion as:

mr2 θ̈ = mgr sin θ
mr̈ = mrθ̇2 − V 0 (r) − mg cos θ

We now place the constraints that r = R and r̈ = ṙ = 0.

dV
0 = mRθ̇2 − − mg cos θ
dr r=R
dV
− = mg cos θ − mRθ̇2
dr r=R

However, recall (1.3). The reaction force is thus:

Fr (θ, θ̇) = mg cos θ − mRθ̇2

The ball will lose contact with the surface the instant that Fr = 0.
0 = mg cos θ − mrθ̇2
mg cos θ = mRθ̇2
R  v 2
cos θ =
g R
v2
=
Rg
By the conservation of energy from when the ball is initially at rest at the top of the
hemisphere:
1
mv 2 = mgR(1 − cos θ)
2
v 2 = 2gR(1 − cos θ)
R cos θ = 2R(1 − cos θ)
3R cos θ = 2R
2
cos θ =
3
Hence, the particle will leave the surface at θ max = cos−1 2
.


3

40
Toby Adkins CP2

Figure 2.14: A point mass attached to the rim of a disk

4. Imagine that a point mass of mass m is attached to the rim of a disk of mass M .
What is the value of the period for small angle oscillations amount the minimum of
the motion?

Let the zero of potential be at the centre of the wheel. Now, assuming that the
wheel rolls without slipping, the position x of the centre of the wheel must satisfy
the condition x = Rθ. We can thus write the position of the smaller mass as:
r = (x − R sin θ, R(1 − cos θ))
= (R(θ sin θ), R(1 − cos θ))
ṙ = R(θ̇ − θ̇ cos θ, θ̇ sin θ)

Considering ṙ2 to nd the kinetic energy:

ṙ2 = R2 (θ̇2 − 2θ̇2 cos θ + θ̇ cos2 θ + θ̇2 sin2 θ)
= 2R2 θ̇2 (1 − cos θ)
1 1
KE = M ẋ2 + m(R2 θ̇2 )(1 − cos θ)
2 2
1 1
= M R2 θ̇2 + m(R2 θ̇2 )(1 − cos θ)
2 2
U = −mgR cos θ
1 1
→ L = M R2 θ̇2 + m(R2 θ̇2 )(1 − cos θ) + mgR cos θ
2 2
We have thus derived a Lagrangian for the system that only has one degree of freedom.
Applying the EL equations:
d   1
M R2 θ̇ + 2mR2 θ̇(1 − cos θ = − (2mR2 θ̇2 ) sin θ − mgR sin θ
dt 2
M R2 θ̈ + 2mR2 θ̈(1 − cos θ) = −mR2 θ̇2 sin θ − mgR sin θ
θ̈ M R2 + 2mR2 (1 − cos θ) = − sin θ(mR2 θ̇2 + mgR)

For small oscillations, let sin θ ≈ θ and cos θ ≈ 1 − 2 .

θ2
Neglecting terms of O(θ2 )

41
Toby Adkins CP2

and above:
θ̈(M R2 ) + mgR sin(θ) = 0
mg
θ̈ + θ=0
MR
Hence the frequency of small angle oscillations about equilibrium is:
r
mg
Ω=
MR

5. A bead of mass m slides around inside a cone of semi-vertical angle α. The bead is
constrained to move on the surface of the cone. Find the frequency of small oscilla-
tions about a stable equilibrium, and nd the value for α for which this frequency is
equal to the frequency at stable equilibrium.

Figure 2.15: A bead sliding around inside a cone

Take the zero of potential to be at the vertex of the cone.

r
tan α =
h
r
h=
tan α
U = mgh
mgr
=
tan α
Similarly,
r
sin α =
x
r
x=
sin α
ṙ
ẋ =
sin α
ṙ2

1 2 2
KE = m + r θ̇
2 sin2 α

42
Toby Adkins CP2

Hence we can write the Lagrangian of the system as:

ṙ2
 
1 mgr
→L= m + r2 θ̇2 −
2 sin2 α tan α
Now, applying the EL equations:
mr̈ mg
2 =− + mrθ̇2
sin α tan α
mg
r̈ = − cos α sin2 α + rθ̇2 sin2 α
sin α
= rθ̇2 sin2 α − g cos α sin α

Evidently, p θ = L = mr2 θ̇ is a conserved quantity; this is the angular momentum of

the bead around the zb axis. Let us re-write the equation of motion in terms of this
quantity:
L
θ̇ =
mr2
L2 sin2 α
r̈ = − g cos α sin α
m2 r3
We have reduced this to a one dimensional problem in the r coordinate.

To nd the frequency of stable circular motion, we need to impose the conditions
that r̈ = 0 , ṙ = 0 , r = ro .
0 = ro θ̇2 sin2 α − g cos α sin(α)
θ̇2 ro sin2 α = g cos α sin α
r
g
→ θ̇c =
ro tan α
Now for small oscillations about equilibrium, we let r = ro + δ .
L2 sin2 α
δ̈ = − g cos α sin α
m2 (ro + δ)3
L2 sin2 α
=  3 − g cos α sin α
m2 ro 3 1 + rδo

Binomially expanding using the fact that ro  δ :

L2 sin2 α
 
δ
δ̈ = 1−3 − g cos α) sin α
m2 ro 3 ro
L2 sin2 α L2 sin2 α
= − 3 δ − g cos α sin α
m2 ro 3 mro 4
L2 sin2 α
= −3 δ
mro 4
Substituing in for L, we obtain:
3g
δ̈ + sin α cos α · δ = 0
ro
r
3g
→Ω= sin α cos α
ro

43
Toby Adkins CP2

For this to equal the frequency of circular motion, we require:

Ω = θ̇c
3g g
sin α cos α =
ro ro tan α
1
sin2 α =
3  
−1 1
→ α = sin √
3

44
3. Special Relativity

This chapter aims to cover the basics of Special Relativity, including:

• The Michaelson-Morley Experiment and Einstein's Postulates

• The Lorentz Transformations

• Space-Time Diagrams

• Relativistic Dynamics

• The Relativistic Doppler Eect

This turns out to be one of the more interesting topics of the mechanics course, as it has
consequences that not many students will have encountered or considered before. However,
some of the concepts can be dicult to grasp, so it is worth working through problems fully
to get to grips with the material. Note that for the entirety of this chapter, the symbol γ
is given by:
1
γ=q
v 2


1− c

where c is the speed of light in a vacuum, and v is the velocity of the moving body.
Sometimes, we use β = vc to simplify some expressions.

45
Toby Adkins CP2

3.1 The Michaelson-Morley Experiment and Einstein's Pos-

tulates
In special relativity, an event is dened as a point in both space and time (more commonly
referred to as 'space-time') that has an associated physical occurrence. Unlike what we
currently think of an event as, such as "meeting up with friends for coee" or "going to
the cinema", events in special relativity are used to refer to a specic instance in time,
such as the emission of light instantaneously from a source, or its reception at another
point in space time. An event can be described by a time coordinate, and a set of spatial
coordinates. For example, we can dene an event E by E (ct, x, y, z). This is using what
is known as a four-vector. It consists of a time component, followed by three spatial com-
ponents. Here it is being used to represent the space-time coordinates of an event, but it
can also be used to represent the energy and momentum of a particle, amongst other things.

A frame is a convenient standard of rest or coordinate system in which we have chosen to

make our measurements. An inertial reference frame is one in which (1.1) is obeyed as it is
not accelerating. For example, imagine a car accelerating away from a set of trac lights
that just turned green. We can observe the scenario from the inertial frame of a stationary
person standing on the pavement, or in the non-inertial frame of the accelerating car. Note
that the term 'stationary' is used in this case in a relative sense; we have dened an object
to be stationary if it has no velocity in the frame of the pavement, but this is not a universal
standard of rest.

3.1.1 The Aether and The Michaelson-Morley Experiment

Now, the question arises: what frame of reference do we measure light relative to? One
such theory was that it propagated through a medium called the 'aether' at the speed of
light. The aether was all permeating and had zero viscosity (properties that these days
would quickly ring alarm bells in the head of any switched on physicist!). The speed of
light would thus be measured based on the relative motion between a body and the aether.

The Michaelson-Morley experiment was designed to test this theory, using the apparatus
shown in Figure (3.1). We assume that the apparatus is moving at a speed v with respect
to the aether.

Consider the time taken for the light to travel along each of the perpendicular paths. First,
t1 along the horizontal path:
l l
t1 = +
c+v c−v
2l
=  2

c 1 − vc2
v2
 
2l
≈ 1+ 2
c c

46
Toby Adkins CP2

Figure 3.1: A schematic drawing of the setup of the M-M experiment

Then, t2 along the vertical path:

2l
t2 = √
c2
− v2
2l
= q
2
c 1 − vc2
1 v2
 
2l
≈ 1+
c 2 c2
Finally, taking the time dierence:
∆t = t1 − t2
lv 2
=
c3
6 0
=

The M-M experiment thus predicted that there would be a time dierence in the arrival of
light at the detector that would cause a shift in the background interference pattern present.
However, no such shift was observed. This null result contributed to the destruction of the
aether model, and it's eventual replacement by Einstein's Postulates.

3.1.2 Einstein's Postulates

Einstein's paper on the Special Theory of Relativity rests on two main postulates:
1. The outcome of any experiment is the same in any inertial frame of rest
2. Light in a vacuum travels at a constant speed c at all times, regardless of one's frame
of observation
The rst is essentially a statement of the idea that no ctitious forces are apparent in iner-
tial frames. The second is really where things start to complicate themselves; this simple
statement that light is constant for all observers has quite remarkable consequences. Note
that a secondary consequence of this postulate is that no information can be transmitted
above the speed of light; it is the universal ceiling (if you like) for information transmission.

47
Toby Adkins CP2

One such consequence (more to follow in later sections) is that events that appear simul-
taneous in one frame may not be so in another. Consider the 'thought experiment' of two
observers; one located equidistant from each ends of a train moving at a relativistic speed
v , and the other on the stationary platform.

Figure 3.2: A diagram showing the lengths involved in the Train thought experiment

Lights at either end of the train (A and B) are turned on. First, consider the time taken
for the light to travel along OA (t1 ) and OB (t2 ) i.e in the rest frame of the observer inside
the train.
OA L
t1 = =
c c
t1 = t2

This means that for the observer on the train observes the lights at A and B to turn on
simultaneously. What about the observer on the platform?
L


OA0 L−v
t1 0 = = c
c c
L


OB 0 L + v
t2 0 = = c
c c
t1 6= t2

This means that the observer on the platform observes the lights turning on at dierent
times (B before A). This, it is clear that whether or not two events are simultaneous will
depend on the frame of reference from which they are being observed.

48
Toby Adkins CP2

3.2 The Lorentz Transformations

At this stage, it is useful to dene two key terms:
• Proper Time - The shortest time interval between two events, as measured in the
frame at which the relevant clock is at rest
• Proper Length - The length of a body as observed in its own reference frame

As we have seen, our perception of length and time can change depending on the frame
that we nd ourselves in. We can relate these quantities in two frames by The Lorentz
Transformations (LT's). An event in frame S dened by (ct, x, y, z) is related to an event
(ct0 , x0 , y 0 , z 0 ) in frame S 0 by:

x0 = γ(x − vt) (3.1)

 vx 
t0 = γ t − 2 (3.2)
c
We can also consider the transformation of energy E and momentum p between these
frames:
E 0 = γ(E − vp) (3.3)
 
vE
0
p =γ p− 2 (3.4)
c

These can also be represented as a matrix, which is useful if we want to perform successive
LT's. It also implies that the Lorentz transformations are in some sense a 'rotation' in
space-time (remark how the matrix is symmetric).
 
γ −βγ 0 0
−βγ γ 0 0
Λ=
 0
 (3.5)
0 1 0
0 0 0 1

We have included the zero entries to impress upon the reader the importance of the fact that
the LT's apply to all directions, even though we are dealing with only the one-dimensional
case at this stage.

Note that we re-obtain the conventional Galilean Transformations for v  c.

γ→1
t0 → t
x0 → x − vt

This to be expected, as the eects of special relativity are negligible at low velocities.

3.2.1 The Interval

The LT's have a particular property in that they preserve a quantity that we call 'the
interval'. The interval s of a particular set of space-time coordinates is dened as:
s2 = c2 t2 − x2 − y 2 − z 2 (3.6)
This is analogous to the dot product of a four-vector with itself. A true rotational matrix
leaves the length of a vector completely unchanged, but in the LT case it is only the interval

49
Toby Adkins CP2

that is an invariant quantity ; a quantity that does not change regardless of one's frame of
reference.

Two observers may nd dierent space and time separations of two events, but they will both
agree on the interval that they calcuate. The use of such invariant quantities becomes more
relevant later on when we start tackling relativistic dynamics, but we have introduced it
here so the reader becomes more familiar with it reading through the subsequent sections.

3.2.2 Time Dilation

One consequence of Einstein's Postulates is the eect of time dilation; that time appears
to pass slower on objects moving relative to the observer. When rst encountered, this
appears as a bit of an odd eect, as it holds for whatever frame you place yourself in. For
example, let observer A be in a relativistic rocket moving past an observer B on the Earth.
From B's frame of reference, the time appears to be passing more slowly on the clock in
A's frame of reference because B observes A to be moving relative to their apparently
stationary frame on the earth. Conversely, A observes time to be passing more slowly on
the clock in B's frame of reference, as from this frame the earth appears to be moving, while
the rocket remains stationary relative to A.

Consider the proper-time interval ∆t0 = t02 − t01 between two events as measured by a clock
that is stationary in frame S 0 , which itself is moving at a velocity v relative to frame S .
Using (3.2), the time interval between these two events as perceived in frame S is given
by:
∆t = t2 − t1
vx02 vx01
   
0 0
= γ t2 − 2 − γ t1 − 2
c c
v
= γ(t02 − t01 ) − γ 2 (x02 − x01 )
c
However, these two events are measured by a clock that is stationary in frame S 0 . This
means that x02 = x01 . Hence, we nd that:
∆t = γ∆t0 (3.7)
Remember when trying to apply this equation that time is dilated on objects moving relative
to your frame of reference.

3.2.3 Length Contraction

Similar to time dilation, length contraction describes how lengths in frames moving relative
to the observer appear contracted. Note that this contraction only occurs in the direction of
motion. For example, if a spacecraft is rushing past the Earth at a relativistic speed, it will
appear 'squashed' in it's direction of motion, but have the same dimension perpendicular
to it.

Consider the proper length interval ∆x = x2 − x1 at rest in frame S . Using (3.2), the
length interval between these two events as perceived in frame S 0 is given by:
∆x = x2 − x1
= γ(x02 − vt02 ) − γ(x01 − vt01 )
= γ(x02 − x01 ) − γ(t02 − t01 )

50
Toby Adkins CP2

However, these measurements are taken at the same time in frame S 0 , meaning that t02 = t01 .
Hence we nd that:
∆x
∆x0 = (3.8)
γ
Remember that when trying to apply this equation that length is contracted on objects
moving relative to your frame of reference.

3.2.4 Relativistic Velocity Addition

Imagine that we have two spacecraft travelling along an axis towards each other. Space-
craft U has velocity u = 0.8c and Spacecraft V has velocity v = 0.9c. If we attempt to nd
their relative velocity by using the 'classical' method of adding their velocities, we obtain
u0rel = 1.7c. However, we know that as a consequence of Einstein's Postulates, nothing
can travel faster than the speed of light.

In Special Relativity, u0rel is known as the mutual velocity of A and B; this does not actu-
ally violate the limitation of the speed of light as no information travels at this velocity. If
we wanted to beam a signal between the two ships and measure their relative velocity, we
would not obtain v rel . So the question becomes: how do we add velocities in relativity?

For this, we want to consider these spacecraft U and V moving away from one another
along the x-axis with velocities u and v respectively. To nd the mutual velocity of the two
objects, we want to transform the rate of change of the distance between the two objects
in the rest frame of a stationary observer into the frame of one of the two spacecraft. For
simplicity's sake, we are going to perform this derivation assuming only one-dimensional
motion, but it can be generalised to three dimensions. Let us transform into the frame of
spacecraft U. Considering (3.2) in their dierential form:
dx0
u rel =
dt0
γ(dx − vdt)
=
γ dt − v dx


c2
dx − vdt
=
dt − v dx
c2
dx
dt −v
= v dx
1− c2 dt

But dx
dt = u. Thus, we obtain a formula for relativistic velocity addition of:

u−v
u rel = (3.9)
1 − u·v
c2

We have gone back to vector notation here to remind the reader that u and v are in fact
vector quantities.

Let us go back to our example of the spacecraft A and B moving towards one another. As
before, we found that the mutual velocity was 1.7c. Now using (3.9):
vA + vB
u rel =
1 + vAc2vB

51
Toby Adkins CP2

Note the signs have changed as the objects are moving towards one another in this problem.
0.8c + 0.9c
u rel =
1 + (0.8)(0.9)
≈ 0.988c

Now, what if A and B were moving away from one another?

vA − vB
u rel =
1 − vAc2vB
0.8c − 0.9c
=
1 − (0.8)(0.9)
≈ −0.357c

As we have shown, (3.9) gives very dierent results to simply adding the velocities; this is
the nature of Special Relativity!

52
Toby Adkins CP2

3.3 Space-Time Diagrams

Space-time diagrams are useful tools for analysing problems of Special Relativity. In
essence, they are a 'graph' of events in space time, with spatial coordinates on the x-
axis, and time coordinates on the y -axis. Each event in time is represented by a world line,
and has an associated light cone.

Figure 3.3: An example space-time diagram

Event A can only eect events that lie inside it's light cone. Intervals inside the light cone
obey (∆s)2 > 0, and are referred to as 'time-like'. On the light cone itself, (∆s)2 = 0.
This is said to be 'light-like'. Elsewhere, (∆s)2 < 0, and this is said to be 'space-like'.

Objects or events that remain xed in space-time have vertical world lines, but objects or
events that move in space-time have slanted world lines. The faster the speed, the less
steep the world line. By convention, world lines that are at 45◦ represent objects or events
travelling at the speed of light. This means that no world line can be at less than 45◦ .

3.3.1 An Example
A stationary observer A res a light pulse at a mirror located at a distance d away. A
second observer B is moving at a relativistic speed v along the line of projection of the
light pulse. What is the time interval between the transmission and reception of the signal
by A from B's frame of reference?

We can solve this problem quite eectively by the use of space-time diagrams in both A's
and B's rest frame.

Let Event 1 be the point at which the light hits the mirror, and Event 2 the time the light
returns to observer A. For Event 1:
x1 = d
ct1 = cto

53
Toby Adkins CP2

Figure 3.4: A space-time diagram in the rest frame of A

Applying the LT's:

x01 = γ(d − cβto )

For Event 2:
x2 = 0
ct2 = 2cto
x02 = −2γβcto

This means that the space-time diagram in the rest frame of B becomes:

Figure 3.5: A space-time diagram in the rest frame of B

This is because the world lines of the observer A and the mirror at a distance d appear to
be moving in the rest frame of B. Now the time interval in B's frame of reference is then

54
Toby Adkins CP2

given by:
tB = t1 + t2
γ γ
= (d − cβto ) + · 2βto
c  c
d
=γ + βto
c
 
d d
=γ +β
c c
p
d (1 + β)2
= ·p
c (1 − β)(1 + β)
s
d 1+β
→ tB = ·
c 1−β

As you can see, the calculation is made easier to visualise and understand what is going
on when we use both the LT's and space-time diagrams in conjunction with one another.

55
Toby Adkins CP2

3.4 Relativistic Dynamics

Unlike with Classical Dynamics, we cannot simply say that if a body is at rest at zero
potential it has zero energy; we have not taken account of the rest mass energy of the
body. In Special Relativity, the energy and momentum of a body are given by:

E = γmc2 (3.10)
p = γmv (3.11)
KE = (γ − 1)mc2 (3.12)
Consider the following:
E 2 − p2 c2 = γ 2 m2 c4 − γ 2 m2 v 2 (3.13)
v2
 
2 2 4
=γ m c 1− 2 (3.14)
c
= m2 c4 (3.15)
This is an invariant quantity because the rest mass m of a body can never change, making
it an important (and useful) result. We can show that it is explicitly frame invariant by
considering (3.4):
vE 2
 
02 02 2 2 2 2
E − p c = γ (E − vp) − γ pc −
c
  2 2 
2 2 2 2 v E 2 2
=γ E −p c − +v p
c2
v2
 
2 2 2 2


=γ E −p c 1− 2
c
= E 2 − p2 c2
= m2 c4

We can actually represent the momentum and energy of a body by four-vector momentum.
Four vector is of the form:
 
µ E
p = , p, 0, 0
c

Again, we have left the zeroes in to demonstrate that the momentum can be in any di-
rection, and not just along the x-axis. As it contains the information for all the energy
and momentum of a system, four-vector momentum is conserved. This means that we
can equate four-vector momenta if the momenta are written in the same frame. Like the
interval s, the dot product of a four-vector with itself yields an invariant quantity:
E2
pµ ·pµ = − p2
c2
= m2 c2

This means that we can equate the dot product of the sum of four-vector momenta, even
if the momenta are written in dierent frames. Do not worry if this seems a little confus-
ing right now; there are plenty of examples to follow that will help elucidate these concepts.

56
Toby Adkins CP2

Like with the LT's, these results reduce to the Classical results for v  c. For example,
consider the kinetic energy:
KE = (γ − 1)mc2
1 v2
 
≈ 1+ − 1 mc2
2 c2
1
= mv 2
2
Some of you might be wondering at this point about how fast we need to be going for these
relativistic eects to matter? Well, if we again consider the kinetic energy:
KE = (γ − 1)mc2
1 v2 3 v2
 
≈ 1+ + − 1 mc2
2 c2 8 c4
1 3 v4
= mv 2 + m 2
2 8 c
For us to have a 1% error in kinetic energy, we require:
3 4
8β
0.01 = 1 2
2β
3
= β2
4
β ≈ 0.115

Thus, a particle has to be going at just over 10% of the speed of light for relativistic eects
to become noticeable.

3.4.1 Photons
So what happens when the particle that we are considering is a photon? By denition, a
photon has zero rest mass energy. This means that the square of the four-momentum for
any photon is always zero. Furthermore, this result leads to:
E hν h
p= = = (3.16)
c c λ
where h is Plank's Constant. This means that a photon still has momentum even though
it's rest mass is zero. This does not make sense classically, but if considered relativistically:
lim p = lim γmv
v→c v→c
= (∞)(o)c

which has a nite value.

3.4.2 Approximations
In relativity, there are a number of useful approximations that we can make to simplify
calculations.
• If a particle is very light, and E  mc2 , then by (3.15):

E ≈ pc

57
Toby Adkins CP2

• For a particle travelling very quickly such that v ≈ c, then:

s r
1−β 2
≈
1+β 1−β
v c
p ≈p
1 − β2 2(1 − β)

The wording of questions will often make it obvious when these are to be used. If you get
stuck on a question, look to see if any approximations such as these can be made.

3.4.3 Example Problems

Below are a range of problems covering dierent aspects of the material in this section.
They should serve as examples that can be referenced when working on harder or more
involved problems.
1. Let us examine the following question: can an isolated electron emit a photon?

Let the original electron have energy Ee , the photon energy Eγ and the energy of
the nal electron be Ee0 .
p
Ee 0 = me 2 c4 + pe 2 c2
Eγ = pγ c

Now considering the conservation of energy:

Eγ + Ee 0 = Ee
p
pγ c + me 2 c4 + pe 2 c2 = me c2

By conservation of momentum, pe = pγ . Hence we obtain:

q
Eγ + me 2 c4 + Eγ 2 = me c2

This relationship can only hold if Eγ = 0, meaning that an isolated electron cannot
emit a photon.
2. In a similar way to the previous problem, we will pose the question: Can the collision
of a positron and electron create a single photon?

Let us rst consider the four-vector momenta:

 
E1
p1µ = , p1 , 0, 0
c
 
µ E2
p2 = , −p2 , 0, 0
c
 
µ E3 E3
p3 = , , 0, 0
c c

By the conservation of four-vector momentum:

p1µ + p2µ = p3µ
p1µ = p3µ − p2µ

58
Toby Adkins CP2

Squaring both sides to obtain the invariant quantities:

E3 E2 E3
m2 c2 = m2 c2 + 2 p2 − 2
c c
E2
→ p2 =
c
However, as an electron is a massive particle - that is, it has rest mass energy - the
nal relation cannot hold. This means that we have shown that the proposed process
is not possible.
3. An electron of energy 9.0 GeV and a positron of energy E collide head-on to produce
a B 0 meson and an anti-B 0 meson, each with a mass of 5.3 Gev/c2 . What is the
threshold energy required to produce the B 0 meson pair?

Using four vector invariance across the laboratory (left hand side) and the centre of
mass (right hand side) frames:
(9 + E)2 − (pc)2 = (2mB c2 )2

In this case, E  me c2 . This means that we can use the rst approximation in
Section (3.4.2). Thus the net momentum is given by:
9 E
p= −
c c
Substituting this result in and re-arranging:
(9 + E)2 − (9 − E)2 = (2mB c2 )2
36E = 4(5.3)2
→ E ≈ 3.12 GeV

The B 0 mesons undergo decay with a mean proper life-time of 1.5 × 10−12 s. Assum-
ing that the particles are produced at threshold energy, what is the mean distance
that the B 0 mesons travel in the lab frame before they decay?

Consider the conservation of energy in the lab frame:

9 + E = 2γmB c2

We can substitute our result for E back in to nd the γ factor for the B 0 mesons.
This gives γ ≈ 1.14. From the lab. frame, the life-time of the particles appears much
longer; we thus have to consider the dilated life-time of the particles. Using (3.7):
τ 0 = γτ
= (1.5 × 10−12 )(1.14)
= 1.71 × 10−12 s

Finding the velocity:

1
γ=q
v2
1− c2
r
v 1
= 1− 2
c γ
≈ 0.485c

59
Toby Adkins CP2

Thus the mean distance they travel is given by:

< r > = (0.485 × 3 × 108 )(1.71 × 10−12 )
= 2.5 × 10−4 m

We can ask one further question (last one!): What is the probability that one of the
photons decays before reaching this mean distance?

As the particles decay, they must obey the population decay equation:
dN
= −λN
dt
This has solution:
N t
= e− τ
No
where No is the initial population. The associated that a particle survives/decays
over a time to are:
to
P (Survive) = e− τ (3.17)
− tτo
P (Decay) = 1 − e (3.18)
Using to = 1.71 × 10−12 (time to travel the mean distance) and τ = 1.71 × 10−12 (the
mean life-time), we obtain:
P (Decay) ≈ 63%

This example covers a lot of the basic concepts involved in relativistic collisions that
produce decaying particles.
4. A photon of unknown energy E collides head-on with a photon of energy ε for ε  E .
What is the minimum value of E required to produce an electron-positron pair?

Again, let us consider the four-vector momenta of the constituents.

 
E E
p1µ = , , 0, 0
c c
ε ε 
p2µ = , − , 0, 0
c c

We have left out the four-vector momenta for the electrons; we don't have to worry
about this as we can just consider them to be created in the zero-momentum (ZMF)/centre-
of-mass frame, and equate invariants as follows:
 2  2
p1µ + p2µ = peµ + peµ
4Eε
4me 2 c2 =
c2
me 2 c4
→E=
ε
What is the speed of the ZMF in this case?

60
Toby Adkins CP2

As the energy of the second photon is so small (ε), we can neglect its energy when
transforming to the ZMF. This means we can easily calculate the energy in the ZMF,
and thus nd it's speed:
s
1−β
EZM F = (E + ε)
1+β
r
2 1−β
2me c ≈ E
2
2
2me c2

→β =1−2
E

5. This is a problem treating the concept of "Compton Scattering". An electron of

energy E = hc λ collides with a stationary electron and is scattered at an angle θ to
the line of incidence. Find an expression for the new wavelength of the photon as a
function of θ.

As usual, let us write the four-vector momenta:

h
p1µ = (1, 1, 0, 0)
λ
mc2
 
µ
p2 = , 0, 0, 0
c
h
p1µ = 0 (1, cos θ, sin θ, 0)
λ
Again, we have left out the momenta of the nal state of the electron as it is incon-
sequential; we can get rid of it with an appropriate squaring of four-vector elements.
p1µ + p2µ = p3µ + p4µ
p1µ + p2µ − p3µ = p4µ

Squaring both sides and re-arranging:

p1µ 2 + p2µ 2 + p3µ 2 + 2p1µ · p2µ − 2p1µ · p3µ − 2p2µ · p3µ = m2 c2
2h2
 
2 2 h h
m c + 2mc − 0 − (1 − cos θ) = mc2 c2
λ λ λλ0

Thus, we obtain a nal result of:

h
λ0 = λ + (1 − cos θ) (3.19)
mc
This holds for any value of θ. For example, the wavelength of a back-scattered
electron is given by:
h
λ0 = λ +(1 − cos π)
mc
2h
=λ+
mc
2h
→ ∆λ =
mc

61
Toby Adkins CP2

6. Suppose that an electron of energy E collides obliquely with a stationary electron,

scattering both at an angle θ with respect to the line of incidence of the two particles.
Find an expression for θ in terms of E and Eo , the rest energy of the electron.

Before the collision:

 
E
p1µ= , p, 0, 0
c
 
µ Eo
p2 = , 0, 0, 0
c

After the collision, both of the electrons have to have the same energy by the con-
servation of energy for both bodies (they have the same mass). We can thus write:
E10 0
 
0
p3µ= , p cos θ, p sin θ, 0
c
 0 
µ E2 0 0
p4 = , p cos θ, −p sin θ, 0
c

By the conservation of momentum and energy:

p = 2p0 cos θ
p
p0 =
2 cos θ
2E 0 = E + Eo
E + Eo
E0 =
2
We can thus write the nal four-momenta as:
 
Eo + E p p
p µfinal = , , ± tan θ, 0
2c 2 2

Using the Lorentz invariant:

 2
2 4 E + Eo  p 2
m c = − (tan2 θ + 1)c2
2 2
 2
E + Eo  p 2
= − c2 sec2 θ
2 2
4m2 c4 = (E + Eo )2 − p2 c2 sec2 θ

Re-arranging for θ:
E 2 + 2EEo + Eo 2 − 4Eo 2
sec2 θ =
E 2 − Eo 2
(E + 3Eo )(E − Eo )
=
(E + Eo )(E − Eo )
r
E + Eo
→ cos θ =
E + 3Eo

62
Toby Adkins CP2

3.5 The Relativistic Doppler Eect

When sources of signals (such as light) move towards or away from an observer, the signal
is shifted based on the rate of separation of the observer and the source. This is known as
the Doppler Shift.

3.5.1 Derivation
Let a source emitting a signal at a period τ be moving away from an observer at a speed
v . Using the LT's, the rst pulse of the signal is at t = 0, x = 0, and the second is at
t = γτ, x = vγτ from the frame of reference of the observer. The observed period is thus
given by:
x
τ 0 = γτ +
c
vγτ
= γτ +
 c v
= γτ 1 +
c
s
1+β
→ τ0 = τ (3.20)
1−β
Thus, for a receding source, τ 0 > τ (known as a red-shift), and for an approaching source,
τ 0 < τ (known as a blue-shift). This can be used to determine the speed at which planets
are receding, for example.

3.5.2 Superluminal Motion

Superluminal motion is where the projected position of an object moving obliquely across
the eld of view onto the plane of the sky appears to move faster than the speed of light.
Consider an object emitted from A that moves with a relativistic speed v along a trajectory
passing through B, as shown in Figure (3.6). Let the observer be located at D.

Figure 3.6: Superluminal Motion

We want to calculate the apparent rate of change of AB'.

AB = βcδt
AC = βcδt cos θ
AD = cδt

63
Toby Adkins CP2

as the emitted photon travels at c. The two photons arriving are separated by:
DC = AD − AC
= cδt(1 − β cos θ)

The projected separation of the object is:

CB = βcδt sin θ

Thus, the apparent speed is given by:

CB
vo = ·c
DC
βcδt sin θ
=
cδt(1 − β cos θ)
β sin θ
→ vo =
1 − β cos θ
Consider an object travelling at v = 0.9c at an angle θ = 10◦ .
(0.9)(cos(10◦ ))
vo =
1 − (0.9)(cos(10◦ ))
≈ 1.37 c

Thus, the projection of the object appears to move faster than c. However, this evidently
does not violate causality as no information is actually transmitted at this speed.

64