2.

Forces

In this section, we describe a number of different forces that arise in Newtonian me-
chanics. Throughout, we will restrict attention to the motion of a single particle. (We’ll
look at what happens when we have more than one particle in Section 5). We start
by describing the key idea of energy conservation, followed by a description of some
common and important forces.

2.1 Potentials in One Dimension

Let’s start by considering a particle moving on a line, so its position is determined by
a single function x(t). For now, suppose that the force on the particle depends only on
the position, not the velocity: F = F (x). We define the potential V (x) (also called the
potential energy) by the equation

dV
F (x) = − (2.1)
dx
The potential is only defined up to an additive constant. We can always invert (2.1)
by integrating both sides. The integration constant is now determined by the choice of
lower limit of the integral,
Z x
V (x) = − dx′ F (x′ )
x0

Here x′ is just a dummy variable. (Do not confuse the prime with differentiation! In
this course we will only take derivatives of position x with respect to time and always
denote them with a dot over the variable). With this definition, we can write the
equation of motion as

dV
mẍ = − (2.2)
dx
For any force in one-dimension which depends only on the position, there exists a
conserved quantity called the energy,

1
E = mẋ2 + V (x)
2

The fact that this is conserved means that Ė = 0 for any trajectory of the particle which
obeys the equation of motion. While V (x) is called the potential energy, T = 21 mẋ2 is
called the kinetic energy. Motion satisfying (2.2) is called conservative.

– 10 –
 It is not hard to prove that E is conserved. We need only differentiate to get
 
dV dV
Ė = mẋẍ + ẋ = ẋ mẍ + =0
dx dx

where the last equality holds courtesy of the equation of motion (2.2).

In any dynamical system, conserved quantities of this kind are very precious. We
will spend some time in this course fishing them out of the equations and showing how
they help us simplify various problems.

An Example: A Uniform Gravitational Field

In a uniform gravitational field, a particle is subjected to a constant force, F = −mg
where g ≈ 9.8 ms−2 is the acceleration due to gravity near the surface of the Earth.
The minus sign arises because the force is downwards while we have chosen to measure
position in an upwards direction which we call z. The potential energy is

V = mgz

Notice that we have chosen to have V = 0 at z = 0. There is nothing that forces us

to do this; we could easily add an extra constant to the potential to shift the zero to
some other height.

The equation of motion for uniform acceleration is

z̈ = −g

Which can be trivially integrated to give the velocity at time t,

ż = u − gt (2.3)

where u is the initial velocity at time t = 0. (Note that z is measured in the upwards
direction, so the particle is moving up if ż > 0 and down if ż < 0). Integrating once
more gives the position
1
z = z0 + ut − gt2 (2.4)
2
where z0 is the initial height at time t = 0. Many high schools teach that (2.3) and
(2.4) — the so-called “suvat” equations — are key equations of mechanics. They are
not. They are merely the integration of Newton’s second law for constant acceleration.
Do not learn them; learn how to derive them.

– 11 –
Another Simple Example: The Harmonic Oscillator
The harmonic oscillator is, by far, the most important dynamical system in all of
theoretical physics. The good news is that it’s very easy. (In fact, the reason that
it’s so important is precisely because it’s easy!). The potential energy of the harmonic
oscillator is defined to be
1
V (x) = kx2
2
The harmonic oscillator is a good model for, among other things, a particle attached
to the end of a spring. The force resulting from the energy V is given by F = −kx
which, in the context of the spring, is called Hooke’s law. The equation of motion is

mẍ = −kx

which has the general solution

r
k
x(t) = A cos(ωt) + B sin(ωt) with ω =
m
Here A and B are two integration constants and ω is called the angular frequency. We
see that all trajectories are qualitatively the same: they just bounce backwards and
forwards around the origin. The coefficients A and B determine the amplitude of the
oscillations, together with the phase at which you start the cycle. The time taken to
complete a full cycle is called the period
2π
T = (2.5)
ω
The period is independent of the amplitude. (Note that, annoyingly, the kinetic energy
is also often denoted by T as well. Do not confuse this with the period. It should
hopefully be clear from the context).

If we want to determine the integration constants A and B for a given trajectory, we

need some initial conditions. For example, if we’re given the position and velocity at
time t = 0, then it’s simple to check that A = x(0) and Bω = ẋ(0).

2.1.1 Moving in a Potential

Let’s go back to the general case of a potential V (x) in one dimension. Although the
equation of motion is a second order differential equation, the existence of a conserved
energy magically allows us to turn this into a first order differential equation,
r
1 2 dx 2
E = mẋ + V (x) ⇒ =± (E − V (x))
2 dt m

– 12 –
This gives us our first hint of the importance of conserved quantities in helping solve
a problem. Of course, to go from a second order equation to a first order equation, we
must have chosen an integration constant. In this case, that is the energy E itself. Given
a first order equation, we can always write down a formal solution for the dynamics
simply by integrating,
Z x
dx′
t − t0 = ± q (2.6)
2 ′ ))
x0
m
(E − V (x

As before, x′ is a dummy variable. If we can do the integral, we’ve solved the problem.
If we can’t do the integral, you sometimes hear that the problem has been “reduced to
quadrature”. This rather old-fashioned phrase really means “I can’t do the integral”.
But, it is often the case that having a solution in this form allows some of its properties
to become manifest. And, if nothing else, one can always just evaluate the integral
numerically (i.e. on your laptop) if need be.

Getting a Feel for the Solutions

Given the potential energy V (x), it is often very simple to figure out the qualitative
nature of any trajectory simply by looking at the form of V (x). This allows us to answer
some questions with very little work. For example, we may want to know whether the
particle is trapped within some region of space or can escape to infinity.

Let’s illustrate this with an example. Consider the cubic potential

V (x) = m(x3 − 3x) (2.7)

If we were to substitute this into the general form (2.6), we’d get a fearsome looking
integral which hasn’t been solved since Victorian times1 .

Even without solving the integral, we can make progress. The potential is plotted
in Figure 2. Let’s start with the particle sitting stationary at some position x0 . This
means that the energy is

E = V (x0 )

and this must remain constant during the subsequent motion. What happens next
depends only on x0 . We can identify the following possibilities
1
Ok, I’m exaggerating. The resulting integral is known as an elliptic integral. Although it can’t
be expressed in terms of elementary functions, it has lots of nice properties and has been studied to
death. 100 years ago, this kind of thing was standard fare in mathematics. These days, we usually have
more interesting things to teach. Nonetheless, the study of these integrals later resulted in beautiful
connections to geometry through the theory of elliptic functions and elliptic curves.

– 13 –
 V(x)

2m

x
−1 +1 +2

−2m

Figure 2: The cubic potential

• x0 = ±1: These are the local maximum and minimum. If we drop the particle at
these points, it stays there for all time.

• x0 ∈ (−1, +2): Here the particle is trapped in the dip. It oscillates backwards
and forwards between the two points with potential energy V (x0 ). The particle
can’t climb to the right because it doesn’t have the energy. In principle, it could
live off to the left where the potential energy is negative, but to get there it would
have to first climb the small bump at x = −1 and it doesn’t have the energy to
do so. (There is an assumption here which is implicit throughout all of classical
mechanics: the trajectory of the particle x(t) is a continuous function).

• x0 > 2: When released, the particle falls into the dip, climbs out the other side,
before falling into the void x → −∞.

• x0 < −1: The particle just falls off to the left.

• x0 = +2: This is a special value, since E = 2m which is the same as the potential
energy at the local maximum x = −1. The particle falls into the dip and starts
to climb up towards x = −1. It can never stop before it reaches x = −1 for at
its stopping point it would have only potential energy V < 2m. But, similarly, it
cannot arrive at x = −1 with any excess kinetic energy. The only option is that
the particle moves towards x = −1 at an ever decreasing speed, only reaching the
maximum at time t → ∞. To see that this is indeed the case, we can consider
the motion of the particle when it is close to the maximum. We write x ≈ −1 + ǫ
with ǫ ≪ 1. Then, dropping the ǫ3 term, the potential is

V (x = −1 + ǫ) ≈ 2m − 3mǫ2 + . . .

– 14 –
 and, using (2.6), the time taken to reach x = −1 + ǫ is
Z ǫ
dǫ′
 
1 ǫ
t − t0 = − √ = − √ log
ǫ0 6ǫ′ 6 ǫ0

The logarithm on the right-hand side gives a divergence as ǫ → 0. This tells us

that it indeed takes infinite time to reach the top as promised.

One can easily play a similar game to that above if the starting speed is not zero. In
general, one finds that the particle is trapped in the dip x ∈ [−1, +1] if its energy lies
in the interval E ∈ [−2m, 2m].

2.1.2 Equilibrium: Why (Almost) Everything is a Harmonic Oscillator

A particle placed at an equilibrium point will stay there for all time. In our last example
with a cubic potential (2.7), we saw two equilibrium points: x = ±1. In general, if
we want ẋ = 0 for all time, then clearly we must have ẍ = 0, which, from the form of
Newton’s equation (2.2), tells us that we can identify the equilibrium points with the
critical points of the potential,
dV
=0
dx
What happens to a particle that is close to an equilibrium point, x0 ? In this case, we
can Taylor expand the potential energy about x = x0 . Because, by definition, the first
derivative vanishes, we have
1
V (x) ≈ V (x0 ) + V ′′ (x0 )(x − x0 )2 + . . . (2.8)
2
To continue, we need to know about the sign of V ′′ (x0 ):

• V ′′ (x0 ) > 0: In this case, the equilibrium point is a minimum of the potential
and the potential energy is that of a harmonic oscillator. From our discussion of
Section 2.1.2, we know that the particle oscillates backwards and forwards around
x0 with frequency
r
V ′′ (x0 )
ω=
m
Such equilibrium points are called stable. This analysis shows that if the ampli-
tude of the oscillations is small enough (so that we may ignore the (x−x0 )3 terms
in the Taylor expansion) then all systems oscillating around a stable fixed point
look like a harmonic oscillator.

– 15 –
 • V ′′ (x0 ) < 0: In this case, the equilibrium point is a maximum of the potential.
The equation of motion again reads
mẍ = −V ′′ (x0 ) (x − x0 )
But with V ′′ < 0, we have ẍ > 0 when x − x0 > 0. This means that if we displace
the system a little bit away from the equilibrium point, then the acceleration
pushes it further away. The general solution is
r
−V ′′ (x0 )
x − x0 = Aeαt + Be−αt with α =
m
Any solution with the integration constant A 6= 0 will rapidly move away from
the fixed point. Since our whole analysis started from a Taylor expansion (2.8),
neglecting terms of order (x − x0 )3 and higher, our approximation will quickly
break down. We say that such equilibrium points are unstable.

Notice that there are solutions around unstable fixed points with A = 0 and
B 6= 0 which move back towards the maximum at late times. These finely tuned
solutions arise in the kind of situation that we described for the cubic potential
where you drop the particle at a very special point (in the case of the cubic
potential, this point was x = 2) so that it just reaches the top of a hill in infinite
time. Clearly these solutions are not generic: they require very special initial
conditions.

• Finally, we could have V ′′ (x0 ) = 0. In this case, there is nothing we can say about
the dynamics of the system without Taylor expanding the potential further.

Yet Another Example: The Pendulum

Consider a particle of mass m attached to the end of a light rod of x

length l. This counts as a one-dimensional system because we need θ

specify only a single coordinate to say what the system looks like at y
length, l

a given time. The best coordinate to choose is θ, the angle that the T

rod makes with the vertical. The equation of motion is m

g
θ̈ = − sin θ (2.9)
l
The energy is mg

1 Figure 3:
E = ml2 θ̇2 − mgl cos θ
2
(Note: Since θ is an angular variable rather than a linear variable, the kinetic energy is
a little different. Hopefully this is familiar from earlier courses on mechanics. However,
we will rederive this result in Section 4).

– 16 –
 There are two qualitatively different motions of the pendulum. If E > mgl, then the
kinetic energy can never be zero. This means that the pendulum is making complete
circles. In contrast, if E < mgl, the pendulum completes only part of the circle before
it comes to a stop and swings back the other way. If the highest point of the swing is
θ0 , then the energy is

E = −mgl cos θ0

We can determine the period T of the pendulum using (2.6). It’s actually best to
calculate the period by taking 4 times the time the pendulum takes to go from θ = 0
to θ = θ0 . We have
T /4 θ0
dθ
Z Z
T =4 dt = 4 p
0 0 2E/ml2 + (2g/l) cos θ
s
θ0
l dθ
Z
=4 √ (2.10)
g 0 2 cos θ − 2 cos θ0
p
We see that the period is proportional to l/g multiplied by some dimensionless num-
ber given by (4 times) the integral. For what it’s worth, this integral turns out to be,
once again, an elliptic integral.

For small oscillations, we can write cos θ ≈ p1 − 12 θ2 and the pendulum becomes a
harmonic oscillator with angular frequency ω = g/l. If we replace the cos θ’s in (2.10)
by their Taylor expansion, we have
s Z s Z s
θ0 1
l dθ l dx l
T =4 p =4 √ = 2π
g 0 2
θ0 − θ2 g 0 1−x 2 g

This agrees with our result (2.5) for the harmonic oscillator.

2.2 Potentials in Three Dimensions

Let’s now consider a particle moving in three dimensional R3 . Here things are more
interesting. Firstly, it is possible to have energy conservation even if the force depends
on the velocity. We will see how this can happen in Section 2.4. Conversely, forces
which only depend on the position do not necessarily conserve energy: we need an extra
condition. For now, we restrict attention to forces of the form F = F(x). We have the
following result:

– 17 –
Claim: There exists a conserved energy if and only if the force can be written in
the form
F = −∇V (2.11)
for some potential function V (x). This means that the components of the force must
be of the form Fi = −∂V /∂xi . The conserved energy is then given by
1
E = mẋ · ẋ + V (x) (2.12)
2

Proof: The proof that E is conserved if F takes the form (2.11) is exactly the same as
in the one-dimensional case, together with liberal use of the chain rule. We have
dE ∂V ∂xi
= mẋ · ẍ + i using summation convention
dt ∂x ∂t
= ẋ · (mẍ + ∇V ) = 0
where the last equality follows from the equation of motion which is mẍ = −∇V .

To go the other way, we must prove that if there exists a conserved energy E taking
the form (2.12) then the force is necessarily given by (2.11). To do this, we need the
concept of work. If a force F acts on a particle and succeeds in moving it from x(t1 )
to x(t2 ) along a trajectory C, then the work done by the force is defined to be
Z
W = F · dx
C

This is a line integral (of the kind you’ve met in the Vector Calculus course). The
scalar product means that we take the component of the force along the direction of
the trajectory at each point. We can make this clearer by writing
Z t2
dx
W = F· dt
t1 dt
The integrand, which is the rate of doing work, is called the power, P = F · ẋ. Using
Newton’s second law, we can replace F = mẍ to get
Z t2 Z t2
1 d
W =m ẍ · ẋ dt = m (ẋ · ẋ) dt = T (t2 ) − T (t1 )
t1 2 t1 dt

where
1
T ≡ m ẋ · ẋ
2
is the kinetic energy. (You might think that K is a better name for kinetic energy. I’m
inclined to agree. Except in all advanced courses of theoretical physics, kinetic energy
is always denoted T which is why I’ve adopted the same notation here).

– 18 –
 So the total work done is proportional to the change in kinetic energy. If we want to
have a conserved energy of the form (2.12), then the change in kinetic energy must be
equal to the change in potential energy. This means we must be able to write
Z
W = F · dx = V (x(t1 )) − V (x(t2 )) (2.13)
C

In particular, this result tells us that the work done must be independent of the tra-
jectory C; it can depend only on the end points x(t1 ) and x(t2 ). But a simple result
(which you will prove in your Vector Calculus course) says that (2.13) holds only for
forces of the form

F = −∇V

as required .

Forces in three dimensions which take the form F = −∇V are called conservative.
You will also see in the Vector Calculus course that forces in R3 are conservative if and
only if ∇ × F = 0.

2.2.1 Central Forces

A particularly important class of potentials are those which depend only on the distance
to a fixed point, which we take to be the origin

V (x) = V (r)

where r = |x|. The resulting force also depends only on the distance to the origin and,
moreover, always points in the direction of the origin,

dV
F(r) = −∇V = − x̂ (2.14)
dr
Such forces are called central. In these lectures, we’ll also use the notation r̂ = x̂ to
denote the unit vector pointing radially from the origin to the position of the particle.
(In other courses, you may see this same vector denoted as er ).

In the vector calculus course, you will spend some time computing quantities such
as ∇V in spherical polar coordinates. But, even without such practice, it is a simple
matter to show that the force (2.14) is indeed aligned with the direction to the origin.
If x = (x1 , x2 , x3 ) then the radial distance is r 2 = x21 + x22 + x23 , from which we can

– 19 –
compute ∂r/∂xi = xi /r for i = 1, 2, 3. Then, using the chain rule, we have
 
∂V ∂V ∂V
∇V = , ,
∂x1 ∂x2 ∂x3
 
dV ∂r dV ∂r dV ∂r
= , ,
dr ∂x1 dr ∂x2 dr ∂x3
dV  x1 x2 x3  dV
= , , = x̂
dr r r r dr
2.2.2 Angular Momentum
We will devote all of Section 4 to the study L
of motion in central forces. For now, we will
just mention what is important about central
forces: they have an extra conserved quantity. x
This is a vector L called angular momentum, x

L = mx × ẋ
Figure 4:
Notice that, in contrast to the momentum p = mẋ, the angular momentum L depends
on the choice of origin. It is a perpendicular to both the position and the momentum.

Let’s look at what happens to angular momentum in the presence of a general force
F. When we take the time derivative, we get two terms. But one of these contains
ẋ × ẋ = 0. We’re left with
dL
= mx × ẍ = x × F
dt
The quantity τ = x × F is called the torque. This gives us an equation for the change
of angular momentum that is very similar to Newton’s second law for the change of
momentum,
dL
=τ
dt
Now we can see why central forces are special. When the force F lies in the same
direction as the position x of the particle, we have x × F = 0. This means that the
torque vanishes and angular momentum is conserved
dL
=0
dt
We’ll make good use of this result in Section 4 where we’ll see a number of important
examples of central forces.

– 20 –
2.3 Gravity
To the best of our knowledge, there are four fundamental forces in Nature. They are

• Gravity

• Electromagnetism

• Strong Nuclear Force

• Weak Nuclear Force

The two nuclear forces operate only on small scales, comparable, as the name suggests,
to the size of the nucleus (r0 ≈ 10−15 m). We can’t really give an honest description of
these forces without invoking quantum mechanics and, for this reason, we won’t discuss
them in this course. (A very rough, and slightly dishonest, classical description of the
strong nuclear force can be given by the potential V (r) ∼ e−r/r0 /r). In this section we
discuss the force of gravity; in the next, electromagnetism.

Gravity is a conservative force. Consider a particle of mass M fixed at the origin. A

particle of mass m moving in its presence experiences a potential energy
GMm
V (r) = − (2.15)
r
Here G is Newton’s constant. It determines the strength of the gravitational force and
is given by

G ≈ 6.67 × 10−11 m3 Kg−1 s−2

The force on the particle is given by

GMm
F = −∇V = − r̂ (2.16)
r2
where r̂ is the unit vector in the direction of the particle. This is Newton’s famous
inverse-square law for gravity. The force points towards the origin. We will devote
much of Section 4 to studying the motion of a particle under the inverse-square force.

2.3.1 The Gravitational Field

The quantity V in (2.15) is the potential energy of a particle of mass m in the presence
of mass M. It is common to define the gravitational field of the mass M to be
GM
Φ(r) = −
r

– 21 –
Φ is sometimes called the Newtonian gravitational field to distinguish it from a more
sophisticated object later introduced by Einstein. It is also sometimes called the grav-
itational potential. It is a property of the mass M alone. The potential energy of the
mass m is then given by V = mΦ.
The gravitational field due to many particles is simply the sum of the field due to
each individual particle. If we fix particles with masses Mi at positions ri , then the
total gravitational field is
X Mi
Φ(r) = −G
i
|r − ri |
The gravitational force that a moving particle of mass m experiences in this field is
X Mi
F = −Gm (r − ri )
i
|r − ri |3
The Gravitational Field of a Planet
The fact that contributions to the Newtonian gravitational potential add in a simple
linear fashion has an important consequence: the external gravitational field of a spher-
ically symmetric object of mass M – such as a star or planet – is the same as that of
a point mass M positioned at the origin.
The proof of this statement is an example of the volume r
integral that you will learn in the Vector Calculus course. We
include it here only for completeness. We let the planet have
density ρ(r) and radius R. Summing over the contribution from R
all points x inside the planet, the gravitational field is given by
Gρ(x)
Z
Φ(r) = − d3 x Figure 5:
|x|≤R |r − x|
It’s best to work in spherical polar coordinates and to choose the polar direction, θ = 0,
to lie in the direction of r. Then r · x = rx cos θ. We can use this to write an expression
for the denominator: |r−x|2 = r 2 +x2 −2rx cos θ. The gravitational field then becomes
Z R Z π Z 2π
ρ(x)x2 sin θ
Φ(r) = −G dx dθ dφ √
0 0 0 r 2 + x2 − 2rx cos θ
Z R Z π
ρ(x)x2 sin θ
= −2πG dx dθ √
0 0 r 2 + x2 − 2rx cos θ
Z R h√
2 1
iθ=π
= −2πG dx ρ(x)x 2 2
r + x − 2rx cos θ
0 rx θ=0
Z R
2πG
=− dx ρ(x)x (|r + x| − |r − x|)
r 0

– 22 –
So far this calculation has been done for any point r, whether inside or outside the
planet. At this point, we restrict attention to points external to the planet. This
means that |r + x| = r + x and |r − x| = r − x and we have
R
4πG GM
Z
Φ(r) = − dx ρ(x)x2 = −
r 0 r

This is the result that we wanted to prove: the gravitational field is the same as that
of a point mass M at the origin.

2.3.2 Escape Velocity

Suppose that you’re trapped on the the surface of a planet of radius R. (This should
be easy). Let’s firstly ask what gravitational potential energy you feel. Assuming you
can only rise a distance z ≪ R from the planet’s surface, we can Taylor expand the
potential energy,

z2
 
GMm GMm z
V (R + z) = − =− 1 − + 2 + ...
R+z R R R

If we’re only interested in small changes in z ≪ R, we need focus only on the second
term, giving

GMm
V (z) ≈ constant + z + ...
R2
This is the familiar potential energy that gives rise to constant acceleration. We usually
write g = GM/R2 . For the Earth, g ≈ 9.8 ms−2.

Now let’s be more ambitious. Suppose we want to escape our

parochial, planet-bound existence. So we decide to jump. How fast
do we have to jump if we wish to truly be free? This, it turns out, is
the same kind of question that we discussed in Section 2.1.1 in the con-
text of particles moving in one dimension and can be determined very
easily using gravitational energy V = −GMm/r. If you jump directly Figure 6:
upwards (i.e. radially) with velocity v, your total energy as you leave
the surface is
1 GMm
E = mv 2 −
2 R
For any energy E < 0, you will eventually come to a halt at position r = −GMm/E,
before falling back. If you want to escape the gravitational attraction of the planet for

– 23 –
ever, you will need energy E ≥ 0. At the minimum value of E = 0, the associated
velocity
r
2GM
vescape =
R
This is the escape velocity.

2.3.3 Inertial vs Gravitational Mass

We have seen two formulae which involve mass, both due to Newton. These are the
second law (1.2) and the inverse-square law for gravity (2.16). Yet the meaning of
mass in these two equations is very different. The mass appearing in the second law
represents the reluctance of a particle to accelerate under any force. In contrast, the
mass appearing in the inverse-square law tells us the strength of a particular force,
namely gravity. Since these are very different concepts, we should really distinguish
between the two different masses. The second law involves the inertial mass, mI

mI ẍ = F

while Newton’s law of gravity involves the gravitational mass, mG

GMG mG
F=− r̂
r2
It is then an experimental fact that

mI = mG (2.17)

Much experimental effort has gone into determining the accuracy of (2.17), most no-
tably by the Hungarian physicist Eötvösh at the turn of the (previous) century. We
now know that the inertial and gravitational masses are equal to within about one part
in 1013 . Currently, the best experiments to study this equivalence, as well as searches
for deviations from Newton’s laws at short distances, are being undertaken by a group
at the University of Washington in Seattle who go by the name Eöt-Wash. A theoret-
ical understanding of the result (2.17) came only with the development of the general
theory of relativity.

2.4 Electromagnetism
Throughout the Universe, at each point in space, there exist two vectors, E and B.
These are known as the electric and magnetic fields. Their role – at least for the
purposes of this course – is to guide any particle that carries electric charge.

– 24 –
 The force experienced by a particle with electric charge q is called the Lorentz force,
 
F = q E(x) + ẋ × B(x) (2.18)

Here we have used the notation E(x) and B(x) to stress that the electric and magnetic
fields are functions of space. Both their magnitude and direction can vary from point
to point.

The electric force is parallel to the electric field. By convention, particles with positive
charge q are accelerated in the direction of the electric field; those with negative electric
charge are accelerated in the opposite direction. Due to a quirk of history, the electron
is taken to have a negative charge given by
qelectron ≈ −1.6 × 10−19 Coulombs
As far as fundamental physics is concerned, a much better choice is to simply say that
the electron has charge 1. All other charges can then be measured relative to this.

The magnetic force looks rather different. It is a velocity dependent force, with
magnitude proportional to the speed of the particle, but with direction perpendicular
to that of the particle. We shall see its effect in simple situations shortly.

In principle, both E and B can change in time. However, here we will consider only
situations where they are static. In this case, the electric field is always of the form
E = −∇φ
For some function φ(x) called the electric potential (or scalar potential or even just the
potential as if we didn’t already have enough things with that name).

For time independent fields, something special happens: energy is conserved.

Claim: The conserved energy is

1
E = mẋ · ẋ + qφ(x)
2
Proof:
Ė = mẋ · ẍ + q∇φ · ẋ = ẋ · (F + q∇φ) = q ẋ · (ẋ × B) = 0
where the last equality occurs because ẋ × B is necessarily perpendicular to ẋ. Notice
that this gives an example of something we promised earlier: a velocity dependent force
which conserves energy. The key part of the derivation is that the velocity dependent
force is perpendicular to the trajectory of the particle. This ensures that the force does
no work. .

– 25 –
2.4.1 The Electric Field of a Point Charge
Charged objects do not only respond to electric fields; they also produce electric fields.
A particle of charge Q sitting at the origin will set up an electric field given by
 
Q Q r̂
E = −∇ = (2.19)
4πǫ0 r 4πǫ0 r 2
where r 2 = x · x. The quantity ǫ0 has the grand name Permittivity of Free Space and
is a constant given by

ǫ0 ≈ 8.85 × 10−12 m−3 Kg−1 s2 C 2

This quantity should be thought of as characterising the strength of the electric inter-
action.

The force between two particles with charges Q and q is given by F = qE with E
given by (2.19). In other words,
qQ r̂
F=
4πǫ0 r 2
This is known as the Coulomb force. It is a remarkable fact that, mathematically,
the force looks identical to the Newtonian gravitational force (2.16): both have the
characteristic inverse-square form. We will study motion in this potential in detail in
Section 4, with particular focus on the Coulomb force in 4.4.

Although the forces of Newton and Coulomb look the same, there is one important
difference. Gravity is always attractive because mass m > 0. In contrast, the electro-
static Coulomb force can be attractive or repulsive because charges q come with both
signs. Further differences between gravity and electromagnetism come when you ask
what happens when sources (mass or charge) move; but that’s a story that will be told
in different courses.

2.4.2 Circles in a Constant Magnetic Field

Motion in a constant electric field is simple: the particle undergoes constant acceleration
in the direction of E. But what about motion in a constant magnetic field B? The
equation of motion is

mẍ = q ẋ × B

Let’s pick the magnetic field to lie in the z-direction and write

B = (0, 0, B)

– 26 –
We can now write the Lorentz force law (2.18) in components. It reads

mẍ = qB ẏ (2.20)
mÿ = −qB ẋ (2.21)
mz̈ = 0

The last equation is easily solved and the particle just travels at constant velocity in
the z direction. The first two equations are more interesting. There are a number of
ways to solve them, but a particularly elegant way is to construct the complex variable
ξ = x + iy. Then adding (2.20) to i times (2.21) gives
B

mξ¨ = −iqB ξ˙
y

which can be integrated to give

ξ = αe−iωt + β x

Figure 7:
where α and β are integration constants and ω is given by
qB
ω=
m
If we choose our initial conditions to be that the particle starts life at t = 0 at the
origin with velocity −v in the y-direction, then α and β are fixed to be
v −iωt

ξ= e −1
ω
Translating this back into x and y coordinates, we have
v v
x= (cos ωt − 1) and y = − sin ωt
ω ω
The end result is that the particle undergoes circles in the plane with angular frequency
ω, known as the cyclotron frequency The time to undergo a full circle is fixed: T = 2π/ω.
In contrast, the size of the circle is v/ω and arises as an integration constant. Circles
of arbitrary sizes are allowed; the only price that you pay is that you have to go faster.

A Comment on Solving Vector Differential Equations

The Lorentz force equation (2.18) gives a good example of a vector differential equation.
The straightforward way to view these is always in components: they are three, coupled,
second order differential equations for x, y and z. This is what we did above when
understanding the motion of a particle in a magnetic field.

– 27 –
 However, one can also attack these kinds of questions without reverting to compo-
nents. Let’s see how this would work in the case of Larmor circles. We start with the
vector equation

mẍ = q ẋ × B (2.22)

To begin, we take the dot product with B. Since the right-hand side vanishes, we’re
left with

ẍ · B = 0

This tells us that the particle travels with constant velocity in the direction of B. This
is simply a rewriting of our previous result z̈ = 0. For simplicity, let’s just assume that
the particle doesn’t move in the B direction, remaining at the origin. This tells us that
the particle moves in a plane with equation

x·B=0 (2.23)

However, we’re not yet done. We started with (2.22) which was three equations. Taking
the dot product always reduces us to a single equation. So there must still be two further
equations lurking in (2.22) that we haven’t yet taken into account. To find them, the
systematic thing to do would be to take the cross product with B. However, in the
present case, it turns out that the simplest way forwards is to simply integrate (2.22)
once, to get

mẋ = qx × B + c

with c a constant of integration. We can now substitute this back into the right-hand
side of (2.22) to find

m2 ẍ = d + q 2 (x × B) × B
= d + q 2 ((x · B)B − (B · B)x)
= −q 2 B 2 x − d/q 2 B 2

where the integration constant now sits in d = qc × B which, by construction, is

perpendicular to B. In the last line, we’ve used the equation (2.23). (Note that if
we’d considered a situation in which the particle was moving with constant velocity
in the B direction, we’d have to work a little harder at this point). The resulting
vector equation looks like three harmonic oscillators, displaced by the vector d/q 2 B 2 ,

– 28 –
oscillating with frequency ω = qB/m. However, because of the constraint (2.23), the
motion is necessarily only in the two directions perpendicular to B. The end result is
d
x= + α1 cos ωt + α2 sin ωt
q2B2
with αi , i = 1, 2 integration constants satisfying αi · B = 0. This is the same result we
found previously.

Admittedly, in this particular example, working with components was somewhat

easier than manipulating the vector equations directly. But this won’t always be the
case — for some problems you’ll make more progress by playing the kind of games that
we’ve described here.

2.4.3 An Aside: Maxwell’s Equations

In the Lorentz force law, the only hint that the electric and magnetic fields are related
is that they both affect a particle in a manner that is proportional to the electric charge.
The connection between them becomes much clearer when things depend on time. A
time dependent electric field gives rise to a magnetic field and vice versa. The dynamics
of the electric and magnetic fields are governed by Maxwell’s equations. In the absence
of electric charges, these equations are given by
∇·E= 0 , ∇·B =0
∂B ∂E
∇×E=− , ∇ × B = c2
∂t ∂t
with c the speed of light. You will learn more about the properties of these equations
in the Electromagnetism and Electrodynamics courses.

For now, it’s worth making one small comment. When we showed that energy is
conserved, we needed both the electric and magnetic field to be time independent.
What happens when they change with time? In this case, energy is still conserved, but
we have to worry about the energy stored in the fields themselves.

2.5 Friction
Friction is a messy, dirty business. While energy is always conserved on a fundamental
level, it doesn’t appear to be conserved in most things that you do every day. If you
slide along the floor in your socks you don’t keep going for ever. At a microscopic level,
your kinetic energy is transferred to the atoms in the floor where it manifests itself as
heat. But if we only want to know how far our socks will slide, the details of all these
atomic processes are of little interest. Instead, we try to summarise everything in a
single, macroscopic force that we call friction.

– 29 –
2.5.1 Dry Friction
Dry friction occurs when two solid objects are in contact. Think of a heavy box being
pushed along the floor, or some idiot sliding in his socks. Experimentally, one finds
that the complicated dynamics involved in friction is usually summarised by the force

F = µR

where R is the reaction force, normal to the floor, and µ is a con- R µR

stant called the coefficient of friction. Usually µ ≈ 0.3, although
it depends on the kind of materials that are in contact. Moreover,
the coefficient is usually, more or less, independent of the velocity.
We won’t have much to say about dry friction in this course. In
mg
fact, we’ve already said it all.
Figure 8:
2.5.2 Fluid Drag
Drag occurs when an object moves through a fluid — either liquid or gas. The resistive
force is opposite to the direction of the velocity and, typically, falls into one of two
categories

• Linear Drag:

F = −γv

where the coefficient of friction, γ, is a constant. This form of drag holds for
objects moving slowly through very viscous fluids. For a spherical object of
radius L, there is a formula due to Stokes which gives γ = 6πηL where η is the
viscosity of the fluid.

• Quadratic Drag:

F = −γ|v|v

Again, γ is called the coefficient of friction. For quadratic friction, γ is usually

proportional to the surface area of the object, i.e. γ ∼ L2 . (This is in contrast to
the coefficient for linear friction where Stokes’ formula gives γ ∼ L). Quadratic
drag holds for fast moving objects in less viscous fluids. This includes objects
falling in air such as, for example, the various farmyard animals dropped by
Galileo from the leaning tower.

– 30 –
 Quadratic drag arises because the object is banging into molecules in the fluid,
knocking them out the way. There is an intuitive way to see this. The force is
proportional to the change of momentum that occurs in each collision. That gives
one factor of v. But the force is also proportional to the number of collisions.
That gives the second factor of v, resulting in a force that scales as v 2 .

One can ask where the cross-over happens between linear and quadratic friction.
Naively, the linear drag must always dominate at low velocities simply because x ≫ x2
when x ≪ 1. More quantitatively, the type of drag is determined by a dimensionless
number called the Reynolds number,

ρvL
R≡ (2.24)
η

where ρ is the density of the fluid while η is the viscosity. For R ≪ 1, linear drag
dominates; for R ≫ 1, quadratic friction dominates.

What is Viscosity?
Above, we’ve mentioned the viscosity of the fluid, η, without really defining it. For
completeness, I will mention here how to measure viscosity.

Place a fluid between two plates, a distance d v

apart. Keeping the lower plate still, move the top plate 11111111111111111111111111111111
00000000000000000000000000000000
00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
at a constant speed v. This sets up a velocity gradient 00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
in the fluid. But, the fluid pushes back. To keep the 00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111 d
00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
upper plate moving at constant speed, you will have to 00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
00000000000000000000000000000000
11111111111111111111111111111111
push with a force per unit area which is proportional to
the velocity gradient, Figure 9:
F v
=η
A d
The coefficient of proportionality, η, is defined to be the (dynamic) viscosity.

2.5.3 An Example: The Damped Harmonic Oscillator

We start with our favourite system: the harmonic oscillator, now with a damping term.
This was already discussed in your Differential Equations course and we include it here
only for completeness. The equation of motion is

mẍ = −kx − γ ẋ

– 31 –
Divide through by m to get

ẍ = −ω02 x − 2αẋ

where ω02 = k/m is the frequency of the undamped harmonic oscillator and α = γ/2m.
We can look for solutions of the form

x = eiβt

Remember that x is real, so we’re using a trick here. We rely on the fact that the
equation of motion is linear so that if we can find a solution of this form, we can take
the real and imaginary parts and this will also be a solution. Substituting this ansatz
into the equation of motion, we find a quadratic equation for β. Solving this, gives the
general solution

x = Aeiω+ t + Beiω− t
p
with ω± = iα ± ω02 − α2 . We identify three different regimes,

• Underdamped: ω02 > α2 . Here the solution takes the form,

x = e−αt AeiΩt + Be−iΩt

√
where Ω = ω 2 − α2 . Here the system oscillates with a frequency Ω < ω0 , while
the amplitude of the oscillations decays exponentially.

• Overdamped: ω02 < α2 . The roots ω± are now purely imaginary and the general
solution takes the form,

x = e−αt AeΩt + Be−Ωt

Now there are no oscillations. Both terms decay exponentially. If you like, the
amplitude decays away before the system is able to undergo even a single oscil-
lation.

• Critical Damping: ω02 = α2 . Now the two roots ω± coincide. With a double root
of this form, the most general solution takes the form,

x = (A + Bt)e−αt

Again, there are no oscillations, but the system does achieve some mild linear
growth for times t < 1/α, after which it decays away.

– 32 –
2.5.4 Terminal Velocity with Quadratic Friction
You can drop a mouse down a thousand-yard mine shaft; and, on arriving
at the bottom, it gets a slight shock and walks away, provided that the ground
is fairly soft. A rat is killed, a man is broken, a horse splashes.

J.B.S. Haldane, On Being the Right Size

Let’s look at a particle of mass m moving in a constant gravitational field, subject to

quadratic friction. We’ll measure the height z to be in the upwards direction, meaning
that if v = dz/dt > 0, the particle is going up. We’ll look at the cases where the
particle goes up and goes down separately.

Coming Down
Suppose that we drop the particle from some height. The equation of motion is given
by
dv
m = −mg + γv 2
dt
It’s worth commenting on the minus signs on the right-hand side. Gravity acts down-
wards, so comes with a minus sign. Since the particle is falling down, friction is acting
upwards so comes with a plus sign. Dividing through by m, we have
dv γv 2
= −g + (2.25)
dt m
Integrating this equation once gives
Z v
dv ′
t=− ′2
0 g − γv /m
p
which can be easily solved by the substitution v = mg/γ tanh x to get
r 
m γ
r
−1
t=− tanh v
γg mg
Inverting this gives us the speed as a function of time
r 
mg γg
r
v=− tanh t
γ m
We now see the effect of friction. As time increases, the velocity does not increase
without bound. Instead, the particle reaches a maximum speed,
mg
r
v→− as t → ∞ (2.26)
γ

– 33 –
This is the terminal velocity. The sign is negative because the particle is falling down-
wards. Notice that if all we wanted was the terminal velocity, then we don’t need to go
through the whole calculation above. We can simply look for solutions of (2.25) with
constant speed, so dv/dt = 0. This obviously gives us (2.26) as a solution. The advan-
tage of going through the full calculation is that we learn how the velocity approaches
its terminal value.

We can now see the origin of the quote we started with. The point is that if we
compare objects of equal density, the masses scale as the volume, meaning m ∼ L3
where L is the linear size of the metric. In contrast, the coefficient of friction usually
scales as surface area, γ ∼ L2 . This means that the terminal velocity depends √ on size.
For objects of equal density, we expect the terminal velocity to scale as v ∼ L. I have
no idea if this is genuinely a big enough effect to make horses splash. (Haldane was a
biologist, so he should know what it takes to make an animal splash. But in his essay
he assumed linear drag rather than quadratic, so maybe not).

Going Up
Now let’s think about throwing a particle upwards. Since both gravity and friction are
now acting downwards, we get a flip of a minus sign in the equation of motion. It is
now
dv γv 2
= −g − (2.27)
dt m
Suppose that we throw the object up with initial speed u and we want to figure out
the maximum height, h, that it reaches. We could follow our earlier calculation and
integrate (2.27) to determine v = v(t). But since we aren’t asking about time, it’s
much better to instead consider velocity as a function of distance: v = v(z). We write

dv dv dz dv γv 2
= =v = −g −
dt dz dt dz m
which can be rewritten as
1 d(v 2 ) γv 2
= −g −
2 dz m
Now we can integrate this equation to get velocity as a function of distance. Writing
y = v 2 , we have
0 h
dy mh  γy iy=0
Z Z
= −2 dz ⇒ log g + = −2h
u2 g + γy/m 0 γ m y=u2

– 34 –
which we can rearrange to get the final answer,
γu2
 
m
h= log 1 +
2γ mg
It’s worth looking at what happens when the effect of friction is small. Naively, it
looks like we’re in trouble here because as γ → 0, the term in front gets very large.
But surely the height shouldn’t go to infinity just because the friction is small. The
resolution to this is that the log is also getting small in this limit. Expanding the log,
we have
u2 γu2
 
h= 1− + ...
2g 2mg
Here the leading term is indeed the answer we would get in the absence of friction; the
subleading terms tell us how much the friction, γ, lowers the attained height.

Linear Drag and Ohm’s Law

Consider an electron moving in a conductor. As we’ve seen, a constant electric field
causes the electron to accelerate. A fairly good model for the physics of a conduc-
tor, known as the Drude model, treats the electron as a classical particle with linear
damping. The resulting equation of motion is
mẍ = −eE − γv
As in the previous example, we can figure out the terminal velocity by setting ẍ = 0,
to get
eE
v=−
γ
In a conductor, the velocity of the electron v gives the current density, j,
j = −env
where n is the density of electrons. This then gives us a relationship between the
current density and the electric field
j = σE
The quantity σ = e2 n/γ is called the conductivity. This equation is Ohm’s law. How-
ever, it’s probably not yet in the form you know and love. If the wire has length L and
cross-sectional area A, then the current I is defined as I = jA. Meanwhile, the voltage
dropped across the wire is V = EL. With this in hand, we can rewrite Ohm’s law as
V = IR
where the resistance is given by R = L/σA.

– 35 –
A 3d Example: A Projectile with Linear Drag
All our examples so far have been effectively one-dimensional. Here we give a three
dimensional example which provides another illustration of how to treat vector differ-
ential equations and, specifically, how to work with vector constants on integration.
We will consider a projectile, moving under gravity, experiencing linear drag. (Think
of a projectile moving very slowly in a viscous liquid). At time t = 0, we throw the
object with velocity u. What is its subsequent motion?

The equation of motion is

dv
m = mg − γv (2.28)
dt
We can solve this by introducing the integrating factor eγt/m to write the equation as
d γt/m

e v = eγt/m g
dt
We now integrate, but have to introduce a vector integration constant – let’s call it c
– for our troubles. We have
m
v= g + ce−γt/m
γ
We specified above that at time t = 0, the velocity is v = u, so we can use this
information to determine the integration constant c. We get
 
m m
v = g + u − g e−γt/m
γ γ

Now we integrate v = dx/dt a second time to determine x as a function of time. We

get another integration constant, b,
 
m m m
x = gt − u − g e−γt/m + b
γ γ γ

To determine this second integration constant, we need some further information about
the initial conditions. Lets say that x = 0 at t = 0. Then we have
 
m m m
u − g 1 − e−γt/m


x = gt +
γ γ γ

We can now look at this in components to get a better idea of what’s going on.
We’ll write x = (x, y, z) and we’ll send the projectile off with initial velocity u =

– 36 –
(u cos θ, 0, u sin θ). With gravity acting downwards, so g = (0, 0, −g), our vector equa-
tion becomes three equations. One is trivial: y = 0. The other two are
m
u cos θ 1 − e−γt/m


x=
γ
 
mgt m mg
1 − e−γt/m


z=− + u sin θ +
γ γ γ

Notice that the time scale m/γ is important. For t ≫ m/γ, the horizontal position is
essentially constant. By this time, the particle is dropping more or less vertically.

Finally, we can revisit the question that we asked in the last example: what happens
when friction is small? Again, there are a couple of terms that look as if they are
going to become singular in this limit. But that sounds very unphysical. To resolve
this, we should ask what γ is small relative to. In the present case, the answer lies
in the exponential terms. To say that γ is small, really means γ ≪ m/t or, in other
words, it means that we are looking at short times, t ≪ m/γ. Then we can expand the
exponential. Reverting to the vector form of the equation, we find
   2 !
m m m γt 1 γt
x = gt + u− g 1−1+ − + ...
γ γ γ m 2 m
 
1 2 γt
= ut + gt + O
2 m

So we see that, on small time scales, we indeed recover the usual story of a projectile
without friction. The friction only becomes relevant when t ∼ m/γ.

– 37 –
Chapter 3

One-Dimensional Conservative
Systems

3.1 Description as a Dynamical System

For one-dimensional mechanical systems, Newton’s second law reads

mẍ = F (x) . (3.1)

A system is conservative if the force is derivable from a potential: F = −dU/dx. The total
energy,
E = T + U = 21 mẋ2 + U (x) , (3.2)
is then conserved. This may be verified explicitly:
dE d h1 2
i
= m ẋ + U (x)
dt dt 2
h i
= mẍ + U ′ (x) ẋ = 0 . (3.3)

Conservation of energy allows us to reduce the equation of motion from second order to
first order: v
u !
dx u2
= ±t E − U (x) . (3.4)
dt m

Note that the constant E is a constant of integration. The ± sign above depends on the
direction of motion. Points x(E) which satisfy

E = U (x) ⇒ x(E) = U −1 (E) , (3.5)

where U −1 is the inverse function, are called turning points. When the total energy is E,
the motion of the system is bounded by the turning points, and confined to the region(s)

27
28 CHAPTER 3. ONE-DIMENSIONAL CONSERVATIVE SYSTEMS

U (x) ≤ E. We can integrate eqn. 3.4 to obtain

r Zx
m dx′
t(x) − t(x0 ) = ± p . (3.6)
2 E − U (x′ )
x0

This is to be inverted to obtain the function x(t). Note that there are now two constants
of integration, E and x0 . Since

E = E0 = 12 mv02 + U (x0 ) , (3.7)

we could also consider x0 and v0 as our constants of integration, writing E in terms of x0

and v0 . Thus, there are two independent constants of integration.

For motion confined between two turning points x± (E), the period of the motion is given
by
xZ
+ (E)
√ dx′
T (E) = 2m p . (3.8)
E − U (x′ )
x− (E)

3.1.1 Example : harmonic oscillator

In the case of the harmonic oscillator, we have U (x) = 12 kx2 , hence

r
dt m
=± . (3.9)
dx 2E − kx2
p
The turning points are x ± (E) = ± 2E/k, for E ≥ 0. To solve for the motion, let us
substitute r
2E
x= sin θ . (3.10)
k
We then find r
m
dt = dθ , (3.11)
k
with solution
θ(t) = θ0 + ωt , (3.12)
p
where ω = k/m is the harmonic oscillator frequency. Thus, the complete motion of the
system is given by
r
2E
x(t) = sin(ωt + θ0 ) . (3.13)
k

Note the two constants of integration, E and θ0 .

3.2. ONE-DIMENSIONAL MECHANICS AS A DYNAMICAL SYSTEM 29

3.2 One-Dimensional Mechanics as a Dynamical System

Rather than writing the equation of motion as a single second order ODE, we can instead
write it as two coupled first order ODEs, viz.
dx
=v (3.14)
dt
dv 1
= F (x) . (3.15)
dt m

This may be written in matrix-vector form, as

   
d x v
= 1 . (3.16)
dt v m F (x)

This is an example of a dynamical system, described by the general form

dϕ
= V (ϕ) , (3.17)
dt
where ϕ = (ϕ1 , . . . , ϕN ) is an N -dimensional vector in phase space. For the model of eqn.
3.16, we evidently have N = 2. The object V (ϕ) is called a vector field. It is itself a vector,
existing at every point in phase space, RN . Each of the components of V (ϕ) is a function
(in general) of all the components of ϕ:

Vj = Vj (ϕ1 , . . . , ϕN ) (j = 1, . . . , N ) . (3.18)

Solutions to the equation ϕ̇ = V (ϕ) are called integral curves. Each such integral curve
ϕ(t) is uniquely determined by N constants of integration, which may be taken to be the
initial value ϕ(0). The collection of all integral curves is known as the phase portrait of the
dynamical system.
In plotting the phase portrait of a dynamical system, we need to first solve for its motion,
starting from arbitrary initial conditions. In general this is a difficult problem, which can
only be treated numerically. But for conservative mechanical systems in d = 1, it is a trivial
matter! The reason is that energy conservation completely determines theq phase portraits.
2


The velocity becomes a unique double-valued function of position, v(x) = ± m E − U (x) .
The phase curves are thus curves of constant energy.

3.2.1 Sketching phase curves

To plot the phase curves,

(i) Sketch the potential U (x).

q

2
(ii) Below this plot, sketch v(x; E) = ± m E − U (x) .
30 CHAPTER 3. ONE-DIMENSIONAL CONSERVATIVE SYSTEMS

Figure 3.1: A potential U (x) and the corresponding phase portraits. Separatrices are shown
in red.

(iii) When E lies at a local extremum of U (x), the system is at a fixed point.
(a) For E slightly above Emin , the phase curves are ellipses.
(b) For E slightly below Emax , the phase curves are (locally) hyperbolae.
(c) For E = Emax the phase curve is called a separatrix .
(iv) When E > U (∞) or E > U (−∞), the motion is unbounded.

(v) Draw arrows along the phase curves: to the right for v > 0 and left for v < 0.

The period of the orbit T (E) has a simple geometric interpretation. The area A in phase
space enclosed by a bounded phase curve is

I xZ
+ (E)
q p
A(E) = v dx = 8
m dx′ E − U (x′ ) . (3.19)
E x− (E)

Thus, the period is proportional to the rate of change of A(E) with E:

∂A
T =m . (3.20)
∂E
3.3. FIXED POINTS AND THEIR VICINITY 31

3.3 Fixed Points and their Vicinity

A fixed point (x∗ , v ∗ ) of the dynamics satisfies U ′ (x∗ ) = 0 and v ∗ = 0. Taylor’s theorem
then allows us to expand U (x) in the vicinity of x∗ :

U (x) = U (x∗ ) + U ′ (x∗ ) (x − x∗ ) + 21 U ′′ (x∗ ) (x − x∗ )2 + 16 U ′′′ (x∗ ) (x − x∗ )3 + . . . . (3.21)

Since U ′ (x∗ ) = 0 the linear term in δx = x − x∗ vanishes. If δx is sufficiently small, we can

ignore the cubic, quartic, and higher order terms, leaving us with

U (δx) ≈ U0 + 21 k(δx)2 , (3.22)

where U0 = U (x∗ ) and k = U ′′ (x∗ ) > 0. The solutions to the motion in this potential are:

δv0
U ′′ (x∗ ) > 0 : δx(t) = δx0 cos(ωt) + sin(ωt) (3.23)
ω
δv0
U ′′ (x∗ ) < 0 : δx(t) = δx0 cosh(γt) + sinh(γt) , (3.24)
γ
p p
where ω = k/m for k > 0 and γ = −k/m for k < 0. The energy is

E = U0 + 21 m (δv0 )2 + 12 k (δx0 )2 . (3.25)

For a separatrix, we have E = U0 and U ′′ (x∗ ) < 0. From the equation for the energy, we
obtain δv0 = ±γ δx0 . Let’s take δv0 = −γ δx0 , so that the initial velocity is directed toward
the unstable fixed point (UFP). I.e. the initial velocity is negative if we are to the right of
the UFP (δx0 > 0) and positive if we are to the left of the UFP (δx0 < 0). The motion of
the system is then
δx(t) = δx0 exp(−γt) . (3.26)
The particle gets closer and closer to the unstable fixed point at δx = 0, but it takes an
infinite amount of time to actually get there. Put another way, the time it takes to get from
δx0 to a closer point δx < δx0 is
 
−1 δx0
t = γ ln . (3.27)
δx
This diverges logarithmically as δx → 0. Generically, then, the period of motion along a
separatrix is infinite.

3.3.1 Linearized dynamics in the vicinity of a fixed point

Linearizing in the vicinity of such a fixed point, we write δx = x − x∗ and δv = v − v ∗ ,

obtaining     
d δx 0 1 δx
= 1 ′′ ∗ + ..., (3.28)
dt δv − m U (x ) 0 δv
32 CHAPTER 3. ONE-DIMENSIONAL CONSERVATIVE SYSTEMS

Figure 3.2: Phase curves in the vicinity of centers and saddles.

This is a linear equation, which we can solve completely.

Consider the general linear equation ϕ̇ = A ϕ, where A is a fixed real matrix. Now whenever
we have a problem involving matrices, we should start thinking about eigenvalues and
eigenvectors. Invariably, the eigenvalues and eigenvectors will prove to be useful, if not
essential, in solving the problem. The eigenvalue equation is

A ψ α = λα ψ α . (3.29)

Here ψα is the αth right eigenvector 1 of A. The eigenvalues are roots of the characteristic
equation P (λ) = 0, where P (λ) = det(λ · I − A). Let’s expand ϕ(t) in terms of the right
eigenvectors of A:
X
ϕ(t) = Cα (t) ψα . (3.30)
α

Assuming, for the purposes of this discussion, that A is nondegenerate, and its eigenvectors
span RN , the dynamical system can be written as a set of decoupled first order ODEs for
the coefficients Cα (t):
Ċα = λα Cα , (3.31)

with solutions
Cα (t) = Cα (0) exp(λα t) . (3.32)

If Re (λα ) > 0, Cα (t) flows off to infinity, while if Re (λα ) > 0, Cα (t) flows to zero. If
|λα | = 1, then Cα (t) oscillates with frequency Im (λα ).
1
If A is symmetric, the right and left eigenvectors are the same. If A is not symmetric, the right and left
eigenvectors differ, although the set of corresponding eigenvalues is the same.
3.4. EXAMPLES OF CONSERVATIVE ONE-DIMENSIONAL SYSTEMS 33

For a two-dimensional matrix, it is easy to show – an exercise for the reader – that

P (λ) = λ2 − T λ + D , (3.33)

where T = Tr(A) and D = det(A). The eigenvalues are then

p
λ± = 12 T ± 12 T 2 − 4D . (3.34)

We’ll study the general case in Physics 110B. For now, we focus on our conservative me-
chanical system of eqn. 3.28. The trace and determinant of the above matrix are T = 0 and
1
D=m U ′′ (x∗ ). Thus, there are only two (generic) possibilities: centers, when U ′′ (x∗ ) > 0,
and saddles, when U ′′ (x∗ ) < 0. Examples of each are shown in Fig. 3.1.

3.4 Examples of Conservative One-Dimensional Systems

3.4.1 Harmonic oscillator

Recall again the harmonic oscillator, discussed in lecture 3. The potential energy is U (x) =
1 2
2 kx . The equation of motion is

d2 x dU
m 2
=− = −kx , (3.35)
dt dx
where m is the mass and k the force constant (of a spring). With v = ẋ, this may be written
as the N = 2 system,
      
d x 0 1 x v
= = , (3.36)
dt v −ω 2 0 v −ω 2 x
p
where ω = k/m has the dimensions of frequency (inverse time). The solution is well
known:
v0
x(t) = x0 cos(ωt) + sin(ωt) (3.37)
ω
v(t) = v0 cos(ωt) − ω x0 sin(ωt) . (3.38)

The phase curves are ellipses:

ω0 x2 (t) + ω0−1 v 2 (t) = C , (3.39)

where C is a constant, independent of time. A sketch of the phase curves and of the phase
flow is shown in Fig. 3.3. Note that the x and v axes have different dimensions.
Energy is conserved:
E = 12 mv 2 + 12 kx2 . (3.40)
Therefore we may find the length of the semimajor and semiminor axes by setting v = 0 or
x = 0, which gives r r
2E 2E
xmax = , vmax = . (3.41)
k m
34 CHAPTER 3. ONE-DIMENSIONAL CONSERVATIVE SYSTEMS

Figure 3.3: Phase curves for the harmonic oscillator.

The area of the elliptical phase curves is thus

2πE
A(E) = π xmax vmax = √ . (3.42)
mk
The period of motion is therefore
r
∂A m
T (E) = m = 2π , (3.43)
∂E k
which is independent of E.

3.4.2 Pendulum

Next, consider the simple pendulum, composed of a mass point m affixed to a massless rigid
rod of length ℓ. The potential is U (θ) = −mgℓ cos θ, hence
dU
mℓ2 θ̈ = − = −mgℓ sin θ . (3.44)
dθ
This is equivalent to    
d θ ω
= , (3.45)
dt ω −ω02 sin θ
p
where ω = θ̇ is the angular velocity, and where ω0 = g/ℓ is the natural frequency of small
oscillations.
The conserved energy is
E= 1
2 mℓ2 θ̇ 2 + U (θ) . (3.46)
Assuming the pendulum is released from rest at θ = θ0 ,
2E
= θ̇ 2 − 2ω02 cos θ = −2ω02 cos θ0 . (3.47)
mℓ2
3.4. EXAMPLES OF CONSERVATIVE ONE-DIMENSIONAL SYSTEMS 35

Figure 3.4: Phase curves for the simple pendulum. The separatrix divides phase space into
regions of rotation and libration.

The period for motion of amplitude θ0 is then

√ Zθ0

8 dθ 4

T θ0 = √ = K sin2 12 θ0 , (3.48)
ω0 cos θ − cos θ0 ω0
0

where K(z) is the complete elliptic integral of the first kind. Expanding K(z), we have
 

2π 1 2 1

9 4 1


T θ0 = 1 + 4 sin 2 θ0 + 64 sin 2 θ0 + . . . . (3.49)
ω0

For θ0 → 0, the period approaches the usual result 2π/ω0 , valid for the linearized equation
θ̈ = −ω02 θ. As θ0 → π2 , the period diverges logarithmically.
The phase curves for the pendulum are shown in Fig. 3.4. The small oscillations of the
pendulum are essentially the same as those of a harmonic oscillator. Indeed, within the
small angle approximation, sin θ ≈ θ, and the pendulum equations of motion are exactly
those of the harmonic oscillator. These oscillations are called librations. They involve
a back-and-forth motion in real space, and the phase space motion is contractable to a
point, in the topological sense. However, if the initial angular velocity is large enough, a
qualitatively different kind of motion is observed, whose phase curves are rotations. In this
case, the pendulum bob keeps swinging around in the same direction, because, as we’ll see
in a later lecture, the total energy is sufficiently large. The phase curve which separates
these two topologically distinct motions is called a separatrix .