7 CIRCULAR MOTION 7.

3 Centripetal acceleration

weight v

cable
m T

Figure 59: Weight on the end of a cable.

Suppose that a weight, of mass m, is attached to the end of a cable, of length

r, and whirled around such that the weight executes a horizontal circle, radius r,
with uniform tangential velocity v. As we have just learned, the weight is subject
to a centripetal acceleration of magnitude v2 /r. Hence, the weight experiences a
centripetal force
m v2
f= . (7.16)
r
What provides this force? Well, in the present example, the force is provided by
the tension T in the cable. Hence, T = m v2 /r.

Suppose that the cable is such that it snaps whenever the tension in it exceeds
a certain critical value Tmax . It follows that there is a maximum velocity with
which the weight can be whirled around: namely,


 r Tmax

vmax = . (7.17)
m
If v exceeds vmax then the cable will break. As soon as the cable snaps, the weight
will cease to be subject to a centripetal force, so it will fly off—with velocity v max —
along the straight-line which is tangential to the circular orbit it was previously
executing.

140
7 CIRCULAR MOTION 7.4 The conical pendulum

θ
h l

T
r
m

mg

Figure 60: A conical pendulum.

7.4 The conical pendulum

Suppose that an object, mass m, is attached to the end of a light inextensible

string whose other end is attached to a rigid beam. Suppose, further, that the
object is given an initial horizontal velocity such that it executes a horizontal
circular orbit of radius r with angular velocity ω. See Fig. 60. Let h be the
vertical distance between the beam and the plane of the circular orbit, and let θ
be the angle subtended by the string with the downward vertical.

The object is subject to two forces: the gravitational force m g which acts ver-
tically downwards, and the tension force T which acts upwards along the string.
The tension force can be resolved into a component T cos θ which acts vertically
upwards, and a component T sin θ which acts towards the centre of the circle.
Force balance in the vertical direction yields

T cos θ = m g. (7.18)

In other words, the vertical component of the tension force balances the weight
of the object.

Since the object is executing a circular orbit, radius r, with angular velocity ω,
it experiences a centripetal acceleration ω2 r. Hence, it is subject to a centripetal
force m ω2 r. This force is provided by the component of the string tension which

141
7 CIRCULAR MOTION 7.5 Non-uniform circular motion

acts towards the centre of the circle. In other words,

T sin θ = m ω2 r. (7.19)

Taking the ratio of Eqs. (7.18) and (7.19), we obtain

ω2 r
tan θ = . (7.20)
g
However, by simple trigonometry,
r
tan θ = . (7.21)
h
Hence, we find 
g
ω= . (7.22)
h
Note that if l is the length of the string then h = l cos θ. It follows that

g
ω= . (7.23)
l cos θ

For instance, if the length of the string is l = 0.2 m and the conical angle is
θ = 30◦ then the angular velocity of rotation is given by


 9.81

ω= = 7.526 rad./s. (7.24)
0.2 × cos 30◦
This translates to a rotation frequency in cycles per second of
ω
f= = 1.20 Hz. (7.25)
2π

7.5 Non-uniform circular motion

Consider an object which executes non-uniform circular motion, as shown in

Fig. 61. Suppose that the motion is confined to a 2-dimensional plane. We can
specify the instantaneous position of the object in terms of its polar coordinates r

142
7 CIRCULAR MOTION 7.5 Non-uniform circular motion

eθ
er

Figure 61: Polar coordinates.

and θ. Here, r is the radial distance of the object from the origin, whereas θ is
the angular bearing of the object from the origin, measured with respect to some
arbitrarily chosen direction. We imagine that both r and θ are changing in time.
As an example of non-uniform circular motion, consider the motion of the Earth
around the Sun. Suppose that the origin of our coordinate system corresponds to
the position of the Sun. As the Earth rotates, its angular bearing θ, relative to the
Sun, obviously changes in time. However, since the Earth’s orbit is slightly ellipti-
cal, its radial distance r from the Sun also varies in time. Moreover, as the Earth
moves closer to the Sun, its rate of rotation speeds up, and vice versa. Hence, the
rate of change of θ with time is non-uniform.

Let us define two unit vectors, er and eθ . Incidentally, a unit vector simply a
vector whose length is unity. As shown in Fig. 61, the radial unit vector er always
points from the origin to the instantaneous position of the object. Moreover, the
tangential unit vector eθ is always normal to er , in the direction of increasing θ.
The position vector r of the object can be written

r = r er . (7.26)

In other words, vector r points in the same direction as the radial unit vector e r ,
and is of length r. We can write the object’s velocity in the form

v = ṙ = vr er + vθ eθ , (7.27)

143
7 CIRCULAR MOTION 7.5 Non-uniform circular motion

whereas the acceleration is written

a = v̇ = ar er + aθ eθ . (7.28)
Here, vr is termed the object’s radial velocity, whilst vθ is termed the tangential ve-
locity. Likewise, ar is the radial acceleration, and aθ is the tangential acceleration.
But, how do we express these quantities in terms of the object’s polar coordinates
r and θ? It turns out that this is a far from straightforward task. For instance, if
we simply differentiate Eq. (7.26) with respect to time, we obtain
v = ṙ er + r ėr , (7.29)
where ėr is the time derivative of the radial unit vector—this quantity is non-
zero because er changes direction as the object moves. Unfortunately, it is not
entirely clear how to evaluate ėr . In the following, we outline a famous trick for
calculating vr , vθ , etc. without ever having to evaluate the time derivatives of the
unit vectors er and eθ .

Consider a general complex number,

z = x + i y, (7.30)
where x and y are real, and i is the square root of −1 (i.e., i2 = −1). Here, x is
the real part of z, whereas y is the imaginary part. We can visualize z as a point in
the so-called complex plane: i.e., a 2-dimensional plane in which the real parts of
complex numbers are plotted along one Cartesian axis, whereas the correspond-
ing imaginary parts are plotted along the other axis. Thus, the coordinates of z
in the complex plane are simply (x, y). See Fig. 62. In other words, we can use
a complex number to represent a position vector in a 2-dimensional plane. Note
that the length of the vector is equal to the modulus of the corresponding complex
number. Incidentally, the modulus of z = x + i y is defined

|z| = x2 + y2 . (7.31)

Consider the complex number e i θ , where θ is real. A famous result in complex

analysis—known as de Moivre’s theorem—allows us to split this number into its
real and imaginary components:
e i θ = cos θ + i sin θ. (7.32)

144
7 CIRCULAR MOTION 7.5 Non-uniform circular motion

Im(z)
z

x Re(z)

Figure 62: Representation of a complex number in the complex plane.

Now, as we have just discussed, we can think of e i θ as representing a vector in

the complex plane: the real and imaginary parts of e i θ form the coordinates of
the head of the vector, whereas the tail of the vector corresponds to the origin.
What are the properties of this vector? Well, the length of the vector is given by
  
 i θ
e  = cos2 θ + sin2 θ = 1. (7.33)
In other words, e i θ represents a unit vector. In fact, it is clear from Fig. 63 that e i θ
represents the radial unit vector er for an object whose angular polar coordinate
(measured anti-clockwise from the real axis) is θ. Can we also find a complex
representation of the corresponding tangential unit vector eθ ? Actually, we can.
The complex number i e i θ can be written
i e i θ = − sin θ + i cos θ. (7.34)
Here, we have just multiplied Eq. (7.32) by i, making use of the fact that i2 = −1.
This number again represents a unit vector, since
  
 i θ
i e  = sin2 θ + cos2 θ = 1. (7.35)
Moreover, as is clear from Fig. 63, this vector is normal to er , in the direction of
increasing θ. In other words, i e i θ represents the tangential unit vector eθ .

Consider an object executing non-uniform circular motion in the complex

plane. By analogy with Eq. (7.26), we can represent the instantaneous position

145
7 CIRCULAR MOTION 7.5 Non-uniform circular motion

i ei θ

Im(z)
ei θ

eθ er

cos θ

sin θ
θ
- sin θ cos θ Re(z)

Figure 63: Representation of the unit vectors er and eθ in the complex plane.

vector of this object via the complex number

z = r e i θ. (7.36)
Here, r(t) is the object’s radial distance from the origin, whereas θ(t) is its angu-
lar bearing relative to the real axis. Note that, in the above formula, we are using
e i θ to represent the radial unit vector er . Now, if z represents the position vector
of the object, then ż = dz/dt must represent the object’s velocity vector. Differ-
entiating Eq. (7.36) with respect to time, using the standard rules of calculus, we
obtain
ż = ṙ e i θ + r θ̇ i e i θ . (7.37)
Comparing with Eq. (7.27), recalling that e i θ represents er and i e i θ represents
eθ , we obtain
vr = ṙ, (7.38)
vθ = r θ̇ = r ω, (7.39)
where ω = dθ/dt is the object’s instantaneous angular velocity. Thus, as desired,
we have obtained expressions for the radial and tangential velocities of the object
in terms of its polar coordinates, r and θ. We can go further. Let us differentiate
ż with respect to time, in order to obtain a complex number representing the
object’s vector acceleration. Again, using the standard rules of calculus, we obtain
z̈ = (r̈ − r θ̇2 ) e i θ + (r θ̈ + 2 ṙ θ̇) i e i θ . (7.40)

146
7 CIRCULAR MOTION 7.5 Non-uniform circular motion

Comparing with Eq. (7.28), recalling that e i θ represents er and i e i θ represents

eθ , we obtain

ar = r̈ − r θ̇2 = r̈ − r ω2 , (7.41)
aθ = r θ̈ + 2 ṙ θ̇ = r ω̇ + 2 ṙ ω. (7.42)

Thus, we now have expressions for the object’s radial and tangential accelerations
in terms of r and θ. The beauty of this derivation is that the complex analysis
has automatically taken care of the fact that the unit vectors e r and eθ change
direction as the object moves.

Let us now consider the commonly occurring special case in which an object
executes a circular orbit at fixed radius, but varying angular velocity. Since the
radius is fixed, it follows that ṙ = r̈ = 0. According to Eqs. (7.38) and (7.39), the
radial velocity of the object is zero, and the tangential velocity takes the form

vθ = r ω. (7.43)

Note that the above equation is exactly the same as Eq. (7.6)—the only difference
is that we have now proved that this relation holds for non-uniform, as well as
uniform, circular motion. According to Eq. (7.41), the radial acceleration is given
by
ar = −r ω2 . (7.44)
The minus sign indicates that this acceleration is directed towards the centre
of the circle. Of course, the above equation is equivalent to Eq. (7.15)—the only
difference is that we have now proved that this relation holds for non-uniform, as
well as uniform, circular motion. Finally, according to Eq. (7.42), the tangential
acceleration takes the form
aθ = r ω̇. (7.45)
The existence of a non-zero tangential acceleration (in the former case) is the
one difference between non-uniform and uniform circular motion (at constant
radius).

147
7 CIRCULAR MOTION 7.6 The vertical pendulum

v’ θ r cos θ
T
r
B
mg cos θ
mg v
A

mg

Figure 64: Motion in a vertical circle.

7.6 The vertical pendulum

Let us now examine an example of non-uniform circular motion. Suppose that

an object of mass m is attached to the end of a light rigid rod, or light string, of
length r. The other end of the rod, or string, is attached to a stationary pivot in
such a manner that the object is free to execute a vertical circle about this pivot.
Let θ measure the angular position of the object, measured with respect to the
downward vertical. Let v be the velocity of the object at θ = 0◦ . How large do we
have to make v in order for the object to execute a complete vertical circle?

Consider Fig. 64. Suppose that the object moves from point A, where its
tangential velocity is v, to point B, where its tangential velocity is v  . Let us,
first of all, obtain the relationship between v and v  . This is most easily achieved
by considering energy conservation. At point A, the object is situated a vertical
distance r below the pivot, whereas at point B the vertical distance below the
pivot has been reduced to r cos θ. Hence, in moving from A to B the object gains
potential energy m g r (1 − cos θ). This gain in potential energy must be offset by
a corresponding loss in kinetic energy. Thus,
1 1 2
m v2 − m v  = m g r (1 − cos θ), (7.46)
2 2
148
7 CIRCULAR MOTION 7.6 The vertical pendulum

which reduces to
2
v  = v2 − 2 r g (1 − cos θ). (7.47)

Let us now examine the radial acceleration of the object at point B. The ra-
dial forces acting on the object are the tension T in the rod, or string, which
acts towards the centre of the circle, and the component m g cos θ of the object’s
weight, which acts away from the centre of the circle. Since the object is execut-
ing circular motion with instantaneous tangential velocity v  , it must experience
an instantaneous acceleration v  2 /r towards the centre of the circle. Hence, New-
ton’s second law of motion yields
m v 2
= T − m g cos θ. (7.48)
r
Equations (7.47) and (7.48) can be combined to give
m v2
T= + m g (3 cos θ − 2). (7.49)
r
Suppose that the object is, in fact, attached to the end of a piece of string,
rather than a rigid rod. One important property of strings is that, unlike rigid
rods, they cannot support negative tensions. In other words, a string can only pull
objects attached to its two ends together—it cannot push them apart. Another
way of putting this is that if the tension in a string ever becomes negative then
the string will become slack and collapse. Clearly, if our object is to execute
a full vertical circle then then tension T in the string must remain positive for
all values of θ. It is clear from Eq. (7.49) that the tension attains its minimum
value when θ = 180◦ (at which point cos θ = −1). This is hardly surprising, since
θ = 180◦ corresponds to the point at which the object attains its maximum height,
and, therefore, its minimum tangential velocity. It is certainly the case that if the
string tension is positive at this point then it must be positive at all other points.
Now, the tension at θ = 180◦ is given by
m v2
T0 = − 5 m g. (7.50)
r
Hence, the condition for the object to execute a complete vertical circle without
the string becoming slack is T0 > 0, or
v2 > 5 r g. (7.51)

149
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

Note that this condition is independent of the mass of the object.

Suppose that the object is attached to the end of a rigid rod, instead of a piece
of string. There is now no constraint on the tension, since a rigid rod can quite
easily support a negative tension (i.e., it can push, as well as pull, on objects
attached to its two ends). However, in order for the object to execute a complete
vertical circle the square of its tangential velocity v  2 must remain positive at all
values of θ. It is clear from Eq. (7.47) that v  2 attains its minimum value when
θ = 180◦ . This is, again, hardly surprising. Thus, if v  2 is positive at this point
then it must be positive at all other points. Now, the expression for v  2 at θ = 180◦
is
2
(v  )0 = v2 − 4 r g. (7.52)
Hence, the condition for the object to execute a complete vertical circle is (v  2 )0 >
0, or
v2 > 4 r g. (7.53)
Note that this condition is slightly easier to satisfy than the condition (7.51). In
other words, it is slightly easier to cause an object attached to the end of a rigid
rod to execute a vertical circle than it is to cause an object attached to the end of
a string to execute the same circle. The reason for this is that the rigidity of the
rod helps support the object when it is situated above the pivot point.

7.7 Motion on curved surfaces

Consider a smooth rigid vertical hoop of internal radius r, as shown in Fig. 65.
Suppose that an object of mass m slides without friction around the inside of this
hoop. What is the motion of this object? Is it possible for the object to execute a
complete vertical circle?

Suppose that the object moves from point A to point B in Fig. 65. In doing so,
it gains potential energy m g r (1 − cos θ), where θ is the angular coordinate of
the object measured with respect to the downward vertical. This gain in potential
energy must be offset by a corresponding loss in kinetic energy. Thus,
1 1 2
m v2 − m v  = m g r (1 − cos θ), (7.54)
2 2
150
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

v’ θ r cos θ
R
r
B
mg cos θ
mg v
A

mg

Figure 65: Motion on the inside of a vertical hoop.

which reduces to
2
v  = v2 − 2 r g (1 − cos θ). (7.55)
Here, v is the velocity at point A (θ = 0◦ ), and v  is the velocity at point B
(θ = θ◦ ).

Let us now examine the radial acceleration of the object at point B. The radial
forces acting on the object are the reaction R of the vertical hoop, which acts
towards the centre of the hoop, and the component m g cos θ of the object’s
weight, which acts away from the centre of the hoop. Since the object is executing
circular motion with instantaneous tangential velocity v  , it must experience an
instantaneous acceleration v  2 /r towards the centre of the hoop. Hence, Newton’s
second law of motion yields

m v 2
= R − m g cos θ. (7.56)
r

Note, however, that there is a constraint on the reaction R that the hoop can
exert on the object. This reaction must always be positive. In other words, the
hoop can push the object away from itself, but it can never pull it towards itself.
Another way of putting this is that if the reaction ever becomes negative then

151
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

the object will fly off the surface of the hoop, since it is no longer being pressed
into this surface. It should be clear, by now, that the problem we are considering
is exactly analogous to the earlier problem of an object attached to the end of a
piece of string which is executing a vertical circle, with the reaction R of the hoop
playing the role of the tension T in the string.

Let us imagine that the hoop under consideration is a “loop the loop” segment
in a fairground roller-coaster. The object sliding around the inside of the loop
then becomes the roller-coaster train. Suppose that the fairground operator can
vary the velocity v with which the train is sent into the bottom of the loop (i.e.,
the velocity at θ = 0◦ ). What is the safe range of v? Now, if the train starts at
θ = 0◦ with velocity v then there are only three possible outcomes. Firstly, the
train can execute a complete circuit of the loop. Secondly, the train can slide
part way up the loop, come to a halt, reverse direction, and then slide back down
again. Thirdly, the train can slide part way up the loop, but then fall off the loop.
Obviously, it is the third possibility that the fairground operator would wish to
guard against.

Using the analogy between this problem and the problem of a mass on the end
of a piece of string executing a vertical circle, the condition for the roller-coaster
train to execute a complete circuit is

v2 > 5 r g. (7.57)

Note, interestingly enough, that this condition is independent of the mass of the
train.

Equation (7.56) yields

2 rR
v = − r g cos θ. (7.58)
m
Now, the condition for the train to reverse direction without falling off the loop
is v  2 = 0 with R > 0. Thus, the train reverses direction when

R = m g cos θ. (7.59)

Note that this equation can only be satisfied for positive R when cos θ > 0. In
other words, the train can only turn around without falling off the loop if the

152
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

mg cos θ
m
R

r cos θ r
θ v
mg

Figure 66: A skier on a hemispherical mountain.

turning point lies in the lower half of the loop (i.e., −90◦ < θ < 90◦ ). The
condition for the train to fall off the loop is
2
v  = −r g cos θ. (7.60)
Note that this equation can only be satisfied for positive v  2 when cos θ < 0. In
other words, the train can only fall off the loop when it is situated in the upper
half of the loop. It is fairly clear that if the train’s initial velocity is not sufficiently
large for it to execute a complete circuit of the loop, and not sufficiently small
for it to turn around before entering the upper half of the loop, then it must
inevitably fall off the loop somewhere in the loop’s upper half. The critical value
of v2 above which the train executes a complete circuit is 5 r g [see Eq. (7.57)].
The critical value of v2 at which the train just turns around before entering the
upper half of the loop is 2 r g [this is obtained from Eq. (7.55) by setting v  = 0
and θ = 90◦ ]. Hence, the dangerous range of v2 is
2 r g < v2 < 5 r g. (7.61)
For v2 < 2 r g, the train turns around in the lower half of the loop. For v2 > 5 r g,
the train executes a complete circuit around the loop. However, for 2 r g < v2 <
5 r g, the train falls off the loop somewhere in its upper half.

Consider a skier of mass m skiing down a hemispherical mountain of radius r,

as shown in Fig. 66. Let θ be the angular coordinate of the skier, measured with
respect to the upward vertical. Suppose that the skier starts at rest (v = 0) on
top of the mountain (θ = 0◦ ), and slides down the mountain without friction. At
what point does the skier fly off the surface of the mountain?

153
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

Suppose that the skier has reached angular coordinate θ. At this stage, the
skier has fallen though a height r (1 − cos θ). Thus, the tangential velocity v of
the skier is given by energy conservation:
1
m v2 = m g r (1 − cos θ). (7.62)
2
Let us now consider the skier’s radial acceleration. The radial forces acting on the
skier are the reaction R exerted by the mountain, which acts radially outwards,
and the component of the skier’s weight m g cos θ, which acts radially inwards.
Since the skier is executing circular motion, radius r, with instantaneous tangen-
tial velocity v, he/she experiences an instantaneous inward radial acceleration
v2 /r. Hence, Newton’s second law of motion yields
v2
m = m g cos θ − R. (7.63)
r

Equations (7.62) and (7.63) can be combined to give

R = m g (3 cos θ − 2). (7.64)

As before, the reaction R is constrained to be positive—the mountain can push

outward on the skier, but it cannot pull the skier inward. In fact, as soon as the
reaction becomes negative, the skier flies of the surface of the mountain. This
occurs when cos θ0 = 2/3, or θ0 = 48.19◦ . The height through which the skier
falls before becoming a ski-jumper is h = r (1 − cos θ0 ) = a/3.

Worked example 7.1: A banked curve

Question: Civil engineers generally bank curves on roads in such a manner that
a car going around the curve at the recommended speed does not have to rely
on friction between its tires and the road surface in order to round the curve.
Suppose that the radius of curvature of a given curve is r = 60 m, and that
the recommended speed is v = 40 km/h. At what angle θ should the curve be
banked?

154
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

R car

centre of curvature

mg
θ

banked curve

Answer: Consider a car of mass m going around the curve. The car’s weight, m g,
acts vertically downwards. The road surface exerts an upward normal reaction R
on the car. The vertical component of the reaction must balance the downward
weight of the car, so
R cos θ = m g.
The horizontal component of the reaction, R sin θ, acts towards the centre of
curvature of the road. This component provides the force m v2 /r towards the
centre of the curvature which the car experiences as it rounds the curve. In other
words,
v2
R sin θ = m ,
r
which yields
v2
tan θ = ,
rg
or ⎛ ⎞
2
v
θ = tan−1 ⎝ ⎠ .
rg
Hence, ⎛ ⎞
−1 ⎝ (40 × 1000/3600)2 ⎠
θ = tan = 11.8◦ .
60 × 9.81
Note that if the car attempts to round the curve at the wrong speed then m v 2 /r =
m g tan θ, and the difference has to be made up by a sideways friction force
exerted between the car’s tires and the road surface. Unfortunately, this does not
always work—especially if the road surface is wet!

155
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

Worked example 7.2: Circular race track

Question: A car of mass m = 2000 kg travels around a flat circular race track of
radius r = 85 m. The car starts at rest, and its speed increases at the constant rate
aθ = 0.6 m/s. What is the speed of the car at the point when its centripetal and
tangential accelerations are equal?

Answer: The tangential acceleration of the car is aθ = 0.6 m/s. When the car
travels with tangential velocity v its centripetal acceleration is a r = v2 /r. Hence,
ar = aθ when
v2
= aθ ,
r
or √
√
v = r aθ = 85 × 0.6 = 7.14 m/s.

Worked example 7.3: Amusement park ride

Question: An amusement park ride consists of a vertical cylinder that spins about
a vertical axis. When the cylinder spins sufficiently fast, any person inside it is
held up against the wall. Suppose that the coefficient of static friction between
a typical person and the wall is μ = 0.25. Let the mass of an typical person be
m = 60 kg, and let r = 7 m be the radius of the cylinder. Find the critical angular
velocity of the cylinder above which a typical person will not slide down the
wall. How many revolutions per second is the cylinder executing at this critical
velocity?

Answer: In the vertical direction, the person is subject to a downward force m g

due to gravity, and a maximum upward force f = μ R due to friction with the
wall. Here, R is the normal reaction between the person and the wall. In order
for the person not to slide down the wall, we require f > m g. Hence, the critical
case corresponds to
f = μ R = m g.

In the radial direction, the person is subject to a single force: namely, the

156
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

mg f

reaction R due to the wall, which acts radially inwards. If the cylinder (and,
hence, the person) rotates with angular velocity ω, then this force must provided
the acceleration r ω2 towards the axis of rotation. Hence,

R = m r ω2 .

It follows that, in the critical case,

 
 
 g  9.81

ω=  = = 2.37 rad/s.
μr 0.25 × 7
The corresponding number of revolutions per second is
ω 2.37
f= = = 0.38 Hz.
2 π 2 × 3.1415

Worked example 7.4: Aerobatic maneuver

Question: A stunt pilot experiences weightlessness momentarily at the top of

a “loop the loop” maneuver. Given that the speed of the stunt plane is v =
500 km/h, what is the radius r of the loop?

Answer: Let m be the mass of the pilot. Consider the radial acceleration of the
pilot at the top of the loop. The pilot is subject to two radial forces: the gravita-
tional force m g, which acts towards the centre of the loop, and the reaction force
R, due to the plane, which acts away from the centre of the loop. Since the pilot

157
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

mg

experiences an acceleration v2 /r towards the centre of the loop, Newton’s second

law of motion yields
v2
m = m g − R.
r
Now, the reaction R is equivalent to the apparent weight of the pilot. In partic-
ular, if the pilot is “weightless” then he/she exerts no force on the plane, and,
therefore, the plane exerts no reaction force on the pilot. Hence, if the pilot is
weightless at the top of the loop then R = 0, giving
v2 (500 × 1000/3600)2
r= = = 1.97 km.
g 9.81

Worked example 7.5: Ballistic pendulum

Question: A bullet of mass m = 10 g strikes a pendulum bob of mass M = 1.3 kg

horizontally with speed v, and then becomes embedded in the bob. The bob is
initially at rest, and is suspended by a stiff rod of length l = 0.6 m and negligible
mass. The bob is free to rotate in the vertical direction. What is the minimum
value of v which causes the bob to execute a complete vertical circle? How does
the answer change if the bob is suspended from a light flexible rod (of the same
length), instead of a stiff rod?

Answer: When the bullet strikes the bob, and then sticks to it, the bullet and bob
move off with a velocity v  which is given by momentum conservation:
m v = (M + m) v  .

158
7 CIRCULAR MOTION 7.7 Motion on curved surfaces

Hence,
mv
v = .
M+m

Consider the case where the bob is suspended by a rigid rod. If the bob and
bullet only just manage to execute a vertical loop, then their initial kinetic energy
(1/2) (M + m) v  2 must only just be sufficient to lift them from the bottom to the
top of the loop—a distance 2 l. Hence, in this critical case, energy conservation
yields
1 2
(M + m) v  = (M + m) 2 g l,
2
which implies
2
v  = 4 g l,
or √ √
(M + m) 4gl 1.31 × 4 × 9.81 × 0.6
v= = = 635.6 m/s.
m 0.01

Consider the case where the bob is suspended by a flexible rod. The velocity v 
of the bob and bullet at the top of the loop is obtained from energy conservation:
1 2 1 2
(M + m) v  = (M + m) v  − (M + m) 2 g l.
2 2
If the bob and bullet only just manage to execute a vertical loop, then the tension
in the rod is zero at the top of the loop. Hence, the acceleration due to gravity
g must account exactly for the required acceleration v  2 /l towards the centre of
the loop:
v  2
= g.
l
It follows that, in this critical case,
2
v  = 5 g l,

or √ √
(M + m) 5gl 1.31 × 5 × 9.81 × 0.6
v= = = 710.7 m/s.
m 0.01

159
8 ROTATIONAL MOTION

8 Rotational motion

8.1 Introduction

Up to now, we have only analyzed the dynamics of point masses (i.e., objects
whose spatial extent is either negligible or plays no role in their motion). Let us
now broaden our approach in order to take extended objects into account. Now,
the only type of motion which a point mass object can exhibit is translational mo-
tion: i.e., motion by which the object moves from one point in space to another.
However, an extended object can exhibit another, quite distinct, type of motion
by which it remains located (more or less) at the same spatial position, but con-
stantly changes its orientation with respect to other fixed points in space. This
new type of motion is called rotation. Let us investigate rotational motion.

8.2 Rigid body rotation

Consider a rigid body executing pure rotational motion (i.e., rotational motion
which has no translational component). It is possible to define an axis of rotation
(which, for the sake of simplicity, is assumed to pass through the body)—this axis
corresponds to the straight-line which is the locus of all points inside the body
which remain stationary as the body rotates. A general point located inside the
body executes circular motion which is centred on the rotation axis, and orien-
tated in the plane perpendicular to this axis. In the following, we tacitly assume
that the axis of rotation remains fixed.

Figure 67 shows a typical rigidly rotating body. The axis of rotation is the line
AB. A general point P lying within the body executes a circular orbit, centred
on AB, in the plane perpendicular to AB. Let the line QP be a radius of this
orbit which links the axis of rotation to the instantaneous position of P at time
t. Obviously, this implies that QP is normal to AB. Suppose that at time t + δt
point P has moved to P  , and the radius QP has rotated through an angle δφ.

160
8 ROTATIONAL MOTION 8.2 Rigid body rotation

A
rigid body

P’ δφ
Q P

B axis of rotation

Figure 67: Rigid body rotation.

The instantaneous angular velocity of the body ω(t) is defined

δφ dφ
ω = lim = . (8.1)
δt→0 δt dt
Note that if the body is indeed rotating rigidly, then the calculated value of ω
should be the same for all possible points P lying within the body (except for
those points lying exactly on the axis of rotation, for which ω is ill-defined). The
rotation speed v of point P is related to the angular velocity ω of the body via
v = σ ω, (8.2)
where σ is the perpendicular distance from the axis of rotation to point P. Thus, in
a rigidly rotating body, the rotation speed increases linearly with (perpendicular)
distance from the axis of rotation.

It is helpful to introduce the angular acceleration α(t) of a rigidly rotating

body: this quantity is defined as the time derivative of the angular velocity. Thus,
dω d2 φ
α= = , (8.3)
dt dt2
where φ is the angular coordinate of some arbitrarily chosen point reference
within the body, measured with respect to the rotation axis. Note that angular

161
8 ROTATIONAL MOTION 8.3 Is rotation a vector?

velocities are conventionally measured in radians per second, whereas angular

accelerations are measured in radians per second squared.

For a body rotating with constant angular velocity, ω, the angular acceleration
is zero, and the rotation angle φ increases linearly with time:
φ(t) = φ0 + ω t, (8.4)
where φ0 = φ(t = 0). Likewise, for a body rotating with constant angular accel-
eration, α, the angular velocity increases linearly with time, so that
ω(t) = ω0 + α t, (8.5)
and the rotation angle satisfies
1 2
φ(t) = φ0 + ω0 t + αt . (8.6)
2
Here, ω0 = ω(t = 0). Note that there is a clear analogy between the above equa-
tions, and the equations of rectilinear motion at constant acceleration introduced
in Sect. 2.6—rotation angle plays the role of displacement, angular velocity plays
the role of (regular) velocity, and angular acceleration plays the role of (regular)
acceleration.

8.3 Is rotation a vector?

Consider a rigid body which rotates through an angle φ about a given axis. It
is tempting to try to define a rotation “vector” φ which describes this motion.
For example, suppose that φ is defined as the “vector” whose magnitude is the
angle of rotation, φ, and whose direction runs parallel to the axis of rotation.
Unfortunately, this definition is ambiguous, since there are two possible directions
which run parallel to the rotation axis. However, we can resolve this problem by
adopting the following convention—the rotation “vector” runs parallel to the axis
of rotation in the sense indicated by the thumb of the right-hand, when the fingers
of this hand circulate around the axis in the direction of rotation. This convention
is known as the right-hand grip rule. The right-hand grip rule is illustrated in
Fig. 68.

162
8 ROTATIONAL MOTION 8.3 Is rotation a vector?

direction of rotation vector

sense of rotation

Figure 68: The right-hand grip rule.

The rotation “vector” φ now has a well-defined magnitude and direction. But,
is this quantity really a vector? This may seem like a strange question to ask,
but it turns out that not all quantities which have well-defined magnitudes and
directions are necessarily vectors. Let us review some properties of vectors. If a
and b are two general vectors, then it is certainly the case that

a + b = b + a. (8.7)

In other words, the addition of vectors is necessarily commutative (i.e., it is in-

dependent of the order of addition). Is this true for “vector” rotations, as we
have just defined them? Figure 69 shows the effect of applying two successive
90◦ rotations—one about the x-axis, and the other about the z-axis—to a six-
sided die. In the left-hand case, the z-rotation is applied before the x-rotation,
and vice versa in the right-hand case. It can be seen that the die ends up in two
completely different states. Clearly, the z-rotation plus the x-rotation does not
equal the x-rotation plus the z-rotation. This non-commutative algebra cannot be
represented by vectors. We conclude that, although rotations have well-defined
magnitudes and directions, they are not, in general, vector quantities.

There is a direct analogy between rotation and motion over the Earth’s surface.
After all, the motion of a pointer along the Earth’s equator from longitude 0 ◦ W to

163
8 ROTATIONAL MOTION 8.3 Is rotation a vector?

x
y
z-axis x-axis

x-axis z-axis

Figure 69: The addition of rotation is non-commutative.

164
8 ROTATIONAL MOTION 8.3 Is rotation a vector?

longitude 90◦ W could just as well be achieved by keeping the pointer fixed and ro-
tating the Earth through 90◦ about a North-South axis. The non-commutative na-
ture of rotation “vectors” is a direct consequence of the non-planar (i.e., curved)
nature of the Earth’s surface. For instance, suppose we start off at (0◦ N, 0◦ W),
which is just off the Atlantic coast of equatorial Africa, and rotate 90 ◦ northwards
and then 90◦ eastwards. We end up at (0◦ N, 90◦ E), which is in the middle of the
Indian Ocean. However, if we start at the same point, and rotate 90 ◦ eastwards
and then 90◦ northwards, we end up at the North pole. Hence, large rotations
over the Earth’s surface do not commute. Let us now repeat this experiment
on a far smaller scale. Suppose that we walk 10 m northwards and then 10 m
eastwards. Next, suppose that—starting from the same initial position—we walk
10 m eastwards and then 10 m northwards. In this case, few people would need
much convincing that the two end points are essentially identical. The crucial
point is that for sufficiently small displacements the Earth’s surface is approxi-
mately planar, and vector displacements on a plane surface commute with one
another. This observation immediately suggests that rotation “vectors” which cor-
respond to rotations through small angles must also commute with one another.
In other words, although the quantity φ, defined above, is not a true vector, the
infinitesimal quantity δφ, which is defined in a similar manner but corresponds
to a rotation through an infinitesimal angle δφ, is a perfectly good vector.

We have just established that it is possible to define a true vector δφ which

describes a rotation through a small angle δφ about a fixed axis. But, how is this
definition useful? Well, suppose that vector δφ describes the small rotation that
a given object executes in the infinitesimal time interval between t and t + δt. We
can then define the quantity
δφ dφ
ω = lim = . (8.8)
δt→0 δt dt
This quantity is clearly a true vector, since it is simply the ratio of a true vector
and a scalar. Of course, ω represents an angular velocity vector. The magnitude of
this vector, ω, specifies the instantaneous angular velocity of the object, whereas
the direction of the vector indicates the axis of rotation. The sense of rotation
is given by the right-hand grip rule: if the thumb of the right-hand points along
the direction of the vector, then the fingers of the right-hand indicate the sense of

165
8 ROTATIONAL MOTION 8.4 The vector product

rotation. We conclude that, although rotation can only be thought of as a vector

quantity under certain very special circumstances, we can safely treat angular
velocity as a vector quantity under all circumstances.

Suppose, for example, that a rigid body rotates at constant angular velocity
ω1 . Let us now combine this motion with rotation about a different axis at con-
stant angular velocity ω2 . What is the subsequent motion of the body? Since we
know that angular velocity is a vector, we can be certain that the combined mo-
tion simply corresponds to rotation about a third axis at constant angular velocity

ω3 = ω1 + ω2 , (8.9)

where the sum is performed according to the standard rules of vector addition.
[Note, however, the following important proviso. In order for Eq. (8.9) to be
valid, the rotation axes corresponding to ω1 and ω2 must cross at a certain
point—the rotation axis corresponding to ω3 then passes through this point.]
Moreover, a constant angular velocity

^ + ωy y
ω = ωx x ^ + ωz ^z (8.10)

can be thought of as representing rotation about the x-axis at angular velocity ω x ,

combined with rotation about the y-axis at angular velocity ω y , combined with
rotation about the z-axis at angular velocity ωz . [There is, again, a proviso—
namely, that the rotation axis corresponding to ω must pass through the origin.
Of course, we can always shift the origin such that this is the case.] Clearly, the
knowledge that angular velocity is vector quantity can be extremely useful.

8.4 The vector product

We saw earlier, in Sect. 3.10, that it is possible to combine two vectors multi-
plicatively, by means of a scalar product, to form a scalar. Recall that the scalar
product a·b of two vectors a = (ax , ay , az ) and b = (bx , by , bz ) is defined

a·b = ax bx + ay by + az bz = |a| |b| cos θ, (8.11)

where θ is the angle subtended between the directions of a and b.

166
8 ROTATIONAL MOTION 8.4 The vector product

ax b |a x b| = a b sin θ

θ
a
Figure 70: The vector product.

Is it also possible to combine two vector multiplicatively to form a third (non-

coplanar) vector? It turns out that this goal can be achieved via the use of the
so-called vector product. By definition, the vector product, a × b, of two vectors
a = (ax , ay , az ) and b = (bx , by , bz ) is of magnitude
|a × b| = |a| |b| sin θ. (8.12)
The direction of a × b is mutually perpendicular to a and b, in the sense given by
the right-hand grip rule when vector a is rotated onto vector b (the direction of
rotation being such that the angle of rotation is less than 180◦ ). See Fig. 70. In
coordinate form,
a × b = (ay bz − az by , az bx − ax bz , ax by − ay bx ). (8.13)

There are a number of fairly obvious consequences of the above definition.

Firstly, if vector b is parallel to vector a, so that we can write b = λ a, then the
vector product a × b has zero magnitude. The easiest way of seeing this is to note
that if a and b are parallel then the angle θ subtended between them is zero,
hence the magnitude of the vector product, |a| |b| sin θ, must also be zero (since
sin 0◦ = 0). Secondly, the order of multiplication matters. Thus, b × a is not
equivalent to a × b. In fact, as can be seen from Eq. (8.13),
b × a = −a × b. (8.14)
In other words, b × a has the same magnitude as a × b, but points in diagram-
matically the opposite direction.

167
8 ROTATIONAL MOTION 8.5 Centre of mass

Now that we have defined the vector product of two vectors, let us find a use
for this concept. Figure 71 shows a rigid body rotating with angular velocity ω.
For the sake of simplicity, the axis of rotation, which runs parallel to ω, is as-
sumed to pass through the origin O of our coordinate system. Point P, whose
position vector is r, represents a general point inside the body. What is the veloc-
ity of rotation v at point P? Well, the magnitude of this velocity is simply

v = σ ω = ω r sin θ, (8.15)

where σ is the perpendicular distance of point P from the axis of rotation, and
θ is the angle subtended between the directions of ω and r. The direction of
the velocity is into the page. Another way of saying this, is that the direction of
the velocity is mutually perpendicular to the directions of ω and r, in the sense
indicated by the right-hand grip rule when ω is rotated onto r (through an angle
less than 180◦ ). It follows that we can write

v = ω × r. (8.16)

Note, incidentally, that the direction of the angular velocity vector ω indicates
the orientation of the axis of rotation—however, nothing actually moves in this
direction; in fact, all of the motion is perpendicular to the direction of ω.

8.5 Centre of mass

The centre of mass—or centre of gravity—of an extended object is defined in much

the same manner as we earlier defined the centre of mass of a set of mutually
interacting point mass objects—see Sect. 6.3. To be more exact, the coordinates
of the centre of mass of an extended object are the mass weighted averages of
the coordinates of the elements which make up that object. Thus, if the object
has net mass M, and is composed of N elements, such that the ith element has
mass mi and position vector ri , then the position vector of the centre of mass is
given by
1 
rcm = mi ri . (8.17)
M i=1,N

168
8 ROTATIONAL MOTION 8.5 Centre of mass

σ v
P

θ r
O

Figure 71: Rigid rotation.

If the object under consideration is continuous, then

mi = ρ(ri ) Vi , (8.18)

where ρ(r) is the mass density of the object, and Vi is the volume occupied by
the ith element. Here, it is assumed that this volume is small compared to the
total volume of the object. Taking the limit that the number of elements goes
to infinity, and the volume of each element goes to zero, Eqs. (8.17) and (8.18)
yield the following integral formula for the position vector of the centre of mass:

1
rcm = ρ r dV. (8.19)
M
Here, the integral is taken over the whole volume of the object, and dV =
dx dy dz is an element of that volume. Incidentally, the triple integral sign in-
dicates a volume integral: i.e., a simultaneous integral over three independent
Cartesian coordinates. Finally, for an object whose mass density is constant—
which is the only type of object that we shall be considering in this course—the
above expression reduces to

1
rcm = r dV, (8.20)
V
169
8 ROTATIONAL MOTION 8.5 Centre of mass

where V is the volume of the object. According to Eq. (8.20), the centre of mass
of a body of uniform density is located at the geometric centre of that body.

geometric centre

a
a
Figure 72: Locating the geometric centre of a cube.

For many solid objects, the location of the geometric centre follows from sym-
metry. For instance, the geometric centre of a cube is the point of intersection
of the cube’s diagonals. See Fig. 72. Likewise, the geometric centre of a right
cylinder is located on the axis, half-way up the cylinder. See Fig. 73.
axis

geometric centre

h/2

Figure 73: Locating the geometric centre of a right cylinder.

As an illustration of the use of formula (8.20), let us calculate the geometric

centre of a regular square-sided pyramid. Figure 74 shows such a pyramid. Let a
be the length of each side.√ It follows, from simple trigonometry, that the height
of the pyramid is h = a/ 2. Suppose that the base of the pyramid lies on the
x-y plane, and the apex is aligned with the z-axis, as shown in the diagram. It
follows, from symmetry, that the geometric centre of the pyramid lies on the z-
axis. It only remains to calculate the perpendicular distance, zcm , between the
geometric centre and the base of the pyramid. This quantity is obtained from the

170
8 ROTATIONAL MOTION 8.5 Centre of mass

y z
geometric centre

a x h
a z cm
x
a a

top view side view

Figure 74: Locating the geometric centre of a regular square-sided pyramid.

z-component of Eq. (8.20):


z dx dy dz
zcm =  , (8.21)
dx dy dz
where the integral is taken over the volume of the pyramid.

In the above integral, the limits of integration for z are z = 0 to z = h, respec-

tively (i.e., from the base to the apex of the pyramid). The corresponding limits of
integration for x and y are x, y = −a (1 − z/h)/2 to x, y = +a (1 − z/h)/2, respec-
tively (i.e., the limits are x, y = ±a/2 at the base of the pyramid, and x, y = ±0
at the apex). Hence, Eq. (8.21) can be written more explicitly as
h +a (1−z/h)/2 +a (1−z/h)/2
0 z dz −a (1−z/h)/2 dy −a (1−z/h)/2 dx
zcm = h +a (1−z/h)/2 +a (1−z/h)/2 . (8.22)
0 dz −a (1−z/h)/2 dy −a (1−z/h)/2 dx

As indicated above, it makes sense to perform the x- and y- integrals before

the z-integrals, since the limits of integration for the x- and y- integrals are z-
dependent. Performing the x-integrals, we obtain
h +a (1−z/h)/2
0 z dz −a (1−z/h)/2 a (1 − z/h) dy
zcm = h +a (1−z/h)/2 . (8.23)
0 dz −a (1−z/h)/2 a (1 − z/h) dy

171
8 ROTATIONAL MOTION 8.6 Moment of inertia

Performing the y-integrals, we obtain

h 2 2
0 a z (1 − z/h) dz
zcm = h . (8.24)
a 2 (1 − z/h)2 dz
0

Finally, performing the z-integrals, we obtain

 h
a2 z2 /2 − 2 z3 /(3 h) + z4 /(4 h2 ) 0 a2 h2 /12 h
zcm = h = 2 = . (8.25)
a2 [z − z2 /(h) + z3 /(3 h)]0 a h/3 4
Thus, the geometric centre of a regular square-sided pyramid is located on the
symmetry axis, one quarter of the way from the base to the apex.

8.6 Moment of inertia

Consider an extended object which is made up of N elements. Let the ith element
possess mass mi , position vector ri , and velocity vi . The total kinetic energy of
the object is written
 1
K= mi vi2 . (8.26)
i=1,N
2
Suppose that the motion of the object consists merely of rigid rotation at angular
velocity ω. It follows, from Sect. 8.4, that
vi = ω × r i . (8.27)
Let us write
ω = ω k, (8.28)
where k is a unit vector aligned along the axis of rotation (which is assumed
to pass through the origin of our coordinate system). It follows from the above
equations that the kinetic energy of rotation of the object takes the form
 1
K= mi |k × ri |2 ω2 , (8.29)
i=1,N
2
or
1
K= I ω2 . (8.30)
2
172
8 ROTATIONAL MOTION 8.6 Moment of inertia

Here, the quantity I is termed the moment of inertia of the object, and is written
 
I= mi |k × ri |2 = mi σi2 , (8.31)
i=1,N i=1,N

where σi = |k × ri | is the perpendicular distance from the ith element to the axis
of rotation. Note that for translational motion we usually write
1
K= M v2 , (8.32)
2
where M represents mass and v represents speed. A comparison of Eqs. (8.30)
and (8.32) suggests that moment of inertia plays the same role in rotational
motion that mass plays in translational motion.

For a continuous object, analogous arguments to those employed in Sect. 8.5

yield 
I= ρ σ2 dV, (8.33)

where ρ(r) is the mass density of the object, σ = |k × r| is the perpendicular

distance from the axis of rotation, and dV is a volume element. Finally, for an
object of constant density, the above expression reduces to
 2
σ dV
I = M  . (8.34)
dV
Here, M is the total mass of the object. Note that the integrals are taken over the
whole volume of the object.

The moment of inertia of a uniform object depends not only on the size and
shape of that object but on the location of the axis about which the object is
rotating. In particular, the same object can have different moments of inertia
when rotating about different axes.

Unfortunately, the evaluation of the moment of inertia of a given body about

a given axis invariably involves the performance of a nasty volume integral. In
fact, there is only one trivial moment of inertia calculation—namely, the moment
of inertia of a thin circular ring about a symmetric axis which runs perpendicular
to the plane of the ring. See Fig. 75. Suppose that M is the mass of the ring, and

173
8 ROTATIONAL MOTION 8.6 Moment of inertia

axis I = M b2

M
ring

Figure 75: The moment of inertia of a ring about a perpendicular symmetric axis.

b is its radius. Each element of the ring shares a common perpendicular distance
from the axis of rotation—i.e., σ = b. Hence, Eq. (8.34) reduces to

I = M b2 . (8.35)

In general, moments of inertia are rather tedious to calculate. Fortunately,

there exist two powerful theorems which enable us to simply relate the moment
of inertia of a given body about a given axis to the moment of inertia of the same
body about another axis. The first of these theorems is called the perpendicular
axis theorem, and only applies to uniform laminar objects. Consider a laminar
object (i.e., a thin, planar object) of uniform density. Suppose, for the sake of
simplicity, that the object lies in the x-y plane. The moment of inertia of the
object about the z-axis is given by
 2
(x + y2 ) dx dy
Iz = M  , (8.36)
dx dy
where we have suppressed the trivial z-integration, and the integral is taken over
the extent of the object in the x-y plane. Incidentally, the above expression fol-
lows from the observation that σ2 = x2 +y2 when the axis of rotation is coincident
with the z-axis. Likewise, the moments of inertia of the object about the x- and
y- axes take the form
 2
y dx dy
Ix = M   , (8.37)
dx dy

174
8 ROTATIONAL MOTION 8.6 Moment of inertia

z
Iz = I x + I y

Figure 76: The perpendicular axis theorem.

 2
x dx dy
Iy = M  , (8.38)
dx dy
respectively. Here, we have made use of the fact that z = 0 inside the object. It
follows by inspection of the previous three equations that
Iz = I x + I y . (8.39)
See Fig. 76.

Let us use the perpendicular axis theorem to find the moment of inertia of a
thin ring about a symmetric axis which lies in the plane of the ring. Adopting the
coordinate system shown in Fig. 77, it is clear, from symmetry, that Ix = Iy . Now,
we already know that Iz = M b2 , where M is the mass of the ring, and b is its
radius. Hence, the perpendicular axis theorem tells us that
2 Ix = I z , (8.40)
or
Iz 1
Ix = = M b2 . (8.41)
2 2
Of course, Iz > Ix , because when the ring spins about the z-axis its elements are,
on average, farther from the axis of rotation than when it spins about the x-axis.

The second useful theorem regarding moments of inertia is called the parallel
axis theorem. The parallel axis theorem—which is quite general—states that if I
is the moment of inertia of a given body about an axis passing through the centre

175
8 ROTATIONAL MOTION 8.6 Moment of inertia

ring

Figure 77: The moment of inertia of a ring about a coplanar symmetric axis.

of mass of that body, then the moment of inertia I  of the same body about a
second axis which is parallel to the first is
I  = I + M d2 , (8.42)
where M is the mass of the body, and d is the perpendicular distance between
the two axes.

In order to prove the parallel axis theorem, let us choose the origin of our
coordinate system to coincide with the centre of mass of the body in question.
Furthermore, let us orientate the axes of our coordinate system such that the z-
axis coincides with the first axis of rotation, whereas the second axis pieces the
x-y plane at x = d, y = 0. From Eq. (8.20), the fact that the centre of mass is
located at the origin implies that
  
x dx dy dz = y dx dy dz = z dx dy dz = 0, (8.43)

where the integrals are taken over the volume of the body. From Eq. (8.34), the
expression for the first moment of inertia is
 2
(x + y2 ) dx dy dz
I=M  , (8.44)
dx dy dz
since x2 + y2 is the perpendicular distance of a general point (x, y, z) from the
z-axis. Likewise, the expression for the second moment of inertia takes the form

[(x − d)2 + y2 ] dx dy dz
I = M  . (8.45)
dx dy dz
176
8 ROTATIONAL MOTION 8.6 Moment of inertia

The above equation can be expanded to give

 2
 [(x + y2 ) − 2 d x + d2 ] dx dy dz
I = M 
dx dy dz
 2 
(x + y2 ) dx dy dz x dx dy dz
= M  − 2 d M 
dx dy dz dx dy dz

dx dy dz
+d2 M  . (8.46)
dx dy dz
It follows from Eqs. (8.43) and (8.44) that

I  = I + M d2 , (8.47)

which proves the theorem.

Let us use the parallel axis theorem to calculate the moment of inertia, I  , of
a thin ring about an axis which runs perpendicular to the plane of the ring, and
passes through the circumference of the ring. We know that the moment of inertia
of a ring of mass M and radius b about an axis which runs perpendicular to the
plane of the ring, and passes through the centre of the ring—which coincides
with the centre of mass of the ring—is I = M b2 . Our new axis is parallel to this
original axis, but shifted sideways by the perpendicular distance b. Hence, the
parallel axis theorem tells us that

I  = I + M b 2 = 2 M b2 . (8.48)

See Fig. 78.

As an illustration of the direct application of formula (8.34), let us calculate

the moment of inertia of a thin circular disk, of mass M and radius b, about an
axis which passes through the centre of the disk, and runs perpendicular to the
plane of the disk. Let us choose our coordinate system such that the disk lies
in the x-y plane with its centre at the origin. The axis of rotation is, therefore,
coincident with the z-axis. Hence, formula (8.34) reduces to
 2
(x + y2 ) dx dy
I=M  , (8.49)
dx dy
177
8 ROTATIONAL MOTION 8.6 Moment of inertia

axis 2 axis 2
I=Mb I=2Mb

M M
ring ring

b b

original axis new axis

Figure 78: An application of the parallel axis theorem.

where the integrals are taken over the area of the disk, and the redundant z-
integration has been suppressed.
 Let us divide the disk up into thin annuli. Con-
sider an annulus of radius σ = x2 + y2 and radial thickness dσ. The area of this
annulus is simply 2π σ dσ. Hence, we can replace dx dy in the above integrals by
2π σ dσ, so as to give
b 3
0 2π σ dσ
I = M b . (8.50)
0 2π σ dσ
The above expression yields
 b
2 π σ4 /4 1
I=M 0
b = M b2 . (8.51)
[2 π σ /2]0
2 2

Similar calculations to the above yield the following standard results:

• The moment of inertia of a thin rod of mass M and length l about an axis
passing through the centre of the rod and perpendicular to its length is
1
I= M l2 .
12
• The moment of inertia of a thin rectangular sheet of mass M and dimensions
a and b about a perpendicular axis passing through the centre of the sheet
is
1
I= M (a2 + b2 ).
12

178
8 ROTATIONAL MOTION 8.7 Torque

• The moment of inertia of a solid cylinder of mass M and radius b about the
cylindrical axis is
1
I = M b2 .
2
• The moment of inertia of a thin spherical shell of mass M and radius b about
a diameter is
2
I = M b2 .
3
• The moment of inertia of a solid sphere of mass M and radius b about a
diameter is
2
I = M b2 .
5

8.7 Torque

We have now identified the rotational equivalent of velocity—namely, angular

velocity—and the rotational equivalent of mass—namely, moment of inertia. But,
what is the rotational equivalent of force?

Consider a bicycle wheel of radius b which is free to rotate around a perpen-

dicular axis passing through its centre. Suppose that we apply a force f, which
is coplanar with the wheel, to a point P lying on its circumference. See Fig. 79.
What is the wheel’s subsequent motion?

Let us choose the origin O of our coordinate system to coincide with the pivot
point of the wheel—i.e., the point of intersection between the wheel and the axis
of rotation. Let r be the position vector of point P, and let θ be the angle sub-
tended between the directions of r and f. We can resolve f into two components—
namely, a component f cos θ which acts radially, and a component f sin θ which
acts tangentially. The radial component of f is canceled out by a reaction at the
pivot, since the wheel is assumed to be mounted in such a manner that it can only
rotate, and is prevented from displacing sideways. The tangential component of
f causes the wheel to accelerate tangentially. Let v be the instantaneous rotation
velocity of the wheel’s circumference. Newton’s second law of motion, applied to

179
8 ROTATIONAL MOTION 8.7 Torque

f sinθ f

O r
P
b

Figure 79: A rotating bicycle wheel.

the tangential motion of the wheel, yields

M v̇ = f sin θ, (8.52)

where M is the mass of the wheel (which is assumed to be concentrated in the

wheel’s rim).

Let us now convert the above expression into a rotational equation of motion.
If ω is the instantaneous angular velocity of the wheel, then the relation between
ω and v is simply
v = b ω. (8.53)
Since the wheel is basically a ring of radius b, rotating about a perpendicular
symmetric axis, its moment of inertia is

I = M b2 . (8.54)

Combining the previous three equations, we obtain

I ω̇ = τ, (8.55)

where
τ = f b sin θ. (8.56)

Equation (8.55) is the angular equation of motion of the wheel. It relates the
wheel’s angular velocity, ω, and moment of inertia, I, to a quantity, τ, which is
known as the torque. Clearly, if I is analogous to mass, and ω is analogous to

180
8 ROTATIONAL MOTION 8.7 Torque

f
τ=fl
θ

b P

Figure 80: Definition of the length of the level arm, l.

velocity, then torque must be analogous to force. In other words, torque is the
rotational equivalent of force.

It is clear, from Eq. (8.56), that a torque is the product of the magnitude of
the applied force, f, and some distance l = b sin θ. The physical interpretation
of l is illustrated in Fig. 80. If can be seen that l is the perpendicular distance of
the line of action of the force from the axis of rotation. We usually refer to this
distance as the length of the lever arm.

In summary, a torque measures the propensity of a given force to cause the

object upon which it acts to twist about a certain axis. The torque, τ, is simply
the product of the magnitude of the applied force, f, and the length of the lever
arm, l:
τ = f l. (8.57)
Of course, this definition makes a lot of sense. We all know that it is far easier
to turn a rusty bolt using a long, rather than a short, wrench. Assuming that we
exert the same force on the end of each wrench, the torque we apply to the bolt
is larger in the former case, since the perpendicular distance between the line of
action of the force and the bolt (i.e., the length of the wrench) is greater.

181
8 ROTATIONAL MOTION 8.7 Torque

Since force is a vector quantity, it stands to reason that torque must also be
a vector quantity. It follows that Eq. (8.57) defines the magnitude, τ, of some
torque vector, τ. But, what is the direction of this vector? By convention, if a
torque is such as to cause the object upon which it acts to twist about a certain
axis, then the direction of that torque runs along the direction of the axis in the
sense given by the right-hand grip rule. In other words, if the fingers of the right-
hand circulate around the axis of rotation in the sense in which the torque twists
the object, then the thumb of the right-hand points along the axis in the direction
of the torque. It follows that we can rewrite our rotational equation of motion,
Eq. (8.55), in vector form:
dω
I = I α = τ, (8.58)
dt
where α = dω/dt is the vector angular acceleration. Note that the direction of α
indicates the direction of the rotation axis about which the object accelerates (in
the sense given by the right-hand grip rule), whereas the direction of τ indicates
the direction of the rotation axis about which the torque attempts to twist the
object (in the sense given by the right-hand grip rule). Of course, these two
rotation axes are identical.

Although Eq. (8.58) was derived for the special case of a torque applied to a
ring rotating about a perpendicular symmetric axis, it is, nevertheless, completely
general.

It is important to appreciate that the directions we ascribe to angular velocities,

angular accelerations, and torques are merely conventions. There is actually no
physical motion in the direction of the angular velocity vector—in fact, all of the
motion is in the plane perpendicular to this vector. Likewise, there is no physical
acceleration in the direction of the angular acceleration vector—again, all of the
acceleration is in the plane perpendicular to this vector. Finally, no physical forces
act in the direction of the torque vector—in fact, all of the forces act in the plane
perpendicular to this vector.

Consider a rigid body which is free to pivot in any direction about some fixed
point O. Suppose that a force f is applied to the body at some point P whose
position vector relative to O is r. See Fig. 81. Let θ be the angle subtended

182
8 ROTATIONAL MOTION 8.7 Torque

f
torque
θ

r
O

pivot

Figure 81: Torque about a fixed point.

between the directions of r and f. What is the vector torque τ acting on the
object about an axis passing through the pivot point? The magnitude of this
torque is simply
τ = r f sin θ. (8.59)
In Fig. 81, the conventional direction of the torque is out of the page. Another
way of saying this is that the direction of the torque is mutually perpendicular
to both r and f, in the sense given by the right-hand grip rule when vector r is
rotated onto vector f (through an angle less than 180◦ degrees). It follows that
we can write
τ = r × f. (8.60)
In other words, the torque exerted by a force acting on a rigid body which pivots
about some fixed point is the vector product of the displacement of the point of
application of the force from the pivot point with the force itself. Equation (8.60)
specifies both the magnitude of the torque, and the axis of rotation about which
the torque twists the body upon which it acts. This axis runs parallel to the
direction of τ, and passes through the pivot point.

183
8 ROTATIONAL MOTION 8.8 Power and work

8.8 Power and work

Consider a mass m attached to the end of a light rod of length l whose other end
is attached to a fixed pivot. Suppose that the pivot is such that the rod is free
to rotate in any direction. Suppose, further, that a force f is applied to the mass,
whose instantaneous angular velocity about an axis of rotation passing through
the pivot is ω.

Let v be the instantaneous velocity of the mass. We know that the rate at which
the force f performs work on the mass—otherwise known as the power—is given
by (see Sect. 5.8)
P = f·v. (8.61)
However, we also know that (see Sect. 8.4)

v = ω × r, (8.62)

where r is the vector displacement of the mass from the pivot. Hence, we can
write
P =ω×r·f (8.63)
(note that a·b = b·a).

Now, for any three vectors, a, b, and c, we can write

a × b · c = a · b × c. (8.64)

This theorem is easily proved by expanding the vector and scalar products in
component form using the definitions (8.11) and (8.13). It follows that Eq. (8.63)
can be rewritten
P = ω · r × f. (8.65)
However,
τ = r × f, (8.66)
where τ is the torque associated with force f about an axis of rotation passing
through the pivot. Hence, we obtain

P = τ·ω. (8.67)

184
8 ROTATIONAL MOTION 8.8 Power and work

In other words, the rate at which a torque performs work on the object upon
which it acts is the scalar product of the torque and the angular velocity of the
object. Note the great similarity between Eq. (8.61) and Eq. (8.67).

Now the relationship between work, W, and power, P, is simply

dW
P= . (8.68)
dt
Likewise, the relationship between angular velocity, ω, and angle of rotation, φ,
is
dφ
ω= . (8.69)
dt
It follows that Eq. (8.67) can be rewritten

dW = τ·dφ. (8.70)

Integration yields 
W= τ·dφ. (8.71)

Note that this is a good definition, since it only involves an infinitesimal rotation
vector, dφ. Recall, from Sect. 8.3, that it is impossible to define a finite rotation
vector. For the case of translational motion, the analogous expression to the
above is 
W = f·dr. (8.72)

Here, f is the force, and dr is an element of displacement of the body upon which
the force acts.

Although Eqs. (8.67) and (8.71) were derived for the special case of the ro-
tation of a mass attached to the end of a light rod, they are, nevertheless, com-
pletely general.

Consider, finally, the special case in which the torque is aligned with the an-
gular velocity, and both are constant in time. In this case, the rate at which the
torque performs work is simply
P = τ ω. (8.73)

185
8 ROTATIONAL MOTION 8.9 Translational motion versus rotational motion

Translational motion Rotational motion

Displacement dr Angular displacement dφ
Velocity v = dr/dt Angular velocity ω = dφ/dt
Acceleration a = dv/dt Angular acceleration α = dω/dt

Mass M Moment of inertia I = ρ|ω ^ ×r|2 dV

Force f = Ma Torque τ ≡ r × f = Iα
 
Work W = f·dr Work W = τ·dφ
Power P = f·v Power P = τ·ω
Kinetic energy K = M v /2 Kinetic energy
2
K = I ω2/2

Table 3: The analogies between translational and rotational motion.

Likewise, the net work performed by the torque in twisting the body upon which
it acts through an angle Δφ is just

W = τ Δφ. (8.74)

8.9 Translational motion versus rotational motion

It should be clear, by now, that there is a strong analogy between rotational mo-
tion and standard translational motion. Indeed, each physical concept used to
analyze rotational motion has its translational concomitant. Likewise, every law
of physics governing rotational motion has a translational equivalent. The analo-
gies between rotational and translational motion are summarized in Table 3.

8.10 The physics of baseball

Baseball players know from experience that there is a “sweet spot” on a baseball
bat, about 17 cm from the end of the barrel, where the shock of impact with the
ball, as felt by the hands, is minimized. In fact, if the ball strikes the bat exactly
on the “sweet spot” then the hitter is almost unaware of the collision. Conversely,
if the ball strikes the bat well away from the “sweet spot” then the impact is felt
as a painful jarring of the hands.

186
8 ROTATIONAL MOTION 8.10 The physics of baseball

pivot

J’ J’

bat
b
centre of mass
h
l v
ball

J J

Figure 82: A schematic baseball bat.

The existence of a “sweet spot” on a baseball bat is just a consequence of ro-

tational dynamics. Let us analyze this problem. Consider the schematic baseball
bat shown in Fig. 82. Let M be the mass of the bat, and let l be its length. Sup-
pose that the bat pivots about a fixed point located at one of its ends. Let the
centre of mass of the bat be located a distance b from the pivot point. Finally,
suppose that the ball strikes the bat a distance h from the pivot point.

The collision between the bat and the ball can be modeled as equal and oppo-
site impulses, J, applied to each object at the time of the collision (see Sect. 6.5).
At the same time, equal and opposite impulses J  are applied to the pivot and the
bat, as shown in Fig. 82. If the pivot actually corresponds to a hitter’s hands then
the latter impulse gives rise to the painful jarring sensation felt when the ball is
not struck properly.

We saw earlier that in a general multi-component system—which includes an

extended body such as a baseball bat—the motion of the centre of mass takes
a particularly simple form (see Sect. 6.3). To be more exact, the motion of the
centre of mass is equivalent to that of the point particle obtained by concentrating
the whole mass of the system at the centre of mass, and then allowing all of the
external forces acting on the system to act upon that mass. Let us use this idea to

187
8 ROTATIONAL MOTION 8.10 The physics of baseball

analyze the effect of the collision with the ball on the motion of the bat’s centre
of mass. The centre of mass of the bat acts like a point particle of mass M which
is subject to the two impulses, J and J  (which are applied simultaneously). If v
is the instantaneous velocity of the centre of mass then the change in momentum
of this point due to the action of the two impulses is simply

M Δv = −J − J  . (8.75)

The minus signs on the right-hand side of the above equation follow from the fact
that the impulses are oppositely directed to v in Fig. 82.

Note that in order to specify the instantaneous state of an extended body we

must do more than just specify the location of the body’s centre of mass. In-
deed, since the body can rotate about its centre of mass, we must also specify its
orientation in space. Thus, in order to follow the motion of an extended body,
we must not only follow the translational motion of its centre of mass, but also
the body’s rotational motion about this point (or any other convenient reference
point located within the body).

Consider the rotational motion of the bat shown in Fig. 82 about a perpendic-
ular (to the bat) axis passing through the pivot point. This motion satisfies
dω
I = τ, (8.76)
dt
where I is the moment of inertia of the bat, ω is its instantaneous angular velocity,
and τ is the applied torque. The bat is actually subject to an impulsive torque (i.e.,
a torque which only lasts for a short period in time) at the time of the collision
with the ball. Defining the angular impulse K associated with an impulsive torque
τ in much the same manner as we earlier defined the impulse associated with an
impulsive force (see Sect. 6.5), we obtain
t
K = τ dt. (8.77)

It follows that we can integrate Eq. (8.76) over the time of the collision to find

I Δω = K, (8.78)

188
8 ROTATIONAL MOTION 8.10 The physics of baseball

where Δω is the change in angular velocity of the bat due to the collision with
the ball.

Now, the torque associated with a given force is equal to the magnitude of the
force times the length of the lever arm. Thus, it stands to reason that the angular
impulse, K, associated with an impulse, J, is simply
K = J x, (8.79)
where x is the perpendicular distance from the line of action of the impulse to the
axis of rotation. Hence, the angular impulses associated with the two impulses,
J and J  , to which the bat is subject when it collides with the ball, are J h and 0,
respectively. The latter angular impulse is zero since the point of application of
the associated impulse coincides with the pivot point, and so the length of the
lever arm is zero. It follows that Eq. (8.78) can be written
I Δω = −J h. (8.80)
The minus sign comes from the fact that the impulse J is oppositely directed to
the angular velocity in Fig. 82.

Now, the relationship between the instantaneous velocity of the bat’s centre of
mass and the bat’s instantaneous angular velocity is simply
v = b ω. (8.81)
Hence, Eq. (8.75) can be rewritten
M b Δω = −J − J  . (8.82)

Equations (8.80) and (8.82) can be combined to yield

 
 Mbh
J =− 1− J. (8.83)
I
’ The above expression specifies the magnitude of the impulse J  applied to the
hitter’s hands terms of the magnitude of the impulse J applied to the ball.

Let us crudely model the bat as a uniform rod of length l and mass M. It
follows, by symmetry, that the centre of mass of the bat lies at its half-way point:

189
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

i.e.,
l
b= . (8.84)
2
Moreover, the moment of inertia of the bat about a perpendicular axis passing
through one of its ends is
1
I = M l2 (8.85)
3
(this is a standard result). Combining the previous three equations, we obtain
   
 3h h
J =− 1− J=− 1− J, (8.86)
2l h0
where
2
h0 = l. (8.87)
3
Clearly, if h = h0 then no matter how hard the ball is hit (i.e., no matter how
large we make J) zero impulse is applied to the hitter’s hands. We conclude that
the “sweet spot”—or, in scientific terms, the centre of percussion—of a uniform
baseball bat lies two-thirds of the way down the bat from the hitter’s end. If
we adopt a more realistic model of a baseball bat, in which the bat is tapered
such that the majority of its weight is located at its hitting end, we can easily
demonstrate that the centre of percussion is shifted further away from the hitter
(i.e., it is more that two-thirds of the way along the bat).

8.11 Combined translational and rotational motion

In Sect. 4.7, we analyzed the motion of a block sliding down a frictionless incline.
We found that the block accelerates down the slope with uniform acceleration
g sin θ, where θ is the angle subtended by the incline with the horizontal. In this
case, all of the potential energy lost by the block, as it slides down the slope, is
converted into translational kinetic energy (see Sect. 5). In particular, no energy
is dissipated.

There is, of course, no way in which a block can slide over a frictional surface
without dissipating energy. However, we know from experience that a round

190
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

object can roll over such a surface with hardly any dissipation. For instance, it
is far easier to drag a heavy suitcase across the concourse of an airport if the
suitcase has wheels on the bottom. Let us investigate the physics of round objects
rolling over rough surfaces, and, in particular, rolling down rough inclines.

Consider a uniform cylinder of radius b rolling over a horizontal, frictional

surface. See Fig. 83. Let v be the translational velocity of the cylinder’s centre
of mass, and let ω be the angular velocity of the cylinder about an axis running
along its length, and passing through its centre of mass. Consider the point of
contact between the cylinder and the surface. The velocity v  of this point is
made up of two components: the translational velocity v, which is common to
all elements of the cylinder, and the tangential velocity vt = −b ω, due to the
cylinder’s rotational motion. Thus,
v  = v − vt = v − b ω. (8.88)
Suppose that the cylinder rolls without slipping. In other words, suppose that
there is no frictional energy dissipation as the cylinder moves over the surface.
This is only possible if there is zero net motion between the surface and the
bottom of the cylinder, which implies v  = 0, or
v = b ω. (8.89)
It follows that when a cylinder, or any other round object, rolls across a rough sur-
face without slipping—i.e., without dissipating energy—then the cylinder’s trans-
lational and rotational velocities are not independent, but satisfy a particular
relationship (see the above equation). Of course, if the cylinder slips as it rolls
across the surface then this relationship no longer holds.

Consider, now, what happens when the cylinder shown in Fig. 83 rolls, with-
out slipping, down a rough slope whose angle of inclination, with respect to the
horizontal, is θ. If the cylinder starts from rest, and rolls down the slope a verti-
cal distance h, then its gravitational potential energy decreases by −ΔP = M g h,
where M is the mass of the cylinder. This decrease in potential energy must be
offset by a corresponding increase in kinetic energy. (Recall that when a cylin-
der rolls without slipping there is no frictional energy loss.) However, a rolling
cylinder can possesses two different types of kinetic energy. Firstly, translational

191
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

vt
centre of mass

cylinder
v
b

vt
surface

Figure 83: A cylinder rolling over a rough surface.

kinetic energy: Kt = (1/2) M v2 , where v is the cylinder’s translational velocity;

and, secondly, rotational kinetic energy: Kr = (1/2) I ω2 , where ω is the cylin-
der’s angular velocity, and I is its moment of inertia. Hence, energy conservation
yields
1 1
M g h = M v2 + I ω 2 . (8.90)
2 2
Now, when the cylinder rolls without slipping, its translational and rotational
velocities are related via Eq. (8.89). It follows from Eq. (8.90) that
2gh
v2 = . (8.91)
1 + I/M b2
Making use of the fact that the moment of inertia of a uniform cylinder about
its axis of symmetry is I = (1/2) M b2 , we can write the above equation more
explicitly as
4
v2 = g h. (8.92)
3
Now, if the same cylinder were to slide down a frictionless slope, such that it fell
from rest through a vertical distance h, then its final translational velocity would
satisfy
v2 = 2 g h. (8.93)
A comparison of Eqs. (8.92) and (8.93) reveals that when a uniform cylinder rolls
down an incline without slipping, its final translational velocity is less than that
obtained when the cylinder slides down the same incline without friction. The
reason for this is that, in the former case, some of the potential energy released

192
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

centre of mass
cylinder

f b
R

Mg

slope

Figure 84: A cylinder rolling down a rough incline.

as the cylinder falls is converted into rotational kinetic energy, whereas, in the
latter case, all of the released potential energy is converted into translational
kinetic energy. Note that, in both cases, the cylinder’s total kinetic energy at the
bottom of the incline is equal to the released potential energy.

Let us examine the equations of motion of a cylinder, of mass M and radius

b, rolling down a rough slope without slipping. As shown in Fig. 84, there are
three forces acting on the cylinder. Firstly, we have the cylinder’s weight, M g,
which acts vertically downwards. Secondly, we have the reaction, R, of the slope,
which acts normally outwards from the surface of the slope. Finally, we have the
frictional force, f, which acts up the slope, parallel to its surface.

As we have already discussed, we can most easily describe the translational

motion of an extended body by following the motion of its centre of mass. This
motion is equivalent to that of a point particle, whose mass equals that of the
body, which is subject to the same external forces as those that act on the body.
Thus, applying the three forces, M g, R, and f, to the cylinder’s centre of mass,
and resolving in the direction normal to the surface of the slope, we obtain
R = M g cos θ. (8.94)
Furthermore, Newton’s second law, applied to the motion of the centre of mass

193
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

parallel to the slope, yields

M v̇ = M g sin θ − f, (8.95)
where v̇ is the cylinder’s translational acceleration down the slope.

Let us, now, examine the cylinder’s rotational equation of motion. First, we
must evaluate the torques associated with the three forces acting on the cylin-
der. Recall, that the torque associated with a given force is the product of the
magnitude of that force and the length of the level arm—i.e., the perpendicular
distance between the line of action of the force and the axis of rotation. Now, by
definition, the weight of an extended object acts at its centre of mass. However,
in this case, the axis of rotation passes through the centre of mass. Hence, the
length of the lever arm associated with the weight M g is zero. It follows that
the associated torque is also zero. It is clear, from Fig. 84, that the line of action
of the reaction force, R, passes through the centre of mass of the cylinder, which
coincides with the axis of rotation. Thus, the length of the lever arm associated
with R is zero, and so is the associated torque. Finally, according to Fig. 84, the
perpendicular distance between the line of action of the friction force, f, and the
axis of rotation is just the radius of the cylinder, b—so the associated torque is
f b. We conclude that the net torque acting on the cylinder is simply
τ = f b. (8.96)
It follows that the rotational equation of motion of the cylinder takes the form,
I ω̇ = τ = f b, (8.97)
where I is its moment of inertia, and ω̇ is its rotational acceleration.

Now, if the cylinder rolls, without slipping, such that the constraint (8.89) is
satisfied at all times, then the time derivative of this constraint implies the follow-
ing relationship between the cylinder’s translational and rotational accelerations:
v̇ = b ω̇. (8.98)
It follows from Eqs. (8.95) and (8.97) that
g sin θ
v̇ = , (8.99)
1 + I/M b2
194
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

M g sin θ
f = . (8.100)
1 + M b2 /I
Since the moment of inertia of the cylinder is actually I = (1/2) M b 2 , the above
expressions simplify to give
2
v̇ = g sin θ, (8.101)
3
and
1
f = M g sin θ. (8.102)
3
Note that the acceleration of a uniform cylinder as it rolls down a slope, without
slipping, is only two-thirds of the value obtained when the cylinder slides down
the same slope without friction. It is clear from Eq. (8.95) that, in the former
case, the acceleration of the cylinder down the slope is retarded by friction. Note,
however, that the frictional force merely acts to convert translational kinetic en-
ergy into rotational kinetic energy, and does not dissipate energy.

Now, in order for the slope to exert the frictional force specified in Eq. (8.102),
without any slippage between the slope and cylinder, this force must be less than
the maximum allowable static frictional force, μ R(= μ M g cos θ), where μ is the
coefficient of static friction. In other words, the condition for the cylinder to roll
down the slope without slipping is f < μ R, or

tan θ < 3 μ. (8.103)

This condition is easily satisfied for gentle slopes, but may well be violated for ex-
tremely steep slopes (depending on the size of μ). Of course, the above condition
is always violated for frictionless slopes, for which μ = 0.

Suppose, finally, that we place two cylinders, side by side and at rest, at the top
of a frictional slope of inclination θ. Let the two cylinders possess the same mass,
M, and the same radius, b. However, suppose that the first cylinder is uniform,
whereas the second is a hollow shell. Which cylinder reaches the bottom of
the slope first, assuming that they are both released simultaneously, and both
roll without slipping? The acceleration of each cylinder down the slope is given
by Eq. (8.99). For the case of the solid cylinder, the moment of inertia is I =

195
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

(1/2) M b2 , and so
2
v̇solid =g sin θ. (8.104)
3
For the case of the hollow cylinder, the moment of inertia is I = M b2 (i.e., the
same as that of a ring with a similar mass, radius, and axis of rotation), and so
1
v̇hollow = g sin θ. (8.105)
2
It is clear that the solid cylinder reaches the bottom of the slope before the hollow
one (since it possesses the greater acceleration). Note that the accelerations of
the two cylinders are independent of their sizes or masses. This suggests that a
solid cylinder will always roll down a frictional incline faster than a hollow one,
irrespective of their relative dimensions (assuming that they both roll without
slipping). In fact, Eq. (8.99) suggests that whenever two different objects roll
(without slipping) down the same slope, then the most compact object—i.e., the
object with the smallest I/M b2 ratio—always wins the race.

Worked example 8.1: Balancing tires

Question: A tire placed on a balancing machine in a service station starts from

rest and turns through 5.3 revolutions in 2.3 s before reaching its final angular
speed. What is the angular acceleration of the tire (assuming that this quantity
remains constant)? What is the final angular speed of the tire?

Answer: The tire turns through φ = 5.3 × 2 π = 33.30 rad. in t = 2.3 s. The
relationship between φ and t for the case of rotational motion, starting from rest,
with uniform angular acceleration α is
1 2
φ= αt .
2
Hence,
2 φ 2 × 33.30
α= 2
= 2
= 12.59 rad./s2 .
t 2.3

196
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

Given that the tire starts from rest, its angular velocity after t seconds takes
the form
ω = α t = 12.59 × 2.3 = 28.96 rad./s.

Worked example 8.2: Accelerating a wheel

Question: The net work done in accelerating a wheel from rest to an angular
speed of 30 rev./min. is W = 5500 J. What is the moment of inertia of the wheel?

Answer: The final angular speed of the wheel is

ω = 30 × 2 π/60 = 3.142 rad./s.
Assuming that all of the work W performed on the wheel goes to increase its
rotational kinetic energy, we have
1
W= I ω2 ,
2
where I is the wheel’s moment of inertia. It follows that
2W 2 × 5500
I= 2 = 2
= 1114.6 kg m2 .
ω 3.142

Worked example 8.3: Moment of inertia of a rod

Question: A rod of mass M = 3 kg and length L = 1.2 m pivots about an axis,

perpendicular to its length, which passes through one of its ends. What is the
moment of inertia of the rod? Given that the rod’s instantaneous angular velocity
is 60 deg./s, what is its rotational kinetic energy?

Answer: The moment of inertia of a rod of mass M and length L about an axis,
perpendicular to its length, which passes through its midpoint is I = (1/12) M L 2 .
This is a standard result. Using the parallel axis theorem, the moment of inertia
about a parallel axis passing through one of the ends of the rod is
 2
 L 1
I =I+M = M L2 ,
2 3
197
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

so
 3 × 1.22
I = = 1.44 kg m2 .
3
The instantaneous angular velocity of the rod is
π
ω = 60 × = 1.047 rad./s.
180
Hence, the rod’s rotational kinetic energy is written
1  2
K= I ω = 0.5 × 1.44 × 1.0472 = 0.789 J.
2

Worked example 8.4: Weight and pulley

Question: A weight of mass m = 2.6 kg is suspended via a light inextensible

cable which is wound around a pulley of mass M = 6.4 kg and radius b = 0.4 m.
Treating the pulley as a uniform disk, find the downward acceleration of the
weight and the tension in the cable. Assume that the cable does not slip with
respect to the pulley.
b
pulley

ω
T

weight

mg

Answer: Let v be the instantaneous downward velocity of the weight, ω the in-
stantaneous angular velocity of the pulley, and T the tension in the cable. Apply-
ing Newton’s second law to the vertical motion of the weight, we obtain
m v̇ = m g − T.

198
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

The angular equation of motion of the pulley is written

I ω̇ = τ,
where I is its moment of inertia, and τ is the torque acting on the pulley. Now, the
only force acting on the pulley (whose line of action does not pass through the
pulley’s axis of rotation) is the tension in the cable. The torque associated with
this force is the product of the tension, T , and the perpendicular distance from
the line of action of this force to the rotation axis, which is equal to the radius, b,
of the pulley. Hence,
τ = T b.
If the cable does not slip with respect to the pulley, then its downward velocity, v,
must match the tangential velocity of the outer surface of the pulley, b ω. Thus,
v = b ω.
It follows that
v̇ = b ω̇.

The above equations can be combined to give

g
v̇ = ,
1 + I/m b2
mg
T = .
1 + m b2 /I
Now, the moment of inertia of the pulley is I = (1/2) M b2 . Hence, the above
expressions reduce to
g 9.81
v̇ = = = 4.40 m/s2 ,
1 + M/2 m 1 + 6.4/2 × 2.6
mg 2.6 × 9.81
T = = = 14.07 N.
1 + 2 m/M 1 + 2 × 2.6/6.4

Worked example 8.5: Hinged rod

Question: A uniform rod of mass m = 5.3 kg and length l = 1.3 m rotates about a
fixed frictionless pivot located at one of its ends. The rod is released from rest at

199
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

an angle θ = 35◦ beneath the horizontal. What is the angular acceleration of the
rod immediately after it is released?
x
pivot
θ

l/2
rod

l
mg

Answer: The moment of inertia of a rod of mass m and length l about an axis,
perpendicular to its length, which passes through one of its ends is I = (1/3) m l 2
(see question 8.3). Hence,
5.3 × 1.32
I= = 2.986 kg m2 .
3
The angular equation of motion of the rod is
I α = τ,
where α is the rod’s angular acceleration, and τ is the net torque exerted on the
rod. Now, the only force acting on the rod (whose line of action does not pass
through the pivot) is the rod’s weight, m g. This force acts at the centre of mass
of the rod, which is situated at the rod’s midpoint. The perpendicular distance x
between the line of action of the weight and the pivot point is simply
l 1.3 × cos 35◦
x = cos θ = = 0.532 m.
2 2
Thus, the torque acting on the rod is
τ = m g x.
It follows that the rod’s angular acceleration is written
τ m g x 5.3 × 9.81 × 0.532
α= = = = 9.26 rad./s2 .
I I 2.986

200
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

Worked example 8.6: Horsepower of engine

Question: A car engine develops a torque of τ = 500 N m and rotates at 3000 rev./min..
What horsepower does the engine generate? (1 hp = 746 W).

Answer: The angular speed of the engine is

ω = 3000 × 2 π/60 = 314.12 rad./s.
Thus, the power output of the engine is
P = ω τ = 314.12 × 500 = 1.57 × 105 W.
In units of horsepower, this becomes
1.57 × 105
P= = 210.5 hp.
746

Worked example 8.7: Rotating cylinder

Question: A uniform cylinder of radius b = 0.25 m is given an angular speed

of ω0 = 35 rad./s about an axis, parallel to its length, which passes through its
centre. The cylinder is gently lowered onto a horizontal frictional surface, and
released. The coefficient of friction of the surface is μ = 0.15. How long does it
take before the cylinder starts to roll without slipping? What distance does the
cylinder travel between its release point and the point at which it commences to
roll without slipping?

Answer: Let v be the velocity of the cylinder’s centre of mass, ω the cylinder’s
angular velocity, f the frictional force exerted by the surface on the cylinder, M
the cylinder’s mass, and I the cylinder’s moment of inertia. The cylinder’s trans-
lational equation of motion is written
M v̇ = f.
Note that the friction force acts to accelerate the cylinder’s translational motion.
Likewise, the cylinder’s rotational equation of motion takes the form
I ω̇ = −f b,

201
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

cylinder

b v
surface

since the perpendicular distance between the line of action of f and the axis
of rotation is the radius, b, of the cylinder. Note that the friction force acts to
decelerate the cylinder’s rotational motion. If the cylinder is slipping with respect
to the surface, then the friction force, f, is equal to the coefficient of friction, μ,
times the normal reaction, M g, at the surface:
f = μ M g.
Finally, the moment of inertia of the cylinder is
1
I= M b2 .
2

The above equations can be solved to give

v̇ = μ g,
b ω̇ = −2 μ g.
Given that v = 0 (i.e., the cylinder is initially at rest) and ω = ω 0 at time t = 0,
the above expressions can be integrated to give
v = μ g t,
b ω = b ω0 − 2 μ g t,
which yields
v − b ω = −(b ω0 − 3 μ g t).

Now, the cylinder stops slipping as soon as the “no slip” condition,
v = b ω,

202
8 ROTATIONAL MOTION 8.11 Combined translational and rotational motion

is satisfied. This occurs when

b ω0 0.25 × 35
t= = = 1.98 s.
3 μ g 3 × 0.15 × 9.81
Whilst it is slipping, the cylinder travels a distance
1
x= μ g t2 = 0.5 × 0.15 × 9.81 × 1.982 = 2.88 m.
2

203
9 ANGULAR MOMENTUM

9 Angular momentum

9.1 Introduction

Two physical quantities are noticeable by their absence in Table 3. Namely, mo-
mentum, and its rotational concomitant angular momentum. It turns out that
angular momentum is a sufficiently important concept to merit a separate discus-
sion.

9.2 Angular momentum of a point particle

Consider a particle of mass m, position vector r, and instantaneous velocity v,

which rotates about an axis passing through the origin of our coordinate system.
We know that the particle’s linear momentum is written
p = m v, (9.1)
and satisfies
dp
= f, (9.2)
dt
where f is the force acting on the particle. Let us search for the rotational equiv-
alent of p.

Consider the quantity

l = r × p. (9.3)
This quantity—which is known as angular momentum—is a vector of magnitude
l = r p sin θ, (9.4)
where θ is the angle subtended between the directions of r and p. The direction of
l is defined to be mutually perpendicular to the directions of r and p, in the sense
given by the right-hand grip rule. In other words, if vector r rotates onto vector
p (through an angle less than 180◦ ), and the fingers of the right-hand are aligned
with this rotation, then the thumb of the right-hand indicates the direction of l.
See Fig. 85.

204
9 ANGULAR MOMENTUM 9.2 Angular momentum of a point particle

l = r p sin θ
p

l particle

origin r

O
Figure 85: Angular momentum of a point particle about the origin.

Let us differentiate Eq. (9.3) with respect to time. We obtain

dl
= ṙ × p + r × ṗ. (9.5)
dt
Note that the derivative of a vector product is formed in much the same manner
as the derivative of an ordinary product, except that the order of the various
terms is preserved. Now, we know that ṙ = v = p/m and ṗ = f. Hence, we obtain
dl p×p
= + r × f. (9.6)
dt m
However, p × p = 0, since the vector product of two parallel vectors is zero. Also,
r × f = τ, (9.7)
where τ is the torque acting on the particle about an axis passing through the
origin. We conclude that
dl
= τ. (9.8)
dt
Of course, this equation is analogous to Eq. (9.2), which suggests that angular
momentum, l, plays the role of linear momentum, p, in rotational dynamics.

For the special case of a particle of mass m executing a circular orbit of ra-
dius r, with instantaneous velocity v and instantaneous angular velocity ω, the
magnitude of the particle’s angular momentum is simply
l = m v r = m ω r2 . (9.9)

205
9 ANGULAR MOMENTUM 9.3 Angular momentum of an extended object

9.3 Angular momentum of an extended object

Consider a rigid object rotating about some fixed axis with angular velocity ω.
Let us model this object as a swarm of N particles. Suppose that the ith particle
has mass mi , position vector ri , and velocity vi . Incidentally, it is assumed that
the object’s axis of rotation passes through the origin of our coordinate system.
The total angular momentum of the object, L, is simply the vector sum of the
angular momenta of the N particles from which it is made up. Hence,

L= mi ri × vi . (9.10)
i=1,N

Now, for a rigidly rotating object we can write (see Sect. 8.4)

vi = ω × r i . (9.11)

Let
ω = ω k, (9.12)
where k is a unit vector pointing along the object’s axis of rotation (in the sense
given by the right-hand grip rule). It follows that

L=ω mi ri × (k × ri ). (9.13)
i=1,N

Let us calculate the component of L along the object’s rotation axis—i.e., the
component along the k axis. We can write

Lk = L · k = ω mi k · ri × (k × ri ). (9.14)
i=1,N

However, since a · b × c = a × b · c, the above expression can be rewritten

 
Lk = ω mi (k × ri ) · (k × ri ) = ω mi |k × ri |2 . (9.15)
i=1,N i=1,N

Now, 
mi |k × ri |2 = Ik , (9.16)
i=1,N

206
9 ANGULAR MOMENTUM 9.3 Angular momentum of an extended object

where Ik is the moment of inertia of the object about the k axis. (see Sect. 8.6).
Hence, it follows that
Lk = Ik ω. (9.17)

According to the above formula, the component of a rigid body’s angular mo-
mentum vector along its axis of rotation is simply the product of the body’s mo-
ment of inertia about this axis and the body’s angular velocity. Does this result
imply that we can automatically write

L = I ω? (9.18)

Unfortunately, in general, the answer to the above question is no! This conclusion
follows because the body may possess non-zero angular momentum components
about axes perpendicular to its axis of rotation. Thus, in general, the angular
momentum vector of a rotating body is not parallel to its angular velocity vector.
This is a major difference from translational motion, where linear momentum is
always found to be parallel to linear velocity.

For a rigid object rotating with angular velocity ω = (ωx , ωy , ωz ), we can

write the object’s angular momentum L = (Lx , Ly , Lz ) in the form

Lx = Ix ωx , (9.19)
Ly = Iy ωy , (9.20)
Lz = Iz ωz , (9.21)

where Ix is the moment of inertia of the object about the x-axis, etc. Here, it is
again assumed that the origin of our coordinate system lies on the object’s axis
of rotation. Note that the above equations are only valid when the x-, y-, and
z-axes are aligned in a certain very special manner—in fact, they must be aligned
along the so-called principal axes of the object (these axes invariably coincide
with the object’s main symmetry axes). Note that it is always possible to find
three, mutually perpendicular, principal axes of rotation which pass through a
given point in a rigid body. Reconstructing L from its components, we obtain

^ + I y ωy y
L = I x ωx x ^ + Iz ωz ^z, (9.22)

207
9 ANGULAR MOMENTUM 9.3 Angular momentum of an extended object

where x ^ is a unit vector pointing along the x-axis, etc. It is clear, from the above
equation, that the reason L is not generally parallel to ω is because the moments
of inertia of a rigid object about its different possible axes of rotation are not
generally the same. In other words, if Ix = Iy = Iz = I then L = I ω, and the
angular momentum and angular velocity vectors are always parallel. However, if
Ix = Iy = Iz , which is usually the case, then L is not, in general, parallel to ω.

Although Eq. (9.22) suggests that the angular momentum of a rigid object is
not generally parallel to its angular velocity, this equation also implies that there
are, at least, three special axes of rotation for which this is the case. Suppose, for
instance, that the object rotates about the z-axis, so that ω = ω z ^z. It follows
from Eq. (9.22) that
L = Iz ωz ^z = Iz ω. (9.23)
Thus, in this case, the angular momentum vector is parallel to the angular velocity
vector. The same can be said for rotation about the x- or y- axes. We conclude
that when a rigid object rotates about one of its principal axes then its angular
momentum is parallel to its angular velocity, but not, in general, otherwise.

How can we identify a principal axis of a rigid object? At the simplest level,
a principal axis is one about which the object possesses axial symmetry. The
required type of symmetry is illustrated in Fig. 86. Assuming that the object
can be modeled as a swarm of particles—for every particle of mass m, located
a distance r from the origin, and subtending an angle θ with the rotation axis,
there must be an identical particle located on diagrammatically the opposite side
of the rotation axis. As shown in the diagram, the angular momentum vectors
of such a matched pair of particles can be added together to form a resultant
angular momentum vector which is parallel to the axis of rotation. Thus, if the
object is composed entirely of matched particle pairs then its angular momentum
vector must be parallel to its angular velocity vector. The generalization of this
argument to deal with continuous objects is fairly straightforward. For instance,
symmetry implies that any axis of rotation which passes through the centre of a
uniform sphere is a principal axis of that object. Likewise, a perpendicular axis
which passes through the centre of a uniform disk is a principal axis. Finally, a
perpendicular axis which passes through the centre of a uniform rod is a principal
axis.

208
9 ANGULAR MOMENTUM 9.4 Angular momentum of a multi-component system

axis of rotation
ω
l1 l2 l2

v1 v2 l
m m l1

r θ θ r

Figure 86: A principal axis of rotation.

9.4 Angular momentum of a multi-component system

Consider a system consisting of N mutually interacting point particles. Such

a system might represent a true multi-component system, such as an asteroid
cloud, or it might represent an extended body. Let the ith particle, whose mass is
mi , be located at vector displacement ri . Suppose that this particle exerts a force
fji on the jth particle. By Newton’s third law of motion, the force fij exerted by
the jth particle on the ith is given by

fij = −fji . (9.24)

Let us assume that the internal forces acting within the system are central forces—
i.e., the force fij , acting between particles i and j, is directed along the line of
centres of these particles. See Fig. 87. In other words,

fij ∝ (ri − rj ). (9.25)

Incidentally, this is not a particularly restrictive assumption, since most forces

occurring in nature are central forces. For instance, gravity is a central force,
electrostatic forces are central, and the internal stresses acting within a rigid
body are approximately central. Suppose, finally, that the ith particle is subject
to an external force Fi .

209
9 ANGULAR MOMENTUM 9.4 Angular momentum of a multi-component system

f Fi
ij

mi
line of centres

ri Fj

mj

rj f
ji

Figure 87: A multi-component system with central internal forces.

The equation of motion of the ith particle can be written

j=i

ṗi = fij + Fi . (9.26)
j=1,N

Taking the vector product of this equation with the position vector r i , we obtain
j=i

ri × ṗi = ri × fij + ri × Fi . (9.27)
j=1,N

Now, we have already seen that

d(ri × pi )
ri × ṗi =. (9.28)
dt
We also know that the total angular momentum, L, of the system (about the
origin) can be written in the form

L= ri × pi . (9.29)
i=1,N
Hence, summing Eq. (9.27) over all particles, we obtain
i=j 
dL
= ri × fij + ri × Fi . (9.30)
dt i,j=1,N i=1,N

210
9 ANGULAR MOMENTUM 9.4 Angular momentum of a multi-component system

Consider the first expression on the right-hand side of Eq. (9.30). A general
term, ri × fij , in this sum can always be paired with a matching term, rj × fji , in
which the indices have been swapped. Making use of Eq. (9.24), the sum of a
general matched pair can be written

ri × fij + rj × fji = (ri − rj ) × fij . (9.31)

However, if the internal forces are central in nature then fij is parallel to (ri − rj ).
Hence, the vector product of these two vectors is zero. We conclude that

ri × fij + rj × fji = 0, (9.32)

for any values of i and j. Thus, the first expression on the right-hand side of
Eq. (9.30) sums to zero. We are left with
dL
= τ, (9.33)
dt
where 
τ= ri × Fi (9.34)
i=1,N
is the net external torque acting on the system (about an axis passing through
the origin). Of course, Eq. (9.33) is simply the rotational equation of motion for
the system taken as a whole.

Suppose that the system is isolated, such that it is subject to zero net external
torque. It follows from Eq. (9.33) that, in this case, the total angular momentum
of the system is a conserved quantity. To be more exact, the components of the to-
tal angular momentum taken about any three independent axes are individually
conserved quantities. Conservation of angular momentum is an extremely useful
concept which greatly simplifies the analysis of a wide range of rotating systems.
Let us consider some examples.

Suppose that two identical weights of mass m are attached to a light rigid rod
which rotates without friction about a perpendicular axis passing through its mid-
point. Imagine that the two weights are equipped with small motors which allow
them to travel along the rod: the motors are synchronized in such a manner that
the distance of the two weights from the axis of rotation is always the same. Let

211
9 ANGULAR MOMENTUM 9.4 Angular momentum of a multi-component system

axle

m m
d d

rod weight

Figure 88: Two movable weights on a rotating rod.

us call this common distance d, and let ω be the angular velocity of the rod. See
Fig. 88. How does the angular velocity ω change as the distance d is varied?

Note that there are no external torques acting on the system. It follows that the
system’s angular momentum must remain constant as the weights move along the
rod. Neglecting the contribution of the rod, the moment of inertia of the system
is written
I = 2 m d2 . (9.35)
Since the system is rotating about a principal axis, its angular momentum takes
the form
L = I ω = 2 m d2 ω. (9.36)
If L is a constant of the motion then we obtain
ω d2 = constant. (9.37)
In other words, the system spins faster as the weights move inwards towards the
axis of rotation, and vice versa. This effect is familiar from figure skating. When
a skater spins about a vertical axis, her angular momentum is approximately a
conserved quantity, since the ice exerts very little torque on her. Thus, if the
skater starts spinning with outstretched arms, and then draws her arms inwards,
then her rate of rotation will spontaneously increase in order to conserve angular
momentum. The skater can slow her rate of rotation by simply pushing her arms
outwards again.

212
9 ANGULAR MOMENTUM 9.4 Angular momentum of a multi-component system

bullet

M
v b
m
d

pivot

rod

Figure 89: A bullet strikes a pivoted rod.

Suppose that a bullet of mass m and velocity v strikes, and becomes embedded
in, a stationary rod of mass M and length 2 b which pivots about a frictionless
perpendicular axle passing through its mid-point. Let the bullet strike the rod
normally a distance d from its axis of rotation. See Fig. 89. What is the instanta-
neous angular velocity ω of the rod (and bullet) immediately after the collision?

Taking the bullet and the rod as a whole, this is again a system upon which
no external torque acts. Thus, we expect the system’s net angular momentum to
be the same before and after the collision. Before the collision, only the bullet
possesses angular momentum, since the rod is at rest. As is easily demonstrated,
the bullet’s angular momentum about the pivot point is
l = mvd : (9.38)
i.e., the product of its mass, its velocity, and its distance of closest approach to the
point about which the angular momentum is measured—this is a general result
(for a point particle). After the collision, the bullet lodges a distance d from the
pivot, and is forced to co-rotate with the rod. Hence, the angular momentum of
the bullet after the collision is given by
l  = m d2 ω, (9.39)
where ω is the angular velocity of the rod. The angular momentum of the rod
after the collision is
L = I ω, (9.40)

213
9 ANGULAR MOMENTUM 9.4 Angular momentum of a multi-component system

where I = (1/12) M (2 b)2 = (1/3) M b2 is the rod’s moment of inertia (about

a perpendicular axis passing through its mid-point). Conservation of angular
momentum yields
l = l  + L, (9.41)
or
mvd
ω= . (9.42)
I + m d2

Worked example 9.1: Angular momentum of a missile

Question: A missile of mass m = 2.3×104 kg flies level to the ground at an altitude

of d = 10, 000 m with constant speed v = 210 m/s. What is the magnitude of the
missile’s angular momentum relative to a point on the ground directly below its
flight path?

θ
v

d r r

O ground

Answer: The missile’s angular momentum about point O is

L = m v r sin θ,
where θ is the angle subtended between the missile’s velocity vector and its posi-
tion vector relative to O. However,
r sin θ = d,
where d is the distance of closest approach of the missile to point O. Hence,
L = m v d = (2.3 × 104 ) × 210 × (1 × 104 ) = 4.83 × 1010 kg m2 /s.

214
9 ANGULAR MOMENTUM 9.4 Angular momentum of a multi-component system

Worked example 9.2: Angular momentum of a sphere

Question: A uniform sphere of mass M = 5 kg and radius a = 0.2 m spins about

an axis passing through its centre with period T = 0.7 s. What is the angular
momentum of the sphere?

Answer: The angular velocity of the sphere is

2π 2π
ω= = = 8.98 rad./s.
T 0.7
The moment of inertia of the sphere is
2
I = M a2 = 0.4 × 5 × (0.2)2 = 0.08 kg m2 .
5
Hence, the angular momentum of the sphere is
L = I ω = 0.08 × 8.98 = 0.718 kg m2 /s.

Worked example 9.3: Spinning skater

Question: A skater spins at an initial angular velocity of ω1 = 11 rad./s with her

arms outstretched. The skater then lowers her arms, thereby decreasing her mo-
ment of inertia by a factor 8. What is the skater’s final angular velocity? Assume
that any friction between the skater’s skates and the ice is negligible.

Answer: Neglecting any friction between the skates and the ice, we expect the
skater to spin with constant angular momentum. The skater’s initial angular
momentum is
L1 = I1 ω1 ,
where I1 is the skater’s initial moment of inertia. The skater’s final angular mo-
mentum is
L2 = I2 ω2 ,
where I2 is the skater’s final moment of inertia, and ω2 is her final angular veloc-
ity. Conservation of angular momentum yields L1 = L2 , or
I1
ω2 = ω2 .
I2
215
9 ANGULAR MOMENTUM 9.4 Angular momentum of a multi-component system

Now, we are told that I1 /I2 = 8. Hence,

ω2 = 8 × 11 = 88 rad./s.

216
10 STATICS

10 Statics

10.1 Introduction

Probably the most useful application of the laws of mechanics is the study of situ-
ations in which nothing moves—this discipline is known as statics. The principles
of statics are employed by engineers whenever they design stationary structures,
such as buildings, bridges, and tunnels, in order to ensure that these structures
do not collapse.

10.2 The principles of statics

Consider a general extended body which is subject to a number of external forces.

Let us model this body as a swarm of N point particles. In the limit that N → ∞,
this model becomes a fully accurate representation of the body’s dynamics.

In Sect. 6.3 we determined that the overall translational equation of motion of

a general N-component system can be written in the form
dP
= F. (10.1)
dt
Here, P is the total linear momentum of the system, and

F= Fi (10.2)
i=1,N

is the resultant of all the external forces acting on the system. Note that F i is the
external force acting on the ith component of the system.

Equation (10.1) effectively determines the translational motion of the system’s

centre of mass. Note, however, that in order to fully determine the motion of the
system we must also follow its rotational motion about its centre of mass (or any
other convenient reference point). In Sect. 9.4 we determined that the overall
rotational equation of motion of a general N-component system (with central

217
10 STATICS 10.2 The principles of statics

internal forces) can be written in the form

dL
= τ. (10.3)
dt
Here, L is the total angular momentum of the system (about the origin of our
coordinate scheme), and 
τ= ri × Fi (10.4)
i=1,N
is the resultant of all the external torques acting on the system (about the origin
of our coordinate scheme). In the above, ri is the vector displacement of the ith
component of the system.

What conditions must be satisfied by the various external forces and torques
acting on the system if it is to remain stationary in time? Well, if the system
does not evolve in time then its net linear momentum, P, and its net angular
momentum, L, must both remain constant. In other words, dP/dt = dL/dt = 0.
It follows from Eqs. (10.1) and (10.3) that

F = 0, (10.5)
τ = 0. (10.6)

In other words, the net external force acting on system must be zero, and the net
external torque acting on the system must be zero. To be more exact:

The components of the net external force acting along any three independent
directions must all be zero;

and

The magnitudes of the net external torques acting about any three indepen-
dent axes (passing through the origin of the coordinate system) must all be
zero.

In a nutshell, these are the principles of statics.

218
10 STATICS 10.2 The principles of statics

It is clear that the above principles are necessary conditions for a general phys-
ical system not to evolve in time. But, are they also sufficient conditions? In other
words, is it necessarily true that a general system which satisfies these conditions
does not exhibit any time variation? The answer to this question is as follows: if
the system under investigation is a rigid body, such that the motion of any com-
ponent of the body necessarily implies the motion of the whole body, then the
above principles are necessary and sufficient conditions for the existence of an
equilibrium state. On the other hand, if the system is not a rigid body, so that
some components of the body can move independently of others, then the above
conditions only guarantee that the system remains static in an average sense.

Before we attempt to apply the principles of statics, there are a couple of

important points which need clarification. Firstly, does it matter about which
point we calculate the net torque acting on the system? To be more exact, if
we determine that the net torque acting about a given point is zero does this
necessarily imply that the net torque acting about any other point is also zero?
Well, 
τ= ri × Fi (10.7)
i=1,N
is the net torque acting on the system about the origin of our coordinate scheme.
The net torque about some general point r0 is simply

τ = (ri − r0 ) × Fi . (10.8)
i=1,N

However, we can rewrite the above expression as

⎛ ⎞
 ⎜

τ = ri × Fi − r0 × ⎝ Fi ⎟
⎠ = τ + r0 × F. (10.9)
i=1,N i=1,N

Now, if the system is in equilibrium then F = τ = 0. Hence, it follows from the

above equation that
τ  = 0. (10.10)
In other words, for a system in equilibrium, the determination that the net torque
acting about a given point is zero necessarily implies that the net torque acting
about any other point is also zero. Hence, we can choose the point about which

219
10 STATICS 10.3 Equilibrium of a laminar object in a gravitational field

we calculate the net torque at will—this choice is usually made so as to simplify

the calculation.

Another question which needs clarification is as follows. At which point should

we assume that the weight of the system acts in order to calculate the contribu-
tion of the weight to the net torque acting about a given point? Actually, in
Sect. 8.11, we effectively answered this question by assuming that the weight
acts at the centre of mass of the system. Let us now justify this assumption. The
external force acting on the ith component of the system due to its weight is

Fi = mi g, (10.11)

where g is the acceleration due to gravity (which is assumed to be uniform

throughout the system). Hence, the net gravitational torque acting on the system
about the origin of our coordinate scheme is
⎛ ⎞
 ⎜

τ= ri × mi g = ⎝ mi ri ⎟
⎠ × g = rcm × M g, (10.12)
i=1,N i=1,N
 
where M = i=1,N mi is the total mass of the system, and rcm = i=1,N mi ri /M
is the position vector of its centre of mass. It follows, from the above equation,
that the net gravitational torque acting on the system about a given point can be
calculated by assuming that the total mass of the system is concentrated at its
centre of mass.

10.3 Equilibrium of a laminar object in a gravitational field

Consider a general laminar object which is free to pivot about a fixed perpendic-
ular axis. Assuming that the object is placed in a uniform gravitational field (such
as that on the surface of the Earth), what is the object’s equilibrium configuration
in this field?

Let O represent the pivot point, and let C be the centre of mass of the ob-
ject. See Fig. 90. Suppose that r represents the distance between points O and
C, whereas θ is the angle subtended between the line OC and the downward

220
10 STATICS 10.3 Equilibrium of a laminar object in a gravitational field

d
O

θ r
h

Mg

Figure 90: A laminar object pivoting about a fixed point in a gravitational field.

vertical. There are two external forces acting on the object. First, there is the
downward force, M g, due to gravity, which acts at the centre of mass. Second,
there is the reaction, R, due to the pivot, which acts at the pivot point. Here, M
is the mass of the object, and g is the acceleration due to gravity.

Two conditions must be satisfied in order for a given configuration of the object
shown in Fig. 90 to represent an equilibrium configuration. First, there must be
zero net external force acting on the object. This implies that the reaction, R, is
equal and opposite to the gravitational force, M g. In other words, the reaction
is of magnitude M g and is directed vertically upwards. The second condition
is that there must be zero net torque acting about the pivot point. Now, the
reaction, R, does not generate a torque, since it acts at the pivot point. Moreover,
the torque associated with the gravitational force, M g, is simply the magnitude
of this force times the length of the lever arm, d (see Fig. 90). Hence, the net
torque acting on the system about the pivot point is

τ = M g d = M g r sin θ. (10.13)

Setting this torque to zero, we obtain sin θ = 0, which implies that θ = 0 ◦ . In

other words, the equilibrium configuration of a general laminar object (which is
free to rotate about a fixed perpendicular axis in a uniform gravitational field) is
that in which the centre of mass of the object is aligned vertically below the pivot
point.

221
10 STATICS 10.3 Equilibrium of a laminar object in a gravitational field

Incidentally, we can use the above result to experimentally determine the cen-
tre of mass of a given laminar object. We would need to suspend the object from
two different pivot points, successively. In each equilibrium configuration, we
would mark a line running vertically downward from the pivot point, using a
plumb-line. The crossing point of these two lines would indicate the position of
the centre of mass.

Our discussion of the equilibrium configuration of the laminar object shown in

Fig. 90 is not quite complete. We have determined that the condition which must
be satisfied by an equilibrium state is sin θ = 0. However, there are, in fact, two
physical roots of this equation. The first, θ = 0◦ , corresponds to the case where
the centre of mass of the object is aligned vertically below the pivot point. The
second, θ = 180◦ , corresponds to the case where the centre of mass is aligned
vertically above the pivot point. Of course, the former root is far more important
than the latter, since the former root corresponds to a stable equilibrium, whereas
the latter corresponds to an unstable equilibrium. We recall, from Sect. 5.7, that
when a system is slightly disturbed from a stable equilibrium then the forces and
torques which act upon it tend to return it to this equilibrium, and vice versa for an
unstable equilibrium. The easiest way to distinguish between stable and unstable
equilibria, in the present case, is to evaluate the gravitational potential energy of
the system. The potential energy of the object shown in Fig. 90, calculated using
the height of the pivot as the reference height, is simply

U = −M g h = −M g r cos θ. (10.14)

(Note that the gravitational potential energy of an extended object can be calcu-
lated by imagining that all of the mass of the object is concentrated at its centre
of mass.) It can be seen that θ = 0◦ corresponds to a minimum of this poten-
tial, whereas θ = 180◦ corresponds to a maximum. This is in accordance with
Sect. 5.7, where it was demonstrated that whenever an object moves in a con-
servative force-field (such as a gravitational field), the stable equilibrium points
correspond to minima of the potential energy associated with this field, whereas
the unstable equilibrium points correspond to maxima.

222
10 STATICS 10.4 Rods and cables

10.4 Rods and cables

Consider a uniform rod of mass M and length l which is suspended horizontally

via two vertical cables. Let the points of attachment of the two cables be located
distances x1 and x2 from one of the ends of the rod, labeled A. It is assumed that
x2 > x1 . See Fig. 91. What are the tensions, T1 and T2 , in the cables?

Let us first locate the centre of mass of the rod, which is situated at the rod’s
mid-point, a distance l/2 from reference point A (see Fig. 91). There are three
forces acting on the rod: the gravitational force, M g, and the two tension forces,
T1 and T2 . Each of these forces is directed vertically. Thus, the condition that zero
net force acts on the system reduces to the condition that the net vertical force is
zero, which yields
T1 + T2 − M g = 0. (10.15)

Consider the torques exerted by the three above-mentioned forces about point
A. Each of these torques attempts to twist the rod about an axis perpendicular
to the plane of the diagram. Hence, the condition that zero net torque acts on
the system reduces to the condition that the net torque at point A, about an
axis perpendicular to the plane of the diagram, is zero. The contribution of each
force to this torque is simply the product of the magnitude of the force and the
length of the associated lever arm. In each case, the length of the lever arm is
equivalent to the distance of the point of action of the force from A, measured
along the length of the rod. Hence, setting the net torque to zero, we obtain
l
x 1 T1 + x 2 T2 − M g = 0. (10.16)
2
Note that the torque associated with the gravitational force, M g, has a minus sign
in front, because this torque obviously attempts to twist the rod in the opposite
direction to the torques associated with the tensions in the cables.

The previous two equations can be solved to give

⎛ ⎞
x2 − l/2 ⎠
T1 = ⎝ M g, (10.17)
x2 − x 1

223
10 STATICS 10.4 Rods and cables

x2

x1 T1 T2
A
l/2

Mg

Figure 91: A horizontal rod suspended by two vertical cables.

⎛ ⎞
l/2 − x1 ⎠
T2 = ⎝ M g. (10.18)
x2 − x 1
Recall that tensions in flexible cables can never be negative, since this would
imply that the cables in question were being compressed. Of course, when cables
are compressed they simply collapse. It is clear, from the above expressions, that
in order for the tensions T1 and T2 to remain positive (given that x2 > x1 ), the
following conditions must be satisfied:
l
x1 < , (10.19)
2
l
x2 > . (10.20)
2
In other words, the attachment points of the two cables must straddle the centre
of mass of the rod.

Consider a uniform rod of mass M and length l which is free to rotate in the
vertical plane about a fixed pivot attached to one of its ends. The other end of
the rod is attached to a fixed cable. We can imagine that both the pivot and the
cable are anchored in the same vertical wall. See Fig. 92. Suppose that the rod is
level, and that the cable subtends an angle θ with the horizontal. Assuming that
the rod is in equilibrium, what is the magnitude of the tension, T , in the cable,
and what is the direction and magnitude of the reaction, R, at the pivot?

224
10 STATICS 10.4 Rods and cables

wall
cable

pivot T
R
φ θ

l
Mg
rod

Figure 92: A rod suspended by a fixed pivot and a cable.

As usual, the centre of mass of the rod lies at its mid-point. There are three
forces acting on the rod: the reaction, R; the weight, M g; and the tension, T .
The reaction acts at the pivot. Let φ be the angle subtended by the reaction with
the horizontal, as shown in Fig. 92. The weight acts at the centre of mass of the
rod, and is directed vertically downwards. Finally, the tension acts at the end of
the rod, and is directed along the cable.

Resolving horizontally, and setting the net horizontal force acting on the rod
to zero, we obtain
R cos φ − T cos θ = 0. (10.21)
Likewise, resolving vertically, and setting the net vertical force acting on the rod
to zero, we obtain
R sin φ + T sin θ − M g = 0. (10.22)
The above constraints are sufficient to ensure that zero net force acts on the rod.

Let us evaluate the net torque acting at the pivot point (about an axis perpen-
dicular to the plane of the diagram). The reaction, R, does not contribute to this
torque, since it acts at the pivot point. The length of the lever arm associated
with the weight, M g, is l/2. Simple trigonometry reveals that the length of the
lever arm associated with the tension, T , is l sin θ. Hence, setting the net torque

225
10 STATICS 10.5 Ladders and walls

about the pivot point to zero, we obtain

l
Mg − T l sin θ = 0. (10.23)
2
Note that there is a minus sign in front of the second torque, since this torque
clearly attempts to twist the rod in the opposite sense to the first.

Equations (10.21) and (10.22) can be solved to give

cos φ
T = M g, (10.24)
sin(θ + φ)
cos θ
R = M g. (10.25)
sin(θ + φ)
Substituting Eq. (10.24) into Eq. (10.23), we obtain
sin(θ + φ) = 2 sin θ cos φ. (10.26)
The physical solution of this equation is φ = θ (recall that sin 2 θ = 2 sin θ cos θ),
which determines the direction of the reaction at the pivot. Finally, Eqs. (10.24)
and (10.25) yield
Mg
T =R= , (10.27)
2 sin θ
which determines both the magnitude of the tension in the cable and that of the
reaction at the pivot.

One important point to note about the above solution is that if φ = θ then the
lines of action of the three forces—R, M g, and T —intersect at the same point,
as shown in Fig. 92. This is an illustration of a general rule. Namely, whenever a
rigid body is in equilibrium under the action of three forces, then these forces are
either mutually parallel, as shown in Fig. 91, or their lines of action pass through
the same point, as shown in Fig. 92.

10.5 Ladders and walls

Suppose that a ladder of length l and negligible mass is leaning against a vertical
wall, making an angle θ with the horizontal. A workman of mass M climbs

226
10 STATICS 10.5 Ladders and walls

wall

S ladder

workman

x R
l
Mg ground

θ
f

Figure 93: A ladder leaning against a vertical wall.

a distance x along the ladder, measured from the bottom. See Fig. 93. Suppose
that the wall is completely frictionless, but that the ground possesses a coefficient
of static friction μ. How far up the ladder can the workman climb before it slips
along the ground? Is it possible for the workman to climb to the top of the ladder
without any slippage occurring?

There are four forces acting on the ladder: the weight, M g, of the workman;
the reaction, S, at the wall; the reaction, R, at the ground; and the frictional
force, f, due to the ground. The weight acts at the position of the workman, and
is directed vertically downwards. The reaction, S, acts at the top of the ladder,
and is directed horizontally (i.e., normal to the surface of the wall). The reaction,
R, acts at the bottom of the ladder, and is directed vertically upwards (i.e., normal
to the ground). Finally, the frictional force, f, also acts at the bottom of the ladder,
and is directed horizontally.

Resolving horizontally, and setting the net horizontal force acting on the ladder
to zero, we obtain
S − f = 0. (10.28)
Resolving vertically, and setting the net vertically force acting on the ladder to
zero, we obtain
R − M g = 0. (10.29)
Evaluating the torque acting about the point where the ladder touches the ground,
we note that only the forces M g and S contribute. The lever arm associated with
the force M g is x cos θ. The lever arm associated with the force S is l sin θ. Fur-

227
10 STATICS 10.6 Jointed rods

thermore, the torques associated with these two forces act in opposite directions.
Hence, setting the net torque about the bottom of the ladder to zero, we obtain

M g x cos θ − S l sin θ = 0. (10.30)

The above three equations can be solved to give

R = M g, (10.31)

and
x
f=S= M g. (10.32)
l tan θ

Now, the condition for the ladder not to slip with respect to the ground is

f < μ R. (10.33)

This condition reduces to

x < l μ tan θ. (10.34)
Thus, the furthest distance that the workman can climb along the ladder before
it slips is
xmax = l μ tan θ. (10.35)
Note that if tan θ > 1/μ then the workman can climb all the way along the ladder
without any slippage occurring. This result suggests that ladders leaning against
walls are less likely to slip when they are almost vertical (i.e., when θ → 90 ◦ ).

10.6 Jointed rods

Suppose that three identical uniform rods of mass M and length l are joined
together to form an equilateral triangle, and are then suspended from a cable, as
shown in Fig. 94. What is the tension in the cable, and what are the reactions at
the joints?

Let X1 , X2 , and X3 be the horizontal reactions at the three joints, and let Y1 , Y2 ,
and Y3 be the corresponding vertical reactions, as shown in Fig. 94. In drawing
this diagram, we have made use of the fact that the rods exert equal and opposite

228
10 STATICS 10.6 Jointed rods

cable
Y1 Y3
T
A l B
X1 X3
X1 X3
θ θ

Mg
Y1 Y3
Y2
rod θ
Mg Mg
X2 X2
C

Y2

Figure 94: Three identical jointed rods.

reactions on one another, in accordance with Newton’s third law. Let T be the
tension in the cable.

Setting the horizontal and vertical forces acting on rod AB to zero, we obtain
X1 − X3 = 0, (10.36)
T + Y1 + Y3 − M g = 0, (10.37)
respectively. Setting the horizontal and vertical forces acting on rod AC to zero,
we obtain
X2 − X1 = 0, (10.38)
Y2 − Y1 − M g = 0, (10.39)
respectively. Finally, setting the horizontal and vertical forces acting on rod BC
to zero, we obtain
X3 − X2 = 0, (10.40)
−Y2 − Y3 − M g = 0, (10.41)
respectively. Incidentally, it is clear, from symmetry, that X1 = X3 and Y1 = Y3 .
Thus, the above equations can be solved to give
T = 3 M g, (10.42)

229
10 STATICS 10.6 Jointed rods

Y2 = 0, (10.43)
X1 = X2 = X3 = X, (10.44)
Y1 = Y3 = −M g. (10.45)

There now remains only one unknown, X.

Now, it is clear, from symmetry, that there is zero net torque acting on rod AB.
Let us evaluate the torque acting on rod AC about point A. (By symmetry, this is
the same as the torque acting on rod BC about point B). The two forces which
contribute to this torque are the weight, M g, and the reaction X2 = X. (Recall
that the reaction Y2 is zero). The lever arms associated with these two torques
(which act in the same direction) are (l/2) cos θ and l sin θ, respectively. Thus,
setting the net torque to zero, we obtain

M g (l/2) cos θ + X l sin θ = 0, (10.46)

which yields
Mg Mg
X=− =− √ , (10.47)
2 tan θ 2 3
√
since θ = 60◦ , and tan 60◦ = 3. We have now fully determined the tension in
the cable, and all the reactions at the joints.

Worked example 10.1: Equilibrium of two rods

Question: Suppose that two uniform rods (of negligible thickness) are welded
together at right-angles, as shown in the diagram below. Let the first rod be
of mass m1 = 5.2 kg and length l1 = 1.3 m. Let the second rod be of mass
m2 = 3.4 kg and length l2 = 0.7 m. Suppose that the system is suspended from
a pivot point located at the free end of the first rod, and then allowed to reach
a stable equilibrium state. What angle θ does the first rod subtend with the
downward vertical in this state?

Answer: Let us adopt a coordinate system in which the x-axis runs parallel to the
second rod, whereas the y-axis runs parallel to the first. Let the origin of our

230
10 STATICS 10.6 Jointed rods

pivot
x

y
l1

l2

coordinate system correspond to the pivot point. The centre of mass of the first
rod is situated at its mid-point, whose coordinates are

(x1 , y1 ) = (0, l1 /2).

Likewise, the centre of mass of the second rod is situated at its mid-point, whose
coordinates are
(x2 , y2 ) = (l2 /2, l1 ).
It follows that the coordinates of the centre of mass of the whole system are given
by
m1 x1 + m2 x2 1 m2 l2 3.4 × 0.7
xcm = = = = 0.138 m,
m1 + m2 2 m1 + m2 2 × 8.6
and
m1 y1 + m2 y2 m1 l1 /2 + m2 l1 5.2 × 1.3/2 + 3.4 × 1.3
ycm = = = = 0.907 m.
m1 + m2 m1 + m2 8.6
The angle θ subtended between the line joining the pivot point and the overall
centre of mass, and the first rod is simply
 
−1 xcm
θ = tan = tan−1 0.152 = 8.65◦ .
ycm
When the system reaches a stable equilibrium state then its centre of mass is
aligned directly below the pivot point. This implies that the first rod subtends an
angle θ = 8.65◦ with the downward vertical.

231
10 STATICS 10.6 Jointed rods

Worked example 10.2: Rod supported by a cable

Question: A uniform rod of mass m = 15 kg and length l = 3 m is supported in

a horizontal position by a pin and a cable, as shown in the figure below. Masses
m1 = 36 kg and m2 = 24 kg are suspended from the rod at positions l1 = 0.5 m
and l2 = 2.3 m. The angle θ is 40◦ . What is the tension T in the cable?

rod θ

pin l1
cable
l2
m1 m2
l

Answer: Consider the torque acting on the rod about the pin. Note that the
reaction at the pin makes no contribution to this torque (since the length of the
associated lever arm is zero). The torque due to the weight of the rod is m g l/2
(i.e., the weight times the length of the lever arm). Note that the weight of the
rod acts at its centre of mass, which is located at the rod’s mid-point. The torque
due to the weight of the first mass is m1 g l1 . The torque due to the weight of
the second mass is m2 g l2 . Finally, the torque due to the tension in the cable is
−T l sin θ (this torque is negative since it twists the rod in the opposite sense to
the other three torques). Hence, setting the net torque to zero, we obtain
l
mg + m1 g l1 + m2 g l2 − T l sin θ = 0,
2
or
[m/2 + m1 (l1 /l) + m2 (l2 /l)] g
T =
sin θ
[0.5 × 15 + 36 × (0.5/3) + 24 × (2.3/3)] × 9.81
=
sin 40◦
= 486.84 N.

232
10 STATICS 10.6 Jointed rods

Worked example 10.3: Leaning ladder

Question: A uniform ladder of mass m = 40 kg and length l = 10 m is leaned

against a smooth vertical wall. A person of mass M = 80 kg stands on the ladder
a distance x = 7 m from the bottom, as measured along the ladder. The foot of
the ladder is d = 1.2 m from the bottom of the wall. What is the force exerted
by the wall on the ladder? What is the normal force exerted by the floor on the
ladder?
wall

S ladder

person

l x R
Mg ground

mg θ
f

Answer: The angle θ subtended by the ladder with the ground satisfies
θ = cos−1 (d/l) = cos−1 (1.2/10) = 83.11◦ .
Let S be the normal reaction at the wall, let R be the normal reaction at the
ground, and let f be the frictional force exerted by the ground on the ladder,
as shown in the diagram. Consider the torque acting on the ladder about the
point where it meets the ground. Only three forces contribute to this torque:
the weight, m g, of the ladder, which acts half-way along the ladder; the weight,
M g, of the person, which acts a distance x along the ladder; and the reaction, S,
at the wall, which acts at the top of the ladder. The lever arms associated with
these three forces are (l/2) cos θ, x cos θ, and l sin θ, respectively. Note that the
reaction force acts to twist the ladder in the opposite sense to the two weights.
Hence, setting the net torque to zero, we obtain
l
mg cos θ + M g x cos θ − S l sin θ = 0,
2
233
10 STATICS 10.6 Jointed rods

which yields
(m g/2 + M g x/l) (0.5 × 40 × 9.81 + 80 × 9.81 × 7/10)
S= = = 90.09 N.
tan θ tan 83.11◦

The condition that zero net vertical force acts on the ladder yields

R − m g − M g = 0.

Hence,
R = (m + M) g = (40 + 80) × 9.81 = 1177.2 N.

Worked example 10.4: Truck crossing a bridge

Question: A truck of mass M = 5000 kg is crossing a uniform horizontal bridge

of mass m = 1000 kg and length l = 100 m. The bridge is supported at its two
end-points. What are the reactions at these supports when the truck is one third
of the way across the bridge?
R truck S
l/3

Mg mg bridge

Answer: Let R and S be the reactions at the bridge supports. Here, R is the
reaction at the support closest to the truck. Setting the net vertical force acting
on the bridge to zero, we obtain

R + S − M g − m g = 0.

Setting the torque acting on the bridge about the left-most support to zero, we
get
M g l/3 + m g l/2 − S l = 0.

234
10 STATICS 10.6 Jointed rods

Here, we have made use of the fact that centre of mass of the bridge lies at its
mid-point. It follows from the above two equations that
S = M g/3 + m g/2 = 5000 × 9.81/3 + 1000 × 9.81/2 = 2.13 × 104 N,
and
R = M g + m g − S = (5000 + 1000) × 9.81 − 2.13 × 104 = 3.76 × 104 N.

Worked example 10.5: Rod supported by a strut

Question: A uniform horizontal rod of mass m = 15 kg is attached to a vertical

wall at one end, and is supported, from below, by a light rigid strut at the other.
The strut is attached to the rod at one end, and the wall at the other, and subtends
an angle of θ = 30◦ with the rod. Find the horizontal and vertical reactions at the
point where the strut is attached to the rod, and the points where the rod and the
strut are attached to the wall.
X1 rod X3
Y1
Y3
θ

X2 mg Y3 X3

Y2 strut

wall

Answer: Let us call the vertical reactions at the joints X1 , X2 , and X3 . Let the
corresponding horizontal reactions be Y1 , Y2 , and Y3 . See the diagram. Here, we
have made use of the fact that the strut and the rod exert equal and opposite
reactions on one another, in accordance with Newton’s third law. Setting the net
vertical force on the rod to zero yields
X1 + X3 − m g = 0.

235
10 STATICS 10.6 Jointed rods

Setting the net horizontal force on the rod to zero gives

Y1 + Y3 = 0.
Setting the net vertical force on the strut to zero yields
X2 − X3 = 0.
Finally, setting the net horizontal force on the strut to zero yields
Y2 − Y3 = 0.
The above equations can be solved to give
−Y1 = Y2 = Y3 = Y,
and
X2 = X3 = X,
with
X1 = m g − X.
There now remain only two unknowns, X and Y.

Setting the net torque acting on the rod about the point where it is connected
to the wall to zero, we obtain
m g l/2 − X3 l = 0,
where l is the length of the rod. Here, we have used the fact that the centre of
gravity of the rod lies at its mid-point. The above equation implies that
X3 = X = m g/2 = 15 × 9.81/2 = 73.58 N.
We also have X1 = m g − X = 73.58 N. Setting the net torque acting on the strut
about the point where it is connected to the wall to zero, we find
Y3 h sin θ − X3 h cos θ = 0,
where h is the length of the strut. Thus,
X 73.58
Y3 = Y = = = 127.44 N.
tan θ tan 30◦

In summary, the vertical reactions are X1 = X2 = X3 = 73.58 N, and the hori-

zontal reactions are −Y1 = Y2 = Y3 = 127.44 N.

236
11 OSCILLATORY MOTION

11 Oscillatory motion

11.1 Introduction

We have seen previously (for instance, in Sect. 10.3) that when systems are per-
turbed from a stable equilibrium state they experience a restoring force which acts
to return them to that state. In many cases of interest, the magnitude of the
restoring force is directly proportional to the displacement from equilibrium. In
this section, we shall investigate the motion of systems subject to such a force.

11.2 Simple harmonic motion

Let us reexamine the problem of a mass on a spring (see Sect. 5.6). Consider
a mass m which slides over a horizontal frictionless surface. Suppose that the
mass is attached to a light horizontal spring whose other end is anchored to an
immovable object. See Fig. 42. Let x be the extension of the spring: i.e., the dif-
ference between the spring’s actual length and its unstretched length. Obviously,
x can also be used as a coordinate to determine the horizontal displacement of
the mass.

The equilibrium state of the system corresponds to the situation where the
mass is at rest, and the spring is unextended (i.e., x = 0). In this state, zero net
force acts on the mass, so there is no reason for it to start to move. If the system
is perturbed from this equilibrium state (i.e., if the mass is moved, so that the
spring becomes extended) then the mass experiences a restoring force given by
Hooke’s law:
f = −k x. (11.1)
Here, k > 0 is the force constant of the spring. The negative sign indicates that
f is indeed a restoring force. Note that the magnitude of the restoring force
is directly proportional to the displacement of the system from equilibrium (i.e.,
f ∝ x). Of course, Hooke’s law only holds for small spring extensions. Hence,
the displacement from equilibrium cannot be made too large. The motion of this

237
11 OSCILLATORY MOTION 11.2 Simple harmonic motion

system is representative of the motion of a wide range of systems when they are
slightly disturbed from a stable equilibrium state.

Newton’s second law gives following equation of motion for the system:

m ẍ = −k x. (11.2)

This differential equation is known as the simple harmonic equation, and its solu-
tion has been known for centuries. In fact, the solution is

x = a cos(ω t − φ), (11.3)

where a, ω, and φ are constants. We can demonstrate that Eq. (11.3) is in-
deed a solution of Eq. (11.2) by direct substitution. Substituting Eq. (11.3) into
Eq. (11.2), and recalling from calculus that d(cos θ)/dθ = − sin θ and d(sin θ)/dθ =
cos θ, we obtain

− m ω2 a cos(ω t − φ) = −k a cos(ω t − φ). (11.4)

It follows that Eq. (11.3) is the correct solution provided



 k

ω= . (11.5)
m

Figure 95 shows a graph of x versus t obtained from Eq. (11.3). The type
of motion shown here is called simple harmonic motion. It can be seen that the
displacement x oscillates between x = −a and x = +a. Here, a is termed the
amplitude of the oscillation. Moreover, the motion is periodic in time (i.e., it
repeats exactly after a certain time period has elapsed). In fact, the period is
2π
T= . (11.6)
ω
This result is easily obtained from Eq. (11.3) by noting that cos θ is a periodic
function of θ with period 2 π. The frequency of the motion (i.e., the number of
oscillations completed per second) is
1 ω
f= = . (11.7)
T 2π
238
11 OSCILLATORY MOTION 11.2 Simple harmonic motion

ωt − φ 0◦ 90◦ 180◦ 270◦

x +a 0 −a 0
ẋ 0 −ω a 0 +ω a
ẍ −ω a
2
0 +ω a2
0

Table 4: Simple harmonic motion.

Figure 95: Simple harmonic motion.

It can be seen that ω is the motion’s angular frequency (i.e., the frequency f
converted into radians per second). Finally, the phase angle φ determines the
times at which the oscillation attains its maximum amplitude, x = a: in fact,
 
φ
tmax =T n+ . (11.8)
2π
Here, n is an arbitrary integer.

Table 4 lists the displacement, velocity, and acceleration of the mass at various
phases of the simple harmonic cycle. The information contained in this table can
easily be derived from the simple harmonic equation, Eq. (11.3). Note that all
of the non-zero values shown in this table represent either the maximum or the
minimum value taken by the quantity in question during the oscillation cycle.

We have seen that when a mass on a spring is disturbed from equilibrium it

239
11 OSCILLATORY MOTION 11.2 Simple harmonic motion

executes simple harmonic motion about its equilibrium state. In physical terms,
if the initial displacement is positive (x > 0) then the restoring force overcom-
pensates, and sends the system past the equilibrium state (x = 0) to negative
displacement states (x < 0). The restoring force again overcompensates, and
sends the system back through x = 0 to positive displacement states. The motion
then repeats itself ad infinitum. The frequency of the oscillation is determined by
the spring stiffness, k, and the system inertia, m, via Eq. (11.5). In contrast, the
amplitude and phase angle of the oscillation are determined by the initial condi-
tions. Suppose that the instantaneous displacement and velocity of the mass at
t = 0 are x0 and v0 , respectively. It follows from Eq. (11.3) that

x0 = x(t = 0) = a cos φ, (11.9)

v0 = ẋ(t = 0) = a ω sin φ. (11.10)

Here, use has been made of the well-known identities cos(−θ) = cos θ and
sin(−θ) = − sin θ. Hence, we obtain

a = x02 + (v0 /ω)2 , (11.11)

and  
−1 v0
φ = tan , (11.12)
ω x0
since sin2 θ + cos2 θ = 1 and tan θ = sin θ/ cos θ.

The kinetic energy of the system is written

1 2 m a2 ω2 sin2 (ω t − φ)
K = m ẋ = . (11.13)
2 2
Recall, from Sect. 5.6, that the potential energy takes the form
1 2 k a2 cos2 (ω t − φ)
U= kx = . (11.14)
2 2
Hence, the total energy can be written
a2 k
E=K+U= , (11.15)
2

240
11 OSCILLATORY MOTION 11.3 The torsion pendulum

fixed support

torsion wire

disk

Figure 96: A torsion pendulum.

since m ω2 = k and sin2 θ + cos2 θ = 1. Note that the total energy is a constant of
the motion, as expected for an isolated system. Moreover, the energy is propor-
tional to the amplitude squared of the motion. It is clear, from the above expres-
sions, that simple harmonic motion is characterized by a constant backward and
forward flow of energy between kinetic and potential components. The kinetic
energy attains its maximum value, and the potential energy attains it minimum
value, when the displacement is zero (i.e., when x = 0). Likewise, the potential
energy attains its maximum value, and the kinetic energy attains its minimum
value, when the displacement is maximal (i.e., when x = ±a). Note that the
minimum value of K is zero, since the system is instantaneously at rest when the
displacement is maximal.

11.3 The torsion pendulum

Consider a disk suspended from a torsion wire attached to its centre. See Fig. 96.
This setup is known as a torsion pendulum. A torsion wire is essentially inexten-
sible, but is free to twist about its axis. Of course, as the wire twists it also causes
the disk attached to it to rotate in the horizontal plane. Let θ be the angle of
rotation of the disk, and let θ = 0 correspond to the case in which the wire is
untwisted.

Any twisting of the wire is inevitably associated with mechanical deformation.

The wire resists such deformation by developing a restoring torque, τ, which acts

241
11 OSCILLATORY MOTION 11.4 The simple pendulum

to restore the wire to its untwisted state. For relatively small angles of twist, the
magnitude of this torque is directly proportional to the twist angle. Hence, we
can write
τ = −k θ, (11.16)
where k > 0 is the torque constant of the wire. The above equation is essentially
a torsional equivalent to Hooke’s law. The rotational equation of motion of the
system is written
I θ̈ = τ, (11.17)
where I is the moment of inertia of the disk (about a perpendicular axis through
its centre). The moment of inertia of the wire is assumed to be negligible. Com-
bining the previous two equations, we obtain
I θ̈ = −k θ. (11.18)

Equation (11.18) is clearly a simple harmonic equation [cf., Eq. (11.2)]. Hence,
we can immediately write the standard solution [cf., Eq. (11.3)]
θ = a cos(ω t − φ), (11.19)
where [cf., Eq. (11.5)] 

k

ω= . (11.20)
I
We conclude that when a torsion pendulum is perturbed from its equilibrium state
(i.e., θ = 0), it executes torsional oscillations about this state at a fixed frequency,
ω, which depends only on the torque constant of the wire and the moment of
inertia of the disk. Note, in particular, that the frequency is independent of the
amplitude of the oscillation [provided θ remains small enough that Eq. (11.16)
still applies]. Torsion pendulums are often used for time-keeping purposes. For
instance, the balance wheel in a mechanical wristwatch is a torsion pendulum in
which the restoring torque is provided by a coiled spring.

11.4 The simple pendulum

Consider a mass m suspended from a light inextensible string of length l, such

that the mass is free to swing from side to side in a vertical plane, as shown in

242
11 OSCILLATORY MOTION 11.4 The simple pendulum

fixed support

pivot point
l
θ
T

mg

Figure 97: A simple pendulum.

Fig. 97. This setup is known as a simple pendulum. Let θ be the angle subtended
between the string and the downward vertical. Obviously, the equilibrium state of
the simple pendulum corresponds to the situation in which the mass is stationary
and hanging vertically down (i.e., θ = 0). The angular equation of motion of the
pendulum is simply
I θ̈ = τ, (11.21)
where I is the moment of inertia of the mass, and τ is the torque acting on the
system. For the case in hand, given that the mass is essentially a point particle,
and is situated a distance l from the axis of rotation (i.e., the pivot point), it is
easily seen that I = m l2 .

The two forces acting on the mass are the downward gravitational force, m g,
and the tension, T , in the string. Note, however, that the tension makes no con-
tribution to the torque, since its line of action clearly passes through the pivot
point. From simple trigonometry, the line of action of the gravitational force
passes a distance l sin θ from the pivot point. Hence, the magnitude of the grav-
itational torque is m g l sin θ. Moreover, the gravitational torque is a restoring
torque: i.e., if the mass is displaced slightly from its equilibrium state (i.e., θ = 0)
then the gravitational force clearly acts to push the mass back toward that state.

243
11 OSCILLATORY MOTION 11.4 The simple pendulum

Thus, we can write

τ = −m g l sin θ. (11.22)
Combining the previous two equations, we obtain the following angular equation
of motion of the pendulum:
l θ̈ = −g sin θ. (11.23)
Unfortunately, this is not the simple harmonic equation. Indeed, the above equa-
tion possesses no closed solution which can be expressed in terms of simple func-
tions.

Suppose that we restrict our attention to relatively small deviations from the
equilibrium state. In other words, suppose that the angle θ is constrained to take
fairly small values. We know, from trigonometry, that for |θ| less than about 6 ◦ it
is a good approximation to write

sin θ  θ. (11.24)

Hence, in the small angle limit, Eq. (11.23) reduces to

l θ̈ = −g θ, (11.25)

which is in the familiar form of a simple harmonic equation. Comparing with

our original simple harmonic equation, Eq. (11.2), and its solution, we conclude
that the angular frequency of small amplitude oscillations of a simple pendulum
is given by 
g
ω= . (11.26)
l
In this case, the pendulum frequency is dependent only on the length of the
pendulum and the local gravitational acceleration, and is independent of the
mass of the pendulum and the amplitude of the pendulum swings (provided that
sin θ  θ remains a good approximation). Historically, the simple pendulum
was the basis of virtually all accurate time-keeping devices before the advent of
electronic clocks. Simple pendulums can also be used to measure local variations
in g.

244
11 OSCILLATORY MOTION 11.5 The compound pendulum

Pivot point P
d
θ

C Centre of mass

Mg

Figure 98: A compound pendulum.

11.5 The compound pendulum

Consider an extended body of mass M with a hole drilled though it. Suppose that
the body is suspended from a fixed peg, which passes through the hole, such that
it is free to swing from side to side, as shown in Fig. 98. This setup is known as a
compound pendulum.

Let P be the pivot point, and let C be the body’s centre of mass, which is located
a distance d from the pivot. Let θ be the angle subtended between the downward
vertical (which passes through point P) and the line PC. The equilibrium state of
the compound pendulum corresponds to the case in which the centre of mass lies
vertically below the pivot point: i.e., θ = 0. See Sect. 10.3. The angular equation
of motion of the pendulum is simply
I θ̈ = τ, (11.27)
where I is the moment of inertia of the body about the pivot point, and τ is the
torque. Using similar arguments to those employed for the case of the simple
pendulum (recalling that all the weight of the pendulum acts at its centre of
mass), we can write
τ = −M g d sin θ. (11.28)

245
11 OSCILLATORY MOTION 11.6 Uniform circular motion

Note that the reaction, R, at the peg does not contribute to the torque, since
its line of action passes through the pivot point. Combining the previous two
equations, we obtain the following angular equation of motion of the pendulum:

I θ̈ = −M g d sin θ. (11.29)

Finally, adopting the small angle approximation, sin θ  θ, we arrive at the simple
harmonic equation:
I θ̈ = −M g d θ. (11.30)
It is clear, by analogy with our previous solutions of such equations, that the
angular frequency of small amplitude oscillations of a compound pendulum is
given by 

M g d
ω= . (11.31)
I

It is helpful to define the length

I
L= . (11.32)
Md
Equation (11.31) reduces to 
g
ω= , (11.33)
L
which is identical in form to the corresponding expression for a simple pendulum.
We conclude that a compound pendulum behaves like a simple pendulum with
effective length L.

11.6 Uniform circular motion

Consider an object executing uniform circular motion of radius a. Let us set up a

cartesian coordinate system whose origin coincides with the centre of the circle,
and which is such that the motion is confined to the x-y plane. As illustrated in
Fig. 99, the instantaneous position of the object can be conveniently parameter-
ized in terms of an angle θ.

246
11 OSCILLATORY MOTION 11.6 Uniform circular motion

a a sinθ

θ
x
a cos θ

Figure 99: Uniform circular motion.

Since the object is executing uniform circular motion, we expect the angle θ to
increase linearly with time. In other words, we can write
θ = ω t, (11.34)
where ω is the angular rotation frequency (i.e., the number of radians through
which the object rotates per second). Here, it is assumed that θ = 0 at t = 0, for
the sake of convenience.

From simple trigonometry, the x- and y-coordinates of the object can be writ-
ten
x = a cos θ, (11.35)
y = a sin θ, (11.36)
respectively. Hence, combining the previous equations, we obtain
x = a cos(ω t), (11.37)
y = a cos(ω t − π/2). (11.38)
Here, use has been made of the trigonometric identity sin θ = cos(θ − π/2). A
comparison of the above two equations with the standard equation of simple har-
monic motion, Eq. (11.3), reveals that our object is executing simple harmonic

247
11 OSCILLATORY MOTION 11.6 Uniform circular motion

motion simultaneously along both the x- and the y -axes. Note, however, that
these two motions are 90◦ (i.e., π/2 radians) out of phase. Moreover, the am-
plitude of the motion equals the radius of the circle. Clearly, there is a close
relationship between simple harmonic motion and circular motion.

Worked example 11.1: Piston in steam engine

Question: A piston in a stream engine executes simple harmonic motion. Given

that the maximum displacement of the piston from its centre-line is ±7 cm, and
that the mass of the piston is 4 kg, find the maximum velocity of the piston when
the steam engine is running at 4000 rev./min. What is the maximum accelera-
tion?

Answer: We are told that the amplitude of the oscillation is a = 0.07 m. Moreover,
when converted to cycles per second (i.e., hertz), the frequency of the oscillation
becomes
4000
f= = 66.6666 Hz.
60
Hence, the angular frequency is

ω = 2 π f = 418.88 rad./sec.

Consulting Tab. 4, we note that the maximum velocity of an object executing

simple harmonic motion is vmax = a ω. Hence, the maximum velocity is

vmax = a ω = 0.07 × 418.88 = 29.32 m/s.

Likewise, according to Tab. 4, the maximum acceleration is given by

amax = a ω2 = 0.07 × 418.88 × 418.88 = 1.228 × 104 m/s2 .

Worked example 11.2: Block and spring

Question: A block attached to a spring executes simple harmonic motion in a

horizontal plane with an amplitude of 0.25 m. At a point 0.15 m away from the

248
11 OSCILLATORY MOTION 11.6 Uniform circular motion

equilibrium position, the velocity of the block is 0.75 m/s. What is the period of
oscillation of the block?

Answer: The equation of simple harmonic motion is

x = a cos(ω t − φ),
where x is the displacement, and a is the amplitude. We are told that a = 0.25 m.
The velocity of the block is obtained by taking the time derivative of the above
expression:
ẋ = −a ω sin(ω t − φ).
We are told that at t = 0 (say), x = 0.15 m and ẋ = 0.75 m/s. Hence,
0.15 = 0.25 cos(φ),
0.75 = 0.25 ω sin(φ).
The first equation gives φ = cos−1 (0.15/0.25) = 53.13◦ . The second equation
yields
0.75
ω= = 3.75 rad./s.
0.25 × sin(53.13◦ )
Hence, the period of the motion is
2π
T= = 1.676 s.
ω

Worked example 11.3: Block and two springs

Question: A block of mass m = 3 kg is attached to two springs, as shown below,

and slides over a horizontal frictionless surface. Given that the force constants
of the two springs are k1 = 1200 N/m and k2 = 400 N/m, find the period of
oscillation of the system.

Answer: Let x1 and x2 represent the extensions of the first and second springs,
respectively. The net displacement x of the mass from its equilibrium position is
then given by
x = x1 + x2 .

249
11 OSCILLATORY MOTION 11.6 Uniform circular motion

k1 k2
m

Let f1 = k1 x1 and f2 = k2 x2 be the magnitudes of the forces exerted by the

first and second springs, respectively. Since the springs (presumably) possess
negligible inertia, they must exert equal and opposite forces on one another. This
implies that f1 = f2 , or
k1 x1 = k2 x2 .
Finally, if f is the magnitude of the restoring force acting on the mass, then force
balance implies that f = f1 = f2 , or

f = keff x = k1 x1 .

Here, keff is the effective force constant of the two springs. The above equations
can be combined to give
k1 x1 k1 k1 k2
keff = = = .
x1 + x 2 1 + k1 /k2 k1 + k2
Thus, the problem reduces to that of a block of mass m = 3 kg attached to a
spring of effective force constant
k1 k2 1200 × 400
keff = = = 300 N/m.
k1 + k2 1200 + 400
The angular frequency of oscillation is immediately given by the standard formula
 
 
 keff  300
 
ω= = = 10 rad./s.
m 3
Hence, the period of oscillation is
2π
T= = 0.6283 s.
ω

250
11 OSCILLATORY MOTION 11.6 Uniform circular motion

Worked example 11.4: Energy in simple harmonic motion

Question: A block of mass m = 4 kg is attached to a spring, and undergoes simple

harmonic motion with a period of T = 0.35 s. The total energy of the system is
E = 2.5 J. What is the force constant of the spring? What is the amplitude of the
motion?

Answer: The angular frequency of the motion is

2π 2π
ω= = = 17.95 rad./s.
T 0.35

Now, ω = k/m for a mass on a spring. Rearrangement of this formula yields
k = m ω2 = 4 × 17.95 × 17.95 = 1289.1 N/m.
The total energy of a system executing simple harmonic motion is E = a2 k/2.
Rearrangement of this formula gives
 
 
2 E

2
 × 2.5
a= = = 0.06228 m.
k 1289.1
Thus, the force constant is 1289.1 N/m and the amplitude is 0.06228 m.

Worked example 11.5: Gravity on a new planet

Question: Having landed on a newly discovered planet, an astronaut sets up a

simple pendulum of length 0.6 m, and finds that it makes 51 complete oscillations
in 1 minute. The amplitude of the oscillations is small compared to the length of
the pendulum. What is the surface gravitational acceleration on the planet?

Answer: The frequency of the oscillations is

51
f= = 0.85 Hz.
60
Hence, the angular frequency is
ω = 2 π f = 2 × π × 1.833 = 5.341 rad./s.

251
11 OSCILLATORY MOTION 11.6 Uniform circular motion

Now, ω = g/l for small amplitude oscillations of a simple pendulum. Rear-
rangement off this formula gives

g = ω2 l = 5.341 × 5.341 × 0.6 = 17.11 m/s2 .

Hence, the surface gravitational acceleration is 17.11 m/s2 .

Worked example 11.6: Oscillating disk

Question: A uniform disk of radius r = 0.8 m and mass M = 3 kg is freely sus-

pended from a horizontal pivot located a radial distance d = 0.25 m from its
centre. Find the angular frequency of small amplitude oscillations of the disk.

Answer: The moment of inertia of the disk about a perpendicular axis passing
through its centre is I = (1/2) M r2 . From the parallel axis theorem, the moment
of inertia of the disk about the pivot point is
3 × 0.8 × 0.8
I  = I + M d2 = + 3 × 0.25 × 0.25 = 1.1475 kg m2 .
2
The angular frequency of small amplitude oscillations of a compound pendulum
is given by  
 
M g d  3 × 9.81 × 0.25
 
ω= 
= = 2.532 rad./s.
I 1.1475
Hence, the answer is 2.532 rad./s.

252
12 ORBITAL MOTION

12 Orbital motion

12.1 Introduction

We have spent this course exploring the theory of motion first outlined by Sir
Isaac Newton in his Principia (1687). It is, therefore, interesting to discuss
the particular application of this theory which made Newton an international
celebrity, and which profoundly and permanently changed humankind’s outlook
on the Universe. This application is, of course, the motion of the Solar System.

12.2 Historical background

Humankind has always been fascinated by the night sky, and, in particular, by
the movements of the Sun, the Moon, and the objects which the ancient Greeks
called plantai (“wanderers”), and which we call planets. In ancient times, much
of this interest was of a practical nature. The Sun and the Moon were impor-
tant for determining the calendar, and also for navigation. Moreover, the planets
were vital to astrology: i.e., the belief—almost universally prevalent in the an-
cient world—that the positions of the planets in the sky could be used to foretell
important events.

Actually, there were only seven “wandering” heavenly bodies visible to ancient
peoples: the Sun, the Moon, and the five planets—Mercury, Venus, Mars, Jupiter,
and Saturn. The ancients believed that the stars were fixed to a “celestial sphere”
which formed the outer boundary of the Universe. However, it was recognized
that the wandering bodies were located within this sphere: e.g., because the
Moon clearly passes in front of, and blocks the light from, stars in its path. It
was also recognized that some bodies were closer to the Earth than others. For
instance, ancient astronomers noted that the Moon occasionally passes in front
of the Sun and each of the planets. Moreover, Mercury and Venus can sometimes
be seen to transit in front of the Sun.

The first scientific model of the Solar System was outlined by the Greek philoso-

253
12 ORBITAL MOTION 12.2 Historical background

pher Eudoxas of Cnidus (409–356 BC). According to this model, the Sun, the
Moon, and the planets all execute uniform circular orbits around the Earth—
which is fixed, and non-rotating. The order of the orbits is as follows: Moon,
Mercury, Venus, Sun, Mars, Jupiter, Saturn—with the Moon closest to the Earth.
For obvious reasons, Eudoxas’ model became known as the geocentric model of
the Solar System. Note that orbits are circular in this model for philosophical
reasons. The ancients believed the heavens to be the realm of perfection. Since
a circle is the most “perfect” imaginable shape, it follows that heavenly objects
must execute circular orbits.

A second Greek philosopher, Aristarchus of Samos (310–230 BC), proposed an

alternative model in which the Earth and the planets execute uniform circular
orbits around the Sun—which is fixed. Moreover, the Moon orbits around the
Earth, and the Earth rotates daily about a North-South axis. The order of the
planetary orbits is as follows: Mercury, Venus, Earth, Mars, Jupiter, Saturn—with
Mercury closest to the Sun. This model became known as the heliocentric model
of the Solar System.

The heliocentric model was generally rejected by the ancient philosophers for
three main reasons:

1. If the Earth is rotating about its axis, and orbiting around the Sun, then the
Earth must be in motion. However, we cannot “feel” this motion. Nor does
this motion give rise to any obvious observational consequences. Hence, the
Earth must be stationary.
2. If the Earth is executing a circular orbit around the Sun then the positions of
the stars should be slightly different when the Earth is on opposite sides of
the Sun. This effect is known as parallax. Since no stellar parallax is observ-
able (at least, with the naked eye), the Earth must be stationary. In order
to appreciate the force of this argument, it is important to realize that an-
cient astronomers did not suppose the stars to be significantly further away
from the Earth than the planets. The celestial sphere was assumed to lie just
beyond the orbit of Saturn.
3. The geocentric model is far more philosophically attractive than the helio-

254
12 ORBITAL MOTION 12.2 Historical background

epicycle

deferant
planet
P
centre of deferant

Earth
+
C E

equant

Figure 100: The Ptolemaic system.

centric model, since in the former model the Earth occupies a privileged
position in the Universe.

The geocentric model was first converted into a proper scientific theory, ca-
pable of accurate predictions, by the Alexandrian philosopher Claudius Ptolemy
(85–165 AD). The theory that Ptolemy proposed in his famous book, now known
as the Almagest, remained the dominant scientific picture of the Solar System for
over a millennium. Basically, Ptolemy acquired and extended the extensive set
of planetary observations of his predecessor Hipparchus, and then constructed a
geocentric model capable of accounting for them. However, in order to fit the
observations, Ptolemy was forced to make some significant modifications to the
original model of Eudoxas. Let us discuss these modifications.

First, we need to introduce some terminology. As shown in Fig. 100, deferants

are large circles centred on the Earth, and epicyles are small circles whose cen-
tres move around the circumferences of the deferants. In the Ptolemaic system,
instead of traveling around deferants, the planets move around the circumfer-
ence of epicycles, which, in turn, move around the circumference of deferants.
Ptolemy found, however, that this modification was insufficient to completely ac-
count for all of his data. Ptolemy’s second modification to Eudoxas’ model was

255
12 ORBITAL MOTION 12.2 Historical background

to displace the Earth slightly from the common centre of the deferants. More-
over, Ptolemy assumed that the Sun, Moon, and planets rotate uniformly about
an imaginary point, called the equant, which is displaced an equal distance in the
opposite direction to the Earth from the centre of the deferants. In other words,
Ptolemy assumed that the line EP, in Fig. 100, rotates uniformly, rather than the
line CP.

Figure 101 shows more details of the Ptolemaic model.2 Note that this dia-
gram is not drawn to scale, and the displacement of the Earth from the centre
of the deferants has been omitted for the sake of clarity. It can be seen that the
Moon and the Sun do not possess epicyles. Moreover, the motions of the inferior
planets (i.e., Mercury and Venus) are closely linked to the motion of the Sun. In
fact, the centres of the inferior planet epicycles move on an imaginary line con-
necting the Earth and the Sun. Furthermore, the radius vectors connecting the
superior planets (i.e., Mars, Jupiter, and Saturn) to the centres of their epicycles
are always parallel to the geometric line connecting the Earth and the Sun. Note
that, in addition to the motion indicated in the diagram, all of the heavenly bodies
(including the stars) rotate clockwise (assuming that we are looking down on the
Earth’s North pole in Fig. 101) with a period of 1 day. Finally, there are epicycles
within the epicycles shown in the diagram. In fact, some planets need as many
as 28 epicycles to account for all the details of their motion. These subsidiary
epicycles are not shown in the diagram, for the sake of clarity.

As is quite apparent, the Ptolemaic model of the Solar System is extremely

complicated. However, it successfully accounted for the relatively crude naked
eye observations made by the ancient Greeks. The Sun-linked epicyles of the
inferior planets are needed to explain why these objects always remain close to
the Sun in the sky. The epicycles of the superior planets are needed to account for
their occasional bouts of retrograde motion: i.e., motion in the opposite direction
to their apparent direction of rotation around the Earth. Finally, the displacement
of the Earth from the centre of the deferants, as well as the introduction of the
equant as the centre of uniform rotation, is needed to explain why the planets
speed up slightly when they are close to the Earth (and, hence, appear brighter
in the night sky), and slow down when they are further away.
2
R.A. Hatch, University of Florida, http://web.clas.ufl.edu/users/rhatch/

256
12 ORBITAL MOTION 12.2 Historical background

Stars
29.46 y

11.86 y

1.88 y

Mars
1y

1y
1y

1y

27 1/3 d
Venus
Earth Mercury Sun
88 d
225 d
Moon

Jupiter

1y

Saturn

1y

Figure 101: The Ptolemaic model of the Solar System.

257
12 ORBITAL MOTION 12.2 Historical background

Ptolemy’s model of the Solar System was rescued from the wreck of ancient
European civilization by the Roman Catholic Church, which, unfortunately, con-
verted it into a minor article of faith, on the basis of a few references in the Bible
which seemed to imply that the Earth is stationary and the Sun is moving (e.g.,
Joshua 10:12-13, Habakkuk 3:11). Consequently, this model was not subject to
proper scientific criticism for over a millennium. Having said this, few medieval
or renaissance philosophers were entirely satisfied with Ptolemy’s model. Their
dissatisfaction focused, not on the many epicycles (which to the modern eye seem
rather absurd), but on the displacement of the Earth from the centre of the defer-
ants, and the introduction of the equant as the centre of uniform rotation. Recall,
that the only reason planetary orbits are constructed from circles in Ptolemy’s
model is to preserve the assumed ideal symmetry of the heavens. Unfortunately,
this symmetry is severely compromised when the Earth is displaced from the
apparent centre of the Universe. This problem so perplexed the Polish priest-
astronomer Nicolaus Copernicus (1473–1543) that he eventually decided to re-
ject the geocentric model, and revive the heliocentric model of Aristarchus. After
many years of mathematical calculations, Copernicus published a book entitled
De revolutionibus orbium coelestium (On the revolutions of the celestial spheres)
in 1543 which outlined his new heliocentric theory.

Copernicus’ model is illustrated in Fig. 102. Again, this diagram is not to scale.
The planets execute uniform circular orbits about the Sun, and the Moon orbits
about the Earth. Finally, the Earth revolves about its axis daily. Note that there is
no displacement of the Sun from the centres of the planetary orbits, and there is
no equant. Moreover, in this model, the inferior planets remain close to the Sun
in the sky without any special synchronization of their orbits. Furthermore, the
occasional retrograde motion of the superior planets has a more natural explana-
tion than in Ptolemy’s model. Since the Earth orbits more rapidly than the supe-
rior planets, it occasionally “overtakes” them, and they appear to move backward
in the night sky, in much the same manner that slow moving cars on a freeway
appears to move backward to a driver overtaking them. Copernicus accounted
for the lack of stellar parallax, due to the Earth’s motion, by postulating that the
stars were a lot further away than had previously been supposed, rendering any
parallax undetectably small. Unfortunately, Copernicus insisted on retaining uni-

258
12 ORBITAL MOTION 12.2 Historical background

29.46 y

11.86 y
Stars

1.88 y

1y
Mars

225 d

Venus
88 d

Mercury Moon
Sun
Earth
27 1/3 d

Jupiter

Saturn

Figure 102: The Copernican model of the Solar System.

259
12 ORBITAL MOTION 12.2 Historical background

form circular motion in his model (after all, he was trying to construct a more
symmetric model than that of Ptolemy). Consequently, Copernicus also had to
resort to epicycles to fit the data. In fact, Copernicus’ model ended up with more
epicycles than Ptolemy’s!

The real breakthrough in the understanding of planetary motion occurred—as

most breakthroughs in physics occur—when better data became available. The
data in question was produced by the Dane Tycho Brahe (1546–1601), who de-
voted his life to making naked eye astronomical observations of unprecedented
accuracy and detail. This data was eventually inherited by Brahe’s pupil and assis-
tant, the German scientist Johannes Kepler (1571–1630). Kepler fully accepted
Copernicus’ heliocentric theory of the Solar System. Moreover, he was just as
firm a believer as Copernicus in the perfection of the heavens, and the conse-
quent need for circular motion of planetary bodies. The main difference was that
Kepler’s observational data was considerably better than Copernicus’. After years
of fruitless effort, Kepler eventually concluded that no combination of circular
deferants and epicycles could completely account for his data. At this stage, he
started to think the unthinkable. Maybe, planetary motion was not circular after
all? After more calculations, Kepler was eventually able to formulate three ex-
traordinarily simple laws which completely accounted for Brahe’s observations.
These laws are as follows:

1. The planets move in elliptical orbits with the Sun at one focus.
2. A line from the Sun to any given planet sweeps out equal areas in equal time
intervals.
3. The square of a planet’s period is proportional to the cube of the planet’s
mean distance from the Sun.

Note that there are no epicyles or equants in Kepler’s model of the Solar System.

Figure 103 illustrates Kepler’s second law. Here, the ellipse represents a plan-
etary orbit, and S represents the Sun, which is located at one of the focii of the
ellipse. Suppose that the planet moves from point A to point B in the same time
it takes to move from point C to point D. According the Kepler’s second law,

260
12 ORBITAL MOTION 12.2 Historical background

C
B
S

A
D

Figure 103: Kepler’s second law.

Planet a(AU) T (yr) a3/T 2

Mercury 0.387 0.241 0.998
Venus 0.723 0.615 0.999
Earth 1.000 1.000 1.000
Mars 1.524 1.881 1.000
Jupiter 5.203 11.862 1.001
Saturn 9.516 29.458 0.993

Table 5: Kepler’s third law. Here, a is the mean distance from the Sun, measured in Astronomical
Units (1 AU is the mean Earth-Sun distance), and T is the orbital period, measured in years.

the areas of the elliptic segments ASB and CSD are equal. Note that this law
basically mandates that planets speed up when they move closer to the Sun.

Table 5 illustrates Kepler’s third law. The mean distance, a, and orbital period,
T , as well as the ratio a3 /T 2 , are listed for each of the first six planets in the Solar
System. It can be seen that the ratio a3 /T 2 is indeed constant from planet to
planet.

Since we have now definitely adopted a heliocentric model of the Solar Sys-
tem, let us discuss the ancient Greek objections to such a model, listed earlier.
We have already dealt with the second objection (the absence of stellar parallax)
by stating that the stars are a lot further away from the Earth than the ancient
Greeks supposed. The third objection (that it is philosophically more attractive
to have the Earth at the centre of the Universe) is not a valid scientific criticism.
What about the first objection? If the Earth is rotating about its axis, and also

261
12 ORBITAL MOTION 12.3 Gravity

orbiting the Sun, why do we not “feel” this motion? At first sight, this objec-
tion appears to have some force. After all, the rotation velocity of the Earth’s
surface is about 460 m/s. Moreover, the Earth’s orbital velocity is approximately
30 km/s. Surely, we would notice if we were moving this rapidly? Of course, this
reasoning is faulty because we know, from Newton’s laws of motion, that we only
“feel” the acceleration associated with motion, not the motion itself. It turns out
that the acceleration at the Earth’s surface due to its axial rotation is only about
0.034 m/s2 . Moreover, the Earth’s acceleration due to its orbital motion is only
0.0059 m/s2 . Nowadays, we can detect such small accelerations, but the ancient
Greeks certainly could not.

Kepler correctly formulated the three laws of planetary motion in 1619. Al-
most seventy years later, in 1687, Isaac Newton published his Principia, in which
he presented, for the first time, a universal theory of motion. Newton then went
on to illustrate his theory by using it to deriving Kepler’s laws from first principles.
Let us now discuss Newton’s monumental achievement in more detail.

12.3 Gravity

There is one important question which we have avoided discussing until now.
Why do objects fall towards the surface of the Earth? The ancient Greeks had
a very simple answer to this question. According to Aristotle, all objects have
a natural tendency to fall towards the centre of the Universe. Since the centre
of the Earth coincides with the centre of the Universe, all objects also tend to
fall towards the Earth’s surface. So, an ancient Greek might ask, why do the
planets not fall towards the Earth? Well, according to Aristotle, the planets are
embedded in crystal spheres which rotate with them whilst holding them in place
in the firmament. Unfortunately, Ptolemy seriously undermined this explanation
by shifting the Earth slightly from the centre of the Universe. However, the coup
de grace was delivered by Copernicus, who converted the Earth into just another
planet orbiting the Sun.

So, why do objects fall towards the surface of the Earth? The first person,
after Aristotle, to seriously consider this question was Sir Isaac Newton. Since

262
12 ORBITAL MOTION 12.3 Gravity

2
f = G m 1 m2 / r

m2
−f
r

f
m1

Figure 104: Newton’s law of gravity.

the Earth is not located in a special place in the Universe, Newton reasoned,
objects must be attracted toward the Earth itself. Moreover, since the Earth is just
another planet, objects must be attracted towards other planets as well. In fact,
all objects must exert a force of attraction on all other objects in the Universe.
What intrinsic property of objects causes them to exert this attractive force—
which Newton termed gravity—on other objects? Newton decided that the crucial
property was mass. After much thought, he was eventually able to formulate his
famous law of universal gravitation:

Every particle in the Universe attracts every other particle with a force directly
proportional to the product of their masses and inversely proportional to the
square of the distance between them. The direction of the force is along the
line joining the particles.

Incidentally, Newton adopted an inverse square law because he knew that this
was the only type of force law which was consistent with Kepler’s third law of
planetary motion.

Consider two point objects of masses m1 and m2 , separated by a distance r.

As illustrated in Fig. 104, the magnitude of the force of attraction between these
objects is
m1 m2
f=G . (12.1)
r2
The direction of the force is along the line joining the two objects.

263
12 ORBITAL MOTION 12.3 Gravity

Let r1 and r2 be the vector positions of the two objects, respectively. The vector
gravitational force exerted by object 2 on object 1 can be written
r2 − r1
f12 = G . (12.2)
|r2 − r1 |3
Likewise, the vector gravitational force exerted by object 1 on object 2 takes the
form
r1 − r2
f21 = G = −f21 . (12.3)
|r1 − r2 |3

The constant of proportionality, G, appearing in the above formulae is called

the gravitational constant. Newton could only estimate the value of this quantity,
which was first directly measured by Henry Cavendish in 1798. The modern
value of G is
G = 6.6726 × 10−11 N m2 /kg2 . (12.4)
Note that the gravitational constant is numerically extremely small. This implies
that gravity is an intrinsically weak force. In fact, gravity usually only becomes
significant if at least one of the masses involved is of astronomical dimensions
(e.g., it is a planet, or a star).

Let us use Newton’s law of gravity to account for the Earth’s surface gravity.
Consider an object of mass m close to the surface of the Earth, whose mass and
radius are M⊕ = 5.97 × 1024 kg and R⊕ = 6.378 × 106 m, respectively. Newton
proved, after considerable effort, that the gravitational force exerted by a spher-
ical body (outside that body) is the same as that exerted by an equivalent point
mass located at the body’s centre. Hence, the gravitational force exerted by the
Earth on the object in question is of magnitude
m M⊕
f=G , (12.5)
R⊕2
and is directed towards the centre of the Earth. It follows that the equation of
motion of the object can be written
m M⊕
m r̈ = −G ^z, (12.6)
R⊕2

264
12 ORBITAL MOTION 12.4 Gravitational potential energy

where ^z is a unit vector pointing straight upwards (i.e., away from the Earth’s
centre). Canceling the factor m on either side of the above equation, we obtain

r̈ = −g⊕ ^z, (12.7)

where
G M⊕ (6.673 × 10−11 ) × (5.97 × 1024 )
g⊕ = = = 9.79 m/s2 . (12.8)
2
R⊕ (6.378 × 10 )6 2

Thus, we conclude that all objects on the Earth’s surface, irrespective of their
mass, accelerate straight down (i.e., towards the Earth’s centre) with a constant
acceleration of 9.79 m/s2 . This estimate for the acceleration due to gravity is
slightly off the conventional value of 9.81 m/s2 because the Earth is actually not
quite spherical.

Since Newton’s law of gravitation is universal, we immediately conclude that

any spherical body of mass M and radius R possesses a surface gravity g given by
the following formula:
g M/M⊕
= . (12.9)
g⊕ (R/R⊕ )2
Table 6 shows the surface gravity of various bodies in the Solar System, estimated
using the above expression. It can be seen that the surface gravity of the Moon is
only about one fifth of that of the Earth. No wonder Apollo astronauts were able
to jump so far on the Moon’s surface! Prospective Mars colonists should note that
they will only weigh about a third of their terrestrial weight on Mars.

12.4 Gravitational potential energy

We saw earlier, in Sect. 5.5, that gravity is a conservative force, and, therefore,
has an associated potential energy. Let us obtain a general formula for this energy.
Consider a point object of mass m, which is a radial distance r from another point
object of mass M. The gravitational force acting on the first mass is of magnitude
f = G M/r2 , and is directed towards the second mass. Imagine that the first
mass moves radially away from the second mass, until it reaches infinity. What

265
12 ORBITAL MOTION 12.4 Gravitational potential energy

Body M/M⊕ R/R⊕ g/g⊕

Sun 3.33 × 105 109.0 28.1
Moon 0.0123 0.273 0.17
Mercury 0.0553 0.383 0.38
Venus 0.816 0.949 0.91
Earth 1.000 1.000 1.000
Mars 0.108 0.533 0.38
Jupiter 318.3 11.21 2.5
Saturn 95.14 9.45 1.07

Table 6: The mass, M, radius, R, and surface gravity, g, of various bodies in the Solar System. All
quantities are expressed as fractions of the corresponding terrestrial quantity.

is the change in the potential energy of the first mass associated with this shift?
According to Eq. (5.33),
∞
U(∞) − U(r) = − [−f(r)] dr. (12.10)
r

There is a minus sign in front of f because this force is oppositely directed to the
motion. The above expression can be integrated to give
GMm
U(r) = − . (12.11)
r
Here, we have adopted the convenient normalization that the potential energy
at infinity is zero. According to the above formula, the gravitational potential
energy of a mass m located a distance r from a mass M is simply −G M m/r.

Consider an object of mass m moving close to the Earth’s surface. The potential
energy of such an object can be written
G M⊕ m
U=− , (12.12)
R⊕ + z
where M⊕ and R⊕ are the mass and radius of the Earth, respectively, and z is the
vertical height of the object above the Earth’s surface. In the limit that z  R ⊕ ,
the above expression can be expanded using the binomial theorem to give
G M⊕ m G M⊕ m
U− + z, (12.13)
R⊕ R⊕2
266
12 ORBITAL MOTION 12.4 Gravitational potential energy

Since potential energy is undetermined to an arbitrary additive constant, we

could just as well write
U  m g z, (12.14)
where g = G M⊕ /R⊕2 is the acceleration due to gravity at the Earth’s surface
[see Eq. (12.8)]. Of course, the above formula is equivalent to the formula (5.3)
derived earlier on in this course.

For an object of mass m and speed v, moving in the gravitational field of a

fixed object of mass M, we expect the total energy,
E = K + U, (12.15)
to be a constant of the motion. Here, the kinetic energy is written K = (1/2) m v 2 ,
whereas the potential energy takes the form U = −G M m/r. Of course, r is the
distance between the two objects. Suppose that the fixed object is a sphere of
radius R. Suppose, further, that the second object is launched from the surface
of this sphere with some velocity vesc which is such that it only just escapes the
sphere’s gravitational influence. After the object has escaped, it is a long way
away from the sphere, and hence U = 0. Moreover, if the object only just escaped,
then we also expect K = 0, since the object will have expended all of its initial
kinetic energy escaping from the sphere’s gravitational well. We conclude that
our object possesses zero net energy: i.e., E = K + U = 0. Since E is a constant of
the motion, it follows that at the launch point
1 2 GMm
E= m vesc − = 0. (12.16)
2 R
This expression can be rearranged to give


2 G M

vesc = . (12.17)
R
The quantity vesc is known as the escape velocity. Objects launched from the sur-
face of the sphere with velocities exceeding this value will eventually escape from
the sphere’s gravitational influence. Otherwise, the objects will remain in orbit
around the sphere, and may eventually strike its surface. Note that the escape
velocity is independent of the object’s mass and launch direction (assuming that
it is not straight into the sphere).

267
12 ORBITAL MOTION 12.5 Satellite orbits

The escape velocity for the Earth is

 
 
 2 G M⊕

2
 × (6.673 × 10−11 ) × (5.97 × 1024 )
vesc = = = 11.2 km/s. (12.18)
R⊕ 6.378 × 106
Clearly, NASA must launch deep space probes from the surface of the Earth with
velocities which exceed this value if they are to have any hope of eventually
reaching their targets.

12.5 Satellite orbits

Consider an artificial satellite executing a circular orbit of radius r around the

Earth. Let ω be the satellite’s orbital angular velocity. The satellite experiences
an acceleration towards the Earth’s centre of magnitude ω2 r. Of course, this
acceleration is provided by the gravitational attraction between the satellite and
the Earth, which yields an acceleration of magnitude G M⊕ /r2 . It follows that
G M⊕
ω2 r = . (12.19)
r2

Suppose that the satellite’s orbit lies in the Earth’s equatorial plane. Moreover,
suppose that the satellite’s orbital angular velocity just matches the Earth’s angu-
lar velocity of rotation. In this case, the satellite will appear to hover in the same
place in the sky to a stationary observer on the Earth’s surface. A satellite with
this singular property is known as a geostationary satellite.

Virtually all of the satellites used to monitor the Earth’s weather patterns are
geostationary in nature. Communications satellites also tend to be geostationary.
Of course, the satellites which beam satellite-TV to homes across the world must
be geostationary—otherwise, you would need to install an expensive tracking
antenna on top of your house in order to pick up the transmissions. Incidentally,
the person who first envisaged rapid global telecommunication via a network of
geostationary satellites was the science fiction writer Arthur C. Clarke in 1945.

Let us calculate the orbital radius of a geostationary satellite. The angular

268
12 ORBITAL MOTION 12.6 Planetary orbits

velocity of the Earth’s rotation is

2π
ω= = 7.27 × 10−5 rad./s. (12.20)
24 × 60 × 60
It follows from Eq. (12.19) that
 1/3 ⎛ ⎞1/3
G M⊕ (6.673 × 10−11 ) × (5.97 × 1024 ) ⎠
rgeo = = ⎝
ω2 (7.27 × 10−5 )2
= 4.22 × 107 m = 6.62 R⊕ . (12.21)
Thus, a geostationary satellite must be placed in a circular orbit whose radius is
exactly 6.62 times the Earth’s radius.

12.6 Planetary orbits

Let us now see whether we can use Newton’s universal laws of motion to derive
Kepler’s laws of planetary motion. Consider a planet orbiting around the Sun. It
is convenient to specify the planet’s instantaneous position, with respect to the
Sun, in terms of the polar coordinates r and θ. As illustrated in Fig. 105, r is the
radial distance between the planet and the Sun, whereas θ is the angular bearing
of the planet, from the Sun, measured with respect to some arbitrarily chosen
direction.

Let us define two unit vectors, er and eθ . (A unit vector is simply a vector
whose length is unity.) As shown in Fig. 105, the radial unit vector er always
points from the Sun towards the instantaneous position of the planet. Moreover,
the tangential unit vector eθ is always normal to er , in the direction of increasing
θ. In Sect. 7.5, we demonstrated that when acceleration is written in terms of
polar coordinates, it takes the form
a = ar er + aθ eθ , (12.22)
where
ar = r̈ − r θ̇2 , (12.23)
aθ = r θ̈ + 2 ṙ θ̇. (12.24)

269
12 ORBITAL MOTION 12.6 Planetary orbits

eθ
er

Planet

θ
Sun

Figure 105: A planetary orbit.

These expressions are more complicated that the corresponding cartesian expres-
sions because the unit vectors er and eθ change direction as the planet changes
position.

Now, the planet is subject to a single force: i.e., the force of gravitational
attraction exerted by the Sun. In polar coordinates, this force takes a particularly
simple form (which is why we are using polar coordinates):
G M m
f=− er . (12.25)
r2
The minus sign indicates that the force is directed towards, rather than away
from, the Sun.

According to Newton’s second law, the planet’s equation of motion is written

m a = f. (12.26)

The above four equations yield

G M
r̈ − r θ̇2 = − , (12.27)
r2
r θ̈ + 2 ṙ θ̇ = 0. (12.28)

270
12 ORBITAL MOTION 12.6 Planetary orbits

δθ
P’

P
S
r

Figure 106: The origin of Kepler’s second law.

Equation (12.28) reduces to

d 2
(r θ̇) = 0, (12.29)
dt
or
r2 θ̇ = h, (12.30)
where h is a constant of the motion. What is the physical interpretation of h?
Recall, from Sect. 9.2, that the angular momentum vector of a point particle can
be written
l = m r × v. (12.31)
For the case in hand, r = r er and v = ṙ er + r θ̇ eθ [see Sect. 7.5]. Hence,

l = m r vθ = m r2 θ̇, (12.32)

yielding
l
h= . (12.33)
m
Clearly, h represents the angular momentum (per unit mass) of our planet around
the Sun. Angular momentum is conserved (i.e., h is constant) because the force
of gravitational attraction between the planet and the Sun exerts zero torque on
the planet. (Recall, from Sect. 9, that torque is the rate of change of angular mo-
mentum.) The torque is zero because the gravitational force is radial in nature:
i.e., its line of action passes through the Sun, and so its associated lever arm is of
length zero.

The quantity h has another physical interpretation. Consider Fig. 106. Sup-
pose that our planet moves from P to P  in the short time interval δt. Here, S

271
12 ORBITAL MOTION 12.6 Planetary orbits

represents the position of the Sun. The lines SP and SP  are both approximately
of length r. Moreover, using simple trigonometry, the line PP  is of length r δθ,
where δθ is the small angle through which the line joining the Sun and the planet
rotates in the time interval δt. The area of the triangle PSP  is approximately
1
δA = × r δθ × r : (12.34)
2
i.e., half its base times its height. Of course, this area represents the area swept
out by the line joining the Sun and the planet in the time interval δt. Hence, the
rate at which this area is swept is given by

δA 1 2 δθ r2 θ̇ h
lim = r lim = = . (12.35)
δt→0 δt 2 δt→0 δt 2 2
Clearly, the fact that h is a constant of the motion implies that the line joining the
planet and the Sun sweeps out area at a constant rate: i.e., the line sweeps equal
areas in equal time intervals. But, this is just Kepler’s second law. We conclude
that Kepler’s second law of planetary motion is a direct manifestation of angular
momentum conservation.

Let
1
r= , (12.36)
u
where u(t) ≡ u(θ) is a new radial variable. Differentiating with respect to t, we
obtain
u̇ θ̇ du du
ṙ = − 2 = − 2 = −h . (12.37)
u u dθ dθ
The last step follows from the fact that θ̇ = h u2 . Differentiating a second time
with respect to t, we obtain
 
d du d2 u d2 u
r̈ = −h = −h θ̇ 2 = −h2 u2 2 . (12.38)
dt dθ dθ dθ
Equations (12.27) and (12.38) can be combined to give
d2 u G M
+ u = . (12.39)
dθ2 h2

272
12 ORBITAL MOTION 12.6 Planetary orbits

This equation possesses the fairly obvious general solution

G M
u = A cos(θ − θ0 ) + , (12.40)
h2
where A and θ0 are arbitrary constants.

The above formula can be inverted to give the following simple orbit equation
for our planet:
1
r= . (12.41)
A cos(θ − θ0 ) + G M /h2
The constant θ0 merely determines the orientation of the orbit. Since we are only
interested in the orbit’s shape, we can set this quantity to zero without loss of
generality. Hence, our orbit equation reduces to
1+e
r = r0 , (12.42)
1 + e cos θ
where
A h2
e= , (12.43)
G M
and
h2
r0 = . (12.44)
G M (1 + e)

Formula (12.42) is the standard equation of an ellipse (assuming e < 1), with
the origin at a focus. Hence, we have now proved Kepler’s first law of planetary
motion. It is clear that r0 is the radial distance at θ = 0. The radial distance at
θ = π is written
1+e
r1 = r 0 . (12.45)
1−e
Here, r0 is termed the perihelion distance (i.e., the closest distance to the Sun)
and r1 is termed the aphelion distance (i.e., the furthest distance from the Sun).
The quantity
r1 − r 0
e= (12.46)
r1 + r 0
is termed the eccentricity of the orbit, and is a measure of its departure from
circularity. Thus, e = 0 corresponds to a purely circular orbit, whereas e →

273
12 ORBITAL MOTION 12.6 Planetary orbits

Planet e
Mercury 0.206
Venus 0.007
Earth 0.017
Mars 0.093
Jupiter 0.048
Saturn 0.056

Table 7: The orbital eccentricities of various planets in the Solar System.

1 corresponds to a highly elongated orbit. As specified in Tab. 7, the orbital

eccentricities of all of the planets (except Mercury) are fairly small.

According to Eq. (12.35), a line joining the Sun and an orbiting planet sweeps
area at the constant rate h/2. Let T be the planet’s orbital period. We expect the
line to sweep out the whole area of the ellipse enclosed by the planet’s orbit in
the time interval T . Since the area of an ellipse is π a b, where a and b are the
semi-major and semi-minor axes, we can write
πab
T= . (12.47)
h/2
Incidentally, Fig. 107 illustrates the relationship between the aphelion distance,
the perihelion distance, and the semi-major and semi-minor axes of a planetary
orbit. It is clear, from the figure, that the semi-major axis is just the mean of the
aphelion and perihelion distances: i.e.,
r0 + r 1
a= . (12.48)
2
Thus, a is essentially the planet’s mean distance from the Sun. Finally, the rela-
tionship between a, b, and the eccentricity, e, is given by the well-known formula
b 
= 1 − e2 . (12.49)
a
This formula can easily be obtained from Eq. (12.42).

Equations (12.44), (12.45), and (12.48) can be combined to give

 
h2 1 1 h2
a= + = . (12.50)
2 G M 1 + e 1 − e G M (1 − e2 )
274
12 ORBITAL MOTION 12.6 Planetary orbits

focus
b

r a
0
r1

Figure 107: Anatomy of a planetary orbit.

It follows, from Eqs. (12.47), (12.49), and (12.50), that the orbital period can be
written
2π
T=√ a3/2 . (12.51)
G M
Thus, the orbital period of a planet is proportional to its mean distance from
the Sun to the power 3/2—the constant of proportionality being the same for all
planets. Of course, this is just Kepler’s third law of planetary motion.

Worked example 12.1: Gravity on Callisto

Question: Callisto is the eighth of Jupiter’s moons: its mass and radius are
M = 1.08 × 1023 kg and R = 2403 km, respectively. What is the gravitational
acceleration on the surface of this moon?

Answer: The surface gravitational acceleration on a spherical body of mass M

and radius R is simply
GM
g= 2 .
R
Hence,
(6.673 × 10−11 ) × (1.08 × 1023 )
g= = 1.25 m/s2 .
(2.403 × 10 )6 2

275
12 ORBITAL MOTION 12.6 Planetary orbits

Worked example 12.2: Acceleration of a rocket

Question: A rocket is located a distance 3.5 times the radius of the Earth above
the Earth’s surface. What is the rocket’s free-fall acceleration?

Answer: Let R⊕ be the Earth’s radius. The distance of the rocket from the centre
of the Earth is r1 = (3.5 + 1) R⊕ = 4.5 R⊕ . We know that the free-fall acceleration
of the rocket when its distance from the Earth’s centre is r0 = R⊕ (i.e., when it is
at the Earth’s surface) is g0 = 9.81 m/s2 . Moreover, we know that gravity is an
inverse-square law (i.e., g ∝ 1/r2 ). Hence, the rocket’s acceleration is
 2
r0 9.81 × 1
g1 = g0 = = 0.484 m/s2 .
r1 (4.5) 2

Worked example 12.3: Circular Earth orbit

Question: A satellite moves in a circular orbit around the Earth with speed v =
6000 m/s. Determine the satellite’s altitude above the Earth’s surface. Determine
the period of the satellite’s orbit. The Earth’s mass and radius are M ⊕ = 5.97 ×
1024 kg and R⊕ = 6.378 × 106 m, respectively.

Answer: The acceleration of the satellite towards the centre of the Earth is v 2 /r,
where r is its orbital radius. This acceleration must be provided by the accelera-
tion G M⊕ /r2 due to the Earth’s gravitational attraction. Hence,
v2 G M⊕
= .
r r2
The above expression can be rearranged to give
G M⊕ (6.673 × 10−11 ) × (5.97 × 1024 )
r= = = 1.107 × 107 m.
v2 (6000) 2

Thus, the satellite’s altitude above the Earth’s surface is

h = r − R⊕ = 1.107 × 107 − 6.378 × 106 = 4.69 × 106 m.

276
12 ORBITAL MOTION 12.6 Planetary orbits

The satellite’s orbital period is simply

2 π r 2 × π × (1.107 × 107 )
T= = = 3.22 hours.
v 6000

Worked example 12.4: Halley’s comet

Question: The distance of closest approach of Halley’s comet to the Sun is 0.57 AU.
(1 AU is the mean Earth-Sun distance.) The greatest distance of the comet from
the Sun is 35 AU. The comet’s speed at closest approach is 54 km/s. What is its
speed when it is furthest from the Sun?

Answer: At perihelion and aphelion, the comet’s velocity is perpendicular to its

position vector from the Sun. Hence, at these two special points, the comet’s
angular momentum (around the Sun) takes the particularly simple form
l = m r u.
Here, m is the comet’s mass, r is its distance from the Sun, and u is its speed.
According to Kepler’s second law, the comet orbits the Sun with constant angular
momentum. Hence, we can write
r0 u0 = r1 u1 ,
where r0 and u0 are the perihelion distance and speed, respectively, and r1 and
u1 are the corresponding quantities at aphelion. We are told that r 0 = 0.57 AU,
r1 = 35 AU, and u0 = 54 km/s. It follows that
u0 r0 54 × 0.57
u1 = = = 0.879 km/s.
r1 35

Worked example 12.5: Mass of star

Question: A planet is in circular orbit around a star. The period and radius of the
orbit are T = 4.3 × 107 s and r = 2.34 × 1011 m, respectively. Calculate the mass
of the star.

277
12 ORBITAL MOTION 12.6 Planetary orbits

Answer: Let ω be the planet’s orbital angular velocity. The planet accelerates
towards the star with acceleration ω2 r. The acceleration due to the star’s gravi-
tational attraction is G M∗ /r2 , where M∗ is the mass of the star. Equating these
accelerations, we obtain
G M∗
ω2 r = .
r2
Now,
2π
T= .
ω
Hence, combining the previous two expressions, we get
4 π2 r3
M∗ = .
G T2
Thus, the mass of the star is
4 × π2 × (2.34 × 1011 )3
M∗ = = 4.01 × 1030 kg.
(6.673 × 10 ) × (4.3 × 10 )
−11 7 2

Worked example 12.6: Launch energy

Question: What is the minimum energy required to launch a probe of mass m =

120 kg into outer space? The Earth’s mass and radius are M⊕ = 5.97 × 1024 kg
and R⊕ = 6.378 × 106 m, respectively.

Answer: The energy which must be given to the probe should just match the
probe’s gain in potential energy as it travels from the Earth’s surface to outer
space. By definition, the probe’s potential energy in outer space is zero. The
potential energy of the probe at the Earth’s surface is
G M⊕ m (6.673 × 10−11 ) × (5.97 × 1024 ) × 120
U=− = = −7.495 × 109 J.
R⊕ (6.378 × 10 )6

Thus, the gain in potential energy, which is the same as the minimum launch
energy, is 7.495 × 109 J.

278
13 WAVE MOTION

13 Wave motion

13.1 Introduction

Waves are small amplitude perturbations which propagate through continuous

media: e.g., gases, liquids, solids, or—in the special case of electromagnetic
waves—a vacuum. Wave motion is a combination of oscillatory and translational
motion. Waves are important because they are the means through which virtually
all information regarding the outside world is transmitted to us. For instance, we
hear things via sound waves propagating through the air, and we see things via
light waves. Now, the physical mechanisms which underlie sound and light wave
propagation are completely different. Nevertheless, sound and light waves pos-
sesses a number of common properties which are intrinsic to wave motion itself.
In this section, we shall concentrate on the common properties of waves, rather
than those properties which are peculiar to particular wave types.

13.2 Waves on a stretched string

Probably the simplest type of wave is that which propagates down a stretched
string. Consider a straight string which is stretched such that it is under uniform
tension T . Let the string run along the x-axis. Suppose that the string is subject
to a small amplitude displacement, in the y-direction, which can vary along its
length. Let y(x, t) be the string’s displacement at position x and time t. What is
the equation of motion for y(x, t)?

Consider an infinitesimal segment of the string which extends from x − δx/2

to x + δx/2. As shown in Fig. 108, this segment is subject to opposing tension
forces, T , at its two ends, which act along the local tangent line to the string.
Here, we are assuming that the string displacement remains sufficiently small
that the tension does not vary in magnitude along the string. Suppose that the
local tangent line to the string subtends angles δθ1 and δθ2 with the x-axis at
x − δx/2 and x + δx/2, respectively—as shown in Fig. 108. Note that these
angles are written as infinitesimal quantities because the string displacement is

279
13 WAVE MOTION 13.2 Waves on a stretched string

y −>
T

δθ 2

δθ1

T x − δ x/2 x + δ x/2 x −>

Figure 108: Forces acting on a segment of a stretched string.

assumed to be infinitesimally small, which implies that the string is everywhere

almost parallel with the x-axis (the string displacement is greatly exaggerated in
Fig. 108, for the sake of clarity).

Consider the y-component of the string segment’s equation of motion. The net
force acting on the segment in the y-direction takes the form

fy (x, t) = T sin δθ2 − T sin δθ1  T (δθ2 − δθ1 ), (13.1)

since sin θ  θ when θ is small. Now, from calculus,

∂y(x − δx/2, t)
= tan δθ1  δθ1 , (13.2)
∂x
∂y(x + δx/2, t)
= tan δθ2  δθ2 , (13.3)
∂x
since the gradient, dy(x)/dx, of the curve y(x) is equal to the tangent of the angle
subtended by this curve with the x-axis. Note that tan θ  θ when θ is small. The
quantity ∂y(x, t)/∂x refers to the derivative of y(x, t) with respect to x, keeping
t constant—such a derivative is known as a partial derivative. Equations (13.1)–
(13.3) can be combined to give
⎛ ⎞
∂y(x + δx/2, t) ∂y(x − δx/2, t) ⎠ ∂2 y(x, t)
fy (x, t) = T ⎝ − = T δx . (13.4)
∂x ∂x ∂x2

280
13 WAVE MOTION 13.2 Waves on a stretched string

Here, ∂2 y(x, t)/∂x2 is the second derivative of y(x, t) with respect to x, keeping t
constant.

Suppose that the string has a mass per unit length μ. It follows that the y
equation of motion of our string segment takes the form
∂2 y(x, t)
μ δx = fy (x, t), (13.5)
∂t2
Here, ∂2 y(x, t)/∂t2 —the second derivative of y(x, t) with respect to t, keeping
x constant—is the y-acceleration of the string segment at position x and time t.
Equations (13.4) and (13.5) yield the final expression for the string’s equation of
motion:
∂2 y T ∂2 y
= . (13.6)
∂t2 μ ∂x2

Equation (13.6) is an example of a wave equation. In fact, all small amplitude

waves satisfy an equation of motion of this basic form. A particular solution of
this type of equation has been known for centuries: i.e.,
y(x, t) = y0 cos (k x − ω t), (13.7)
where y0 , k, and ω are constants. We can demonstrate that (13.7) satisfies (13.6)
by direct substitution. Thus,
∂y
= y0 ω sin (k x − ω t), (13.8)
∂t
∂2 y
2
= −y0 ω2 cos (k x − ω t), (13.9)
∂t
and
∂y
= −y0 k sin (k x − ω t), (13.10)
∂x
∂2 y
2
= −y0 k2 cos (k x − ω t). (13.11)
∂x
Substituting Eqs. (13.9) and (13.11) into Eq. (13.6), we find that the latter equa-
tion is satisfied provided
ω2 T
= . (13.12)
k2 μ
281
13 WAVE MOTION 13.2 Waves on a stretched string

Equation (13.7) describes a pattern of motion which is periodic in both space

and time. This periodicity follows from the well-known periodicity property of the
cosine function: namely, cos(θ + 2 π) = cos θ. Thus, the wave pattern is periodic
in space,
y(x + λ, t) = y(x, t), (13.13)
with periodicity length
2π
λ= . (13.14)
k
Here, λ is known as the wavelength, whereas k is known as the wavenumber. The
wavelength is the distance between successive wave peaks. The wave pattern is
periodic in time,
y(x, t + T ) = y(x, t), (13.15)
with period
2π
T= . (13.16)
ω
The wave period is the oscillation period of the wave disturbance at a given point
in space. The wave frequency (i.e., the number of cycles per second the wave
pattern executes at a given point in space) is written
1 ω
f= = . (13.17)
T 2π
The quantity ω is termed the angular frequency of the wave. Finally, at any given
point in space, the displacement y oscillates between +y0 and −y0 (since the
maximal values of cos θ are ±1). Hence, y0 corresponds to the wave amplitude.

Equation (13.7) also describes a sinusoidal pattern which propagates along the
x-axis without changing shape. We can see this by examining the motion of the
wave peaks, y = +y0 , which correspond to

k x − ω t = n 2 π, (13.18)

where n is an integer. Differentiating the above expression with respect to time,

we obtain
dx ω
= . (13.19)
dt k

282
13 WAVE MOTION 13.2 Waves on a stretched string

Figure 109: A sinusoidal wave propagating down the x-axis. The solid, dotted, dashed, and dot-
dashed curves show the wave displacement at four successive and equally spaced times.

In other words, the wave peaks all propagate along the x-axis with uniform speed
ω
v= . (13.20)
k
It is easily demonstrated that the wave troughs, y = −y0 , propagate with the
same speed. Thus, it is fairly clear that the whole wave pattern moves with speed
v—see Fig. 109. Equations (13.14), (13.17), and (13.20) yield

v = fλ : (13.21)

i.e., a wave’s speed is the product of its frequency and its wavelength. This is true
for all types of (sinusoidal) wave.

Equations (13.12) and (13.20) imply that



T
v=  . (13.22)
μ
In other words, all waves that propagate down a stretched string do so with the
same speed. This common speed is determined by the properties of the string: i.e.,
its tension and mass per unit length. Note, from Eq. (13.7), that the wavelength

283
13 WAVE MOTION 13.3 General waves

λ is arbitrary. However, once the wavelength is specified, the wave frequency

f is fixed via Eqs. (13.21) and (13.22). It follows that short wavelength waves
possess high frequencies, and vice versa.

13.3 General waves

By analogy with the previous discussion, a general wave disturbance propagating

along the x-axis satisfies
∂2 y 2
2 ∂ y
=v , (13.23)
∂t2 ∂x2
where v is the common wave speed. In general, v is determined by the properties
of the medium through which the wave propagates. Thus, for waves propagating
along a string, the wave speed is determined by the string tension and mass
per unit length; for sound waves propagating through a gas, the wave speed
is determined by the gas pressure and density; and for electromagnetic waves
propagating through a vacuum, the wave speed is a constant of nature: i.e.,
c = 3 × 108 m/s2 .

One solution of Eq. (13.23) is

y(x, t) = y0 cos [k (x − v t)]. (13.24)
This is interpreted as a (sinusoidal) wave of amplitude y0 and wavelength λ =
2 π/k which propagates in the +x direction with speed v. It is easily demonstrated
that another equally good solution of Eq. (13.23) is
y(x, t) = y0 cos [k (x + v t)]. (13.25)
This is interpreted as a (sinusoidal) wave of amplitude y0 and wavelength λ =
2 π/k which propagates in the −x direction with speed v.

Equation (13.23) is a linear partial differential equation (PDE): i.e., it is in-

variant under the transformation y → a y + b, where a and b are arbitrary
constants. One important mathematical property of linear PDEs is that their so-
lutions are superposable: i.e., they can be added together and still remain solu-
tions. Thus, if y1 (x, t) and y2 (x, t) are two distinct solutions of Eq. (13.23) then

284
13 WAVE MOTION 13.4 Wave-pulses

x −>

Figure 110: A wave-pulse propagating down the x-axis. The solid, dotted, and dashed curves show
the wave displacement at three successive and equally spaced times.

a y1 (x, t) + b y2 (x, t) (where a and b are arbitrary constants) is also a solution—

this can be seen from inspection of Eq. (13.23). To be more exact, if

y1 (x, t) = a1 cos [k1 (x − v t)] (13.26)

represents a wave of amplitude a1 and wavenumber k1 which propagates in the

+x direction, and
y2 (x, t) = a2 cos [k2 (x + v t)] (13.27)
represents a wave of amplitude a2 and wavenumber k2 which propagates in the
−x direction, then
y(x, t) = y1 (x, t) + y2 (x, t) (13.28)
is a valid solution of the wave equation, and represents the two aforementioned
waves propagating in the same region without affecting one another.

13.4 Wave-pulses

As is easily demonstrated, the most general solution of the wave equation (13.23)
is written
F(x − v t), (13.29)

285
13 WAVE MOTION 13.4 Wave-pulses

where F(p) is an arbitrary function. The above solution is interpreted as a pulse

of arbitrary shape which propagates in the +x direction with speed v, without
changing shape—see Fig. 110. Likewise,

G(x + v t) (13.30)

represents another arbitrary pulse which propagates in the −x direction with

speed v, without changing shape. Note that, unlike our previous sinusoidal wave
solutions, a general wave-pulse possesses a definite propagation speed but does
not possess a definite wavelength or frequency.

What is the relationship between these new wave-pulse solutions and our pre-
vious sinusoidal wave solutions? It turns out that any wave-pulse can be built up
from a suitable linear superposition of sinusoidal waves. For instance, if F(x − v t)
represents a wave-pulse propagating down the x-axis, then we can write
∞
F(x − v t) = F̄(k) cos [k (x − v t)] dk, (13.31)
0

where we have assumed that F(−p) = F(p), for the sake of simplicity. The above
formula is basically a recipe for generating the propagating wave-pulse F(x − v t)
from a suitable admixture of sinusoidal waves of definite wavelength and fre-
quency: F̄(k) specifies the required amplitude of the wavelength λ = 2 π/k com-
ponent. How do we determine F̄(k) for a given wave-pulse? Well, a mathematical
result known as Fourier’s theorem yields

2 ∞
F̄(k) = F(p) cos (k p) dp, (13.32)
π 0
The above expression essentially tells us the strength of the wavenumber k com-
ponent of the wave-pulse F(x − v t). Note that the function F̄(k) is known as the
Fourier spectrum of the wave-pulse F(x − v t).

Figures 111 and 112 show two different wave-pulses and their associated
Fourier spectra. Note how, by combining sinusoidal waves of varying wavenum-
ber in different proportions, it is possible to build up wave-pulses of completely
different shape.

286
13 WAVE MOTION 13.4 Wave-pulses

Figure 111: A propagating wave-pulse, F(x − v t), and its associated Fourier spectrum, F̄(k).

287
13 WAVE MOTION 13.4 Wave-pulses

Figure 112: A propagating wave-pulse, F(x − v t), and its associated Fourier spectrum, F̄(k).

288
13 WAVE MOTION 13.5 Standing waves

13.5 Standing waves

Up to now, all of the wave solutions that we have investigated have been propa-
gating solutions. Is it possible to construct a wave solution which does not prop-
agate? Suppose we combine a sinusoidal wave of amplitude y0 and wavenumber
k which propagates in the +x direction,

y1 (x, t) = y0 cos (k x − ω t), (13.33)

with a second sinusoidal wave of amplitude y0 and wavenumber k which propa-

gates in the −x direction,

y2 (x, t) = y0 cos (k x + ω t). (13.34)

The net result is

y(x, t) = y1 (x, t) + y2 (x, t) = y0 [cos (k x − ω t) + cos (k x + ω t)] . (13.35)
Making use of the standard trigonometric identity
   
x+y x−y
cos x + cos y = 2 cos cos , (13.36)
2 2
we obtain
y(x, t) = 2 y0 cos (k x) cos (ω t). (13.37)
The pattern of motion specified by the above expression is illustrated in Fig. 113.
It can be seen that the wave pattern does not propagate along the x-axis. Note,
however, that the amplitude of the wave now varies with position. At certain
points, called nodes, the amplitude is zero. At other points, called anti-nodes,
the amplitude is maximal. The nodes are halfway between successive anti-nodes,
and both nodes and anti-nodes are evenly spaced half a wavelength apart.

The standing wave shown in Fig. 113 can be thought of as the interference
pattern generated by combining the two traveling wave solutions y1 (x, t) and
y2 (x, t). At the anti-nodes, the waves reinforce one another, so that the oscillation
amplitude becomes double that associated with each wave individually—this is
termed constructive interference. At the nodes, the waves completely cancel one
another out—this is termed destructive interference.

289
13 WAVE MOTION 13.5 Standing waves

node

anti−node

Figure 113: A standing wave. The various curves show the wave displacement at different times.

Most musical instruments work by exciting standing waves. For instance,

stringed instruments excite standing waves on strings, whereas wind instruments
excite standing waves in columns of air. Consider a guitar string of length L.
Suppose that the string runs along the x-axis, and extends from x = 0 to x = L.
Since the ends of the string are fixed, any wave excited on the string must satisfy
the constraints
y(0, t) = y(L, t) = 0. (13.38)
It is fairly clear that no propagating wave solution of the form y 0 cos [k (x ±
v t)] can satisfy these constraints. However, a standing wave can easily satisfy
the constraints, provided two of its nodes coincide with the ends of the string.
Since the nodes in a standing wave pattern are spaced half a wavelength apart,
it follows that the wave frequency must be adjusted such that an integer number
of half-wavelengths fit on the string. In other words,
λ
L=n , (13.39)
2
where n = 1, 2, 3, . . .. Now, from Eqs. (13.21) and (13.22),


T
fλ =  , (13.40)
μ
290
13 WAVE MOTION 13.6 The Doppler effect

where T and μ are the tension and mass per unit length of the string, respectively.
The above two equations can be combined to give

n 
T
 .
f= (13.41)
2L μ
Thus, the standing waves that can be excited on a guitar string have frequencies
f0 , 2 f0 , 3 f0 , etc., which are integer multiples of

1 
T
 .
f0 = (13.42)
2L μ
These frequencies are transmitted to our ear, via sound waves which oscillate in
sympathy with the guitar string, and are interpreted as musical notes. To be more
exact, the frequencies correspond to notes spaced an octave apart. The frequency
f0 is termed the fundamental frequency, whereas the frequencies 2 f0 , 3 f0 , etc. are
termed the overtone harmonic frequencies. When a guitar string is plucked an
admixture of standing waves, consisting predominantly of the fundamental har-
monic wave, is excited on the string. The fundamental harmonic determines the
musical note which the guitar string plays. However, it is the overtone harmonics
which give the note its peculiar timbre. Thus, a trumpet sounds different to a
guitar, even when they are both playing the same note, because a trumpet excites
a different mix of overtone harmonics than a guitar.

13.6 The Doppler effect

Consider a sinusoidal wave of wavenumber k and angular frequency ω propagat-

ing in the +x direction:

y(x, t) = y0 cos (k x − ω t). (13.43)

The wavelength and frequency of the wave, as seen by a stationary observer, are
λ = 2 π/k and f = ω/2 π, respectively. Consider a second observer moving with
uniform speed vo in the +x direction. What are the wavelength and frequency of
the wave, as seen by the second observer? Well, the x-coordinate in the moving
observer’s frame of reference is x  = x − v0 t (see Sect. 4.9). Of course, both

291
13 WAVE MOTION 13.6 The Doppler effect

observers measure the same time. Hence, in the second observer’s frame of ref-
erence the wave takes the form

y(x  , t) = y0 cos (k x  − ω  t), (13.44)

where
ω  = ω − k vo . (13.45)
Here, we have simply replaced x by x  + vo t in Eq. (13.43). Clearly, the moving
observer sees a wave possessing the same wavelength (i.e., the same k) but a
different frequency (i.e., a different ω) to that seen by the stationary observer.
This phenomenon is called the Doppler effect. Since v = ω/k, it follows that the
wave speed is also shifted in the moving observer’s frame of reference. In fact,

v  = v − vo , (13.46)

where v  is the wave speed seen by the moving observer. Finally, since v = f λ,
and the wavelength is the same in both the moving and stationary observers’
frames of reference, the wave frequency experienced by the moving observer is
 
 vo
f = 1− f. (13.47)
v
Thus, the moving observer sees a lower frequency wave than the stationary ob-
server. This occurs because the moving observer is traveling in the same direction
as the wave, and is therefore effectively trying to catch it up. It is easily demon-
strated that an observer moving in the opposite direction to a wave sees a higher
frequency than a stationary observer. Hence, the general Doppler shift formula
(for a moving observer and a stationary wave source) is
 
 vo
f = 1∓ f, (13.48)
v
where the upper/lower signs correspond to the observer moving in the same/opposite
direction to the wave.

Consider a stationary observer measuring a wave emitted by a source which is

moving towards the observer with speed vs . Let v be the propagation speed of the
wave. Consider two neighbouring wave crests emitted by the source. Suppose

292
13 WAVE MOTION 13.6 The Doppler effect

that the first is emitted at time t = 0, and the second at time t = T , where
T = 1/f is the wave period in the frame of reference of the source. At time t, the
first wave crest has traveled a distance d1 = v t towards the observer, whereas
the second wave crest has traveled a distance d2 = v (t − T ) + vs T (measured
from the position of the source at t = 0). Here, we have taken into account the
fact that the source is a distance vs T closer to the observer when the second wave
crest is emitted. The effective wavelength, λ  , seen by the observer is the distance
between neighbouring wave crests. Hence,
λ  = d1 − d2 = (v − vs ) T. (13.49)
Since v = f  λ  , the effective frequency f  seen by the observer is
f
f = , (13.50)
1 − vs /v
where f is the wave frequency in the frame of reference of the source. We con-
clude that if the source is moving towards the observer then the wave frequency
is shifted upwards. Likewise, if the source is moving away from the observer
then the frequency is shifted downwards. This manifestation of the Doppler effect
should be familiar to everyone. When an ambulance passes us on the street, its
siren has a higher pitch (i.e., a high frequency) when it is coming towards us than
when it is moving away from us. Of course, the oscillation frequency of the siren
never changes. It is the Doppler shift induced by the motion of the siren with
respect to a stationary listener which causes the frequency change.

The general formula for the shift in a wave’s frequency induced by relative
motion of the observer and the source is
⎛ ⎞
1 ∓ v o /v
f = ⎝ ⎠ f, (13.51)
1 ± vs /v
where vo is the speed of the observer, and vs is the speed of the source. The
upper/lower signs correspond to relative motion by which the observer and the
source move apart/together.

Probably the most notorious use of the Doppler effect in everyday life is in
police speed traps. In a speed trap, a policeman fires radar waves (i.e., electro-
magnetic waves of centimeter wavelength) of fixed frequency at an oncoming

293
13 WAVE MOTION 13.6 The Doppler effect

car. These waves reflect off the car, which effectively becomes a moving source.
Hence, by measuring the frequency increase of the reflected waves, the policeman
can determine the car’s speed.

Worked example 13.1: Piano range

Question: A piano emits sound waves whose frequencies range from fl = 28 Hz to

fh = 4200 Hz. What range of wavelengths is spanned by these waves? The speed
of sound in air is v = 343 m/s.

Answer: The relationship between a wave’s frequency, f, wavelength, λ, and

speed, v, is
v = f λ.
Hence, λ = v/f. The shortest wavelength (which corresponds to the highest
frequency) is
v 343
λl = = = 8.1667 × 10−2 m.
fh 4200
The longest wavelength (which corresponds to the lowest frequency) is
v 343
λh = = = 12.250 m.
fl 28

Worked example 13.2: Middle C

Question: A steel wire in a piano has a length of L = 0.9 m and a mass of m =

5.4 g. To what tension T must this wire be stretched so that its fundamental
vibration corresponds to middle C: i.e., the vibration possess a frequency f =
261.6 Hz.

Answer: The fundamental standing wave on a stretched wire is such that the
length L of the wire corresponds to half the wavelength λ of the wave. Hence,

λ = 2 L = 1.80 m.

294
13 WAVE MOTION 13.6 The Doppler effect

The propagation speed of waves on the wire is given by

v = f λ = 261.6 × 1.80 = 470.88 m/s.
Furthermore, the string’s mass per unit length is
m 5.4 × 10−3
μ= = = 6.00 × 10−3 kg/m.
L 0.9
Now, the relationship between the wave propagation speed, v, the mass per unit
length, μ, and the tension, T , of a stretched wire is


T
v=  .
μ
Thus,
T = v2 μ = (470.88)2 × 6.00 × 10−3 = 1.330 × 103 N.

Worked example 13.3: Sinusoidal wave

Question: A wave is described by

y = A sin (k x − ω t),
where A = 4 cm, k = 2.65 rad./m, and ω = 4.78 rad./s. Moreover, x is in meters
and t is in seconds. What are the wavelength, frequency, and propagation speed
of the wave?

Answer: We identify A as the wave amplitude, k as the wavenumber, and ω as

the angular frequency. Now, k = 2 π/λ, where λ is the wavelength. Hence,
2π 2 × π
λ= = = 2.371 m.
k 2.65
Furthermore, ω = 2 π f, where f is the frequency. Hence,
ω 4.78
f= = = 0.7608 Hz.
2π 2 × π
Finally, v = f λ, where v is the propagation speed. Thus,
v = 0.7608 × 2.371 = 1.804 m/s.

295
13 WAVE MOTION 13.6 The Doppler effect

Worked example 13.4: Truck passing stationary siren

Question: A truck, moving at vo = 80 km/hr, passes a stationary police car whose

siren has a frequency of f = 500 Hz. What is the frequency change heard by the
truck driver as the truck passes the police car? The speed of sound is v = 343 m/s.

Answer: The truck’s speed is

80 × 1000
vo = = 22.22 m/s.
3600
When the truck is moving towards the police car, the siren’s apparent frequency
is    
vo 22.22
f1 = 1 + f= 1+ × 500 = 532.39 Hz.
v 343
When the truck is moving away from the police car, the siren’s apparent frequency
is    
vo 22.22
f2 = 1 − f= 1− × 500 = 467.61 Hz.
v 343
Hence, the frequency shift is
Δf = f1 − f2 = 532.39 − 467.61 = 64.79 Hz.

Worked example 13.5: Ambulance and car

Question: An ambulance is traveling down a straight road at speed vs = 42 m/s.

The ambulance approaches a car which is traveling on the same road, in the same
direction, at speed vo = 33 m/s. The ambulance driver hears his/her siren at a
frequency of f = 500 Hz. At what frequency does the driver of the car hear the
siren? The speed of sound is v = 343 m/s.

Answer: The apparent frequency f  of a sound wave is given by

⎛ ⎞
1 − vo /v ⎠
f = ⎝ f,
1 − vs /v
where vo is the speed of the observer (i.e., the car driver), vs is the speed of the
source (i.e., the ambulance), v is the speed of sound, and f is the wave frequency

296
13 WAVE MOTION 13.6 The Doppler effect

in the frame of reference of the source. We have chosen a minus sign in the
numerator of the above formula because the observer is moving away from the
source, leading to a downward Doppler shift. We have chosen a minus sign in
the denominator of the above formula because the source is moving towards the
observer, leading to a upward Doppler shift. Hence,
⎛ ⎞
 1 − 33/343 ⎠
f =⎝ × 500 = 514.95 Hz.
1 − 42/343

297