INTERMEDIATE

DYNAMICS
Second Edition

Patrick Hamill
Contents

Part 1. Principles of Mechanics 7

Chapter 1. A Brief Review of Introductory Concepts 9

1.1. Kinematics 9
1.2. Newton’s Second Law 11
1.3. Work and Energy 13
1.4. Momentum 14
1.5. Rotational Motion 15
1.6. Summary 22
1.7. Problems 23

Chapter 2. Kinematics 29
2.1. Galileo Galilei (Optional Historical Note) 30
2.2. The Principle of Inertia 31
2.3. Basic Concepts in Kinematics 32
2.4. The Position of a Particle on a Plane 43
2.5. Unit Vectors 44
2.6. Kinematics in Two Dimensions 47
2.7. Kinematics in Three Dimensions 51
2.8. Summary 60
2.9. Problems 61

Chapter 3. Newton’s Laws: Determining the Motion 69

3.1. Isaac Newton(Optional Historical Note) 70
3.2. The Law of Inertia 71
3.3. Newton’s Second Law and the Equation of Motion 74
3.4. Newton’s Third Law: Action Equals Reaction 78
3.5. Is Rotation Absolute or Relative? 81
3.6. Determining the Motion 83
3.7. Numerical Method to Determine the Motion (Optional) 94
iii
iv CONTENTS

3.8. Summary 98
3.9. Problems 99

Chapter 4. The Lagrangian Method 109

4.1. The Equation of Motion by Inspection 109
4.2. The Lagrangian 111
4.3. Lagrange’s Equations 117
4.4. Degrees of Freedom 121
4.5. Generalized Momentum 123
4.6. Generalized Force (Optional) 126
4.7. The Calculus of Variations (Optional) 128
4.8. Hamilton’s Equations (Optional) 134
4.9. Summary 138
4.10. Problems 140

Chapter 5. The Conservation of Energy 149

5.1. The Work-Energy Theorem 149
5.2. Work Along a Path. The Line Integral 151
5.3. Potential Energy 155
5.4. Force, Work, and Potential Energy 167
5.5. The Conservation of Energy 172
5.6. Energy Diagrams 176
5.7. Solving for the Motion: The Energy Integral 178
5.8. The Kinetic Energy of a System of Particles 181
5.9. Work on an Extended Body. Pseudowork 183
5.10. Summary 184
5.11. Problems 187

Chapter 6. Conservation of Linear Momentum 195

6.1. The Law of Conservation of Momentum 195
6.2. The Motion of a Rocket 196
6.3. Collisions 200
6.4. Inelastic Collisions. The Coefficient of Restitution 210
6.5. Impulse 211
6.6. Momentum of a System of Particles 212
6.7. Relative Motion and the Reduced Mass 214
6.8. Collisions in Center of Mass Coordinates (Optional) 216
6.9. Summary 221
6.10. Problems 223

Chapter 7. Conservation of Angular Momentum 233

7.1. Definition of Angular Momentum 233
7.2. Conservation of Angular Momentum 235
 CONTENTS v

7.3. Angular Momentum of a System of Particles 237

7.4. Rotation of a Rigid Body about a Fixed Axis 244
7.5. The Moment of Inertia 247
7.6. The Gyroscope 250
7.7. Angular Momentum is an Axial Vector 253
7.8. Summary 255
7.9. Problems 257
Chapter 8. Conservation Laws and Symmetries 265
8.1. Symmetry 265
8.2. Symmetry and the Laws of Physics 267
8.3. Symmetries and Conserved Physical Quantities 268
8.4. Are the Laws of Physics Symmetrical? 270
8.5. Strangeness 272
8.6. Symmetry Breaking 274
8.7. Summary 274
8.8. Problems 274

Part 2. The Gravitational Field 277

Chapter 9. The Gravitational Field 279
9.1. Newton’s Law of Universal Gravitation 280
9.2. The Gravitational Field 282
9.3. The Gravitational Field of an Extended Body 286
9.4. The Gravitational Potential 289
9.5. Field Lines and Equipotential Surfaces 293
9.6. The Newtonian Gravitational Field Equations 294
9.7. The Equations of Poisson and Laplace 298
9.8. Einstein’s Theory of Gravitation (Optional) 299
9.9. Summary 306
9.10. Problems 308

Part 3. The Mechanics of Particles 315

Chapter 10. Central Force Motion: The Kepler Problem 317
10.1. Johannes Kepler (Optional Historical Note) 318
10.2. Kepler’s Laws 321
10.3. Central Forces 321
10.4. The Equation of Motion 328
10.5. Energy and the Effective Potential Energy 331
10.6. Solving the Radial Equation of Motion 336
10.7. The Equation of the Orbit 337
10.8. The Equation of an Ellipse 342
vi CONTENTS

10.9. Kepler’s Laws Revisited 348

10.10. A Perturbed Circular Orbit 355
10.11. Resonances 361
10.12. Summary 362
10.13. Problems 363
Chapter 11. Harmonic Motion 373
11.1. Springs and Pendulums 373
11.2. Solving the Differential Equation (Optional) 376
11.3. The Damped Harmonic Oscillator 383
11.4. The Forced Harmonic Oscillator 389
11.5. Coupled Oscillators 402
11.6. Summary 409
11.7. Problems 410
Chapter 12. The Pendulum 417
12.1. A Simple Pendulum with Arbitrary Amplitude 418
12.2. The Physical Pendulum 424
12.3. The Center of Percussion 429
12.4. The Spherical Pendulum 434
12.5. Summary 445
12.6. Problems 447
Chapter 13. Accelerated Reference Frames 453
13.1. A Linearly Accelerating Reference Frame 453
13.2. A Rotating Coordinate Frame 455
13.3. Fictitious Forces 457
13.4. Centrifugal Force and the Plumb Bob 460
13.5. The Coriolis Force 462
13.6. The Foucault Pendulum 469
13.7. Summary 476
13.8. Problems 476

Part 4. The Mechanics of Extended Bodies 481

Chapter 14. Statics (Optional) 483
14.1. Basic Concepts 483
14.2. Couples, Resultants and Equilibrants 487

14.3. Reduction to the Simplest Set of Forces 489

14.4. The Hanging Cable 489
14.5. Stress and Strain 494
14.6. The Centroid (Optional) 497
14.7. The Center of Gravity (Optional) 499
 CONTENTS vii

14.8. Equilibrium of Fluids 501

14.9. D’Alembert’s Principle and Virtual Work (Optional) 507
14.10. Summary 511
14.11. Problems 513

Chapter 15. Rotational Kinematics 519

15.1. Orientation of a Rigid Body 520
15.2. Orthogonal Transformations 522
15.3. The Euler Angles 530
15.4. Euler’s Theorem 534
15.5. Infinitesimal Rotations 545
15.6. Summary 547
15.7. Problems 549

Chapter 16. Rotational Dynamics 551

16.1. Angular Momentum 552
16.2. Kinetic Energy 556
16.3. Properties of the Inertia Tensor 558
16.4. The Euler Equations of Motion 571
16.5. Torque-Free Motion 572
16.6. The Spinning Top 574
16.7. Summary 582
16.8. Problems 585

Chapter 17. Waves 589

17.1. A Wave in a Stretched String 590
17.2. Direct Solution of the Wave Equation 593
17.3. Standing Waves 595
17.4. Traveling Waves 597
17.5. Standing Waves as a Special Case of Traveling Waves 600
17.6. Energy 603
17.7. Momentum 607
17.8. Summary 609
17.9. Problems 610

Chapter 18. Small Oscillations (Optional) 615

18.1. Introduction 615
18.2. Statement of the Problem 615
18.3. Normal Modes 621
18.4. Matrix Formulation 629
18.5. Normal Coordinates 632
18.6. Coupled Pendulums: An Example 634
18.7. Many Degrees of Freedom 638
viii CONTENTS

18.8. Transition to Continuous Systems 643

18.9. Summary 647
18.10. Problems 650

Part 5. Special Topics 653

Chapter 19. The Special Theory of Relativity 655
19.1. Albert Einstein (Optional Historical Note) 656
19.2. Experimental Background 657
19.3. The Postulates of Special Relativity 659
19.4. The Lorentz Transformations 660
19.5. The Addition of Velocities 668
19.6. Simultaneity and Causality 670
19.7. The Twin Paradox 673
19.8. Minkowski Space-Time Diagrams 675
19.9. 4-Vectors 679
19.10. Relativistic Dynamics 685
19.11. Summary 688
19.12. Problems 689
Chapter 20. Classical Chaos (Optional) 695
20.1. Configuration Space and Phase Space 696
20.2. Periodic Motion 698
20.3. Attractors 700
20.4. Chaotic Trajectories and Liapunov Exponents 702
20.5. Poincaré Maps 703
20.6. The Henon Heiles Hamiltonian 705
20.7. Summary 707
20.8. Problems 708
Chapter A. Formulas and Constants 713
Chapter B. Answers to Selected Problems 717
Preface

Although this book begins at an introductory level, by the end of

the book the student will have been exposed to all of the subject matter
usually found in an intermediate mechanics course as well as a few less
common advanced topics.
Organization
This book is divided into five parts, preceded by a short chapter
consisting of a review of a few essential introductory concepts. This
chapter can be skipped by well prepared students, or assigned as read-
ing for students who only need a quick refresher.
Part I (Chapters 2 through 8) is called “Principles of Mechanics.”
This part covers all the basic concepts in mechanics. The topics of kine-
matics, dynamics and the conservation principles are treated in depth.
The first chapter in this part is called “Kinematics.” This is the tradi-
tional starting point for courses in intermediate mechanics. Here, the
student is exposed to relations between acceleration, velocity and po-
sition in Cartesian, plane polar, cylindrical, and spherical coordinates.
A few simple concepts from vector analysis are introduced. A number
of reasonably difficult projectile problems are included in the problems.
The next chapter considers Newton’s laws. It includes a discussion on
“Determining the Motion” in which the student learns techniques for
integrating Newton’s second law to obtain the position as a function of
time. This is done for constant forces, and for forces that are functions
of time, of velocity, and of position. A short section called “Solving for
the Motion by Numerical Methods” gives a flavor for the use of compu-
tational techniques in physics. The role of computers in physics is not
emphasized in this course. However, I realize that many instructors
want to expose their students to computational methods, so I have in-
cluded a few discussions of numerical techniques. Furthermore, nearly

1
2 PREFACE

every chapter has a number of “Computational Projects.” However,

this text does not emphasize the role of computers in physics because
I find that teaching the traditional material of intermediate mechanics
takes most of two semesters and does not give enough time to delve
into computational physics. Furthermore, many universities have in-
cluded a computational physics course in the undergraduate physics
curriculum.
Chapter 4 is called “The Lagrangian Approach.” I think it is im-
portant for physics students to be exposed to the Lagrangian early on
in mechanics. The Lagrangian is presented, at first, as a simple tech-
nique for generating the equation of motion. Later in the chapter, I go
through a derivation of the Lagrange equations using the Calculus of
Variations. This section need not be covered if the instructor feels it
is too advanced. The chapter ends with an optional discussion of the
Hamiltonian and Hamilton’s equations. There are several reasons why
I chose to present the Lagrangian early in the course, perhaps the most
important is that it gives the student a simple (almost “cookbook”)
technique for obtaining the equations of motion for a complicated dy-
namical system. For example, the Lagrangian technique allows one to
easily determine the equations of motion for a double pendulum, for
a spherical pendulum, or for coupled oscillators. More importantly,
it allows one to introduce the concepts of generalized momentum and
ignorable coordinates and leads to the relation between conservation
laws and symmetries. Furthermores, it lets the student know that
this course is not simply a rehash of concepts learned in introductory
physics.
The following four chapters cover the conservations of energy, linear
momentum, and angular momentum, followed by a short chapter on
the relation between symmetry and conservation laws. The chapter on
the conservation of energy discusses potential energy and the concept
of effective potential. Potential energy naturally leads to a discussion
of the gradient of a scalar field. There is a section on the way the “Del”
operator can be expressed in cylindrical and spherical coordinates; this
allows one to discuss coordinate transformations in general. (I be-
lieve that introducing concepts from vector calculus as required by the
physics is much more effective than stuffing all of the vector concepts
into a single introductory chapter.) The chapters on the conservation
of linear and angular momentum cover the usual topics (rockets, col-
lisions, etc.) as well as some less usual topics such as the fact that
angular momentum is an axial vector.
Part II is a very short introduction to field theory. Since field theory
is treated exhaustively in courses on electromagnetism, this study is
 PREFACE 3

limited to a single chapter on the gravitational field. Additional vector

concepts are introduced here and the student is exposed to Gauss’s law
and the equations of Poisson and Laplace.
Next is Part III (Chapters 10 - 13) called “The Mechanics of Par-
ticles.” This important section covers central force motion, harmonic
motion and motion in accelerated reference frames. The student is ex-
posed to the techniques of expansions in power series and the idea of
using successive approximations to solve otherwise intractable physics
problems. Chapter 10 deals with central force motion, as illustrated by
the Kepler problem. Also considered is the stability of circular orbits
which shows the student how to deal with small perturbations. Specif-
ically, we imagine a comet striking a planet in a previously perfectly
circular orbit and analyze the subsequent motion to determine stability
and the frequency of radial oscillations.
The next chapter is called “Simple Harmonic Motion.” Damped and
driven harmonic oscillators are treated in depth. A rather thorough dis-
cussion on how to solve second order differential equations is included
here. Coupled oscillators and normal modes are touched upon.
The third chapter in this part (Chapter 12) is on the motion of a
pendulum. We begin with the motion of a simple pendulum of arbi-
trary amplitude and introduce elliptic integrals. Next we consider the
physical pendulum, centers of oscillation and percussion, the spherical
pendulum, and the conical pendulum. To spend a whole chapter on
the pendulum may seem excessive, but it is a simple, easily visualized
physical system that allows one to introduce many useful mathematical
techniques without having to spend time explaining the motion.
The fourth chapter in this part is called “Accelerated Reference
Frames” in which we (mainly) consider motion on the surface of the
rotating Earth. Coriolis forces and the Foucault pendulum are treated.
Once again, perturbation theory is used to solve these problems.
Part IV (Chapters 14-18) is called “The Mechanics of Extended
Bodies.” Portions of the material in this part may be too advanced
for some classes, but the instructor should try to cover the chapters
on rotational kinematics, rotational dynamics, and waves. This part
begins with an optional chapter on statics that includes a discussion
of d’Alembert’s principle and virtual work. The next chapter (on rota-
tional kinematics) introduces orthogonal transformations. We obtain
the Euler angles and consider Euler’s theorem. The following chapter
(Chapter 16) on rotational dynamics introduces the inertia tensor and
some simple methods from tensor analysis. The next chapter in this
part is an introduction to wave motion. This topic is not considered
4 PREFACE

in great detail since it is treated extensively in the undergraduate elec-

tromagnetic theory class, nevertheless it does show the student how to
solve a partial differential equation by separation of variables. The last
chapter in this part treats small oscillations. It is rather advanced and
the chapter is denoted as optional.
The last part of the book is called “Special Topics” and consists
of two chapters. The first of these is on special relativity. The second
chapter is an introduction to classical chaos. Neither of these chapters
is at all exhaustive; they are simply intended to give the student a
flavor of these interesting subjects.
A One-Semester Course
Many instructors will find that the intermediate mechanics course
in their department has been reduced to one semester. If such is the
situation at your institution, you may wish to cover the material in
Chapters 2, 4, 6, 7, 11, 13, 15, 16, 17 and include some material from
the skipped chapters, in particular, Sections 3.6, 5.6, and 12.2
Exercises and Problems
Learning physics requires doing physics, so I have included a large
number of “exercises.” These are found at the end of nearly every
section. Most of them are fairly easy. Some are merely “plug-ins” to
get the student to look at a formula and (hopefully) to think about
it. Others ask the student to fill in the missing steps in a derivation.
A few require a bit of clever thinking. Nearly all have answers given.
I hope that students studying this book will solve every one of these
exercises. At the end of each chapter is a collection of problems that
are of the degree of difficulty to be expected from a course at this level.
Most of these will require significant effort on the part of the student.
However, I believe that a student who has read the chapter and worked
the exercises will be prepared to attack the problems.
 PREFACE 5

Acknowledgements
I thank my colleagues at San Jose State University and NASA Ames
Research Center, particularly Drs. Alejandro Garcia and Michael Kauf-
man. I am especially indebted to the many students in my mechanics
courses whose influence on this book cannot be overestimated.
 Part 1

Principles of Mechanics
Chapter 1
A Brief Review of
Introductory Concepts

This introductory chapter is intended to remind you of some of

the basic concepts from your introductory physics course. You should
probably read through this chapter quickly, work out some of the ex-
ercises and try a few of the problems at the end of the chapter. If you
remember all of this material, go on to Chapter 2, but if you find that
you have forgotten some of these basic concepts, you should go back
to your introductory physics textbook and review the material there.
Most of the standard introductory physics texts are well written and
contain many instructive figures and diagrams. It is a good idea to go
back to that text whenever you are exposed to the same material on a
more advanced level.
As you may recall, the mechanics section of your introductory
physics book covered the following topics:
• Kinematics
• Newton’s Second Law
• Work and Energy
• Momentum
• Rotational Motion
We now very briefly review some concepts from each of these items.

1.1. Kinematics
Kinematics is defined as the study of motion. Essentially, kine-
matics involves determing the relationships between position, velocity,
acceleration, and time.
9
10 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

Position is denoted by the vector r and the change in position (or

displacement) can be written as ∆r. Velocity is defined as the displace-
ment with respect to time, so the average velocity is given by
∆r
< v >=
∆t
where ∆t is the time during which the object moved a distance ∆r.
As the time interval becomes very small, we replace the difference
(represented by ∆) with the derivative and write
dr
v= . (1.1)
dt
The change in velocity with respect to time is called the acceleration
and is given by
dv
a= . (1.2)
dt
Clearly, we can use the definitions of acceleration and velocity to
write the inverse relations:
Z Z
dv = adt,

and Z Z
dr = vdt.

1.1.1. Motion in a Straight Line at Constant Acceleration.

For motion in a straight line, we do not need to use vector notation.
Letting the position to be represented by x, if the acceleration is con-
stant the integrals above are easy to evaluate, leading to the familiar
relations
v(t) = at + v0 , (1.3)
and
1
x(t) = at2 + v0 t + x0 , (1.4)
2
where v0 and x0 are the initial velocity and the initial position.
You probably memorized Equations (1.3) and (1.4) in your intro-
ductory physics course. You should, however, keep in mind that they
are only valid if the acceleration is constant. In this course, we shall
frequently be concerned with non-constant accelerations and these re-
lations cannot be used.
Two other useful equations can be obtained from (1.3) and (1.4)
as follows. Solve one equation for t and substitute it in the other to
obtain
2a(x − x0 ) = v 2 − v02 ,
 1.2. NEWTON’S SECOND LAW 11

or solve one equation for a and substitute it in the other equation to

obtain
1
x − x0 = (v + v0 )t.
2

Comment 1. In this book you will find a large number of

“exercises.” They are not difficult. They are intended to
give you a chance to review and understand the concepts
in the preceding section. It would be a good idea to work
out the solution for each exercise as you read through
the text. At the end of each chapter you will find a set
of problems that are significantly more difficult than the
exercises. However, if you solve the exercises, you will be
well prepared for solving the problems. Many students
find working the exercises to be a good preparation for
the examinations.

Exercise 1.1. You were driving your new Ferrari at 62 mph (= 100
km/hr) when you spotted a police car. Naturally, you hit the brakes.
You slowed to 31 mph, covering a distance of 50 m. (a) What is your
constant acceleration? (b) How much time did it take to slow to 31
mph? Answers: (a) -5.79 m/s2 (b) 2.4 sec.

1.2. Newton’s Second Law

The study of the relation between the forces acting on a body and
the motion of the body is called dynamics. In your introductory course
you were exposed to dynamics in the form of Newton’s Second Law.
This law states that a body of mass m that is acted upon by a force F
will accelerate at
a = F/m.
This relationship is usually remembered as
F = ma. (1.5)
As you shall see, this form of Newton’s second law is only valid if the
mass is constant.
12 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

Figure 1.1. A block sliding down an inclined plane

with friction. Sketch (a) shows the forces acting on the
block. Sketch (b) shows the free-body diagram. Sketch
(c) shows the x-and y-axes inclined so that they are par-
allel and perpendicular to the plane.

Worked Example 1.1. Determine the acceleration of a block

of mass m sliding down an inclined plane of angle θ. Assume the
coefficient of sliding friction is µ. See Figure 1.1.
Solution: The forces acting on the block are gravity mg (down-
wards), the normal force N (perpendicular to the plane), and the
frictional force µN (parallel to the plane). These forces are illus-
trated in Figure 1.1(a). The free-body diagram with all of the
forces acting at the center of mass of the block is shown in Figure
1.1(b). In a problem such as this one, it is convenient to assume the
axes are parallel to and perpendicular to the surface of the plane
as indicated in Figure 1.1(c).
There is no acceleration perpendicular to the plane so the net
force in that direction must be zero. Consequently, N = mg cos θ.
The net force down the plane is Fd = mg sin θ − µN . The acceler-
ation of the block is
Fd
a= = g sin θ − µg cos θ.
m

Exercise 1.2. Two blocks of mass M1 and M2 are tied together.

They are sitting on a smooth frictionless surface as shown in Figure
1.2. A force F is applied to the free string attached to M1 . What
is the tension in the string between the two blocks? Answer: T =
M2 F/(M1 + M2 ).

Exercise 1.3. A block of mass 25 kg is held in place on an inclined

plane of angle 30◦ as shown in Figure 1.3. The coefficient of static
 1.3. WORK AND ENERGY 13

Figure 1.2. Two blocks on a smooth frictionless surface

connected by a massless string.

friction is 0.4. a) Draw the free body diagram. What forces act on the
block? b) What is the tension in the string? c) If the string is cut,
what is the acceleration of the block? Answers (b) T = 37.63 N (c)
a = 1.51 m/s2 .

30o

Figure 1.3. A block on an inclined plane. What is the

acceleration of the block if the string is cut?

1.3. Work and Energy

Imagine pushing a box along a horizontal surface, such as a table-
top. The force you are applying can be denoted by F. As you might
expect, there will be opposing forces such as friction, air resistance,
etc., but for now we are only interested in the force you are exerting.
If you push the box from x1 to x2 , the work you do on it is
Z x2
W = F · dx. (1.6)
x1
When we do work on an object, we often change its energy. The
energy of an object is either kinetic energy (energy of motion) or po-
tential energy (energy of position). For a particle of mass m moving
with speed v the kinetic energy (denoted by T ) is given by
1
T = mv 2 . (1.7)
2
14 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

The potential energy of an object of mass m raised a height h above

the surface of the Earth is given by
V = mgh. (1.8)
We shall often be interested in the potential energy of a stretched
or compressed spring. The potential energy of such a system is given
by
1
V = kx2 , (1.9)
2
where k is the spring constant and x represents the amount the spring
is stretched or compressed.
Power is defined as the work done per unit time. Suppose a force F
acts on a particle for an infinitesimal time interval, from t to t+dt. The
work done during this interval is dW = F · dx. Therefore,the power is

dW F · dx
P = = .
dt dt
Since dx/dt ≡ v, another expression for power is P = F · v.

Exercise 1.4. A rock is thrown upward from the top of a 30 m

building with a velocity of 5 m/s. Determine its velocity (a) When it
falls back past its original point, (b) When it is 15 m above the street,
and (c) Just before it hits the street. Answer: (a) -5 m/s, (c)24.76 m/s.

Exercise 1.5. A horse drags a 100 kg sled a distance of 4 km in

20 minutes. The horse exerts one horsepower, of course. What is the
coefficient of sliding friction between the sled and the ground? Answer:
µk = 0.23.

1.4. Momentum
A moving particle is characterized by having a particular momen-
tum. When we use the term “momentum” we usually are referring to
the linear momentum, not to be confused with the angular momentum,
which we will define in a few moments.
 1.5. ROTATIONAL MOTION 15

The momentum is simply defined as mass times velocity and is

denoted by the letter p. Thus,
p = mv. (1.10)
If the mass of a body is constant, the time derivative of the mo-
mentum is
dp d(mv) dv
= =m = ma
dt dt dt
But according to Newton’s second law, F = ma, so Newton’s second
law can be expressed as
dp
F= (1.11)
dt
Here F is the net or total vector sum of all forces acting on the body.
Consequently, we appreciate that if the net force is zero, the time de-
rivative of the momentum is zero. That is, the momentum of the body
is constant. This is called the law of conservation of momentum.

Exercise 1.6. A 1500 kg car traveling east at 40 km/hr turns a

corner and speeds up to a velocity of 50 km/hr due north. What is the
change in the car’s momentum? Answer: 26, 700 kg m/s at 38.7◦ West
of North.

1.5. Rotational Motion

The motion of a body rotating about a fixed axis is mathematically
identical to one dimensional linear motion. Recall that kinematics
studies the relationship between position, velocity, and acceleration.
Rotational kinematics deals with angular position (θ), angular velocity
(ω), and angular acceleration (α), where
dθ
ω≡ ,
dt
and
dω d2 θ
α≡ = 2.
dt dt
You will discover that rotational motion can be very complicated.
To keep things simple for the moment, consider the special case of
a symmetrical body rotating about a fixed axis, such as the wheel
illustrated on the left side of Figure 1.4.
16 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

Figure 1.4. A wheel mounted on a fixed axis. The

wheel is allowed to rotate but not to translate.

For a fixed, stationary axis, the center of the wheel is at rest. All
other points are moving in circles around it. If you looked straight down
the axis you would see the circle shown on the right side of Figure 1.4.
Point P is on the rim of the wheel. The angular position of P is given
by the angle between some fixed line and the radius vector to P. If the
wheel is turning, after a time dt point P will have moved a distance
ds to P 0 . Recall from geometry that ds = rdθ where dθ (in radians of
course!) is the angle subtended by the arc P P 0 = ds. Point P moves
with speed
ds rdθ
v= = = rω.
dt dt
The speed of point P is called the “tangential speed” because instan-
taneously P is moving tangent to the rim of the wheel. It is sometimes
convenient to write the tangential speed as vT . Then
vT = rω. (1.12)
Taking the time derivative of vT yields the tangential acceleration
aT ,
d2 θ
 
dv d dθ
aT = = r = r 2,
dt dt dt dt
dω
where we used the fact that r is constant. But dt
= α, so
aT = rα. (1.13)

Exercise 1.7. A wheel initially spinning at ω0 = 50.0 rad/s comes

to a halt in 20.0 seconds. Determine the constant angular acceleration
and the number of revolutions it makes before stopping. Answers: -2.5
rad/s2 ; 79.6 rev.
 1.5. ROTATIONAL MOTION 17

1.5.1. Rotational Dynamics. Rotational dynamics is the anal-

ysis of the rotational motion of a body subjected to external torques.
Consider a body constrained to rotate about a fixed axis as shown
in Figure 1.5. I drew the body in the shape of a plane lamina for
simplicity. Let a force F act on the body at a point on its rim. Let
us assume (again for simplicity) that F is perpendicular to the axis
of rotation. The point of application of F is specified by the vector r
whose origin is at the axis of rotation.
The body cannot accelerate linearly because the axis is fixed. The
applied force causes the body to rotate about the axis. The tendency
of a force to cause a rotation is called the moment of the force, or more
commonly, the torque. Just as you can think of a force as a pull, you
can think of a torque as a twist.
The ability of a force to produce a rotation depends not only on the
magnitude of the force, but also on its direction and on the location of
the point where the force is applied to the body.
To define torque draw a line having the direction of the force and
passing through the point of application. This is called the “line of
action of the force.” (See Figure 1.6.) Next draw a line that starts at
the axis of rotation and intersects the line of action at a 90◦ angle.
This line is called the lever arm. The lever arm is the shortest distance
from the axis of rotation to the line of action of the force. Its length
is r sin θ where θ is the angle between r and F. The magnitude of the

Axis of
rotation

r
F

Figure 1.5. Illustration of torque. The laminar body

is free to rotate around the fixed axis of rotation. The
vector r, with origin at the axis, specifies the point of
application of the force.
18 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

Figure 1.6. Definition of lever arm. The axis of rota-

tion is perpendicular to the plane of the figure.

torque is given by the relation:

Torque = Force times Lever Arm.
This elementary definition of torque is complicated and cumber-
some. Fortunately, vector concepts allow us to simplify the definition.
Torque (N) is the cross product of r and F, thus:

N ≡ r × F. (1.14)
In Equation (1.14) r is the vector from the axis of rotation to the
point of application of the force. The direction of the torque is perpen-
dicular to the plane defined by r and F and it is usually represented
as a vector along the axis of rotation.

Exercise 1.8. Show that a given force applied at any point along
the line of action yields the same torque.
Exercise 1.9. An athlete is holding a 2.5 meter pole by one end.
The pole makes an angle of 60◦ with the horizontal. The mass of the
pole is 4 kg. Determine the torque exerted by the pole on the athlete’s
hand. (The mass of the pole can be assumed to be concentrated at the
center of mass.) Answer: Torque = 24.5 N m.
 1.5. ROTATIONAL MOTION 19

1.5.2. Statics. This section is an overview of some of the basic

concepts of statics.
Newton’s first law tells us that if no external force acts on a body,
it will move at constant velocity. That is, zero force implies zero ac-
celeration. (This is also a consequence of the Newton’s second law.)
Consequently, a body that is not accelerating must have no net exter-
nal force acting on it. Similarly, a body that is rotating at a constant
angular velocity has no angular acceleration. Such a body has no net
external torque acting on it.
A body that has no linear acceleration and no angular acceleration
is said to be in equilibrium. The conditions for equilibrium are:
X
Fi = 0, (1.15)
and
X
Ni = 0, (1.16)
where Fi and Ni are the net external forces and torques acting on the
system.
A body in equilibrium has constant linear velocity and constant
angular velocity. A particular but important special case occurs when
both the linear and angular velocity are zero. This is called static
equilibrium.
The general conditions for equilibrium are given by Equations (1.15)
and (1.16). I will state these conditions in words for a situation in which
all of the forces applied to a body lie in the same plane (but it is not
difficult to generalize). If a body is in equilibrium then:
1. The algebraic sum of the components of the forces in each of two
mutually perpendicular directions is zero, and,
2. The algebraic sum of the torques about any point in the plane
of the forces is zero.
The first condition is obvious. The second condition will be proved
in Chapter 7. It is very useful for solving simple statics problems
because it allows us to take the torques about any convenient point.
This point does not even have to be in the body.

Worked Example 1.2. The captain of the swimming team

(who weighs 600 N) is posing on the end of a 4 m diving board
that weighs 150 N. The board is bolted to two pillars that are 1.5
m apart. The first pillar is at the end of the board. Determine the
tension (or compression) acting on the bolts. See Figure 1.7.
20 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

Figure 1.7. What forces are exerted on the bolts that

hold the diving board on the pillars?

Solution: The board is in equilibrium so

forces up = forces down
and
torques clockwise (CW) = torques counterclockwise (CCW).
The first condition leads to
600 + 150 = F1 + F2 .
The second condition leads to
600 × 4 + 150 × 2 = F2 × 1.5.
Therefore, F2 = 1800 N (up) and F1 = –1050N (down).

1.5.3. Rotational Kinetic Energy. The rotational kinetic en-

ergy of an extended rigid body is obtained (at least conceptually) by
summing over the kinetic energy of each particle in the body. If the
body is rotating about a fixed axis and if the body is symmetrical
about this axis, the rotational kinetic energy is given by a fairly simple
formula. Recall that every particle in the body is moving in a circle,
and that the linear velocity v of the particle is related to the angular
velocity of the body, ω, according to v = ωr⊥ where r⊥ is the per-
pendicular distance from the axis of rotation to the particle. So the
kinetic energy of the particle is T = 21 mv 2 = 12 m(ωr⊥ )2 . The total
energy of the body will be the sum over all of the particles (molecules)
in the body. If we denote the mass of the ith particle by mi and its
 1.5. ROTATIONAL MOTION 21

perpendicular distance from the axis by ri⊥ , we have

X1
2
T = mi ri⊥ ω2
i
2
2
P
The quantity i mi ri⊥ is called the moment of inertia of the body and
is denoted I. Therefore the rotational kinetic energy can be expressed
as

1
T = Iω 2 . (1.17)
2
Note the similarity with the translational kinetic energy T = 21 mv 2 .

Exercise 1.10. A disk of mass M and radius R is initially at rest. It

is acted upon by a constant torque N for a time T. Determine the final
angular velocity of the disk and the work required to spin it up to this
final state. If the energy was supplied by a motor, what was the average
power output of the motor? Answer: Average Power = N 2 T /M R2 .

1.5.4. Angular Momentum. In general, the equations for linear

motion go over to the equations for rotational motion if we substitute
I for m and ω for v. Thus, for example, the magnitude of the angular
momentum of a single particle of mass m moving in a circle of radius
r with speed v will be

l = Iω, (1.18)
2
where I = mr .
This relation can also be expressed as l = mr2 ω or as l = mvr.
But angular momentum is a vector, so
l = Iω
where ω is the angular velocity vector, directed along the axis of rota-
tion.
For an extended rigid body the total angular momentum is usually
denoted by L and, as you might expect, is given by
L = Iω
2
P
where now I = i mi ri⊥ .
22 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

The rotational analogue of Newton’s second law is obtained by sub-

stituting torque for force, moment of inertia for mass and angular ac-
celeration for linear acceleration. This yields
N = Iα. (1.19)
Recall that Newton’s second law can be expressed as
dp
F= .
dt
Similarly, for rotational motion we can write
dL
N= .
dt
Just as linear momentum is constant if the net force is zero, so too,
angular momentum is constant if the net torque is zero.

Exercise 1.11. A spinning ice skater speeds up by pulling her

hands to the side of her body. Approximate the ice skater by a cylinder
of radius 10 cm and mass 30 kg. Her hands are two point masses of
0.25 kg each and her (massless) arms extend 90 cm from her torso (or
1 m from the axis of rotation). If she was initially rotating at 2 rev/s,
what is her final angular velocity? Answer: 8.4 rev/s.

1.6. Summary
The position of a particle is given by the vector r. Velocity is
defined by
dr
v= ,
dt
and the acceleration is
dv
a= .
dt
Newton’s second law for a particle with constant mass is
F = ma. (1.20)
Work is defined by the integral of the dot product of force and
displacement, Z
W = F · dr. (1.21)
 1.7. PROBLEMS 23

Kinetic energy is given by

1
T = mv 2 ,
2
for linear motion, and
1
T = Iω 2 ,
2
for rotational motion.
Potential energy of an object raised a height h above the Earth’s
surface is
V = mgh.
The potential energy stored in a spring is
1
V = kx2 .
2
The linear momentum of a particle of mass m moving at velocity v
is defined as
p = mv. (1.22)
The relation between angular velocity and linear velocity is given
by the cross product:
v = ω × r, (1.23)
where ω is a vector pointing along the axis of rotation whose magnitude
is equal to the angular velocity.
Torque is defined by
N = r × F. (1.24)
The angular momentum of a particle can be written as
l = r × p. (1.25)
The rotational version of Newton’s Second Law is
dl
N= . (1.26)
dt

1.7. Problems
Problem 1.1. Two ships are sailing in a thick fog. Initially, ship
A is 10 miles north of ship B. Ship A sails directly east at 30 miles
per hour. Ship B sails due east at constant speed vB then turns and
sails due north at the same speed. After two hours, the ships collide.
Determine vB .
24 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

Problem 1.2. You carefully observe an object moving along the

x−axis and determine that its position as a function of time is given
by
x(t) = 2t − 3t2 + t3 .
(a) What is the position at time t = 2 seconds?
(b) What is the velocity at time t = 2 seconds?
(c) What is the acceleration at time t = 2 seconds?
(d) How far did it travel between times t = 0 and t = 2 seconds?
(Note: Distance, not displacement! It might be helpful to plot x vs t
for 0 ≤ t ≤ 2.)
Problem 1.3. The first train leaves the station and accelerates at a
constant rate to its maximum speed of 100 km/hr, reaching this speed
at a distance of 2 km from the station. Five minutes later, a second
train leaves the station and accelerates to 100 km/hr in 4 km. What is
the distance between the two trains when they both reach maximum
speed?
Problem 1.4. A rocket is trying to land on a barge in ocean. The
rocket engines deliver an upward force of 36,000 N and the rocket is
observed to be descending at a constant safe speed of 3 m/s when it is
100 m above the barge. But suddenly there is a malfunction and the
rocket engine thrust is reduced to 30,000 N.
(a) What is the mass of the rocket?
(b) What is the acceleration of the rocket after the malfunction?
(c) Assume this acceleration is maintained constant during the final
descent. What is the speed of the rocket when it contacts the barge?
Problem 1.5. You travel a distance d in time t. (a) If you travel
at speed v1 for half the time and at v2 for the other half of the time,
what is your average speed? (b) If you travel at speed v1 for half the
distance and at speed v2 for the other half of the distance, what is your
average speed?
Problem 1.6. There is a long straight road out in the desert and
it goes through a small town that has just one police car. The police
car accelerates at 2 m/sec2 unitl it reaches a maximum speed of 200
km/hour. A car full of escaped criminals skpeeds through the town
at its top speed which is 150 km/hour. The police car, starting from
rest, gives chase. How far from the town do the police catch up to the
criminals?
Problem 1.7. A police car is at rest on the side of the road when
a wild teenager comes speeding by at 75 miles per hour. The police
 1.7. PROBLEMS 25

car starts immediately and accelerates at 8 miles per hour per second.
At that same moment the teenager steps on the gas, but his car only
accelerates at 2 miles per hour per second. (a) How far from the starting
point does the police car overtake the speeder? (b) How fast are they
going at that time? (c) Why is the speed you calculated for the police
car unrealistic?
Problem 1.8. A brick is on a wooden plank that is resting on a
table. One end of the plank is slowly raised so that it forms an angle
θ with the horizontal table top. When θ = 60◦ the brick starts to slide
down the plane. (a) Draw the free body diagram. (b) Determine the
coefficient of static friction between brick and plank. (c) If the coef-
ficient of sliding friction is one half of the coefficient of static friction,
determine the acceleration of the block.
Problem 1.9. A block of mass M is on an inclined plane with a
coefficient of sliding friction µ. At a given instant of time the block is
at some point P on the plane and is moving up the plane with a speed
v0 . (a) Obtain the time for the block to reach its highest point relative
to P. (b) Obtain the time for the block to slide back down to point P.
(c) Obtain an expression for the velocity of the block at the time it
returns to P.
Problem 1.10. A toy rocket burns fuel for 1.5 seconds. During
that time, it accelerates upwards at 60 m/s2 . It then coasts upwards
to some maximum altitude before falling back down. Determine the
maximum altitude reached.
Problem 1.11. Atwood’s machine consists of two weights (M1 and
M2 ) suspended at the ends of a string that passes over a pulley. Assume
massless, inextensible strings and a frictionless pulley. Let M1 =6 kg
and M2 = 5.5 kg. The masses are released from rest. Determine the
distance descended by the 6 kg mass when its velocity reaches 0.5 m/s.
Problem 1.12. (a) Determine the rotational kinetic energy of a
wheel of your bicycle when your linear speed is 20 km/hour. You
may assume the wheel is a hoop of mass 1.5 kg and radius 30 cm. (b)
Compare your result with the translational kinetic energy of the wheel.
(c) Give a theoretical explanation for the relationship found in parts
(a) and (b).
Problem 1.13. The frictional force between water and seabed in
shallow seas causes an increase in the day by about 1 ms/century.
Determine the torque that causes this change. Assume the Earth is a
uniform sphere.
26 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

Problem 1.14. A meter stick has a pivot at one end. It is found

to be in static equilibrium when acted upon by three forces that act
at different points along the meter stick and act in different directions.
We conclude that the net torque about the pivot is zero. Prove that
the net torque about any other (arbitrary) point is also zero.
Problem 1.15. To build the pyramids it was necessary to pull
heavy stones up inclined planes. Suppose a 2000 kg stone was dragged
up a 20◦ incline at a speed of 0.25 m/s by a gang of 20 laborers. The
coefficient of kinetic friction between the stone and the incline was 0.4.
How much power was exerted by each laborer?
Problem 1.16. Two wooden blocks, both of mass 3 kg, are sliding
in the same direction at 2 m/s on a frictionless horizontal surface. Let
M1 = M2 = 3 kg and assume their speed is 2 m/s. A third block of
mass 1 kg is also sliding in the same direction at a speed of 10 m/s,
and it collides with the trailing 3 kg block. The third block is covered
with a sticky, gooey substance, so it sticks to the trailing block. This
combination block catches up with and collides elastically with the
leading 3 kg block. Determine the final speed of the leading block.
(Note that this is a one-dimensional problem. You may find it easier
to solve the problem if you insert numerical values sooner rather than
later.)
Problem 1.17. You are driving along at 60 mph (=100 km/hr).
Your car’s wheels have a radius of 35 cm. (a) Determine the angular
velocity of the wheels. (b) What is the angular displacement of a wheel
when you travel 1 km?
Problem 1.18. Aeronautical engineers have developed “tip jet”
helicopters in which small jet engines are attached to the tips of the
rotor. One such helicopter was powered by two ramjets. For a ramjet
to develop thrust, it needs to be moving through the air quite rapidly,
and is not efficient until it is moving at about 1000 km/hr. Assume
the rotor diameter is 10 meters. Determine the angular speed of the
rotor when the ramjet is moving through the air at 1000 km/hr.
Problem 1.19. Figure 1.8 shows an angle iron of mass M hang-
ing from the pivot at point P. The three segments have equal length.
Determine whether or not this is a stable equilibrium orientation.

Problem 1.20. A solid ball of mass M and radius R rolls down an

inclined plane. What is its translational speed when it has descended
a vertical distance h?
 1.7. PROBLEMS 27

Figure 1.8. An angle iron. Determine equilibrium.

Problem 1.21. A disk of mass 72 kg and radius 50 cm is rotating

at 2000 rpm. (a) Determine its angular momentum. (b) If acted upon
by a retarding force of 20 N acting tangent to the rim of the disk,
determine the time required to stop the disk.

COMPUTATIONAL PROJECTS: The following problems can

be solved easily if you know a computer language such as MATLAB,
Python, FORTRAN, or C++. A few can also be solved using a spread
sheet program such as EXCEL. Although some of the computational
projects in this book can be solved analytically, they will require a
significant amount of brainless labor that can be done much more con-
veniently using a computer.
Computational Project 1.1. The position of a particle as a
function of time is given by x = 5t3 − 2t (meters). Plot the position
as a function of time for the interval t = −5 sec to t = +10 sec. Using
the relationship v = ∆x/∆t, obtain the average velocity at one second
intervals. Plot the average velocity as a function of time and on the
same graph, plot the analytical expression for the velocity.
Computational Project 1.2. This is a more realistic version
of problem 1.4. In that problem a teenager driving at 33.5 m/s (∼75
mph) speeds past a parked policeman. The policeman and the teenager
then accelerate at given constant rates and you are required to deter-
mine how far from the starting point the policeman catches up to the
teenager. To make the problem somewhat more realistic, assume the
teenager accelerates at aT = kT e−bT t and the policeman accelerates at
aP = kP e−bP t , where kT = 2.0 m/s2 and kP = 3.5 m/s2 . By plotting
the positions as functions of time for various values of bT and bP obtain
reasonable values for these constants. (Make sure your answers are
28 1. A BRIEF REVIEW OF INTRODUCTORY CONCEPTS

reasonable. Obviously, the policeman will catch the teenager, but your
answer is not reasonable if the distance required is hundreds of miles!)
Chapter 2
Kinematics

Kinematics in one dimension is fairly simple because position, ve-

locity and acceleration can be treated as scalars. Motion in a straight
line and rotation about a fixed axis are examples of one-dimensional
motion. The motion of a projectile in a constant gravitational field can
be resolved into two linked one-dimensional problems. Going to two
or three dimensions complicates the problem significantly because now
you need to express the basic quantities as vectors. In this chapter you
will learn the relations between the position, velocity and acceleration
vectors in the three main coordinate systems used in physics: Cartesian
coordinates, cylindrical coordinates and spherical coordinates. (Rota-
tional motion in three dimensions is considerably more difficult than
linear motion and will be left to Chapters 13 and 16.)
Although some of the material presented in this chapter will be
familiar to you, you will also find many new concepts. These concepts
are used throughout the course, so please make sure you understand
this chapter thoroughly. Be aware that many students find this material
rather difficult.
I think it is important for you as a physicist to know something
about the people upon whose shoulders you are standing, so in this
book I have included a few sections marked “Optional Historical Note.”
You will encounter very little physics in these sections, but you might
find them interesting and helpful in placing a few famous physicists in
historical context. The first Historical Note describes the life of the
person who invented your profession.

29
30 2. KINEMATICS

2.1. Galileo Galilei (Optional Historical Note)

Galileo Galilei (1564 - 1642) was a brilliant but difficult man whose
studies of the physical world launched a scientific and cultural revolu-
tion. This cantankerous Italian genius was the first modern scientist.
He rejected authority and based his conclusions on observation, exper-
iment, and rational analysis. His ideas were opposed by the powers
of the Church, and he ended his days under house arrest. Neverthe-
less, his view of the universe and his methods for discovering scientific
truth eventually won out and today every physicist is an intellectual
descendent of Galileo.1
Galileo did not invent the telescope, but he built one and was prob-
ably the first person to make a scientific study of what he saw in the
sky. In short order, he discovered that the Moon has mountains and
valleys, that the Milky Way is made up of individual stars, that Jupiter
has satellites orbiting it, and that the Sun rotates and has spots on its
surface. He became the first “celebrity scientist.” He got into trouble,
however, because he was outspoken and had little patience with peo-
ple who put faith above reason. He was particularly offensive towards
those who believed the Earth was the center of the universe and that
the Sun, planets and stars revolved around it. This “geocentric” uni-
verse was based on the philosophy of Aristotle which had been adapted
to Christianity by Thomas Aquinas and was subsequently accepted as
the official philosophy of the Church. Therefore, Galileo’s outspoken
attacks on the geocentric theory were considered by many to be hereti-
cal. Galileo was eventually called to Rome and ordered by Church
authorities to neither teach nor defend the Copernican theory that the
Sun is the center of the Universe. Shortly thereafter Galileo wrote a
book in the form of a dialogue between three men. In this book he
ridiculed the Aristotelians and even put some arguments used by the
Pope in the mouth of a character named “Simplicio” who was pictured
as a bit of a dunce. Galileo was called back to Rome by the Inquisition
and was eventually condemned to life imprisonment, to take place in
own home. He was 70 years old at the time. He continued to work,
publishing his last book, “Discourses and Mathematical Demonstra-
tions Relating to Two New Sciences” in 1638. This book describes
much of the work in physics that he had carried out over the previous
30 years. He died at age seventy eight.
Galileo’s life is an instructive example of the conflict between a
brilliant individual and the powers that rule society. There is no doubt
1Anexcellent biography of Galileo is “Galileo’s Daughter” by Dava Sobel,
Walker and Co., New York, 1999.
 2.2. THE PRINCIPLE OF INERTIA 31

that Galileo changed the world and our understanding of it. Some
people condemn the Church authorities for trying to stop scientific
progress; others are more understanding of the authorities who saw
their entire world view threatened by a revolutionary iconoclast.

2.2. The Principle of Inertia

Among his many accomplishments, Galileo was arguably the first
person to truly understand motion. He performed a series of experi-
ments with pendulums of different lengths, as well as falling bodies and
balls rolling down inclined planes. His experiments with these simple
devices allowed him to grasp the essential aspects of motion, something
that had eluded the greatest minds of previous ages. Prior to Galileo’s
work many respected philosophers (including Aristotle) described mo-
tion in ways that modern scientists consider to be either meaningless
or just plain silly. For example, they said, “Water flows downhill be-
cause it is trying to return to its natural place, which is the ocean.”
Or they said, “Smoke rises because smoke is air and the natural place
for air is up in the sky.” They explained the motion of the stars and
planets with the preposterous statement that, “The celestial bodies are
perfect, therefore they move about the Earth in perfect circles.”
A crucial element in this erroneous theory of motion was the idea
that an object will move to its “natural place” and then remain at
rest in that place. This was called, of course, “natural motion.” For
something to move away from its natural place, it was necessary to give
it a push or a pull. This kind of motion, requiring a force, they called
“violent motion.”
Galileo’s most significant insight was that objects do not tend to
come to rest, but rather they tend to keep on moving. In fact they tend
to keep on moving at a constant speed in a straight line. (We call this
uniform motion.) The property of material bodies to maintain their
state of motion is called inertia. As you recall from your introductory
course in physics, the law of inertia states that:
A body will remain in uniform motion
as long as no net external force acts on it.
This basic principle has come to be known as Newton’s first law.
We will return to it later.
Although his equipment was crude, Galileo’s experiments showed
him that time is an essential component of motion. Mechanical clocks
had not yet been invented, but Galileo devised a number of ingenious
ways to estimate time intervals. He counted his pulse or watched the
32 2. KINEMATICS

oscillations of a pendulum or weighed the amount of water dripping

out of a container.
Time is a very difficult concept to comprehend. Everyone has an
intuitive, qualitative understanding of time. If you think about it, you
will realize that your personal definition of time has some connection
to the motion of material bodies. The length of a day is related to the
rotation of the Earth, and a year is the Earth’s orbital period. Most
people describe an hour and a minute in terms of the motion of the
hands on a clock. Scientists define the second in terms of the vibrations
of certain molecules. Is it possible to define time without reference to
motion? If all objects in the universe were perfectly still, would time
exist? Newton avoided these philosophical questions by assuming time
to be an absolute parameter that is continuously changing at a constant
rate. Einstein gave an even simpler definition when he said, “Time is
what you measure with a clock.” We shall adopt Newton’s point of
view at present, then come back for a deeper look at the question of
time when we consider Einstein’s theory of relativity in Chapter 19.2
As mentioned above, kinematics is the study of the relationships
between position, velocity, acceleration, and time. Galileo and his dis-
ciples grappled with these concepts. Fortunately, we can use many
powerful mathematical tools that were not available to Galileo. For ex-
ample, the calculus was invented by Newton and Leibniz after Galileo’s
death, and vector analysis was not developed until about 150 years ago.

2.3. Basic Concepts in Kinematics

Let me remind you that the general relations between position (r),
velocity (v), and acceleration (a) are given by
dr
v= = ṙ,
dt
dv
a= = v̇ = r̈.
dt
One-dimensional motion refers to any kind of motion for which the
position can be described in terms of a single parameter. For example,
the position of a car driving down a curving road can be specified
by its distance from the starting point. Similarly, the position of a
bead sliding along a wire can be specified by the distance to the bead
from some given point on the wire. A single number is enough to
2Some physicists believe that time, in the Newtonian sense of one event happen-
ing before another event, does not apply in the realm of quantum mechanics. For
an interesting discussion of this problem see the article by Charles Seife, “Quantum
Physics: Spooky Twins Survive Einsteinian Torture,” Science, 294, 1265, (2001).
 2.3. BASIC CONCEPTS IN KINEMATICS 33

completely specify the position of the bead, no matter how the wire
may be twisted or curled. However, the linear distance may not be the
most convenient parameter to use. For example, the position of a bead
sliding on a circular hoop might best be described in terms of an angle.
The simplest example of one-dimensional motion is a particle mov-
ing along a straight line, such as the x-axis. The acceleration is
dv
a= .
dt
Here I wrote a and v rather than a and v because if the motion is one-
dimensional there is no need to use vectors. Multiplying both sides by
dt gives the following very simple differential equation:
dv = adt.
To solve, you integrate both sides:
Z vf Z tf
dv = adt,
v0 t0

where the limits indicate that the time runs from t0 to tf (that is,
“initial” time to “final” time), and the initial and final values of the
velocity are v0 and vf .
The quantity t0 represents the time when the particle was at the
initial position x0 . Since the starting time for any process is arbitrary
(it just depends on when you push the button on the stopwatch) it is
nearly always set equal to zero (t0 = 0). R
Now you can go no further. You cannot evaluate the integral adt
because you do not have an expression for the acceleration as a function
of time. But if you do know the acceleration as a function of time,
a = a(t), then you can carry out the integration. When you do so,
you will obtain an expression for the velocity as a function of time,3
v = v(t). Once you have determined the velocity, you can continue the
analysis and write

dx = v(t)dt,
which follows immediately from the definition of velocity.
Integrating once again, you will obtain an expression for the posi-
tion as a function of time,
x = x(t),
or, more explicitly,
x = x(t, v0 , x0 ).
3To be precise, this will give you an expression for the velocity in terms of the
time, the initial time and the initial velocity, v = v(t, t0 , v0 ).
34 2. KINEMATICS

2.3.1. Motion in One Dimension with Constant Accelera-

tion. As discussed in Section 1.1.1, if the acceleration is constant, we
can integrate dv = adt to obtain
v(t) = v0 + at, (2.1)
and
1
x(t) = x0 + v0 t + at2 . (2.2)
2
I am sure you are very familiar with Equations (2.1) and (2.2) from
your introductory physics course. In fact, you probably memorized
them. However, it is more important at this stage in your physics
career to understand the process for obtaining these equations.

Exercise 2.1. Starting with x = v0 t + 12 at2 and v = v0 + at, obtain

the relations
1
x = (v + v0 )t, (2.3)
2
and
2ax = v 2 − v02 . (2.4)
Are these expressions valid if the acceleration is not constant?

I must emphasize that the formulas obtained above (Equations 2.1

through 2.4) are only valid for constant acceleration. If the acceleration
is a function of time or velocity or position, then you must go through
the entire procedure to obtain the correct expressions for v(t) and x(t).
2.3.2. Projectile Motion (Motion in a Plane). A projectile is
any object that is launched with some initial velocity and then moves
under the action of the gravitational force which imparts to it a con-
stant downward acceleration.
The horizontal motion is motion at constant velocity, and therefore
x = x0 + v0x t,
and the vertical motion is motion at constant acceleration, and there-
fore
1
y = y0 + v0y t − gt2 ,
2
where g = 9.8 m/s.
It is helpful to remember that at the top of the trajectory the ver-
tical velocity (vy ) is zero. This gives us a useful way to determine
 2.3. BASIC CONCEPTS IN KINEMATICS 35

the time for the projectile to reach the top of its trajectory. Since
vy = v0y + ay t we have
0 = v0y − gttop
so
ttop = v0y /g.
Worked Example 2.1. Consider the projectile motion of a
cannon ball. Assume the initial speed is v0 and the cannon is aimed
at an angle θ above the horizontal. Determine the range.
Solution: The range (R) is the horizontal distance traveled
by the projectile along a flat horizontal plane; it is the value of
x when y = 0. The initial velocity components are v0x = v0 cos θ
and v0y = v0 sin θ. Since the total time of flight is twice the time
required to reach the top of the trajectory, the range is given by
x(t = 2ttop ). It is easy to obtain a formula for the range in terms
of the initial velocity as follows:
vy = v0y − gt.
But vy = 0 when t = ttop so 0 = v0y − gttop = v0 sin θ − gttop .
Consequently,
ttop = (v0 sin θ) /g.
Therefore,
R = v0x (2ttop ) = 2v0 cos θ (v0 sin θ) /g,
v02
R=2 sin θ cos θ. (2.5)
g

Our analysis of projectile motion made two assumptions: (1) The

acceleration of gravity is constant, and (2) There are no horizontal
forces acting on the projectile. These assumptions mean that we are
assuming a flat, non-rotating Earth, and that there is no air resis-
tance acting on the body. These seemingly preposterous assumptions
(a flat, motionless, airless Earth!) illustrate an important technique
in physics. When faced with a difficult problem, we solve a similar,
simpler problem. Frequently, the solution to the simple problem is
sufficiently accurate. The solution of the simple problem is often re-
ferred to as the “zeroth order” approximation to the solution of the
real problem. Later in this course you will learn techniques for includ-
ing the factors that were ignored in the zeroth order approximation to
get results that are nearer and nearer to the exact solution. You will
appreciate more fully how this technique works when you study projec-
tile motion on a rotating Earth in Chapter 13. As you can imagine, a
36 2. KINEMATICS

working physicist is continually faced with very complicated problems.

It is part of a physicist’s education to learn what is fundamentally im-
portant and what can be left until later as a refinement (or “higher
order approximation”). For example, you can estimate the trajectory
of a thrown baseball with reasonable accuracy by assuming a flat, non-
rotating Earth, but if you want to calculate the trajectory of a rocket,
those factors will have to be included.

Exercise 2.2. Determine the elevation angle for a projectile such

that the range will be maximized. (Hint: the maximum value of a
function is determined by setting its derivative to zero.) Answer: 45◦ .
Exercise 2.3. An airplane traveling at 900 km/hr at 5,000 m
altitude is directly over the target when it drops a bomb. How far
from the target does the bomb hit? Answer: ' 8 km.

In working problems involving projectile motion in a uniform grav-

itational field, keep in mind that the horizontal motion has constant
velocity (vx = constant) and the vertical motion has constant accel-
eration (ay = constant). Projectile motion is a combination of two
independent one-dimensional motions. These are related to one an-
other through the time. If you eliminate the time from the equations,
you obtain an equation for the trajectory of the projectile as shown in
the following example.

Worked Example 2.2. Prove that the path of a projectile in

a uniform gravitational field is a parabola
Solution: The equation of a parabola whose axis is parallel to
the y axis in Cartesian coordinates is
y = ax2 + bx + c.
In section 2.3.2 we showed that for a projectile
1
y = − gt2 + v0y t + y0 ,
2
x = v0x t + x0 .
Therefore,
x − x0
t= ,
v0x
 2.3. BASIC CONCEPTS IN KINEMATICS 37

and consequently
g (x − x0 )2 x − x0
y=− 2
+ v0y + y0 ,
2 v0x v0x
g v0y v0y
y = − 2 (x2 − 2xx0 + x20 ) + x− x0 ,
2v0x v0x v0x
g g v0y gx2 v0y
y = − 2 x2 + [( 2x0 + )]x + (− 20 − x0 + y0 ).
2v0x 2v0x v0x 2v0x v0x
This has the required form as can more easily be appreciated if we
set x0 = 0 and y0 = 0 to obtain
g 2 v0y
y=− x + x.
2v0x v0x

Although projectile problems tend to be straightforward, sometimes

it is necessary to use a more subtle approach, as illustrated in the
following example.
Worked Example 2.3. A cannon that fires shells with a muz-
zle speed v0 is mounted on the top of a cliff a height h above a level
plain. Show that the angle at which thepcannon should be aimed
to give maximum range is θ = sin−1 (v0 / 2(v02 + gh)).
Solution: I am including this problem to show that a seem-
ingly simple projectile problem can require considerable ingenuity
to obtain a solution. The obvious approach would be to attempt
to solve the problem by using the fact that if t is the time of flight
(tf ), then the value of x is the range (R) and the value of y is
−h (where we assume that the coordinate origin (0, 0) is at the
cannon). That is
R = (v0 cos θ) (tf ),
1
−h = (v0 sin θ) (tf ) − g(tf )2 .
2
You can solve the first equation for tf , obtaining tf = R/(v0 cos θ).
Plugging into the second equation leads to
 2
gR 1
− h = R tan θ − 2
. (2.6)
2v0 cos2 θ
Now, you could solve this quadratic equation to obtain an expres-
sion for R as a function of θ. Then taking the derivative of R with
respect to θ and setting it equal to zero gives an expression for θ
38 2. KINEMATICS

that maximizes R. However, the resultant expression is so compli-

cated that you will find it extremely difficult to solve for θ. (You
might want to give it a try.)
However, the problem can be solved fairly simply by using a
clever trick. Instead of solving for R and taking its derivative
respect to θ, note that
dR dR dh
= .
dθ dh dθ
This expression will be zero if dhdθ
= 0. So going back to the expres-
sion above for −h, equation (2.6), we have
  2 
dh d gR 1
− = R tan θ − = 0,
dθ dθ 2v02 cos2 θ
gR
0 = 1 − 2 tan θ,
v0
2
v0
R= .
g tan θ
Plugging this into equation (2.6) yields
2
v02 v02 v02 v02

g
−h = − 2 = − .
g 2v0 cos2 θ g tan θ g 2g sin2 θ
Therefore
v02 v02
= + h,
2g sin2 θ g
v0
sin θ = p 2 .
2(v0 + gh)

2.3.3. Rotation About a Fixed Axis. Rotational motion about

a fixed axis (as discussed in Section 1.5) is one dimensional motion
described by the variables θ, ω and α where
dθ
ω= ,
dt
dω d2 θ
α= = 2.
dt dt
Worked Example 2.4. A physics student is carefully observ-
ing a wheel while it is being spin balanced. The wheel speeds up for
10 seconds and subsequently rotates at a constant angular speed
of 50 rad/sec. Going home, the student decides to write an expres-
sion for the acceleration of the wheel that would account for this
 2.3. BASIC CONCEPTS IN KINEMATICS 39

behavior, and comes up with

α = bte−ct (rad/s2 ).
Obtain an expression for the angular velocity (ω) as a function of
time. What are reasonable values for b and c?
Solution: Using the definition α = dωdt
,
Z ω(t) Z t
dω = αdt
ω0 t=0
Z t
ω(t) − ω0 = bte−ct dt
0
Note that ω0 = 0. Integrating by parts,
 t
1 −ct
ω(t) = b 2 e (−ct − 1)
c 0
b
= − 2 (e−ct )(ct + 1) − e0 (1)
 
c
b 
= 2 1 − e−ct (ct + 1) ,

c
Note that at time t = 0 the angular speed is zero. It increases
rapidly for a time, and then asymptotically approaches a maximum
value of b/c2 . If the time to spin up to this value is 10 seconds, then
at t = 10 the term e−ct ' 0, suggesting a value of c near unity. Let
c = 1. The final angular speed of the wheel is 50 radians/second so
b/c2 = 50, and b ' 50. It is instructive to plot α vs t and ω vs t.

2.3.4. The Relation between Linear and Rotational Mo-

tion. Imagine a child has tied a string to a rock and is whirling it
around in a circle. If the stone makes one revolution every T seconds,
it has an angular speed given by
2π
ω= .
T
It also has a linear speed that is tangent to the circular path (as is
obvious if the child lets go of the string). If the speed is constant, the
linear (or tangential) speed is given by
2πr
vT = ,
T
where r is the radius of the circular path. Clearly, the linear and
angular speeds are related by
vT = ωr.
40 2. KINEMATICS

We shall usually represent the angular velocity by a vector ω whose

magnitude is the angular speed ω and whose direction is perpendicular
to the plane of the motion. (If you curl the fingers of your right hand
in the direction of the motion, your thumb will point in the direction
of the vector ω. See Figure 2.1.)
Thus, the tangential velocity vT is related to the angular velocity
ω by

vT = ω × r. (2.7)

Figure 2.1. The direction of the vector angular velocity

is given by a right hand rule. The fingers of the right
hand point in the direction of the motion and the thumb
points in the direction of the vector ω.

It is interesting to note that rotational motion introduces two differ-

ent kinds of acceleration. To appreciate this, consider a rotating wheel.
If the wheel speeds up or slows down, it has an angular acceleration
given by α = dω/dt. Now recall that linear acceleration is defined as
a change in the linear velocity vector. Therefore, a car moving in a
circle at constant speed is accelerating. This acceleration is due to the
change in the direction of the velocity vector. Similarly, a point on
a rotating body is accelerating even if the body rotates at a constant
angular speed. Figure 2.2 illustrates the change in the velocity vector
of a point on the rim of a wheel that is rotating about a fixed axis with
constant angular speed ω.
 2.3. BASIC CONCEPTS IN KINEMATICS 41

v(t+Δt)

Δv v(t)
Δv
v(t) v(t+Δt) Δθ
Δθ

Figure 2.2. Illustration of centripetal acceleration.

The sketch on the left represent the motion of a par-
ticle, and the sketch on the right is the hodograph, a plot
of the velocity vectors.

By definition4,
dv v(t + ∆t) − v(t) ∆v
a= = lim = lim .
dt ∆t→0 ∆t ∆t→0 ∆t

Tip-to-tail addition of the vector v(t + ∆t) and the vector −v(t)
shows that ∆v is a vector pointing towards the center of the circle.
To help you see this, on the left side of Figure 2.2, I drew the vector
∆v at a midpoint between the vectors v(t) and v(t + ∆t). You can
appreciate that the vector ∆v is pointing approximately towards the
center of the circle. As ∆t → 0, the vector ∆v points more and more
closely towards the center.
The same concept is shown in a different way on the right hand side
of Figure 2.2. If the angular velocity is constant then the speed (that
is, the magnitude of the velocity vector) is a constant equal to v, but
the direction of the vector changes with time. The tip of the velocity
vector traces out a circle, as indicated on the right hand side of Figure
2.2. (A plot showing the path of the tip of the velocity vector is called a
hodograph.) For circular motion at constant speed, the hodograph is a
circle of radius v. The vector ∆v is a chord of this circle. Transferring
the vector ∆v back to the left hand side of the figure shows that it
points toward the center of the circle.
Consequently, a particle moving at a constant speed in a circular
path is accelerating towards the center of the circle. We can determine
the magnitude of this acceleration by considering the hodograph again.
In the limit of infinitesimally small times, ∆θ → dθ, and ∆v → dv,
4In the rest of this section, the velocity v should be expressed as vT but I left
off the subscript to keep the notation simple. This is common practice and one
must distinguish between tangential and linear velocity from the context
42 2. KINEMATICS

and the chord approaches the subtended arc. From the basic relation
ds = rdθ, but applied to the hodograph, we see that in the limit, the
magnitude of ∆v is dv = vdθ. Therefore,

dv vdθ
a= = .
dt dt
But dθ = ds/r, so

v ds v2
a= = .
r dt r
That is, if a particle is moving in a circle at constant speed, it is
accelerated towards the center with a “centripetal” acceleration given
by
v2
ac = .
r
Since v = rω, this can be written in the equivalent form
ac = ω 2 r.
Note that a point in a rotating body can experience three different
types of acceleration:
(1) linear acceleration (if the body as a whole is accelerating),
(2) tangential acceleration (aT = rα; if the angular velocity of
the body is changing),
(3) centripetal acceleration (ac = v 2 /r = ω 2 r; due to the rota-
tion of the body).

Exercise 2.4. (a)Determine the centripetal acceleration of the Earth

due to its orbital motion about the Sun. Assume the orbit is circular.
(b) Determine the centripetal acceleration of a point on the equator
of Earth due to the rotational motion of the Earth. Answers: (a)
5.95 × 10−3 m/s2 , (b) 3.37 × 10−2 m/s2 .
Exercise 2.5. Show that the angles denoted ∆θ on the right and
left hand plots of Figure 2.2 are the same.
 2.4. THE POSITION OF A PARTICLE ON A PLANE 43

2.4. The Position of a Particle on a Plane

The position of a point P on a flat surface can be specified in a
variety of ways. One usually arbitrarily selects an origin and draws a
pair of perpendicular reference lines through it. (These are denoted x
and y in Figure 2.3.)

y P
5

3 r
2

1
q
x
1 2 3 4 5

Figure 2.3. The position of the point P in rectangu-

lar coordinates is (4, 5) and in polar coordinates it is
(6.4, 0.9) where the angular measure is given in radians.

The most useful way to describe the position of P is by a vector

(r) drawn from the origin to P, as in Figure 2.3. The magnitude of r is
denoted by r and its direction is given by θ, the angle the vector makes
with the x axis.
To describe the vector r we can either use its components (namely,
x and y, the Cartesian coordinates), or its magnitude and direction (r
and θ, the polar coordinates). These are equivalent descriptions of r,
so (x, y) and (r, θ) are related. A glance at Figure 2.3 shows that these
relations are
x = r cos θ,
y = r sin θ.
The inverse relations (r and θ in terms of x and y) are
p
r = x2 + y 2 ,
θ = tan−1 (y/x).
Equations such as these that relate two sets of coordinates are called
“transformation equations.”
In your study of mechanics, you will only be using three different
coordinate systems, namely: Cartesian, cylindrical, and spherical coor-
dinates. There are, however, many other coordinate systems, some be-
ing quite specialized and appropriate for only a few types of problems.
44 2. KINEMATICS

In early editions of his book “Mathematical Methods for Physicists,”

George Arfken5 included a long discussion describing fourteen differ-
ent coordinate systems. Just to give you a flavor of these, consider
the “bipolar coordinates” η and ζ that are related to the Cartesian
coordinates, x and y by the following transformation equations:

a sinh η a sin ζ
x= and y = .
cosh η − cos ζ cosh η − cos ζ
You will be happy to know that we will not be using this particular
set of coordinates! Fortunately, most coordinate transformations are
much simpler than this one.

Exercise 2.6. A particle moves along the x axis at a constant speed

of 3 m/s. Determine dr/dt and dθ/dt. Answer: (3 m/s, zero)
Exercise 2.7. A particle travels in a circle of radius 3 m at a
constant speed of 5 m/s. Determine dr/dt and dθ/dt. Answer: (zero,
5/3 rad/s)
Exercise 2.8. In bipolar coordinates a particle is at η = 0.5, ζ
= 0.5. Assuming a = 1 (meter) determine its position in plane polar
coordinates r, θ. Answer: (2.832 m, 0.744 rad)

2.5. Unit Vectors

Unit vectors are the building blocks of kinematics. They are vectors
that point in the increasing direction of the coordinates and have a
length of one unit. (Unit vectors are unitless and have unit length.)
Figure 2.4 illustrates the three unit vectors ı̂, ̂, k̂ associated with the
Cartesian coordinate system (x, y, z).
The coordinate system in Figure 2.4 is a right-handed coordinate
system. This means that if you point the fingers of your right hand
along the x-axis and bend them towards the y-axis, your thumb will
be pointing in the direction of the z-axis.
5“Mathematical Methods for Physicists,” 2nd ed., George Arfken, Academic
Press, New York, 1970. The discussion was dropped in later editions of the book
because high speed computers allow one to solve most physics problems using sim-
ple coordinate systems. Many of the more exotic coordinate systems that were
applicable only to one or two specific problems have fallen into disuse.
 2.5. UNIT VECTORS 45

z
^k
^i y
^j
x

Figure 2.4. The unit vectors ı̂, ̂, k̂ in the Cartesian

coordinate system. The coordinate system is right-
handed.

Note that the Cartesian unit vectors are mutually perpendicular or

orthogonal.
In your vector analysis course you learned the definition of the cross
product (or vector product) of two vectors. Recall that if
c=a×b
then the magnitude of the vector c (denoted |c| or c ) is
| c |=| a | | b | sin(a, b), (2.8)
where sin(a, b) is the sine of the angle between a and b, and the direc-
tion of c is given by the right hand rule.
The definition of the dot product (or scalar product) of two vectors
is:
a · b = | a | | b | cos(a, b). (2.9)
The cross (or vector) product and the dot (or scalar) product are the
two most common ways of multiplying vectors. Later in this course you
will learn another type of vector multiplication which involves mathe-
matical objects called dyadics.
Consider the dot product of the unit vectors. It is quite obvious
from Equation (2.9) that

ı̂ · ı̂ = 1, ̂ · ̂ = 1, k̂ · k̂ = 1,
ı̂ · ̂ = 0, ı̂ · k̂ = 0, ̂ · k̂ = 0.
These relations describe the orthogonality of the unit vectors. Similarly,
from the definition of the cross product, Equation (2.8), we find that

ı̂ × ı̂ = 0, ̂ × ̂ = 0, k̂ × k̂ = 0,
46 2. KINEMATICS

and
ı̂ × ̂ = k̂, ̂ × ı̂= −k̂
̂ × k̂ = ı̂, k̂ × ̂= −ı̂
k̂ × ı̂ = ̂, ı̂ × k̂= −̂
The last six relations are easily remembered with the help of the
mnemonic of Figure 2.5.

^ ^
i i

^
+ ^
j
^
k
- ^
j
k

Figure 2.5. A mnemonic for the cross product of unit

vectors. The sketches indicate, for example, that ı̂ × ̂
= + k̂, and ̂ × ı̂ = - k̂.

Exercise 2.9. Two vectors are given by A = 2ı̂ +3̂ +4k̂ and
B = 3ı̂ − 2 ̂ Determine A + B, A − B, A · B and A × B.
Exercise 2.10. Calculate the angle between the vector V=2ı̂+4̂+6k̂
and the vector W=2ı̂+2̂+2k̂. Answer: 22.21◦
Exercise 2.11. The direction angles of a vector are the angles
between the vector and the x, y, and z axes. Determine the direction
angles of the vector r = 6ı̂+5̂−2k̂. Answer: α = 41.0◦ , β = 51.67◦ ,
γ = 104.3◦ .
Exercise 2.12. A vector forms the diagonal of a cube of side a.
Express it in terms of the unit vectors.
Exercise 2.13. Show that the cross product of two vectors A and
B can
 be written as
 the following determinant:
 ı̂ ̂ k̂ 
 
 Ax Ay Az 
 
 Bx By Bz 
Exercise 2.14. Let A and B be the sides of a parallelogram. Show
that the area of the parallelogram is A×B.
 2.6. KINEMATICS IN TWO DIMENSIONS 47

2.6. Kinematics in Two Dimensions

2.6.1. Cartesian Coordinates. In two dimensions, the position
of a point can be specified by two numbers. When using Cartesian
coordinates we usually set z = 0 and define the plane of motion to be
the x-y plane. (As long as the motion is confined to a flat plane, we
can let the plane be the z = 0 surface with no loss of generality.) In
Cartesian coordinates the position of a particle is given by
r =xı̂ + y̂.
The definition of velocity is v = ṙ = dr/dt, so
d
v= (xı̂ + y̂).
dt
The Cartesian unit vectors (ı̂, ̂, k̂) are constant in magnitude and
direction so their time derivatives are zero. Consequently, writing ẋ for
dx/dt and ẏ for dy/dt, the velocity is given by
v = ẋı̂ + ẏ̂.
Using the definition of acceleration, a = v̇ we obtain,
a =ẍı̂ + ÿ̂.

2.6.2. Plane Polar Coordinates. Let us now go through the

same analysis as above but for the plane polar coordinates (r, θ). That
is, let us derive expressions for position, velocity and acceleration in
plane polar coordinates. Although the analysis is complicated, it is
very important for you to understand exactly what I am doing. The
ideas presented in this section are the basis for the rest of this chapter.
The first thing we need to do is to define unit vectors for the polar
coordinates. In Figure 2.6 the point P is located at (r, θ) where r is
the length of the vector r and is equal to the distance from the origin
to the point P. The angle between r and the x-axis is denoted θ. At
the tip of vector r you can see the two unit vectors r̂ and θ̂. These
have unit length. The direction of r̂ is the direction of increasing r.
The direction of θ̂ is the direction of increasing θ. This means that an
increase in θ will cause the vector r to rotate in the direction of θ̂.
As long as their magnitudes and directions are unchanged, vectors
are allowed to “slide.” We can slide r̂ and θ̂ to the origin and represent
them on a unit circle as shown in Figure 2.7.
48 2. KINEMATICS

y ^ ^r
q
P
r

^
j q
^ x
i

Figure 2.6. Definition of the plane polar unit vectors

r̂ and θ̂. The unit vector r̂ points in the direction of
increasing r, and θ̂ points in the direction of increasing
θ.

The unit vectors r̂ and θ̂ are not constant because they are not
associated with a set of fixed axes. Rather, they are associated with
a vector r that can change with time. If r changes in magnitude, this
does not affect r̂ or θ̂ but if r changes in direction, then the direction
of the unit vectors r̂ and θ̂ will change. This means that r̂ and θ̂ are
functions of θ. In mathematical form we express this as
r̂ = r̂(θ),
and
θ̂ = θ̂(θ).
The position of a point in plane polar coordinates is given quite
simply by
r = rr̂.

^
j
^ ^
q r
q ^
i

Figure 2.7. The unit vectors r̂ and θ̂ and the unit cir-
cle. All the unit vectors originate at the center of the
circle and all have magnitude unity, so they end on the
circumference of the circle.
 2.6. KINEMATICS IN TWO DIMENSIONS 49

If you take the derivative of this equation with respect to time, you
obtain an expression for the velocity in polar coordinates.
dr d dr dr̂
v= = (rr̂) = r̂ + r .
dt dt dt dt
The last term comes from the product rule of differentiation and reflects
the fact that the unit vector r̂ may be changing with time. Although
the magnitude of r̂ is always unity, it may change in direction. Since r̂
is a function of θ, from the calculus the differential of r̂ is
dr̂
dr̂ = dθ.
dθ
So,
dr̂ dr̂ dθ dr̂
= = θ̇.
dt dθ dt dθ
But what is dr̂/dθ? Looking at Figure 2.7 you see that r̂ and θ̂ can be
resolved into Cartesian components as follows:
r̂ = 1 cos θı̂ + 1 sin θ̂,
where I explicitly included the number 1 to emphasize that the radius of
the unit circle is 1. You obtain a similar expression for θ̂ by inspection
of Figure 2.7. Dropping the “1” we write:
r̂ = cos θı̂ + sin θ̂, (2.10)
θ̂ = − sin θı̂ + cos θ̂. (2.11)
Therefore,
dr̂ d
= (cos θı̂ + sin θ̂) = − sin θı̂ + cos θ̂ = θ̂.
dθ dθ
Furthermore,
dθ̂
= − cos θı̂ − sin θ̂ = −r̂.
dθ
You may wish to memorize the following relations, as you will be using
them frequently in this and other courses.

dr̂ dθ̂
= θ̂ and = −r̂. (2.12)
dθ dθ
Having determined dr̂/dθ and dθ̂/dθ we can obtain dr̂/dt and dθ̂/dt
as follows:
dr̂ dr̂ dθ̂
= = θ̇θ̂,
dt dθ dt
and
dθ̂ dθ̂ dθ
= = −θ̇r̂.
dt dθ dt
50 2. KINEMATICS

As noted previously, in polar coordinates the velocity is given by

d d dr dr̂
v = r = (rr̂) = r̂ + r ,
dt dt dt dt
so
v =ṙr̂ + rθ̇θ̂. (2.13)
This equation for the velocity can be differentiated to obtain the ex-
pression for acceleration in polar coordinates:
a = (r̈ − rθ̇2 )r̂ + (rθ̈ + 2ṙθ̇)θ̂. (2.14)
We shall frequently refer to the expressions for velocity and acceleration
in terms of polar coordinates, so you should either memorize Equations
2.12, 2.13 and 2.14 or mark this page so you can find them easily.

Worked Example 2.5. A turntable rotates at a constant an-

gular speed ω. An ant crawls directly toward the rim along a radial
line at a constant speed b. You observe the ant from above. From
your point of view, the ant is moving in a spiral. Write an expres-
sion for the velocity and acceleration of the ant in polar coordinates.
Solution: Given ṙ = b and θ̇ = ω. Therefore, r = bt + r0 and
θ = ωt + θ0 . Note that θ̈ = 0 and r̈ = 0.
r = rr̂ = (bt + r0 )r̂
d dr̂
v = [(bt + r0 )r̂] = br̂ + (bt + r0 )
dt dt
v = br̂ + (bt + r0 )ω θ̂,
and
d
a= [br̂ + (bt + r0 )ω θ̂]
dt
dr̂ dθ̂
= b + bω θ̂ + (bt + r0 )ω
dt dt
= bω θ̂ − (bt + r0 )ω 2 r̂
= −(bt + r0 )ω 2 r̂ + 2bω θ̂.

Exercise 2.15. Derive Equation (2.14) from (2.13).

Exercise 2.16. Given r = a(ı̂ sin ωt+̂cos ωt), where ω is a con-
stant. (a) Determine the magnitude of the velocity. (b) Prove the
velocity is perpendicular to r. (c) Describe the motion. Answer: (a)
aω.
 2.7. KINEMATICS IN THREE DIMENSIONS 51

Exercise 2.17. The position of a certain particle is described by

r = 4t (meters), θ = 0.2t (radians). Determine its velocity as a func-
tion of time in (a) polar coordinates and (b) Cartesian coordinates. An-
swers: (a) v=4r̂+0.8tθ̂. (b) v = (4 cos 0.2t−0.8t sin 0.2t)ı̂+(4 sin 0.2t+
0.8t cos 0.2t)̂.
Exercise 2.18. You are analyzing the motion of a red dot painted
on the rim of a wheel of radius r = 5.0 cm. It returns to the same
point every 2 seconds. At t = 0 the particle is at (r, θ) = (5, 0).
Express the position and velocity vectors in Cartesian and polar co-
ordinates. Answers: r = 5r̂ = (5 cos πt)ı̂+(5 sin πt)̂, v = 5π θ̂ =
(−5π sin πt)ı̂+(5π cos πt)̂.

2.7. Kinematics in Three Dimensions

We now generalize to motion in three dimensions. We shall describe
three different coordinate systems: Cartesian coordinates, cylindrical
coordinates, and spherical coordinates. These are the most commonly
used three-dimensional coordinate systems in physics.
You will find that the cylindrical and spherical coordinates rely
heavily on the plane polar coordinates described in the previous section.

2.7.1. Cartesian Coordinates. The position of a particle in a

three-dimensional Cartesian coordinate system is given by
r =xı̂+y̂ + z k̂.
In an inertial (nonaccelerating) coordinate system the three unit
vectors (ı̂, ̂, k̂) are constant in magnitude and direction so their time
derivatives are zero. This makes it easy to obtain expressions for the ve-
locity and acceleration in terms of Cartesian coordinates. The velocity
is the time derivative of position.
dr
v= = ẋı̂ + ẏ̂ + ż k̂.
dt
Similarly, the acceleration is obtained by taking the time derivative of
the velocity.
dv
a= = ẍı̂+ÿ̂+z̈ k̂.
dt
52 2. KINEMATICS

2.7.2. Cylindrical Coordinates. To describe cylindrical coordi-

nates, visualize a particle located at some arbitrary point in a three
dimensional Cartesian coordinate system, as in Figure 2.8. Draw a
perpendicular from the x-y plane (that is, the plane z = 0) up to the
position of the particle. This perpendicular has magnitude z and di-
rection k̂ so we denote it by the vector z k̂. The point at which the
perpendicular intersects the z = 0 plane can be specified by its plane
polar coordinates. (But when using cylindrical coordinates we generally
use the symbols ρ and φ rather than r and θ.) The set of coordinates,
ρ, φ, z are called cylindrical coordinates. Figure 2.8 gives a graphical
representation of this coordinate system. Note that the vector from
the origin to the point is still denoted r and also note that φ is mea-
sured from the x-axis to the projection of r onto the x-y plane, which
is denoted ρ.

z
P
^
zk
r
y
x φ ρ

Figure 2.8. The position of point P can be specified

either in terms of the Cartesian coordinates x, y, z or the
cylindrical coordinates ρ, φ, z.

The next task is to determine unit vectors for the cylindrical coordi-
nate system. This is quite simple because the new coordinate system is
made up of old coordinates whose properties you have already learned.
You can appreciate from Figure 2.9 that ρ̂ and φ̂ are the same as the
plane polar unit vectors r̂ and θ̂, and the unit vector in the z direction
is just k̂ as in the Cartesian coordinate system.
Although k̂ is constant in magnitude and direction, ρ̂ and φ̂ will
change in direction as the angle φ varies. That is, ρ̂ and φ̂ are functions
of φ. We write:
ρ̂ = ρ̂(φ),
and
φ̂ = φ̂(φ).
The set ρ̂, φ̂, k̂ form an orthogonal set. The orthogonality of these
unit vectors leads the following set of easily proved properties:
 2.7. KINEMATICS IN THREE DIMENSIONS 53

ρ̂ · ρ̂ = 1, φ̂ · φ̂ = 1, k̂ · k̂ = 1
ρ̂ · φ̂ = 0, ρ̂ · k̂ = 0, φ̂ · k̂ =0
ρ̂ × φ̂ = k̂, φ̂ × k̂ = ρ̂, k̂ × ρ̂ = φ̂
The relationships between cylindrical coordinates and Cartesian co-
ordinates (the transformation equations) are:
x = ρ cos φ y = ρ sin φ z = z.
The inverse relationships are:
−1 y
p  
ρ= x +y 2 2 φ = tan z = z.
x
The unit vectors ρ̂ and φ̂ can be expressed in terms of ı̂ and ̂ as
follows:

ρ̂ = cos φı̂ + sin φ̂, (2.15)

φ̂ = − sin φı̂ + cos φ̂. (2.16)
The position of a particle in cylindrical coordinates is specified by

r = ρρ̂ + z k̂. (2.17)

The velocity is obtained by taking the time derivative of the position
vector:

dρ̂ dφ
v = ṙ = d
dt
(ρρ̂ + z k̂) = ρ̇ρ̂ + ρ dρ̂
dt
+ ż k̂ =ρ̇ρ̂ + ρ( dφ dt
) + ż k̂,
or
v =ρ̇ρ̂ + ρφ̇φ̂ + ż k̂. (2.18)
Similarly, the acceleration is given by
a = r̈ =(ρ̈ − ρφ̇2 )ρ̂ + (ρφ̈ + 2ρ̇φ̇)φ̂ + z̈ k̂. (2.19)

r
y
^ ^
ρ k
x φ
φ
^
ρ

Figure 2.9. Definition of the cylindrical unit vectors ρ̂,

φ̂, and k̂.
54 2. KINEMATICS

(You should be able to carry out the intermediate steps in obtaining

this expression for a. Don’t just nod your head and think, “Of course
I can do it.” Pick up your pencil and work it out! The exercise will
make the concepts stick in your mind.)

Worked Example 2.6. A bead slides on a wire bent into a

helix. The position of the bead as a function of time is given by
ρ = a, φ = ωt, z = bt2 , where a, ω, b are constants. Determine
an expression for the velocity and acceleration of the bead as a
function of time.
Solution: Recall that in three dimensions using cylindrical
coordinates, we have
r = ρ̂ + z k̂,
v = ρ̇ρ̂ + ρφ̇φ̂ + ż k̂,
a = (ρ̈ − ρφ̇2 )ρ̂ + (ρφ̈ + 2ρ̇φ̇)φ̂ + z̈ k̂.
For this particular problem
ρ = a ∴ ρ̇ = 0 ⇒ ρ̈ = 0
φ = ωt ∴ φ̇ = ω ⇒ φ̈ = 0
z = bt2 ∴ ż = 2bt ∴ z̈ = 2b
Consequently,
v = ρ̇ρ̂ + ρφ̇φ̂ + ż k̂ =0 + aω φ̂ + 2btk̂
= aω φ̂ + 2btk̂
and
a = (ρ̈ − ρφ̇2 )ρ̂ + (ρφ̈ + 2ρ̇φ̇)φ̂ + z̈ k̂
= 0 − aω 2 ρ̂ + (0 + 0)φ̂ + 2bk̂

= −aω 2 ρ̂ + 2bk̂

Exercise 2.19. Carry out the steps to obtain Equation (2.19).

Exercise 2.20. The position of a particle is given by ρ = 3t2 ,
φ = 2t and z = 12t, (distances in meters, angles in radians). What is
the acceleration at time t = 2 seconds? Answer: a = (−42ρ̂ + 48φ̂)
m/s2 .
 2.7. KINEMATICS IN THREE DIMENSIONS 55

2.7.3. Spherical Coordinates. The last set of coordinates we

will consider are called spherical coordinates. In these coordinates the
position of a particle is specified by the vector r which has length r and
direction (or orientation) specified by two angles, θ and φ, as shown
Figure 2.10. These angles are defined as follows: The “polar” angle θ
is measured from the z-axis to the vector r. The “azimuthal” angle φ
is measured from the x-axis to the projection of r onto the x-y plane.
The component of r along the z-axis is r cos θ. The component of r in
the x-y plane (denoted ρ) has a length of r sin θ. The x component of
r is, therefore, r sin θ cos φ and the y component is r sin θ sin φ.

z
^
r
z =r cos θ ^
φ
θr
^ ^
θ
k y =r sinθ sinφ
x = r sinθ cosϕ ^
ρ y
φ
ρ=rsinθ
x
Figure 2.10. Spherical coordinates and the unit vec-
tors r̂, θ̂, φ̂. Note that the Cartesian components of the
vector r are x = r sin θ cos φ, y = r sin θ sin φ and
z = r cos θ. The unit vectors r̂, θ̂, φ̂ point in the direc-
tion of increasing r,θ, φ, respectively. Also shown are the
cylindrical unit vectors ρ̂ and k̂.

What are the appropriate unit vectors? The unit vector r̂ is in the
direction of increasing r. The unit vector θ̂ is in the direction that the
tip of r will move if θ increases. In Figure 2.11 I sketched a quarter
circle to illustrate the motion of r as θ changes. The unit vector θ̂ is
tangent to this circle and has the direction of increasing θ. The unit
vector φ̂ lies in the x-y plane and gives the direction that the tip of
ρ̂ will move if φ increases. By sliding φ̂ down to the x-y plane, you
can see that φ̂ in spherical coordinates is the same as the φ̂ of the
cylindrical coordinates.
Although it may not be immediately obvious, the spherical unit
vectors r̂, θ̂, φ̂ form an orthogonal set. Figure 2.10 shows these three
mutually perpendicular unit vectors at the tip of the vector r. This is
56 2. KINEMATICS

φ
ρ

Figure 2.11. The unit vectors r̂, θ̂, and φ̂

probably the easiest way to remember them. In that figure, both r̂

and θ̂ lie in the shaded plane that contains r and the z-axis, and φ̂ is
perpendicular to that plane.
The relationships (transformation equations) between the spherical
coordinates and the Cartesian coordinates are:
x = r sin θ cos φ, (2.20)
y = r sin θ sin φ,
z = r cos θ.
The inverse transformation equations are:
1
r = (x2 + y 2 + z 2 ) 2 ,
p !
x 2 + y2
θ = tan−1 ,
z
y
φ = tan−1 .
x
Equations (2.20) are extremely useful and you might want to mem-
orize them. (However, they are easy to remember if you can visualize
Figure 2.10.) You will use these equations throughout this course as
well as in many upper division and graduate physics courses.
Figure 2.12 shows a unit circle that contains both r and the z-
axis. Its intersection with the x-y plane gives ρ̂, the unit vector for the
cylindrical coordinates. From the figure,
r̂ = cos θk̂ + sin θρ̂. (2.21)

Recalling that ρ and φ are the same as the plane polar coordinates
(formerly denoted r and θ), and using Equation (2.25),
ρ̂ = cos φı̂ + sin φ̂.
Consequently,
r̂ = sin θ cos φı̂ + sin θ sin φ̂ + cos θk̂. (2.22)
 2.7. KINEMATICS IN THREE DIMENSIONS 57

Figure 2.12. The unit vectors r̂, θ̂, φ̂ for spherical co-
ordinates and the unit circle. Also shown are k̂ and ρ̂.
In the sketch on the right, θ̂ has been translated from
the tip of r̂ to the origin.

We can also express θ̂ and φ̂ in terms of the Cartesian unit vectors

(ı̂,̂, k̂). From the unit circle on the right in Figure 2.12, it is clear that
θ̂ = cos θρ̂ − sin θk̂.
Substituting for ρ̂ we obtain
θ̂ = cos θ cos φı̂ + cos θ sin φ̂ − sin θk̂. (2.23)
Finally, φ̂ is the same as for the cylindrical coordinates, so
φ̂ = − sin φı̂ + cos φ̂. (2.24)
Now that we have determined r̂, θ̂, φ̂, we can evaluate their time
derivatives. Expressions for the velocity and acceleration are somewhat
more difficult to obtain in terms of spherical coordinates than in cylin-
drical coordinates because the unit vectors r̂ and θ̂ are functions of
both θ and φ. (It is clear from Figure 2.10 that a change in either θ or
φ will cause a change in the direction of r̂ and θ̂.) On the other hand,
φ̂ is a function only of φ. Thus we write, in the usual mathematical
notation,
r̂ = r̂(θ, φ),
θ̂ = θ̂(θ, φ),
φ̂ = φ̂(φ).
Recall that if a quantity, (say f ), is a function of several variables such
as x, y and z, then f = f (x, y, z), and by the rules of calculus, the
58 2. KINEMATICS

differential quantity df is given by:

∂f ∂f ∂f
df = dx + dy + dz.
∂x ∂y ∂z
(This simple relationship is used over and over again in physics. If you
have already taken a course in thermodynamics, you are aware how
important it is.)
Going back to the discussion of the unit vectors, you can appreciate
that if r̂ = r̂(θ, φ), then
∂r̂ ∂r̂
dr̂ = dθ + dφ,
∂θ ∂φ
and consequently
dr̂ ∂r̂ dθ ∂r̂ dφ
= + ,
dt ∂θ dt ∂φ dt
or
dr̂ ∂r̂ ∂r̂
= θ̇ + φ̇. (2.25)
dt ∂θ ∂φ
Therefore,
∂r̂
= cos θ cos φı̂ + cos θ sin φ̂ − sin θk̂. (2.26)
∂θ
But comparison with Equation (2.23) shows us that the right hand side
of this equation is simply θ̂.
Therefore,
∂r̂
= θ̂.
∂θ
We also need to determine ∂r̂/∂φ. Expressing r̂ in terms of the Carte-
sian unit vectors (Equation 2.22) and taking the derivative with respect
to φ we obtain
∂r̂
= − sin θ sin φı̂ + sin θ cos φ̂
∂φ
= sin θ(− sin φı̂ + cos φ̂),
or
∂r̂
= sin θφ̂.
∂φ
So finally, the expression for dr̂/dt (Equation 2.25) can be written

dr̂
= θ̇θ̂ + sin θφ̇φ̂.
dt
 2.7. KINEMATICS IN THREE DIMENSIONS 59

Similarly, we obtain the following two relations (see Problem 2.36):

dθ̂ ∂ θ̂ ∂ θ̂
= θ̇ + φ̇ = −θ̇r̂ + cos θφ̇φ̂, (2.27)
dt ∂θ dφ
dφ̂ ∂ φ̂ ∂ φ̂
= θ̇ + φ̇ = 0 − φ̇ρ̂ = −φ̇ρ̂. (2.28)
dt ∂θ ∂φ
Using Figure 2.12 you can show that,
ρ̂ = sin θr̂ + cos θθ̂,
so we can also write Equation (2.28) as
dφ̂
= −φ̇ sin θr̂ − φ̇ cos θθ̂.
dt
Having obtained the time derivatives of the three unit vectors r̂, θ̂, φ̂
it is straightforward to determine expressions for the velocity and ac-
celeration in spherical polar coordinates. Since
dr d dr dr̂
v= = (rr̂) = r̂ + r , (2.29)
dt dt dt dt
the velocity is
v = ṙr̂ + rθ̇θ̂ + r sin θφ̇φ̂. (2.30)
Taking the time derivative of both sides and using Equations (??) and
(??) leads to the following expression for the acceleration in spherical
polar coordinates.
a = (r̈ − rθ̇2 − rφ̇2 sin2 θ)r̂ + (rθ̈ + 2ṙθ̇ − rφ̇2 sin θ cos θ)θ̂
(2.31)
+(rφ̈ sin θ + 2ṙφ̇ sin θ + 2rθ̇φ̇ cos θ)φ̂.

Worked Example 2.7. Prove that the unit vectors in spheri-

cal coordinates are orthogonal. You may assume the orthogonality
of the Cartesian unit vectors.
Solution: The unit vectors for the spherical coordinates can
be expressed in terms of the Cartesian unit vectors by
r̂ = sin θ cos φı̂ + sin θ sin φ̂ + cos θk̂
θ̂ = cos θ cos φı̂ + cos θ sin φ̂ − sin θk̂
φ̂ = − sin φı̂ + cos φ̂
60 2. KINEMATICS

If two vectors are orthogonal (mutually perpendicular) their dot

product is zero. So
   
r̂ · θ̂= sin θ cos φı̂ + sin θ sin φ̂ + cos θk̂ · cos θ cos φı̂ + cos θ sin φ̂ − sin θk̂
= sin θ cos φ cos θ cos φ + sin θ sin φ cos θ sin φ − cos θ sin θ
= (sin θ cos θ)(cos2 φ + sin2 φ) − cos θ sin θ
= sin θ cos θ − cos θ sin θ = 0
∴ r̂ and θ̂ are perpendicular.
Similarly
 
r̂ · φ̂= sin θ cos φı̂ + sin θ sin φ̂ + cos θk̂ · (− sin φı̂ + cos φ̂)
= − sin θ cos φ sin φ + sin θ sin φ cos φ = 0
∴ r̂ and φ̂ are perpendicular.
Finally
 
θ̂ · φ̂= cos θ cos φı̂ + cos θ sin φ̂ − sin θk̂ · (− sin φı̂ + cos φ̂)
= − cos θ cos φ sin φ + cos θ sin φ cos φ = 0
∴ θ̂ and φ̂ are perpendicular.

Exercise 2.21. A particle is at (3,4,5) meters in Cartesian coordi-

nates. Determine r, θ, φ. Answer: (7.07, π/4, 0.927)

2.8. Summary
In this chapter you learned how to express the position, velocity,
and acceleration of a particle in two dimensions in Cartesian coordi-
nates and plane polar coordinates, and in three dimensions in Cartesian
coordinates, cylindrical coordinates and spherical coordinates.
In three dimensions the position, velocity, and acceleration in Carte-
sian coordinates are given by the simple relations:
r = xı̂ + y̂ + z k̂,
v = ẋı̂ + ẏ̂ + ż k̂,
a = ẍı̂ + ÿ̂ + z̈ k̂.
 2.9. PROBLEMS 61

These relationships for two dimensions in plane polar coordinates are:

r = rr̂,
v = ṙr̂ + rθ̇θ̂,
a = (r̈ − rθ̇2 )r̂ + (rθ̈ + 2ṙθ̇)θ̂.
In three dimensions using cylindrical coordinates, the relationships are:
r = ρρ̂ + z k̂,
v = ρ̇ρ̂ + ρφ̇φ̂ + ż k̂,
a = (ρ̈ − ρφ̇2 )ρ̂ + (ρφ̈ + 2ρ̇φ̇)φ̂ + z̈ k̂.
Finally, in spherical coordinates the position, velocity and acceleration
are given by:
r = rr̂,
v = ṙr̂ + rθ̇θ̂ + r sin θφ̇φ̂.
a = (r̈ − rθ̇2 − rφ̇2 sin2 θ)r̂ + (rθ̈ + 2ṙθ̇ − rφ̇2 sin θ cos θ)θ̂
+ (rφ̈ sin θ + 2ṙφ̇ sin θ + 2rθ̇φ̇ cos θ)φ̂.

2.9. Problems
Problem 2.1. A particle initially at rest undergoes an acceleration
given by a = 3e−0.5t m/s2 . Determine the terminal velocity.
Problem 2.2. The nitrogen atom in an ammonia molecule can be
assumed to oscillate in simple harmonic motion, so that its position
at any time t is given by z = A cos ωt where A and ω are constants.
Obtain expressions for its velocity and acceleration as functions of time.
Problem 2.3. This is the story of Freddy Physics who got a ticket
for going through a red light. He explained to the judge that if he
slammed on the brakes when he saw the signal light turn yellow, he
would not be able to stop before the intersection, and if he continued
at a constant speed the light would turn red before he reached the
intersection. In this problem you will determine how far from the light
Freddy was when the light turned yellow. Suppose he is traveling at
the speed limit, v0 , that the light remains yellow for a time t0 and
that Freddy’s reaction time is τ . His car decelerates at a rate a. (a)
Determine the distance d1 he will travel in time t0 if he continues at the
speed limit (b) Determine the distance d2 required to stop. (c) Show
that he is bound to go through the red light if t0 < τ + v0 /2|a|
62 2. KINEMATICS

Problem 2.4. Assume a particle is given an acceleration a = 10e−2t

m/s2 . The velocity at time t = 0 is zero. (a) Obtain expressions for the
velocity and position of the particle as functions of time. (b) Determine
the velocity as t → ∞.
Problem 2.5. An object has an acceleration that is inversely pro-
portional to the velocity squared: a = 9/v 2 m/s2 . Assume the object
is initially at rest at the origin. Determine its position at time t = 10
sec.
Problem 2.6. In some circumstances the acceleration is a function
of the velocity. (An example is the motion of an object in a resistive
medium such as air.) Assume a = −kv 2 where k is a constant. Deter-
mine the velocity and position as a function of time. What is the value
of the velocity if the time is very large?
Problem 2.7. Mary is playing outfield and John is playing third
base, 30.0 meters away. Mary can throw a baseball at 20.0 m/s. Ignor-
ing air resistance, determine the two launch angles she can use to get
the ball to John.
Problem 2.8. A cannon launches a shell at 250 m/s at an elevation
of 40◦ . Assuming level ground, determine the vertical and horizontal
components of the velocity one second before the shell hits the ground.
Problem 2.9. When great athletes or ballet dancers leap into the
air, they appear to “hang” at the top of the trajectory. (The reason
for this is that the vertical speed of a projectile is smallest near the top
of its path and greatest near the ground.) Show that half of the total
time for the jump is spent in the top one quarter of the trajectory.
Problem 2.10. The British Navy sends a gunboat to quell a revolt
on a British colony. This colony happens to be on an island exactly
3 km wide with a 1.5 km high mountain running down its center, as
shown in Figure 1.10. The gunboat is anchored offshore of the western
side of the island, 2.5 km from the center. On the eastern shore a
distance of 1.5 km from the center, the rebels have a cannon that fires
shells at 200 m/s. The gunboat’s cannon fires shells at 300 m/s. Find
the angle at which the British should aim their cannon. Show that the
rebels cannot sink the British ship. See Figure (2.13). Answer: 77.08o .

Problem 2.11. A man is at the base of a hill that slopes upward

at a constant angle of 10 ◦ . He aims a rifle up the hill at an angle of
30 ◦ above the horizontal. If the initial speed of the bullet is 300 m/s,
determine how far up the hill the bullet will hit. You may neglect the
fact that the initial vertical position of the bullet is not zero.
 2.9. PROBLEMS 63

300 m/s 1.5 km

200 m/s

2.5 km 1.5 km

Figure 2.13. The British gunboat and the rebel cannon

of Problem 2.10. Do the shells of each one reach the
other?

Problem 2.12. A projectile is launched at an angle θ0 with initial

speed v0 . (a) Obtain an expression for y as a function of x, y = y(x).
(b) Obtain an expression for the speed and direction of the projectile
as a function of its position.
Problem 2.13. David is practicing with his sling before going out
to meet Goliath. The sling is 1.3 meters long and David whirls it in
a horizontal circle over his head at 3 rev/sec. The circular path of
the sling is 2.0 meters above the ground. How far does the stone fly
horizontally before hitting the ground?
Problem 2.14. In projectile problems one usually assumes the
gravitational acceleration is directed straight down. Now assume the
gravitational acceleration is a constant equal to 9.8 m/s2 but pointing
at an angle 30◦ below the horizontal. Obtain an expression for the
range of a projectile launched at 20◦ to the horizontal with an initial
velocity of 300 m/s.
Problem 2.15. Determine the distance traveled along its parabolic
path when a projectile moves from the origin to the top of its trajectory.
It is launched with initial velocity v0 at an angle θ.
Problem 2.16. James Bond want to impress Miss Moneypenny.
They go out to an empty field and Bond fires his pistol at an elevation
angle θ1 , then an instant later, he fires it again at a lower angle θ2 .
The bullets collide in midair. Show that the time interval between the
shots is
2v sin(θ1 − θ2 )
∆t = .
g cos θ1 + cos θ2
64 2. KINEMATICS

Problem 2.17. A projectile fired from y = 0 has reached the top

of its trajectory. Its coordinates at this point are (xt , yt ). Obtain a
formula for the magnitude and direction of the initial velocity in terms
of xt , yt .
Problem 2.18. A particle is moving in a circular path of radius
3 meters with angular speed ω = 10 rad/sec. Place the origin of the
Cartesian coordinate system at the center of the circle.
(a) Obtain expressions for its Cartesian coordinates (x, y) as func-
tions of time.
(b) Express r and v in Cartesian coordinates.
(c) Determine the speed from your expression for v.
Problem 2.19. Assume the orbit of Earth is perfectly circular.
Using tabulated values for the radius of the Earth’s orbit and the ap-
propriate masses, determine the ratios of the centripetal accelerations
of the Earth and the Sun. From this, determine the radius of the Sun’s
orbit around the center of mass of the system.
Problem 2.20. The position (x, y) of a particle (in meters) is given
by
x = 3 sin 5t
y = 3 cos 5t.
(a) Plot the path of the particle. (b) Determine the tangential accel-
eration and (c) the centripetal acceleration of the particle.
Problem 2.21. Consider a “surface skimming” satellite which is in
a circular orbit whose radius is equal to the Earth’s radius. (a) Obtain
an expression for the period of such a satellite in terms of the accelera-
tion and radius. (b) Look up the radius of the Earth and evaluate the
acceleration, speed and period of a surface skimming satellite.
Problem 2.22. A spinning top is rotating at 20 rev/sec about
its axis. The axis itself is oriented at 15◦ from the vertical and is
precessing at 0.5 rev/sec about the vertical. What is the total vector
angular velocity ω? (Express in Cartesian coordinates.)
Problem 2.23. A satellite is in a circular orbit around the Earth.
As shown in Problem
p 2.21 the period, radius, and acceleration are
related by τ = 2π r/a. Furthermore, it is known from Newton’s law
of universal gravitation that the acceleration decreases with distance
from the center of Earth according to
1
a ∝ 2.
r
 2.9. PROBLEMS 65

If the orbital radius increases by a small amount ∆r, the period will
increase by ∆τ. Show that
∆τ 3 ∆r
= .
τ 2 r
Problem 2.24. Prove the triangle inequality which states that for
two vectors A and B, |A + B| ≤ |A| + |B| .
Problem 2.25. Show that A · (B × C) is the volume of a paral-
lelepiped whose edges can be represented by the vectors A, B, C.
Problem 2.26. Evaluate (A × B) · C if A = 3ı̂+2̂+4k̂, B = 3ı̂+
2k̂ and C = 4k̂. Also show that you get the same result by evaluating
A · (B × C).
Problem 2.27. Show that
 
 Ax Ay Az 
 
(A × B) · C =  Bx By Bz  .
 Cx Cy Cz 
Problem 2.28. Prove the “BAC minus CAB” rule:
A × (B × C) = B(A · C) − C(A · B).
Problem 2.29. Derive the law of cosines for triangles, i.e., show
that if the sides of a triangle have lengths a, b, c, and the angle between
a and b is θ, then
c2 = a2 + b2 − 2ab cos θ.
Problem 2.30. The position of a particle is given by r = 2t (me-
ters), θ = 5t (radians). (a) Plot the path of the particle. (b) Obtain
an expression for the speed as a function of time.
Problem 2.31. Suppose the position of a particle as a function of
time is given by r = 1 + sin t and θ = 1 − e−t . Obtain v and a in terms
of r̂ and θ̂.
Problem 2.32. A particle moves in a plane such that r = 1 − cos θ
m/s and θ̇ = 4 rad/s. Determine v and a.
Problem 2.33. A particle moves in a plane in such a way√ that
r = 2 + sin t (m) and the magnitude of the velocity is v = 2 cos t
(m/s). Find a formula for θ = θ(t).
Problem 2.34. For plane polar coordinates, determine
d3 r
.
dt3
66 2. KINEMATICS

Problem 2.35. Let êν be a unit vector parallel to the velocity

vector v. Therefore, the acceleration is given by
   
dv dêν
a= êν + v .
dt dt
(a) Show that (dêν /dt) is a vector perpendicular to êν . (b) Show that
if r is the radius of curvature of the path of the motion, then (dêν /dt)
= v/r. (c) Finally, show that if ên is a unit vector directed towards the
center of curvature on the concave side of the path, that we can write
·v2
ên .
a =at êν + an ên = vêν +
r
Problem 2.36. Derive Equations (2.27) and (2.28).
Problem 2.37. A bug is crawling on the curved surface of a cylin-
der of radius 25 cm. A scientist determines that the bug’s coordinates
vary according to φ = 4.0t and z = 0.3t2 . Write an expression for the
bug’s velocity in cylindrical coordinates and evaluate its speed at time
t = 2 seconds.
Problem 2.38. A particle of mass m moves on the curved sur-
face of a cylinder of radius R. Express the acceleration in cylindrical
coordinates.
Problem 2.39. In cylindrical coordinates, the divergence of a vec-
tor F is given by
1 ∂(ρFρ ) 1 ∂Fφ ∂Fz
∇·F= + + .
ρ ∂ρ ρ ∂φ ∂z
Evaluate ∇ · F for
√
F(ρ, φ, z) = ρρ̂ + z sin φφ̂ + ρz k̂.
Problem 2.40. An object initially at (0,0,1) meters in Cartesian
coordinates, moves in the y direction at a constant speed v. (a) Obtain
an expression for its position as a function of time in spherical coordi-
nates. (b) Determine ṙ and θ̇ as functions of time. (c) Show that the
speed, expressed in spherical coordinates, is constant.
Problem 2.41. A satellite moves with constant speed about the
Earth in a circular orbit of radius a inclined at 90◦ to the equator, so
that the satellite goes over the poles. A system of Cartesian coordinates
fixed in the Earth has its origin at the center of the Earth, the z-axis
passing through the North Pole, and the x-axis in the equatorial plane
and passing through the Greenwich meridian. Obtain expressions for
 2.9. PROBLEMS 67

the position of the satellite (as functions of time) in Cartesian coordi-

nates. The angular speed of the satellite is ω(rad/s) and the rotation
rate of the Earth is Ω (rad/s).
Problem 2.42. Show that the element of arc length in spherical
coordinates is

1/2
ds = dr2 + r2 dθ2 + r2 sin2 θdφ2 .
Problem 2.43. Derive the equations of transformation between
cylindrical and spherical coordinates.
Problem 2.44. Show that the unit vectors of the spherical coor-
dinate system can be expressed in terms of Cartesian coordinates as
follows:
xı̂+y̂ + z k̂
r̂=
(x2 + y 2 + z 2 )1/2
z(xı̂+y̂)− (x2 + y 2 ) k̂
θ̂=
(x2 + y 2 )1/2 (x2 + y 2 + z 2 )1/2
−yı̂+x̂
φ̂=
(x2 + y 2 )1/2
Problem 2.45. An object is moving in such a way that its position
in cylindrical coordinates is given by ρ = a − bt, φ = 21 kt2 , z = 0. (a)
Describe the path. (b) Find the speed as a function of time.

Problem 2.46. A proton with velocity v = aρ̂ + bk̂ enters a uni-

form magnetic field oriented in the +z direction, so that B =B0 k̂. (a)
Describe the subsequent motion of the proton. (b) Otain an expression
for its position as a function of time in cylindrical coordinates.
Problem 2.47. An object moves at constant speed. Prove that
any non-zero acceleration must be perpendicular to the velocity.
Problem 2.48. (a) Show that the component of the acceleration
in the direction of the velocity can always be expressed as
v·a
aq = .
v
(b) Find the component of the acceleration perpendicular to the veloc-
ity.
Computational Projects
68 2. KINEMATICS

Computational Project 2.1. A certain (imaginary) planet is

in orbit about the Sun. It follows an elliptical path that is given by
4
r= ,
1 + 0.2 cos θ
where r is measured in Astronomical Units (AU). Plot the orbit of the
planet in polar coordinates.
Computational Project 2.2. Two cannons are located at
(essentially) the same point. The first cannon fires a projectile at an
initial speed of v1 = 250 m/s at an elevation angle of θ1 = 64.1◦ . After
a time interval ∆t, the second cannon fires its projectile at v2 = 300
m/s and θ2 = 47.3◦ . Determine the time interval ∆t such that the two
projectiles collide in midair. Check your answer against the analytical
result
2v1 v2 sin(θ1 − θ2 )
∆t = .
g v1 cos θ1 + v2 cos θ2
Computational Project 2.3. A volcanic eruption throws a
boulder vertically into the air with an initial speed of 50 m/s. (a)
Write a computer program to plot the position and velocity of the
boulder as a function of time. (b) Now assume the boulder is ejected
at an angle of ten degrees to the vertical. Plot the two components of
position and velocity as functions of time and also plot the trajectory
of the boulder (on an x, y plot). Ignore the effect of air resistance.
Computational Project 2.4. In Worked Example 2.3 we ob-
tained an expression for the launch angle that would give the greatest
range for a cannon mounted on the top of a hill. (a) Validate that
expression by varying the launch angle θ. (b) Determine how the opti-
mum launch varies as the height of the hill is increased from zero.
Computational Project 2.5. Worked Example 2.6 describes a
bead sliding on a helical wire. Plot the trajectory of the bead.
Computational Project 2.6. In a shotput competition an ath-
lete throws a heavy metal ball as far as possible. A champion athlete
can throw the “put” a distance of 20 meters. Assume the ball is thrown
with an initial speed of 13.66 m/s at an angle of 45◦ . If it is thrown
from a height of one meter above the ground, the ball’s initial position
is (0,1). Using the relations
x = x0 + v0x t,
y = y0 + v0y t − (1/2)gt2 ,
obtain x(t) and y(t). Plot the trajectory for y > 0. Also plot trajectories
for angles ranging from 20◦ to 70◦ .
Chapter 3
Newton’s Laws: Determining
the Motion

This chapter (and much of the rest of this book) deals with dy-
namics, that is, the relation between the forces acting on a body and
its motion. A force is an interaction between a body (or particle) and
its environment, usually described as a push or a pull in a specified
direction. In this book you will encounter a number of familiar forces,
such as the gravitational force and forces exerted by springs as well as
a few less familiar ones.
On a very basic level, all forces are manifestations of the “funda-
mental forces,” namely, the gravitational force, the nuclear (or strong)
force, and the electroweak force. The electroweak force is often thought
of as two different forces, the electromagnetic force and the weak nu-
clear force. The electromagnetic force, in turn, is often thought of as
the electric force and the magnetic force.1 All the known forces in
nature are ultimately related to these fundamental forces. For exam-
ple, the force exerted by your muscles can be traced back to electrical
forces.
Dynamics is neatly summarized by Newton’s three laws. You stud-
ied Newton’s laws in your introductory mechanics course. By this time,
you surely know the laws by heart and you know how to use them to
solve reasonably complicated physics problems. However, it is impor-
tant for you as a physicist to have a thorough understanding of these
fundamental statements about the nature of the physical universe. In

1To learn more about the fundamental forces you might read the interesting
article by Charles Seife, “Can the Laws of Physics be Unified?” Science, 309, 82
(2005).
69
70 3. NEWTON’S LAWS: DETERMINING THE MOTION

this chapter I will discuss some aspects of Newton’s Laws that you
may not have considered before. I hope this will give you a greater
appreciation for the scope and significance of the laws.

3.1. Isaac Newton(Optional Historical Note)

In some people’s minds, Sir Isaac Newton was the greatest physicist
who ever lived. Others might give that honor to Albert Einstein, but
no one would deny the tremendous insight and genius of Isaac Newton.
Newton was born in 1642, the year Galileo died.2 Newton’s father
died before Isaac was born, and his mother moved to her brother’s
farm, where the boy grew up. When the time came, his uncle sent
him to Cambridge. While he was studying at Cambridge, a plague
swept through England, and the university authorities sent the students
home for a year. Isaac went back to the farm where he spent his time
thinking about the properties of the physical universe and doing simple
but very clever experiments. After the year was up, the plague had run
its course, and Newton returned to Cambridge. We can only imagine
the scene when he and his major professor met:

Professor: Hello Isaac. Welcome back. Did you spend

your time fruitfully while the University was closed?
Newton: Yes sir, I believe I did.
Professor: Very good, Isaac. Precisely what did you
do?
Newton: Well, sir, I proved the binomial theorem,
invented calculus, designed and built a reflecting tele-
scope, derived the law of universal gravitation, devel-
oped a theory of optics, and determined three fun-
damental laws of nature governing the motion of any
physical object.

According to legend, the professor quit his job and left the position
to Newton.
Newton did not formulate his laws out of the blue. He was very fa-
miliar with the work of his predecessors, especially Galileo and Kepler.
(As he stated, “If I have seen further than others, it is by standing on

2Newton was born on 25 December 1642, according to the Julian calendar

then in use in England. In 1753, England adopted the Gregorian calendar making
Newton’s birth date 4 January 1643.
 3.2. THE LAW OF INERTIA 71

the shoulders of giants.”) In fact, Newton’s first law, the law of iner-
tia, was formulated and demonstrated by Galileo. It is possible that
Galileo got the idea from Rene Descartes. However, the second and
third laws were first formulated by Isaac Newton.3
Although Newton discovered the law of universal gravitation when
he was very young, he did not publish it until many years later because
his early calculations did not give the correct value for the period of
the Moon. Urged by his friend, the architect Sir Christopher Wren,
Newton carried out the calculations more carefully, obtained the right
answer, and convinced himself that his law was correct.
Newton invented the reflecting telescope and carried out a large
number of experimental investigations, particularly in optics. He was
also interested in alchemy and theology. But his greatest contribution
was in formulating physics as an exact mathematical science. His ge-
nius was immediately recognized and he soon became one of the most
famous men in Europe. After Newton, people realized that natural
processes take place in a manner that can be analyzed and predicted.
People began to think of the universe as a “clockwork” mechanism in
which the future development of any system could (in principle) be
determined from a knowledge of its present state. The poet Alexander
Pope, eulogized Newton with the couplet,
Nature and Nature’s Laws lay hid in Night,
God said, Let Newton be! and All was Light.
Personally, Newton was not easy to deal with. It was said of him
that, “He suffered people poorly, and fools not at all.” Since his position
at Cambridge was a semi-clerical post, he was not allowed to marry. It
appears that the life of a bachelor suited him well. He died in 1727 at
the ripe old age of eighty four.4

3.2. The Law of Inertia

The law of inertia, or Newton’s first law,5 can be stated as follows:
3Newton expressed the laws of motion in his famous book, Philosophia Naturalis
Principia Mathematica published in 1687. A fairly recent English version is, The
Principia: Mathematical Principles of Natural Philosophy translated by I. Bernard
Cohen and Anne Whitman, University of California Press, Berkeley, 1999.
4A recent very readable biography is Isaac Newton by James Gleick, Pantheon
Books, New York, 2003.
5This law was first expounded by Galileo. Newton gave Galileo full credit for
it, but over the years we have grown accustomed to calling it “Newton’s First Law.”
If you happen to be a purist, you might want to know Newton’s precise words.
He said, “Every body perseveres in its state of rest, or of uniform motion in a right
72 3. NEWTON’S LAWS: DETERMINING THE MOTION

A body in motion will remain in uniform mo-

tion and a body at rest will remain at rest
unless acted upon by a net external force.
You have certainly thought about this law and you realize that it
says that in the absence of external forces a body at rest will remain
at rest. This is a very profound concept. Like many profound con-
cepts, after you have accepted it, it appears completely obvious. It is
important to note that internal forces cannot change the motion of a
system. If you sit on a chair and pull up on the seat, you remain at
rest no matter how hard you pull. You cannot lift yourself into the air.
The reason is, of course, that your pull is an internal force.
You might object that an exploding bomb is an example of internal
forces causing a system to accelerate. After the explosion, pieces of
the bomb are flying in all directions. This is true. But the center of
mass of the bomb remains at rest. The position of an extended body
is described by the position of its center of mass and if you trace the
trajectories of all the pieces, you will find that the center of mass did
not accelerate.
Perhaps the most amazing thing about the law of inertia is the idea
that a moving body will continue to move even though no external
force is applied to it. This was hard for people in Newton’s time to
understand. It is still hard for some people to understand. They say,
“When I push a trunk across the floor, as soon as I quit pushing, the
trunk quits moving.” Of course, you and I know the trunk quits moving
because there is a force acting on it - the force of friction. Furthermore,
you are familiar with games such as air hockey in which the puck slides
across the table at a (nearly) constant velocity because the frictional
force has (very nearly) been eliminated. You have also been exposed to
countless TV scenes of astronauts floating in space, so you do not have
the conceptual difficulties your great-grandfather had in accepting that
an object, not subjected to any force, will move forever in a straight
line at a constant velocity.
Consider a moving body with no external forces acting on it. Ac-
cording to the first law, this body will move at a constant speed in a
straight line. If you think about it for a moment, you will realize that
this will only be true in certain specific reference frames. For example,
line, unless it is compelled to change that state by forces impressed thereon.” (The
Principia by Isaac Newton, translated by Andrew Motte from the Latin. Pub. D.
Adee, 1848.)
I urge you to express Newton’s laws in your own words. In fact, sometimes I
state them one way and sometimes another. The crucial thing is not the words we
use but the concept expressed by those words.
 3.2. THE LAW OF INERTIA 73

if you are riding in a bus and someone places a tennis ball in the aisle,
it will remain at rest in the aisle as long as the bus is moving at a con-
stant velocity. But if the bus accelerates the ball will roll. If the bus
slows down, the ball rolls forward. If the bus turns left, the ball rolls
to the right. If I were somebody who did not know very much physics,
I might conclude that the ball does not obey Newton’s first law. You
realize, of course, that I am describing the motion of the ball relative
to the bus and the bus is, as I said, accelerating. So you would refute
my statement by pointing out that relative to the ground the ball did
obey the first law. It did maintain a state of uniform motion in the
reference frame of the Earth. I might then point out that the Earth is
also an accelerating reference frame because it is rotating. How would
you respond?
If we follow such arguments to their logical conclusion we will decide
that the first law is only valid in a non-accelerating reference frame.
(This is also true of Newton’s other two laws.) A non-accelerating
frame is called an inertial reference frame. Does such a reference frame
actually exist? Newton stated that his laws were valid in a reference
frame that was “at rest with respect to the fixed stars.” But we know
the stars are all in motion! It is probably best to treat the concept
of an inertial reference frame as a useful idealization. For many prob-
lems, the Earth can be treated as if it were at rest. This approximation
breaks down when the rotation of the Earth must be considered. Then
we usually assume that the inertial reference frame is a nonrotating
reference frame with origin at the center of the Earth. That approx-
imation breaks down if we need to include the orbital motion of the
Earth around the Sun. If necessary, we can take the origin of coordi-
nates to be at the center of the Sun, or even the center of the galaxy.

The physical property called “inertia” is associated with the fact

that a moving body tends to preserve its state of motion. The ex-
pression, “preserving its state of motion” means the body has constant
velocity. A locomotive has more inertia than a ping-pong ball. The
word inertia thus appears to be a synonym for mass. Since it requires a
great force to change the motion of an object with great mass, one often
hears the expression, “Mass is a measure of the inertia of a body.” But
the tendency of a body to maintain its state of motion also depends on
its velocity. It is easier to deflect a slowly moving five gram ping-pong
ball than a speeding five gram bullet. In this case, inertia appears to
be a synonym for momentum.
It is very difficult to define fundamental quantities such as mass,
distance, time, charge, and so on. Similarly, inertia is usually described
74 3. NEWTON’S LAWS: DETERMINING THE MOTION

by the somewhat vague expression that it is the tendency of a body to

maintain its state of motion. (We do not have a formula for inertia!)
On the other hand, the law of inertia (Newton’s First Law) is perfectly
well defined.

3.3. Newton’s Second Law and the Equation of Motion

Newton’s second law can be stated in the form
The rate of change of the momentum of an
isolated body is equal to the net external
force applied to it.
In equation form this is written:

dp
= F, (3.1)
dt
where F is the net or total force acting on the body. (You may prefer
to write it as Fnet or as ΣF.) Now momentum is defined as p =mv, so
if the mass is constant, Newton’s second law takes on the familiar form

F = ma.
Keep in mind that Newton’s second law is applicable only in an inertial
reference frame.
Newton’s second law expressed in the form a = F/m is sometimes
referred to as the “equation of motion.” The reason is that if the forces
are known, we can determine the acceleration of the body, and once
the acceleration has been determined we can integrate to obtain the
“motion,” that is, the velocity and position as functions of time. Ac-
cording to classical mechanics, a knowledge of the position and velocity
of all the particles in a system allow one to determine the forces and
hence to predict the future development of the system.6 Therefore, an
6This is a wonderful fact of nature. As stated by Landau and Lifshitz on the
very first page of their excellent advanced mechanics book, “If all the coordinates
and velocities are simultaneously specified, it is known from experience that the
state of the system is completely determined and that its subsequent motion can,
in principle, be calculated.” (Emphasis added.) They are suggesting that there is
no fundamental reason why the motion of a system depends only on position, ve-
locity and time; we must accept it as an experimental fact. (Reference: Mechanics,
Course in Theoretical Physics, Volume 1, L. D. Landau and E. M. Lifshitz, Perg-
amon Press, New York, 1976. Page 1.) Interestingly, in electrodynamics there is a
quantity called the “Abraham-Lorentz force” that is usually expressed as a function
of acceleration. This force leads to some strange behavior that has been referred to
as “philosophically repugnant” by David Griffiths (Introduction to Electrodynamics,
3rd Ed, Prentice Hall, 1999, page 467).
 3.3. NEWTON’S SECOND LAW AND THE EQUATION OF MOTION 75

expression for the acceleration in terms of the velocity, position and

time is called an equation of motion. Newton’s second law is one way
to obtain the equation of motion. In this course you will also learn
other ways to obtain the equation of motion.
Much of our work in this course will involve obtaining and solving
the equation of motion for a variety of different physical systems. The
equation of motion a = a(x, v, t) = F/m can be obtained directly from
an analytical expression for F. As you will learn in Chapter 4, the
equation of motion can be obtained without explicitly using Newton’s
second law. However, for now, we will stick to the second law.
Although we shall not be particularly concerned with philosoph-
ical questions, it should be mentioned that over the years since the
publication of Newton’s great work, “Principia,” there has been much
debate and speculation on the meaning of the second law. Clearly it is
a statement of how a body reacts when a force is applied to it. If the
mass is constant, the second law tells us that the acceleration of the
body is proportional to the force and inversely proportional to the mass
(a = F/m). From this point of view, mass becomes the proportionality
constant in the relationship a ∝ F. Some thinkers have gone so far as
to state that the second law is just a definition of mass. But if mass
is nothing more than a measure of inertia, one cannot explain the Law
of Universal Gravitation (F = −G (mM/r2 ) r̂) which states that the
gravitational force between two bodies of masses m and M is propor-
tional to the product of the masses of the bodies. The gravitational
law implies that mass is the source of the gravitational force. Perhaps
the “mass” in the gravitational force law is not the same as the “mass”
in Newton’s second law. Perhaps we are using the same word for two
different things. In that case, the question could be resolved by using
different symbols and different names for the two things. Thus, for
example, we could define the quantities mG and mI as:
mG = gravitational mass = the property by which
a body exerts a gravitational force on other bod-
ies, and Newton’s gravitational law would be ex-
pressed as F = −G (mG MG /r2 ) r̂
and
mI = inertial mass = the property by which a body
resists a change in its state of motion, so that New-
ton’s second law would be expressed as F = mI a.
However, many very careful experiments (some of the most famous
were by Baron Eötvos in Hungary over a hundred years ago), showed
76 3. NEWTON’S LAWS: DETERMINING THE MOTION

that mG and mI are equal to within one part in 20 million. This leads
one to conclude that gravitational mass and inertial mass are indeed
the same thing. The fact that inertial mass and gravitational mass are
equal is called “The Principle of Equivalence.” Albert Einstein used
this equivalence as a basic postulate of his theory of General Relativity
(1915).7
Newton’s second law is a cornerstone of classical physics because
it can be used to calculate the acceleration of a body, given the force
acting on it. Once the acceleration of a body is known, the laws of
kinematics determine its velocity and position at any later time. This
means that if you know the net force acting on a body, you can calculate
its position at any future time. The ability to predict the motion is the
power of the second law.
Recall again that dynamics is the study of how a force affects the
motion of a body. As you might suspect, dynamics usually involves
accelerating bodies. However, in some situations there may be forces
acting on a body but nevertheless the acceleration is zero. Consider,
for example, a body acted upon by two equal and opposing forces. The
effects of these forces cancel out and the body does not accelerate. Zero
acceleration and zero velocity is an important special case in dynamics
and is called statics. Statics is of particular interest to civil engineers
who want to make sure the structures they design, such as bridges and
skyscrapers, will have zero acceleration. Statics was treated briefly in
Section 1.5.2 and will be dealt with in greater detail in Chapter 14.
The principle of superposition states that if two or more forces act
on a particle, the net effect is that due to a single force equal to the
vector sum of all the forces. You will be exposed to the principle of
superposition in other areas of physics. For example, the net electric
field at a point is the vector sum of all the electric fields acting at that
point.
In our discussion of Newton’s laws we considered interactions be-
tween particles. When we apply Newton’s second law to an extended
body the acceleration a refers to the acceleration of the center of mass
of the body. A net external force F applied at any point of an extended

7Ifyou are interested in some of the philosophical implications of mass, force,

inertia and Newton’s laws, you might enjoy the three articles by Franck Wilczek
entitled, “Whence the Force of F = ma?” These articles were published in Physics
Today in the issues of December 2004, July 2005, and October 2005.
The article “Drop Test”by Adrian Cho (Science, 6 March, 2015, page 1096)
describes three experiments being performed to measure the equivalence of inertial
mass and gravitational mass to better than one part in ten trillion.
 3.3. NEWTON’S SECOND LAW AND THE EQUATION OF MOTION 77

rigid body will cause the center of mass of the body to accelerate ac-
cording to a = F/m. This is often not at all intuitive. For example,
consider a tricycle with the pedals in a vertical position as shown in
Figure 3.1. If you pull forward on the top pedal, the bicycle will ob-
viously move forward. But what happens if you pull forward on the
bottom pedal? It might seem that this would propel the tricycle back-
ward, but Newton’s second law tells us that if we pull forward, the
tricycle moves forward. (If you do not believe me, try it! You will be
surprised by the motion of the pedal itself. This is not at all the same
as what would happen if you were riding the tricycle. In that case, the
force your foot exerts on the pedal is an internal force.)

Figure 3.1. What direction does the tricycle move

when you pull on the bottom pedal as shown?

Exercise 3.1. A box of mass 5 kg is placed on an inclined plane

of angle 35◦ . A force of 10 N parallel to the plane in a direction up the
plane is applied to the box. The coefficient of sliding friction is 0.07.
Determine the acceleration of the box. Answer: -3.05 m/s2 (down the
plane).

Exercise 3.2. A crate of mass 50 kg is sitting on the flat bed of

a truck. The truck accelerates at 0.2 m/s2 . The coefficient of static
friction is 0.15. Does the crate slide? Answer: No.

Exercise 3.3. A spool of thread is lying on its side on a table and

is free to roll. You pull on the free end of the string. Does the spool
move towards you? Does it matter whether the string comes from the
bottom side of the spool or the top side? (Try it!)
78 3. NEWTON’S LAWS: DETERMINING THE MOTION

3.4. Newton’s Third Law: Action Equals Reaction

Using Isaac Newton’s terminology of “action” and “reaction”, we
can formulate his third law as follows:
To every action there is always opposed an equal reaction.
In different words, if one body exerts a force on a second body, the
second body exerts an equal but oppositely directed force of the same
kind on the first body. In applying the third law, it is important to
remember that the two forces involved act on different bodies.
The third law states that the force one body exerts on another
body (the “action”) is equal and opposite to the force the second body
exerts on the first (the “reaction”). The law does not, however, state
that these forces must lie along the same line. Thus, in Figure 3.2 we
see that in both case (a) and case (b) the forces are equal and opposite,
but in case (b) the forces do not act along the same line. When the
action-reaction forces lie along the same line we say the third law is
obeyed in its strong form. Otherwise, we say it is obeyed in the weak
form.

(a) Strong Form (b) Weak Form

Figure 3.2. Illustrating the strong form and the weak

form of Newton’s third law. The arrows represent the
forces acting on the particles. In both cases the forces
are equal and opposite, but in the strong form the forces
act along the line joining the particles.

Newton’s third law is intimately related to the law of conservation of

linear momentum. Consider, for example, an isolated system consisting
of two particles that exert equal and opposite forces on each other:
F1 = −F2 .
By the second law, the forces can be expressed as changes in the mo-
mentum of the particles:
dp1 dp2
=− .
dt dt
 3.4. NEWTON’S THIRD LAW: ACTION EQUALS REACTION 79

Therefore,
d
(p1 + p2 ) = 0.
dt
That is, the total momentum of an isolated system is constant.
Is the third law always obeyed? This is a question that physicists
have debated for many years. We can state unequivocally that the third
law is always obeyed in purely mechanical systems, but the situation
is not quite as clear when we consider the electromagnetic interaction
between charged particles, as illustrated by the following example.

Worked Example 3.1. Consider the action-reaction pair of

forces for two charged particles interacting through the magnetic
force. See Figure 3.3. Is the third law obeyed?
Solution: Figure 3.3 shows two moving charged particles. The
problem is to determine if the forces they exert on each other are
equal and opposite.
You remember from your introductory physics course in elec-
tricity and magnetism that the force on a charge q1 moving at
velocity v1 in a magnetic field B12 is
F = q1 v1 ×B12 .
The magnetic field B12 acting on q1 is due to the motion of charge
q2 and is given bya
B12 = (q2 v2 × r)(µ0 /4πr3 )
where r is the vector from q2 to q1 .
Now it is easy to see that B12 6= 0, but B21 = 0 because v1 ⊥ r.
Therefore, the force on q1 is non-zero, but the force on q2 is zero.
We might therefore conclude that the third law has failed! This
result is actually cited in many physics texts as evidence that the
third law is not universal and does not apply to electromagnetic
forces.
But wait! Our conclusion was reached too hastily! Upon further
(and much deeper) analysis one finds that the two moving charged
particles set up electromagnetic fields that have momentum and
energy associated with them. When one takes into consideration
the forces acting on the particles due to the rate of change of elec-
tromagnetic momentum, then the third law is found to hold. Fur-
thermore, if the charges are part of electric currents in two nearby
circuits, it is easy to show that the forces exerted by the circuits
on one another are equal and opposite. Consequently, it is safe to
80 3. NEWTON’S LAWS: DETERMINING THE MOTION

conclude that the third law has the same general range of validity
as the first two.b
aThis relation is based on the Biot-Savart law which, strictly speaking, is not
valid for point charges. However, the correct expressions reduce to the given
formula as long as the velocity of the particle is much less than the speed of
light. (D. J. Griffiths, Introduction to Electrodynamics, 3rd Ed., Prentice Hall,
1999, p. 439.
bRoald K. Wangsness, Electromagnetic Fields, 2nd Ed., Wiley and Sons, New
York, 1986, pages 219 and 359. For a different point of view, see “Classical
Dynamics” by Jerry Marion and Stephen Thornton, 3rd Ed., Harcourt, Brace,
Jovanovich, 1988, page 45.

q1 q2
r12
+ +
v1
v
2

Figure 3.3. Two moving charged bodies give rise to an

apparent counter example to Newton’s third law. For
this arrangement v1 × r12 = 0 whereas v2 × r21 6= 0.

The point of the preceding example is that you, as a physicist, must

be very careful in applying the third law. As an illustration, let me
give you a puzzle. You may have heard it before. It is the story of a
donkey hitched to a cart. The donkey pulls on the cart. The cart pulls
back on the donkey. By Newton’s third law, action equals reaction, and
these two forces are equal and opposite. Thus it would appear that the
forces cancel out. How, then, is it possible for the cart to move?
Give up? Well, the answer is that the forces do not cancel because
they are acting on different bodies! Suppose a (single) body is acted
upon by two forces of equal magnitude but opposite directions. You
can then say that these forces cancel. But when you are adding up
the forces that are acting on a particular body, you certainly cannot
include forces that are acting on a different body! You would never
say the force the Sun exerts on the Earth is cancelled by the force the
Earth exerts on the Sun. These forces are indeed equal and opposite,
 3.5. IS ROTATION ABSOLUTE OR RELATIVE? 81

but they act on different bodies. Similarly, for the cart and donkey
problem, the donkey exerts a force on the cart and the cart exerts an
equal and opposite force on the donkey. Considering only the cart, the
forces on the cart are the force exerted by the donkey and frictional
forces (the road on the wheels, etc.). If the force exerted by the donkey
is greater than the frictional force, the cart accelerates forward. You
should also think about the forces acting on the donkey and explain
why the donkey is accelerating forward and not backward. Note that
the only forces that can make an object accelerate are external forces.
You cannot pick yourself up by pulling on your shoestrings!
In determining action-reaction pairs, it is useful to remember that
they are always the same kind of force. For example, a book on a
tabletop is pulled downwards by the gravitational attraction of the
Earth. It is easy to make the error that the reaction is the upward
normal force of the table. But this is not the same kind of force. The
reaction is the upward gravitational pull of the book on the Earth.
What is the reaction force to the normal force exerted by the table?

3.5. Is Rotation Absolute or Relative?

Linear motion is relative. A person in a moving bus is at rest
relative to the bus, but has non-zero velocity relative to the ground.
The velocity of an object depends on (is relative to) the reference frame
in which it is viewed.
Galileo suggested that a person in the hold of ship sailing in per-
fectly smooth water would not be able to determine whether or not the
ship was moving. According to Newton (and later Einstein) it is impos-
sible to distinguish between one inertial reference frame and another.
If you were to place a TV camera inside a closed box, there would be no
physical phenomenon you could observe and no experiment that you
could carry out that would tell you whether the box was at rest, or
moving at a constant velocity.
Is rotational motion also relative? Newton pondered this question
and decided that rotation is absolute. He described an experiment in
which he suspended a bucket full of water from a long rope. He turned
the bucket, twisting the rope. Upon releasing the bucket, he observed
the water level while the rope unwound and the bucket rotated. He
noted that initially, when the bucket and water were at rest, the surface
of the water was flat. Next, the bucket began to rotate, but the water
was still at rest (for a while) and the surface of the water was still flat.
Finally, the water took on the rotation of the bucket and the surface
of the water became concave. (I am sure none of this surprises you.)
82 3. NEWTON’S LAWS: DETERMINING THE MOTION

Now in the initial and final stages, the water was not moving relative
to the bucket, but the surface was flat in the initial stage and concave
in the final stage. Therefore, the behavior of the surface was not due to
the motion of the water relative to the bucket. Also, when the bucket
was rotating but the water was not rotating, the surface remained
flat. Newton concluded that the concavity of the surface was due to
the absolute rotation of the water and not its motion relative to the
bucket.
If you placed the bucket of water in a closed box and set the box
rotating, a TV camera inside the box would show the surface of the
water to be concave. Thus there is a physical phenomenon that can be
used to determine whether or not a reference frame is rotating. There
is no other reference frame (Newton decided) in which the surface of
the water is flat, so rotation is not relative.8
Ernst Mach was fascinated by this simple experiment. However,
he believed that all motion was relative, including rotational motion.
Mach claimed that rotation was relative to all the mass in the universe.
He reasoned that a bucket of water at rest would have a flat surface,
just as Newton observed. But, he asked, would the surface be curved
if the bucket were spun in a totally empty universe? In other words,
what would the bucket be spinning relative to? From Mach’s point
of view, it is the rest of the universe that causes the surface to be
curved, and the same effect would be obtained by spinning the entire
universe around a stationary bucket. You cannot determine whether
the water is rotating and the universe is at rest, or the water is at
rest and the whole universe is rotating around it! Einstein’s theory of
general relativity does not include this “total relativity” of Mach, but
later in his life Einstein tried to analyze the consequences of imposing
it on his theory.
Although the question of whether or not rotation is relative or abso-
lute is still being debated,9 we shall go along with Newton and assume
that rotation is absolute. This point of view is supported by a re-
cent in-depth study10 of Mach’s principle that shows there are many
effects that cannot be explained by a rotating universe and a stationary
8Newton stated that centrifugal force is the feature that distinguishes absolute
motion from relative motion.
9An interesting article on this subject is “Total Relativity: Mach 2004” by
Frank Wilczek, Physics Today, April 2004, page 10. Newton’s bucket is also used
as a starting point for the very readable book, The Fabric of the Cosmos: Space,
Time and the Texture of Reality by Brian Greene, Knopf, New York, 2004.
10H. Hartman and C. Nissim-Sabat, “On Mach’s critique of Newton and Coper-
nicus,” Am. J. Physics, 71, 1163 (2003).
 3.6. DETERMINING THE MOTION 83

bucket. For example, Mach’s principle cannot explain how two buck-
ets, rotating in opposite directions, can both have concave surfaces.
Finally, we might note that in Quantum Mechanics we find that rota-
tional motion is quantized whereas linear motion is not. A body can
have any linear velocity whatever, but a body can only have certain
discrete values for rotational motion.

3.6. Determining the Motion

Recall that “determining the motion” means obtaining an equation
for the position of the body as a function of time. This is done by
solving the equation of motion (which could simply be Newton’s second
law). The equation of motion is a second order differential equation
for the position, so solving for the motion involves carrying out two
integrations. In this section you will learn how to obtain expressions
for the velocity and position of a particle subjected to several different
kinds of force. Specifically, we will consider the following situations:
Force is constant: F = const
Force is a function of time: F = F (t)
Force is a function of velocity: F = F (v)
Force is a function of position: F = F (x)
Unless otherwise noted (and to keep things simple) the motion will
be one dimensional and the object that is moving will be a particle of
constant mass. The generalization to two or three dimensions is quite
straightforward.
Constant Force
If the force is constant, from F = ma we appreciate that the accel-
eration is constant. The equation of motion is is
ẍ = F/m = constant = a,
and integrating dv = adt yields
v(t) = v0 + at. (3.2)
Integrating again we obtain
1
x(t) = x0 + v0 t + at2 . (3.3)
2
Equation (3.3) giving x = x(t) is the solution of the equation of
motion. Thus, we have “determined the motion.”
You learned Equations (3.2) and (3.3) in your introductory physics
course. Perhaps you did not fully appreciate at that time that these
84 3. NEWTON’S LAWS: DETERMINING THE MOTION

equations are valid only for a constant force. The rest of this chapter
deals with forces that are not constant.

Exercise 3.4. It is observed that the position of an object of mass

3 kg is given by x(t) = 3t + 6t2 meters. Determine the force acting on
it. Answer: 36 N.

Force as a Function of Time

A somewhat more complicated situation arises when the force F can
be expressed as a function of time, i.e., F = F (t). Then the equation
of motion is

F (t)
a= .
m
or
dv 1
= F (t).
dt m
Separating variables and integrating:
Z v
1 t
Z
dv = F (t)dt.
v0 m 0
In writing this last equation I was careless and used the same symbol (v)
for a variable of integration and a limit of integration. (This is the kind
of thing that drives mathematicians crazy!) You can usually get away
with this sloppy notation, but sometimes it can get you into trouble.
So I will rewrite the equation above and use double primes for the
variables of integration and single primes for the limits of integration.
Z v0 =v(t0 ) Z 0
00 1 t
dv = F (t00 )dt00 .
v0 m 0

Integrating and rearranging a bit gives

Z 0
0 1 t
v(t ) = v0 + F (t00 )dt00 . (3.4)
m 0
To evaluate the integral in Equation (3.4) you need an explicit ex-
pression for the force as a function of time. Suppose that you do have
such an expression. Then integrating once again you obtain the posi-
tion as a function of time. Since v(t0 ) is given by Equation (3.4) you
 3.6. DETERMINING THE MOTION 85

can substitute and integrate to obtain

Z x(t0 ) Z t" Z 0 #
1 t
dx = v0 + F (t00 )dt00 dt0 .
x0 0 m 0

So,
Z t "Z t0
#
1
x(t) = x0 + v0 t + F (t00 )dt00 dt0 .
m 0 0
Although this expression looks rather forbidding, the procedure for
1
R
finding x = x(t) is not at all difficult: you simply integrate Rm F (t)dt
to get the velocity as a function of time and then integrate v(t)dt to
obtain the position as a function of time (and don’t worry about all
those primes and double primes).

Worked Example 3.2. A particle of mass m is acted upon

by a force F (t) = Ae−bt . The particle is initially at rest at x = 0.
Determine the velocity and position of the particle at time t = 1/b.
Solution: From F = ma
dv
a = F/m = (A/m)e−bt =
dt
so
Z t t
−bt A −bt  A  −bt 
v − v0 = (A/m)e dt = − e  = − e −1
0 mb 0 mb
A
where v0 = 0. For t = 1/b the velocity is v = mb (1 − 1e )
To determine the position we must use the general expression
for the velocity, not its final value. That is,
Z t=1/b Z t=1/b
A  −bt 
x − x0 = vdt = − e − 1 dt
0 0 mb
 1/b
A −bt A
= 2
e + t
mb mb 0
A −1 A 1 A
= e + −
mb2 mb b mb2
A 1
= ,
mb2 e
where x0 = 0.

Exercise 3.5. A force F = 3t2 N acts on a particle for 3 s after

which the particle moves freely. The particle is initially at the origin
86 3. NEWTON’S LAWS: DETERMINING THE MOTION

with zero velocity. Determine its position at time t=5 s. Assume the
mass of the particle is 0.1 kg. Answer: 742.5 m.
Exercise 3.6. A certain electromagnetic wave has its electric field
oriented along the z-axis. The magnitude of the field varies in time
according to E = E0 cos ωt. Initially, an electron is at rest at the
origin. Determine the motion of the electron. (Recall that the force
exerted on a charged body in an electric field is given by F = qE.)
Answer: x = −(qE0 /ω 2 m)(cos ωt − 1).

Force as a Function of Velocity

In nature many forces depend on the velocity of the body. For
example, a particle moving through a fluid, such as a marble falling
through water or a baseball thrown through the air, will experience a
resistive force. This force can usually be assumed to be proportional to
the velocity for small speeds and proportional to the velocity squared
for higher speeds. (Of course, this is just an approximation; the actual
dependence of the resistive force on the velocity is more complicated.
See, for example, the fluid mechanics books by Batchelor11 or by Lan-
dau and Lifshitz.12) Another velocity dependent force is the Lorentz
force. This is the force on a charged particle moving with velocity v in
a region of space where there is an electric field E and a magnetic field
B. You probably remember from your introductory course in electric-
ity and magnetism that the Lorentz force is given by F = q(E+v × B).
(The magnetic force was considered in Worked Example 3.1 in connec-
tion with Newton’s third law. It presents us with a rather complicated
situation because it depends on the direction of the velocity as well as
on its magnitude. In this section we will only consider forces that act
in a constant direction.)
Consider the one dimensional problem of determining the motion of
a particle if the force F is a function of the magnitude of the velocity.
That is,
F = F (v).
The equation of motion is
dv F (v)
= .
dt m
11G. K. Batchelor, An Introduction to Fluid Dynamics, Cambridge University
Press, 1970.
12L. D. Landau, and E. M. Lifshitz, Fluid Mechanics, Vol 6 of Course of The-
oretical Physics, Pergamon Press, Oxford, 1959.
 3.6. DETERMINING THE MOTION 87

Separating variables and integrating:

Z v(t)
1 t
Z
dv
= dt.
v0 F (v) m 0
So,
Z v(t)
dv
t=m .
v0 F (v)
Given an explicit expression for F (v) you can carry out the integration
and obtain an equation for t involving v and v0 (as well as some other
parameters such as the mass and any other constants that appear in
the expression for the force). You have now obtained an expression of
the form t = t(v). This can usually be inverted to yield the desired
form, v = v(t). To be specific, after inverting you will have
v = v(t, v0 , α)
where α is the set of constant parameters mentioned above. You then
use the definition
dx
v= ,
dt
and integrate again to obtain the position as a function of time. That
is,
Z x(t) Z t
dx = v(t, v0 , α)dt,
x0 0
or Z t
x(t) = x0 + v(t, v0 , α)dt.
0
This procedure is a bit complicated so I will give you an example.
Please go through the steps carefully.

Worked Example 3.3. Imagine you are paddling a canoe.

When you appproach the dock you quit paddling and let the re-
sistance of the water bring you to a stop. Assume this force is
proportional to the first power of the speed. Determine the mo-
tion.
Solution: By Newton’s second law, F = m dv dt
. According to
the assumption, F = −bv where b is the constant of proportionality
and the negative sign indicates that it is a retarding force. Then,
−bv = m dvdt
, or dv
v
= − mb dt. Hence,
Z v(t)
b t
Z
dv
=− dt,
v0 v m 0
88 3. NEWTON’S LAWS: DETERMINING THE MOTION

and
v(t) bt
ln v|v(t)
v0 = ln =− ,
v0 m
or
v(t) = v0 e−bt/m .
Thus we have obtained an expression for the velocity at any given
time; half of our job is finished. Integrating again we will obtain
the position as a function of time. Starting with the definition of
velocity, v = dx
dt
, we write
dx
= v(t) = v0 e−bt/m .
dt
Consequently,
Z x(t) Z t
dx = v0 e−bt/m dt.
x0 0
Integrating,
h vm it v0 m −bt/m
0 −bt/m
x(t) − x0 = − e =− (e − e0 ),
b 0 b
or
v0 m
x(t) = x0 + (1 − e−bt/m ).
b
It is interesting to consider the limiting value of our result by letting
t → ∞.13 This leads to the rather un-intuitive result for the velocity of
the canoe because it indicates that the velocity of the canoe [v0 e−bt/m ]
does not go to zero until t → ∞. The canoe is in motion for an infinite
time! How far do you suppose it will travel in this infinite amount
of time? To answer this question consider the expression for position
and set t = ∞ to obtain x(t = ∞) = x0 + v0 b/m. Note that this is
a finite distance. So, even though it takes an infinitely long time for
the canoe to stop, it only goes a finite distance. Does this result make
sense? Yes, it does. The retarding force gets smaller and smaller as
the velocity decreases. Perhaps you might complain that the problem
is poorly posed and it should ask for the time required for the velocity
to fall below some particular value. For example, once the canoe is
moving at some very small velocity, such as one inch per hour, then for
all intents and purposes it is stopped.
If an object is moving in air at less than about 20 m/s (≈ 45 mph),
it is safe to assume the retarding force is proportional to the first power
of the speed. For objects moving at higher speeds (but less than the

13It
is a good idea to subject your answers to a “sanity check” by seeing how
they behave in limiting cases such as t = 0 and t = ∞.
 3.6. DETERMINING THE MOTION 89

speed of sound) it is more realistic to assume the resistive force is

proportional to the speed squared. That is, F = −Dv 2 . The equation
of motion can then be written in the form
dv
m = −Dv 2 .
dt
The proportionality constant D depends on the size and shape of the
body and the density of the fluid through which it is moving. A rea-
sonable formula for calculating D is
1
D = CD Aρ,
2
where the “drag coefficient” CD is a unitless parameter of the order
unity which depends on the shape of the body. In practical applications
the value CD = 0.2 is often used. The quantity ρ is the air density, and
A is the cross-sectional area of the object.
If you apply this law to a body falling through air under the action of
the force of gravity, the resistive force is upwards and the gravitational
force is downwards, so the equation of motion becomes
dv
m = +Dv 2 − mg.
dt
In applying this last equation you will have to be very careful with
the signs. If the body is rising, then both gravity and the force of air
resistance are acting downwards.
Worked Example 3.4. A particle of mass m is acted upon by
the gravitational force (−mg) and a retarding force given by bv 2 .
It is dropped from rest at an initial height x0 . Obtain expressions
for its velocity and position as a function of time.
Solution In problems such as this one, it is often convenient to
express relations in terms of the “terminal velocity” which is the
speed when the two forces are equal and opposite and the object
is no longer accelerating. In this problem the net force is
F = mg − bv 2
where we let “down” be positive. The terminal velocity is obtained
by setting F = 0, so p
vT = mg/b.
Now we solve for the velocity as a function of time using
dv F b
a= = = g − v2
dt m m
90 3. NEWTON’S LAWS: DETERMINING THE MOTION

or
dv b gm b
= ( − v 2 ) = (vT2 − v 2 )
dt m b m
Therefore,
Z v(t)
b t
Z
dv
2 2
= dt
v0 =0 vT − v m 0
which yields
1 v b
tanh−1 = t
vT vT m
and consequently,
r
vT bt bg
v(t) = vT tanh = vT tanh t.
m m
The position is given by
Z t Z t r
bg
x(t) − x0 = − vdt = vT tanh tdt,
0 0 m
where the minus sign is due to the decrease in height of the object
with time. A udu substitution leads to
Z √ gb
m u= m t
x(t) − x0 = − tanh udu
b u=0
and finally
m gb
x(t) = x0 − log cosh t.
b m

Exercise 3.7. A body is dropped from rest from a hot air balloon.
Make the unrealistic assumption that the resistive force is proportional
at all times to the velocity, i.e., fR = −bv. (a) What is its terminal
velocity vT ? (b)How much time is required for the body to reach a
velocity of 0.9 vT ? Answer: (b) 2.3m/b.
Exercise 3.8. A body is falling under the effects of gravity and a
retarding force proportional to the
p square of the speed. Determine its
terminal velocity. Answer: vT = gm/b

Force as a Function of Position

The most important forces you will encounter in this course are
forces that depend on position. These forces are most easily treated
using conservation of energy, and you will do that later. However, for
 3.6. DETERMINING THE MOTION 91

the sake of completeness, let us obtain the motion by a straightforward

integration of the equation of motion.
If F = F (x), the second law is
d2 x
m = F (x).
dt2
This differential equation is not hard to solve if you first separate the
variables. Use the chain rule to write
d2 x dv dv dx dv
2
= = = v. (3.5)
dt dt dx dt dx
Therefore,
dv 1
v = F (x),
dx m
or
1
vdv = F (x)dx.
m
Having separated the variables, you can now integrate to obtain
1 x
Z
1 2 v
2
v |v0 = F (x)dx. (3.6)
m x0
You will need an explicit expression for F (x) to carry out the integral
on the right. A bit of algebra will then yield an expression for the
velocity as a function of position
v = v(x). (3.7)
But what you really want is an expression for the position as a function
of time, x = x(t). Writing dx/dt for v in Equation (3.7) and rearranging
generates a differential equation involving x and t, namely,
dx
= dt.
v(x)
Therefore, Z x Z t
dx
= dt.
x0 v(x) 0
That is, Z x
dx
t= .
x0 v(x)
Assuming this integral can be carried out, you will obtain an expression
of the form t = t(x). Finally, this must be inverted to yield x = x(t),
a mathematical procedure that may involve a significant amount of
algebra.
Simple Harmonic Motion: A particle that oscillates back and
forth periodically is undergoing simple harmonic motion, (SHM).
92 3. NEWTON’S LAWS: DETERMINING THE MOTION

This particularly important type of motion is generated by a force that

is a function of position. Specifically, the force that leads to SHM is a
force that is always directed toward a fixed point and is proportional
to the distance from the body to the fixed point.

k
m

Figure 3.4. A mass m connected to a massless spring

of constant k on a frictionless surface.

Worked Example 3.5. A prime example of simple harmonic

motion is the motion of a block of mass m attached to a spring of
force constant k, and sliding on a frictionless horizontal surface, as
illustrated in Figure 3.4. Obtain and solve the equation of motion.
Interpret the expression for position as a function of time.
Solution: The force exerted by the spring is F = −kx where
x is the amount the spring has been stretched. This is equal to
the displacement of the block from the equilibrium position. The
equation of motion is
mẍ = −kx. (3.8)
Going through the steps from Equation (3.5) to (3.6) you obtain
2 x
Z
2 2
v = v0 + (−kx)dx.
m x0
k
∴ v 2 = v02 − (x2 − x20 )
m
or r
k
v = v02 − (x2 − x20 ).
m
2 k 2
Note that v0 + m x0 is a constant whose value depends on the initial
conditions. Denoting this constant by C you can write
r
dx k
v= = C − x2 .
dt m
 3.6. DETERMINING THE MOTION 93

Separate variables again and integrate to obtain the motion.

Z x Z t
dx
q = dt,
k 2
x0 C−m x 0

or r
x
k t
Z Z
dx
p = dt.
x0 mC/k − x2 m 0
The integral on the left can be found in any table of integrals,
yielding x r t
x  k 
sin−1 p  = t .

mC/k  m 
x0 0
Therefore
r
−1 x −1 x0 k
sin p − sin p = t.
mC/k mC/k m
The second term on the left is just a constant; call it β. Then
r
−1 x k
sin p = t + β,
mC/k m
or r r !
mC k
x= sin t+β .
k m
This is usually written in the easily remembered form
x = A sin (ωt + β) . (3.9)
The constant r r
mC mv02 + kx20
A= =
k k
is the amplitude. The quantity
p
ω = k/m
is the (angular) frequency of the oscillation. The constant β is
called the “phase constant” and is related to the position of the
oscillator at time t = 0. By adding π/2 to β one can express x(t)
in terms of the cosine rather than the sine.

Exercise 3.9. Verify Equation (3.9) by direct substitution into the

equation of motion.
94 3. NEWTON’S LAWS: DETERMINING THE MOTION

3.7. Numerical Method to Determine the Motion (Optional)

3.7.1. Closed Form Solutions. It is usually fairly easy to deter-
mine the forces acting on a body and to write the equation of motion.
However, it is not always possible to solve the equation of motion in
closed form. That is, it is not always possible to write down an ana-
lytical solution in which the position is expressed as a function of time
in terms of simple functions. For example, x = 3t2 , and x = 4 cos 5t
are analytical solutions. However, there are ways to describe the mo-
tion that do not involve analytical expressions. For example, you might
measure the position of an object in the laboratory and draw up a table
giving its position at various times. This would give you a collection
of numbers that you could use to determine the position of the body
at any given time, but you would not have a solution in closed form.
Similarly, a graph of x vs t is a visual representation of position as a
function of time, but once again it is not in closed form. A closed form
(or analytical solution) is a relationship between x and t in terms of
elementary mathematical functions, namely powers, roots, logarithms,
exponentials, and trigonometric functions.
If you have the equation of motion for a system and you know the
initial conditions, you can always write a computer program that will
yield a “numerical solution” giving the position and velocity of the
body at any desired time. With the advent of large, fast computers,
numerical solutions have become very common in physics. In fact, there
is a whole branch of physics called “computational physics” in which
you use a computer to solve physics problems by applying numerical
methods.14 Many important problems cannot be solved analytically; in
recent years more and more work has gone into developing sophisticated
techniques for obtaining numerical solutions using computers.
You may be wondering why we bother to study the techniques for
obtaining analytical solutions instead of just writing a computer pro-
gram to solve the problem. There are many reasons. For one thing,
numerical techniques for solving physics problems are often based on
the analytic solution of a similar (usually simpler) problem. Further-
more, numerical techniques are nearly always verified by determining
whether or not they can reproduce a known analytic solution. Knowing
analytical techniques for solving the equations of motion is extremely
14An excellent computational physics textbook is “Numerical Methods for
Physics, 2nd Edition, ” by Alejandro L. Garcia, CreateSpace Publishing 2015.
 3.7. NUMERICAL METHOD TO DETERMINE THE MOTION (OPTIONAL) 95

helpful in obtaining and interpreting numerical solutions. It might also

be mentioned that numerical solutions have a number of disadvantages.
A numerical solution, as the name implies, is just a number and not a
formula. Numerical solutions yield a specific answer to the problem at
hand and very rarely lead to general formulas.
In this book, I emphasize obtaining analytical solutions for prob-
lems that have closed form solutions. You will find that the techniques
developed here will be used frequently in your other physics courses
and in your professional career. If you end up working as a Computa-
tional Physicist you will often apply the material you are learning in
this course.
You have seen that, in principle, the equation of motion can be
solved analytically for a number of different types of forces, specifically
for forces that are functions of time, velocity or position.
All of the procedures described above basically boil down to in-
tegrating the equation of motion twice, leading to expressions of the
form:

v = v(t, v0 , x0 ),
x = x(t, v0 , x0 ).

These give the position and velocity as functions of time and the ini-
tial conditions, v0 and x0 . If you are lucky, these expressions are in
closed form, that is, in terms of familiar functions such as logarithms,
exponentials, trigonometric functions, etc. (You are comfortable with
these so-called “elementary functions,” but if you were confronted with
an expression involving the gamma function or an elliptic integral, you
might not feel so happy.)
If v(t) and x(t) are given in closed form, you have a great deal
of knowledge about the motion. First of all, you can determine the
velocity and position at any time by simply plugging the time into
the equation. Furthermore, you can simply look at the equation and
get a very good idea about the motion. For example, if you obtain
the solution x = A cos ωt you immediately know that the motion is
oscillatory with amplitude A and angular frequency ω. Generally, it is
easy to manipulate the solution to obtain more information about the
system. For example, squaring the expression for the velocity to get v 2
immediately leads to an equation for the kinetic energy as a function
of time. If the motion is two-dimensional and you obtain solutions x(t)
and y(t) you can easily combine them to get an equation for the path
(or orbit) of the body.
96 3. NEWTON’S LAWS: DETERMINING THE MOTION

A closed form analytic solution of the equation of motion contains

all the essential physical information about the system. (In this sense,
it is similar to the quantum mechanical wave function from which one
can extract all knowledge of the system.) However, it often happens
that such a solution is not available. Perhaps we are not able to solve
the equation of motion due to a lack of mathematical knowledge, or
perhaps the equation of motion is actually insolvable. In such a case,
we are forced to use a different technique, namely, we are forced to
evaluate a numerical solution.
Since numerical solutions are seldom as useful as analytical solu-
tions, you should make sure that you cannot find an analytical solution
before you start writing a computer program. An analytical solution
can usually be found in much less time than it takes to write, debug,
and run a computer program. With this warning out of the way, I
will now describe an elementary technique for obtaining a numerical
solution to the equation of motion.15
A numerical solution of a differential equation essentially involves
replacing the differential equation with a difference equation. For ex-
ample, suppose you wanted to solve
dx
= f (x, y),
dy
You would replace this expression, in which dx and dy are infinitesimal
quantities, with the expression
∆x
= f (x, y).
∆y
Here ∆x and ∆y are small quantities but are not infinitesimal. To
generate a solution recall the definition of derivative:
dx x(y + ∆y) − x(y)
= lim .
dy ∆y→0 ∆y
15Physicists frequently use numerical techniques for solving differential equa-
tions which are difficult or impossible to solve analytically. This is especially true if
they are analyzing fluid flow. Fluid dynamics is, of course, a branch of Newtonian
mechanics and the flow of a fluid is controlled by Newton’s laws. However, the
forces involved are very complicated. Some are velocity dependent quantities that
depend on the pressure exerted by other parts of the fluid. Other forces include
viscous drags and the force of gravity. When one writes Newton’s second law for a
fluid, it may have a large number of terms in it. Often, the only practical way to
solve this equation and thus determine the motion of the fluid is to use numerical
techniques and a high speed computer. (The second law, when applied to fluid
flow, is called the Navier-Stokes equation.)
 3.7. NUMERICAL METHOD TO DETERMINE THE MOTION (OPTIONAL) 97

Replace dx/dy = f (x, y) by the approximate relation

x(y + ∆y) − x(y)

' f (x, y),
∆y

and rearrange to get

x(y + ∆y) ' x(y) + f (x, y)∆y.

If you know x and y and can calculate f (x, y) you can use this rela-
tionship to determine x for a slightly larger value of y, namely, y + ∆y.
Note that the relationship is not exact. However, the approximation
becomes better and better as ∆y becomes smaller and smaller.
Let us apply this technique to the equation of motion. We want
to solve for x(t) numerically, given an initial value x0 and an explicit
expression for the acceleration, a = a(x, v, t). The procedure involves
carrying out the following sequence of operations:
1. Select ∆t to be a small time step.
2. Set x = x0 and t = t0 .
3. Determine a = a(x, v, t).
4. Obtain new time from t = t + ∆t.
5. Obtain new velocity from v = v + a · ∆t.
6. Obtain new position from x = x + v · ∆t.
7. Go to Step 3.
You go through this procedure repeatedly. At each step you obtain
new values for x, v, and t.
Consider a simple but important numerical technique called the
Euler method. This is a technique for obtaining the numerical solution
of an equation of the form

d2 x
= f (x, v, t).
dt2
If the function on the right-hand side is the force divided by the mass,
this is just Newton’s second law.
Write the second-order differential equation above as two first-order
equations. These are,

dx
= v(x, t),
dt
dv
= f (x, v, t).
dt
98 3. NEWTON’S LAWS: DETERMINING THE MOTION

Solve as follows: let x1 and v1 be the initial values of x and v. Let the
initial time be t1 and let τ be a small time step. Then
x2 = x1 + v1 τ
v2 = v1 + f (x1 , v1 , t1 )τ
The next step is to replace x1 by x2 and v1 by v2 and let t1 be replaced
by t2 = t1 + τ. Then evaluate
x3 = x2 + v2 τ,
v3 = v2 + f (x2 , v2 , t2 )τ.
You can appreciate that, in general, the technique consists in replacing
xn by xn+1 and vn by vn+1 and repeating as many times as desired.
This is the essence of the Euler method.
Some years ago, Alan Cromer16 noted that the Euler method, as
expressed above, is unstable. That is, the answer diverges further and
further from the correct solution. However, Cromer discovered that the
equations could be made stable by a simple but crucial change, namely
by calculating v first and using the new value of v to determine x. The
so-called “Euler-Cromer” algorithm is
v2 = v1 + f (x1 , v1 , t1 ) · τ,
x2 = x1 + v2 τ.
This simple scheme allows you to numerically determine the velocity
and position of a particle at any future time if you know its acceleration
f (x, v, t) at a prior time.
When carrying out a numerical integration, you should check fre-
quently to determine if the solution is stable. For example, you might
check at each time step to make sure the energy or angular momentum
is conserved.

3.8. Summary
A number of important concepts were presented in this chapter. Ba-
sically, you learned (or reviewed) Newton’s three laws and you learned
how to determine the motion for a mechanical system (assuming you
know the force).
Newton’s three laws are:
(1) The law of Inertia. (A body tends to preserve its state of
motion.)
16Alan Cromer, “Stable solutions using the Euler approximation,” American
Journal of Physics, 49, 455 (1981). In this paper Cromer explains why one algo-
rithm is stable and the other is not.
 3.9. PROBLEMS 99

(2) The second law, F = dp dt

, can often be written as F = ma. It
gives the relation between the net external force acting on a
body and its acceleration.
(3) Action equals reaction. (Two bodies always exert equal and
opposite forces on each other.)
If the mass is constant, Newton’s second law can be expressed as
F
a= .
m
We call such a relation an “equation of motion” because it can be
integrated to determine the velocity v and the position x as functions
of time. Obtaining an expression for x = x(t) is called “determining
the motion.”
To determine the motion you need to integrate the equation of
motion twice, thus:
Z t
v(t) = adt
0
Z t
x(t) = v(t)dt
0
We have described the techniques for determining the motion for four
different kinds of forces. These forces are: (1) constant force, (2) force
a function of time, (3) force a function of velocity, and (4) force a
function of position. Make sure you understand the proper procedure
for each of these cases.
Finally, you were exposed to the method of determining the mo-
tion using numerical techniques on a computer. The “Euler-Cromer”
algorithm is a simple and useful way to obtain x = x(t).

3.9. Problems
Problem 3.1. A particle of mass 3 kg is acted upon by the two
forces (in newtons)
F1 = 6ı̂,
and
F2 = 3ı̂ + 3̂.
Determine the magnitude and direction of the acceleration.
Problem 3.2. A pail is filled with oil to a depth of 10 cm. A
steel marble of mass 0.2 kg is released from rest 50 cm above the top
surface. Assuming the oil exerts a constant resistive force of 2.4 N on
the marble, determine the speed with which it reaches the bottom of
the pail. Answer: 3.06 m/s
100 3. NEWTON’S LAWS: DETERMINING THE MOTION

Problem 3.3. An automobile starts from rest and accelerates at

a constant rate for 1 km, covering that distance in 20 seconds. Ignore
air resistance. Determine the coefficient of static friction between the
tires and the pavement.
Problem 3.4. A sailor on board a ship of mass 2 × 106 kg moving
at 0.2 m/s throws a heavy rope to a man standing on the wharf who
drops the noose around a bollard. After the rope is taut, the ship
is brought to rest, stretching the rope 0.5 m. Find the average pull
sustained by the rope.
Problem 3.5. A particle of mass m is observed to have a veloc-
ity given by v = A cos αx where A and α are constants. Obtain an
expression for the force F = F (x).
Problem 3.6. (a) Two stars of masses M1 and M2 attract one
another with the gravitational force F = −GM1 M2 /r2 . Determine the
ratio of their accelerations. (b) Assume both stars move in circles about
their common center of mass. Determine the radii of these circles in
terms of M1 , M2 , and r. (Actually the stars will move in elliptical orbits
with the center of mass at their common focal point.)
Problem 3.7. You are searching for an inertial reference frame.
Determine the accelerations of the following reference frames. (a) A
reference frame fixed to the surface of Earth at latitude 37◦ . (b) A
reference frame with origin at the center of Earth with one axis always
pointing towards the Sun. (c) A reference frame with origin at the
center of the Sun with one axis always pointing towards the center of
the galaxy. (The Sun is about 3/5 of the distance from the center of
our galaxy to the edge. It takes some 200 million years to “orbit” the
galaxy. The diameter of the Milky Way is about 105 ly.)
Problem 3.8. Atwood’s Machine consists of two masses m1 and
m2 tied together by a light string that passes over a smooth pulley.
Assume the pulley is massless (so it has zero moment of inertia). (a)
Obtain an equation for the acceleration of the masses and the tension
in the string. (b) Obtain an expression for the acceleration if the pulley
is not smooth and has moment of inertia I and radius R.
Problem 3.9. One normally learns Newton’s three laws as state-
ments about the nature of the physical universe. However, some people
prefer to consider only the third law as a law of nature and the first
two as definitions. Assume that you subscribe to this point of view.
Write a short essay (one or two paragraphs) explaining why you are
interpreting Newton’s laws in this manner.
 3.9. PROBLEMS 101

Problem 3.10. The gravitational force between two masses is

F = −G (M1 M2 /r2 ) and the electrostatic force between two oppo-
site charges is F = −k (Q1 Q2 /r2 ) . (a) Show that all massive objects
attracted to the Earth will have the same acceleration independent of
mass. (b) Show that two charged objects attracted to a point charge Q
will not have the same acceleration unless both have the same charge
to mass ratio.
Problem 3.11. A friend of yours, hearing that the universe is
expanding at an increasing rate, theorizes that there is a universal re-
pulsive force between any two masses, as well as the usual gravitational
force. According to your friend, this hypothetical repulsive force de-
creases with the inverse of R rather than the inverse square of R and
is given by
M1 M2
F = G0
R
where G0 = 6.67 × 10−31 Nm/kg2 . Determine the separation between
two bodies when the repulsive force would become dominant. Could
this explain the expansion of the universe? Would a galaxy the size
of the Milky Way be stable under such a force? (The diameter of the
Milky Way is about 100,000 ly.)
Problem 3.12. A certain satellite in a circular orbit around Earth
has a period T. Suppose the gravitational force were not exactly an
inverse square force, but rather depended on distance as 1/rα where
α = 2 +  where  is a small number. If the value of  were 10−4
how would this affect the period of the satellite? Would this be a
measurable quantity? (You may assume that in SI units the universal
gravitational constant has the same numerical value, although the units
would have to be different. Let the distance from the center of the
perfectly spherical Earth to the satellite be 7000 km.)
Problem 3.13. Suppose that gravitational mass and inertial mass
are different but that they are proportional to one another. Would
heavy objects and light objects then fall at the same rate? As an
explicit example, assume that inertial mass is twice as great as gravita-
tional mass. Determine the acceleration of a freely falling object near
the surface of the Earth.
Problem 3.14. Consider the following three systems: (1) A book
is sitting on a table which is at rest on the surface of Earth. (2) A
rocket is taking off from the surface of Earth. A large cloud of burned
fuel forms a contrail behind it. (3) A donkey that is hitched to a
cart is trotting down a road. (a) Identify the action-reaction pairs in
102 3. NEWTON’S LAWS: DETERMINING THE MOTION

these three cases. (b) What is the force that is causing the rocket to
accelerate? (c) What is the force on the donkey that keeps it in motion?
Problem 3.15. The Aristotelians noticed that if you push a box
across the room, it moves with a constant speed, and if you push it
harder the speed is greater, but as soon as you quit pushing, the box
stops. This suggests that the equation of motion should be F = mv.
Furthermore, the Aristotelians claimed that heavier objects fall to the
ground faster than light objects. (Thus, an object four times heavier
than a light object will fall to the ground in one fourth the time.)
(a) Show that these assumptions lead to the absurd conclusion that
the Earth exerts the same gravitational force on all bodies, regardless
of their mass. (b) Describe a simple experiment to show that the
Earth does not exert the same gravitational attraction on all bodies,
regardless of their mass.
Problem 3.16. Two carts of masses M and 2M have springs at-
tached to either end as shown in Figure 3.5. Cart 1 has mass M and
cart 2 has mass 2M. They are on a frictionless segment of track of
length L with barriers at the ends. The two carts are brought together
at the center of the track, the springs in contact are compressed, and
the carts are released from rest. Cart 1 moves left, hits the barrier at
x = 0 and bounces back. Cart 2 moves right, hits the barrier at x = L
and bounces back. The collisions are completely elastic. Where do the
carts meet? The dimensions of the carts are much smaller than the
length of track, so the carts can be treated as particles. (b) Where do
the carts meet if cart 2 is four times more massive than cart 1?

1 2
0 L/2 L

Figure 3.5. Two carts with attached springs on a fric-

tionless track. If the carts are squeezed together at the
midpoint, where will they meet? The mass of cart 2 is
twice the mass of cart 1. (See Problem 3.16)

Problem 3.17. The velocity of a particle of mass m varies with

distance x as
v(x) = v0 + bx
 3.9. PROBLEMS 103

where b and v0 are constants. (a) Find the force acting on the particle,
F (x). (b) Determine x(t). (c) Obtain an expression for the force as a
function of time, F (t).
Problem 3.18. A certain object is losing mass at a rate κ (kg/sec).
It is acted upon by a constant force F. (a) Determine its acceleration
and its position as a function of time. Assume that at t = 0, the mass
is m0 , the velocity is zero and the object is at x = 0.
 (b)
 Show that
1 F
as κ → 0, the expression for x(t) reduces to x = 2 m0 t2 , that is,
x = 21 at2.
Problem 3.19. An object of mass 2 kg is thrown upwards with an
initial velocity of 4 m/s. It is acted upon by the constant gravitational
force of the Earth and a mysterious force of 2 newtons that always
opposes the motion of the object. (This is like the force of air resistance,
except we are assuming the force is constant.) Determine the position
of the object at time t = 0.5 s. (Be careful; by this time the object is
falling downwards and the mysterious force is acting upwards!)
Problem 3.20. A builder plans to use a chain of length 1.5 meters
and linear density 7 kg/m to suspend a 220 kg mass from an overhead
beam. Determine the forces at either end and at the middle of the
chain. Obtain an equation for the force on a link at any point of the
chain. If the maximum tension that can be supported by a link is 2200
N, where will the chain break?
Problem 3.21. A bathroom scale is placed in an elevator. A man
weighing 180 lbs stands on the scale and observes that as the elevator
ascends, the scale reads 200 lbs while the elevator is accelerating and
140 lbs when it is slowing to a stop. What are the acceleration and the
deceleration of the elevator? (g=32.2 ft/s2 )
Problem 3.22. A bullet initially traveling at 400 m/s passes through
a 2 cm thick slab of wood and emerges at speed vf . When fired into a
thick block of the same wood, the bullet is found to penetrate 10 cm.
Determine vf .
Problem 3.23. A physics student designed an apparatus for her
experimental physics class. The apparatus consisted of a block m =
0.25 kg mass attached to a compressed spring of constant k mounted on
a cart of mass M = 2 kg, as shown in Figure 3.6. The car is pulled by a
string that passes over a pulley to a 10 kg hanging weight. The student
used photogate timers to determine the acceleration of the mass on the
spring when it was 15 cm from its equilibrium position. She found this
104 3. NEWTON’S LAWS: DETERMINING THE MOTION

acceleration to be -2 m/s2 relative to the lab bench. (a) What is the

spring constant? (b) What is the acceleration of the block relative to
the cart? (You may assume there is no friction in the system and the
spring is massless.)

m
M=2kg

10

Figure 3.6. A block on an accelerated cart.

Problem 3.24. Johannes Kepler determined that the period of a

planet squared is proportional to the cube of the semi-major axis of
the orbit. Let us assume planets move in circular orbits, in which case
the semi-major axis is equal to the radius of the orbit. Using this fact,
deduce that the gravitational force obeys an inverse square law. (Note:
You cannot use Newton’s law of universal gravitation!) Also show that
the velocity of a planet decreases proportionally
√ with the inverse square
root of its distance from the sun (v ∝ 1/ r).
Problem 3.25. A singly charged lithium ion is accelerated by an
electric field E =10ı̂ volts/meter applied between two plates ten cen-
timeters apart. After passing through a small hole in the negative
plate, the ion enters a region of space where there is a magnetic field
B =5̂̂ millitesla. The ion then moves in a circular path. What is the
radius of this circle?(You may assume the mass of the Li ion is 7 amu.)
Problem 3.26. A particle of mass 0.1 kg is acted upon by a force
given by
 2
t−C
F =A+B
2
where A = 1 N, B = −1 N/s2 and C = 2 s. Determine the speed and
position of the particle at time t = 4 s, if it is initially at the origin
with a velocity +6 m/s.
Problem 3.27. A particle of mass m, initially at rest, is subjected
to a force F (t) = Ae−αt cos(βt). Determine v(t) and x(t).
 3.9. PROBLEMS 105

Problem 3.28. An object with initial velocity v0 is slowed by a

force given by F = −Aebv where A and b are constants. Obtain an
expression for the velocity as a function of time.
Problem 3.29. An airplane accelerates down the runway. Assume
the jet engines exert a constant force on the airplane and assume air
resistance is proportional to the velocity squared. Obtain an expression
for the velocity as a function of time. What is the terminal velocity?
Problem 3.30. A falling body is subjected to the gravitational
force (acting down) and the retarding force of air resistance (acting
Dv 2 where D is a constant.
up). Suppose the retarding force is given byp
Show that in the limit of small t (t << m/Dg) the velocity and
position are given by
1
v ' −gt, x ' − gt2 ,
2
p
and in the limit of large t (t >> m/Dg)
p m p
v ' − mg/D, x' ln 2 − mg/Dt.
D
Hint: Expand the hyperbolic functions.
Problem 3.31. You throw a ball vertically upward with an initial
velocity v0 . Assume that air resistance generates a force given by Dv 2
opposing the motion. (a) Determine the terminal velocity (note that
this involves the falling motion of the ball). (b) Determine the height
attained by the ball. (c) Determine the speed of the ball when it returns
to your hand (in general this will not be the terminal velocity).
Problem 3.32. A race car of mass 800 kg reaches a velocity of
200 kilometers per hour when suddenly the engine quits operating and
the brakes go out! The only thing slowing the car down is air resis-
tance (rolling friction is assumed negligible). Assume the drag force is
proportional to the square of the velocity (F = −Dv 2 ). After the car
coasts 500 meters it is moving at 5 km/h. What is the drag coefficient
D?
Problem 3.33. A skydiver has a mass of 100 kg (including the
parachute). When the parachute is open, the retarding force is pro-
portional to the velocity to the first power. Furthermore, it is known
experimentally that the parachute exerts a retarding force per unit area
of 90 N/m2 when the speed is 3 m/s. (a) It is desired that the skydiver
not exceed a safe speed of 1.5 m/s upon reaching the ground. What is
the required area of the parachute? (b) Assuming an initial speed of
106 3. NEWTON’S LAWS: DETERMINING THE MOTION

zero, how much time does it take for the skydiver to reach a speed of
90% of the terminal velocity?
Problem 3.34. A bullet of mass m is fired vertically upward with
an initial velocity v0 . Assume the resistive force of the air is given by
mkv where m is the mass, v is the velocity and k is a constant. How
much time is required for the bullet to reach the top of its path?
Problem 3.35. A physics student on the roof of a tall building
drops a ball of diameter 10 cm and mass 0.3 kg. The air density is 1.2
kg/m3 . (a) What is the terminal speed? (b) Obtain an expression for
the speed of the ball as a function of how far it has fallen. (c) If the
roof is 100 meters above the ground, what is the speed of the ball when
it hits the sidewalk?
Problem 3.36. An object is dropped from an airplane that is flying
at an altitude of 6000 m at a speed of 1000 km/hr. Determine the
horizontal distance traveled by the object by the time it hits the ground.
For the sake of making the calculations somewhat simpler, you may
assume that the resistive force of the air is proportional to the first
power of the velocity and that the terminal velocity of the object is 98
m/s. (Justify neglecting small terms.)
Problem 3.37. An object of mass m in a uniform gravitational
field is dropped from some initial height. Assume that the force of air
resistance is proportional to the first power of the velocity (∝ bv). Show
that the distance y it has fallen in time t is
mg  m −bt/m

y= t+ e −1 .
b b
Problem 3.38. Assume the force of air resistance is given by Fair =
2
Dv where v is the velocity and D is a constant equal to 0.01 kg/m.
An object of mass 2 kg initially at rest, is dropped from a height of
1000 m. The gravitational force can be assumed constant. Determine
how far the object has fallen in 20 s.
Problem 3.39. An object of mass 10 kg is dropped from a high
place. Assume air resistance is proportional to the velocity squared
(= Dv 2 ) where D = 0.01 Ns2 /m2 . (a) Evaluate the terminal velocity.
(b) Determine the time required for the ball to reach 0.9vT .
Problem 3.40. A bead on a long straight wire is repelled from the
origin x = 0 by a force proportional to its distance to that point. The
bead is initially at rest at position xo . Determine its motion.
Problem 3.41. A particle of mass m kg is at rest at x0 . It is
subjected to a force of magnitude F = k/x where k = 4 Nm. the
 3.9. PROBLEMS 107

force is directed along the positive x-axis. Determine the speed of the
particle when it passes through the point x = 2x0 .
Problem 3.42. A block of mass m on a frictionless horizontal
surface is attached to two springs in series. The springs have force
constants k1 and k2 . Determine
p pthe angular frequency of oscillation of
the block. Answer: ω = 1/m (k1 k2 )/(k1 + k2 ).
Problem 3.43. An asteroid, initially at rest, falls from a great
distance (x0 ) towards a star. You are asked to determine the time
required for it to hit the star. The star can be assumed to be at rest
at the origin at all times. Evaluate the time required for the asteroid
to reach x = 0, that is, assume the star is a point object.
Problem 3.44. A mass m is attached to a spring of constant k.
Determine the motion if it is initially at the unstretched
R position and
is given an impulse J. (Impulse is defined by J = Fdt.)
Problem 3.45. (a) Obtain a general relation for the motion of a
particle of mass m acted upon by a repulsive force F = +kx. Assume
the particle is initially at the position x = x0 . Note that the solution
will have the form x = Aeβt + Be−βt . (b) It may surprise you to know
that for this force there is a possible solution in which the particle
moves to the origin and remains at rest there. In this case, what is the
initial value of the velocity?
Problem 3.46. An asteroid of mass m is initially very far from
Earth and has zero velocity relative to Earth. It falls under the action
of the gravitational force, F = −GmM/z 2 where M is the mass of
Earth. Determine the speed of the asteroid as a function of z, its
distance from the center of Earth.
Problem 3.47. An enterprising scientist drills a hole straight through
the Earth, from North Pole to South Pole, and drops an object of mass
m in the hole at the North Pole. How long does it take for the object
to emerge from the hole at the South Pole? You may assume that there
is no air in the hole. Hint: The gravitational force inside a uniform
sphere is given by F = −GM 0 m/r2 where M 0 is the mass contained
within a sphere of radius r and r is the distance from the center of the
sphere to the object m.
Computational Projects
Computational Project 3.1. Write a program to plot the ve-
locity and position of the canoe of Worked Example 3.3. Assume the
mass of the canoe plus person is 200 kg. Use b = 150 kg/s as a first
108 3. NEWTON’S LAWS: DETERMINING THE MOTION

estimate and get plots for various values of b. Use a reasonable choice
for the initial velocity.
Computational Project 3.2. Demonstrate that the Euler-Cromer
method is stable but the Euler method is not. Do this by assuming
a particle moves in one dimension (x) under the influence of a force
F = −kx. This will lead to simple harmonic motion. Compare the
total energy as a function of time as obtained using the Euler method
and the Euler-Cromer method. Also, on the same plot, present the to-
tal energy as a function of time as obtained from the exact (analytical)
expression. You will appreciate that the Euler method does not con-
serve energy, the total energy obtained by the Euler-Cromer method
oscillates around a constant value, and the analytical expression gives
a constant value for the energy at all times.
Computational Project 3.3. A projectile is fired perpendicu-
lar to the Earth’s surface with an initial velocity of 600 m/s. Assume a
constant gravitational force acts on the projectile so its acceleration is
9.8 m/s2 downwards. Write a program to determine the position of the
particle as a function of time. Next, consider the same problem, but
now include the effect of air resistance, assuming it is a retarding force
that can be expressed as fair = 2.5 × 10−4 v 2 where v is the velocity of
the projectile.
Chapter 4
The Lagrangian Method

In this chapter you will learn a different approach for solving physics
problems. This approach is associated with Lagrange,1 although many
other physicists and mathematicians contributed to it, including Leib-
niz and Euler.
We shall first introduce a function called the Lagrangian, and then
show how it can be used to determine the equations of motion.
A mechanical system is fully described by the Lagrangian. This
means that if you know the Lagrangian for a system, you can deter-
mine the equation of motion, the momentum, and all other relevant
mechanical quantities.
So far, we have been using Newton’s second law to determine the
equation of motion. The Lagrangian method is a very useful alterna-
tive technique for determing this equation, but before discussing the
Lagrangian, it is helpful to review how Newton’s second law leads to
an expression for the equation of motion.

4.1. The Equation of Motion by Inspection

We have considered the equation of motion to be an equation for
the acceleration as a function of position, velocity and time. That is,
a = a(x, v, t),
or
ẍ = ẍ(x, ẋ, t).

1JosephLouis Lagrange was a French physicist and mathematician who lived

from 1736 to 1813.
109
110 4. THE LAGRANGIAN METHOD

In the preceding chapter you learned how to solve the equation of

motion for a variety of different force laws. But how can you determine
the equation of motion?
The equation of motion for a system can often be determined from
Newton’s second law in the form F = ma. For example, the equation of
motion for a simple pendulum is easily obtained by drawing the forces
acting on the bob and resolving them into appropriate components, as
illustrated in Figure 4.1. A glance at the figure will remind you that the
gravitational force can be resolved into a component along the string
(mg cos θ) and a component (mg sin θ) tangent to the circular path of
the bob. The position of the bob can be specified by the arc length s,
measured from the lowest point in the motion. Consequently, F = ma
becomes
−mg sin θ = ms̈.
This equation involves two variables, s and θ. But s = lθ, where l is
the length of the string. So the equation of motion can be written in
terms of the single parameter θ as:
g
θ̈ = − sin θ. (4.1)
l

l
θ
T

s mgcosθ
mg
mgsinθ
Figure 4.1. Forces acting on a simple pendulum. The
length of the string is l.

I went through this trivial example to remind you that it is easy

to determine the equation of motion for a simple system, such as a
pendulum.
But now consider the problem of determining the equation of mo-
tion for the double pendulum shown in Figure 4.2. For simplicity,
assume the double pendulum is constrained to oscillate in the plane of
the page so the positions of the bobs can be described by the two angles
θ1 and θ2 . Observe that the motion of m2 depends on the motion of
m1 so the tension in the lower string is not equal to m2 g cos θ2 . (The
 4.2. THE LAGRANGIAN 111

reason is that the point of support of the lower pendulum is accelerat-

ing so m2 does not behave like a pendulum in an inertial system.) It
would be very difficult to obtain the accelerations of the two masses by
applying Newton’s second law. A different approach is needed.
I will now describe a very powerful but fairly simple technique called
the Lagrangian approach. This will allow you to obtain the equations of
motion for complicated problems such as the double pendulum. First,
however, I would like to tell you a bit more about the Lagrangian.

Exercise 4.1. Draw a force diagram for each of the masses of the
double planar pendulum. Consider the motion of mass m2 . What is
wrong with writing m2 s̈2 = −m2 g sin θ2 ?

y
x
l
1
θ1
m1

l2
θ m2
2

Figure 4.2. A double planar pendulum.

4.2. The Lagrangian

By studying Lagrangian dynamics you will develop a greater un-
derstanding of the framework of mechanics and an appreciation of how
the physical universe behaves. Lagrangian dynamics goes to the very
core of the science of mechanics. In fact, as you will see in Section 4.7,
the Lagrangian approach can be used to derive Newton’s laws.
If you take a graduate courses in mechanics, you will find yourself
deeply immersed in Lagrangian methods. Here I will only use it as
a technique - an easy way to obtain the equations of motion for a
complicated system. Nevertheless, later in this chapter I will present
an elementary derivation of the technique, showing how Lagrange’s
112 4. THE LAGRANGIAN METHOD

equations are obtained using the calculus of variations. You will learn
that Lagrangian dynamics is based on a profound statement about
the nature of the physical world called Hamilton’s principle. Later (in
Chapter 8) you will see how the Lagrangian is related to symmetries
in physical systems, and to the conservation laws of physics.
The Lagrangian method, as described here, is applicable to systems
in which all the forces are conservative. Consequently, we will not
consider systems with frictional forces. Although there are ways of
incorporating dissipative forces into the Lagrangian formulation, they
involve advanced methods and we do not consider them in this book.
So if friction is present, you must go back to Newton’s second law.
The study of Lagrangian dynamics begins with the definition of a
physical quantity called the Lagrangian. It is denoted by L and is de-
fined as the difference between the kinetic energy (T ) and the potential
energy (V ):

L = T − V. (4.2)

As an application, I will evaluate the Lagrangian for the simple

pendulum of Figure 4.1. Assume the origin of coordinates is at the
point of support. In Cartesian coordinates the potential energy is V =
mgy. The kinetic energy is T = 12 m(ẋ2 + ẏ 2 ). Therefore the Lagrangian
for the simple pendulum is

1
L = T − V = m(ẋ2 + ẏ 2 ) − mgy. (4.3)
2

There are two coordinates in this expression for L, namely x and y. As

you will see shortly, this leads to two equations of motion. Normally it
is desirable to have as few equations of motion as possible. Therefore,
it is desirable to use a coordinate system (such as r, θ) that will min-
imize the number of equations of motion. Physicists frequently need
to transform from one set of coordinates to another. This is done by
using transformation equations. Consider transforming from the Carte-
sian coordinates, x, y, z, to a different set of coordinates, say q1 , q2 , q3 .
The transformation equations are a set of equations having the general
form

x = x(q1 , q2 , q3 , t)
y = y(q1 , q2 , q3 , t)
z = z(q1 , q2 , q3 , t).
 4.2. THE LAGRANGIAN 113

The q’s are called “generalized coordinates.” In practice they will often
be Cartesian, cylindrical, or spherical coordinates, but they are not
limited to these familiar coordinates.2
In the example of the pendulum, the transformation equations from
x and y to θ are
x = l sin θ,
y = −l cos θ,
where the coordinate origin is at the point of suspension and θ is mea-
sured from the negative y-axis (this is not the way we usually define
θ!)
Differentiating these equations with respect to time,
ẋ = lθ̇ cos θ,
ẏ = lθ̇ sin θ.
Plugging these expressions into the Lagrangian for the pendulum (Equa-
tion 4.3) leads to
1 1
L = m(lθ̇ cos θ)2 + m(lθ̇ sin θ)2 + mgl cos θ
2 2
1 2 2
= ml θ̇ (cos2 θ + sin2 θ) + mgl cos θ
2
1 2 2
= ml θ̇ + mgl cos θ. (4.4)
2
Equation (4.4) gives the same information as Equation (4.3) that was
previously obtained for the Lagrangian in terms of x and y, except that
now L depends only on θ.
All I have done so far is to write the Lagrangian in Cartesian co-
ordinates and then transform to polar coordinates. In doing so, the
Lagrangian went from being a function of two variables to being a
function of one variable. This procedure illustrates a very important
point: the Lagrangian should be expressed in terms of the least possible
number of coordinates.
Let me give you a simple “cookbook” procedure for obtaining the
Lagrangian in terms of a set of generalized coordinates q1 , q2 , q3 .
1. If at all possible, write the Lagrangian L = T −V in terms
of Cartesian coordinates. In Cartesian coordinates the translational
2The generalized coordinates are normally a “minimal set” of coordinates. For
example, in Cartesian coordinates the simple pendulum requires two coordinates
(x and y), but in polar coordinates only one coordinate (θ) is required. So θ is the
appropriate generalized coordinate for the pendulum problem.
114 4. THE LAGRANGIAN METHOD

kinetic energy is
1
T = m ẋ2 + ẏ 2 + ż 2 .


2
2. Write the transformation equations:
x = x(q1 , q2 , q3 , t),
y = y(q1 , q2 , q3 , t),
z = z(q1 , q2 , q3 , t).
3. Take derivatives to obtain expressions for ẋ, ẏ, ż in terms
of q’s, q̇’s and t.
4. Write L = T − V in terms of the q’s, q̇’s and t.
You might wonder why I am emphasizing that you should first write
the Lagrangian in Cartesian coordinates. The answer is simple. The
translational kinetic energy in Cartesian coordinates always has the
simple form T = 21 m (ẋ2 + ẏ 2 + ż 2 ) , whereas it can be very compli-
cated in other coordinate systems. Furthermore, the potential energy
is often (but not always) easier to express in Cartesian coordinates.
A very simple (but important) system is a block of mass m on a
frictionless surface and connected to a spring of constant k. See Figure
4.3. It is easy to appreciate that the Lagrangian for this system is
1 1
L = mẋ2 − kx2 . (4.5)
2 2

k
m

Figure 4.3. A mass m on a frictionless surface acted

upon by a spring of constant k.

Exercise 4.2. A rock of mass m is dropped from a height z0 . Write

the Lagrangian. Answer: L = 12 mż 2 − mgz.

Worked Example 4.1. Determine the Lagrangian of the dou-

ble planar pendulum.
 4.2. THE LAGRANGIAN 115

Solution: See Figure 4.2. In Cartesian coordinates the kinetic

and potential energies are:
1 1
T = m1 (ẋ21 + ẏ12 ) + m2 (ẋ22 + ẏ22 ),
2 2
V = m1 gy1 + m2 gy2 ,
where (x1 , y1 ) are the coordinates of the bob of mass m1 and (x2 , y2 )
are the coordinates of the other bob. The origin of coordinates is
at the point of suspension. The transformation equations from the
Cartesian coordinates to the angles θ1 and θ2 are:
x1 = l1 sin θ1
x2 = l1 sin θ1 + l2 sin θ2
y1 = −l1 cos θ1
y2 = −l1 cos θ1 − l2 cos θ2
The velocity components are obtained by taking derivatives of
position with respect to time:
ẋ1 = l1 θ̇1 cos θ1
ẋ2 = l1 θ̇1 cos θ1 + l2 θ̇2 cos θ2
ẏ1 = l1 θ̇1 sin θ1
ẏ2 = l1 θ̇1 sin θ1 + l2 θ̇2 sin θ2
Therefore,
ẋ21 + ẏ12 = l12 θ̇12 (cos2 θ1 + sin2 θ1 ) = l12 θ̇12 ,
and
 2
ẋ22 + ẏ22 2
= (l1 θ̇1 cos θ1 + l2 θ̇2 cos θ2 ) + l1 θ̇1 sin θ1 + l2 θ̇2 sin θ2

= l12 θ˙12 + l22 θ˙22 + 2l1 l2 cos(θ1 − θ2 ).

Consequently
1 1
L = T − V = (m1 + m2 )l12 θ̇12 + m2 (l22 θ̇22 + 2l1 l2 θ̇1 θ̇2 cos{θ1 − θ2 })
2 2
+m1 gl1 cos θ1 + m2 g(l1 cos θ1 + l2 cos θ2 ). (4.6)
I hope you appreciate that obtaining the Lagrangian is a straight-
forward procedure although it may be somewhat tedious. This ex-
ample also illustrates the value of using Cartesian coordinates as
your starting point.
116 4. THE LAGRANGIAN METHOD

y
x
s

Figure 4.4. A disk of mass m and radius R rolls down

a perfectly rough plane.

Worked Example 4.2. Determine the Lagrangian for a disk

of mass m and radius R rolling down an inclined plane of angle
α. See Figure 4.4. The plane is perfectly rough so the disk rolls
without slipping.
Solution: Since the disk is rolling, its total kinetic energy is
rotational kinetic energy plus translational kinetic energy. That is,
T = Trot + Ttrans . The rotational kinetic energy is not, of course,
expressed in Cartesian coordinates; it is Trot = 12 I θ̇2 . (The moment
of inertia is I and the angle through which the disk has rotated is
θ.) The moment of inertia of a disk is 12 mR2 . Therefore,
 
1 1 1
Trot = mR θ̇2 = mR2 θ̇2 .
2
2 2 4
As always, Ttrans = 12 m(ẋ2 + ẏ 2 ). Let s be the distance measured
down the plane. It is convenient to use s as the generalized coor-
dinate. The transformation equations are
x = s cos α,
y = −s sin α.
The translational velocity components are
ẋ = ṡ cos α,
ẏ = −ṡ sin α.
Consequently the translational kinetic energy is
1 1
Ttrans = m(ẋ2 + ẏ 2 ) = mṡ2 .
2 2
 4.3. LAGRANGE’S EQUATIONS 117

I expressed the translational kinetic energy in terms of ṡ and the

rotational kinetic energy in terms of θ̇. But these two quantities are
related. As the disk rolls down the plane, the relationship between
the angle θ through which it has rotated, and the distance s it has
moved down the plane is
s = Rθ.
(A relationship between coordinates, such as this one, is called
a “constraint.” A constraint allows us to express one coordinate
in terms of the others and to describe the problem with one less
coordinate.) Therefore, ṡ = Rθ̇. Consequently, Trot = (1/4)mṡ2 ,
and the total kinetic energy can be written in terms of the single
parameter ṡ :
1 1 3
T = Ttrans + Trot = mṡ2 + mṡ2 = mṡ2 .
2 4 4
The potential energy is
V = mgy = −mgs sin α.
The Lagrangian is defined as L = T − V, so,
3
L = mṡ2 + mgs sin α. (4.7)
4

Exercise 4.3. A pendulum is made of a bob of mass m but the

string is replaced by a spring of constant k. Write the Lagrangian in
terms of the length of the spring and the angle it makes with the
vertical. Let the unstretched length of the spring be l0 . Answer:
1 1 1
L = ml˙2 + ml2 θ̇2 + mgl cos θ − k(l − l0 )2 .
2 2 2

4.3. Lagrange’s Equations

As you have seen, the Lagrangian for a system can be obtained
fairly easily by using the definition L = T − V. The algebra may get a
bit complicated at times, but basically all you need to do is express the
translational kinetic energy and the potential energy in terms of Carte-
sian coordinates (if possible) and then transform to the appropriate set
of “generalized” coordinates.
By this time you are probably wondering what the Lagrangian is
used for. What good is it? The simplest answer is that the Lagrangian
118 4. THE LAGRANGIAN METHOD

allows us to determine the equations of motion for a mechanical system.

In fact, the Lagrangian technique is usually the easiest way to find the
equations of motion for a complicated system. (As I mentioned earlier,
the Lagrangian method is much more that just a useful technique, but
for the present let’s take a purely utilitarian approach and consider the
Lagrangian as a tool for generating the equations of motion.)
Suppose the Lagrangian is expressed as a function of a single gen-
eralized coordinate q, the corresponding generalized velocity q̇, and the
time t. That is, L = L(q, q̇, t). Then the equation of motion is given by:
 
d ∂L ∂L
− = 0. (4.8)
dt ∂ q̇ ∂q
This is called Lagrange’s equation. You should memorize it. Note
particularly that the first partial derivative is taken with respect to the
generalized velocity (q̇) and the second partial derivative is taken with
respect to the generalized position (q).
If the Lagrangian is a function of n coordinates, q1 , q2 , ..., qn , there
are n equations of motion, each having the form of Equation (4.8).
That is, the equations of motion are
 
d ∂L ∂L
− = 0 ; i = 1, 2, ..., n. (4.9)
dt ∂ q̇i ∂qi
Consider, for example, the mass on the spring of Figure 4.3. As
noted in Equation 4.5, the Lagrangian is L = 12 mẋ2 − 12 kx2 . For this
system the generalized coordinate q is just the Cartesian coordinate x.
There are no other coordinates. Therefore, there is only one Lagrange
equation, namely,  
d ∂L ∂L
− = 0.
dt ∂ ẋ ∂x
But
∂L
= mẋ,
∂ ẋ
and
∂L
= −kx.
∂x
Consequently, Lagrange’s equation is
d
(mẋ) + kx = 0,
dt
or
mẍ + kx = 0.
Of course, you get the same result immediately by applying Newton’s
second law.
 4.3. LAGRANGE’S EQUATIONS 119

To illustrate the process for a slightly more complicated system,

the following worked example asks for the equation of motion for a
simple pendulum and the next one asks for the equations of motion for
a double pendulum.

Worked Example 4.3. Determine the equation of motion for

a simple pendulum.
Solution: The Lagrangian for a pendulum is L = 21 ml2 θ̇2 +
mgl cos θ. (See Equation 4.4.) In this expression, the single coor-
dinate is θ and Lagrange’s equation (Eq. 4.8) becomes
 
d ∂L ∂L
− = 0.
dt ∂ θ̇ ∂θ
But the partial of L with respect to the angular velocity θ̇ is
 
∂L ∂ 1 2 2
= ml θ̇ + mgl cos θ = ml2 θ̇,
∂ θ̇ ∂ θ̇ 2
and the partial of L with respect to the angular position θ is
 
∂L ∂ 1 2 2
= ml θ̇ + mgl cos θ = −mgl sin θ.
∂θ ∂θ 2
Therefore, Lagrange’s equation is
d  2 2
ml θ̇ + mgl sin θ = 0,
dt
or,
g
θ̈ = − sin θ.
l
This is, of course, exactly the same equation of motion obtained
earlier by elementary methods (Equation 4.1).

Worked Example 4.4. Determine the equations of motion

for the double planar pendulum illustrated in Figure 4.2.
Solution: Recall that the double pendulum was too compli-
cated to analyze using Newton’s second law. The Lagrangian for
the double pendulum is given by Equation (4.6) which is repeated
120 4. THE LAGRANGIAN METHOD

here:
1 1
L = (m1 + m2 )l12 θ̇12 + m2 [l22 θ̇22 + 2l1 l2 θ̇1 θ̇2 cos(θ1 − θ2 )]
2 2
+m1 gl1 cos θ1 + m2 g(l1 cos θ1 + l2 cos θ2 ).
This problem involves two coordinates, θ1 and θ2 , and therefore
there are two Lagrange equations, namely:
 
d ∂L ∂L
− = 0, (4.10)
dt ∂ θ̇1 ∂θ1
 
d ∂L ∂L
− = 0.
dt ∂ θ̇2 ∂θ2
To facilitate writing them, first evaluate the derivatives:
∂L
= −m2 l1 l2 θ̇1 θ̇2 sin(θ1 − θ2 ) − m1 gl1 sin θ1 − m2 gl1 sin θ1 ,
∂θ1
∂L
= m2 l1 l2 θ̇1 θ̇2 sin(θ1 − θ2 ) − m2 gl2 sin θ2 ,
∂θ2
∂L
= (m1 + m2 )l12 θ̇1 + m2 l1 l2 θ̇2 cos(θ1 − θ2 ),
∂ θ̇1
∂L
= m2 l22 θ̇2 + m2 l1 l2 θ̇1 cos(θ1 − θ2 ).
∂ θ̇2
Plugging into Equations (4.10) yields two coupled equations of mo-
tion:
d h 2 2
i
m1 l1 θ̇1 + m2 l1 θ̇1 + m2 l1 l2 θ̇2 cos(θ1 − θ2 )
dt
+m2 l1 l2 θ̇1 θ̇2 sin(θ1 − θ2 ) + (m1 + m2 )gl1 sin θ1 = 0,
and
d  2

m2 l2 θ̇2 + m2 l1 l2 θ̇1 cos(θ1 − θ2 )
dt
−m2 l1 l2 θ̇1 θ̇2 sin(θ1 − θ2 ) + m2 gl2 sin θ2 = 0.
These equations can be further simplified and written in the
form
h i
2 2
(m1 + m2 )l1 θ̈1 + m2 l1 l2 θ̈2 cos(θ1 − θ2 ) + θ̇2 sin(θ1 − θ2 )
+(m1 + m2 )gl1 sin θ1 = 0,
and
h i
m2 l22 θ̈2 +m2 l1 l2 θ̈1 cos(θ1 − θ2 ) − θ̇12 sin(θ1 − θ2 ) +m2 gl2 sin θ2 = 0.
 4.4. DEGREES OF FREEDOM 121

(I’m sure you now appreciate why I did not attempt to solve
the double pendulum problem by using F = ma!)

Exercise 4.4. Find the equation of motion for a disk rolling down
a perfectly rough inclined plane of angle α. (The Lagrangian was ob-
tained in Worked Example 4.2.) Answer: s̈ = 32 g sin α.
Exercise 4.5. Find the equation of motion for a sphere rolling
down a perfectly rough inclined plane of angle α. (Use the Lagrangian
technique.) The moment of inertia of a sphere is 52 mR2 . Answer: s̈ =
5
7
g sin α.
Exercise 4.6. Using the Lagrangian technique, determine the equa-
tion of motion for a body of mass m falling in a constant gravitational
field.
Exercise 4.7. Find the Lagrangian and the equation of motion for
two astronomical bodies on a collision course that are attracting one
another according to Newton’s Law of Universal Gravitation. (Place
the origin of coordinates at the point where the bodies will collide.)
m1 m2
Answer: m1 r¨1 = −G (1+m 2 (where r1 is the distance from the
1 /m2 )r1
center of mass to body m1 ).

4.4. Degrees of Freedom

The Lagrangian technique often involves changing from one set of
coordinates (usually x, y, z) to another (say, r, θ, φ). (It turns out that
coordinate transformations are central to the Lagrangian method; they
are considered in excruciating detail in graduate courses in mechanics.
If you want to know more about this subject, I recommend the excel-
lent, but advanced, books by Goldstein3 and by Fetter and Walecka.4)
3Goldstein, H. Classical Mechanics, Addison-Wesley Publishing Co., Reading,
MA: 1950. Note that there are three editions. The first edition is the shortest
and probably the best. The second edition included a significant amount of new
material but unfortunately it had many typographical errors which made it rather
difficult to read. The third edition (by H. Goldstein, C. Poole, and J. Safko) has
corrected the errors from the second edition and contains some interesting new
material.
4Fetter, A. L., and Walecka, J. D., Theoretical Mechanics of Particles and
Continua, McGraw-Hill, NY, 1980.
122 4. THE LAGRANGIAN METHOD

In discussing coordinates it is important to realize that every phys-

ical system has a particular number of degrees of freedom. The number
of degrees of freedom is the number of independent coordinates needed
to completely specify the position of every part of the system. To de-
scribe the position of a free particle you must specify the values of three
coordinates (say, x, y, and z). Thus a free particle has three degrees
of freedom. For a system of two free particles you need to specify the
positions of both particles. Each particle has three degrees of freedom,
so the system as a whole has six degrees of freedom. In general a me-
chanical system consisting of N free particles will have 3N degrees of
freedom.
For many systems, however, the number of independent coordinates
is much less than 3N. For example, the position of a particle on a flat
surface such as a tabletop can be described with two coordinates, such
as x and y. A condition that specifies the value of a coordinate or gives
a relation between coordinates is called a constraint. For the particle
on the tabletop, the constraint is
z = constant.
Similarly, a pendulum bob is constrained by the string to remain a
constant distance from the point of support. That is, the bob is con-
strained to the surface of a sphere whose radius equals the length of
the string. Similarly, a bar of soap slipping in a sink is constrained to
the surface of the sink. If the sink is a hemispherical bowl of radius a,
the equation of constraint is
x 2 + y 2 + z 2 = a2 .
This relationship indicates that there are only two independent coordi-
nates because the third coordinate (say z) is related to x and y through
p
z = a2 − x 2 − y 2 .
Constraints are important to us because each constraint reduces
the number of degrees of freedom by one.
Going back to the Lagrangian method, recall that there is one La-
grange equation for each coordinate. (See Equation 4.9.) It is obviously
beneficial to reduce the number of equations. You can do this by us-
ing constraints to get rid of as many nonindependent coordinates as
possible. When you describe the system in terms of independent coor-
dinates (say q1 , q2 , ..., qn ), you have minimized the number of Lagrange
equations. These independent coordinates are referred to as generalized
coordinates and are usually denoted by qi . A system with n degrees
of freedom can be described in terms of the n generalized coordinates
 4.5. GENERALIZED MOMENTUM 123

q1 , · · ·, qn . For such a system, the Lagrangian will be a function of

the generalized coordinates, the generalized velocities, and possibly the
time. That is,
L = L(q1 , q2 , · · ·qn ; q̇1 , q̇2 , · · ·q̇n ; t).
In this case, there will be n Lagrange equations of motion.5

4.5. Generalized Momentum

Consider again the problem of a mass m connected to a spring of
constant k as illustrated in Figure 4.3. If the mass is moving in the x
direction with speed ẋ it has momentum px = mẋ. We have seen that
the Lagrangian for this system is
L = 21 mẋ2 − 12 kx2 .
Taking the derivative of the Lagrangian with respect to ẋ yields
∂L ∂ 1 2 1 2


= mẋ − kx = mẋ.
∂ ẋ ∂ ẋ 2 2

But mẋ is just the linear momentum! Therefore, for this system, the
linear momentum is related to the Lagrangian by
∂L
px = .
∂ ẋ
Next, consider the problem of a simple pendulum consisting of a
mass m hanging from a string of length l. According to Equation (4.4),
the Lagrangian for this system is
L = 12 ml2 θ̇2 + mgl cos θ.
Taking the derivative of the Lagrangian with respect to θ̇ yields
∂L
= ml2 θ̇.
∂ θ̇
2
But ml θ̇ is the angular momentum of the pendulum! In this case the
derivative of the Lagrangian with respect to the angular velocity θ̇ is
the angular momentum.
In the first case, the Lagrangian was expressed in terms of x and
ẋ. That is, L = L(x, ẋ). The generalized coordinate was x and the
generalized velocity was ẋ. In the second case, the Lagrangian was a
function of θ and θ̇, that is, L = L(θ, θ̇). Here the generalized coordinate
was θ and the generalized velocity was θ̇. These two cases illustrate the
5I must warn you that physicists are somewhat careless in their usage of the
term “generalized coordinate.” You will often hear it applied in situations in which
the qi are not all independent.
124 4. THE LAGRANGIAN METHOD

general functional form of the Lagrangian for a one variable system,

L = L(q, q̇, t). That is, the Lagrangian depends on the generalized
position (q), the generalized velocity (q̇), and possibly the time (t).
The linear momentum of a mass on a spring was given by ∂L/∂ ẋ
and the angular momentum for the pendulum was given by ∂L/∂ θ̇. It
is reasonable to define a quantity called the “generalized momentum”
by
∂L
pi = . (4.11)
∂ q̇i
The generalized momentum pi is associated with the generalized co-
ordinate qi . These two quantities are called conjugates. Thus, for the
mass on a spring, the generalized momentum conjugate to x is mẋ and
for the pendulum, the generalized momentum conjugate to θ is ml2 θ̇. In
the first case the generalized momentum is the linear momentum and in
the second case the generalized momentum is the angular momentum.
Although the quantity ∂L/∂ q̇ is called the generalized momentum,
and is denoted by the symbol pi , be aware that it is frequently not
anything you would normally think of as a momentum. For example,
when we discussed a disk rolling down an inclined plane (see Worked
.2
Example 4.2) we found the Lagrangian to be L = 34 ms + mgs sin α.
Consequently the generalized momentum is
∂L 6
ps = = mṡ.
∂ ṡ 4
which is not what one might expect for linear momentum.
4.5.1. Ignorable Coordinates. We now come to a very interest-
ing point. For a mechanical system the Lagrangian might not depend
explicitly on a particular coordinate, say qi . If qi does not appear ex-
plicitly in the Lagrangian, then qi is called an ignorable coordinate.6
In such a case, it is obvious that ∂L/∂qi = 0. Recall that Lagrange’s
equation states that
 
d ∂L ∂L
− = 0.
dt ∂ q˙i ∂qi
But if ∂L/∂qi = 0, then
 
d ∂L
= 0.
dt ∂ q˙i
6When we say that q
i does not appear explicitly in the Lagrangian we mean that
if the Lagrangian is expressed in terms of the minimal set of generalized coordinates,
the quantity qi is not present. (Nevertheless, the generalized velocity q̇i may appear
in the Lagrangian.)
 4.5. GENERALIZED MOMENTUM 125

The term in parenthesis is the generalized momentum pi , so this equa-

tion states that dpi /dt = 0. If the time derivative of a quantity is
zero, the quantity is a constant. Therefore, if the coordinate qi does
not appear explicitly in the Lagrangian, then the conjugate generalized
momentum pi is constant.
I will summarize all of this in a single phrase which is easy to
remember, but perhaps not so easy to understand:
If the coordinate qi is ignorable, the
conjugate momentum pi is constant.
In studying a physical system, whenever you encounter a quantity
that remains constant during the motion, you have a very useful tool
for solving problems. Such constant quantities are called “constants of
the motion” or “first integrals” for reasons we shall discuss later.

Worked Example 4.5. The Lagrangian for a certain system

is given by L = (1/2)mṙ2 + (1/2)mr2 φ̇2 , where r and φ are gen-
eralized coordinates and m is constant. Determine the generalized
momenta pr and pφ . Is either of these a constant of the motion?
Solution: The generalized coordinates are r and φ, so the
conjugate generalized momenta are
∂L ∂L
pr = and pφ =
∂ ṙ ∂ φ̇
Therefore,
 
∂L ∂ 1 2 1 2 2
pr = = mṙ + mr φ̇ = mṙ
∂ ṙ ∂ ṙ 2 2
and  
∂L ∂ 1 2 1 2 2
pφ = = pφ = mṙ + mr φ̇ = mr2 φ̇
∂ φ̇ ∂ φ̇ 2 2
Since the Lagrangian does not depend on φ (it is ignorable), we
conclude that pφ is constant.

Exercise 4.8. Determine the generalized momenta pθ1 and pθ2 for
the double planar pendulum. Is either one of them a constant? Answer:
pθ2 = m2 l22 θ̇2 + m2 l1 l2 θ̇1 cos(θ1 − θ2 ). No.
Exercise 4.9. (a) Determine the generalized momentum for a sphere
of mass m and radius R rolling down an inclined plane of angle α. Let
126 4. THE LAGRANGIAN METHOD

the distance down the plane be given by s. (b) Is the generalized

momentum a constant? Answer: (a) ps = (5/2)mṡ. (b) No.

4.6. Generalized Force (Optional)

Lagrangian mechanics introduced the concepts of generalized coor-
dinates and generalized momenta. So it should not surprise you that
there is another quantity called the generalized force. The generalized
force is denoted by Qj and is defined in terms of a quantity called the
“virtual work.” However, before we can consider these concepts fur-
ther, we need to introduce yet another new concept, the “virtual dis-
placement” δx. The virtual displacement is an imaginary infinitesimal
displacement that is similar to the real infinitesimal displacement de-
noted dx, except that time is frozen for the virtual displacement. That
is, δx is imaginary, infinitesimal and instantaneous. Furthermore, it
satisfies any constraints on the system.
Now real work is defined in terms of forces and displacements and
we can write
X
dW = F·dr = Fx dx + Fy dy + Fz dz = Fi dxi .
i

Similarly, as you might suspect, the virtual work is given by

X
δW = Fi δxi .
i

Now since xi = xi (q1 , q2 , · · · qn , t) and by the rules of calculus

∂xi ∂xi ∂xi dxi
dxi = dq1 + dq2 + · · · dqn + dt,
∂q1 ∂q2 ∂qn dt
we conclude that the virtual displacement is given by
∂xi ∂xi ∂xi X ∂xi
δxi = δq1 + δq2 + · · · δqn = δqj .
∂q1 ∂q2 ∂qn j
∂q j

Note that the dt has disappeared because time is frozen during a virtual
displacement. Consequently, the virtual work can be expressed as
!
X X ∂xi X X ∂xi
δW = Fi δqj = Fi δqj .
i j
∂qj j i
∂qj
 4.6. GENERALIZED FORCE (OPTIONAL) 127

If we define the generalized force as

X ∂xi
Qj = Fi
i
∂qj
then we can write the virtual work in the form
X
δW = Qj δqj
j

and it has the same form as our usual expression for work, as a force
times a displacement.
Under certain circumstances, the generalized force can be obtained
from a scalar potential function V as
∂V
Qj = − .
∂qj
This relation is analogous to the well known relation between real
forces and the potential energy, F = −∇V.

Worked Example 4.6. Assume that for a certain system, the

generalized force Q can be obtained from the potential V according
to
∂V
Qi = − ,
∂qi
where the potential is assumed to be independent of velocity. Show
that for this situation, Lagrange’s equation can be written in the
so-called “Nielsen form”
d ∂T ∂T
− = Qi . (4.12)
dt ∂ q˙i ∂qi
Solution: According to the statement of the problem, the poten-
tial does not depend on velocity, so
∂L ∂ ∂T
= (T − V ) =
∂ q˙i ∂ q˙i ∂ q˙i
Similarly,
∂L ∂ ∂T ∂V ∂T
= (T − V ) = − = + Qi
∂qi ∂qi ∂qi ∂qi ∂qi
128 4. THE LAGRANGIAN METHOD

Therefore the Lagrange equation can be expressed as follows:

d ∂L ∂L
− = 0
dt ∂ q˙i ∂qi
 
d ∂T ∂T
− + Qi = 0
dt ∂ q˙i ∂qi
d ∂T ∂T
− = Qi .
dt ∂ q˙i ∂qi

4.7. The Calculus of Variations (Optional)

I am now going to discuss a branch of mathematics known as the
Calculus of Variations. I am primarily doing this so that you will un-
derstand how Lagrange’s equations are derived. However, the calculus
of variations is an interesting subject in its own right.
The calculus of variations considers problems involving maximums
and minimums for certain definite integrals. In doing so, it arrives at
answers to various interesting questions, such as: (1) What is the shape
of the curve of given length that encloses the largest area? (A circle.)
(2) What curve gives the shortest distance between two points in a
plane? (A straight line.) (3) What curve gives the shortest distance
between two points on a sphere? (A segment of a great circle.) (4)
What is the shape of a curve between two points, one higher than the
other, such that a bead slides from one point to the other in minimum
time? (A cycloid.)
In such problems you are asked to maximize or minimize some
quantity. The quantity itself will be expressed as an integral. You
may have to exert considerable cleverness to figure out this integral, as
it is usually not obvious.

y 2
ds
dy
dx
1
x
Figure 4.5. The quantity ds is an element of the path
from 1 to 2. Note that ds2 = dx2 + dy 2 .
 4.7. THE CALCULUS OF VARIATIONS (OPTIONAL) 129

Consider a specific example. Suppose you want to determine the

equation of the curve between two points in a plane such that the
distance along the curve is a minimum. (You already know the answer:
It is a straight line, and its equation is y R= mx + b.) Now the distance
2
between two points along a curve s is 1 ds. But ds2 = dx2 + dy 2
(Figure 4.5) so
s 2
Z 2 Z 2 Z x2 
p dy
I= ds = dx2 + dy 2 = 1+ dx
1 1 x1 dx (4.13)
Z x2 q
= 1 + (y 0 )2 dx.
x1
0
Here y ≡ dy/dx. (In the rest of this chapter, a prime will represent
differentiation with respect to x.) The integral I is the distance be-
tween points 1 and 2. The problem is to find the curve y = y(x) that
minimizes I.
I will come back to this problem in a moment, but first I want to
consider the calculus of variations in a general way. Problems in the
calculus of variations always lead to definite integrals of the form
Z x2 Z x2
dy
I= Φ(x, y, )dx = Φ(x, y, y 0 )dx.
x1 dx x1

The solution of the problem is the function y = y(x) that makes the
integral an extremum.7 Note that Φ is a function of a function be-
cause Φ is a function of y, and y itself is a function of x. Φ is called a
“functional” to distinguish it from an ordinary function.
Keep in mind that the integral I is a line integral between limits
x1 and x2 . It is the distance from point 1 to point 2 along a particular
path y = y(x). Obviously, if you use a different function y1 = y1 (x),
you will integrate along a different path and get a different value for
the integral.
Figure 4.6 illustrates the problem of the shortest distance between
two points in a plane. The three paths y1 (x), y2 (x), and y3 (x) have
different path lengths. The problem asks us to find the shortest of all
the possible paths between the fixed end points. Let us assume that
y0 (x) is the shortest path, and let us further assume that y1 (x) is a
path that only differs infinitesimally from y0 (x). That is, at each point
7By “extremum” we mean a maximum or a minimum, that is, values of a
function where the derivative is zero. The derivative of a function is also zero at an
inflection point, but in the calculus of variations we are always interested in either
the maximum or the minimum of the integral of a functional.
130 4. THE LAGRANGIAN METHOD

y y1(x)
y (x)
2

y3(x)

x x x
1 2

Figure 4.6. Three possible paths between two fixed end points.

on the path (except at the end points) y1 (x) is slightly different from
y0 (x). We can write the relationship between y1 (x) and y0 (x) in the
form
y1 (x) = y0 (x) + 1 η(x)
where 1 is a small quantity and η(x) is a function of x. Note that η(x)
is completely arbitrary except that it must be zero at the endpoints.
That is,
η(x1 ) = η(x2 ) = 0,
because all the paths between x1 and x2 must meet at the endpoints.
Remember that the problem is to find the minimal path y0 (x). I
will denote this minimal path simply by y(x) from now on. In general,
a nearby path is a member of a family of paths that are described by
the relationship
Y = Y (x, ) = y(x) + η (x) , (4.14)
where y(x) is the minimal path. Here  is taken as a continuous vari-
able. Different paths correspond to different values of . The minimal
path is obtained when  = 0.
The path length along any particular member of this family of
curves is
Z x2   Z x2
dY
I = I() = Φ x, Y, dx = Φ (x, Y, Y 0 ) dx.
x1 dx x1
Here the integral I is explicitly expressed as a function of . The desired
curve has  = 0; it is the curve that minimizes I. Therefore, the
condition on I is  
dI
= 0.
d =0
Observe that
d x2
Z x2 
∂Φ ∂Y 0
Z 
dI 0 ∂Φ ∂Y
= Φ(x, Y, Y )dx = + dx.
d d x1 x1 ∂Y ∂ ∂Y 0 ∂
 4.7. THE CALCULUS OF VARIATIONS (OPTIONAL) 131

From Equation (4.14)

∂Y
= η(x).
∂
Also,
∂Y 0
 
∂ dY ∂ 0
= = (y + η 0 ) = η 0 .
∂ ∂ dx ∂
If  = 0 then Y = y and Y 0 = y 0 , so
  Z x2  
dI ∂Φ ∂Φ 0
= η + 0 η dx = 0. (4.15)
d =0 x1 ∂y ∂y
The formula for integration by parts is
Z b Z b
b
udv = uv|a − vdu.
a a
Applying this to the second term in Equation (4.15) leads to
Z x2  x2 Z x2
∂Φ dη ∂Φ d ∂Φ
0
dx = 0
η − η dx.
x1 ∂y dx ∂y x1 x1 dx ∂y 0
But η = 0 at the limits x1 and x2 so the first term drops out, and
equation (4.15) becomes
Z x2  
∂Φ d ∂Φ
0= − ηdx.
x1 ∂y dx ∂y 0
Since η 6= 0 (except at the end points) this integral can be zero if and
only if the term in parenthesis is zero.8 That is,
∂Φ d ∂Φ
− = 0. (4.16)
∂y dx ∂y 0
This is called the Euler-Lagrange equation. It gives the condition on Φ
that minimizes the integral I.
To appreciate how to use the Euler-Lagrange equation, let us return
to the problem of the shortest distance between two points. We found
that for this problem the functional Φ is (Equation 4.13)
s  2
dy
1
Φ= 1+ = 1 + y 02 2 .
dx
8You might argue that
Z b
g(x)η(x)dx = 0
a
can be satisfied even if g(x) is not zero everywhere for some function η(x). But
that argument does not hold for an arbitrary function η(x).
132 4. THE LAGRANGIAN METHOD

Plugging this functional into the Euler-Lagrange equation (4.16) gives

∂
1 d ∂
1
1 + y 02 2 − 1 + y 02 2
= 0,
∂y dx ∂y 0
 
d 1
1
02 − 2 0
0− 1+y (2y ) = 0,
dx 2
" #
d y0
p = 0.
dx 1 + y 02
If the derivative of a function is zero, the function is constant, so

y0
p = const = C1 .
1 + y 02
Squaring both sides and rearranging,

y 02 = (C1 )2 (1 + y 02 ). (4.17)
02
Solving for y ,
y 02 (1 − C12 ) = C12
s
C12
y0 = = constant = m.
1 − C12
But y 0 = dy/dx, so
dy
= m.
dx
A final integration yields

y = mx + b,
the equation for a straight line! Thus we have proved that a straight
line is the shortest distance between two points. (Whew!)
4.7.1. Hamilton’s Principle. By this time you are surely won-
dering what any of this has to do with physics. The answer lies in
Hamilton’s principle. Hamilton’s principle states that the behavior
of any physical system will minimize the time integral of the
Lagrangian. In other words, a physical system that evolves over time
from t1 to t2 will follow a “path” q = q(t) such that the integral
Z t2
I= L(q, q̇, t)dt
t1

is minimized. The condition for this integral to be minimized is given by

the Euler-Lagrange equation of the calculus of variations. By replacing
 4.7. THE CALCULUS OF VARIATIONS (OPTIONAL) 133

q’s for y’s and t for x and changing the order of the terms, the Euler-
Lagrange equation (Equation 4.16) becomes
d ∂L ∂L
− = 0,
dt ∂ q̇ ∂q
which you recognize R as Lagrange’s equation!
The integral Ldt is called the “action.” Using that terminology,
Hamilton’s principle is:
The time development of a dynamical system will minimize the action.
Hamilton’s principle is sometimes referred to as, “The fundamental
principle of mechanics.”9
To make this a bit more explicit, consider a simple physical process,
such as the motion of a falling rock. This physical system can be
considered to evolve from some initial situation at time t1 to a different
final situation at time t2 . Initially the rock is at height h and has zero
velocity. Therefore the initial conditions are (q, q̇) = (h, 0). The√final
conditions (just before it hits the ground) are q = 0 and q̇ = 2gh.
Hamilton’s principle states that the behavior of the rock (its position
and
R t2 velocity at any instant of time) is such as to minimize the quantity
t1
L(q, q̇, t)dt. This leads to the variant of the Euler-Lagrange equation
called the Lagrange equation. Finally, the Lagrange equation generates
the equation of motion, which for the falling rock, is simply d2 q/dt2 =
−g.
4.7.2. Relation to Newton’s Second Law. As mentioned above,
Newton’s second law can be derived from the Lagrange equation. As
an illustration, consider the one-dimensional motion of a particle of
mass m. Assume the particle is acted upon by a conservative force F .
Recall that in one dimension, if F is conservative it can be obtained
from the potential energy V = V (x) by F = −dV /dx. The Lagrangian
is
1
L = T − V = mẋ2 − V (x),
2
and Lagrange’s equation is
    
d ∂L ∂L d ∂ 1 2 ∂ 1 2
0 = − = mẋ − V (x) − mẋ − V (x)
dt ∂ ẋ ∂x dt ∂ ẋ 2 ∂x 2
d ∂V (x)
= (mẋ) + = mẍ − F.
dt ∂x
9You may have studied Fermat’s principle in optics which states that the path
of a ray of light from one point to another is such as to minimize the time of flight.
This is a special case of Hamilton’s principle.
134 4. THE LAGRANGIAN METHOD

So
F = mẍ
as expected.
Thus, Newton’s second law is a consequence of Lagrange’s equations
and Lagrange’s equations are a consequence of Hamilton’s principle.10
I suppose you might conclude that Hamilton’s principle is more basic
and important than Newton’s laws. Nevertheless, it is true that New-
ton’s laws are the basic relations from which all of physics was derived.
Hamilton’s principle is a beautiful, sophisticated, and elegant way of
describing the way nature behaves, but it is not very easily understood.
Newton’s laws, on the other hand, can be grasped by anyone with an
elementary knowledge of mathematics.

4.8. Hamilton’s Equations (Optional)

In this section I will briefly discuss another formulation of the laws
of mechanics. First, however, I have to define a quantity called the
Hamiltonian. The Hamiltonian is a function of position and momen-
tum. For a single particle the one-dimensional Hamiltonian is defined
as:
H(p, q, t) = pq̇ − L(q, q̇, t), (4.18)
where p stands for the generalized momentum and q for the generalized
coordinate. Note that the Lagrangian is a function of velocity, position,
and time, L = L(q, q̇, t). The Hamiltonian is a function of momentum,
position and time, H = H(p, q, t). If you are asked to determine a
Hamiltonian, make sure your final expression does not explicitly con-
tain any velocities!
For more than one dimension and/or more than one particle the
Hamiltonian is: X
H= pi q̇i − L. (4.19)
i
More explicitly, for a system of N particles in three dimensions,

3N
X
H(p1 , p2 , · · · , p3N ; q1 , q2 , · · · q3N ; t) = pi q̇i −L(q1 , · · · q3N ; q̇1 , · · · q̇3N ; t).
i=1

10You might ask: “Why does nature always minimize the time integral of the
Lagrangian?” I do not know the answer to this question. I know that Hamilton’s
principle is an expression of the way the physical universe is put together. I know
that is how nature behaves, but I cannot tell you why it behaves that way. (Physics
tells you how, not why.)
 4.8. HAMILTON’S EQUATIONS (OPTIONAL) 135

Having obtained the Hamiltonian you can use it to obtain the equa-
tions of motion of the system. Recall that the equation of motion as
expressed by Newton’s second law as well as by Lagrange’s equation
is a second-order differential equation. On the other hand the Hamil-
tonian formulation yields two first-order equations for the momentum
and position. They are:
∂H ∂H
ṗi = − and q̇i = + . (4.20)
∂qi ∂pi
These are Hamilton’s equations of motion and they are completely
equivalent to Newton’s second law and to the Lagrange equations of
motion.
There are a number of different ways of deriving Hamilton’s equa-
tions. One way is to start with Equation (4.19) and write the differen-
tial of H as
X ∂L ∂L

∂L
dH = pi dq̇i + q̇i dpi − dqi − dq̇i − dt.
i
∂qi ∂ q̇i ∂t
Using the definition of generalized momentum to cancel the first and
last term in parenthesis, and using Lagrange’s equation to write
d ∂L d ∂L
= pi = ṗi =
dt ∂ q̇i dt ∂qi
we obtain
X ∂L
dH = (q̇i dpi − ṗi dqi ) − dt.
i
∂t
But the differential of H as obtained from H = H(pi , qi , t) is
X  ∂H ∂H

∂H
dH = dpi + dqi + dt.
i
∂pi ∂qi ∂t
A term-by-term comparison of the two expressions for dH yields Equa-
tions (4.20) as well as
∂H ∂L
=− .
∂t ∂t
For the sake of a specific example, consider the Hamiltonian of a
particle of mass m moving vertically in a uniform gravitational field
g. (For example, you might want to determine the Hamiltonian and
Hamilton’s equations for a ball of mass m thrown vertically upward.)
Begin with the Lagrangian. It is
1
L = T − V = mż 2 − mgz.
2
136 4. THE LAGRANGIAN METHOD

By the definition of generalized momentum,

∂L
p= = mż.
∂ ż
Therefore,
p
ż = .
m
Equation (4.18) then gives
p2
H = pż − L = pż − + mgz
2m
p p2
= p − + mgz
m 2m
p2
= + + mgz.
2m
Note that I was careful to replace ż by p/m. Inserting this Hamiltonian
into Hamilton’s equations yields the following equations of motion:

∂ p2
 
∂H
ṗ = − =− + mgz = −mg
∂z ∂z 2m
∂ p2
 
∂H p
q̇ = ż = + = + mgz = .
∂p ∂p 2m m
The second equation reads ż = p/m or p = mż which is just the
definition of momentum. The first equation states that ṗ = −mg or
d d
p = mż = mz̈ = −mg.
dt dt
That is, z̈ = −g, as expected.
Now
p2 m2 v 2 1
= = mv 2 = T,
2m 2m 2
and
V = mgz,
so
H = T + V = E = total energy.
Therefore, in this case, H = E. In fact, the Hamiltonian is nearly
always the total energy expressed in terms of generalized momentum
and position. Students frequently assume that H is always equal to
E. This is not true. The following conditions determine when H is
constant and when it is equal to E.
 4.8. HAMILTON’S EQUATIONS (OPTIONAL) 137

(1) If the Lagrangian does not depend explicitly on time, the

Hamiltonian is constant but not necessarily equal to the en-
ergy.
(2) If the transformation equations do not depend on time and
the potential energy does not depend on velocity, then the
Hamiltonian is equal to the total energy but it may not be
constant.
(3) If the constraints and the transformation equations and the
potential energy are all time independent, then the Hamilton-
ian is equal to the total energy and is constant.
The third condition is satisfied by many systems, and frequently
when requested to write the Hamiltonian (especially in quantum me-
chanical problems) a physics student will simply write H = p2 /2m + V.
However, this can be dangerous because there are problems in which
the Hamiltonian is not equal to the total energy. Fortunately, you prob-
ably will not encounter any of these situations in your undergraduate
courses in physics.
Students are often asked to determine the Hamiltonian for a partic-
ular system. A common error is to write an expression for H that in-
volves generalized velocities. This is not the way a Hamiltonian should
be written. The final expression for H must only contain momenta and
positions (and possibly the time) because H = H(p, q, t).
Finally, I would like to mention that if you transform to a new co-
ordinate system, you will have a new set of generalized momenta and
generalized coordinates, say Pi and Qi . If these new coordinates main-
tain the form of Hamilton’s equations unchanged, then they are called
“canonical conjugates.” Although we will not consider this subject any
further, I can assure you that you will be hearing much more about
canonical conjugates in your advanced courses.

Worked Example 4.7. (a) Write the Hamiltonian in polar

coordinates for a planet (mass m) in the gravitational field of a
star (mass M ). (Assume the planet is a particle moving in a planar
orbit. The star may be assumed to remain at rest.)
Solution: The Lagrangian for this system in polar coordinates
(r, θ) can be obtained from the fact that the velocity is
v = ṙr̂ + rθ̇θ̂
so
1 1  
T = mv 2 = m ṙ2 + r2 θ̇2 .
2 2
138 4. THE LAGRANGIAN METHOD

The potential energy is obtained from Newton’s Law of Universal

Gravitation
GM m
V =− .
r
Therefore,
1 1 GM m
L = T − V = mṙ2 + mr2 θ̇2 + .
2 2 r
The generalized momenta are
∂L
pr = = mṙ,
∂ ṙ
∂L
pθ = = mr2 θ̇.
∂ θ̇
The Hamiltonian is, therefore,
H = Σpi q̇i − L
= pr ṙ + pθ θ̇ − L
 
1 2 1 2 2 GM m
= pr ṙ + pθ θ̇ − mṙ + mr θ̇ + .
2 2 r
Replacing ṙ by pr /m and θ̇ by pθ /mr2 we obtain
pr pθ 1  pr 2 1 2  pθ 2 GM m
H = pr + pθ 2 − m − mr −
m mr 2 m 2 mr2 r
p2r p2θ 1 p2r 1 p2θ GM m
= + − − −
m mr2 2 m 2 mr2 r
2 2
1 pr 1 p θ GM m
= + 2
− .
2 m 2 mr r
Exercise 4.10. Write the Hamiltonian for a free particle. Answer:
2
p /2m.
Exercise 4.11. A student writes the Hamiltonian for a particle
falling in a uniform gravitational field as H = 21 mż 2 + mgz. What is
wrong with this?

4.9. Summary
To determine the equation(s) of motion do one of the following:
(1) Draw the free body diagram and apply F = ma. (That is,
determine the equation of motion by inspection.)
(2) Write the Lagrangian and apply Lagrange’s equations.
 4.9. SUMMARY 139

(3) Write the Hamiltonian and apply Hamilton’s equations.

The Lagrangian is given by
L = L(qi , q̇i , t) = T − V
and the Lagrange equations are
 
d ∂L ∂L
− =0; i = 1, 2, ..., n.
dt ∂ q̇i ∂qi
The qi and q̇i are called generalized coordinates and generalized veloc-
ities. The generalized momentum is defined by
∂L
pi = .
∂ q̇i
If the generalized coordinate qi does not appear in the Lagrangian, it is
called an ignorable coordinate. The generalized momentum conjugate
to an ignorable coordinate is a constant.
The number of generalized coordinates is equal to the number of
degrees of freedom. Each constraint reduces by one the number of
degrees of freedom (and hence, the number of generalized coordinates).
Lagrange’s equations are obtained from the calculus of variations.
The calculus of variations allows one to determine the extremum con-
ditions for a given parameter (such as time, distance, area, etc.). To
determine the function y = y(x) that will minimize or maximize the
quantity Z
Φ(x, y, y 0 )dx,
one applies the Euler-Lagrange equation
∂Φ d ∂Φ
− = 0.
∂y dx ∂y 0
The most difficult aspect of such problems is determining the functional
Φ(x, y, y 0 ).
Similarly, Lagrange’s equations are obtained from Hamilton’s prin-
ciple. This principle states that the time evolution of a mechanical
system will minimize the action, defined by
Z
Action ≡ Ldt.

Applying the technique of the calculus of variations to minimize the

action leads to Lagrange’s equations.
The Hamiltonian is defined by
X
H≡ pi q̇i − L.
140 4. THE LAGRANGIAN METHOD

The equations of motion (Hamilton’s equations) are

∂H ∂H
ṗi = − , and q̇i = +
∂qi ∂pi

4.10. Problems
Note: Problems 4.15 through 4.20 are based on optional section 4.7.
Problem 4.1. Determine the Lagrangian for a particle sliding on
the inner surface of a hemisphere of radius a (the “bar of soap in a
sink” problem). See Figure 4.7.

θ
ϕ y

x m

Figure 4.7. A particle of mass m slides on the inside

of a perfectly smooth hemispherical surface.

Problem 4.2. A bead of mass m, acted upon by a uniform gravita-

tional force, can slide freely on a frictionless circular hoop of radius R.
The hoop is oriented vertically and is rotating about a vertical axis with
constant angular velocity ω. Assume the hoop is massless. Determine
the Lagrangian for the bead.
Problem 4.3. A certain molecule is made up of four atoms. Three
of them have mass M and are located at the vertices of an equilateral
triangle of side a. The fourth atom has mass m and is free to move along
a line perpendicular to the triangle and passing through its center. As-
sume this atom is attracted to the other atoms by a force proportional
to the distance between them. (You can imagine the atom of mass m
is attached by three springs of force constant k to the atoms of mass
M.) Determine the Lagrangian for the system.
 4.10. PROBLEMS 141

Problem 4.4. A heavy ring of mass m can slide on a frictionless

rod. The rod is attached to a wall at one end with a bracket and the
other end hangs downward, so the constant angle between the rod and
the wall is α. One end of a massless spring of constant k is attached to
the bracket and the other end is connected to the ring. Determine the
Lagrangian for the system and the equation of motion. See Figure 4.8.
y x
s

m
Figure 4.8. The ring connected to the spring is free to
slide on the inclined rod.

Problem 4.5. A cart of mass M slides freely (no friction) on el-

evated horizontal tracks. Hanging from the cart is a pendulum with
string of length l and bob of mass m. Determine the Lagrangian for
this system. Obtain the equations of motion. By considering whether
or not any coordinates are ignorable, determine any constants of the
motion. See Figure 4.9.

θ l
m

Figure 4.9. A cart sliding freely on raised rails has a

pendulum hanging from it.

Problem 4.6. Obtain the equations of motion for a spherical pen-

dulum, that is, a pendulum that is not constrained to oscillate in a
plane.
Problem 4.7. A wedge of mass M and angle α is resting on a
frictionless horizontal plane. A rectangular box of mass m is on the
142 4. THE LAGRANGIAN METHOD

wedge. Since all surfaces are frictionless, the box will slide down the
wedge and the wedge will slide in the opposite direction on the plane.
Determine the accelerations of the two bodies.
Problem 4.8. A system of two masses and three springs is illus-
trated in Figure 4.8. Write the Lagrangian. Determine the generalized
momenta for this system. (Hint: Measure the positions of the masses
from their equilibrium points.)

k k k
1 2 3
M M
1 2

Problem 4.9. A toy cart is made up of a block of mass M and

four cylindrical wheels of radius R and mass m. It is rolling down a
fixed plane of angle α. Let the cart have a length d and let the wheels
be at positions ±d/2 relative to the center of mass of the cart. Assume
the center of mass lies on a line joining the centers of the wheels. See
Figure 4.10. Determine the Lagrangian for the system.

Figure 4.10. A block with four wheels rolls down a

plane. Problem 4.9.

Problem 4.10. A rope passes over a frictionless pulley. A monkey

of mass M is hanging from one end of the rope and a bunch of bananas
of mass m is attached to the other end. Since M > m, the monkey
is descending and the bananas are ascending and getting further and
further away from the monkey. Naturally, the monkey begins to climb
the rope. See Figure 4.11. You are asked to obtain the motion (x =
x(t)) for the monkey under two different assumptions. (a) Assume the
monkey climbs at a constant velocity relative to the rope. (b) Assume
the monkey exerts a constant force on the rope.
Problem 4.11. A wedge of mass M and angle α is free to slide on
a frictionless horizontal surface. On the wedge is mounted a mass and
spring arrangement that allows the mass m to slide up and down on
 4.10. PROBLEMS 143

Figure 4.11. The monkey and the bananas ( Problem 4.10).

the wedge. All surfaces are frictionless and the spring is massless. (a)
Write the Lagrangian for the system. (b) Obtain expressions for the
generalized momenta. (c) Obtain the equations of motion. See Figure
4.12.

X y
s m
M α
Figure 4.12. A block and spring on a sliding wedge.
The dot in the middle of the spring represents the equi-
librium position of the block, and is a distance X from
some fixed point. The stretch of the spring is measured
from the equilibrium position. The position of the block
is given by x and y where x = X + s sin α.

Problem 4.12. A pendulum is made of a bob of mass M but

instead of a string, the bob is suspended from a massless spring of
constant k. The system can swing back and forth and also move up
144 4. THE LAGRANGIAN METHOD

and down, but it is restricted to motion in a plane. Determine the

equations of motion.
Problem 4.13. Figure 4.13 shows two equal masses m. One is
hanging and the other is on a frictionless surface. They are connected
by an ideal horizontal string that passes over an ideal massless pulley.
The mass on the surface is connected to a fixed point by a massless
spring of constant k. Place the origin of coordinates at the pulley.
(a) Write the Lagrangian for the system in terms of a single variable.
(b) Obtain the equation of motion. (c) Determine the frequency of
oscillation of the system.

k
m

Figure 4.13. Two masses, a frictionless surface, a

spring, and an ideal pulley.

Problem 4.14. A marble of mass M and radius R rolls back and

forth on the bottom of a semi-circular bowl of radius a. (a) Write the
Lagrangian for the system. (b) Obtain the equation of motion. (c)
Determine the period with which the marble rolls back and forth. You
may assume the amplitude of the oscillations is small.
Problem 4.15. It is desired to find the equation for the shortest
distance between two points on a sphere. Determine the functional for
this problem. (Use spherical coordinates.)
Problem 4.16. The speed of light in a medium of index of refrac-
tion n is v = c/n = ds/dt. The time for light to travel from point A to
point B is
Z B
ds
.
A v
Obtain the law of reflection and the law of refraction (Snell’s law) by
using Fermat’s principle of least time.
 4.10. PROBLEMS 145

Problem 4.17. Determine the relation y=y(x) such that the fol-
lowing integral is an extremum. What is the shape of the curve?
Z x2
1/2
(x)(1 + y 02 )

dx.
x1

Problem 4.18. A bead slides down a wire. You want to determine

the shape the wire must have for the bead to slide from its initial
position to its final position in the least amount of time. (This is
called the “brachistochrone” problem.) Note that the quantity to be
minimized is the time and that dt = ds/v where the velocity v is given
by conservation of energy. I do not expect you to actually solve this
problem; it will be sufficient if you write the functional that describes
the system.11
Problem 4.19. Two pegs are at the same height above the ground.
A rope is draped over them, with both ends of the rope resting on the
ground. We wish to know the shape of the rope between the pegs.
The quantity that is minimized in this case is the potential energy.
Determine the appropriate functional.
Problem 4.20. A vase is designed by drawing a curve and then
rotating it about the z-axis, forming a “surface of revolution.” It is
desired that the vase have the minimum possible surface area. Obtain
the functional for this problem. See Figure 4.14 to get an idea of the
geometry.

axis

ds Area of shaded
Curve of
region is
length s
2πxds
Figure 4.14. A curve of length s is revolved about an
axis to generate a surface of revolution.

11This problem was apparently first devised by one the Bernoulli brothers who
sent it out as a “challenge problem” to the scientists of that day. It is said that
Newton solved the problem in a few hours. (The calculus of variations had not
been invented yet!)
146 4. THE LAGRANGIAN METHOD

Problem 4.21. In quantum mechanics, one replaces the momen-

h ∂
tum p by the “momentum operator” i 2π ∂x
. Using this replacement,
write the quantum mechanical Hamiltonian operator for a free particle.
(The quantity h/2π is usually written as ~.)
Problem 4.22. Obtain the Schrödinger equation as follows: Write
the quantum mechanical version of the Hamiltonian for a particle whose
potential energy is V (x). Let H = E, and allow the operators to act
on the function Ψ(x).
Problem 4.23. Write the Hamiltonian for a planar pendulum and
obtain the Hamilton’s equations of motion for the system.
Problem 4.24. Write the Hamiltonian for a mass on a spring (as
in Figure 4.3) and obtain the Hamilton’s equations of motion for the
system.

Computational Projects
Computational Techniques: To solve the first computational prob-
lem below you will have to use a Computer Algebra System (CAS) such
as Mathematica or Maple. We do not have the space nor the time to
get into the details of these excellent programs, but you should learn
some of the basics of one such system. They are easily available. For
example, Matlab incorporates symbolic computation using the Maple
kernel. (Maple is a CAS that was developed primarily at the University
of Waterloo in Canada and more recently at ETH in Zurich - Einstein’s
Alma Mater.) In Matlab’s version of Maple, if you want to obtain, say,
an expression for y = cos2 x + sin2 x you simply declare x, y and z as
symbolic variables, using the command
syms x y z
then write
y = cos(x)ˆ2+sin(x)ˆ2
Then write
simplify(y)
You will obtain (of course) y=1.
Another example, is to write
z = sin(x)/cos(x)
then write
simple(z)
 4.10. PROBLEMS 147

to get
z = tan(x)
Of course, you will have to consult the “help” files to get further infor-
mation, but let me assure you that using Maple is not difficult.
Computational Project 4.1. Use a symbolic manipulator such
as Maple to obtain the Hamiltonian for the double planar pendulum.
Computational Project 4.2. Solve for the motion of the double
pendulum. Assume the two bobs have the same mass and that the two
strings are of length 0.1 m. Compute examples of the motion for several
initial conditions.
Computational Project 4.3. A 2 kg object is launched hori-
zontally from a 5 m cliff with an initial velocity of 20 m/s. As you
know, it will follow a parabolic path. It hits the ground a distance 20.2
m horizontally from the launch point and will have been in the air for
1.01 s. The purpose of this project is to evaluate
R the action for this
process. The action is defined by the integral Ldt where L = T − V.
After you have written a computer program to evaluate the action for
the actual path of the projectile, evaluate the action for the following
hypothetical path between the same two end points and taking the
same amount of time. The hypothetical path is a straight line from
(0,5) to (20.2,0) at a constant speed of 20.6 m/s. You will find that the
action for the hypothetical path is greater than the action for the actual
path since the actual path for any dynamical system always minimizes
the action. You may wish to construct other paths between the end
points and show that their action is greater than that for the actual
path. (Suggestion: Use small time steps [about 1/1000 of a second],
evaluate L at each time step and estimate the integral by summing the
quantities L(t)∆t.)
Chapter 5
The Conservation of Energy

This chapter analyzes the conservation of energy in a much more

rigorous manner than the review in Section 1.3. Here you will also be
exposed to a number of other useful concepts such as energy diagrams,
the concept of metrics, and the del operator. The techniques presented
here are used quite frequently, not only in this course, but also in other
advanced physics courses.

5.1. The Work-Energy Theorem

When you push an object from one place to another, you do work
on it. To be precise, the work done on a particle by a force F during a
displacement from point 1 to point 2 along some given path, is defined
as Z r2
W = F·ds, (5.1)
r1
where F is the force acting on the particle during the displacement
ds. If there is no displacement, no work is done. By definition, the
displacement ds is the displacement of the point of application of the
force. For a particle, this is equal to the displacement of the particle.
For an extended rigid body ds refers to the displacement of the center
of mass.1
1The definition of work is often written as
Z
W = F·dr.

This definition may cause some confusion because dr represents the displacement
of the body, but it is not necessarily in the r̂ direction. It is wrong to write dr =drr̂
unless the displacement happens to be in the r̂ direction. For example, in spherical
coordinates,
dr = drr̂+rdθθ̂ + r sin θdφφ̂.
149
150 5. THE CONSERVATION OF ENERGY

It is reasonable to assume that doing work on a body will increase

its kinetic energy. For an extended body, this would include the kinetic
energy of rotation as well as the kinetic energy of translation. Since a
particle cannot rotate, it is easiest to consider, for the time being, the
work done on a particle.
The relation between work and kinetic energy is expressed by the
work-energy theorem:
Work-Energy Theorem. The increase in kinetic energy
of an object is equal to the total work done on it.

This theorem refers to the total or net work done by all the forces
acting on the particle.
The proof of this theorem is very simple. Let F be the net force
acting on a particle of mass m. Then using the definition v =ds/dt
and Newton’s second law in the form F = m dv dt
, we can write
Z r2 Z t2 Z t2
dv
W = F·ds = F · v dt = m · v dt.
r1 t1 t1 dt
Now
d 2 d dv dv dv
(v ) = (v · v) = · v + v· =2 · v.
dt dt dt dt dt
Therefore  
dv 1 d 2 d 1 2
m · v = m (v ) = mv .
dt 2 dt dt 2
So
Z r2 Z t2   Z T2
d 1 2
W = F·ds = mv dt = d(T ) = T2 − T1 = ∆T,
r1 t1 dt 2 T1

and the theorem is proved.

There are two things in this proof that I would like you to note.
One is the fact that
d 2 dv
v =2 · v.
dt dt
Keeping this relationship in mind will help you in solving a number
of problems. The other thing I would like you to be aware of is the
way the limits of integration change when the variable of integration
changes. Be sure to change the limits whenever you make a change of
variables!
To avoid confusion, in this chapter I will use ds for the infinitesimal vector
displacement.
 5.2. WORK ALONG A PATH. THE LINE INTEGRAL 151

5.2. Work Along a Path. The Line Integral

Equation (5.1) states that the work done by a force on a particle as
it is displaced from one point to another depends on the path between
the two points. In some cases, the work does not depend on the path;
in those cases the force is called “conservative” and has certain special
properties that we will study shortly. In general, however, the work
done along one path is different from the work done along some other
path. Therefore, to evaluate work, you will often have to evaluate a
path or line integral.
An example of a line integral is the work done by a force F as a
particle moves along a path described by the curve C:
Z
W = F·ds.
C
I will describe two ways to evaluate line integrals. The first is to
express the force and the displacement in Cartesian coordinates. Thus:
F = Fx ı̂+Fy ̂+Fz k̂
and
ds =dxı̂ + dy̂ + dz k̂.
So Z Z Z Z
W = F·ds = Fx dx + Fy dy + Fz dz. (5.2)
C C C C
Here the three integrals are evaluated along the curve C. Note that in
the first integral the force component Fx must be expressed in terms
of x, and likewise Fy is expressed in terms of y and Fz in terms of z,
where the relations between x, y, and z are determined by the path.

Worked Example 5.1. A particle is acted upon by the force

F = a(x + y)ı̂ + b(y − x)̂
Determine the work done by this force as the particle moves along
a straight line from the origin to the point (2, 1).
Solution: Using the definition of work,
Z Z Z
W = F·ds = Fx dx + Fy dy
ZC C
Z C

= a(x + y)dx + b(y − x)dy

C C
152 5. THE CONSERVATION OF ENERGY

Along the straight line, the relation between x and y is y = mx

where m = 1/2. In the first integral, replace y by x/2 (thus express-
ing y as a function of x) and in the second integral write x = 2y
(thus expressing x as a function of y). Then,
Z 2 Z 1
1
W = a(x + x)dx + b(y − 2y)dy
x=0 2 y=0
Z 2 Z 1
3
= axdx − bydy
x=0 2 y=0
1
= 3a − b.
2
The technique used in this example requires having an equation for
the curve in Cartesian coordinates. Obviously this procedure can
immediately be generalized to other coordinate systems.
It is interesting to note that if the path is first along the x axis
to (2,0) then up along the y axis to (2,1) we have the same initial
and end points, but the work done is quite different.
Z 2 Z 1
3
W = a(x + 0)dx + b(y − 2)dy = 2a − b.
0 0 2

The second approach is, in principle, the same as the first approach,
but it may be easier to apply in certain situations. This approach is
based on the fact that for motion along a smooth curve, the position
can be specified by a single variable, say λ. This single independent
variable can be the distance along the curve from some starting point,
or the time, or some other parameter (as the angle θ in the exampleR
below). Then writing both F and ds in terms of λ, the integral F·ds
reduces to a single integral over one variable. Thus, if F = F(λ) and
s = s(λ), then
Z Z
ds
W = F·ds = F(λ)· dλ.
C C dλ
The last integral can be expressed in a variety of ways. For example,
if F is given in Cartesian coordinates, we write
Z λ2  
dx dy dz
W = Fx + Fy + Fz dλ. (5.3)
λ1 dλ dλ dλ
This expression shows the close relationship to the first method. Nev-
ertheless, the second method is often more convenient, as illustrated
by the following example.
 5.2. WORK ALONG A PATH. THE LINE INTEGRAL 153

Worked Example 5.2. Consider the force

−y x
F= 2 2
ı̂ + 2 ̂.
x +y x + y2
R
Evaluate C F·ds along a semicircular path from (-1,0) to (+1,0),
as shown in the figure below.
Solution: The shape of the path suggests using polar coordi-
nates and the angle θ as a parameter. Since the radius of the semi-
circle is unity, we have x = cos θ, and y = sin θ and x2 + y 2 = 1.
Note that in terms of θ, Fx = − sin θ/(sin2 θ + cos2 θ) and similarly
for Fy .
Z θ=0   Z 0
dx dy
sin2 θ + cos2 θ dθ


W = Fx + Fy dθ =
θ=π dθ dθ π
Z 0
= dθ = −π
π

Y
+1

X
-1 +1

Figure 5.1. A semicircular path from (-1,0) to (1,0).

Worked Example 5.3. A particle is in a region of space where

the force on it is given by
F = xyı̂ − x2 ̂.
The particle is dragged along a path that is described by
s = s(t) = 3tı̂ + 2t2 ̂,
where t is the time. The force is in Newtons and the displacement
is in meters. Determine the work done by the force during the
time interval 0 ≤ t ≤ 1. Do this two ways, first by treating t as
a parameter (as in the second method described above), and then
by evaluating the work with equation (5.2). The force field and
the trajectory are shown in Figure 5.2. Note that as the particle
moves along the trajectory the force on it due to the force field is
154 5. THE CONSERVATION OF ENERGY

changing. We are not considering the work done by the force that
is dragging the particle through this field.
Solution: Treating t as a parameter, we write F and ds in
terms of t. Since s=xı̂+y̂ we appreciate that x = 3t and y = 2t2 .
Furthermore,
ds
ds = dt = (3ı̂ + 4t̂)dt
dt
Then the force can be expressed in terms of t as
F = (3t)(2t2 )ı̂ − (3t)2 ̂ = 6t3 ı̂ − 9t2 ̂.
Consequently,
Z Z t=1
3 2
W = F·ds = (6t ı̂ − 9t ̂) · (3ı̂ + 4t̂) = (18t3 − 36t3 )dt
t=0
18  4 1 9
=−t 0 = − J.
4 2
Now let us solve the problem using equation (5.2). Note that
we need to have expressions for dx and for dy. Since x = 3t and
y = 2t2 we can solve for t and obtain y as a function of x; that is,
t = x/3, so y = 2(x/3)2 = (2/9)x2 . Then
Z 3 Z 2 Z 3 Z 2 Z 3   Z 2 
2 2 2 9
W = Fx dx + Fy dy = xydx − x dy = x x dx − y dy
x=0 y=0 0 0 0 9 0 2
 3  2
2 x4 9 y2 81 36 9
= − = − =− J
9 4 0 2 2 0 18 4 2
Obviously, the two methods yield the same answer.

Force Field

Figure 5.2. The force field as a function of position for

the problem considered in Worked Example 5.3.
 5.3. POTENTIAL ENERGY 155

Exercise 5.1. A charged particle in an electric field E is acted

upon by a force qE. Assume the field is constant and evaluate the
work done by this force for the closed path illustrated in Figure 5.3(a).
Does the result depend on whether you traverse the path clockwise or
counterclockwise? Answers: zero, no.

y y
(a) (b)
18
2

x x
0 1 4 0 3

Figure 5.3. (a) A closed path. The corners are at

(1,0), (1,2), (4,2) and (4,0). (b) A parabolic path going
through the origin and the point (3,18).

Exercise 5.2. Evaluate the work done by the force F = yı̂+x̂

N along the parabolic path illustrated in Figure 5.3(b) as the particle
moves from (0,0) to (3,18) meters. Answer: 54 J.

5.3. Potential Energy

Forces are vectors. Although vectors were only invented about 100
years ago, it is hard to imagine trying to do physics without them.
Nevertheless, it is often much easier to deal with scalar quantities. As
you probably know, there are certain forces that can be represented
in terms of a scalar quantity called the potential energy. Examples of
forces that can be represented in terms of potential energies are the
electrostatic force, the gravitational force and the force exerted by a
spring. Scalars will usually make your life much simpler, so it is a good
idea to use them whenever possible.
For example, the force on a body of mass m at a height z near
the surface of Earth is F = −mg. If we define the scalar function
V = −mgz we can write the magnitude of the force as
dV
F =− .
dz
156 5. THE CONSERVATION OF ENERGY

As a second example, consider a stretched spring acting on an object

of mass m as illustrated in Figure 4.3. The force exerted by the spring
is F = −kxı̂. If V = 12 kx2 , then
dV
= kx = −F,
dx
and
dV
F =−
.
dx
As a third example, consider the electrostatic force:
Q1 Q2 r̂
F= .
4π0 r2
Q1 Q2 r̂
We can define V = 4π0 r
such that the electrostatic force is given by
dV
F =−
.
dr
You appreciate that in each case the derivative of V is the negative
of the magnitude of the force. (As you probably realized, the quantity
V is the potential energy. But don’t worry about the physical meaning
of V just yet. I’ll get to that pretty soon. Just consider V to be a
function whose derivative is related to the force.)
Replacing F by −dV /dr or −dV /dx in the examples above only
gives the magnitude of the force but not its direction. Is there some
way to obtain both the magnitude and the direction of the force from
the potential energy? The answer is yes. However, before continuing
with the discussion, I need to make a small digression to consider the
properties of the mathematical object called “del.”

5.3.1. The Del Operator. The del operator2 is defined as

∂ ∂ ∂
∇ ≡ ı̂ + ̂ + k̂ . (5.4)
∂x ∂y ∂z
Del is an operator, that is, it is a mathematical object that acts on a
function to produce a different function.
We shall be interested in three different kinds of operations involv-
ing del, and these are:
∇f The gradient of a scalar function f
∇ · F The divergence of a vector function F
∇ × F The curl of a vector function F.

2The symbol ∇ is also called “nabla”

 5.3. POTENTIAL ENERGY 157

Note that the gradient will generate a vector, the divergence (be-
ing a dot product) will generate a scalar, and the curl (being a cross
product) will generate a vector.

5.3.2. The Gradient. First consider the gradient, ∇f, where f =

f (x, y, z) is a scalar function. Using the definition of del, we have
∂f ∂f ∂f
gradient of f = grad f = ∇f = ı̂ + ̂ + k̂ . (5.5)
∂x ∂y ∂z
For example, if f = f (x, y, z) = 3x + 2y 2 then
 
∂ ∂ ∂
3x + 2y 2 = 3ı̂+4y̂


grad f = ı̂ + ̂ + k̂
∂x ∂y ∂z
The gradient of the scalar function f = 3x + 2y 2 produced the vector
3ı̂+4y̂. The fact that taking the gradient of a scalar function generates
a vector is one of the most important properties of the gradient.
The geometrical interpretation of the gradient is quite interesting.
To best appreciate it consider a specific example. Figure 5.4 is a contour
map, as on a regular geographic map, and it represents the shape of a
region containing two mountains. The curves on the figure are contour
lines or altitude “isopleths,” that is, lines drawn through points having
the same value of height or altitude.3 The map in our figure shows
two peaks, one higher than the other. You could (in principle at least)
determine an equation of the form h = h(x, y) that would give the
height above sea level at any point (x, y) on the map.
If you were at some spot on the mountain, say at the point marked
A, you could climb towards the peak directly (a very steep climb) or
you could walk along the isopleth passing through A. That would be a
level walk. If you walk along a contour line your altitude above sea level
does not change (dh = 0). By looking at the sketch, you can figure out
which of the three arrows points along the path of steepest ascent. It
is, of course, the arrow pointing in the direction in which the altitude
isopleths are closest together so that in moving a given distance in
that direction, you cross the largest number of isopleths. This is the
direction of the gradient, ∇h. The gradient does not necessarily point
toward the peak of the mountain. (What is the direction of ∇h at
point B?)

3In general an isopleth is a line drawn on a map through all points hav-
ing the same numerical value. You are familiar with the temperature isopleths
(“isotherms”) and pressure isopleths (“isobars”) drawn on weather maps.
158 5. THE CONSERVATION OF ENERGY

A
x

Figure 5.4. Contour plot of two mountains(altitude isopleths).

It is not difficult to prove that the gradient ∇h is a vector point-

ing in the direction of steepest ascent and whose magnitude gives the
“steepness” along this path of steepest ascent.4
Let us begin by considering the differential of an arbitrary scalar
function. You should recall from your calculus class that the differential
of a function of several variables, such as u = u(x, y, z), is given by
∂u ∂u ∂u
du = dx + dy + dz.
∂x ∂y ∂z
Keep this expression for du in mind for a moment. Now consider the
dot product of ∇u with the infinitesimal displacement vector ds. Since
ds = dxı̂ + dy̂ + dz k̂,
the dot product is
  
∂u ∂u ∂u 
∇u · ds = ı̂ + ̂ + k̂ · dxı̂ + dy̂ + dz k̂ ,
∂x ∂y ∂z
∂u ∂u ∂u
= dx + dy + dz.
∂x ∂y ∂z
But this is precisely the expression for du. Therefore,
du = ∇u · ds. (5.6)
The relationship given by equation (5.6) is so important that some
people consider it to be the definition of du. (I guess it just depends
4By “steepness” I mean the change of h for a given displacement.
 5.3. POTENTIAL ENERGY 159

on whether you think of du or ∇u as the more fundamental quan-

tity.) I will refer to relationship (5.6) as “the vector definition of the
differential.”
By Equation (5.6)
Z 2 Z 2
∇u · ds = du = u2 − u1 ,
1 1

which is the change in u due to the displacement from point 1 to point

2.
∇u and ds are both vectors. Their dot product, du, is a function
of the angle between the vectors ∇u and ds. From the definition of
dot product, if ds is perpendicular to ∇u, then ∇u · ds = 0, and so,
du = 0. Since ds is a displacement, this tells you that for a displacement
perpendicular to ∇u there is no change in u. Therefore, the lines of
constant u (contour lines) must be perpendicular to ∇u because a
displacement along a contour line results in no change in the value
of the function. (Actually, I should say surfaces of constant u are
perpendicular to ∇u because we have defined u = u(x, y, z) to be a
three dimensional function.)
Similarly, the largest change in du = ∇u · ds comes when the dis-
placement ds is parallel to ∇u, that is, in the direction of ∇u. The
largest change in du for a given displacement is along the path of steep-
est ascent, so ∇u points in the direction of steepest ascent. Further-
more, du = |∇u||ds| cosθ, and the maximum value of du occurs when
cos θ = 1, so  
 du 
 
 ds  = |∇u| .
max
That is, the magnitude of ∇u gives the value of the greatest increase
in u per unit displacement.
To summarize the preceding discussion you could state that the
direction of the gradient indicates the direction of greatest change of a
function and the magnitude of the gradient gives the maximum change
in the function per unit displacement.

Exercise 5.3. The surface of a sphere is described by x2 +y 2 +z 2 =

constant. Show that the gradient of this function points in the radial
direction.
Exercise 5.4. Evaluate the gradient of f = 5x3 +6y 2 +2z at (1,2,3).
Answer: 15ı̂+24̂+2k̂.
160 5. THE CONSERVATION OF ENERGY

Exercise 5.5. A family of coaxial cylinders is described by

f (x, y, z) = x2 + y 2 = constant.
Show that ∇f points away from the axis.
Exercise 5.6. The pressure in a body of still water is given by
P = ρgz + P0 where P0 is atmospheric pressure and z is the depth
below the surface. Show that the direction of maximum increase in
pressure is straight down. (If you got the wrong sign it is because you
did not consider that z is measured positive downward!)

5.3.3. The Relationship between Force and Potential En-

ergy. As described above, the potential energy is a function whose
derivative with respect to position gives the magnitude of the force.
But the expressions F = −dV /dr and F = −dV /dx only gave the
magnitude of the forces, but not their direction. The description is
not complete. It turns out (as you probably suspect by now) that the
gradient of the potential energy is related to the force. Specifically,
the force is equal to the negative gradient of the potential energy. In
symbols,
F = −∇V. (5.7)
This equivalence means you can work with the scalar function V in-
stead of the vector function F. This is a great advantage in solving
problems because it simplifies the calculations significantly. (You prob-
ably recollect from your introductory electricity class that determining
the electric field E by direct application of Coulomb’s law required
resolving E into components and then integrating over the charge dis-
tribution. You soon discovered it was much easier to first evaluate
the scalar function V, the electric potential, and then determine E by
differentiation.)
Not all forces can be represented by a potential energy function.
Forces that do not have a potential energy associated with them are
called nonconservative forces. Fortunately, most of the important forces,
such as the gravitational force and the electrical force, are conserva-
tive. All forces acting on the molecular and atomic level are believed
to be conservative. On the other hand, forces such as friction and air
resistance are nonconservative and do not allow one to take advantage
of the simplifications afforded by the potential energy.
How can you determine whether or not a force is conservative? The
answer is quite simple. A force is conservative if and only if its curl is
zero.
 5.3. POTENTIAL ENERGY 161

F is conservative iff ∇ × F = 0.
This statement is easy to prove. If F is conservative, then it can
be represented by the negative gradient of the potential energy, i.e.,
F = −∇V. Consequently,
∇ × F = ∇ × (−∇V ) = −∇ × (∇V ) = −curl ( grad V ) .
But the curl of the gradient of any function is zero. Therefore, if we
can write F as the gradient of V, then ∇ × F = 0.
For the sake of completeness and because the steps in the proof are
important to understand, I will now prove that curl grad V = 0 for
any function V . Using Cartesian coordinates,
   
∂ ∂ ∂ ∂V ∂V ∂V
∇ × ∇V = ı̂ + ̂ + k̂ × ı̂ + ̂ + k̂
∂x ∂y ∂z ∂x ∂y ∂z
 

 ı̂ ̂ k̂ 

=  ∂/∂x ∂/∂y ∂/∂z 
 ∂V /∂x ∂V /∂y ∂V /∂z 
 2
∂ 2V
 2
∂ 2V
 2
∂ 2V
  
∂ V ∂ V ∂ V
= ı̂ − − ̂ − + k̂ − .
∂y∂z ∂z∂y ∂x∂z ∂z∂x ∂x∂y ∂y∂x
But
∂ 2V ∂ 2V
= ,
∂y∂z ∂z∂y
because the order of taking partial derivatives does not affect the an-
swer. Therefore, each term in parenthesis is zero and
∇ × ∇V = 0.
Thus, we have proved that
curl grad (any scalar function) = 0.
5.3.4. Del in Other Representations. So far we have been us-
ing Cartesian coordinates, but it is often useful to work in some other
set of coordinates, such as cylindrical coordinates or spherical coordi-
nates. I am going to show you how to obtain representations for ∇ in
these coordinate systems. However, before doing so we must discuss
coordinate transformations.5
Coordinate Transformations: Suppose you want to transform
from the Cartesian coordinates x, y, z to a different coordinate system
5If you would like to have more information on this subject, an excellent refer-
ence is Chapter 2 of Mathematical Methods for Physicists, 5th Ed.” Arfken G, and
Weber H, Academic Press, NY, 2001.
162 5. THE CONSERVATION OF ENERGY

in which the coordinates are q1 , q2 , q3 . In the Cartesian coordinate sys-

tem the location of a point is specified by the intersection of three
planes, x = constant, y = constant and z = constant, as shown in Fig-
ure 5.5(a). For example, the point (1,2,0) is located at the intersection
of the planes x = 1, y = 2, z = 0.

z
y=const q 3=const
x=const q2=const
y t
z=const q 1=cons

x
(a) (b)
Figure 5.5. (a) A Cartesian coordinate system. (b) A
general coordinate system.

In Figure 5.5(b) I generalized the situation and specified the loca-

tion of the point by the intersection of three surfaces (q1 = constant,
q2 = constant, q3 = constant). Although it is not necessary, it is usu-
ally convenient to consider only orthogonal coordinate systems. For an
orthogonal coordinate system the surfaces are always perpendicular at
the lines of intersection. (All of the coordinate systems used in this
book are orthogonal.)
You realize, of course, that there must be a set of transformation
equations relating the Cartesian coordinates to the “new” coordinates.
The transformation equations can be written as:
x = x(q1 , q2 , q3 ),
y = y(q1 , q2 , q3 ),
z = z(q1 , q2 , q3 ).
The inverse transformations are:
q1 = q1 (x, y, z),
q2 = q2 (x, y, z),
q3 = q3 (x, y, z).
 5.3. POTENTIAL ENERGY 163

Just as ı̂,̂, k̂ are defined as the unit vectors of the Cartesian system, so
too you can define the quantities ê1 , ê2 , ê3 as the unit vectors of the new
coordinate system. These are defined so that ê1 points in the direction
of increasing q1 and is perpendicular to the surface q1 = constant, and
similarly for ê2 and ê3 .
The distance between two nearby points is ds. In the Cartesian sys-
tem ds is related to dx, dy, dz by the generalized Pythagorean relation
ds2 = dx2 + dy 2 + dz 2 .
If you express this distance in terms of the new coordinates you will
get a fairly complicated expression that reduces to the form
X
ds2 = h2ij dqi dqj i, j = 1, 2, 3. (5.8)
ij

To see how this comes about, note that since x is a function of q1 , q2 ,

and q3 , you can write
∂x ∂x ∂x
dx = dq1 + dq2 + dq3 .
∂q1 ∂q2 ∂q3
Squaring this expression leads to
 2      
2 ∂x 2 ∂x ∂x ∂x ∂x
dx = dq1 + dq1 dq2 + dq1 dq3 +···.
∂q1 ∂q1 ∂q2 ∂q1 ∂q3
Since Problem (5.4) asks you to go through this process, I did not write
out the whole expression, which has nine terms (some terms are the
same, so there are only six different terms). You can obtain similar
expressions for dy 2 and dz 2 . Each term involves a product of the form
dqi dqj . If you add and combine the three terms dx2 + dy 2 + dz 2 to
get ds2 you will obtain a sum of nine different terms involving dqi dqj
and it will have the form of Equation (5.8) above. The coefficients of
dqi dqj (the h2ij in Equation 5.8) are called scale factors. The set of nine
quantities hij (i, j = 1, 2, 3) is called the metric or metric tensor for the
coordinate system. It is easy to appreciate that
∂x ∂x ∂y ∂y ∂z ∂z
h2ij = + + .
∂qi ∂qj ∂qi ∂qj ∂qi ∂qj
A tensor can be expressed in matrix form, so the metric tensor can
be written as
 2 2 2 
h11 h12 h13
 h221 h222 h223 
h231 h232 h233
164 5. THE CONSERVATION OF ENERGY

For an orthogonal coordinate system, the scale factors have the

property
hij = 0 for i 6= j,
2
and the expression for ds (Equation 5.8) reduces to
ds2 = h21 dq12 + h22 dq22 + h23 dq32 , (5.9)
where I made a slight change in notation and replaced hii by hi .
With this expression for ds2 it is clear that for an orthogonal co-
ordinate system, the element of displacement along a given coordinate
is
dsi = hi dqi .
In vector form an infinitesimal displacement is
ds =h1 dq1 ê1 + h2 dq2 ê2 + h3 dq3 ê3 .
The element of area in the new system of coordinates is
dσij = dsi dsj = hi hj dqi dqj ,
and the element of volume is
dτ = ds1 ds2 ds3 = h1 h2 h3 dq1 dq2 dq3 . (5.10)
Now we can obtain an expression for del in generalized coordinates.
In Cartesian coordinates, ∂u/∂x describes how the function u = u(x)
varies for a given change in x. A small displacement along the x-axis
leads to a small change in u given by
 
∂u
du = dx.
∂x
Similarly, a small displacement along the q1 -axis (denoted by ds1 ) leads
to a change in u given by
 
∂u
du = ds1 .
∂s1
But
∂u ∂u
= .
∂s1 h1 ∂q1
Thus the definition of del in Cartesian coordinates
∂ ∂ ∂
∇ ≡ı̂ + ̂ + k̂ ,
∂x ∂y ∂z
leads to the following general definition of del for any coordinate system
1 ∂ 1 ∂ 1 ∂
∇ = ê1 + ê2 + ê3 . (5.11)
h1 ∂q1 h2 ∂q2 h3 ∂q3
 5.3. POTENTIAL ENERGY 165

Exercise 5.7. For cylindrical and spherical coordinates, sketch the

three planes that define the location of point (1,2,0). Label the axes
and the planes clearly.
Exercise 5.8. Sketch the element of area dσ12 for Cartesian coor-
dinates.

5.3.5. Cylindrical Coordinates . Let us apply these general

concepts to obtain an expression for del in cylindrical coordinates. Be-
gin with the transformation equations for cylindrical coordinates:
x = ρ cos φ,
y = ρ sin φ,
z = z.
Then,
dx = dρ cos φ − ρ sin φdφ,
dy = dρ sin φ + ρ cos φdφ,
dz = dz,
and
ds2 = dx2 + dy 2 + dz 2 , (5.12)
= dρ2 + ρ2 dφ2 + dz 2 .
By inspection of this last equation and comparing it to Equation (5.9),
the scale factors for cylindrical coordinates are

h21 = 1, h22 = ρ2 , h23 = 1.

The unit vectors are ρ̂, φ̂, k̂, so using Equation (5.11) yields the follow-
ing expression for ∇ :

∂ 1 ∂ ∂
∇ = ρ̂ + φ̂ + k̂ .
∂ρ ρ ∂φ ∂z
The element of volume in cylindrical coordinates is obtained immedi-
ately from Equation (5.10) as
dτ = ρdρdφdz.
166 5. THE CONSERVATION OF ENERGY

Geometrically this is the infinitesimal parallelepiped obtained from a

displacement of dρ along ρ̂, a displacement of ρdφ along φ̂, and a dis-
placement of dz along k̂. For the purposes of remembering this expres-
sion, it is probably best to visualize the volume element as illustrated
in Figure 5.6.

Exercise 5.9. Fill in the missing steps in the derivation of Equation

(5.12).

z
dz

dρ ρdϕ

y
r
x ρdϕ
dρ

Figure 5.6. Volume element in cylindrical coordinates.

The projected shaded area in the xy plane is the base
of the volume element. Note that the edge marked ρdφ
is an arc of a circle of radius ρ. The sides of the volume
element become flatter as they shrink to zero.

5.3.6. Spherical Coordinates . The relationships between Carte-

sian coordinates and spherical coordinates are
x = r sin θ cos φ,
y = r sin θ sin φ,
z = r cos θ.
Consequently,

dx = dr sin θ cos φ + dθr cos θ cos φ − dφr sin θ sin φ,

dy = dr sin θ sin φ + dθr cos θ sin φ + dφr sin θ cos φ,
dz = dr cos θ − dθr sin θ.
 5.3. POTENTIAL ENERGY 167

In Exercise 5.10 you will show that these relations lead to:

ds2 = dx2 + dy 2 + dz 2
= dr2 + r2 dθ2 + r2 sin2 θdφ2 . (5.13)
By inspection, the elements of the metric (the scale factors) for spher-
ical coordinates are
h21 = 1, h22 = r2 , h23 = r2 sin2 θ.
The unit vectors are r̂, θ̂, φ̂ so using Equation (5.11) the gradient is
expressed as:
∂ 1 ∂ 1 ∂
∇ = r̂ + θ̂ + φ̂ .
∂r r ∂θ r sin θ ∂φ
From Equation (5.10), the volume element (illustrated in Figure 5.7)
is
dτ = h1 h2 h3 dq1 dq2 dq3 = r2 sin θdrdθdφ.

dr
z dϕ
rs in θ rs in θ d ϕ
r rd θ
θ dθ

x ϕ
dϕ rs in θd ϕ

Figure 5.7. Volume element in spherical coordinates.

The volume element has curved edges but in the limit of
infinitesimal quantities the volume is given by (base ×
width × height).

Exercise 5.10. Fill in the missing steps between the two equations
in (5.13).
168 5. THE CONSERVATION OF ENERGY

5.4. Force, Work, and Potential Energy

To begin our consideration of the relation between force, work, and
potential energy let us recall the difference between conservative forces
and nonconservative forces. Work done by a nonconservative force
(such as friction) is at least partially converted into a non-recoverable
form of energy (such as thermal energy which is quickly dissipated
to the environment as heat). You cannot define a potential energy
function for a nonconservative force.
Think about lifting a rock. Imagine we raise it very slowly so the
kinetic energy (T = 12 mv 2 ) is essentially zero at all times. Call the
force that our muscles exert on the rock Fus . Then the work done by
our muscles is Z x2
Fus dx
x1
But Fus involves the tension in our muscles, the friction in our joints,
and so on. I’m pretty sure nobody has come up with a mathematical
expression for this force. Then how in the world can we evaluate the
integral? (Now comes the clever part.) To evaluate the integral we use
the fact that to lift the rock we exerted a force equal and opposite to
the force of gravity.6

Fus = −Fgrav .
Therefore, Z x2 Z x2
Fus dx = − Fgrav dx,
x1 x1
where the gravitational force is mg, directed downward.
When we raised the rock, the change in its potential energy was
Z x2
V2 − V1 = − Fgrav dx
x1
= negative of work done by the gravitational force
= (positive) work done by us.
Thus, the increase in potential energy is equal to the work done by
us, just as you would expect.
The work done by the nonconservative force (us) plus the work
done by the conservative force (gravity) is zero and since there is no
6Tobe precise, we need to exert a force slightly greater than the force of gravity
for an instant to get the rock moving, and when it reaches the top we exert a force
slightly less than the force of gravity so the rock will slow down and stop. Overall,
the average force we exert on the rock is equal to the gravitational force.
 5.4. FORCE, WORK, AND POTENTIAL ENERGY 169

net work done, the work-energy theorem tells us there is no increase in

kinetic energy.
Now suppose we drop the rock. There are no nonconservative forces
acting. The work done by the conservative gravitational force equals
the decrease in potential energy, i.e.,
Z Z
Wg = Fgrav dx = mgdx = mgh = −∆V.

According to the work-energy theorem, the work done by the gravita-

tional force is equal to the increase in the kinetic energy. Therefore, the
decrease in potential energy is equal to the increase in kinetic energy.
Generalizing to any conservative force (F = −∇V ), the work done
by the force during a displacement from r1 to r2 is
Z r2 Z r2 Z r2
W = F·ds = −∇V ·ds = − dV = V (r1 ) − V (r2 ).
r1 r1 r1

Note the use of the vector definition of the differential (see Equation
5.6). Furthermore, ∆V ≡ V (r2 ) − V (r1 ), so
W = −∆V.
That is, the work done by the conservative force will cause the potential
energy to decrease by the same amount. This might be the opposite
of what you expected. You have known for a long time that doing
work on an object will increase its potential energy. But please note
carefully that I said the work done by the conservative force causes an
equal decrease in the potential energy. When you raise a body from
the ground to a height h the conservative force (gravity) does negative
work and the potential energy increases. When the body falls back to
the ground, the conservative force does positive work and the potential
energy decreases.
You might wonder where the zero point of potential energy should
be situated. Recall that the change in potential energy is equal to the
negative of the work done by the conservative forces. Thus, you are
dealing only with the changes in V and the actual numerical values of
V at the end points of the process make no difference. The choice of
the zero level of V is completely arbitrary (but remember that once
you have made the choice you must stick with it).
For example Figure 5.8 shows a mass m at three different locations:
at the bottom of a well of depth d, on level ground, and on top of a
hill of height h.
If you chose the zero level of potential energy to be the level ground,
then V2 = 0; the particle in the well has negative potential energy
V1 = −mgd, and the particle on the hill has positive potential energy,
170 5. THE CONSERVATION OF ENERGY

3
h
2

dd
11

Figure 5.8. Illustration that potential energy is always

relative to an arbitrary value.

V3 = +mgh. The work done by the nonconservative force to take the

particle from the bottom of the well to the top of the hill is
W = V3 − V1 = mg(h + d).
If you select the zero of potential energy at the bottom of the well,
the values of V1 , V2 , V3 all change, but the work required to take the
particle to the top of the hill is still W = mg(h + d).
The quantity W = mg(h + d) is the work you have to do to take
the particle from the bottom of the well to the top of the hill. This
is not the work done by the conservative gravitational force. That
force was directed opposite to the displacement. The work done by the
gravitational field as the particle went from z1 = −d to z3 = +h is
W = −mg(h + d).
When dealing with potential energy, it is often convenient to define
some arbitrary reference position r0 as the position of zero potential
energy. In general,
Z r
V (r) − V (r0 ) = − F·ds.
r0

If V (r0 ) = 0, then Z r
V (r) = − F·ds.
r0
(The reference position is often taken to be at infinity.)

Worked Example 5.4. A particle has a potential energy

given by V = V (x) = kx2 ex/a (in joules), where k = 3 J/m2
and a = 1 m. What is the force on the particle at x = 0? What is
the force on the particle at x = 2.0 m? What is the work done by
 5.4. FORCE, WORK, AND POTENTIAL ENERGY 171

the conservative force during a displacement from x = 0 to x = 2

m?
Solution: Since this is a one dimensional problem, the relation
F = −∇V reduces to F = − dV dx
. Therefore,
dV d
= − (3x2 ex ) = − 6xex + 3x2 ex = −ex (6x + 3x2 )
 
F =−
dx dx
Therefore, for x = 0, F = 0 and for x = 2,
F = −e2 (12 + 12) = −24e2 = −177 newtons.
The work is
W = −∆V = − [V (x = 2) − V (x = 0)] = − 3(22 )e2 − 0 = −88.7 J.
 

Figure 5.9. A rope hanging off of a frictionless platform.

Worked Example 5.5. A rope of mass 2 kg and length 3 m is

partially hanging off the edge of a platform, as shown in Figure ??.
Initially, one meter of the rope is on the platform and two meters
are hanging down. Determine the work required to slowly pull the
rope onto the platform. (Assume that the platform is frictionless.)
Solution: The force required to pull the rope onto the platform
is equal to the weight of the portion of the rope hanging off the
edge. Initially this is 2 meters and finally it is zero meters. The
mass per unit length of the rope is λ = (2/3) kg/m and the weight
of the hanging portion is λgx where x is the length hanging off the
edge. That is F = λgx. The work done to pull it up is
Z Z 2  2 2
x 2 4
W = F dx = λgxdx = λg = (9.8) = 13.1 J.
0 2 0 3 2
This problem can also be solved by determining the change in po-
tential energy of the rope. Taking the platform as the zero of
potential energy we have Vf inal = 0 because the entire rope is on
172 5. THE CONSERVATION OF ENERGY

the platform. The initial value of the potential energy is

Z 0 Z 0
Vinitial = dmgx = gλ xdx = −2λg.
x=−2 −2
So the work done by the conservative force (gravity) is
Wg = −∆V = −(Vf inal − Vinitial ) = −2g(2/3) = −13.1J, (5.14)
and the work done by the applied force is W = −Wg = +13.1J.
A third approach is to simply determine the change in potential
energy of the center of mass of the hanging portion of the rope.
Once again, Vf inal = 0 whereas
Vcm(initial) = mghcm = (2/3)(2)(9.8)(−1) = −13.1J
and once again Wg = −∆V = −13.1 J, and the work done by the
applied force is 13.1 J.

Exercise 5.11. A particle of mass m is in a uniform gravitational

field where the acceleration is g. Is this field conservative?
Exercise 5.12. Show that the gravitational force, as given by New-
ton’s law of universal gravitation, F = − (GM1 M2 /R2 ) r̂, is conserva-
tive.
Exercise 5.13. A particular force can be expressed as F = 3x2 ı̂+2y 2 ̂.
Is this force conservative? (yes)
Exercise 5.14. Two positive charges (Q1 and Q2 ) are separated
by a distance r1 . They are pushed toward one another until their
separation is r2 . What is the change in potential energy? How much
work was done by external forces in pushing the charges together?
Recall that the electrostatic force is given by F = (Q1 Q2 )/(4πε0 r2 )r̂.
Answer: Q1 Q2 (r1 − r2 )/4π0 r1 r2 .
Exercise 5.15. The force exerted by a spring is F = −kx where
k is a constant of proportionality and x is the displacement from equi-
librium. Obtain an expression for the potential energy of a spring.
Answer: If x0 = 0, V = (1/2)kx2 .
Exercise 5.16. The gravitational potential energy of an object of
mass m at a distance r from the center of the Earth is V = −GmME /r
where ME is the mass of the Earth. From this obtain Newton’s Law
of Universal Gravitation.
Exercise 5.17. What is the change in potential energy of a 1 kg
mass when it is taken from the surface of the Earth to infinity? (Ignore
 5.5. THE CONSERVATION OF ENERGY 173

the effect of any other bodies in the universe.) Answer: 6.24 × 107
joules.

5.5. The Conservation of Energy

The work-energy theorem states that the net work done on a parti-
cle is equal to the increase in its kinetic energy. If Wnet is the algebraic
sum of all the work, then
∆T = Tf − Ti = Wnet .
This principle is never violated. (However, it is easy to think of situa-
tions in which it does not appear to be true.)
Suppose the only force acting on the particle is conservative. In
that case, the work done can be written as W = −(Vf − Vi ) = Vi − Vf .
Combining this with the work-energy theorem leads to
Tf − Ti = Vi − Vf
or
Tf + Vf = Ti + Vi .
This equation is the simplest expression of the law of conservation of
energy. It states that if the only work performed on a system is done
by conservative forces, then the sum of the initial kinetic energy plus
the initial potential energy (before the work was performed) is equal
to the sum of the final kinetic energy plus the final potential energy
(after the work was performed).
The sum of kinetic energy and potential energy is called the total
energy. It is denoted by E :
E = T + V.
The conservation law simply states that the total energy is constant.
This statement holds as long as the only forces acting on the
system are conservative forces.
In high school physics courses one often hears the energy conser-
vation principle expressed as, “Energy can neither be created nor de-
stroyed but it can be converted from one form to another.” A more
sophisticated version of this statement might be, “In mechanical sys-
tems the effect of a conservative force is to transform potential energy
into kinetic energy, or kinetic energy into potential energy.”
The situation gets a little more complicated when nonconservative
forces are involved. Suppose you pick a book from the floor and place
it on a table, at height h above the floor. The initial energy of the
174 5. THE CONSERVATION OF ENERGY

book was zero because it had zero kinetic energy to begin with (and
the floor is the zero point of potential energy). When the book is on
the table, once again it has zero kinetic energy and its total energy is
E = V = mgh. Where did this energy come from? Clearly you are
the ultimate source of the energy of the book. The net work done on
the book (by you and gravity) was zero, so the book’s kinetic energy
did not increase. The increase in potential energy was equal to the
negative of the work done by the gravitational force, but this energy
can be recovered (if the book falls off the table, for example). The
work done by you is not recoverable because the force you exerted was
nonconservative. If you could evaluate the chemical energy supplied
to your muscles, do you think it would be equal to mgh? (Actually it
would not, because energy is also internally converted into heat in a
working person’s body.)
Mechanics primarily deals with problems involving conservative
forces, so the conservation of energy principle is nearly always written in
the simple form E = T +V = constant, or equivalently, Ti +Vi = Tf +Vf
where subscript i represents some initial state of the system and sub-
script f represents some final state. If you wish to use the energy
principle when nonconservative forces are acting you can include them
as a work term. That is, if Wnc is the work done on the system by
nonconservative forces as the system evolves from initial state i to final
state f , then
Tf + Vf = Ti + Vi + Wnc
Z f
= Ti + Vi + Fnc · ds.
i

In solving problems, make sure you have the right sign on Wnc , de-
pending on whether the nonconservative work increases or decreases
the total energy of the system. This is usually obvious from the state-
ment of the problem or from a consideration of the behavior of the
system.

Exercise 5.18. A 40 kg child climbs 5 meters up a flagpole then

slides to the bottom. The child is moving at 3 m/s upon reaching the
bottom. What was the work done by friction? Answer: 1780 J.
Exercise 5.19. Evaluate the escape velocity of an object from the
Earth. (This means that the object is initially on the surface of the
Earth and ends up an infinite distance away.) If the object just barely
 5.5. THE CONSERVATION OF ENERGY 175

escapes, it reaches infinity with zero velocity. Ignore any other bodies
in the universe. Answer: 11.2 km/sec.
Exercise 5.20. Evaluate the velocity required for a body at 1 AU
from the Sun to escape from the solar system. Assume the solar system
consists only of the Sun (a good approximation).

5.5.1. A Reflection on the Conservation of Energy. Conser-

vation laws are extraordinarily useful for solving physics problems. To
determine the speed of a falling object it is easier to apply the law
of conservation of energy than to integrate the equation of motion.
However, conservation laws are much more than mere computational
aids; they are fundamental laws of nature that tell us how the phys-
ical universe behaves.7 Consider, for example, the universe defined
as the total collection of all material objects. There are no external
forces doing work on the universe (because there is nothing outside of
it!). Therefore, the total energy of the universe must be a constant
quantity. Similarly, the total linear momentum and the total angular
momentum of the universe are also constant quantities.
It has been suggested that conservation laws are not laws of nature,
but just mental constructs designed to help us comprehend and deal
with the physical world. The law of conservation of energy is often
cited as an example of this point of view. In your introductory physics
course you learned the law of conservation of energy in the form:
kinetic energy + potential energy = constant.
When you studied thermodynamics you learned that mechanical energy
could be converted into heat and you modified the law to:
kinetic energy + potential energy + thermal energy = constant.
Then you studied electricity and magnetism and learned of other forms
of energy associated with the electromagnetic fields. These also had to
be included in the conservation law. And so on...
This procedure gives the impression that every time we find a new
form of energy we simply add it to the left hand side of the equation and
7A physicist who finds that a conservation law has been violated knows that it
is either a mistake or a very important discovery. For example, it was believed for
many years that parity was always conserved. The discovery of a reaction in which
parity did not remain constant caused a furor among physicists and eventually led
to a much deeper insight into the nature of the physical world. (We will consider
the conservation of parity briefly in Chapter 8.)
176 5. THE CONSERVATION OF ENERGY

define energy as that quantity whose total value is constant. However,

such an approach tends to trivialize the conservation laws and we shall
not adopt it. We shall consider the conservation laws to be fundamental
to the structure of the physical universe. As shown in Chapter 8, the
conservation laws are related, on a very profound level, to symmetries
found in nature. Thus, energy conservation is related to a symmetry
in time and momentum conservation is related to a symmetry in space.
(I really don’t expect you to understand what I am talking about here
- but don’t worry, it will all become clear as we go along!)

5.6. Energy Diagrams

Imagine a particle that is initially at the top of a perfectly smooth
hill, as illustrated in Figure 5.10 which shows the particle perched on
a peak at point A. There are peaks at A, C, and E and valleys at B
and D. The only force acting on the particle is gravity.
The particle is given an infinitesimal push to the right. It slides
down the hill towards B speeding up as it goes. It climbs the hill at
C, slowing down somewhat, then speeding up as it descends into the
valley at D. Finally it climbs another big hill and stops on the tip of
the hill at point E.
In other words, the motion is what you would expect from an ordi-
nary amusement park roller coaster ride, if you could find a completely
frictionless roller coaster.

A B C D E x

Figure 5.10. A particle on the top of a frictionless hill.

If given an infinitesimal shove, the particle will slide past
B, C, D, and end up on the tip of E.

You could consider the sketch in Figure 5.10 to be a diagram of

potential energy versus position because gravitational potential energy
 5.6. ENERGY DIAGRAMS 177

(mgh) is proportional to h, the height of the particle above zero level.

Figure 5.11 shows the same situation, but now I labeled the vertical
axis V (x) and I drew a line intersecting the vertical axis at a value
labelled E. This line gives the total energy. We call such a plot an
energy diagram.

V(x)
E

x
A B C D E

Figure 5.11. An energy diagram. The vertical axis is

energy (traditionally labeled as potential energy) and the
horizontal axis is position. Note that the horizontal line
at energy E indicating the total energy is constant and
equal to the same value for all x.

This plot is a very simple energy diagram, but it illustrates a num-

ber of important points. First of all, note that the total energy is
constant (the horizontal line indicates the particle has the same total
energy whatever the value of x, the position of the particle). Secondly,
note that I knew where to draw the total energy line because I knew
that at point A the particle had no kinetic energy (TA = 0) and the
total energy at that location was equal to the potential energy.
Since E = T + V at any point, the kinetic energy is given by the
difference between the total energy and the potential energy at that
point. The energy diagram allows you to determine the kinetic energy
for any value of x by simply noting the “distance” between the line
for the total energy and the potential energy curve. The velocity is
proportional to the square root of T so this also gives an estimate of
how fast the particle is moving.
It is very important for you to know how to interpret energy di-
agrams. Consider, for example, the energy diagram of Figure 5.12
in which I plotted the potential energy of a mass m connected to a
spring of constant k. (Recall that for this system the potential energy
is V = 21 kx2 .) The total energy is E. The mass will therefore oscillate
178 5. THE CONSERVATION OF ENERGY

V(x)

forbidden forbidden
region region

-x 1 x=0 +x
1

Equilibrium point

Figure 5.12. Energy diagram for a mass on a spring.

between the points labelled ±x1 . These two points are called “turning
points.” Even though the actual physical system is a block of mass m
sliding back and forth horizontally, you can represent it by the round
dot sliding up and down in a “potential well,” as shown in the up-
per part of the figure. If the dot starts from rest at a turning point
(say −x1 ) it will slide down to x = 0, picking up speed and reaching
maximum velocity at the bottom, then slowing down as it climbs the
potential hill and coming to a stop at the other turning point (at +x1 ).
The physical motion of the block is given by the projection of the dot
onto the horizontal (x) axis. At the turning points the velocity of the
block is zero. The regions beyond the turning points are labeled “for-
bidden region.” A particle in the forbidden region would have V (x) > E
(because the turning point is the location where V (x) = E). The al-
lowed region is where V (x) < E. It is obvious that V (x) > E is
impossible because T = E − V, and T would be negative if V > E.
Since T = 12 mv 2 , a negative value of the kinetic energy requires either a
negative mass or an imaginary value for the velocity. There is no such
thing as a negative mass, and an imaginary value for velocity is not
allowed because velocity is an observable quantity and all observable
physical quantities are real.8
8Physicists
frequently use complex quantities to make the mathematics of a
problem easier, but only the real part has physical significance. Keeping this simple
fact in mind will help you to avoid confusion, especially in your study of electro-
magnetic waves where the electric and magnetic fields are expressed as complex
 5.7. SOLVING FOR THE MOTION: THE ENERGY INTEGRAL 179

5.7. Solving for the Motion: The Energy Integral

Let us consider the motion of a particle of mass m and total energy
E in a region where the potential energy has a form such as illustrated
in Figures 5.11 or 5.12 and all the forces are conservative. (We limit
ourselves to one dimensional motion for simplicity.) To obtain an ex-
pression for the position as a function of time, we can start with the
fact that the total energy is constant:
T + V = constant = E.
Using the definition of kinetic energy write
1 2
mv + V (x) = E.
2
Then,
2
v2 = (E − V (x)) ,
m
or
r
dx 2p
= E − V (x).
dt m
This fairly simple differential equation can be solved by separation of
variables. Since E is a constant and V is a function only of x,
r
dx 2
p = dt.
E − V (x) m
This equation can be integrated, at least in principle. The expression
Z x r
dx 2
p = t (5.15)
x0 E − V (x) m
is called the energy integral.
You can go no further unless you have an explicit expression for
V (x). If such an expression is given, carry out the integration over x
and obtain a mathematical expression involving x and x0 and E. That
is, obtain
t = t(x, x0 , E).
Finally, invert this equation and solve for x in terms of x0 , E, and t to
get an expression of the form
x = x(t) = x(x0 , E, t).
functions whose real parts correspond to the actual measurable quantities. This
concept is also important in quantum mechanics where you must keep in mind that
measurable physical quantities are expressed as real numbers even though the wave
functions are complex.
180 5. THE CONSERVATION OF ENERGY

Solving for the motion means obtaining an expression for x as a function

of time and various constants. The constants are usually the initial
position and the initial velocity. Here x(t) is given in terms of initial
position and total energy, which is not quite the same thing, but is
obviously equivalent.
Note that solving for the motion using the energy only requires one
integration, whereas starting with the equation of motion requires two
integrations. The reason is that the conservation of energy equation
can be expressed as
1 2
mv = E − V (x)
2
which is essentially an expression for the velocity. If we start with
the equation of motion, the first integration gives the velocity. For
this reason, the conservation of energy equation is often called a “first
integral.”
Since the total energy E remains constant (as long as the forces
are conservative), E is called a “constant of the motion.” As you have
noted, knowing E makes the problem easier to solve. In general, know-
ing the constants of the motion makes problem solving much easier.

Worked Example 5.6. A particle of mass m = 0.1 kg moves

in a one dimensional potential field given by V = −2x + 3x2 J.
Determine the motion if the total energy is 1 J. Assume that at
time t = 0 the particle is at x = 1/3 m.
Solution: Begin with equation (??):
r Z x Z x(t)
2 dx dx
t= p = √
m x0 E − V (x) 1/3 1 + 2x − 3x2
x(t)
1 ±(1 − 3x)
= − √ sin−1

 .
3 2 
1/3

The last expression came from evaluating the standard form inte-
gral
2ax − b
Z
dx 1
√ = −√ sin−1 √ .
ax2 + bx + c −a b2 − 4ac
 5.8. THE KINETIC ENERGY OF A SYSTEM OF PARTICLES 181

Therefore,
r  
2 1 −1 1 − 3x
t = − √ sin ±
m 3 2
1 − 3x √
± = − sin 60t
2
1 2 √
x = ± sin 60t.
3 3
The particle oscillates sinusoidally in a potential well between the
turning points at x = − 31 and x = +1.

Exercise 5.21. A particle moves in a constant potential energy

field, V = V0 = constant. Assume E is known. Determine the position,
velocity and acceleration as a function
p of time, that is, x = x(t), v =
v(t), and a = a(t). Answer: x(t) = 2(E − V0 )/mt + x0 .
Exercise 5.22. Show that a particle in a region of space where the
potential energy is constant will move with zero acceleration.

5.8. The Kinetic Energy of a System of Particles

Nearly everything that has been said in this chapter has referred
to a single particle or to a rigid body that can be treated as a particle
located at the center of mass of the body. But a rigid body can rotate
as well as translate. We now show that the energy of a system of
particles can be expressed as the energy of translation of the center of
mass, plus the energy of the particles relative to the center of mass.
We shall show that
TT otal = TCM + TwrtCM
where TCM is the translational kinetic energy of a particle whos mass
is the total mass of the body and which is moving with the center of
mass of the body, and TwrtCM is the kinetic energy of all the particles
with respect to the center of mass. (This is esentially what we assumed
when we considered a disk rolling down an inclined plane and wrote
1 1
T = Ttrans + Trot = M v 2 + Iω 2 .
2 2
Keep in mind that a rigid body is a collection of particles that
maintain fixed distances from one another.
182 5. THE CONSERVATION OF ENERGY

Before going any further, you should refresh your memory on the
definition of center of mass. In particular, recall that for a collection
of N particles located at ri , the position of the center of mass, denoted
rc , is PN
mi ri
rc = Pi=1N
. (5.16)
i=1 m i
So
XN
M rc = mi ri , (5.17)
i=1
P
where M = mi is the total mass.

mi
z i

r'
i
r cm
i
rc
y

Figure 5.13. A system of particles. Particle i with

mass mi is located at position ri . The center of mass
(cm) is at rc . The arrow marked r0i is the position of
particle i with respect to the center of mass.

We now show that the total kinetic energy of a system of particles

can be written as the sum of the kinetic energy of translation of a
particle of mass M moving with the center of mass and the kinetic
energy due to the motion of all the particles relative to the center of
mass. To prove this, begin with the fact that the kinetic energy of
particle i is
Ti = 21 mi vi2 = 21 mi (ṙi ·ṙi ),
where the vector ri is measured from the origin. As shown in Figure
5.13 the position vector of particle i can be described in terms of the
position of the center of mass by
ri = rc + r0i ,
where r0i is the position of particle i relative to the center of mass.
Therefore, the kinetic energy of particle i can be expressed as
Ti = 12 mi (ṙc + ṙ0i ) · (ṙc + ṙ0i ) = 21 mi ṙc2 + 2ṙc · ṙ0i + ṙi02


 5.9. WORK ON AN EXTENDED BODY. PSEUDOWORK 183

The total kinetic energy is obtained by adding the kinetic energies of

all the particles in the system.
N
X X X X
T = Ti = 12 ṙc2 mi + ṙc · mi ṙ0i + 1
m ṙ02 .
2 i i
i
The middle term is zero because
X d X d X
mi ṙ0i = mi r0i = mi (ri − rc ),
dt dt
and
X X X
mi (ri − rc ) = mi ri − mi rc = M rc − rc M = 0.
Consequently, the total kinetic energy of a system of particles is given
by X X
T = 21 ṙc2 mi + 1
m ṙ02 ,
2 i i
or X
T = 12 M vc2 + 1
m v 02 .
2 i i
(5.18)
If you imagine a particle of mass M at the center of mass, then the
total kinetic energy of a system of particles is equal to the sum of the
translational kinetic energy of M plus the kinetic energy of all of the
particles relative to the center of mass, and the proposition is proved.

Exercise 5.23. Two particles are on a collision course along the

x-axis. In the laboratory coordinate system a particle of mass 2 kg
moves towards the right with a speed of 2 m/s, and a particle of mass
1 kg moves to the left with a speed of 1 m/s. (a) Determine the total
kinetic energy of the system. (b) Determine the velocity of the center
of mass. (c) Determine the kinetic energy of each particle relative to
the center of mass. (d) Verify Equation (5.18). Answers: (a) 4.5 J (b)
1 m/s.

5.9. Work on an Extended Body. Pseudowork

The work done on a particle was defined by Equation (5.1),
Z r2
W = F·ds,
r1
where F was the force acting on the particle as it moved from r1 to r2 .
The displacement ds can be interpreted either as the displacement of
184 5. THE CONSERVATION OF ENERGY

the particle or as the displacement of the point of application of the

force, since the two are equivalent. Furthermore, it is clear that the
source of the increase in the particle’s kinetic energy is the agent that
is producing the force. However, the situation gets more complicated
when you consider the work done when a force acts on an extended
body.
A simple example of an external force acting on an extended body
is an ice skater who puts her hand on a wall and pushes herself away
from it. The external force is the reaction force of the wall on her hand.
(She pushes on the wall and the wall pushes back on her with an equal
and opposite force.) The force acting on the skater stops as soon as her
hand loses contact with the wall. The point of application of the force
did not move at all. Did the wall do work on the skater? The answer
is obvious from an energy point of view: The energy of the wall did
not change! The source of the kinetic energy imparted to the skater
was chemical energy from her own body. This chemical energy was
converted into mechanical energy through the action of her muscles.
Nevertheless, the only unbalanced external force acting on the skater
was the reaction force the wall exerted on her hand.
To determine the appropriate form of the equation for work in the
case of an extended body, we begin with Newton’s second law in the
form
Fe = mac ,
where ac is the acceleration of the center of mass of the extended body
and Fe is the external force. Multiplying both sides of this equation by
an infinitesimal displacement of the center of mass, dsc leads to
dvc dsc
Fe dsc = mac dsc = m dsc = mdvc = mvc dvc .
dt dt
Integrating from initial point i to final point f, gives
Z f Z f    
1 2 1 2
Fe dsc = mvc dvc = mv − mv = ∆T.
i i 2 c f 2 c i
Therefore, the change
R f in kinetic energy of the extended body is equal
to the expression i Fe dsc . This looks suspiciously like the work. But
it isn’t exactly like the work because the point of application of the
force did not move. Furthermore, the body exerting the force was not
the source of the energy increase. For this reason, some authors prefer
to call this expression the “pseudowork” and they call the equation
Z f
Fe · dsc = ∆T
i
 5.10. SUMMARY 185

the “energy equation” to distinguish it from the work-energy theorem

in which the external force is the source of the energy.9

5.10. Summary
In this chapter you have been exposed to a number of mathematical
concepts and a number of physical concepts. In this summary the two
sets of concepts are listed separately, first the math, then the physics.

5.10.1. Mathematical Concepts. R

Line Integral. By definition, Work = F·ds. The integral is a line
integral, and it must be evaluated along the path taken by the particle.
You have seen two ways to evaluate a line integral: (1) Express the
force and the differential of displacement in component form, thus
Z Z Z Z
W = F·ds = Fx dx + Fy dy + Fz dz,
C C C C

or (2) Express the force and displacement in terms of some parameter

λ, and evaluate
Z Z
ds
W = F(λ)·ds = F(λ)· dλ.
C C dλ
Del Operator. The del operator in Cartesian coordinates is de-
fined as
∂ ∂ ∂
∇ = ı̂ + ̂ + k̂ .
∂x ∂y ∂z
Operations involving del include
∇f The gradient of the scalar function f
∇ · F The divergence of the vector function F
∇ × F The curl of the vector function F.
This chapter concentrated on del operating on a scalar function.
This generates a vector called the gradient that is denoted by ∇f .
Geometrically, think of the gradient as a vector pointing in the direction
of the greatest increase in f, whose magnitude gives the rate of increase
of f in that direction.

9A very interesting article on this subject is the paper by Bruce Sherwood,

Pseudowork and Real Work Am. J. Phys, 51, 597-602, 1983.
186 5. THE CONSERVATION OF ENERGY

Metric. The differential displacement vector ds has different forms

in different coordinate systems. In terms of the generalized coordinates
q1 , q2 , q3 ,
X
ds2 = h2ij dqi dqj , i, j = 1, 2, 3.
ij

The scale factors hij depend on the geometrical properties of the coor-
dinate space and collectively are called the metric.
Del and Volume Element in Other Representations. Using
the transformation properties of coordinates we derived expressions for
the volume element and del in terms of generalized coordinates as

dτ = ds1 ds2 ds3 = h1 h2 h3 dq1 dq2 dq3 ,

and
1 ∂ 1 ∂ 1 ∂
∇ = ê1 + ê2 + ê3 .
h1 ∂q1 h2 ∂q2 h3 ∂q3
In cylindrical coordinates these yield
dτ = ρdρdφdz,
∂ 1 ∂ ∂
∇ = ρ̂ + φ̂ + k̂ ,
∂ρ ρ ∂φ ∂z
and in spherical coordinates they yield
dτ = r2 sin θdrdθdφ,
∂ 1 ∂ 1 ∂
∇ = r̂ + θ̂ + φ̂ .
∂r r ∂θ r sin θ ∂φ

5.10.2. Physical Concepts. I will now summarize the many im-

portant physical concepts introduced in this chapter.
Work. The work done on a particle by a force F during a displace-
ment from r1 to r2 is
Z r2
W = F·ds
r1

As noted above, the integral is a line integral and must be evaluated

along the path followed by the particle. However, if the force is conser-
vative, then the value of the integral depends only on the end points.
Work-Energy Theorem. The total work done by all of the ex-
ternal forces equals the increase in kinetic energy of the particle.
W = ∆T = Tf − Ti .
 5.10. SUMMARY 187

Potential Energy. If a force is conservative (∇ × F = 0) we can

associate with it a potential energy V. The relation between force and
potential energy is
F = −∇V.
The work done by a conservative force can be expressed in terms of the
change of potential energy, thus

W = −∆V = −Vf + Vi .

Conservation of Energy. For conservative forces, equating the

two expressions for work leads to the conservation law

Tf + Vf = Ti + Vi .

Energy Diagrams. An energy diagram is a plot of the potential

energy (V ) as a function of position (x). The total energy is represented
by a constant horizontal line on such a diagram. The distance between
the total energy line and the potential energy curve (E − V ) is equal
to the kinetic energy and is proportional to the speed of the particle
squared.
Solving for the Motion using Energy. When the force is a
function of position, energy methods are the easiest way to determine
the motion. Basically, this involves evaluating an integral of the form
Z x r
dx 2
p = t.
x0 E − V (x) m

Energy of a System of Particles. The kinetic energy of a system

of particles can be expressed as the sum of the kinetic energy of the
center of mass and the kinetic energy with respect to the center of
mass:
X
T = 12 M vc2 + 1
m v 02
2 i i
= Tc + Twrt .
cm

Work on an Extended Body. Although the equations for the

work done on an extended body are the same as the equations for the
work done on a particle, the two concepts are not the same. In particu-
lar, the source of energy may be different in the two cases. Nevertheless,
the correct answer is obtained by simply replacing the displacement of
the particle by the displacement of the center of mass of the extended
body.
188 5. THE CONSERVATION OF ENERGY

5.11. Problems
Problem 5.1. The force R on a particle is given by F =3xı̂+2y̂ N.
Determine the line integral F·ds for the straight line path that starts
at the origin and ends at the point (3,6) m. Answer: 49.5 J.
Problem 5.2. In this problem we assume that somehow an elec-
tron is dragged past a proton in a straight line path. The electron is
subject to the electrostatic force and to whatever external force keeps
it moving in a straight line at constant speed. (a) Evaluate the work
done by the electrostatic force as the electron is taken from x1 to x2
along the trajectory of Figure 5.14. (b) What is the work done by the
external force required to keep the electron on the straight line path?
(c) Evaluate the work done by the electrostatic force if the electron is
taken around a semicircular path with the proton at the center.

x ^r
x1 x2
0
q
- q

r ds = dx
d
F
+

Figure 5.14. An electron is dragged past a proton from

x1 to x2 along the straight line path.

Problem 5.3. The magnitude of the force on a particle is F = − kr

where r is the distance from the origin to the particle. The force is
directed towards the origin at all times. (This is an example of a
central force.) Determine the work done when the particle is moved
along a semicircular path of radius R from the origin to point (2R, 0).
Determine the work done when the particle is moved from the origin
to the point (2R, 0) along a straight line path. (Note: The origin is not
at the center of the semicircular path.)
Problem 5.4. Derive Relations (5.8) and (5.10) from the transfor-
mation equations and the expression for ds2 in Cartesian coordinates.
Problem 5.5. (a) Determine the metric for plane polar coordi-
nates. (b) Obtain an expression for ds in polar coordinates, as well as
the element of area.
Problem 5.6. Determine ds2 , the scale factors, the vector ds, the
volume element and the ê vectors for the paraboloidal coordinates
 5.11. PROBLEMS 189

u, v, φ :
x = uv cos φ,
y = uv sin φ,
1 2
u − v2 .


z =
2
Problem 5.7. The “elliptic cylindrical” coordinates, u, v, z, are
defined by
x = a cosh u cos v,
y = a sinh u sin v,
z = z.
Determine the metric. Write an expression for ds.
Problem 5.8. Consider the u, v, z coordinates defined by
1 2
u − v2 ,


x =
2
y = uv,
z = z.
(a) Determine the metric for these coordinates. (b) Evaluate ∇ in these
coordinates.
Problem 5.9. The prolate spheroidal coordinates η, θ, φ are related
to the Cartesian coordinates by the following transformation equations
x = a sinh η sin θ cos φ,
y = a sinh η sin θ sin φ,
z = a cosh η cos θ.
Determine the expression for ∇ using these coordinates.
Problem 5.10. (a) Using the expression for ∇ given by Equation
(5.11), obtain ∇ · V in cylindrical coordinates. Note that the deriva-
tives also act on the unit vectors, as in going from Equation (2.8) to
∂V
Equation (2.9). Answer: ∇ · V = ρ1 ∂ρ ∂
(rVρ ) + ρ1 ∂φφ + ∂V∂z
z
. (b) Show that
the same result is obtained
h from i
∇ · V = h1 h12 h3 ∂q∂1 (h2 h3 V1 ) + ∂q∂2 (h1 h3 V2 ) + ∂q∂3 (h1 h2 V3 ) .

Problem 5.11. Using spherical coordinates, determine the volume

of a sphere. Pay particular attention to the limits on the integrals over
θ and φ.
190 5. THE CONSERVATION OF ENERGY

Problem 5.12. Near the surface of the Earth the potential energy
is mgh. Use this fact to obtain an equation for the position of a falling
particle as a function of time. Show that the acceleration of the particle
is g.
Problem 5.13. Upon finding that the top of a certain mountain
is a perfect cone, a surveyor describes it mathematically in cylindrical
coordinates by the formula z = h0 − ρ. (a) Sketch the contour lines.
(b) Show that the direction of steepest ascent everywhere points toward
the summit.
Problem 5.14. A mountain rises above a flat plain. The height of
the mountain above the plane is described by the relation
z(x, y) = 2000 exp − x2 + 2y 2 /8000


meters.
(a)How high is the mountain? (b) If you are standing at x = 20 m, y =
10 m, how high above the plain are you? (c) At that location, what is
the direction of quickest descent? (d) What is the maximum rate of
descent at this point? (That is, how many meters do you descend for
every meter of horizontal displacement?) Answer: (d) -13.12 m/m.
Problem 5.15. The temperature in a certain location is given by
T = T0 − A(2x2 + y 2 + z 2 ) (in kelvin). Distances are in meters and
the constant A has the value 0.5 K/m2 . The value of T0 is 300 K. (a)
At point (1,2,2) what is the direction and rate (in K/m) of maximum
temperature increase? (b) If you move one meter in that direction
what is the temperature at the new position? (c) The temperature
at the new position as calculated by the formula is not equal to the
value obtained by multiplying the rate of temperature decrease by the
displacement. Explain this discrepancy.
Problem 5.16. (a) Determine the center of mass of the system
composed of three particles of masses 1, 2, 3 kg located at (0,1), (1,0)
and (1,1) respectively. (b) Assume the 1 kg particle has a velocity
of 1̂ m/s, the 2 kg particle has a velocity of 2ı̂ m/s and the 3 kg
particle has a velocity of 3(ı̂+̂) m/s. Determine the total kinetic energy
of the system. (c) What is the velocity of the center of mass? (d)
Determine the kinetic energy of the particles relative to the center of
mass. Answers: (a) (5/6)ı̂+(4/6)̂, (b) 31.5J, (c) 136
ı̂+ 10
6
̂ m/s, (d) 9.08
J.
Problem 5.17. Two positive charges (Q1 and Q2 ) are located on
the x-axis at positions x = ±a. A negative charge (−q) of mass m is
constrained to move along the y-axis. The electrostatic potential of the
 5.11. PROBLEMS 191

system is
qQ1 qQ2
V =− −
4π0 r1 4π0 r2
where r1 and r2 are the distances from the negative charge to the
positive charges. The negative charge is released from rest at y = b.
Obtain an expression for the maximum speed of the negative charge.
Where does this occur?
Problem 5.18. A football player of mass M decides to go bungee
jumping. He suits up in a halter with a long elastic cord of unstretched
length b. After tying the free end of the rope to a high bridge he leaps
out into space. A few moments later he finds himself suspended far
above the ground. He climbs the bungee cord back up to the bridge.
Determine the ratio of the work done climbing the elastic rope to the
work done in climbing an inelastic rope of length b. You may assume
the elastic rope obeys Hooke’s law and has a force constant k.
Problem 5.19. A girl on a swing is pushed harder and harder by
her older brother. Eventually, she is swinging in an arc whose highest
point is 1.5 m above its lowest point. Assume the swing has massless
ropes 2.8 m long and that the mass of the girl (plus swing seat) is 30
kg. What is the tension in the ropes when the girl passes through the
lowest point?
Problem 5.20. A block of mass 5 kg is on a horizontal surface.
The coefficient of sliding friction is 0.5. The mass has a speed of 10 m/s
and is moving towards a stationary spring that is 3 meters away. The
spring constant is 4000 N/m. The block strikes the spring, compresses
it, and bounces back. How far from the spring does the block come to
rest?
Problem 5.21. A radioactive nucleus decays by emitting an alpha
particle with energy 6.0 MeV. The total energy released in the disinte-
gration is 6.2 MeV. Determine the mass of the recoiling nucleus (ignore
relativistic effects).
Problem 5.22. A particle of mass m and total energy E is moving
in a one-dimensional potential given by V (x) = −bx. Determine the
motion.
Problem 5.23. A rail gun is a device that accelerates a projectile
to extremely high speeds using magnetic forces. Suppose a particle of
mass m is fired vertically from the surface of the Earth. Ignoring air
resistance, show that the maximum height reached by the particle is
192 5. THE CONSERVATION OF ENERGY

given by
GM m
−R
|E|
where R is the radius of the Earth and M is its mass. (Assume that
its total energy E is negative.)
Problem 5.24. Prove that for a particle the rate of change of
kinetic energy dT
dt
is equal to the dot product of the force acting on it
and the instantaneous velocity of the particle, that is, F · v.
Problem 5.25. Let h = r − R be the position of a particle above
the surface of Earth where R is the radius of the Earth. Prove that in
the limit
hR
the gravitational potential reduces to mgh.
Problem 5.26. A particle of mass m is in a one-dimensional po-
tential energy field given by
2
V (x) = −Ae−ax ,
where A and α are constants. (a) Plot the energy diagram. (b) Deter-
mine the turning points if the particle has a total energy E = −0.5A.
(c) Determine the turning points if the particle has a total energy
E = −Ae−1 . (d) Assume the particle has zero energy and is located at
x = −∞. It is given an infinitesimal shove towards the origin. What is
its velocity when it passes through the origin?
Problem 5.27. In a certain region of space the potential energy
can be expressed as
A
V = −p ,
x2 + y 2 + z 2
where A is a constant, and the origin is excluded. (a) Obtain an expres-
sion for the force. (b) Determine the work required to take a particle
from (x1 , y1 , z1 ) to ∞. (c) Express the force in spherical coordinates.
Problem 5.28. The potential energy of a vibrating diatomic mol-
ecule as a function of the separation (s) between the two atoms is
approximately given by the “Morse Function,”
V (s) = V0 (1 − e−(s−s0 )/δ )2 − V0 ,
where s0 , δ, V0 are constant parameters. (a) Obtain an expression for
the force on the atoms. (b) Determine the separation between the
atoms when the potential energy is a minimum. (c) What is the
 5.11. PROBLEMS 193

minimum value of the potential energy? Answer (a) − (2V0 /δ) (1 −

e−(s−s0 )/δ )e−(s−s0 )/δ .
Problem 5.29. Coulomb’s law tells us that the electrostatic force
between two charged particles is
Q1 Q2
F= r̂
4πε0 r2
where Q1 and Q2 are the charges on the particles and r is the sepa-
ration between them. The quantity 4πε0 is constant. Show that the
electrostatic force is conservative. Obtain an expression for the poten-
tial energy.
Problem 5.30. A particle is subjected to a force given by
F =K[(2x + y)ı̂ + (x + 2y)̂].
(a) Prove that this force is conservative. (b) Obtain an expression for
the potential energy.
Problem 5.31. Newton’s law of universal gravitation states that
the force between two particles of masses m1 and m2 is
Gm1 m2
F=− r̂,
r2
where G is a constant and r is the distance between the particles. Use
this force expression to obtain the gravitational potential energy of the
system. Select an appropriate point for the zero of potential energy.
Problem 5.32. By evaluating the work required to bring three
masses m1 , m2 , and m3 from infinity to their final positions at r1 , r2 ,
and r3 determine the potential energy of the system. (The only force
acting on the particles is their mutual gravitational attraction.)
Problem 5.33. A particle of mass 2 kg moves along the x axis. Its
potential energy as a function of position is V (x) = −3x + x2 joules.
(Here x is measured in meters.) The particle passes through the origin
with a speed of 4m/s. (a) Sketch the potential energy as a function
of x. (b) Determine the speed of the particle when it is at x = 1. (c)
Where are the turning points?
Problem 5.34. Consider the potential
a b
V = − + 2.
r r
(a) Plot the potential as a function of r. (b) Obtain an expression for
the force. (c) Obtain an expression for the turning points as a function
of a, b, and the total energy E.
194 5. THE CONSERVATION OF ENERGY

Problem 5.35. A particle moving in a potential given by V (x, y) =

ax2 +by 2 is a two-dimensional oscillator with different force constants in
the two directions. (This is called a nonisotropic oscillator.) Determine
the motion, assuming that at time t = 0 the particle is passing through
the origin with velocity v =v0x ı̂ +v0y ̂.
Computational Projects
Computational Project 5.1. Assume that the potential energy
is given by V (x) = −(1/x6 ) + (1/x12 ). Integrate Equation (??) numer-
ically to determine the position of a particle of unit mass and total
energy 10−4 energy units. Plot x = x(t) and v = v(t) for a particle
coming in from infinity. (Infinity is approximately at x = 5.)
Chapter 6
Conservation of Linear Momentum

We now turn our attention to problems that can be solved using the
law of conservation of linear momentum. Examples of such problems
include the motion of a rocket, collisions in one and two dimensions,
and the behavior of a system when an impulsive force acts on it.

6.1. The Law of Conservation of Momentum

The law of conservation of linear momentum is based on Newton’s
second law: The rate of change of momentum of a system is equal to
the net external force acting on the system:
dP
= F,
dt
where F is the net external force and P is the total linear momentum.
Therefore, if F = 0, then P =Σi mi vi = constant.1 In words, the law
of conservation of linear momentum is:
If no net external force acts on a system,
the total momentum of the system is constant.
For example, in the collision of two bodies, if external forces are negli-
gible, the conservation of momentum principle can be expressed as
Pfinal = Pinitial ,
where Pinitial is the total momentum of the two bodies before the colli-
sion and Pfinal is the total momentum after the collision. Many prob-
lems can be solved using this simple relation.
1In Chapter 8 you will be exposed to a more general derivation of the law of
conservation of linear momentum.
195
196 6. CONSERVATION OF LINEAR MOMENTUM

If a net external force does act on a particle, and if the mass of a

body changes with time, then Newton’s second law must be expressed
in the form
dP dv dm
F= =m +v . (6.1)
dt dt dt
Note that dv/dt is the acceleration of the body; in this case, one cannot
write Newton’s second law in the form F =ma.

Exercise 6.1. A bullet of mass m and velocity v is fired at a wooden

block of mass M at rest on a frictionless surface. The bullet embeds
itself in the block. (a) Determine the speed of bullet plus block after
the impact. (b) Is kinetic energy conserved? Explain. (c) Evaluate the
loss of kinetic energy. Answers: (a) mv/(M +m), (c) ∆T = − 12 m+M mM 2
v .

6.2. The Motion of a Rocket

As an application of the law of conservation of linear momentum to
a system of variable mass, consider the motion of a rocket. Imagine the
rocket to be in empty space, far from any stars or any other material
bodies so there are no external forces acting on it. The pilot decides
to fire the rocket engines for a short time, say dt. The rocket engines
burn an amount of fuel dM . The burned fuel is expelled from the
rocket exhaust tubes at a velocity u with respect to the rocket. How
much faster is the rocket traveling after this fuel burn?
Let the initial mass of the rocket be M and let V be its initial ve-
locity, relative to some inertial reference frame. The initial momentum
is, therefore,
Pi = M V.
After the burn, the mass of the rocket is M − dM and its velocity
is V + dV. The mass of the burned fuel is dM and its velocity in
the inertial reference frame is the vector sum V + u. The total final
momentum of the rocket plus burned fuel is
Pf = (M − dM )(V + dV) + dM (V + u).
See Figure 6.1
Since there is no net external force acting on the rocket/fuel system,
the total momentum must be constant; that is, the momentum after
the burn must be equal to the momentum before the burn. Therefore,
(M − dM )(V+dV) + dM (V + u) = M V.
 6.2. THE MOTION OF A ROCKET 197

Figure 6.1. A rocket before and after it fires an amount

of fuel dM.

Next we carry out the indicated multiplications and discard all second
order differentials. (The term dM · dV is a second order differential.
The product of one infinitesimal quantity with another infinitesimal
quantity generates a very small quantity indeed!) This procedure yields
(and you should verify this result yourself):
M dV = −udM.
Dividing both sides by dt gives
dV dM
M = −u . (6.2)
dt dt
Consider the left hand side of (6.2). Since dV/dt is the acceleration of
the rocket and M is its mass, the left hand side is mass times acceler-
ation. Therefore, the right hand side looks like the force on the rocket,
usually called the thrust. The thrust depends on the rate at which fuel
is burned (dM/dt) and on the velocity with which the burned fuel is
ejected, u. It is convenient to express Equation (6.2) in scalar form as
dV dM
M = −u .
dt dt
The right hand side appears to have the wrong sign, but recall that
dM/dt is negative. Since M is changing, this equation is not of the
form F = ma. In fact, when solving problems, it is usually safer to
write
dM
dV = −u .
M
(What would you write if the rocket is slowing down by firing its retro-
rockets?)
To generate a large thrust, a rocket motor is designed to burn fuel
very rapidly (large dM/dt) and to expel it at as high a speed as possible
(large u). The whole purpose of burning the fuel is to convert it to a
198 6. CONSERVATION OF LINEAR MOMENTUM

gas, thus greatly increasing its volume and causing it to be expelled

through the rocket tubes at the highest possible velocity.

Worked Example 6.1. A rocket of total mass m0 in inter-

planetary space is coasting at speed v0 . It approaches an interest-
ing asteroid and must slow down to the speed of the asteroid which
happens to be v0 /2. Determine the amount of fuel that must be
burned to achieve this reduction in speed.
Solution: Since the rocket is slowing down we write
dM
dV = u .
M
Integrating
1 v0 /2
Z Z mf
dM
dV = ,
u v0 m0 M
1 mf
(v0 /2 − v0 ) = ln ,
u m0
v0 mf
− = ln ,
2u m0
mf = m0 e−v0 /2u .
The amount of fuel that must be burned is
m0 − mf = m0 (1 − e−v0 /2u ).

Worked Example 6.2. The rocket equation was obtained

by considering the difference in momentum between an initial and
final state, assuming a small amount of fuel was burned. Obtain
the same result by showing that the total momentum is constant.
Assume the rate of fuel burn is constant and equal to ṁ.
Solution: At any given moment the total momentum of the
system is the momentum of the rocket (including unburned fuel)
and the momentum of the expelled (burned) fuel. We shall solve
the problem using scalar quantities. Let the mass of the rocket
at any instant be MR and the mass of burned fuel be mf . The
ejection speed of the fuel in inertial space is VR − u where u is
a positive quantity. The rate of change of mass of the rocket is
dMR /dt = −ṁ and the rate of change of the mass of burned fuel is
 6.2. THE MOTION OF A ROCKET 199

dmf /dt = +ṁ. To determine the momentum of the burned fuel at

time τ we need to keep in mind that while it was being burned, the
rocket was moving at a changing speed. Therefore, the momentum
of the burned fuel at time τ is
Z τ Z τ
Pf (τ ) = ṁ(VR (t) − u)dt = ṁ (VR − u)dt
0 0
The momentum of the rocket at time τ is
PR (τ ) = MR (τ )VR (τ )
The total momentum is Ptot = PR + Pf so the rate of change of
momentum is
 Z τ 
dPtot d
= MR (τ )VR (τ ) + ṁ (VR − u)dt
dt dt 0
d
= (MR VR ) + ṁ(VR − u)
dt
dMR dVR
= VR + MR + ṁVR − ṁu = 0
dt dt
But ṁ = − dM
dt
R
, so
dVR dMR
MR = −u .
dt dt
Note that in this expression the ejection speed u is positive and
rate of mass decrease (dMR /dt) is negative.

Exercise 6.2. A rocket of initial mass 500 kg burns fuel at a rate of

5 kg/sec. The exhaust speed of the gases is 300 m/s. What is the initial
acceleration of the rocket? (Ignore gravity.) What is its acceleration
after one minute? Answers: 3 m/s2 , 7.5 m/s2 .

Exercise 6.3. Joe and Bill are railroad men who are riding side
by side on flatcars rolling on parallel tracks. The tracks are straight
and perfectly horizontal. There are absolutely no frictional forces or
air resistance. It starts to snow. Joe sweeps the snow off his flatcar as
soon as it lands, sweeping it off the side, perpendicular to the direc-
tion of motion of the flatcar. Bill, the lazy one, simply lets the snow
accumulate on his flatcar. Who travels further in the same interval
of time? Answer this question conceptually and also mathematically.
(The snow falls perfectly vertically.)
200 6. CONSERVATION OF LINEAR MOMENTUM

Exercise 6.4. Beginning with Newton’s second law in the form

F = dP/dt and the definition of derivative, generalize equation (6.2)
to include the effect of an external force. Answer: M dV
dt
+ u dM
dt
= F.
Exercise 6.5. Consider a rocket that is rising into the air. The
burning exhaust gases stream out through the nozzles and push down
on the air below. But what pushes upward on the rocket? (An equiv-
alent question: When you inflate a balloon and then release it, it flies
wildly about the room. What is pushing the balloon?)

6.3. Collisions
Another application of the principle of conservation of linear mo-
mentum is in the analysis of collisions. Generating and studying col-
lisions between elementary particles are a principal way physicists ex-
plore the underlying properties of nature. A collision between two
bodies often involves a strong, short range interactive force between
the bodies. When two extended bodies (such as automobiles or bil-
liard balls) come in contact, we idealize the collision and assume no
forces act except during the instant of contact. Taking the two bod-
ies as the entire system, these forces are internal forces. During an
ideal collision, no external forces are acting. Consequently, the total
momentum of the system is constant.
In a contact collision, there is a very short range repulsive force that
acts while the surfaces of the two bodies are touching.2 A glancing
collision, such as illustrated in Figure 6.2, occurs when the velocity
vectors of the two bodies are not aligned along the line of centers.
The distance b between the initial velocity vectors is called the impact
parameter.
A collision may not involve the actual physical contact of the two
bodies. The force exerted by one body on the other may be a long range
force such as the force of electric repulsion between an alpha particle
and the nucleus as in Rutherford’s experiment, or the force of gravity
when a comet in a hyperbolic orbit approaches from infinity, swings
2Theorigin of these forces is the repulsion between the electrons in one body
and the electrons in the other. When the two “electron clouds” start to overlap,
there is a repulsive Coulomb force between them. Fortunately, we do not need
detailed information about the force between the bodies because we can solve the
problem using the law of conservation of momentum in which internal forces play
no part.
 6.3. COLLISIONS 201

Figure 6.2. A glancing collision.

about the Sun, and travels back out to infinity. The same physics
applies regardless of the range of the forces.3
As a basic collision problem, consider the situation illustrated in
Figure 6.3 where body M1 with velocity V1 makes a glancing collision
with body M2 that is initially at rest. (We can always find a coordinate
system in which one body is initially at rest.) The two bodies move
off with velocities V10 and V20 at angles θ1 and θ2 relative to V1 , as
shown in the figure. To keep things simple, assume the bodies are not
rotating.
There are no external forces acting on the system so the law of
momentum conservation states that

Pi = Pf . (6.3)
That is, the initial momentum and the final momentum are equal. This
is a vector equation so it is equivalent to the three scalar equations
Pxi = Pxf , Pyi = Pyf , Pzi = Pzf ,
(if two vectors are equal, their components must be equal). It is con-
venient to place the origin of coordinates at the original position of
body M2 and to let the x-axis be defined by the direction of V1 . Select
the z-axis perpendicular to the plane of the motion, i.e., perpendicular
to the plane containing V10 and V20 . Then Pzf = 0. By momentum
conservation, Pzi = 0. Therefore the problem is two-dimensional; the
motion takes place entirely in the xy-plane.
3We often refer to colliding bodies as particles, even though they may be as-
tronomical objects. It is appropriate to use the term particle when we are dealing
with long range forces. In a glancing collision, the surfaces of two extended bodies
come in contact and they should not be called particles. Nevertheless, physicists
are a bit careless in the usage of this term, and in dealing with collisions the word
particle is often used when, strictly speaking, it should not be.
202 6. CONSERVATION OF LINEAR MOMENTUM

Figure 6.3. Illustration of the parameters involved

when two bodies undergo a glancing collision.

The momentum conservation equations in the xy-plane can be writ-

ten
0 0
Pxi = Pxf , or M1 V1x = M1 V1x + M2 V2x ,
0 0
Pyi = Pyf , or 0 = M1 V1y − M2 V2y .

In terms of the angles θ1 and θ2 these two equations are

M1 V1 = M1 V10 cos θ1 + M2 V20 cos θ2 , (6.4)

0 = M1 V10 sin θ1 − M2 V20 sin θ2 . (6.5)

Note the minus sign on the last term.

Conservation of momentum led to two equations. These equations
involve the seven parameters M1 , M2 , V1 , V10 , V20 , θ1 , θ2 . Obviously five
of these parameters must be “known” quantities so that you can solve
for two unknowns. In most collision problems you will know the speed
of the incoming particle, V1 . You will also probably know the masses
of the particles, or at least their ratio (which is sufficient). This leaves
you with four unknowns, namely the final speeds, V10 and V20 , and the
final directions, θ1 and θ2 . Two of these must be determined (perhaps
experimentally) before you can solve the problem for the other two.
If, however, the collision is elastic, then there is one more equation
you can use, namely the equation expressing the conservation of kinetic
energy. By definition, an elastic collision is one in which the kinetic
 6.3. COLLISIONS 203

energy is conserved.4 That is, Ti = Tf , or in terms of the problem,

1 1 1
M1 V12 = M1 V102 + M2 V202 . (6.6)
2 2 2
This condition gives you a third equation, allowing you to solve for
three unknowns. Frequently, the unknown quantities will be the final
velocities and the direction of one particle, for example, V10 , V20 , and θ2 .
The quantities you will need to know are (usually) the velocity of the
incoming particle and its deflection angle, as well as the masses.

Figure 6.4. Rutherford’s experiment. An alpha parti-

cle is scattered through the angle θ by a gold nucleus.
This collision experiment was crucial in determining the
structure of atoms.

To help you visualize a typical problem, Figure 6.4 illustrates Ruther-

ford’s famous experiment in which he bombarded the nuclei of gold
atoms with alpha particles emitted by a radioactive substance. The
gold nuclei were essentially at rest and the alpha particles approached
with a known velocity. After interacting with a gold nucleus the alpha
particles hit a screen painted with fluorescent material. This caused a
tiny flash of light to appear at the point where an alpha particle hit the
screen. The pinpoints of light were observed by eye by Rutherford’s
graduate students who tabulated the final positions of all the alpha
particles. In this problem the known quantities were M1 , M2 , V1 and
θ1 . The unknown quantities were V10 , V20 , and θ2 . (You will solve this
problem after studying central force problems; see Problem 10.20.)
4A collision is usually elastic if the two bodies do not come in contact, or if
neither body is deformed by the collision. Thus the interactions between celestial
bodies and between elementary particles are often elastic. Collisions of billiard balls
are frequently assumed to be elastic since the deformation is negligible.
204 6. CONSERVATION OF LINEAR MOMENTUM

Going back to the general case, I will now show you how to manip-
ulate momentum conservation and kinetic energy conservation, Equa-
tions (6.4), (6.5), and (6.6), to solve for the unknown quantities in
terms of the known quantities. Let me warn you that the algebra is a
bit tedious so you might want to get up and get a cup of coffee before
we start. Also, I am not going to give you all the intermediate steps,
so while you are up you had better get your pencil because you will
not understand the final result unless you can work out all the steps.
To begin, rewrite Equations (6.4) and (6.5), placing all the terms
with subscript 1 on one side. Thus:

M1 V1 − M1 V10 cos θ1 = M2 V20 cos θ2 , (6.7)

M1 V10 sin θ1 = M2 V20 sin θ2 . (6.8)
Next square both equations and add them together to eliminate θ2 .
You obtain
M12 V12 − 2M12 V1 V10 cos θ1 + M12 V102 = M22 V202 . (6.9)
This equation, together with the kinetic energy equation (6.6) give you
two equations in the two unknowns V10 and V20 . (The initial velocity V1
and the angle θ1 are assumed known.)
Now eliminate V20 from Equations (6.6) and (6.9). If you take the
expression for M22 V202 given by Equation (6.9) and plug it into Equation
(6.6), you obtain the following equation for V10 :
[(M1 + M2 )] V102 − [2M1 V1 cos θ1 ] V10 + (M1 − M2 )V12 = 0. (6.10)
 

All the terms in square brackets are known, so this is a quadratic

equation for V10 whose solution is
q
0
2M1 V1 cos θ1 ± (2M1 V1 cos θ1 )2 − 4 (M1 + M2 ) (M1 − M2 ) V12
V1 = ,
2 (M1 + M2 )
or
s 2
0 M 1 M1 V1 M1 − M2 2
V1 = V1 cos θ1 ± cos2 θ1 − V .
M1 + M2 M1 + M2 M1 + M2 1
Dividing through by V1 you obtain a nicer looking expression:
 s 
0
 2
V1 M1 cos θ1 ± cos2 θ1 − 1 + M2  .
= (6.11)
V1 M1 + M2 M1
 6.3. COLLISIONS 205

Let us interpret this result by discussing several special cases. Use

your experience with colliding bodies to visualize each situation.5 Case
1: Head-on Collisions: If two objects hit “head-on” then the motion
is one-dimensional and θ1 = θ2 = 0. The velocity vector of the incom-
ing particle, V1 , is pointed directly at the center of body M2 . Assume
M1 is moving from left to right. Recall from experience that if the
masses are equal, M1 stops and M2 moves off to the right. If M1 > M2
both masses move off to the right with M2 moving faster than M1 . If
M1 < M2 , then M1 bounces back (to the left) and M2 moves off to
the right. All of this information is incorporated in Equation (6.11).
I will now show you how to “read” Equation (6.11) to extract this
information.
Since we are imagining a head-on collision, the angle θ1 is zero.
Setting cos θ1 = 1 in equation(6.11), and doing a little bit of algebra,
leads to
M1 ± M2
V10 = V1 . (6.12)
M1 + M2
Subcase 1.1: M1 = M2
Consider first the situation in which M1 = M2 . Equation (6.12)
then reduces to
V10
 
1 1
= [1 ± 1] = .
V1 2 0
That is, V10 is either equal to V1 or it is equal to zero. From experi-
ence we expect V10 to be zero (the incoming object stops). What does
V10 = V1 correspond to? It corresponds to a miss, that is, no collision
at all. It tells us that the final velocity of M1 is equal to its initial
velocity. Furthermore, Equations (6.6) and (6.9) both tell us that for
a miss, V20 = 0. The condition θ1 = θ2 = 0 is equally satisfied by a
head-on collision or a miss. (Note how much physical information is
coded into the equations!)
We are not interested in the situation where one particle misses the
other, so let us set V10 = 0. What is V20 , the final velocity of the particle
that was initially at rest? Again, from experience we expect it to be
equal to V1 , the initial velocity of the incoming object. And, indeed,
Equation (6.7) for M1 = M2 and θ1 = 0 yields
V1 − V10 = V20
But V10 = 0, so,
V20 = V1 .
5You can make a rough check of the results we obtain by carrying out experi-
ments on a smooth surface with coins of the same or different masses.
206 6. CONSERVATION OF LINEAR MOMENTUM

This equality corresponds to our experience that the struck object flies
off with the velocity the incoming particle had before the collision.
Subcase 1.2 M1 6= M2
Now allow the two colliding bodies to have different masses, but
still assume a head-on collision so θ1 = θ2 = 0. Again, Equation (6.11)
reduces to Equation (6.12). Selecting the plus sign gives V10 = V1 ,
implying no collision at all. This case is of no interest, so select the
minus sign. That is,
M1 − M2
V10 = V1 . (6.13)
M1 + M2
Therefore, if M1 > M2 , the velocity V1 0 is positive. Plugging Equation
(6.13) into (6.7) and keeping in mind that θ1 = θ2 = 0, you get the
velocity acquired by M2 :
M1
V20 = 2V1 . (6.14)
M1 + M2
Equations (6.13) and (6.14) agree with your experience and intuition.
If a heavy (massive) object collides with a light (less massive) object,
the heavy object will keep on moving in its original direction. If a
billiard ball strikes a ping-pong ball, M1 >> M2 , and
. M1
V10 = V1 = V1 .
M1
That is, the billiard ball keeps on moving at essentially its original
velocity. What happens to the ping-pong ball? According to Equation
(6.14), if M1 >> M2 , then V20 = 2V1 . That is, the light object moves
off with a speed of twice the speed of the incoming heavy object.
It is easy to appreciate that if M1 < M2 , the situation is simply
reversed. If a light object hits a heavy object then M1 < M2 and V10 is
negative. That is, the light object bounces back. For example, if you
throw a ball against a wall, then M1 is the mass of the ball and M2
is the mass of the Earth (assuming the wall is attached to the Earth).
So M1 << M2 and according to Equation (6.13), V10 , the final velocity
of the ball, will be V10 = −V1 . The ball bounces back with its initial
speed. Similarly, V20 = 0. The Earth stands still.

Worked Example 6.3. A spacecraft flyby of a planet can

be used to give the spacecraft a “boost” and increase its velocity
by the process known as the “slingshot effect.” To underscore the
physics of this process, this example is unrealistic because we will
 6.3. COLLISIONS 207

assume a spacecraft approaches Mars, swings around the planet,

and heads off in the opposite direction. This would be (essentially)
a head-on collision. A realistic problem would have the spacecraft
approach the planet at some angle to its velocity vector and be
“scattered” at another angle.
In this problem we assume the spacecraft has a mass of 2000
kg and that Mars has a mass of 6.4×1023 kg. The initial speed
of the spacecraft is -12 km/sec and the initial speed of Mars is
+23.36 km/sec. (The speeds are relative to a coordinate system
at rest with respect to the Sun.) (a) Determine the final speed of
the spacecraft. (b) Evaluate the ratio of the final to initial kinetic
energy of the spacecraft.
Solution: As mentioned, this is (essentially) a two body head-
on collision. To analyze it we transform to a reference frame in
which one particle (Mars, in this case) is at rest. Then the initial
velocities are:
V2 = 0,
V1 = 23.36 + 12 = 35.36 km/s.
Equations (6.13) and (6.14) give the final velocities.
2000 − 6.4 × 1023 .
 
0 M1 − M2
V1 = V1 = 35.36 = −35.36 km/s.
M1 + M2 2000 + 6.4 × 1023
 
0 M1 2000 .
V2 = 2V1 = (2)(35.36) = 0.
M1 + M2 2000 + 6.4 × 1023
These are the speeds in a reference frame in which Mars is at rest.
Transforming back to the “inertial” reference frame we find the
speed of the spacecraft is
vs0 = V10 − vmars = −35.36 − 23.36 = −58.72 km/s.
The ratio of kinetic energies is
1
Tf ms vs2 (58.72)2
= 12 = = 23.9.
Ti m v 02
2 s s
(12)2
(Note: This problem is artificial because it assumes a one di-
mensional situation in which the spacecraft’s velocity vector and
the planet’s velocity vector are opposite to one another. Such a
situation could be set up but would require thrusts to adjust the
spacecraft’s speed.)
208 6. CONSERVATION OF LINEAR MOMENTUM

Exercise 6.6. Fill in the missing steps to obtain Equation (6.12)

from (6.11).
Exercise 6.7. (a) A ball of mass m traveling at speed v hits a ball
of mass 3m at rest. Determine the final velocities of the two balls. (b)
A ball of mass 3m traveling at speed v hits a ball of mass m at rest.
Determine the final velocities of the two balls. Assume head-on elastic
collisions. Answers: (a) -v/2 and v/2, (b) v/2 and 3v/2.

Case 2: Glancing Collisions: Let us now analyze a glancing

collision. The final directions of the two objects are given by nonzero
angles θ1 and θ2 , shown in Figure 6.3. As before, there are three possi-
bilities, M1 = M2 , M1 > M2 , and M1 < M2 . Once again the discussion
is based on Equation (6.11). Note that cos2 θ ranges from 1 to 0 so the
term under the radical varies from a minimum of (M2 /M1 )2 − 1 to a
maximum of (M2 /M1 )2 .
Subcase 2.1: M1 = M2
Let us begin by assuming M1 = M2 . Equation (6.11) then reduces
to
V10 1h p i 
0
= cos θ1 ± cos2 θ1 = . (6.15)
V1 2 cos θ1
Thus, the final velocity of M1 is either zero or V1 cos θ1 . The solution
V10 = 0 can be discarded because it implies a head-on collision, as
considered previously. Recall that V1 and θ1 are assumed known, so
the lower solution of Equation (6.15) yields the first of our unknown
quantities, the velocity of M1 after the collision:
V10 = V1 cos θ1 .
Plugging this into Equation (6.6) yields
V12 = (V1 cos θ1 )2 + V202 ,
or
2
(V20 ) = V12 (1 − cos2 θ1 ) = V12 sin2 θ1 , (6.16)
so
V20 = V1 sin θ1 .
This gives us the second unknown quantity, V20 . The third “unknown”
is θ2 and it can be obtained immediately from the conservation of mo-
mentum along the y−axis, Equation (6.5), as
V10
θ2 = sin−1 .
V1
 6.3. COLLISIONS 209

Worked Example 6.4. Prove that in any glancing elastic

collision between bodies of equal mass, the sum of the deflection
angles is θ1 + θ2 = π/2.
Solution: Given V10 = V1 cos θ1 and V20 = V1 sin θ1 . Plug into
the conservation of momentum equation V10 sin θ1 = V20 sin θ2 to get
V1 cos θ1 sin θ1 = V1 sin θ1 sin θ2
cos θ1 = sin θ2
But cos θ1 = sin π2 − θ1 , so θ2 = π2 − θ1, or

π
θ1 + θ2 = .
2

Exercise 6.8. Show that the null solution for Equation (6.15) im-
plies a head-on collision.
Exercise 6.9. Equation 6.16 leads to the two solutions V20 = ±V1 sin θ1 .
Using the facts that θ1 + θ2 = π/2 and V10 = V1 cos θ1 , show that only
the positive solution is obtained. (The negative solution corresponds
to θ1 = 0 and V10 = V1 , that is, a miss.)

Subcase 2.2: M1 > M2

If the mass of body 1 is greater than the mass of body 2, you expect
the heavier object (M1 ) to continue moving in the forward direction.
That is, you expect θ1 to be less than π/2. To appreciate that this is
true, note that the quantity under the radical in Equation (6.11) must
be positive. (Otherwise the velocity V10 would be a complex number
and this is not possible because physically measurable quantities must
be real.) The quantity under the radical is non-negative for M1 > M2
only if
2 M22
cos θ1 ≥ 1 − 2 .
M1
This means that the angle θ1 can range from zero to some maximum
value θmax given by
1
M22 2

−1
θmax = cos 1− 2 .
M1
210 6. CONSERVATION OF LINEAR MOMENTUM

In the limit M2 /M1 → 1 you obtain θmax = cos−1 (0) = π/2. In the
limit M2 /M1 → 0 you obtain θmax = cos−1 (1) = 0. Therefore,
π
0 < θ1 < .
2
This tells you that if the incoming object is the more massive body, it
will be scattered through an angle θ1 smaller than π/2.
Subcase 2.3: M1 < M2
The case of a glancing collision in which M1 < M2 can also be
analyzed with Equation (6.11). It is left as an exercise to show that if
the target M2 is much more massive than M1 , then the incoming body
will bounce off with its original speed (but with a change in direction).
Note that this case reduces to Subcase 2.2 if you interchange the two
bodies.
In conclusion, you have seen that an elastic collision between two
bodies can be analyzed using the conservation of linear momentum
and the conservation of kinetic energy. The equations obtained are
amazingly complex for such a simple problem. An important benefit
you should get from the analysis of collisions is an appreciation for how
to extract physical meaning from mathematical relations. Be aware
that in a collision the conservation of momentum always holds, but
conservation of kinetic energy may not.

Exercise 6.10. Show that if M1 << M2 the speed of M1 after the

collision is (essentially) the same as its speed before the collision.
Exercise 6.11. A billiard ball with velocity v strikes a second bil-
liard ball at rest with a glancing collision. The first ball is observed to
emerge from the collision at an angle of 20◦ . Determine the speed and
direction of the second ball.

6.4. Inelastic Collisions. The Coefficient of Restitution

In the previous section the collision between the two bodies was
elastic: There was no loss of kinetic energy. Of course, this is not
always the case. It is customary to represent the gain or loss of kinetic
energy by a quantity called the “Q value,” defined by
Q = Tf − Ti ,
where Tf is the total final kinetic energy and Ti is the total initial
kinetic energy.
 6.5. IMPULSE 211

For an elastic collision, Q = 0. This condition is generally not met;

most collisions are either endoergic in which kinetic energy is lost, or
exoergic in which kinetic energy is gained. For example, in the collision
of two putty balls having equal but opposite momenta, all of the kinetic
energy is lost. This collision is completely inelastic and Q is negative.
On the other hand, a collision between two molecules may involve an
exothermic chemical reaction in which chemical energy is transformed
into mechanical energy. For such a collision, Q is positive.
A closely related concept is the coefficient of restitution. This
was originally described by Isaac Newton who observed that for any
head-on collision of two non-rotating bodies the ratio of relative final
velocities to relative initial velocities is a constant. That is, if V1 and
V2 are the initial velocites and V10 and V20 are the final velocities, then
Newton’s Rule is
|V20 − V10 |
e= .
|V2 − V1 |
For an elastic collision, e = 1. This is easily demonstrated from
Equations (6.13) and (6.14). For a completely inelastic collision in
which all energy is lost, e = 0.
If the collision is a glancing collision, then the velocities to be used
in Newton’s formula are the velocity components along the line joining
the two bodies.6

Exercise 6.12. Prove that e = 0 for a completely inelastic collision

and e = 1 for an elastic collision. You may assume a head-on collision.
Exercise 6.13. Using Equations (6.13) and (6.14) show that Q = 0
for a head-on elastic collision. (This is obviously true from the defi-
nition of elastic collision; the purpose of the exercise is to give you
experience in manipulating the relations.)

6.5. Impulse
When a bat hits a baseball, a large force acts for a short period of
time. Such a blow gives rise to an impulse. By definition, an impulse
6Thecoefficient of restitution actually depends on various other factors, such
as the medium in which the collision occurs, but Newton’s formula is a good
approximation.
212 6. CONSERVATION OF LINEAR MOMENTUM

is the time integral of a force. Denoting the impulse by J, we can write

Z τ
J= Fdt,
0
where τ is the time during which the force acts. In general, it would be
difficult to evaluate the integral on the right because the force is usually
an unknown, and probably complicated, function of time. Nevertheless,
it is easy to evaluate J because from Newton’s second law, F = dp dt
, so
Z τ Z τ Z pf
dp
J= Fdt = dt = dp = pf − pi = ∆p.
0 0 dt pi
That is, the impulse is simply equal to the change in momentum.

Exercise 6.14. A force F = 3 sin 5t N acts on a particle of mass

2 kg that was initially at rest. The force acts during the time interval
from t = 0 to t = π/10 s. What is the final velocity of the particle?
Answer: 0.3 m/s.
Exercise 6.15. A stationary block sitting on a frictionless surface
is acted upon by a force (in newtons) given by
F = 2t for 0 ≤ t ≤ 2,
F = 4 for 2 ≤ t ≤ 5,
F = −t for 5 ≤ t ≤ 7.
(Times in seconds.) Determine the final momentum of the block.
Answer: 4 kg m/s.

6.6. Momentum of a System of Particles

In this book we have claimed more than once that internal forces
do not affect the total linear momentum of a mechanical system. For
example if a bomb explodes, pieces fly off in all directions, but the total
momentum is unchanged. We now prove the claim by considering a sys-
tem of N particles and determining the effect of internal and external
forces on the momentum. The particles have masses m1 , m2 , · · · , mN
and are located at positions r1 , r2 , · · · , rN .
All of the particles exert forces on each other. These forces are
internal forces. Denote by Fij the force exerted on particle i by particle
j.
Fij = force acting on particle i, due to particle j.
 6.6. MOMENTUM OF A SYSTEM OF PARTICLES 213

Newton’s second law, applied to particle i is, then,

dpi (e)
X
= Fi + Fij ,
dt j=1,N
j6=i

(e)
where Fi is the external force acting on i. Note that the summation
over the internal forces is subject to the condition j 6= i because a
particle cannot exert a force on itself.
There is one such equation for each particle. Adding all N equations
yields
N N N N
X dpi dPtot X (e) X X
= = Fi + Fij ,
i
dt dt i i j6=i
P
where Ptot = pi is the total momentum of the system. The last term
(the double sum) is zero because by Newton’s third law, Fij = −Fji .
By writing out a few terms you will see that the double sum consists
of pairs of terms that cancel each other out. Therefore,
dPtot X (e) (e)
= Fi = Ftot (6.17)
dt i
(e)
where Ftot is the total (or net, or resultant) external force acting on the
system. Thus, we have shown that the internal forces have no effect on
the total momentum.
This result has another important consequence. If the masses of
the particles are constant, we can write
dPtot X
= mi r̈i ,
dt i

and Equation (6.17) can be written

X (e)
mi r̈i = Ftot .
i

The definition of center of mass (rc ) for a system of particles having

total mass M is (see Equation 5.17)
X
M rc = mi ri .
i

Differentiating twice with respect to time gives

X
M r̈c = mi r̈i . (6.18)
i
214 6. CONSERVATION OF LINEAR MOMENTUM
P (e)
But i mi r̈i = Ftot so
(e)
M r̈c = Ftot . (6.19)
This important results states that the center of mass of a system of
particles moves like a particle of mass M acted upon by the resultant
of the sum of all the external forces, regardless of their point of ap-
plication. A corollary is that if a system is not acted upon by a net
external force, the center of mass will move with constant velocity.

m1
r2-r1
r1
m2
r2
O

Figure 6.5. Relative positions of two particles. Parti-

cles m1 and m2 are located at r1 and r2 with respect to
the origin. The relative coordinate r gives the position
of m2 with respect to m1 .

6.7. Relative Motion and the Reduced Mass

When we study the motion of two interacting bodies such as a star
and a planet, or an electron and a nucleus, we are often interested in
their relative motion and do not care about their motion with respect
to an inertial reference frame. For such problems it is convenient to
introduce the concepts of relative coordinate and reduced mass. The
introduction of these quantities allows us to replace the two-body prob-
lem (with two equations of motion) by a single one-body problem (and
only one equation of motion).7
Consider a system consisting of two particles (m1 and m2 ) that are
exerting equal and opposite forces on each other. Assume no external
forces are acting. Let F be the force on m2 due to m1 . Then the force
on m1 is −F. Consequently, the equations of motion of the two particles
are:
m1 r̈1 = −F,
7This
is a very important simplification allowing us to obtain (for example)
the motion of a planet relative to the Sun. There is no such simplification for a
system of three bodies, although physicists have been trying to solve the “three
body problem” for hundreds of years. Some mathematicians claim the problem is
actually unsolvable.
 6.7. RELATIVE MOTION AND THE REDUCED MASS 215

and
m2 r̈2 = +F.
Figure 6.5 shows two particles, m1 and m2 , at positions r1 and r2
with respect to the inertial origin O. The “relative” vector r gives the
position of m2 with respect to m1 and (by tip-to-tail addition) is given
by
r = r2 − r1 .
Differentiating the relative coordinate twice with respect to time yields
r̈ = r̈2 − r̈1 .
Substituting for r̈2 and r̈1 from the equations of motion gives
 
F F m1 + m2
r̈ = + = F
m2 m1 m1 m2
or  
m1 m2
r̈ = F.
m1 + m2
The quantity mm11+m
m2
2
is called the “reduced mass.” It is denoted by µ.
So the equation of motion for the relative coordinate is
µr̈ = F. (6.20)
As an example, consider a system composed of a star and a planet.
The star is much more massive than the planet and for all intents
and purposes the star remains at rest and the planet orbits around it.
If the mass of the star is m1 and the mass of the planet is m2 and
if m1 >> m2 , then µ ∼ = m2 . On the other hand, for a binary star
system both masses may be approximately equal and the two stars
orbit around their common center of mass. If both stars have the same
mass, say m, the reduced mass is µ = 21 m.
Consider the Sun-Earth system. The relative coordinate gives the
distance from the Sun to the Earth. F is the force the Sun exerts on the
Earth. In an analysis of this system you should write µr̈ = F and not
m2 r̈ = F. The reason is that the Sun is accelerating, so a coordinate
system with origin at the Sun is not an inertial coordinate system
and Newton’s second law does not hold. (Actually for the Sun and
Earth, m2 and µ are so nearly equal that the error in using m2 r̈ = F
is negligible.)

Exercise 6.16. Where is the center of mass of the Sun-Earth sys-

tem? (Look up the necessary values.) Answer: 4.5 × 105 meters from
center of the Sun.
216 6. CONSERVATION OF LINEAR MOMENTUM

Exercise 6.17. Determine the reduced mass of the Sun-Jupiter

system and the reduced mass of the Earth-Moon system. Answer: For
Earth-Moon, µ = 7.26 × 1022 kg.

6.8. Collisions in Center of Mass Coordinates (Optional)

When studying two body collisions, physicists frequently use a co-
ordinate system that is moving with the center of mass because it is
often an easier and quicker way to solve the problem. This approach
is particularly useful when studying collisions between elementary par-
ticles. If you specialize in high energy physics, you will become very
familiar with center-of-mass coordinates.
Let me remind you that if there are no external forces acting on
the system, the center of mass moves with constant velocity (Equation
6.19). In that case, a coordinate system moving with the center of mass
is an inertial (i.e., non-accelerating) coordinate system.

m1 v'1cm
v1cm ϕ v2cm
m1 ϕ m2
m2
v'2cm
Figure 6.6. A collision as seen in the center-of-mass
coordinate system.

From the center-of-mass point of view, in a two body collision both

bodies are moving, approaching one another and the center of mass.
They collide at the center of mass. After the collision, they recede from
the center of mass.
In the center-of-mass frame, the velocities of the two particles before
the collision are denoted v1cm and v2cm . After the collision the particles
0 0
recede in opposite directions with velocities v1cm and v2cm . The angle
between the incoming and outgoing directions is denoted φ, as shown
 6.8. COLLISIONS IN CENTER OF MASS COORDINATES (OPTIONAL) 217

in Figure 6.6. Note that there is now a single angle, so we have already
achieved some simplification of the problem.
Figure 6.7 shows the positions of the two particles and their center
of mass relative to the origin O of an arbitrary inertial frame. The
center of mass (indicated by the small open circle) lies on the line
joining the two particles and is at rc relative to O. The coordinates
r1cm and r2cm give the positions of the particles relative to the center
of mass. Note that
r1cm = r1 −rc
r2cm = r2 −rc

Figure 6.7. The center-of-mass coordinates. r1cm and

r2cm give the positions of m1 and m2 with respect to the
center of mass. The open circle specifies the position of
the center of mass.

The total initial momentum in the center of mass system is zero.

To prove this, take the time derivative of the definition of the center of
mass, rc = (m1 r1 + m2 r2 )/M, where M = m1 + m2 :
M ṙc = (m1 ṙ1 + m2 ṙ2 ) = (m1 ṙ1cm + m1 ṙc + m2 ṙ2cm + m2 ṙc ),
so,
M ṙc − (m1 + m2 )ṙc = m1 ṙ1cm + m2 ṙ2cm ,
0 = m1 v1cm + m2 v2cm .
Therefore, the total initial momentum in the center of mass system is
p1cm + p2cm = 0. (6.21)
The total final momentum must also be zero,
p01cm + p02cm = 0. (6.22)
218 6. CONSERVATION OF LINEAR MOMENTUM

The energy equation in the cm system is

p21cm p22cm p02

1cm p02
+ = + 2cm + Q, (6.23)
2m1 2m2 2m1 2m2

where the factor Q represents any energy gained or lost during the
collision. For the rest of this section I will assume elastic collisions so
Q = 0.
Let us go back to the basic collision of Section 6.3 in which a particle
of mass m1 and speed V1 strikes a particle of mass m2 at rest, but now
we analyze the problem in the center of mass coordinate system. I will
use capital letters (such as V1lab ) for speeds in the laboratory frame,
and small letters (such as v1cm ) for speeds in the center of mass frame.
The speed of the center of mass in the lab frame is denoted Vc and
is obtained by taking the time derivative of the definition of center of
mass,
m1 r1 + m2 r2
rc = ,
M
m1 ṙ1 + m2 ṙ2
∴ Vc = .
M
Since ṙ1 = V1lab and ṙ2 = V2lab = 0 the velocity of the center of mass
is
m1
Vc = V1lab . (6.24)
M
In the laboratory system the origin of coordinates is located at
the initial position of the particle at rest. After the collision the two
particles move off at angles θ1 and θ2 . In other words, the laboratory
frame is described by Figure 6.3.
The transformation between laboratory frame velocities and center
of mass velocities is illustrated in Figure 6.8, which shows the rela-
0 0
tionship between v1cm and V1lab , as seen in the laboratory frame of
reference. In constructing the figure I used the vector relationship

V0 1lab = v0 1cm + Vc . (6.25)

This equation tells you that to convert velocities from the cm frame to
the lab frame you simply add the velocity of the center of mass.
We now obtain some useful relationships between the velocities in
the two coordinate systems. The velocity of the center of mass in the
lab system is given by Equation (6.24). Then, according to Figure 6.8
 6.8. COLLISIONS IN CENTER OF MASS COORDINATES (OPTIONAL) 219

Vc
v'1cm
V'1lab
ϕ
θ1
direction of V1lab
Figure 6.8. Relationship between velocities in the cen-
ter of mass system and the laboratory system for the
final velocities of particle number 1.

and Equation (6.25),

v1cm = V1lab − Vc , (6.26)
 m1 
= V1lab 1 − ,
M
= V1lab (m2 /M ),
and similarly for the other velocities.
Looking at Figure 6.8 you can see that the velocity components are
given by
0 0
V1lab sin θ1 = v1cm sin φ,
and
0 0
V1lab cos θ1 = v1cm cos φ + Vc .
These two equations allow you to determine the relationship between
θ1 and φ. Dividing one equation by the other yields
0
v1cm sin φ sin φ
tan θ1 = 0
= 0
. (6.27)
v1cm cos φ + Vc cos φ + (Vc /v1cm )
Although this gives an expression for θ1 in terms of φ, it is not in a very
convenient form. As you will see in a moment, it is possible to express
θ1 in terms of φ and the masses of the particles. To do so, begin by
0
expressing both Vc and v1cm in terms of the relative velocity, vrel :
vrel = V2lab −V1lab = V02lab −V01lab .
(This equation implies that the relative velocity is constant. However,
that is only true for an elastic collision.)
Since V2lab = 0, you can write
vrel = −V1lab .
220 6. CONSERVATION OF LINEAR MOMENTUM

Use Equation (6.24) to form the ratio

m1
Vc V1lab m1
= M =− ,
vrel −V1lab M
so  m  m
1 1
Vc = |Vc | = − vrel  = vrel .
 
M M
0
Similarly, you can obtain an expression for v1cm as follows:
m m2 0 
1
0
v1cm 0
= V1lab −Vc = V01lab − 0
V1lab + V ,
M M 2lab
0
 m1  m2 0 m2 0 0
= V1lab 1− − V2lab = (V1lab − V2lab ),
M M M
m2
= (−vrel ) ,
M
and consequently,
0 m2
v1cm = vrel .
M
0
Having obtained expressions for Vc and v1cm you can write Equation
(6.27) as
sin φ sin φ
tan θ1 = (m1 /M )vrel
= . (6.28)
cos φ + ( ) cos φ + (m1 /m2 )
(m2 /M )vrel

It is left as a problem to show that

sin φ
tan θ2 = . (6.29)
1 − cos φ
If V2lab = 0 and V1lab and θ1 are known lab frame variables, you can
determine the center of mass coordinate φ from (6.28). Then Equation
(6.29) yields θ2 . The value of v1cm is obtained from Equation (6.26).
0 0
Since vrel = −V1lab you can determine v1cm from v1cm = (m2 /M )vrel .
0 0
Finally, v2cm = −(m1 /m2 )v1cm . If desired, you can then transform back
to laboratory coordinates.

Worked Example 6.5. Using the center-of-mass coordinate

system, show that in an elastic collision between two particles of
equal mass, the scattering angles in the laboratory system add up
to π/2. (That is, show that θ2 + θ1 = π/2.)
Solution: If m1 = m2 , Equation (6.28) can be written as
sin φ φ
tan θ1 = = tan .
cos φ + 1 2
 6.9. SUMMARY 221

(The last step requires the “half angle” trigonometric identity.)

This result indicates that φ/2 = θ1 . Similarly, Equation 6.29 leads
to
sin φ φ
tan θ2 = = cot .
1 − cos φ 2
Combining these relations gives
tan θ2 = cot θ1 .
Another trigonometric identity is
tan θ1 + tan θ2
tan(θ1 + θ2 ) =
1 − tan θ1 tan θ2
so
tan θ1 + tan θ2 tan θ1 + tan θ2
tan(θ1 + θ2 ) = = → ∞.
1 − tan θ1 cot θ1 0
Therefore,

θ2 + θ1 = π/2.
This result indicates that if the two masses are equal, the angle
between the two outgoing particles in the lab system is a right
angle.

Exercise 6.18. Use the center-of-mass system to show that if

m1 << m2 the scattering angle in the cm system is (almost) equal
to the scattering angle in the lab system. That is, show that θ1 ≈ φ.

6.9. Summary
The law of conservation of linear momentum is applicable to nu-
merous problems. In this chapter we have been particularly interested
in two problems: the motion of a rocket and the collision of two masses.
Conservation of momentum is based on Newton’s second law. Since
F =dp/dt, if the net external force is zero, the momentum is constant.
A system such as a rocket (or a conveyor belt) whose mass is chang-
ing with time but which is not acted upon by external forces, is ana-
lyzed by requiring that the initial and final momenta be equal. For the
rocket, this leads to
dV dM
M = −u ,
dt dt
222 6. CONSERVATION OF LINEAR MOMENTUM

or,
dM
dV = −u ,
M
where u is the speed of the ejected gases relative to the rocket.
The analysis of a glancing collision of two objects is also based on
the conservation of linear momentum. If the masses are given, the
problem can be expressed in terms of the parameters V1 , V10 , V20 , θ1 , θ2 .
Momentum conservation yields two equations. Thus you need to be
given three of these parameters. In the case of elastic collisions, con-
servation of kinetic energy gives you an additional equation and you
only need to be given two of the parameters.
Assuming an elastic collision and that V1 and θ1 are given, the
conservation laws lead to an equation for V10 , the final velocity of M1 .
 s 
0
 2
V1 M1 cos θ1 ± cos2 θ1 − 1 + M2  .
=
V1 (M1 + M2 ) M1

The final velocity of M2 is given by

V202 = [M12 V12 − 2M12 V1 V10 cos θ1 + M12 V102 ]/M22 .
Finally, θ2 can be determined from Equation (6.8),
M2 V20
sin θ1 = sin θ2 .
M1 V10
A sharp blow involves a force F that acts during a short time in-
terval τ. The impulse J of such a force is the time integral of the force
and is equal to the change in momentum. Thus,
Z τ
J= Fdt = ∆p = pf −pi .
0
A system of particles is acted upon by internal and
Pexternal forces.
The center of mass moves like a particle of mass M = mi acted upon
by the total external force:
(e)
M r̈c = Ftot .
When studying the relative motion of two particles, it is convenient
to introduce the reduced mass µ, given by
m1 m2
µ= .
m1 + m2
Then, the equation of motion for the relative coordinate r is F =µr̈.
Collisions are frequently studied in the center-of-mass coordinate
system in which the total initial and final momenta are zero. For
 6.10. PROBLEMS 223

elastic collisions, the scattering angles in the lab and cm systems are
related by
sin φ
tan θ1 = 0
,
cos φ + (Vc /v1cm )
sin φ
tan θ2 = ,
1 − cos φ
where θ1 and θ2 are the scattering angles in the laboratory coordinate
system and φ is the (single) scattering angle in the center-of-mass sys-
tem.

6.10. Problems
Problem 6.1. A 90 kg railroad worker is on a handcar of mass 200
kg. The handcar is moving at 5 m/s when it passes under a tree. (a)
The railroad worker leaps upwards, grabs a limb and hangs on. Does
the speed of the handcar change? If so, determine its final velocity. (b)
Now consider the converse problem. The empty handcar is moving at
5 m/s when it passes under a tree and a 90 kg railroad worker drops
out of the tree onto the handcar. In this case, does the speed of the
handcar change? If so, determine its final velocity.
Problem 6.2. Prove that in a one dimensional elastic collision
between two bodies, the relative velocity between the bodies has the
same magnitude before and after the collision, but the opposite sign.
Problem 6.3. (The ballistic pendulum.) A ballistic pendulum
can be used to determine the speed of the bullet fired from a rifle by
determining its effect when it hits and is embedded in a pendulum.
Consider a ballistic pendulum that consists of a suspended block of
wood of mass M . A bullet of mass m and initial velocity v is fired
into and becomes embedded in the block. To block swings upward a
height h. (See Figure 6.9) Derive an equation for v in terms of the given
quantities.

Problem 6.4. A raindrop is falling through fog and is picking up

tiny water droplets as it falls. (a) Justify that the rate of change of
mass of the raindrop is proportional to r2 v, where r is the radius of the
raindrop and v is its downward velocity. (b) Prove that the acceleration
of the drop is g/7. (Ignore air resistance.)
Problem 6.5. A box filled with sand is placed on a sled and slides
down an ice covered hill (so that friction is negligible). Sand is leaking
224 6. CONSERVATION OF LINEAR MOMENTUM

Figure 6.9. The ballistic pendulum.

out of a hole in the box at a constant rate. The slope of the hill is α.
(a) Show that the equation of motion of the box (plus sled) is just
dv
m = mg sin α.
dt
(b) Now assume that the sand is somehow thrown out of the box with
a velocity −v, that is a velocity in the direction opposite to the motion
of the box but with the same speed as the box. Show that the equation
of motion in this case is
dv dm
m =v + mg sin α.
dt dt
Problem 6.6. A jet boat (or jet ski) operates on the following
principle: Water is drawn through an inlet into a turbine and pumped
out at high speed through a smaller opening. The manufacturer of a
jet boat states that the turbine draws 50 gallons of water per second
and expels it at a pressure of 80 psi. The manufacturer states that the
thrust developed is over 2000 lbs. Determine whether or not this is a
realistic value. (Hint: Look up Bernoulli’s equation.)
Problem 6.7. A spherical asteroid of mass m0 is moving freely
in interstellar space with velocity v0 . It runs into a dust cloud whose
uniform density is ρd . Assume that every particle of dust that hits the
asteroid sticks to it. (a) Obtain an expression for the velocity of the
asteroid as a function of time. (b) Obtain an expression for the force
exerted on the asteroid by the dust as a function of time. Answer (a)
h i−3/4
4/3
v = v0 m0 m0 + 4kv 3K
0 m0
2/3 t where k = ρd π and K = m/r3 .

Problem 6.8. During a war being waged in Antarctica, an armored

car of mass 2000 kg with a machine gun mounted on its roof drives onto
a frozen lake at 30 km/hr. The ice is, of course, perfectly frictionless,
and the truck will continue to slide in a straight line at a constant
velocity directly into the enemy camp unless it can make a 90◦ turn
 6.10. PROBLEMS 225

and slide to safety. G. I. Joe (who studied physics in college) jumps

to the roof, swivels the machine gun, and begins firing in a direction
perpendicular to the motion. The bullets have a mass of 500 grams
each and leave the gun with a velocity of 800 m/s. The gun fires at
a rate of 200 bullets per minute. How long must G. I. Joe fire the
machine gun for the car’s motion to be deviated by 90◦ ? Since the
bullets are being fired so rapidly, you can assume the mass decrease is
continuous. Answer: 17.55 sec.
Problem 6.9. The man on the flying trapeze is hanging by his
knees from the cross bar. The woman trapeze artist stands on the
circus floor. The trapeze starts at an angle of 60◦ from the vertical.
At the bottom of the swing, the man grabs the woman and they both
swing upward. To what angle will the trapeze swing with both artists
on it? For simplicity, assume the man is a point mass mM and the
woman is a point mass mW . The rope has length l and negligible mass.
Problem 6.10. A rocket motor is undergoing a “bench test.” It is
attached to a fixed support by four large springs of constant 106 N/m.
The motor burns fuel at a rate of 50 kg/s. When the motor is running
the springs are observed to stretch 1.5 cm. Determine the exhaust
speed of the burned fuel.
Problem 6.11. Most airports have a conveyer belt behind the
check-in counter for luggage to be transported to the airplane. Bags
of mass m are dropped onto the belt a rate of k per second. (You can
assume the mass increase is constant.) What is the increased power
load during the time the bags are being placed on the belt? Show that
the extra power required is twice the rate of increase of kinetic energy.
Explain what is happening to the “missing” power.
Problem 6.12. The “people mover” at the San Francisco airport
is essentially a horizontal, very long conveyor belt running at speed
v. Suppose that initially there are no people on the belt. The motor
driving the belt is drawing W watts of electrical power. A flight arrives
and passengers (all having the same mass) step onto the belt, one after
another at one second intervals, so that the mass being carried increases
at a constant rate k kg/sec. Assume the people stepping onto the belt
had an initial speed of v/2. If the belt is to continue running at the
same speed, how much extra electrical power must be supplied to the
motor?
Problem 6.13. A physics student is holding a vertical hanging
chain by its top link. The bottom link is just touching the top surface
226 6. CONSERVATION OF LINEAR MOMENTUM

of a scale. The student lets go of the chain and observes the reading
on the scale while the chain is dropping. The student claims that
the reading on the scale is three times the weight of the length of
the chain on the scale. The Laboratory Instructor doubts the result
obtained by the student. Show that the student’s observation is correct.
(The reading of the scale is the force exerted by the scale to stop the
downward motion of the chain.)
Problem 6.14. Many years ago on a cold winter morning in Chicago,
Bonnie and Clyde stole an armored truck full of money. (The mass of
the truck was 2000 kg and its top speed was 240 km/hr or 66.6 m/s.)
Officer Dick Tracy and his driver spotted them and gave chase. (The
police car had a mass of 1500 kg and its top speed was also 240 km/hr.)
As luck would have it, the two vehicles ran off the bank onto Lake
Michigan which was covered with perfectly frictionless ice, so the two
vehicles continued to move at constant speed and maintained a con-
stant separation. Dick Tracy grew impatient, so he opened the moon
roof, stood up, and started to shoot at the armored truck with his
machine gun. The machine gun fired 120 bullets per minute with a
muzzle speed of 1000 m/s. Each bullet had a mass of 0.05 kg. All of
the bullets hit and were embedded into the armored truck. After one
minute of this, what was the speed of the police car and what was the
speed of the truck? Answer: 62.6 m/s, 69.4 m/s.
Problem 6.15. An Atwood’s machine uses two containers filled
with water on either side of the ideal, frictionless pulley. Initially,
both buckets contain the same amount of water and weigh the same.
However, one of the containers has a small hole in it, and water is
leaking out at a rate k (kg/s). The leaking container will, therefore,
move upward. Obtain an expression for the velocity of this container.
Answer: v = −gt+(2mg/k) ln [(2m)/(2m − kt)] , where m is the initial
mass of the buckets plus water.
Problem 6.16. A rocket of mass 40,000 kg is in empty space.
Determine its velocity increase after burning all its fuel if the mass of
the fuel is 90% of the mass of the rocket. The rate of fuel burn is
constant. The speed of the exhaust gas relative to the rocket is 3000
m/s.
Problem 6.17. A rocket of mass 1500 kg is designed to expel
exhaust gas at 1000 m/s. Determine the minimum required burn rate
if the rocket is to rise from the surface of the Earth. What burn rate
is required for it to have an initial acceleration of 1 m/s2 ?
 6.10. PROBLEMS 227

Problem 6.18. A small rocket is launched from the surface of the

Earth. Since it does not rise very high, we are justified in assuming g =
const. Obtain an equation for the height reached when all of the fuel
is burned. The mass of the fuel is m and the initial mass of rocket plus
fuel is M0 . The exhaust speed of the gases is u. Assume the exhaust
speed and the burn rate are constant. Ignore air resistance.
Problem 6.19. Johnny Whizz, the inventor, designs an automobile
that will just hover at the surface of the Earth. It has four rocket
motors, one in each wheel well. The rocket motors expel burned fuel
with an exhaust velocity of 1000 m/s. Assume the mass of the fuel is
80% of the total mass of the car. Evaluate the maximum time the car
can hover above the ground.
Problem 6.20. Rockets are not always launched straight up. Con-
sider a rocket launcher consisting of a ramp inclined at 30◦ above the
horizontal. The mass of the rocket is 4000 kg of which 3000 kg are fuel.
The exhaust speed is 1000 m/s and the fuel is burned at a constant
rate of 200 kg/s. You may assume a flat, airless, nonrotating Earth.
(a) Determine the velocity and direction of the rocket at the time all
the fuel is burned. (b) Determine its position at this time.
Problem 6.21. A billiard ball is placed in contact with the upper
surface of a bowling ball and they are dropped from a height h onto a
cement floor. (Since they fall at the same rate, they are essentially in
contact but you might like to think of the billiard ball as lagging the
bowling ball by an infinitesimal amount.) The bowling ball hits the
ground and bounces back up, immediately striking the billiard ball.
How high does the billiard ball rise? You may assume the bowling
ball has a mass 20 times greater than a billiard ball. All collisions are
elastic.
Problem 6.22. A particle of mass 5 kg and initial speed 5 m/s
undergoes a head-on elastic collision with a particle of mass 3 kg with
initial speed −3 m/s where the negative sign indicates that it is ap-
proaching the first particle. Determine the final speeds of the two
particles.
Problem 6.23. An air hockey puck of mass 2M collides in a glanc-
ing collision with a puck of mass M. The heavier puck had an initial
velocity of 3 m/s and comes off at an angle of 30o after the collision.
Determine the velocity and direction of the lighter puck. Assume an
elastic collision.
Problem 6.24. Consider a glancing collision between two objects
of nearly equal mass (M2 = M1 +δ where δ is a small quantity). Obtain
228 6. CONSERVATION OF LINEAR MOMENTUM

an expression for V10 /V1 and show that it reduces to 1 or cos θ as δ → 0,

as in Subcase 2.1. What is the value of V10 /V1 in the extreme of δ being
a large value (that is, M2 >> M1 )?
Problem 6.25. When a spacecraft carries out a flyby maneuver
near a planet, the speed of the spacecraft can be dramatically increased.
The Cassini spacecraft approached Jupiter at a speed of 9.36 km/sec as
measured relative to the Sun. Jupiter’s orbital speed is 13.02 km/sec.
As it approached the planet, the spacecraft velocity formed an angle of
about 37o to Jupiter’s velocity vector. Cassini was deflected through
an angle of 52.8o . Determine the increase in speed of the spacecraft due
to this slingshot maneuver.
Problem 6.26. At time t = 0 three particles of masses m1 = 1
g, m2 = 2 g and m3 = 3 g are at rest and lined up along the x-axis at
positions x1 = −1 cm, x2 = 0 cm, and x3 = +2 cm. The forces acting
on the particles are F1 =-3̂ dynes, F2 = 0, and F3 =+12̂ dynes. (a)
Determine the position of the center of mass at t = 0. (b) Determine the
positions of the three particles at t = 10 s. (c) Determine the position
of the center of mass at t = 10 s from Equation (5.17). (d) Determine
the position of the center of mass at t = 10 s by using Equation (6.19).
Problem 6.27. Two masses, m1 and m2 , undergo a completely
inelastic collision. Show that the loss of kinetic energy is equal to 12 µv 2
where µ is the reduced mass and v is the relative velocity of the two
masses.
Problem 6.28. Consider a planet-star system with masses MS and
MP . The planet is located at a distance r from the star. As seen from
an inertial coordinate system, the star is located at RS . (a) Determine
the acceleration (in inertial space)of the star and the planet due to
their mutual gravitational attraction. (b) Determine the acceleration
of the planet with respect to the star. (Note that it is not equal to
−(GMS /r2 )r̂ but it reduces to this value if MS >> MP .)
Problem 6.29. A ball is dropped from a height h. The coefficient
of restitution is e. Determine that the time required for the ball to come
to rest is s  
2h 1 + e
t= ,
g 1−e
and that the total distance the ball travels is
1 + e2
 
d=h .
1 − e2
 6.10. PROBLEMS 229

Hint: Express the time as an infinite series. Note that the sum of a
geometric series of the form a, ar, ar2 , ar3 , · · · is S = a/(1 − r).
Problems 6.30 to 6.32 are based on Optional Section 6.8
Problem 6.30. Show that in the center of mass system, the con-
servation of kinetic energy (equation 6.23) reduces to
2 0
P1cm P2
= 1cm
2µ 2µ
if Q = 0.
Problem 6.31. Show that in an elastic collision between two par-
ticles the relative velocity does not change.
Problem 6.32. Show that the scattering angle φ in center of mass
coordinates is related to the angle θ2 in the lab coordinates by
sin φ
tan θ2 = .
1 − cos φ
(In other words, derive Equation 6.29.)
Computational Projects
Computational Techniques: The Runge-Kutta Method
In Section 3.7 I described the Euler Cromer algorithm for solving
ordinary differential equations (ODE’s). Here I will describe the Runge-
Kutta method which is more accurate and perhaps more elegant, but
less transparent. It is based on the truncated Taylor Series expansion
for a function g(t):

dg 
g(t + τ ) = g(t) + τ
dt  ξ

The second term is usually evaluated at t, but to make the solu-

tion more accurate, the Runge-Kutta technique evaluates it half way
through the time step. That is ξ = t + τ /2.
An economical way of expressing the ODE is to define two vectors,
x and f in the following way. Assume a two dimensional system so the
position is given by x and y and the velocity by vx and vy . Then
x(t) = [x(t) y(t) vx (t) vy (t)]
f (x,t) = [vx (t) vy (t) ax (t) ay (t)]
and the ODE can be written
dx
= f (x(t), t).
dt
230 6. CONSERVATION OF LINEAR MOMENTUM

The second-order Runge-Kutta algorithm is obtained by first defining

a new vector x∗ by:
1
x∗ = x∗ (t + τ /2) = x(t) + τ f (x(t), t).
2
Then
x(t + τ ) = τ f (x∗ , t + τ /2).
It is not too difficult to appreciate that this is equivalent to the trun-
cated Taylor Series. However, the most common ODE solver is not the
second-order Runge-Kutta, but the fourth-order Runge-Kutta, which
is also based on the Taylor Series but in a much less transparent way.
In vector form, the value of x at time t + τ is given by
1
x(t + τ ) = x(t) + τ (F1 + 2F2 + 2F3 + F4 ) ,
6
where
F1 = f (x, t),
1 1
F2 = f (x + τ F1 , t + τ ),
2 2
1 1
F3 = f (x + τ F2 , t + τ ),
2 2
1
F4 = f (x + τ F3 , t + τ ).
2
Computational Project 6.1. A man is standing on a stationary
railroad flatcar loaded with large rocks, all having the same mass of 10
kg. He is able to throw the rocks at a speed of 2 m/s, relative to himself.
There are 50 rocks on the flatcar. Determine the speed of the flatcar
after the man has thrown all of them straight back off the end of the
car. The mass of the man plus empty flatcar is 1000 kg. Naturally, this
advanced propulsion system depends on the development of completely
frictionless railroad cars. Solve this problem numerically, and also solve
it using equation 6.2. Compare your answers. Why do they not agree?
Which answer do you think is correct?
Computational Project 6.2. Write a program to determine
the altitude reached by a rocket launched from the surface of Earth.
Assume the rate at which fuel is burned is proportional to the amount
of fuel left and the exhaust speed of the ejected gases is constant. Plot
the position of the rocket as a function of time for dM/dt = −0.01M if
the initial mass of the fuel is 50,000 kg and u = 2500 m/s. The mass of
the empty rocket is 10,000 kg. Ignore air resistance, but keep in mind
that the gravitational force decreases with the distance from the center
of Earth.
 6.10. PROBLEMS 231

Computational Project 6.3. A rocket is launched from the

surface of the Earth. The initial mass of the rocket is 1000 kg, and it
is 90% fuel. The exhaust gases have a speed (relative to the rocket) of
250 m/s, and the burn rate is 50 kg/s. The resistance of the air can
be expressed as a retarding force given by F = −0.5Cd ρAv 2 where the
drag coefficient Cd is 0.35 and the density of the air can be assumed to
be constant and equal to 1.20 kg/m3 . The cross sectional area of the
rocket, A, is 0.8m2 . Do not assume g is constant, but you may set it
equal to 9.8 m/ sec2 at the ground. Determine the altitude reached by
the rocket when all of its fuel is burned. Plot altitude as a function of
time. Explain the shape of the curve.
Computational Project 6.4. Solve Computational Project 6.3
using a realistic profile for air density. (You can obtain tables of air
density as a function of altitude using the United States Standard At-
mosphere or the Smithsonian Meteorological Tables. These can be
found in your library or on the internet.)
Computational Project 6.5. A 5 MeV alpha particle is ap-
proaching a gold nucleus. The impact parameter is 1 Å. You may
assume the gold nucleus is initially at rest. Plot the trajectory of the
two particles and determine the angles at which both are scattered.
Computational Project 6.6. You are required to design a two
stage rocket that will accelerate a 5000 kg payload to Earth’s escape
velocity. Assume that 95% of the mass of the rockets is fuel. Assume
the exhaust velocity of the rocket motors is 2000 m/s. Investigate the
possible ranges of masses for the two stages and determine the con-
figuration that will minimize the take-off weight. Determine why a
single-stage rocket, burning the same amount of fuel, cannot accom-
plish the same objective.
Chapter 7
Conservation of Angular
Momentum

In this chapter we consider the conservation of angular momen-

tum. Although some of the concepts discussed here were mentioned in
Chapter 1, you will now be exposed to them in much greater detail.
We begin by applying the conservation of angular momentum to a
particle, and then generalize to a rigid body rotating about a fixed axis.
To keep the analysis simple, it will be limited to symmetrical bodies
with fixed rotational axes. In Chapters 15 and 16 you will study the
general rotational motion of a rigid body.

7.1. Definition of Angular Momentum

Suppose a particle of mass m is located at position r and is moving
with velocity v. Its linear momentum is p =mv. Its angular momentum
is given by

l ≡ r × p. (7.1)
See Figure 7.1. According to the definition of cross product, the angular
momentum is perpendicular to the plane defined by r and p. By the
right hand rule, the angular momentum of the particle illustrated in
Figure 7.1 points into the page.
The magnitude of the angular momentum is given by
l = mvr sin θ.
From the figure it is easy to appreciate that r sin θ is a constant; call it
b. Then l = mvb.
233
234 7. CONSERVATION OF ANGULAR MOMENTUM

Keep in mind that the angular momentum depends on the choice of

the origin of the coordinate system (because the vector r is part of the
definition). When working with angular momentum, remember that
once you have chosen an origin, you must not change it.

Figure 7.1. Illustrating angular momentum for a par-

ticle. A particle of mass m moving with velocity v is
located at r. The linear momentum of the particle is
p =mv and its angular momentum is l = r × p, a vector
pointing into the page.

An important special case is the angular momentum of a particle

moving in a circular path. If the center of the circle is chosen as the
origin, the angle between r and v is π/2 and
l = mvr sin(π/2) = mvr.
But for circular motion, v = ωr, where ω is the angular velocity. (See
Equation 1.12.) Consequently,
l = mr2 ω

Exercise 7.1. Evaluate the magnitude and direction of the angular

momentum vector for a 1000 kg automobile driving down a straight
road at 100 km/hr, with respect to a point 20 meters to the side of
the road. What is the angular momentum with respect to a point 20
meters on the other side of the road? Answer: l = 5.56 × 105 kg m2 /s.
Directions = down and up.
 7.2. CONSERVATION OF ANGULAR MOMENTUM 235

7.2. Conservation of Angular Momentum

7.2.1. Torque. Before we express the law of conservation of an-
gular momentum, let us recall the definition of torque. Suppose that
an extended rigid body is mounted on a fixed axis and is free to rotate
about that axis. (See Figures 1.5 and 1.6.) Let a force F be applied at
some point. The torque exerted by the force is
N = r × F,
where r is the vector from the origin (usually the axis of rotation) to
the point of application of the force. It is obvious from this definition
that the torque depends on the location of the axis or rotation as well
as the point of application of the force, and, of course, the magnitude
and direction of the force. However, if the net torque is zero then the
torque does not depend on the location of the axis of rotation. This
useful fact is explored in the following worked example.

Worked Example 7.1. Prove that for a body in equilibrium

the net torque about any point is zero.
Solution: Consider the relation between the torque about
some point (call it O) and the torque about some other point (O0 )
displaced by a distance d from O. The total torque about O is
X X
NO = NiO = riO × Fi ,
where riO is the vector from O to the point of application of force
Fi . See Figure 7.2.
Similarly, the torque about O0 is
X X
NO 0 = NiO0 = riO0 × Fi ,
where riO0 is the vector from O0 to the point of application of force
Fi . But as shown in Figure 7.2
riO = d + riO0
so X X X
NiO0 = riO0 × Fi = (riO − d) × Fi (7.2)
or X X X
NiO0 = NiO − d× Fi .
P
For a body in equilibrium, the net force is zero ( Fi = 0). There-
fore, if the torque about a particular point is zero, then the torque
about any other point will also be zero. This is a useful fact. For
236 7. CONSERVATION OF ANGULAR MOMENTUM

a body in equilibrium, the torque on the body with respect to any

point is zero.

Figure 7.2. A laminar body acted upon by various

forces (only one of them is shown). If the sum of the
forces is zero and the torque about point O is zero, then
the torque about any other point O0 is also zero.

7.2.2. Torque and Angular Momentum. In Section 1.5.4, we

noted that the rate of change of the angular momentum is equal to the
applied torque. We now prove this assertion, first for a particle, and
then, in the next section, for a collection of particles.
For a particle, l = r × p and so,
dl d dr dp
= (r × p) = ×p+r× .
dt dt dt dt
The first term is zero because
dr
× p = v × p =m(v × v) =0.
dt
Since dp/dt = F, and since the torque is defined by N = r × F, the
equation for dl/dt reduces to
dl dp
=r× = r × F = N. (7.3)
dt dt
That is, the rate of change of the angular momentum of a particle is
equal to the net torque exerted on it by external forces. If there is no
net external torque acting on a particle, the time rate of change of its
angular momentum is zero. That is,

dl
if N = 0 then dt
= 0 and l = constant.
 7.3. ANGULAR MOMENTUM OF A SYSTEM OF PARTICLES 237

If the torque acting on the particle is zero, its angular momentum

is conserved.

Exercise 7.2. A particle of mass m has an initial velocity v = v0 ı̂.

d2 l
It is subjected to a constant force F = f0 ̂. Evaluate dt 2 at the initial
d2 l
time. Answer: dt2 = v0 f0 k̂.

7.3. Angular Momentum of a System of Particles

Having defined the angular momentum of a single particle, let us
consider the angular momentum of a collection of particles. Since the
angular momentum of the i-th particle is li = ri × pi , the total angular
momentum of N particles is just the vector sum of the angular momenta
of all the particles. That is,
N
X N
X
L= li = (ri × pi ).
i=1 i=1

You will not be surprised to find that the internal torques these
particles exert on each other do not affect the total angular momentum.
That is, just as internal forces do not affect the total linear momentum,
so too, internal torques do not affect the total angular momentum.
Suppose the N particles have masses m1 , m2 , · · · , mN and at some
instant of time, these particles are located at positions r1 , r2 , · · · , rN .
All of the particles exert (internal) forces on each other. The force on
particle i, due to particle j, will be denoted Fij . Particle i may also
(e)
be subjected to an external force (denoted Fi ). Consequently, the
equation of motion for particle i is,
N
X
(e)
mi r̈i = Fi + Fij .
j=1
j6=i

Crossing ri into this equation of motion yields

(e)
X
ri × (mi r̈i ) = ri × Fi + ri × Fij . (7.4)
j6=i
238 7. CONSERVATION OF ANGULAR MOMENTUM

Now note that the rate of change of the angular momentum vector of
particle i is
dli d
= (ri × mi ṙi )
dt dt
= mi (ṙi × ṙi ) + mi (ri × r̈i )
= ri × mi r̈i .
Therefore, the left hand side of Equation 7.4 is just the rate of change
of the angular momentum, so
dli (e)
X
= ri × Fi + ri × Fij .
dt j6=i

There is one such equation for each particle. Adding all these equations
gives !
N N
X dli X  (e)
 X X
= ri × Fi + ri × Fij . (7.5)
i
dt i=1 i j6=i
But
X dli d X d
= li = L,
i
dt dt i dt
where L is the total angular momentum of the system, defined as the
vector sum ofthe angular momenta  of all the particles. Furthermore,
P  (e) (e)
i ri × Fi is the total torque Ntot on the system due to external
forces. Consequently, Equation (7.5) can be written as
dL (e)
XX
= Ntot + ri × Fij .
dt i j6=i

The last term, containing the double summation, is equal to zero,

as we now prove. Consider two particles, i and j. The corresponding
terms in the double sum are:
ri × Fij + rj × Fji .
By Newton’s third law, Fij = −Fji , so we can write this pair of terms
as
ri × Fij − rj × Fij = (ri − rj ) × Fij = rij × Fij ,
where rij is the relative coordinate and points from particle j to par-
ticle i. If Newton’s law is obeyed in the strong form,1 then the force
1ALagrangian analysis of the constancy of angular momentum does not invoke
the strong form of Newton’s third law. See, for example, “A Student’s Guide to
Lagrangians and Hamiltonians” by Patrick Hamill, Cambridge University Press,
2014, pages 33-36.
 7.3. ANGULAR MOMENTUM OF A SYSTEM OF PARTICLES 239

between the particles is directed along rij and so rij and Fij are paral-
lel. Therefore, the cross product rij × Fij is zero, and all the terms in
the double sum cancel out pairwise, leaving
dL (e)
= Ntot . (7.6)
dt
That is, the rate of change of the total angular momentum is equal
to the net external torque on the system. In particular, the law of
conservation of angular momentum can be expressed as follows:
If the net external torque acting on a system of
particles (or an extended body) is zero, the total
angular momentum of the system will remain constant.
This is the law of conservation of angular momentum. It is one of
the basic physical principles governing material objects and, as far as
we know, it is never violated. When you study quantum mechanics
and are introduced to the concepts of the orbital and the spin angular
momentum of electrons, you will begin to appreciate the power of this
conservation law.
Furthermore, the law of conservation of angular momentum is a
very useful tool for solving physics problems. There is a wide variety
of problems in which no external torque acts on a particle or system
of particles. Then the angular momentum is constant and the angular
momentum before some event is equal to the angular momentum after
the event.
For example, suppose a comet is attracted to some star, approaches
it, swings about the star in a hyperbolic path, and then travels back out
to “infinity.” The force exerted by the star on the comet is along the
line joining them, so the torque is zero. Therefore, the initial angular
momentum is equal to the final angular momentum, where “initial”
and “final” can refer to any two points along the trajectory.
A common problem found in introductory physics textbooks has
a running child leap onto a playground merry-go-round. This type of
problem can be solved by setting the total initial angular momentum
(the value for child plus merry-go-round before the child jumps aboard)
equal to the total final angular momentum (the value after the child
has jumped on).

Exercise 7.3. A little girl (mass = 20 kg) runs at 3 m/s and jumps
onto the rim of a playground merry-go-round that consists of a 50 kg
240 7. CONSERVATION OF ANGULAR MOMENTUM

disk with radius 1 m. The frictionless merry-go-round was initially at

rest. What is the final angular velocity of the system (child plus disk)?
Answer: 4/3 rad/sec.

Another type of angular momentum problem involves bodies upon

which a net external torque is acting. Consider, for example, a spool
with thread wrapped around it. A tension in the thread exerts a torque
on the cylinder causing its angular momentum to increase.

Figure 7.3. The Earth-Moon system, looking down on

the North Pole. The Earth rotates counterclockwise and
the Moon orbits the Earth counterclockwise. The Moon
exerts a force on the tidal bulges of the Earth. The force
on the further bulge is somewhat smaller than the force
on the nearer bulge so there is a net torque on the Earth
acting clockwise in the figure, causing the Earth to slow
down. The figure is extremely out of scale. The Earth-
Moon distance is roughly 60 Earth radii, and the bulges
are only a few degrees away from the Moon-Earth line.

An interesting example in which a torque causes a change in angular

momentum is the effect of the Moon on the rotation rate of the Earth.
As you know, the Moon raises tidal bulges in the oceans of the Earth, a
large one on the face of Earth facing the Moon and a smaller one on the
opposite side.2 These liquid bulges are dragged along with the rotating
Earth so they are not directly “under” the Moon, but somewhat ahead
of the position of the Moon by a small amount. The Moon pulls on the
tidal bulges as illustrated in Figure 7.3. The force on the nearer bulge
is greater than the force on the further bulge so there is a net torque
on the Earth. The effect of this torque is to slow down the rotation
2The reason there are two tidal bulges is the subject of Problem 10.6.
 7.3. ANGULAR MOMENTUM OF A SYSTEM OF PARTICLES 241

rate of the Earth. As the Earth rotates more and more slowly, the day
grows longer. In several billion years the day will be so long that it will
be equal to the orbital period of the Moon. That is, the day will be as
long as the month.
If we consider Earth-Moon to be an isolated system, there are no ex-
ternal torques on it, and the total angular momentum must be constant.
We have seen that the angular momentum of the Earth is decreasing,
so how can the total angular momentum remain constant? The answer
is that the angular momentum of the Moon is increasing. This requires
an increase in the distance between the Earth and the Moon. That is,
the Earth is slowing down, losing rotational angular momentum, and
the Moon is receding from the Earth, gaining orbital angular momen-
tum in such a way as to conserve the total angular momentum of the
system.3

Worked Example 7.2. You have invented a certain device

that is to be used in outer space (so gravity can be neglected). It
consists of a small weight of mass m connected to a massless string
of length d and it is spun around in a circle whose center is at a fixed
point. The initial angular velocity is ω0 . The string is pulled in until
it is half of its initial length. (a) Evaluate the change in angular
momentum. (b) Compare the final angular velocity to the initial
angular velocity. (c) Evaluate the change in kinetic energy.(d) How
much work was done in pulling the mass closer to the center?
Solution: (a) The initial angular momentum is l = md2 ω0 .
The force is along the string so the torque = N = r × F =0 and
angular momentum is conserved. Therefore ∆l = 0.
(b) By conservation of angular momentum
lf = li =⇒ md2f ωf = md2 ω0 .
But df = 0.5d, so
d2
ωf = ω0 = 4ω0.
(d/2)2

3The tidal interaction between Earth and Moon has many fascinating aspects.
It has been used to explain the eccentricity of the Moon’s orbit, but recent studies
seem to indicate that the gravitational force of Jupiter may be more important.
If you are interested, you might look up the paper by Matija Cuk, “Excitation of
lunar eccentricity by planetary resonances,” Science, 318, 244 (2007),
242 7. CONSERVATION OF ANGULAR MOMENTUM

(c)The change in kinetic energy is

∆T = Tf − T0 .
Since v = ωr and T = (1/2)mv 2 , the kinetic energy can be ex-
pressed as T = (1/2)m (ωr)2 . Therefore, T0 = 21 md2 ω02 and Tf =
1
2
m(d/2)2 ωf2 . Consequently,
1 1
∆T = m(d/2)2 (4ω0 )2 − md2 ω02
2  2

1 2 2 1 3
= md ω0 (16) − 1 = md2 ω02 = 3T0 .
2 4 2
(d) By the work energy theorem, the work done is 3T0 =
3
2
md2 ω02 .

Exercise 7.4. An Atwood’s machine has two weights tied to the

ends of a string that passes over a pulley of radius 2 cm. The two
masses are 500 gr and 700 gr. Assuming the string does not slip,
determine the rate of change of angular momentum of the pulley.The
moment of inertia of the pulley is 100 gr cm2 . (Note: Since the masses
are accelerating, you cannot assume tension equals mg.) Answer: 8000
dyne cm.
Exercise 7.5. A child’s top of mass M is spinning at an angular
velocity ω. When it slows down, it leans over a bit so that the axis
is at an angle θ with the vertical. The center of mass is a distance R
from the tip. Determine the gravitational torque with respect to the
point of contact with the floor. What is the direction of this torque?
Draw a diagram indicating the direction of the rate of change of angular
momentum. (This is why the top precesses.) Answer: N = RM g sin θ
perpendicular to ω and g.
Exercise 7.6. A dumbbell consists of two point masses M con-
nected by a massless rod of length d. It is rotating about an axis through
its center of mass at a rate ω0 . One of the masses comes loose and flies
off. Determine the angular velocity of the remaining mass.

7.3.1. Angular Momentum Relative to the Center of Mass.

Recall that the kinetic energy of a system of particles can be expressed
as the sum of the kinetic energy of the center of mass plus the kinetic
energy of all the particles with respect to the center of mass. (See
 7.3. ANGULAR MOMENTUM OF A SYSTEM OF PARTICLES 243

Section 5.8.) A similar relation holds for the linear momentum of a

system of particles. (See Section 6.6.) As we now demonstrate, the
angular momentum can also be expressed as the angular momentum
of the center of mass, plus the angular momentum about the center of
mass.
The total angular momentum of a system of particles is
X
L= ri × mi ṙi .
i

The position vector of particle i can be expressed as

ri = rc + r0i ,
where rc is the vector to the center of mass and r0i is the position of
particle i with respect to the center of mass (see Figure 5.13). Using
this relationship, the angular momentum is
X
L = mi (rc + r0i ) × (ṙc + ṙ0i )
i
X X
= mi rc × (ṙc + ṙ0i ) + mi r0i × (ṙc + ṙ0i )
i i
X X X X
= mi rc × ṙc + mi rc × ṙ0i + mi r0i × ṙc + mi r0i × ṙ0i .
i i i i

The quantities rc and ṙc do not depend

P on i and can be taken out of
the summations. Also note that i mi = M (the total mass). Conse-
quently,
! !
X X X
L =M (rc × ṙc ) + rc × mi ṙ0i − ṙc × mi r0i + mi (r0i × ṙ0i ) .
i i i
0
= 0 and i mi ṙ0i =
P P
But the two middle terms are zero because i mi ri
0 by the definition of the center of mass. (The first of these is propor-
tional to the position of the center of mass with respect to itself and
the second is proportional to the velocity of the center of mass with
respect to itself.) The expression for L then reduces to
X
L =M (rc × ṙc ) + mi (r0i × ṙ0i ) ,
i

which can be written

L = Lc + L0 .
Here Lc = M (rc × ṙc ) is the angular momentum of a particle of mass
M locatedPat the center of mass and moving with the center of mass,
and L = i mi r0i × ṙ0i is the sum of the angular momenta of all the
0
244 7. CONSERVATION OF ANGULAR MOMENTUM

particles relative to the center of mass. That is what we set out to

prove.

mi r0i = 0.
P
Exercise 7.7. Prove that i

7.4. Rotation of a Rigid Body about a Fixed Axis

A rigid body is a system of particles in which all the particles are
constrained to remain at constant distances from one another. If the
body is rotating about a fixed axis of rotation, the particles move in
circular paths centered on the axis. These paths lie in planes perpen-
dicular to the axis of rotation. All the particles have the same angular
velocity. The angular velocity ω is a vector directed along the rota-
tion axis whose magnitude is equal to the rotation rate. The sense of
ω is given by the right hand rule, as illustrated in Figure 7.4. The
magnitude of ω is
2π
|ω| = ω = = 2πf,
T
where T is the period of revolution and f is the frequency.

Figure 7.4. The direction of the vector angular veloc-

ity ω is given by the right hand rule. If the fingers of
your right hand curl in the direction of the motion, your
thumb points in the direction of ω.

As noted in Equation (1.23), the linear velocity vi of particle i is

related to the angular velocity ω by
vi = ω × ri⊥ ,
where ri⊥ is a vector in the plane of the circular path from the axis of
rotation to the particle. Since the angle between ω and ri⊥ is 90◦ , the
 7.4. ROTATION OF A RIGID BODY ABOUT A FIXED AXIS 245

magnitude of vi is ri⊥ ω, as expected, and the angular momentum of

2
particle mi is li =mi ri⊥ ω. I wrote the distance from the axis of rotation
as ri⊥ to emphasize that it is the perpendicular distance from the axis
and for future convenience.
Now recall that angular momentum depends on the choice of origin.
If we place the origin at the center of the circular path of some particle,
say mi , then it will not be at the center of the path of another particle,
say mj , that is displaced somewhat along the axis of rotation. We need
an expression for the angular momentum of a particle relative to an
arbitrary origin. However, to keep our analysis simple, let us restrict
the location of the origin to a point on the axis of rotation. This is not
the most general position, but it is sufficient for our purposes at this
time.
Our generalization of the origin to an arbitrary location of the axis
of rotation is illustrated in Figure 7.5.
The angular momentum of particle i can be expressed in terms of
the angular velocity as follows:
li = ri × pi = mi (ri × vi ) = mi ri × (ω × ri ) . (7.7)
Using the “BAC-CAB” rule for the vector triple product,4 this becomes

li =mi ri2 ω − (ri · ω) ri .

 

From Figure 7.5, we appreciate that if the z axis is along ω, then

ri = ri cos θi k̂ + ri sin θi êi
where θi is the angle between ri and ω, and êi is a unit vector in the
plane of the circular path and pointing (instantaneously) towards mi .
Using this expression for ri and doing a little bit of algebra, you can
easily show that

li = mi ri2 ω sin2 θi k̂−mi ri2 ω sin θi cos θi êi . (7.8)

Now I am going to place an important restriction on the analysis.
I will only consider bodies that are symmetrical about the rotation
axis. (When this restriction is lifted in Chapter 16 you will need to
use tensors to describe the motion. I don’t want to talk about tensors
just yet, so let’s assume the rotating body is symmetrical about the
axis of rotation.) In a symmetrical body, for every particle i there is
an identical particle j diametrically opposite to it at the same distance
from the axis of symmetry. Consider the ith particle which has mass
mi and angular momentum li , as shown in Figure 7.6. The jth particle
4The “BAC - CAB” rule is A × (B × C) = B(A · C) − C(A · B).
246 7. CONSERVATION OF ANGULAR MOMENTUM

^
ei
ri vi
mi

ri
i
k

Figure 7.5. The origin is located on the axis of rota-

tion. The vector ri specifies the location of particle mi
relative to the origin. ri is resolved into a component
along the rotation axis (ri cos θi k̂), and a component per-
pendicular to the rotation axis (ri sin θi êi ). The unit vec-
tor êi points towards the instantaneous position of the
particle.

is identical, except for the direction of its angular momentum, lj . The

components of li and lj parallel to the rotation axis add together, but
the components perpendicular to the rotation axis cancel. So adding
equations of the form (7.8) to get the total angular momentum of a
symmetrical body yields
X
L= mi ri2 ω sin2 θi k̂.
i

The quantity ri sin θi is the perpendicular distance from the axis of

rotation to the ith particle. If this distance is denoted by ri⊥ we can
write !
X
2
L= mi ri⊥ ω k̂. (7.9)
i

Exercise 7.8. Fill in the missing steps between Equations (7.7)

and (7.8).
Exercise 7.9. A spinning disk of radius 10 cm and mass 15 gr is
rotating at 50 rad/s. A sticky piece of chewing gum (mass = 3 gr)
falls on the rim of the disk. Determine the final angular velocity of the
system. Answer: 35.7 rad/sec.
 7.5. THE MOMENT OF INERTIA 247

Figure 7.6. The angular momentum of a particle (rela-

tive to the given origin) has components in the direction
of the axis of rotation and perpendicular to it, as shown
in the panel on the left. The angular momentum of a
symmetric particle has an equal but opposite component
perpendicular to the rotation axis.

7.5. The Moment of Inertia

2
P
The quantity i mi ri⊥ in Equation (7.9) depends only on the mass
of the body and how that mass is distributed with respect to the axis
of rotation. For any given body rotating about some specific axis, this
quantity is a characteristic constant. We call it the moment of inertia
of the body5 and denote it by the letter I.
X
2
I= mi ri⊥ . (7.10)
i

We can generalize Equation (7.10) to a continuous body of density ρ,

and write Z Z Z Z Z Z
2 2
I= r⊥ dm = ρr⊥ dτ. (7.11)
body body

5Although you studied the moment of inertia in your introductory mechanics

course, you will find out that this topic is more complicated than you might expect.
In discussing the general motion of a rigid body we need to introduce a quantity
called the inertia tensor. For the moment, however, we will restrict ourselves to
the simple situation of a symmetrical rigid body rotating about a fixed axis and
the scalar quantity I is all we need.
248 7. CONSERVATION OF ANGULAR MOMENTUM

where dm = ρdτ is a mass element of the body and r⊥ is the perpen-

dicular distance from mass element dm to the axis of rotation.
Using the definition for moment of inertia (Equation 7.10), we can
write the angular momentum (Equation 7.9) in the form
L =Iω.
Keep in mind that this relationship is based on the body being sym-
metrical about the rotation axis. Only then will the angular momentum
be parallel to the angular velocity. If the body is not symmetric about
the axis of rotation, the angular momentum vector will not be parallel
to the angular velocity vector. (Such a body will tend to “wobble”
about the axis.)
We now return to the relationship between the torque and the rate
of change of angular momentum. If a torque N acts on a body there
will be a change in its angular momentum given by
dL
N= .
dt
Replacing L by Iω,
d
N = (Iω) .
dt
For a body with constant moment of inertia, this becomes
dω
N=I
dt
or
N = Iα. (7.12)
This has the same form as F = ma with I replacing m, α replacing a,
and N replacing F.
The conservation of angular momentum for a rigid body when no
external torques are acting on it is just Li = Lf . For a symmetrical
body this can be written as
Ii ωi =If ωf .
Observe the analogy between the expression for linear momentum,
p =mv, and the expression for angular momentum, L =Iω.
7.5.1. Two Theorems. Two useful theorems for calculating the
moment of inertia are the parallel axis theorem and the perpendicular
axis theorem.
The parallel axis theorem states that the moment of inertia of a
body about a given axis is equal to the moment of inertia about a
parallel axis through the center of mass plus M d2 where M is the mass
of the body and d is the distance between the axes. The parallel axis
 7.5. THE MOMENT OF INERTIA 249

theorem is a very valuable labor saving device. If we write Ik for the

moment of inertia about the parallel axis and Ic for the moment of
inertia about an axis that passes through the center of mass, we have
Ik = Ic + M d2 .
The perpendicular axis theorem can be applied to plane laminar
bodies. Consider two mutually perpendicular axes in the plane of the
lamina. Let the moments of inertia about these two axes be I1 and I2 .
Then the moment of inertia about an axis perpendicular to the lamina
and passing through the point of intersection of the first two axes, is
I3 = I1 + I2 .

Figure 7.7. A rod of length l with axis passing through

the center of mass. The mass element dm = λdx is a
distance x from the axis.

Worked Example 7.3. (a) Determine the moment of inertia

of a cylindrical rod of mass M and length l about an axis passing
through the center of mass of the rod (line AA), as illustrated in
figure 7.7. (b) Determine the moment of inertia about line BB at
the end of the rod.
Solution: (a) The mass per unit length (linear mass density)
is λ = M/l. The moment of inertia about the axis (AA) is
Z Z Z Z Z l/2  3 l/2
2 2 2 x
I = r⊥ dm = r⊥ λdx = λ x dx = λ
−l/2 3 −l/2
λ 3  1
= 2l /8 = M l2 .
3 12
250 7. CONSERVATION OF ANGULAR MOMENTUM

(b) The moment of inertia about BB can be obtained using the

parallel axis theorem
Ik = Ic + M d2
 2
1 l 1
= M l2 + M = M l2 .
12 2 3

Exercise 7.10. Determine the moment of inertia of a rod of length

l and mass M about an axis perpendicular to the axis of the rod and
a distance a beyond one end of the rod. Answer: M a2 + M al + 13 M l2 .
Exercise 7.11. A force of 2 N is applied tangentially on the rim
of a disk of mass 15 kg and radius 20 cm. The disk is mounted on a
frictionless fixed axis perpendicular to the plane of the disk and passing
through its center. Determine the angular acceleration of the disk.
Answer: 4/3 rad/sec2 .
Exercise 7.12. The moment of inertia of a uniform square plate
of side a about a perpendicular axis through its center is (1/6)M a2 .
Determine the moment of inertia about an axis in the plane of the
plate. Assume this axis passes through the center of the plate. Answer:
(1/12)M a2 .
RRR 2
Exercise 7.13. From the definition I = r⊥ dm, obtain the
moment of inertia for a ring or hoop of mass M and radius R, relative
to an axis through its center and perpendicular to the plane of the ring.
Answer: M R2 .
Exercise 7.14. A meter stick of mass 0.25 kg is mounted horizon-
tally on a perpendicular axis at the 20 cm mark. A downward force of
3 N is applied at the 80 cm mark. Determine the angular acceleration
of the ruler. Answer: 44.12 rad/s2 .

7.6. The Gyroscope

Figure 7.8 illustrates a simple gyroscope. For convenience let us
place the origin of coordinates at the base (point Q in the figure). The
gyroscope is symmetrical, so the angular momentum vector and the
angular velocity vector are parallel and point along the axis of rotation.
The gravitational force acts at the center of mass, which is also on the
 7.6. THE GYROSCOPE 251

axis of rotation, a distance R from Q. The torque about Q due to this

force is
NQ = R×M g.
dL
Since dt = N we have
dL
= R×M g =RM g sin θφ̂.
dt
The torque points in a direction perpendicular to both R and M g (as
well as L). In spherical coordinates this torque points in the direction
of the unit vector φ̂. Hence, N = N φ̂. Further, note that L =Iω and
I is a constant, so
dL dω dω
N= =I = I φ̂.
dt dt dt
This tells us that the angular momentum vector (as well as the angular
velocity vector) is changing. It is not changing in magnitude because
the direction of N is perpendicular to the direction of the angular
momentum. Consequently, the angular momentum vector must be
changing in direction, as indicated in the rightmost sketch in the figure.

Figure 7.8. A gyroscope. The middle sketch shows

the vectors involved. Note that the torque (R×M g) is
perpendicular to both R and M g. The sketch to the right
illustrates that the tip of the angular momentum vector
traces out a circle.
The gyroscope precesses, that is, the axis of rotation changes in
direction in such a way that the tip of the angular momentum vector
traces out a circle, as indicated in the figure. The precession rate, Ωp ,
is the speed with which the tip of the angular momentum vector goes
around this circle. That is,
dφ
Ωp = .
dt
252 7. CONSERVATION OF ANGULAR MOMENTUM

Figure 7.9 gives three different views of the precession of the angu-
lar momentum vector around the vertical. The leftmost figure shows
the angular momentum vector and the circle traced out by the tip of
this vector. Note that as time passes, the angular momentum goes
from L to L + ∆L. The magnitude of the angular momentum does not
change, only its direction, so ∆L lies in the plane of the circle (and
is perpendicular to both L and L + ∆L). The middle figure is a side
view, indicating that θ is the angle between L and the vertical, and
that the horizontal component of L is L sin θ. The rightmost figure is a
top view, looking down on the circular path, indicating that ∆L is the
chord of the arc subtended by the angle ∆φ. Setting the chord equal
to the arc we have
∆L = L sin θ∆φ.
Divide by ∆t and let ∆t → dt, to obtain
dL dφ
= L sin θ .
dt dt
dL dφ
Now recall that dt
= RM g sin θ, and dt
= Ωp so
RM g sin θ = L sin θΩp ,
and consequently
RM g RM g
Ωp = = .
L Iω
It is interesting to note that the precession rate increases as the top
slows down.
Since the angular momentum vector and the angular velocity vector
(ω) lie along the same line, they vary together. The angular velocity
vector is precessing at the same rate as the angular momentum vector.
A gyroscope also “nods” as it precesses. This is called nutation.
We shall consider it in Chapter 16.

Exercise 7.15. A gyroscope is made of a very light axle and a

heavy ring. The ring (of mass M and radius a) is connected to the
axle by nearly massless spokes. The ring is a distance d from the point
of the axis in contact with the floor. Determine the precession rate of
this gyroscope. (Answer: Ωp = gd/a2 ω)
 7.7. ANGULAR MOMENTUM IS AN AXIAL VECTOR 253

z z
Lsinθ
ΔL ΔL
ΔΦ
L L+ΔL
L Lsinθ
θ

Precessing Angular Side View Top View

Momentum Vector

Figure 7.9. Left figure: The tip of the precessing angu-

lar momentum vector traces out a circle. Middle figure:
Side view showing the angle θ between the angular mo-
mentum vector and the vertical. The horizontal compo-
nent of angular momentum is L sin θ. Right figure: Top
view showing that ∆L = L sin θ∆φ.

7.7. Angular Momentum is an Axial Vector

Angular momentum is a vector, but it is a somewhat different type
of vector than familiar “ordinary” vectors such as the velocity vector
and the position vector. The difference between the angular momentum
vector and an ordinary vector is related to its behavior under a coordi-
nate transformation. But before describing this difference, I have to tell
you that there are two kinds of coordinate transformations. They are
called “proper” transformations and (of course) “improper” transfor-
mations, depending on whether or not the coordinate axes retain their
cyclic order. We usually draw a Cartesian coordinate system such that
ı̂ × ̂ = k̂ Such a coordinate system is called “right handed.” Using
your right hand, if you point your fingers along the x-axis and bend
them into the y-axis, your thumb will be pointing in the direction of
the z-axis. If you rotate a right handed coordinate system, you still get
a right handed coordinate system, even though the axes will now be
pointing in different directions. However, if you reflect a right handed
coordinate system, you end up with a left handed coordinate system.
This is illustrated in Figure 7.10 where the right handed coordinate
system reflected in the xz-plane is transformed into a left handed coor-
dinate system. This is, of course, an example of an improper coordinate
transformation.
Consider the behavior of a vector under a transformation of the
coordinate system. To begin, consider the behavior of the position
vector r. The components of r may or may not change sign during
a coordinate transformation. That detail is not important. However,
254 7. CONSERVATION OF ANGULAR MOMENTUM

Figure 7.10. The reflection of a Cartesian coordinate

system. The reflection is in the xz-plane. Note that the
axis perpendicular to the reflecting plane changes direc-
tion.

what is important is whether or not the components of other vectors

have the same sign changes as the components of r under the same
transformation.
We call r a “polar” or “true” vector. Feynman6 called it an “honest”
vector. Many other vectors, such as velocity, momentum and force, are
polar vectors. They behave the same way as r under a coordinate
transformation such as a rotation or a reflection.
For example, if you reflect the vector r it changes direction. (More
precisely, the component of r perpendicular to the plane of reflection,
changes sign.) See Figure 7.11. Under a similar reflection, velocity and
force behave the same way.
There are other vectors, particularly those related to rotations, that
do not transform like r. These are called “axial” or “pseudo” vectors.
Recall that angular velocity ω was defined to be a vector lying along
the axis of rotation. The sense or direction of ω was given by the right
hand rule (see Figure 7.3). If you observe a rotating wheel in a mirror,
you see that the angular velocity vector does not change direction as
shown in Figure 7.12. Angular momentum is another axial vector. The
component perpendicular to the plane of reflection does not change
direction. Therefore, angular velocity and angular momentum do not
transform like the “honest” vector r. This means that ω and L are
axial vectors. Axial vectors behave differently than r under improper
transformations.
6Richard
P. Feynman, The Feynman Lectures on Physics, Addison-Wesley,
Reading, Mass., 1963.
 7.8. SUMMARY 255

Figure 7.11. The reflection of the position vector r in a mirror.

You will study the transformation properties of vectors in much

more detail in a course in mathematical methods. At this stage, it is
sufficient for you to realize that there are different kinds of vectors and
that they are distinguished by their transformation properties.

Figure 7.12. The angular velocity vector does not

change directions under reflection.

7.8. Summary
The angular momentum of a particle is defined as
l=r×p
and, consequently, it depends on the choice of origin. Since
dl d
= (r × p) = r × F = N
dt dt
256 7. CONSERVATION OF ANGULAR MOMENTUM

we appreciate that if N = 0 then l = constant. This relation summa-

rizes the law of conservation of angular momentum.
The angular momentum of a system of particles (or an extended
body) is
L =Σli
and the law of conservation of angular momentum can be written
dL (e)
= Ntot
dt
(e)
where Ntot is the net torque due to all external forces. Internal forces
(assuming they obey the third law in the strong form) do not contribute
to the rate of change of angular momentum.
The angular momentum of a system of particles can be written
as the sum of the angular momentum of the center of mass and the
angular momentum relative to the center of mass, that is,
L = Lc +L0 ,
where
Lc = M (rc ×ṙc ),
and
L0 = Σmi (r0i ×ṙ0i ).
The angular momentum of a symmetrical body rotating about an axis
of symmetry is
 
2
L = Σmi ri⊥ ω.
i
The quantity in parenthesis is the moment of inertia of the body and
is denoted I. Therefore,
L =Iω,
and we can write
(e) dL d d
Ntot = = Iω = Iω = Iα.
dt dt dt
(e)
If Ntot = 0, the angular momentum is constant and consequently
Lf = Li =⇒ If ωf = Ii ωi .
The motion of a gyroscope illustrates the concepts presented in
this chapter. The torque due to gravity causes the angular momentum
vector to precess at a rate
RM g
Ωp = .
Iω
 7.9. PROBLEMS 257

7.9. Problems
Problem 7.1. Two particles of equal charge and equal mass travel
at the same speed v but in opposite directions. Their initial trajecto-
ries are straight lines separated by a distance b. Due to their mutual
repulsion, they will reach a point of closest approach and then move
away from each other. The electrical force between the charges is given
by F = (q 2 /4π0 r2 )r̂. The vector rr̂ is drawn from one particle to the
other. Although it is unrealistic to do so, ignore the magnetic force
between the moving charges. (a) Draw the trajectories of the two
particles, indicating the electric forces acting on them. (b) Explain
whether or not the angular momentum of the system is constant. (c)
Do your answers depend on your choice of coordinate origin?
Problem 7.2. A uniform rod of length d and mass M is moving
at a constant velocity v in the direction of its length. (a) Obtain its
angular momentum relative to an arbitrary origin and show that this
is the same as the angular momentum of a particle of mass M moving
along the trajectory of the rod. (b) If the rod is inclined at an angle α
to its velocity vector, what is the position of the equivalent particle?
Problem 7.3. Prove that the total angular momentum of a system
of two particles is independent of the displacement of the origin only if
the total linear momentum is zero or if the displacement of the origin
is parallel to the total linear momentum.
Problem 7.4. In proving the law of conservation of angular mo-
mentum for an extended body (or system of particles) we assumed
that Newton’s third law is obeyed in the strong form. In this problem
you will investigate this assumption for the case of particles interacting
through magnetic forces. (The magnetic force between moving charged
particles is not directed along the line joining the particles.) The or-
bital motion of an electron around the nucleus is equivalent to a small
current loop, which we usually call a “magnetic moment.” A magnetic
moment generates a magnetic field. Another electron orbiting around
another nucleus will feel a force due to this magnetic field. Let us gen-
eralize the problem to two current loops of arbitrary shape that carry
currents I1 and I2 . Considering small elements of the current loops,
I1 dl1 and I2 dl2 , the force on I1 dl1 due to I2 dl2 is
dF12 = I1 dl1 × B
where
I2 dl2 × r12
I
µ0
B= .
4π |r12 |3
2
258 7. CONSERVATION OF ANGULAR MOMENTUM

Show that the force between the two current loops obeys Newton’s third
law in the strong form. (Hint: this involves showing that dl1 · r12 / |r12 |3
is an exact differential.)
Problem 7.5. An electron is fired horizontally with speed voy be-
tween parallel charged electrical plates. Assume the plates are oriented
horizontally and that the top plate is negatively charged and the bot-
tom plate is positively charged, so the electron feels a uniform force
downward. Denote this force by Fe . Using Cartesian coordinates, cal-
culate the angular momentum as a function of time and show that the
time rate of change of angular momentum is equal to the torque exerted
on the electron by the electric field. Evaluate the angular momentum
(and torque) relative to an origin located at the position of the elec-
tron when it first enters the electric field. See Figure 7.13. Answer:
l = −(1/2)v0y Fe t2 ı̂.

Figure 7.13. An electron between charged plates.

Problem 7.6. In analyzing the collision of two particles we stated

that the conservation of linear momentum (p =constant) led to three
equations describing the motion. Since there are no external torques
acting on the system, it might seem that conservation of angular mo-
mentum (l =constant) would lead to three additional equations. Show
that this is not true, that is, that conservation of angular momentum
does not lead to any new relations.
Problem 7.7. A roll of rather heavy gift wrapping paper is placed
on a frictionless roller and suspended by a bracket of length b as shown
in Figure 7.13. The roll is in contact with the wall. The coefficient of
static friction between wall and paper is µ. A portion d of the paper
is hanging down. Assume the total length of paper on the roll is D
and that it has a linear mass density λ. Determine the length d such
that the paper will unroll on its own. At the instant the roll begins to
unwind on its own, it has a radius R.
 7.9. PROBLEMS 259

Figure 7.14. A roll of paper that will unwind under its

own weight.

Problem 7.8. Professor Ptolemy of the Astronomy Department

claims that there is a planet with the same mass as Earth that is
directly on the other side of the Sun in an orbit of the same radius as
Earth’s. This planet, he says, has the same mass as Earth, but it is
cylindrical in shape. The axis of the cylinder lies in the orbit plane
and its length is one Earth diameter. Determine the torque the Sun
exerts on this planet, as a function of the angle between the axis of
the cylinder and a line from the planet to the Sun. You may assume
the angle between the axis of the cylinder and the line to the Sun is
small. (Hint: Determine the gravitational force on the planet, then
find the center of gravity and finally calculate the torque. By the way,
if there were a planet on the other side of the Sun from us, we would
be able to see it at times, due to the eccentricity of Earth’s orbit. But
in this problem you may assume circular orbits. Hint #2: The center
of gravity is the point in the body where the gravitational force acts.
In this problem it is not at the same point as the center of mass.)
Problem 7.9. A neutron star whose mass is 1.5 times the mass
of the Sun has collapsed to a sphere of radius 10 km. If the original
angular velocity was one revolution per month (before the collapse),
what is its final angular velocity? Assume the initial density of the
star was equal to the mean density of the Sun.
Problem 7.10. Prove the parallel axis theorem.
Problem 7.11. Prove the perpendicular axis theorem.
Problem 7.12. (a) Derive an expression for the moment of inertia
of a disk of mass M and radius R about an axis perpendicular to its
plane and passing through its center of mass. (b) Obtain an expression
260 7. CONSERVATION OF ANGULAR MOMENTUM

for the moment of inertia about a parallel axis that is tangent to the
edge of the disk.
Problem 7.13. Determine the moment of inertia of an annular
cylinder (or ring) about an axis perpendicular to the plane of the ring
and passing through its center. The ring has uniform density, mass M,
inner radius R1 and outer radius R2 .
Problem 7.14. Derive an expression for the moment of inertia of
a uniform, solid sphere about a diameter.
Problem 7.15. Determine the moment of inertia of a flat disk of
mass M and radius R about an axis tangent to the edge of the disk
and lying in the plane of the disk.
Problem 7.16. Assume the Earth is a sphere of mass M and radius
R. Remove the Southern Hemisphere. By direct integration determine
the moment of inertia of the remaining hemisphere about an axis per-
pendicular to the equatorial plane and passing through the North Pole.
Problem 7.17. A disk of mass M and moment of inertia I has a
hole in its center. An axle, which is slightly smaller in diameter than
the hole, passes through the hole. The axle is horizontal and the plane
of the disk is vertical. The disk spins smoothly on the axle. There is
friction between the disk and the axle, and the coefficient of friction
is µ. Assume the disk has an initial angular speed ω0 . Determine the
number of turns and the time for the disk to stop. (Hint: the point of
contact is not over the center of mass.)
Problem 7.18. A dumbbell is made up of a rod of length l and
mass m connected on either end to spheres of mass M and radius
R. (a) Determine the moment of inertia of the dumbbell about an axis
perpendicular to the rod and going through its center. (b) You suspend
the dumbbell from its center of mass and apply a force of 1.5 N at the
center of mass of one of the spheres. The force is in the horizontal plane
and perpendicular to the rod. Determine the angular acceleration of
the system. Assume M = 1 kg, R = 10 cm, l = 20 cm and m = 0.1
kg. Answer: (b) 0.1 rad/s2 .
Problem 7.19. The gyroscope of Figure 7.6 is made up of a disk
of mass M and radius a and an axis of negligible mass that is free to
swivel in any manner about the bottom end. It is inclined at an angle
θ = 90◦ so the axis is horizontal and the gyroscope is precessing. The
disk spins about the axis with angular speed ω. Write an expression
for the total angular velocity vector in Cartesian coordinates in terms
of M, a, R, ω, the time, and any other appropriate parameters.
 7.9. PROBLEMS 261

Problem 7.20. A gyroscope that is free to move about the support

point is in the upright position and is spinning at an angular velocity ω k̂
(see Figure 7.6). Assume the axis has a length d and negligible mass.
The disk of mass M and radius a is mounted at d/2. A frictionless ball
bearing device allows you to grasp the top of the axis and pull it to a
horizontal orientation. After pulling the gyroscope down, you hold it
stationary with the axis horizontal. (a) How much work did you do to
rotate the gyroscope axis? (b) What is the final angular velocity of the
gyroscope? (c) What force do you need to exert on the tip of the axis
to keep it stationary? You may assume that ω is large so Ωp is small
enough that the energy associated with precession can be neglected.

Figure 7.15. The speed of the particle m does not in-

crease as the string winds up on the cylindrical rod.

Problem 7.21. A string is partially wound around a cylindrical

rod of radius a, as shown in Figure 7.15. A particle of mass m is tied
to the end of the string. The mass m is given an initial velocity v0 and
the string winds itself onto the rod. Make the unrealistic assumption
that there is no gravitational force. Show that the the linear velocity
of the mass is constant. Do this is in the following two ways: (a)
Using the conservation of energy, and (b) Using the fact that the rate
of change of angular momentum is equal to the torque. (Hint: there is
a torque on the particle because the rod does not have zero radius and
the instantaneous velocity of the particle is perpendicular to the string
at all times.}
Problem 7.22. A pendulum consists of a bob of mass M and a
massless string of length l. The pendulum is initially at rest. It is
struck by a bullet of mass m and velocity v. Let m = 0.1M. Assume
the pendulum hangs vertically and the bullet travels horizontally. (a)
Determine the initial angular velocity of the pendulum if the bullet be-
comes embedded in the bob. (b) Determine the initial angular velocity
of the pendulum bob if the bullet bounces back off the bob elastically.
262 7. CONSERVATION OF ANGULAR MOMENTUM

Problem 7.23. Generalize the derivation of dL dt

= N to the case of
a moving origin. If the origin (Q) is at rQ , and LQ and NQ are the
total angular momentum and torque with respect to Q show that
dLQ
= NQ
dt
as long as the point Q has zero acceleration or is accelerating along
the line from Q to the center of mass. You may assume the internal
forces obey the third law in the strong form so the total internal torque
vanishes.
Problem 7.24. Consider a symmetrical rigid body to be composed
of N particles of masses mi (i = 1, N ). Assume the body is rotating
about a fixed axis with angular speed ω. Prove that the kinetic energy
is 21 Iω 2 where I is the moment of inertia.
Problem 7.25. A particle moves in the field of an attractive central
force whose potential is given by V = −k/r. Show that the angular
momentum does not change.

Figure 7.16. Two disks collide and stick together. See

Problem 7.26.

Problem 7.26. Figure 7.16 is a view from above, showing two

identical disks of mass M and radius R on a frictionless surface. One
disk is at rest, the other is rotating counterclockwise with angular ve-
locity ω and is moving with a linear velocity v = 21 ωR. It makes a
grazing collision with the second disk at point P. After the collision
the two disks stick together. Determine the final angular momentum
of the system with respect to P.
Computational Projects
Computational Project 7.1. A child’s playground merry-go-
round has been fitted with frictionless bearings. If air resistance can be
 7.9. PROBLEMS 263

ignored, this “ideal merry-go-round” will rotate indefinitely. Suppose

that it is spinning at 3 rad/sec when it starts to snow. The rate at
which snow accumulates on the disk is 1 gram/second. The disk has a
radius of 1.5 m and a mass of 40 kg. Determine the angular speed of
the merry-go-round after 30 minutes of steady snowfall. Plot ω vs t.
Chapter 8
Conservation Laws and Symmetries

8.1. Symmetry
Nature exhibits many symmetries, from the the structure of atoms
to the patterns of snowflakes. As physicists, we exploit the concept
of symmetry to understand the nature of physical reality. On a more
mundane level, we find symmetry concepts to be a valuable aid in
solving physics problems.
All of us have an intuitive concept of what is meant by symmetry.
We recognize spatial symmetry in flowers and snowflakes and in the
bodies of most living creatures. We say these things “look the same”
when viewed from different points of view. For example, a snowflake
looks the same when rotated through 60 degrees (known as hexagonal
symmetry). A human body looks the same when viewed in a mirror.
A cylindrical vase looks the same when rotated through an arbitrary
angle. Although its mirror image may be indistinguishable from the
vase itself, it will probably look different when it is turned upside down.
A physicist would say the vase has rotational symmetry and symmetry
under reflection, but it does not have “right side up - upside down”
(axial) symmetry.
From the examples above, you can appreciate that a symmetry is
related to an invariance under some particular operation. In simpler
words, a symmetry means that a system is not changed when something
is done to it. If there is no way we can tell whether or not a vase has
been rotated through some angle, then it has rotational symmetry. If a
pendulum oscillates in exactly the same way on one side of the room as
on the other, then it has translational symmetry. A mechanical system
that does not change in time has temporal symmetry.

265
266 8. CONSERVATION LAWS AND SYMMETRIES

Among the simpler symmetry operations are spatial translations,

rotations, reflections, and translations in time. Other symmetry oper-
ations are not as familiar to you. For example, “charge conjugation”
symmetry refers to interchanging positive and negative charges. Does
an “antihydrogen atom” with a negative proton and a positive electron
behave the same as an ordinary hydrogen atom? If so, it has charge con-
jugation symmetry. Another way of expressing invariance under charge
conjugation is to call it matter-antimatter symmetry. This means that
matter and antimatter particles are attracted to one another by the
gravitational and electromagnetic forces and appear to be very much
the same except for the fact that when they come into contact they
annihilate one another!
An interesting symmetry is called “chirality”1 This is the symme-
try between your left hand and the mirror image of your right hand.
There is no combination of translations or rotations that will transform
one hand into the image of the other, but a mirror reflection does al-
low such a transformation. This is important in chemistry since some
molecules are chiral, that is, they exist in both “left hand” and “right
hand” forms, which may have different properties, such as their effect
on polarized light. It is worth noting that proteins in sugars contain
the “right-handed” form of the the alanine molecule. This molecule
also comes in a mirror imaged or “left-handed” form. Biological sys-
tems (such as you and me) can only digest the right-handed alanine as
a result of some evolutionary quirk. However, it is possible to manu-
facture “left-handed” sugar that tastes the same as “ordinary” sugar,
but it is not digested. This product is useful for diabetics and people
who are avoiding sugar for medical reasons. It has the drawback of
being expensive to produce.
For a simple example, consider a translation operation. Suppose
a particle is lying on a smooth, infinite, featureless table. If we slide
the particle to some other point on the table, nothing has changed.
The system has symmetry under horizontal translations. On the other
hand, if we move the particle vertically, its potential energy changes.
The state of the system is different. The system exhibits symmetry
for translations along the two horizontal axes but not for translations
along the vertical axis.
If a system that is moved forward in time behaves the same as
before, we say it is symmetric with respect to a translation in time.
(Invariance with respect to time is an important property of a clock.)

1The
name comes from the Greek word for “hand.” The word “chiral” is
pronounced “kai-ral” and rhymes with “spiral.”
 8.2. SYMMETRY AND THE LAWS OF PHYSICS 267

There is an even more complex symmetry operation involving time

called “time reversal.” Time reversal is analogous to spatial reflection.
Imagine a box full of Mexican jumping beans. Suppose you videotape
the beans while they are jumping wildly all over the place inside the
box. If you show the tape to your family, they will not be able to
tell whether it is running forward or backward. A box full of Mexican
jumping beans is symmetric under time reversal. But if you videotape
your friend jumping off a diving board into a pool, you will certainly
know if the tape is played in reverse.

8.2. Symmetry and the Laws of Physics

The laws of physics are a set of relations between physical quan-
tities (usually expressed in mathematical form) that describe how the
universe behaves. For example, F = dp dt
and the conservation of energy
are laws of physics.
Do the laws of physics exhibit symmetry? That is, do the laws
of physics remain constant under symmetry operations? Suppose I use
some equipment and verify that for a constant mass system the relation
F = ma is true in my laboratory. If I perform a spatial translation by
taking my equipment to your laboratory, will F = ma still be true? I
am sure you will agree that Newton’s second law is invariant under a
space translation. But is F = ma invariant under a reflection? If you
look at a physical system in a mirror, will Newton’s law be F = ma or
will it be F = −ma? If you recall that both F and a are polar (“honest”)
vectors, you can easily convince yourself that in the reflected system the
law is still F = ma. Would F = ma hold in a time reversed system?
(The answer is yes.) You see that this question of the symmetry of
physical laws is rapidly becoming complicated!
Perhaps the most surprising thing about symmetry and the laws of
physics is the relationship between symmetries and conservation laws.
This relationship is expressed in Noether’s theorem.2 The theorem tells
us that:
For every symmetry there is a
corresponding constant of the motion.
In the next section we shall explore this relationship using La-
grange’s equations.

2Emmy Noether (1882-1932) was one of the most important mathematical

physicists of the 20th century. She proved the theorem bearing her name in 1918.
268 8. CONSERVATION LAWS AND SYMMETRIES

8.3. Symmetries and Conserved Physical Quantities

Consider a physical system. If it is symmetrical with respect to a
rotation, then it is unchanged by the rotation. The system is the same
after it was rotated as it was before.
A physical system can be described in many ways; for example, you
could generate a table giving all of the physical properties of the system
(velocity, position, etc.) as functions of time. However, a much better
and simpler way to describe a physical system is to give its Lagrangian.
If you can write down the Lagrangian (and a set of initial conditions)
you know all the essential mechanical properties of a physical system.
You can use Lagrange’s equations to determine the equations of motion
and you can solve them to evaluate how the system will evolve in time.
Suppose the Lagrangian of some system does not contain a particu-
lar coordinate. Specifically, suppose that the angle φ does not appear in
the Lagrangian. Remember that we call such a coordinate “ignorable”
and as we showed in Chapter 4, the generalized momentum conjugate
to an ignorable coordinate is constant. The generalized momentum
conjugate to the angle φ is the angular momentum, so for this system,
the angular momentum is constant.
That is, if a system is symmetrical with respect to a rotation, then
the Lagrangian of the system does not depend in any way on φ and φ
will not appear in the Lagrangian. Consequently, the angular momen-
tum associated with rotations through φ will be conserved. Similarly,
if a system is symmetrical with respect to a translation, then its linear
momentum will be conserved.
Let us consider these concepts in more detail.

8.3.1. Conservation of Linear Momentum. Imagine a particle

in a homogeneous region of space. By spatial homogeneity we mean
that the space is the same at all points. Obviously, a particle in this
space will exhibit translational symmetry. We now show analytically
that its momentum must remain constant.
The Lagrangian for the particle cannot depend on position, so in-
stead of L = L(q, q̇, t) the Lagrangian is just L = L(q̇, t). (I am consid-
ering a single generalized coordinate for simplicity.) Lagrange’s equa-
tion is
d ∂L ∂L
− = 0.
dt ∂ q̇ ∂q
Since L does not depend on q, this relation reduces to
d ∂L
= 0.
dt ∂ q̇
 8.3. SYMMETRIES AND CONSERVED PHYSICAL QUANTITIES 269

Therefore,
∂L
= constant.
∂ q̇
But recall that
∂L
= p = generalized momentum.
∂ q̇
Therefore, p = constant and the proposition is proved.

Exercise 8.1. Show that if a system is symmetric under rotations

through some angle φ then the angular momentum conjugate to φ is
constant.

8.3.2. Conservation of Energy. Perhaps the most surprising

of the symmetry/conservation relationships involves temporal symme-
tries. A system that is symmetrical with respect to a translation in
time exhibits conservation of energy. If the system does not depend on
time in any way, the Lagrangian is not a function of time. Therefore,
instead of L(q, q̇, t) the Lagrangian will be L = L(q, q̇). Consequently,
∂L/∂t = 0. Recall that if the constraints and the transformation equa-
tions and the potential energy are all time independent, the Hamilton-
ian is equal to the total energy. If a system does not depend explicitly
on time, we expect these conditions to be met. Therefore, it is suffi-
cient to prove that the Hamiltonian is constant if the Lagrangian is not
a function of time.
But if L does not depend on t, its total time derivative will be
dL ∂L dq̇ ∂L dq
= + .
dt ∂ q̇ dt ∂q dt
By Lagrange’s equation,
∂L d ∂L
= ,
∂q dt ∂ q̇
so
dL ∂L dq̇ dq d ∂L
= + .
dt ∂ q̇ dt dt dt ∂ q̇
Now note that
 
d ∂L ∂L dq̇ dq d ∂L
q̇ = + ,
dt ∂ q̇ ∂ q̇ dt dt dt ∂ q̇
270 8. CONSERVATION LAWS AND SYMMETRIES

so  
dL ∂L dq̇ ∂L dq̇ d ∂L
= − + q̇ .
dt ∂ q̇ dt ∂ q̇ dt dt ∂ q̇
Therefore,
 
dL d ∂L
= q̇ ,
dt dt ∂ q̇
or  
d ∂L
L − q̇ = 0.
dt ∂ q̇
The definition of the Hamiltonian is
H = pq̇ − L,
∂L
where p = ∂ q̇
. Therefore the quantity in brackets is −H and
d
[−H] = 0.
dt
Consequently,
H = constant.
That is, for a system in which L is independent of the time, the Hamil-
tonian (and hence the energy) is constant.

8.4. Are the Laws of Physics Symmetrical?

So far we have seen that if a physical system exhibits some sort of
symmetry, it will be characterized by having a corresponding conserved
quantity. (Thus, we showed that if the system is symmetrical with
respect to translations, then its linear momentum is constant, and if it
is symmetrical with respect to time, its energy is constant.)
However, we have not shown that the laws of physics are invari-
ant under symmetry operations, such as translations, reflections, in-
versions, and so on.
The discussion so far could lead you to believe that the laws of
physics are invariant under any kind of transformation, but this is not
true. For example, Galileo noted that the laws of physics are not invari-
ant under a change of scale. To use his simple example, the bones of
a dog would be shaped differently if dogs were as big as houses. Feyn-
man pointed out that a model cathedral built out of toothpicks has
very different properties from a full sized cathedral. There are trans-
formations (such as change in scale) under which physical relationships
do not remain invariant. Aeronautical engineers are well aware that a
scale model airplane in a wind tunnel does not behave the same as the
full scale airplane. However, you probably expect the laws of physics
 8.4. ARE THE LAWS OF PHYSICS SYMMETRICAL? 271

to be invariant under a reflection or a translation. In fact, the force

of gravity, the electromagnetic force, and the strong nuclear force all
exhibit symmetry under reflections and translations. We have seen in
Section 3.5 that a rotating water filled bucket exhibits different behav-
ior than a bucket full of water at rest. In Chapter 13 we will see that
F = ma takes on a more complicated form in a rotating coordinate
frame.
A less well known lack of symmetry is related to the parity opera-
tion.
8.4.1. Non-conservation of Parity. The parity operation is es-
sentially a quantum mechanical operation but in classical terms we can
think of it as a reflection followed by a 180o rotation about the reflected
axis is called the parity operation. The effect of this inversion opera-
tion is to reverse each coordinate axis; in a parity operation all three
coordinates change sign:
(x, y, z) → (−x, −y, −z).
A physical system can be assigned a value for parity, either even or
odd, depending on how it behaves under a parity operation.3
This can be more easily understood in terms of a specific example.
Consider a classical system described by a position vector r and an an-
gular momentum L. The position vector has odd parity because under
the parity operation r → −r. Likewise, the linear momentum vector
p = mv has odd parity. The angular momentum vector, however, has
even parity because it is given by L = r × p. (The spin vector S also
has even parity.)
It seems reasonable that the parity of a physical system should
remain constant. The invariance of the parity of a physical system is
referred to as “conservation of parity.” For example, in the reaction
π − + d → n + n,
the total parity of the products (the two neutrons) is the same as the
total parity of the pion and deuteron. (When we speak of the parity of a
particle, we mean the parity of the quantum mechanical wave function
that describes the particle.)
In the 1950’s there was a great deal of interest in interactions in-
volving elementary particles. Physicists were surprised to discover that
3In quantum mechanics the wave function describing a system is characterized
by its parity; for example, the parity of a system with orbital angular momentum
number l is (−1)l . If the wave function is unchanged under an inversion, we say the
system has “even” parity, but if it is transformed into the negative of the original
wave function, we say it has “odd” parity.
272 8. CONSERVATION LAWS AND SYMMETRIES

parity is not conserved for particles undergoing a particular decay pro-

cess called beta decay. This discovery was related to a problem that
high energy physicists called the tau-theta puzzle. The particle known
as the τ meson was observed to decay into three pions, and the final
state had odd parity. The particle known as the θ meson decayed into
two pions and the final state had even parity. But the tau and theta
mesons were identical in all respects except for their decay products.
They had to be the same particle! The tau-theta puzzle was solved in
1956 by T. D. Lee and C. N. Yang who suggested that the tau and
theta mesons were indeed the same particle but that parity was not
conserved in the decay process. Their idea was verified by a series of
clever experiments carried out by C. S. Wu.
High energy physicists began to question some other conservation
laws. They denoted the parity operation by P. The operation called
charge conjugation in which a particle is transformed into its antipar-
ticle was denoted by C. It was found that in nature there are certain
reactions involving elementary particles in which CP is found to be
violated. (The CP operation involves changing all the particles into
antiparticles and then reflecting all the axes.)4

8.5. Strangeness
Conservation laws are used in an unusual but interesting way by
high energy physicists. In studying the elementary particles, it is found
that some reactions never occur. There is no apparent reason why a
reaction such as
π− + p → π◦ + Λ
is never observed. It does not violate any of the everyday conserva-
tion laws. But it does not happen. Using the principle that “what is
not forbidden is required,” high energy physicists decided that there
must be a conservation principle at work. They called it “conserva-
tion of strangeness.” The way it works is this: Each particle is given
a “strangeness quantum number.” For example, the strangeness of a
π particle is 0, that of a proton is also 0, and the strangeness of a Λ
particle is -1. Note that the total strangeness on the right hand side
4However, relativistic quantum mechanics shows that CPT must be invariant
in any physical process. Here the letter T stands for the time reversal process. It is
interesting to note that CP violation was confirmed experimentally at the Stanford
Linear Accelerator Center (SLAC) in 2001 based on the observation of 32 million
decay events by a 1,200 ton detector. The violation of CP explains why the universe
is predominantly made of ordinary matter rather than being 50% anti-matter. (See
Colin Macilwain, “Physicists show what really matters,” Nature, 412, 105, 12 July
2001.)
 8.5. STRANGENESS 273

of the reaction is not equal to the total strangeness on the left hand
side. Therefore, in this reaction, strangeness is not conserved and the
reaction is “forbidden.” This is not any weirder than requiring that
linear momentum be conserved during a collision. However, we feel at
home with momentum conservation because we have an analytical ex-
pression for momentum and we know that conservation of momentum
implies that no net external forces are acting on the system. We do
not have an analytical expression for strangeness nor do we know the
symmetry implied by strangeness conservation. We can be sure, how-
ever, that there is some symmetry in nature that requires strangeness
to be conserved.5
As an example of the importance of strangeness, Figure 8.1 is a
plot of strangeness vs charge for the eight spin 1/2 particles called
baryons. Note the neutron and proton on the top of the plot, both
having strangeness 0, but charge 0 for the neutron and +1 for the pro-
ton. (The lines of constant charge are skewed.) A plot just like this one
can be generated for the eight spin zero particles called mesons. These
plots (known as “The Eightfold Way”) are the basis of the standard
model of particle physics.6

n p
s=0

s=-1 Λ
Σ- Σ0 Σ+

s=-2
Ξ- Ξ0

q=-1 q=0 q=+1

Figure 8.1. The Eightfold Way: a plot of the strange-

ness and charge of the eight baryons. Note that the
charge axis is skewed.

5Ifyou wish to delve further into these questions, a good source is, Brehm, J.
and Mullin, W. J., Introduction to the Structure of Matter. John Wiley and Sons,
1989, New York, NY. Chapter 16.
6An interesting article on the history of the standard model by one of its
founders is Gerard ’t Hooft, “The making of the standard model”, Nature, 448,
271-273 (19 July 2007).
274 8. CONSERVATION LAWS AND SYMMETRIES

8.6. Symmetry Breaking

You may have heard the expression “symmetry breaking.” This is
used to describe a situation in which a symmetry ceases to exist. Con-
sider, for example, the freezing of a liquid. An atom in a liquid is as
likely to move in one direction as in another. There are no preferred
direction for the coordinate axes. There is isotropy and spatial ho-
mogeneity so the system exhibits symmetry under translation. If the
liquid freezes, the situation changes drastically because now the crystal
does have preferred directions. The space is no longer isotropic. The
symmetry between different directions has been “broken.” Theoreti-
cians interested in the evolution of the early universe believe that at
the high temperatures shortly after the big bang, the electromagnetic
interaction and the weak interaction were indistinguishable. But as the
universe cooled down there was some sort of “phase transition” caus-
ing these two forces to have vastly different magnitudes and to act on
vastly different spatial scales. This symmetry breaking is sometimes
referred to as a “freezing out” of the different kinds of forces.7

8.7. Summary
The point of this chapter was to introduce you to the idea that con-
servation laws are a consequence of symmetries in the physical universe.
We showed that momentum conservation is related to the homogeneity
of space (symmetry under translation) and energy conservation is a
consequence of temporal symmetry. The technique we used to demon-
strate this was to consider Lagrange’s equation when some coordinate
is ignorable. An ignorable coordinate in the Lagrangian implies a sym-
metry with respect to that variable.
Finally, we considered examples from elementary particle physics in
which the consequences of conservation laws are evident even if we do
not have a mathematical expression for the conservation law or even
know the form of the conserved quantity.

8.8. Problems
Problem 8.1. Consider a double pendulum (as in Chapter 4).
What physical quantities are conserved?
Problem 8.2. (a) Prove that if a certain coordinate qi does not
appear in the Lagrangian, it is also ignorable in the Hamiltonian. (b)
7For
a good discussion see Whitten, E., “When Symmetry Breaks Down,”
Nature, 429, 507-508, (3 June 2004).
 8.8. PROBLEMS 275

Show that
∂H ∂L
=− .
∂qi ∂qi
Problem 8.3. Does the law
dL
N=
dt
exhibit symmetry under reflections? Explain.
Problem 8.4. Are your footprints chiral? Does the pattern formed
by your footprints when you are walking exhibit chirality? Would chi-
rality be exhibited if the pattern were of infinite length? (Hint: What
is a “glide reflection”?)
Problem 8.5. The “Baryon Decuplet” consists of four ∆ parti-
cles with charges from -1 to +2 and strangeness (s)=0, three Σ∗ with
charges from -1 to +1 and s=-1, two Ξ∗ with charges -1, 0 and s=-2,
and one Ω with charge -1 and s=-3. Plot the decuplet on a graph like
that of Figure 8.1.
Problem 8.6. The position vector r has components (1,2,3). Ex-
press the components of this vector after (a) an inversion and (b) a
reflection in the xy plane. (c) Let r0 be the vector obtained by re-
flecting r in a plane perpendicular to the unit vector n̂. Show that
r0 = r − 2(r · n̂)n̂.
Problem 8.7. An electrmagnet consists of a current carrying coil
wrapped around an iron core, as illustrated in Figure ??. (a) What
is the direction of B in the mirror image? (b) If a positive charge is
moving towards the mirror, it will feel a force F = qv × B directed out
of the page. What is the direction of the force in the reflected system?
276 8. CONSERVATION LAWS AND SYMMETRIES

B
q
v

Figure 8.2.
 Part 2

The Gravitational Field

Chapter 9
The Gravitational Field

Classical field theory is primarily a study of electromagnetic and

gravitational fields. This chapter is an elementary introduction to field
theory and limited to a few aspects of the gravitational field.
Perhaps the most thoroughly studied field is the electromagnetic
field. You will learn about it in your E&M course. However, field the-
ory is used to study many other physical phenomena. For example,
field theory is used in fluid dynamics and in studies of elasticity, not
to mention such important reasearch areas in physics as the unified
field theory (which hopes to unite gravity and electromagnetism) and
quantum field theory. Quantum field theory arose from studies of the
power radiated by an atom when it transitions to a lower energy state.
An interesting aspect of quantum field theory is that it associates par-
ticles with the fields. Relativistic quantum field theory has become an
important tool in high energy physics. As you might imagine, field
theory is quite complex, both conceptually and mathematically. You
may be relieved to know that in this chapter we will limit ourselves to
a few fairly simple ideas.1
An interesting and important aspect of field theory is the forma-
tion and propagation of waves. When you study the electromagnetic
field you will spend a significant amount of time and mental energy
learning about electromagnetic waves. These waves are of great prac-
tical importance, particularly since X-rays, visible light, radio waves,
1To learn more about field theory and the closely related subject called potential
theory, see L. D. Landau and E. M. Lifshitz, The Classical Theory of Fields: Course
of Theoretical Physics. Vol 2.Pergammon Press, 1975. A very old but very good
book that was first published in 1929 is Oliver Dimon Kellogg, Foundations of
Potential Theory. J. Springer, 1929. (Available as a republication from Dover
Publications.)
279
280 9. THE GRAVITATIONAL FIELD

and infrared radiation are all electromagnetic waves. The gravitational

field also gives rise to waves. Einstein’s general theory of relativity pre-
dicts that gravity waves are generated by the acceleration of massive
bodies. Thus, gravity waves are expected to be produced by super-
nova explosions and by orbiting pulsars. Recently, direct detection of
gravitational waves was made with a very large (and complex) system
called LIGO.2 We will not be considering gravitational waves here.

9.1. Newton’s Law of Universal Gravitation

The concept of gravitation was put on a mathematical basis by
Isaac Newton when he formulated the law of universal gravitation.
He postulated that every particle in the universe attracts every other
particle with a force that is proportional to the product of their masses
and inversely proportional to the square of the distance between them.
In equation form, Newton’s law of universal gravitation states that the
force on a particle of mass m2 due to a particle of mass m1 is
m1 m2
F21 = − G r̂, (9.1)
|r2 − r1 |2
where G is a constant of proportionality, known as the universal gravi-
tational constant and found experimentally3 to be approximately equal
to 6.67×10−11 Nm2 /kg2 . As shown in Figure 9.1, the vectors r1 and
r2 are the position vectors of the two particles, and r̂ is a unit vector
pointing from m1 to m2 . This unit vector is given by
r2 − r1
r̂ = .
|r2 − r1 |
A slightly different way to express the law of gravitation is
r2 − r1
F21 = − Gm1 m2 . (9.2)
|r2 − r1 |3
Although this is more complicated, it avoids confusion as to the direc-
tion of the force. Recall that F21 is the force on m2 due to m1 and that
2LIGO stands for Laser Interferometer Gravitational-Wave Observatory. On
September 14, 2015, LIGO observed gravitational waves from the merger of
two black holes. More details are found in the article by B. P. Abbott
and many co-authors in Physical Review Letters, Volume 116, February, 2016,
DOI:10.1103/PhysRevLett.116.061102, and in the article by Adrian Cho in Science
Magazine, Feb 11, 2016 (DOI: 10.1126/science.aaf4041)
3A recent measurement of G by Guglielmo Tino and collaborators at the Uni-
versity of Florence, Italy, yielded a value of G = 6.67191(99) × 10−11 m3 kg−1 s−2 .
See the article “Precision measurement of Newtonian gravitational constant using
cold atoms,” by G. Rosi et al., Nature 510, 518-521, 2015.
 9.1. NEWTON’S LAW OF UNIVERSAL GRAVITATION 281

m1 ^
r
r2- r1
r1
m2
r2

Figure 9.1. A body of mass m1 is located at r1 and a

body of mass m2 is located at r2 . The relative position
of m2 with respect to m1 is r2 − r1 .

the negative sign in the equation indicates that the force is attractive.
By Newton’s third law, the force on particle m1 due to particle m2 is
r1 − r2
F12 = − Gm2 m1 .
|r2 − r1 |3
What is the force law for extended bodies? If the distance between
two bodies is much larger than the dimensions of the bodies (as for
astronomical objects), it is usually safe to assume that the bodies are
particles and the quantity |r2 − r1 | is the distance between the centers
of the two bodies. If an extended body cannot be treated as a particle,
you will need to determine the gravitational field of the extended body,
as discussed below in Section 9.3. (For example, if a satellite is in a
near-Earth orbit, the Earth cannot be considered a point mass.)

9.1.1. Universality of the Law of Gravitation. According to

Newton’s Law of Gravitation, every object (every atom) in universe
is exerting a force on every other object in the universe. It is true
that the force decreases as 1/r2 (inverse square force law), so the force
exerted on us by the distant stars is entirely negligible. Nevertheless,
a person on Earth is exerting a tiny force on all the planets and stars
and galaxies.4

9.1.2. Action at a Distance. Newton’s law of universal gravita-

tion is a prime example of the concept of “action at a distance” which
means that body A exerts a force on B, no matter how far apart the
bodies happen to be. Naturally, the question arises, “How can bodies
that are separated by great distances exert forces on one another?”
Newton stated, “Hypothesis non fingo,” which means, “I make no hy-
potheses.” This is a perfect example of physics telling us how nature
4The British physicist P.A.M. Dirac is reputed to have expressed this aspect
of the law of gravitation with the poetic sentiment, “When you pluck a flower, you
move a distant star.”
282 9. THE GRAVITATIONAL FIELD

works but making no assumptions (or hypotheses) as to why it works

that way.
The biggest problem with the action at a distance concept is that
it implies that the force between two bodies is propagated instanta-
neously. That is, if body A is moved to another position, body B will
instantaneously notice a change in the force acting on it, even if B is
halfway across the universe. This is obviously impossible. We believe
(but cannot yet prove) that the gravitational force is transmitted at
the speed of light. The conceptual difficulties inherent in Newton’s law
of universal gravitation expressed in terms of action at a distance are
effectively dealt with when one considers the gravitational field, rather
than the gravitational force.

Exercise 9.1. Determine the force exerted by Mars on the Sun.

What is the acceleration of Mars? What is the acceleration of the
Sun?
Exercise 9.2. A body falls in the gravitational field of the Earth.
Show that its acceleration is independent of its mass.
Exercise 9.3. A particle of mass m is located at (1,2), a particle
of mass 2m is located at (3, 4), and a particle of mass 3m is located
at (−2, −2). Determine the gravitational force acting on the particle of
mass 2m. Answer: F = −Gm2 (0.24ı̂ + 0.26̂).
Exercise 9.4. Prove that the gravitational force is conservative by
evaluating the curl of Newton’s Law of Universal Gravitation.

9.2. The Gravitational Field

By definition, a field is a physical quantity that is defined at ev-
ery point in some region of space. For example, the temperature is a
physical quantity and using a thermometer you could determine the
temperature at every point in this room. This scalar field might be
described by an equation T = T (x, y, z), or by a table of values, or by
a drawing showing the temperature isopleths of the room. Similarly,
the water in a river has a velocity at every point. The region of space is
the river and the physical quantity is velocity. This vector field might
be represented graphically or perhaps by an equation for the velocity as
a function of position. Suppose that somehow you could determine the
 9.2. THE GRAVITATIONAL FIELD 283

gravitational force on a mass at all points near a planet. This would

define the gravitational force field of the planet. Note that in all of
these examples, the value of the field depends on position.
To define the gravitational field more explicitly, consider a region
of space that is empty except for two particles. One particle has mass
M and it is located at a fixed point in space. At some other point you
place a particle of much smaller mass m. Assume you have a way to
determine the force acting on the small mass, wherever it may happen
to be. The small mass will be the “test body.”
Now imagine you move this test body from one point to another
and you measure the magnitude and the direction of the gravitational
force acting on it at every location. If you represent the force by a
vector, you will end up with a lot of force vectors all pointing towards
M. Unfortunately, as indicated by Equation (9.1), these force vectors
depend on the mass of the test body so they are not a good measure of
the gravitational field of M. To avoid any dependence on the test body,
you can define the gravitational field of M as the force per unit test
mass at all points in space around M. Furthermore, since you don’t
want m to have any influence on the field, you take the limit as m → 0.
These concepts are expressed more simply using mathematics. Let
F be the force on m due to M and let g be the gravitational field of
M. Then, by definition, if m is at r and M is at r0 ,
r − r0
 
F
g(r)= lim = −GM . (9.3)
m→0 m |r − r0 |3
The notation g = g(r) reminds us that g is a function of position.

M = source point
r-r'
r'
P = field point
r

Figure 9.2. Particle M at the source point r0 generates

a gravitational field everywhere. The gravitational field
at point P (r) (the “field point”) is given by Equation
9.3.

The location of body M is called the “source point,” and its position
is denoted by r0 . The point specified by r is called the “field point.” It
is the place where the field is evaluated. See Figure 9.2. Note that the
test mass does not enter into Equation (9.3), the expression for g(r).
284 9. THE GRAVITATIONAL FIELD

In introductory physics books it is usual to place the origin of co-

ordinates at the mass M. This gives a simpler but less descriptive ex-
pression for the field:
r r̂
g(r) = − GM 3 = −GM 2 , (9.4)
|r| r
where r is the location of the field point.
Let me be very clear on the difference between the force and the
field. The force F depends on the masses of both of the particles and
their positions at some instant of time. The field g depends only on
the mass of the source particle M and the relative position of the field
point. The force F describes an interaction between two objects, but
the field g is a property of a point in space.5
The point mass M is the source of the gravitational field. The field
extends throughout all space. It is not easy to grasp the field concept;
it is particularly difficult to describe exactly what is filling all of space.
We know that if we place a test mass at any point in the field it will
feel a force, so we think of the mass M as producing something which
exerts a force on any other mass in this space. The field is everywhere,
but it exerts a force only when a material body is placed in the field.
At the surface of Earth the gravitational field is approximately given
by g = − 9.8 m/s2 k̂. Here k̂ is a unit vector pointing upward at the
surface. For points far above the surface, the field is approximately
g(r) = −(GME /r2 )r̂ where the origin is at the center of the Earth and
ME is the mass of the Earth. (Do not confuse the field vector g(r)
with the constant value 9.8 m/s2 which is usually abbreviated by g.)
The field approach to gravity is quite different from the action at
a distance approach. In the action at a distance approach we consider
the force one body exerts directly on another. When using the field
concept, we think of the interaction of two bodies (M and m) as a
two step process in which mass M generates a field and then mass m
interacts with the field, rather than interacting with M directly. Thus
the field approach decouples the sources from the test body used to
determine the field. If two different source arrangements generate the
same field at some point, a test body at that point will feel the same
force.
It is tempting to consider the field to be nothing but a convenient
way of expressing the force, but it turns out that the field has an
5Sometimes,I like to change the definition of field slightly and state, “A field
is a region of space in which some physical quantity is defined at every point.”
This focuses the mind on the space rather than on the physical quantity. It is not,
however, the standard definition of field.
 9.2. THE GRAVITATIONAL FIELD 285

existence of its own, independent of the force. A force is just a force,

but a field has energy, momentum, and angular momentum as well as
the ability to exert a force. A field can propagate through space as a
wave and can exist independent of the source. (The field can persist
even after the source has ceased to exist.)
Although the field approach seems more complicated because it
introduces a new concept (“the field”), it does resolve a number of
difficulties with the action at a distance concept. For example, in the
action at a distance scenario, forces are transmitted instantaneously.
In the field concept, the response of the field to changes in the source
propagate at a finite speed. Since the source (M ) generates the field, it
is obvious that if the source is moved, the field will change. However,
this change need not occur instantaneously at all points in space. As
mentioned before, we believe that changes in the gravitational field are
propagated at the speed of light.
Furthermore, the field approach leads to a conceptual framework for
understanding how the force is transmitted. In quantum field theory
the interaction between particles can be described as an interchange of
“virtual” particles. The electromagnetic force, for example is believed
to be transported by virtual photons. This is often represented graphi-
cally by “Feynman diagrams.” For example, the electromagnetic inter-
action between two electrons (“electron-electron scattering”) is repre-
sented by the Feynman diagram of Figure 9.3. In the figure, the two
electrons are represented by the straight lines and the wavy line repre-
sents a virtual photon where the word “virtual” is used because these
photons are not (and cannot be) observed. The photon transmits the
electromagnetic force. It is emitted by one of the electrons and ab-
sorbed by the other. Both particles are deflected from their original
paths. Similarly, the gravitational force is assumed to be transmitted
by undetected particles called “gravitons.” Since we have our hands full
with just Newton’s theory of gravitation, we will not consider any of
these advanced concepts in this book. However, if you are interested in
delving deeper into quantum field theory, I recommend you read a very
short but very profound book by Richard Feynman called “QED.”6

Exercise 9.5. Two particles of mass M are separated by a dis-

tance a. Determine the gravitational field at a distance z along the

6Richard Feynman, QED: The Strange Theory of Light and Matter. Princeton
University Press, Princeton, NJ, 1985.
286 9. THE GRAVITATIONAL FIELD

e e

Figure 9.3. A Feynman diagram for electron-electron

scattering. The wavy line represents the virtual photon.

perpendicular bisector of the line joining the masses. Answer: g =

3/2
−2GM zẑ/ [(a/2)2 + z 2 ]
Exercise 9.6. A point mass M is located at the origin. Another
point mass M is located at (0,4). Determine the gravitational field at
(0,3) and at (3,0). Answer: Field at (3,0) is g = GM [−0.14ı̂ + 0.03̂] .
Exercise 9.7. Determine the gravitational field due to the Earth
and the Sun at the position of the Moon, (a) at the time of a lunar
eclipse, (b) at the time of a solar eclipse. (c) Determine the force on
the Moon at these two times.

9.3. The Gravitational Field of an Extended Body

The gravitational field produced by an extended body can be de-
termined by treating the body as a collection of point masses and sum-
ming the fields due to each individual particle. (Gravitational fields
obey the principle of superposition and add vectorially.) Although this
is a perfectly reasonable thing to do, it is not a practical thing to do.
So we assume that the material of the extended body is continuous.
This assumption breaks down on the atomic level where there are huge
differences in mass density between the small, heavy nuclei and the
empty space that comprises most of the volume of atoms. However, a
volume element that contains billions of atoms is still very small on the
macroscopic scale and we are justified in thinking of it as containing a
continuous distribution of matter with a well defined density.
Figure 9.4 shows an infinitesimal volume element (dτ 0 ) in a contin-
uous body. The mass contained in this volume element is dm = ρdτ 0
where ρ = ρ(r0 ) is the density. (For a surface, the element of mass is
σdA, for a line it is λds and for an extended body it is ρdτ, where σ, λ, ρ
are the mass per unit area, per unit length, and per unit volume.) The
 9.3. THE GRAVITATIONAL FIELD OF AN EXTENDED BODY 287

density may vary from one part of the body to another so it is written
as ρ = ρ(r0 ) to remind us that ρ is a function of position.
In Figure 9.4 the point P is the field point. Note that the source
point is at r0 and the field point is at r. The infinitesimal portion of
the gravitational field at r due to an infinitesimal mass element dm at
r0 is
r − r0
dg(r) = − Gdm .
|r − r0 |3
or
r − r0
dg(r) = − Gρ(r0 )dτ 0 ,
|r − r0 |3
where dτ 0 is so small that ρ(r0 )dτ 0 can be considered a particle, yet
large enough to treat the mass as continuous. To obtain the field at P
due to the whole body, we integrate over the body. That is,
0
0 r−r
Z
0
g(r) = −G ρ(r ) 0 3 dτ . (9.5)
body |r − r |

dm' = ρdτ'
r - r'
dg
P
r'
r

Figure 9.4. An extended body. The mass element

dV 0 = ρdτ 0 is located at r0 (the source point) and it
produces an infinitesimal gravitational field dg at point
P located at r (the field point).

For an extended body with sufficient symmetry, it is usually not

too difficult to evaluate this integral. To find the field due to an arbi-
trarily shaped body, it is usually best to first determine the potential
and then obtain the field, using the technique described in Section 9.4
below. Therefore, I will not emphasize the direct integration of Equa-
tion (9.5). Nevertheless, you should have some experience with this
technique so I will work through two examples. You can get some
additional experience by working out the exercises below. It might
help you to review the section on the electric field of continuous charge
distribution in your introductory physics textbook, as the techniques
are exactly the same. It is usually easier to solve such problems by
288 9. THE GRAVITATIONAL FIELD

resolving the vector dg into its components before carrying out the
integration.

Worked Example 9.1. A long thin rod of length L lies along

the x axis. The end points of the rod are at x0 = 0 and x0 = L.
Assume the rod can be represented as a continuous mass distri-
bution with linear density λ mass per unit length. Determine the
gravitational field at the point x = 4L.
Solution: In this problem, the mass element is dm = λdx0
where the variable x0 represents the distance from the origin to the
mass element. The infinitesimal field at P due to this portion of
the rod is
r − r0 x − x0
dg = −Gdm = Gdm
|r − r0 |3 |x − x0 |3
(x − x0 )ı̂ dx0
= −Gλdx0 = −Gλı̂ .
|x − x0 |3 (x − x0 )2
The field point is x = 4L, so integrating we have
Z x0 =L L
dx0

1 1
g = −Gλı̂ 0 2
= −Gλ 0
ı̂ = − λGı̂.
x0 =0 (4L − x ) 4L − x 0 12L

Worked Example 9.2. Determine the gravitational field at

a point P a distance +z along the axis of a very thin disk of mass
M, uniform density, and radius R. See Figure 9.5.
Solution: Treat the disk as if it had zero thickness. Then
the appropriate mass element is σdA where σ = M/πR2 . As indi-
cated in Figure 9.5 the field at point P due to mass element dm is
the vector dg pointing towards dm. The distance from the source
point dm to the field point is denoted D. Consider all the mass
elements lying on a ring a distance r from the axis. The vectors
from P to these mass elements form a cone of vectors with vertex
at P. The horizontal components of these vectors will cancel. The
components along the axis will sum. The resultant field is given
by
Z Z Z
Gdm σdA
g = (dg) cos θk̂ = − 2
cos θk̂ = −G cos θk̂.
D D2
 9.4. THE GRAVITATIONAL POTENTIAL 289

But cos θ = z/D = z/(r2 + z 2 )1/2 , and dA = rdφdr, so

Z R Z 2π Z R
(dr)(rdφ) rdr
g = −σGk̂ z 2 2 3/2
= −σGk̂(2π)z 2 2 3/2
r=0 φ=0 (r + z ) 0 (r + z )
Z u=R2 +z2  
du 1 1
= −σGk̂(π)z = −2πσGk̂z −√
u=z 2 u3/2 z z 2 + R2
 
2GM z
= − 2 1− 2 k̂.
R (z + R2 )1/2

P
dg

θ d
dr
D
r
φr

Figure 9.5. The gravitational field of a thin disk is de-

termined by integration of the elementary field vectors
dg.

Exercise 9.8. Determine the gravitational field at a point z on the

axis of a ring of mass M and radius R. Answer: −GM z/(z 2 + R2 )3/2 .
Exercise 9.9. Show that a disk can be treated as a collection of
concentric rings. Write an integral expression for g using the result of
Exercise 9.8 and show that it yields the expression obtained in Worked
Example 9.2.

9.4. The Gravitational Potential

The gravitational potential is a scalar quantity related to the grav-
itational field which is a vector quantity. It is usually easier to work
with the potential than the field; if you know the potential you can
always determine the field.
Recall the discussion in Section 5.3.3 proving that if F is a conser-
vative force, then ∇ × F = 0. It is easy to show that the gravitational
290 9. THE GRAVITATIONAL FIELD

force is conservative. (See Exercise 9.4 above.) Using the fact that the
curl of the gradient of any scalar function is zero, we conclude that
since ∇ × F = 0, then F can be expressed as the gradient of some
scalar function. That is, we can define the potential energy V = V (r)
such that
F = −∇V.
Then
Z r2 Z r2 Z 2
F · dr = − ∇V · dr = − dV = V1 − V2 .
r1 r1 1

Thus, the difference in potential energy between two points is defined

by
Z r2
V2 − V1 = − F · dr.
r1

In working with the gravitational potential energy, it is often convenient

to define the zero of potential energy at infinity, so let r1 = ∞ and
V1 = 0. Writing r for r2 we have
Z r
V =− F · dr.
∞

If F is the gravitational force acting on test body m located a distance

r from point mass M, you can write
Z r
GM m Mm
V (r) = − − 2 r̂·dr = −G . (9.6)
∞ r r
This is the gravitational potential energy of a system of two point
masses.
When using the field approach, it is convenient to define a quantity
called the “gravitational potential” which we will denote by the sym-
bol Φ. The gravitational potential is obtained from the gravitational
potential energy by dividing by m, the mass of the test body. The
potential energy (V ) of two interacting masses is an example of action
at a distance and depends on the presence of two bodies. The potential
(Φ) is defined at every point in space and depends on the presence of
just one body. Thus, Φ = Φ(r) is a field. The potential is related to
the potential energy by
V
Φ(r) = lim ,
m→0 m

where m refers to an infinitesiml test mass located at r.

 9.4. THE GRAVITATIONAL POTENTIAL 291

Dividing Equation (9.6) by m shows that the gravitational potential

at r due to a particle of mass M located at the origin, is
GM
Φ(r) = − . (9.7)
r
The potential of an extended body is obtained from
ρ(r0 )dτ 0
Z
Φ(r) = −G 0
. (9.8)
body |r − r |

where r is the field point and r0 is the location of the source point
dm = ρ(r0 )dτ 0 .

Figure 9.6. The potential at P due to a shell of radius

a. The mass element dm is the mass of the element of
area dA = a2 sin θdθdφ. If σ is the surface mass density,
dm = σdA.

Worked Example 9.3. Determine the gravitational potential

at a point outside of a spherical shell of mass M and radius a. See
Figure 9.6
Solution: The mass per unit area (σ) of the (uniform) shell is
the total mass of the shell divided by its surface area, or
M
σ= .
4πa2
Selecting the mass element is more or less an “art” and you will
become proficient at making a good choice after you have worked
292 9. THE GRAVITATIONAL FIELD

out a number of problems. For a mass shell, pick a small element

of area on the surface whose mass is
dm = σdA = σa2 sin θdθdφ.
Integrating over φ generates a ring on the surface of the shell, as
indicated in the figure. (Integrating over φ is equivalent to letting
the line a sin θ rotate around r. Note that all points on this ring
are equidistant from the field point.) The potential at P due to
the entire shell is given by the double integral

2π π
σa2 sin θdθdφ
Z Z Z
σdA
Φ = −G = −G ,
surface |r − r0 | φ=0 θ=0 |r − r0 |
Z 2
σa sin θdθ
Φ = −2πG ,
|r − r0 |
where the factor 2π came from integrating over φ. Using the law
of cosines, the figure shows that since |r0 | = a,
√
|r − r0 | = R = a2 + r2 − 2ar cos θ,
and R2 = a2 + r2 − 2ar cos θ. Therefore 2RdR = 2ar sin θdθ and
the integral takes the form
Z π 2 Z R=r+a
σa sin θdθ 2 (R/ar)dR
Φ = −2πG = −2πGσa
0 R R=r−a R
Z r+a
−2πGσa
= −Gσa2 (2π/ar) dR = [(r + a) − (r − a)]
r−a r
4πa2 σ GM
= −G =− ,
r r
where we used the fact that as θ goes from 0 to π, R ranges from
r − a to r + a.
Our result shows that for points outside the shell, the potential
of a shell of mass M is the same as the potential of a point mass
M at the origin. As a problem you can show that the field inside
the shell is zero.

Exercise 9.10. Determine the potential on the axis of a disk of

thickness t. Assume that t << z so the thin disk approximation is still
valid. (Answer: Φ = −2πρtG[(z 2 + R2 )1/2 − z].)
 9.5. FIELD LINES AND EQUIPOTENTIAL SURFACES 293

Figure 9.7. The force vectors are represented on the

left. The strength of the force is represented by the
length of the arrows. On the right, the field lines give
the direction of the force and the density of field lines at
any point is proportional to the strength of the field.

9.5. Field Lines and Equipotential Surfaces

Going back to the gravitational field (g) due to a point mass M ,
recall that g is the force per unit mass on an infinitesimal test body (See
Equation 9.3). You could, in principle, transport the test body to every
point in space around M and measure the force on it at each location.
Then you could represent the field g in a number of different ways. For
example, you could make up a large table giving the magnitude and
direction of g at every point. Or you could represent g by an equation.
(It would be Equation 9.4.) Or you could represent g graphically. The
graphical representation would show many arrows pointing toward M.
The field at some given point might be represented by drawing a vector
of appropriate length at that point, as illustrated by the sketch on the
left in Figure 9.7. Note that the vectors are longer at points close to
the mass and shorter at points further away. But this is not a very
good way to represent the field. The sketch on the right in Figure
9.7 illustrates a much more convenient way. In this representation the
lines give the direction of the force and the areal density of the lines
at a point gives the magnitude of the force at that point. Imagine the
lines are drawn in such a way that the number of lines crossing a unit
area perpendicular to the field is numerically equal to the field at that
location. If the field at some point has a magnitude of 15 N/kg, it
could be represented by drawing 15 lines (per meter squared) at that
point. The same number of lines go through the surface of any sphere
centered on M. Since the area of a sphere increases as r2 , the density
294 9. THE GRAVITATIONAL FIELD

of lines per unit area falls off as 1/r2 , just like the gravitational field.
The lines representing the field are traditionally called “lines of force”
although most modern books refer to them with the better terminology
of “field lines.”
Another graphical representation of the field consists in sketching
the surfaces on which the potential is constant. For a point mass,
the equipotential surfaces are concentric spheres centered on the mass.
As we will show in a moment, the field and the potential are related
by g = −∇Φ, and you know from Chapter 5 that the gradient of
Φ is perpendicular to the surfaces of constant Φ. Therefore, the two
representations (field lines and equipotential surfaces) are equivalent.

9.6. The Newtonian Gravitational Field Equations

The relationship between potential energy and force is F = −∇V
(Equation 5.7). Dividing both sides of this equation by m yields the
relation between the gravitational field g(r) and the gravitational po-
tential Φ(r).
g = −∇Φ. (9.9)
(In many books - particularly older books - the gravitational potential
is defined as the negative of Φ and the relation equivalent to Equation
(9.9) does not have the minus sign.)
The curl of a gradient is zero. Therefore,
∇ × g = −∇ × (∇Φ) = 0. (9.10)
(I am sure you expected this result because you know that g is a con-
servative field.)
Having found the curl of g, it is reasonable to go on and determine
the divergence of g. The process is a bit involved, so please follow
carefully. Consider a mass point M. Draw a surface S around it, as
illustrated in Figure 9.8. Let dS be an infinitesimal patch of the surface.
Let n̂ be a unit vector perpendicular to dS and let r be the vector from
M to dS.
Let the origin of our coordinate system be located at the point mass
M . Then the gravitational field of M at the position of the patch of
surface dS can be written as
GM
g = − 2 r̂,
r
so
GM GM
g · n̂ = − 2 r̂ · n̂ = − 2 cos θ
r r
where θ is the angle between r̂ and n̂ (see Figure 9.8).
 9.6. THE NEWTONIAN GRAVITATIONAL FIELD EQUATIONS 295

^
n dS
S ^
n
dS
θ
r r
M dS

Figure 9.8. A closed surface S encloses a point mass M.

The unit vector n̂ is perpendicular to the surface at the
location of dS, an infinitesimal portion of the surface. Let
dS⊥ be the projection of dS perpendicular to r. Note that
dS ·r̂ =dSn̂ · r̂ =dS cos θ = dS⊥ . The sketch on the right
views the areas dS and dS⊥ edge-on. The angle θ is the
angle between the radial vector and n̂, or equivalently,
between dS and dS⊥ .

Note that the solid angle7 dΩ subtended by dS at M is

dS⊥ dS cos θ
dΩ = = , (9.11)
r2 r2
where dS⊥ is the perpendicular projection of dS on r. Next consider
the surface integral
I I
GM
I = g · n̂dS = − 2
cos θdS
S S r
I I
dS cos θ
= −GM = −GM dΩ
S r2 S
= −4πGM.
This result is very interesting. It tells us that the surface integral I is
proportional to the mass M inside the surface. This fact, in itself, is
not too surprising. However, the result also tells us that it does not
matter where the mass is, as long as it is inside the closed surface. If
there are several masses (m1 , m2 , · · · ) inside S, then, by the principle

7The solid angle is the three dimensional angle formed at the vertex of a cone.
Let point O be at the origin and a surface of area dA be a distance r from the
origin. Then, if dA is perpendicular to r, the solid angle subtended by dA at O is
dΩ = dA/r2 . The units of solid angle are steradians. The solid angle subtended by
a spherical shell at a point inside the shell is 4π steradians.
296 9. THE GRAVITATIONAL FIELD

of superposition,
I X X
g · n̂ dS = − 4πGmi = −4πG mi = −4πGMenc , (9.12)
S
where Menc is the total mass enclosed by S. The equation
I
g · n̂ dS = −4πGMenc
S
is called Gauss’s Law. It is particularly useful in the study of electro-
statics.
Assume mass is continuously distributed in a volume V bounded
by a closedR surface S. The element of mass is ρdτ , so the mass enclosed
is Menc = V ρdτ. Therefore,
I Z
g · n̂ dS = − 4πGρdτ. (9.13)
S V
You may recall from your vector analysis course that Gauss’s divergence
theorem states that for any vector A,
I Z
A · n̂ dS = ∇ · A dτ. (9.14)
S V
Therefore,
R the left hand side of equation (9.13) can be expressed as
V
∇ · gdτ, and consequently
Z Z
∇ · g dτ = − 4πGρdτ,
V V
or
∇ · g = − 4πGρ. (9.15)
Thus, the divergence of g is proportional to ρ, the mass density. An
important theorem called the Helmholtz theorem states that for any
vector field, if one knows the divergence and the curl of the field, then
one can determine the field itself. For this reason, ∇ · g and ∇ × g are
called the “sources” of g. Equations (9.10) and (9.15) indicate that the
source of the gravitational field is the mass density ρ. In other words,
masses generate gravitational fields.
Worked Example 9.4. Use Gauss’s law to determine the
gravitational field outside of an infinitely long cylinder of radius a
with constant linear mass density λ.
Solution: To solve we construct a Gaussian surface which is a
closed surface that has the same symmetry as the mass distribution.
We select this surface so that on the surface g · n̂ is either constant
or zero. The field point lies at an arbitrary point on this surface.
 9.6. THE NEWTONIAN GRAVITATIONAL FIELD EQUATIONS 297

P = field point

r
a

Figure 9.9. A Gaussian surface. The outer cylinder

is the Gaussian surface appropriate for determining the
field of the cylindrical mass. Note that the field point is
on the Gaussian surface.

For this problem the appropriate Gaussian surface is a cylinder of

length L and radius r, where r > a.
On the curved face of this “tin can” surface, the field is directed
towards the axis of the mass element, and g is constant. On the
two end faces, g and n̂ are perpendicular to each other. Therefore,
Gauss’s law gives
I Z Z
g · n̂ dS = g · n̂ dS + g · n̂ dS = −4πGMenc .
S side ends
Now Z Z
g · n̂ dS = −g dS = −g(2πr)L,
side
and Z
g · n̂ dS = 0.
ends
The mass enclosed by the Gaussian surface is
Menc = λL.
Therefore,
4πGλL 2Gλ
g=− =− .
2πrL r
By symmetry, this is pointing toward the axis of the cylinder so
2Gλ
g=− ρ̂
r
where ρ̂ is the unit vector in cylindrical coordinates.

Exercise 9.11. Show that the solid angle subtended by a spherical

shell at any interior point is 4π steradians.
298 9. THE GRAVITATIONAL FIELD

Exercise 9.12. Use Gauss’s law to determine the gravitational field

of a point mass.
Exercise 9.13. Maxwell’s equations for the electric field (E) and
the magnetic field (B) can be written
∂B
∇ · E = ρ/0 ∇×E=−
∂t
∂E
∇·B = 0 ∇ × B = µ0 J+0 µ0
∂t
where ρ is the charge density and J is the current density. The quanti-
ties 0 and µ0 are constants. What are the sources of the electric field?
What are the sources of the magnetic field?

9.7. The Equations of Poisson and Laplace

Recall that g = − ∇Φ. Taking the divergence of both sides,
∇ · g = −∇ · ∇Φ= −∇2 Φ.
Using the relation ∇ · g = −4πGρ leads to
∇2 Φ = 4πGρ. (9.16)
This is called “Poisson’s equation.”
Poisson’s equation is a second order partial differential equation
whose solution depends on the boundary conditions. In a region of
space where there is no mass, the density ρ is zero, and Poisson’s
equation reduces to
∇2 Φ = 0. (9.17)
This very important special case is called “Laplace’s equation.”
Most of the time you will want to determine the field in a region of
space where ρ = 0, for example, in empty space outside of a star, and
you will use Laplace’s equation. If, however, you want to determine Φ
inside the star then you have to solve Poisson’s equation.
Since Poisson’s equation is an inhomogeneous differential equation,
its solution can be expressed as the sum of two parts:
• The general solution of the homogeneous differential equation
(Laplace’s equation).
• A particular solution of the inhomogeneous differential equa-
tion (Poisson’s equation).
 9.8. EINSTEIN’S THEORY OF GRAVITATION (OPTIONAL) 299

In your undergraduate course in electromagnetic theory you will

learn techniques for solving the Laplace equation under a variety of
different conditions. In your graduate course in electromagnetic theory
you will learn a number of methods for solving the Poisson equation.
Poisson’s equation, ∇2 Φ = 4πGρ is the “field equation” of the New-
tonian gravitational field. If you solve for Φ you know the potential
everywhere and you can determine the gravitational field g everywhere.
In the next section we shall consider the field equations for the gravi-
tational field according to Einstein.

9.8. Einstein’s Theory of Gravitation (Optional)

We cannot leave this chapter on the gravitational field without a
word or two about Einstein’s theory of gravitation, which is referred to
as The General Theory of Relativity. As you know, this is one of the
greatest achievements in physics and supersedes Newton’s theory of
gravitation. For example, it was known that the orbit of Mercury pre-
cesses at the extremely precisely measured rate of 5601 seconds of arc
per century. Newton’s theory, including the effects of the other planets
which perturb the motion of Mercury, yielded a precession rate of 5558
seconds of arc per second. The discrepancy (43 seconds per century)
was explained with great precision by Einstein’s theory. Note that
we are talking about a difference of less than one degree per century!
There are numerous other examples of situations in which Einstein’s
theory gives the correct value whereas Newton’s theory is slightly off.
In that case, you might ask, why do we bother studying Newton’s
theory? The answer is that calculations in Einstein’s theory are ex-
tremely complicated, and that Newton’s theory gives results that are
extremely close to the correct values.
This section will give you a rough idea of Einstein’s theory.8 You
will probably not be using the theory of general relativity unless you
decide to specialize in cosmology - the study of the structure of the
universe. Nevertheless, as a physicist, you should have some idea about
the theory.
You may be aware that in 1905 Einstein developed the theory called
the “Special Theory of Relativity.” (We will consider this theory in
Chapter 19.) The special theory deals with the relationship between
two inertial reference frames. That is, it considers two reference frames
that are moving at constant velocities. Clearly, the laws of physics have

8Avery readable book on the topic is “Relativity, Gravitation and Cosmology”

by Robert J. A. Lambourne, Cambridge University Press, 2010.
300 9. THE GRAVITATIONAL FIELD

to be the same in the two reference frames. In other words, a trans-

formation from one reference frame to the other should leave physical
equations and relationships unchanged. The theory was extremely suc-
cessful and was particularly applicable to electromagnetism. However,
Einstein was unsatisfied with one aspect of the special theory, and that
was that Newton’s law of universal gravitation did not transform as it
should. Furthermore, Einstein was interested in developing a theory
that would be applicable to accelerated (that is, non-inertial) reference
frames. He finally developed the “general” theory of relativity that
solved both of these problems. To do so, however, he had to delve
deeply into very advanced concepts in geometry.
Einstein’s theories required understanding how to transform from
one reference frame to another, a procedure that has much in common
with transforming from one coordinate system to another. Recall that
in Section (5.3.4) we introduced the concept of the “metric tensor”
that allowed us to transform from Cartesian coordinates to spherical
or cylindrical coordinates. Letting hij represent the elements (or com-
ponents) of the metric tensor, we wrote the differential (infinitesimal)
displacement as X
ds2 = h2ij dqi dqj ,
i,j
(see Equation 5.8). In Sections 5.3.5 and 5.3.6 we used this relation
to obtain expressions for the volume element and for del in cylindrical
and spherical coordinates.
In special relativity, Einstein had to use a 4-dimensional coordinate
system with x, y, z being the usual Cartesian space coordinates and the
fourth coordinate being ct where c is the constant speed of light. Note
that ct has the dimensions of distance, but it depends only on t. Thus,
time is the fourth dimension. Including an extra dimension requires a
somewhat more complicated expression for the displacement as
3
X
2
ds = ηµν dxµ dxν .
µν=0

There are two things to note about this expression. First of all it is
customary to use upper and lower indices which are called “contravari-
ant” and “covariant”. The difference need not concern us at present,
but it is important to appreciate that dxµ does NOT mean dx raised to
the power µ! It is just a way to denote the component9. Secondly, note
9Superscript
indices are denoted contravariant and subscript indices are de-
noted covariant. The difference between covariant and contravariant vectors is the
way they transform. The components of a covariant vector transform according
 9.8. EINSTEIN’S THEORY OF GRAVITATION (OPTIONAL) 301

that the indices µ and ν range from 0 to 3. Here 1,2,3 represent the
spatial coordinates, dx, dy, dz and zero denotes the “time” coordinate
cdt. Thus, dx3 represents dz and dx0 represents cdt. The quantity ds is
the “line element” in 4-dimensional “spacetime.” In special relativity,
the metric ηµν is quite simple. It is called the “Minkowski metric10”
and in matrix form, it can be written as
 
1 0 0 0
 0 −1 0 0 
ηµν =  0 0 −1 0 

0 0 0 −1
Now Minkowski spacetime is described as “flat” because Euclidean
geometry is valid in this space. This means, for example, that parallel
lines never meet and that the angles in a triangle add up to 180 degrees.
Note that the geometry of the surface of a sphere is non-Euclidean.
Parallel lines that start out perpendicular to the equator, meet at the
pole. The angles of a triangle may not sum to 180 degrees. In 3-
dimensional space, the line element ds on a flat surface is
ds2 = dx2 + dy 2 + dz 2 ,
but on the surface of a sphere of radius R,
ds2 = R2 dθ2 + R2 sin2 θdφ2 .
Not all surfaces that we might think of as curved are curved in the
mathematical sense. For example, the curved side of a cylinder is
mathematically flat because (as you can easily show) parallel lines never
meet, triangle have 180 degrees, etc.
Bernhard Riemann11 used the concept of line element to generalize
geometry. He expressed the generalized line element in a curved 3-
dimensional space as
3
X
2
ds = gij dxi dxj ,
i,j=1

∂xj
to A0i = ∂x0j Aj , whereas the components of a contravariant vector transform as
∂x0j
A0i = ∂xj Aj .
10This metric was developed by Hermann Minkowski who was one of Einstein’s
mathematics professors. Minkowski is credited for having developed the concept of
spacetime.
11Bernhard Riemann (1826-1866), who was a student of Carl Friedrich Gauss, is
considered one of the worlds greatest mathematicians. Among other achievements,
he developed the geometry of curved surfaces.
302 9. THE GRAVITATIONAL FIELD

where the gij are the elements of the metric tensor (or “metric coef-
ficients”). The metric relates the coordinate differentials (the dxi ) to
a length ds in the space under consideration. Therefore, the metric is
related to the geometry of the space. Once the metric is known, the ge-
ometry of the space is entirely determined. As we shall see, the metric
can not only tell us whether the space is flat or curved, it can actually
give us the curvature of the space.12 In fact the curvature is given by
a complicated quantity called the Riemann tensor that is defined as

l ∂Γlik ∂Γlij X m l X
Rijk ≡ − + Γik Γmj − Γm l
ij Γmk
∂xj ∂xk m m

where
 
1 X il ∂glk ∂gjl ∂gjk
Γljk = g + k + .
2 l ∂xj ∂x ∂xl

The Riemann tensor is rank 4 (as you can tell by noting that there
are four indices) and Γlmk is a rank 3 tensor. Obviously, this is a very
complicated quantity. You will probably never have to evaluate it un-
less you become a theoretical physicist specializing in general relativity!
l
Nevertheless, it is worth noting that for a flat space, Rijk = 0 every-
where. Now so far we have been considering the geometry of ordinary
3 dimensional space (as indicated by our use of latin letters as indices).
Einstein’s job was to generalize to 4 dimensional spacetime. Thus, in
a curved Riemannian 3-D space,
3
X
ds2 = gij dxi dxj ,
i,j=1

where ds2 is positive and the gij are functions of the coordinates,
whereas in a flat 4-D Minkowski spacetime
3
X
2
ds = ηµν dxµ dxν
µ,ν=0

where the ηµν are constants equal to (1, −1, −1, −1) and ds2 may be
positive, negative or zero. In order to treat curved spacetime, the
constant ηµν are replaced by the functional quantities gµν . That is, in

12It
might be mentioned that most of the metrics that are encountered in
physics are orthogonal, and as we have seen in Chapter 5, this means the metric
tensor can be expressed as a diagonal matrix.
 9.8. EINSTEIN’S THEORY OF GRAVITATION (OPTIONAL) 303

a curved 4 dimensional spacetime the line element is

3
X
2
ds = gµν dxµ dxν
µ,ν=0

In a flat region of spacetime, the metric reduces to the Minkowski met-

ric. It turns out that even if spacetime is curved overall, it is possible
to define a small neighborhood about any point in which the spacetime
can be considered to be flat. (This is similar to our experience on the
curved surface of the Earth; in a small neighborhood of your location,
the Earth can be treated as if it were flat.) Thus, in a small enough
neighborhood of any point, special relativity is valid. Another way of
saying this is that special relativity applies locally but not globally.
We now consider some consequences of assuming that spacetime
is curved. Einstein was aware of two problems with Newton’s law of
universal gravitation with respect to special relativity. First, as we
mentioned before, the gravitational force law does not transform as it
should under a change in inertial reference frame, and secondly the
concept of action at a distance and the instantaneous transmission of
gravitational forces is not consistent with special relativity which is
based on the postulate that nothing can travel faster than the speed of
light.
Einstein’s new theory required that the laws of physics be expressed
as tensors that are invariant under transformations from one acceler-
ated reference frame to another. Newton’s law of universal gravitation
does not meet this condition. Another property that must be met is
that the laws must reduce to the well known forms of classical physics
in the appropriate limit, just as special relativity reduces to classical
physics when the speeds of the reference frames are much small than
c.
According to Newtonian gravity, the source of the gravitational field
is mass. Recall the Poisson equation states that
∇2 Φ = 4πGρ,
where ρ is the mass per unit volume. Furthermore, in Newtonian me-
chanics, mass is a conserved quantity. But according to relativity the-
ory, mass and energy are interchangeable quantities. The problem is
to replace the Poisson equation with tensors and to find a substitute
for mass on the right hand side. In special relativity it is shown that
the mass-energy-momentum relationship is
E 2 = p 2 c2 + m 2 c4 ,
304 9. THE GRAVITATIONAL FIELD

where p is the momentum. Einstein decided that this quantity could

serve in the role of the mass of Newtonian mechanics. The expression
had the added benefit that it could be written in tensor form as the
Energy-Momentum tensor T µν . This is a 4×4 symmetric tensor which
has 16 components, although only 10 of them are independent. These
components are:
T 00 = local energy density (including mass energy),
T 0i = momentum density in i direction times the speed of light,
T ij = rate of flow per unit area of ith component of momentum
perpendicular to the j direction.
It can be shown that conservation of energy and momentum requires
that X
∇µ T µν = 0.
µ
Examples of the energy-momentum tensor often involve electro-
magnetic fields and are beyond the scope of our study. However, two
simple situations can be noted. For a region of space where the only
constituents are non-interacting particles of mass m we find
T µν = ρU µ U ν
where U is the four-dimensional analog of velocity. For a region of
space filled with an ideal fluid with pressure P we have
T µν = (ρ + P/c2 )U µ U ν − P g µν .
Finally, for a region of empty space (no particles and no electromagnetic
fields)
T µν = 0.
The tensor T µν will replace 4πGρ on the right hand side of Poisson’s
equation. Next we consider what to put on the left-hand side of the
δ
equation. Recall that the rank 4 Riemann tensor Rαβγ describes the
curvature of spacetime. Using a mathematical procedure known as
“contracting” we can define a rank 2 tensor Rαβ that has some of the
properties of the Riemann tensor. Then a third quantity, R, called the
Ricci scalar, can be obtained by evaluating
X
R= g αβ Rαβ .
α,β

Einstein put these together to form the following rank 2 tensor

1
Gµν = Rµν − g µν R,
2
referred to as the “Einstein tensor.” Note that all the terms on the
right hand side are functions of the metric g µν . It turns out that in the
 9.8. EINSTEIN’S THEORY OF GRAVITATION (OPTIONAL) 305

“Newtonian limit” this tensor reduces to the left hand side of Poisson’s
equation. It is, of course, a sign of Einstein’s genius that he realized
this correspondence and expressed the Einstein tensor as proportional
to the energy-momentum tensor. It turns out that the proportionality
constant is K = 8πG/c2 where G is the universal gravitational constant
in Newton’s law of gravity. Consequently,
8πG
Gµν = − 2 T µν . (9.18)
c
This is, essentially, the Einstein expression for the gravitational field. It
is a set of 10 coupled second order differential equations for the metric
g µν and is referred to as “Einstein’s field equations.”13
We still have not answered the question about how a mass particle
will move in curved spacetime. Recall our discussion of the calculus of
variations in Chapter 4 in which we showed that the shortest distance
between two points in a plane is a straight line and the shortest distance
between two points on the surface of a sphere is a section of a great
circle (or geodesic). It is an interesting consequence of the general rela-
tivity that a particle in a gravitational field will move along a geodesic.
At first Einstein thought that this would have to be incorporated into
his theory as a postulate, but some years after he had published his
theory he realized that geodesic motion follows from the condition
X
∇µ T µν = 0.
µ

So geodesic motion is implicitly included in the theory.

The motion of a particle in curved spacetime is often described
in terms of the analogy of a bowling ball placed on a soft mattress,
thus deforming the surface of the mattress. A marble rolling along the
mattress will roll into the depression, as if attracted to the bowling
ball.
John Wheeler summarized the rules of general relativity in the fol-
lowing way:
Matter tells space how to curve.
Space tells matter how to move.
Our discussion has been rather abstract, so it might be helpful to
present a specific metric tensor that was evaluated by Karl Schwarzschild
shortly after Einstein published his theory. Schwarzschild considered
13For the purists, I must admit that Equation 9.18 is a simplified version of
the Einstein Field equations because I have left off the “cosmological constant.”
You may have noted that there are 16 field equations, but only 10 of them are
independent.
306 9. THE GRAVITATIONAL FIELD

Figure 9.10. A bowling ball on a mattress as an analog

to curved spacetime.

the gravitational field in the empty space outside of a star. In empty

space the energy-momentum tensor components are all zero. Solving
the Einstein field equations for the components of the metric tensor,
Schwarzschild found
1 − 2GM 0 0 0
 
c2 r
1
 0 − 1− 2GM 0 0 
gµν =  c2 r .
0 0 −r2 0 
0 0 0 −r2 sin2 θ

The quantity 2GMc2

is a distance. It is called the Schwarzschild radius
(RS ). Note that the g11 element of the matrix becomes infinite at the
Schwarzschild radius. It turns out that if a star shrinks to a radius less
than RS , it becomes a black hole from which not even light can escape.
The analysis of Schwarzschild led to other important consequences,
such as predicting the correction for the precession of planetary orbits
that we discussed previously.
The theory of general relativity is, as we have seen, basically a field
theory of gravitation. It has been submitted to numerous experimental
tests and has been proved correct over and over. It is unfortunate that
it is so mathematically complex that it is only accessible to a small
community of physicists, astronomers, and mathematicians.

9.9. Summary
A field is defined as a physical quantity whose value can be deter-
mined at every point in some given region of space. The gravitational
field g(r) due to a particle of mass M located at point r0 can be defined
in terms of the force it exerts on an infinitesimal point mass m located
 9.9. SUMMARY 307

at r,
r − r0
 
F
g(r)= lim = −GM .
m→0 m |r − r0 |3
The field produced by an extended body can be determined by evalu-
ating the expression
r − r0
Z
g(r) = −G ρ(r0 ) 0 |3
dτ 0 .
body |r − r
To evaluate the integral it is usually necessary to resolve the vectors
into components, using the symmetry of the problem as a guide.
The gravitational potential Φ(r) is given by
ρ(r0 )dτ 0
Z
Φ(r) = −G 0
.
body |r − r |

A field can be represented graphically by drawing the field lines (as

illustrated in Figure 9.7) or by drawing equipotential surfaces.
The Helmholtz theorem tells us that we can determine a vector field
if we know its “sources” defined as the gradient and curl of the field.
For the gravitational field these are:
∇ × g = 0,
∇ · g = −4πGρ.
The first of these relations indicates that the gravitational field is con-
servative, thus allowing us to define a potential. The second equation
is Gauss’s law and in integral form it is expressed as:
I
g · n̂ dS = −4πGMenc
S

where Menc is the mass enclosed by the surface S. In terms of the

potential, this leads to
∇2 Φ = 4πGρ
which is called Poisson’s equation. In regions of space where ρ = 0 this
reduces to Laplace’s equation
∇2 Φ = 0.
Finally, although Newton’s theory of gravitation has been super-
seded by Einstein’s theory of general relativity, we still use Newtonian
mechanics for most practical problems because it is much easier to
apply and the differences are miniscule.
308 9. THE GRAVITATIONAL FIELD

9.10. Problems
Problem 9.1. A neutron star has a mass of 1030 kg and a radius
of 5 kilometers. A body is dropped from a height of 20 cm above the
surface. Determine the speed of the body when it hits the surface.
Problem 9.2. The Earth is suddenly brought to a standstill. Eval-
uate the time required for it to collide with the Sun. You can assume
the Sun does not move and that the center of mass of the system is at
the center of the Sun. How good is this approximation?
Problem 9.3. Determine the period of a surface skimming satellite
in a circular orbit about a uniform, perfectly spherical planet of radius
R and density ρ. Determine the period of such a satellite about a
different planet which has radius 2R but the same density as the first.
Explain your result.
Problem 9.4. Consider an infinitely long, straight string whose
linear mass density is λ (mass per unit length). By direct integration,
determine the gravitational field a distance r from the string.
Problem 9.5. By direct integration, find the field at a point on
the axis of symmetry of a cylinder of length L, radius R and uniform
density ρ. (Hint: Let the mass element be an infinitesimally thin disk
and use the result of Worked Example 9.2.)
Problem 9.6. By direct integration determine the gravitational
field a distance z above an infinite flat surface whose mass per unit
area is σ. Check your result by using the result of Worked Example
9.2,
Problem 9.7. Consider an infinite string with linear mass density
λ. A particle of mass m is a distance d from the string. Determine the
force on the particle.
Problem 9.8. Using the technique of Worked Example 9.3, de-
termine the potential and the field of a spherical shell of mass M and
radius a, at an interior point. Answer: g = 0 and Φ = −GM/a =
constant.
Problem 9.9. Assuming that the mass cylinder of Worked Exam-
ple 9.4 has a constant mass density ρ, determine the field at a point
inside the cylinder.
Problem 9.10. A sphere of radius R and constant mass density
ρ has a spherical cavity of radius r where r = R/2. The center of the
cavity is a distance R/4 from the center of the sphere. A particle of
 9.10. PROBLEMS 309

mass m is located at an outside point, a distance z from the center

of the larger sphere and along the line going through the centers of
the sphere and the cavity. Determine the force on the particle. (Note:
z > R.)
Problem 9.11. Consider a planet of mass M and radius R. Assume
the planet is spherical and has a constant density. By direct integration
determine the gravitational field and the gravitational potential at all
points inside and outside the planet. (Assume the potential is zero at
infinity and there are no other bodies in the universe.)
Problem 9.12. Suppose an intergalactic gas cloud were found to
have a density given by
M b2
ρ= .
2πr(r2 + b2 )2
Here M and b are constants related to the size and mass of the cloud.
Note that the density extends out to infinity, but the density is infinitely
small for very large values of r. Determine the gravitational field and
the gravitational potential as functions of r.
Problem 9.13. (a) Determine the gravitational potential at a point
on the axis of a ring of mass of radius a and mass m. (b) Determine the
potential for this ring at an off-axis point located a distance r from the
center of the ring and making an angle θ with the axis. You may assume
r >> a and you can disregard any terms of order (a/r)3 or smaller.
Hint: the angle γ between two lines whose directions are specified by
θ1 , φ1 and θ2 , φ2 is given by cos γ = cos θ1 cos θ2 + sin θ1 sin θ2 cos(φ1 −
φ2 ).
Problem 9.14. Determine the gravitational potential inside and
outside of a constant density sphere of radius R and mass M. Plot Φ
vs r. (Answer: Inside the sphere Φ = −GM r2 /R3 .)
Problem 9.15. Suppose a tunnel were dug through the Earth
along a diameter. Show that an object dropped into this tunnel would
undergo simple harmonic motion. Determine the period of the motion.
(Air resistance is, of course, neglected.) Answer: 84 minutes.
Problem 9.16. A particle is at an arbitrary point P inside a mass
shell. (a) Draw the line through P so that it passes through the center
of the shell (that is, the line is a diameter). Construct a “double cone”
with vertices at P as shown in Figure 9.11. Show that the resultant
gravitational force on the particle due to the two surface elements A1
and A2 is zero. (b) Generalize to any line through P and show that
310 9. THE GRAVITATIONAL FIELD

there is no net force acting on the particle at any interior point. (This
is the way Newton showed that there was no force on a particle inside
a shell.)

A2
P

r2

r1
A1

Figure 9.11. A spherical mass shell and two cones with

common vertex at P.

Problem 9.17. Use Gauss’s law to determine the field inside and
outside of: (a) A sphere of uniform mass density, (b) A homogeneous
hollow spherical shell. Let the mass and radius be M and R in both
cases.
Problem 9.18. Use Gauss’s law to determine the field inside and
outside an infinite cylinder of radius R and uniform mass density ρ.
Express your answer in terms of r, the distance from the axis.
Problem 9.19. Use Gauss’s law to determine the field above and
below an infinite plane of surface mass density σ.
Problem 9.20. An infinite mass plane gives rise to a constant
gravitational field, and a sphere of mass M gives rise to the gravita-
tional field given by Newton’s law of universal gravitation. Consider
an object that looks like Saturn, but instead of rings, the equatorial
plane of the sphere is a uniform mass plane of infinite extent and mass
density σ (per unit area). Let the radius of the sphere be R. Deter-
mine the work done as a particle of mass m is moved from the “north
pole” of the sphere to a distance h above it (h > R). (The principle
of superposition tells us the field of two mass distributions is just the
vector sum of the field due to each.)
Problem 9.21. The gravitational field in some region of space is
given by g = −kr3 r̂ where k is a constant. What is the mass density
ρ? What is ρ if the field is given by g = −(k/r2 )r̂?
 9.10. PROBLEMS 311

Problem 9.22. A collision between Earth and an asteroid of diam-

eter 200 meters would cause widespread damage and loss of life. Space
scientists are interested in devising ways to deflect an approaching as-
teroid. Not too long ago, two NASA engineers14 suggested placing a
spacecraft near such an asteroid and use the gravitational attraction
between spacecraft and asteroid to tow it away. This “gravitational
tractor” would hover a distance d from the asteroid. Its engines would
be directed as shown in Figure 9.12 so that they would not blast the
surface of the asteroid. The angle φ is the half angle of the rocket
jet; you may assume a value of 20◦ for this angle. (a) Show that the
minimum thrust required to just balance the gravitational attraction
of the asteroid is
GM m/d2
T =
cos[sin−1 (r/d) + φ]
where r is the radius of the asteroid. (b) Evaluate T for a 200 meter
diameter asteroid (r = 100 m), having a density of 2X103 kg/m3 , if a
20 tonne spacecraft hovers at d = 2r. (c) Show that the velocity change
imparted to the asteroid per second is
∆v = Gm/d2 .

Figure 9.12. A “gravitational tractor.” The shaded ar-

eas are the exhaust plumes from the rocket motors which
are tilted away from the asteroid so as not to push on it.
The angle φ is the half angle of the plume.

COMPUTATIONAL PROJECTS
14E. T. Lu and S. G. Love, Gravitational Tractor for Towing Asteroids, Nature,
438, 177-178 (2005).
312 9. THE GRAVITATIONAL FIELD

Computational Technique: Numerical Integration

Two simple numerical integration schemes are called the “Trape-
zoidal Rule” and “Simpson’s Rule.” These are two applications of
a more general method called the “Newton-Cotes Technique.” Both
schemes assume you know the value of the integrand at the end points
and at a number of intermediate points.
The trapezoidal rule assumes the integrand f (x) varies linearly be-
tween the “data” points. Thus, if we know f (x) at xi and at xi+1 ,
then Z xi+1
1
f (x)dx = (xi+1 − xi )(fi+1 + fi )
xi 2
The total integral from initial to final values of x will be the sum of
such terms. (Note that the area under the f (x) curve is broken up into
trapezoids. The results are often not very accurate unless the x values
are very closely spaced.)
Simpson’s rule uses a quadratic fit to a series of three equally spaced
points, separated by equal distances d. Then
Z xi+1
d
f (x)dx = (fi + 4fi+1 + fi+2 )
xi 3
Note that the number of values of x must be odd.

Computational Project 9.1. A sphere of radius 1 meter has a

density that varies as ρ(r) = 5r3 for 0 ≤ r ≤ 1. Obtain the mass of
the sphere analytically and numerically with the trapezoidal rule and
Simpson’s rule for nine equally spaced values of r. Compare results of
the three calculations.
Computational Project 9.2. Write a computer program to de-
termine the position and velocity of a projectile fired vertically from
the Earth’s surface with an initial speed of 3000 m/s. Plot the position
and velocity as a function of time. The Earth is assumed to be a sphere.
You may neglect air resistance, the rotation of the Earth and the effect
of any other astronomical bodies, but note that the acceleration due
to gravity varies with distance from the center of the Earth.
Computational Project 9.3. Plot the potential for two equal
point masses. (To make things easier, assume you are using a system of
units in which the masses are unity, they are separated by unit distance
and the gravitational constant is unity.) You can let the two particles
lie on the x axis.
 9.10. PROBLEMS 313

Computational Project 9.4. Imagine a planet in the shape of a

cube with side 105 m. Draw a Cartesian coordinate system with origin
at the center of the cube and axes passing through the centers of the
faces. Obtain an integral expression for the potential at a point 106 m
from the origin along one of the axes. Write a computer program to
carry out the integration and obtain a numerical answer.
 Part 3

The Mechanics of Particles

Chapter 10
Central Force Motion: The Kepler
Problem

The orbital motion of a planet around the Sun was one of the first
important problems to be analyzed in terms of Newton’s three laws.
The gravitational force attracting a planet to the Sun is a central force.
The motion of a planet is a prime example of the more general problem
of the behavior of a particle acted upon by a central force.
Although we shall be primarily concerned with the motion of plan-
ets and satellites, the techniques you will learn in this chapter are ap-
plicable to any kind of central force. In this chapter, as well as learning
the laws governing the motion of celestial bodies, you will be exposed
to the concept of effective potential energy, and you will appreciate
how constants of the motion are used in solving physics problems.
Historically, the quantitative analysis of orbital motion began with
Kepler’s realization that the motion of planets can be described by
three empirical laws. An important point made in this chapter is that
Kepler’s laws of planetary motion can be derived theoretically from
Newton’s laws of motion. Additionally, Newton’s laws gives us a much
deeper understanding of Kepler’s laws. This application of Newton’s
ideas amazed and fascinated the “natural philosophers” of his era and
was one of the most important events in the history of science.1

1Much of the material in this chapter is an introduction to celestial mechanics.

Many excellent celestial mechanics books are available, including A. E. Roy, Orbital
Motion, Adam Hilger Press, Bristol, 1988, and J. M. A. Danby, Fundamentals of
Celestial Mechanics, William-Bell Inc, Richmond, VA, 1992. A more recent book
that is particularly well written is by C. D. Murray and S. F. Dermott, Solar System
Dynamics, Cambridge University Press, Cambridge,1999.
317
318 10. CENTRAL FORCE MOTION: THE KEPLER PROBLEM

10.1. Johannes Kepler (Optional Historical Note)

Johannes Kepler lived from 1571 to 1630. He was born into a poor
family in a small town in Germany. His father was a professional sol-
dier who spent much of his time away from home and his mother was
a quarrelsome woman who was accused of witchcraft in her old age. It
would be fair to say that Kepler had an unhappy childhood. Neverthe-
less, he was very intelligent and an excellent student. The prince who
ruled the region sent him to study at a Lutheran seminary. Eventu-
ally, Kepler graduated from the University of Tubingen. Originally he
planned to study divinity, but before he was ordained, the authorities
in the seminary convinced him that he was not cut out for the clergy.
With the help of his advisors, he was appointed to a position teaching
mathematics in Graz, Austria. He was a quiet, introspective person
and not greatly interested in teaching; in fact, he was probably a ter-
rible teacher for he had the reputation of interrupting his own lectures
to silently mull over some idea that had just occurred to him.
In Graz, Kepler had an idea that changed the course of his life.
This idea, which became an obsession with him, was (he thought) a
glimpse into the mind of God: A vision of the basic structure of the
universe. It was, so to speak, revealed to him why there are only five
planets, and why they are in their particular orbits around the sun.
Kepler’s mind-boggling inspiration was this: there are only five planets
(besides Earth) because there are only five “perfect” solids, and the
orbits of the planets correspond to spheres circumscribed about the
perfect solids when they are nested, one inside the other.
The “perfect” solids (or “simple polyhedrons”) are the geometrical
figures formed from regular polygons. A regular polygon is one with
equal sides. Thus, for example, a tetrahedron (or equilateral pyra-
mid) is made up of four equilateral triangles. A cube is made up of
six squares. Similarly, an octahedron is an eight sided solid made of
equilateral triangles, an icosahedron is made of twenty equilateral tri-
angles and the dodecahedron is composed of twelve pentagons. Many
other solids can be constructed from polygons; for example a “soccer
ball” shape can be made up of pentagons and hexagons. This is the
structure of Carbon-60, the so-called “Bucky balls,” named in honor
of Buckminster Fuller who studied the properties of such structures.
There are, however, only five “perfect” solids whose faces are a single
type of regular polygon. There is a very neat proof that there can only
be five such solids; this proof can be traced back to the ancient Greeks.
 10.1. JOHANNES KEPLER (OPTIONAL HISTORICAL NOTE) 319

If you are interested, it is reproduced in the book Cosmos, by Carl

Sagan.2
Kepler’s idea was that if the perfect solids were placed one inside
the other, and each was circumscribed with a sphere, then the sun
would be at the center of the system and each of the five planets would
orbit the sun in a circular orbit whose radius would be equal to the
radius of the corresponding circumscribed sphere. Kepler built models
of these nested solids and their circumscribed spheres, but he could not
prove his theory because he did not have enough information on the
distances of the five (known) planets to the Sun. Figure 10.1 illustrates
the idea behind Kepler’s model.

Figure 10.1. A nested tetrahedron and cube with in-

scribed and circumscribed spheres. In Kepler’s model the
nested perfect solids (from innermost outwards) were an
octahedron, an icosahedron, a dodecahedron, a tetrahe-
dron and a cube.

At that time, the best astronomical data in Europe were in the

observatory of an eccentric Danish astronomer named Tycho Brahe.3
2Carl Sagan, Cosmos, Random House, New York, 1980 (Appendix 2).
3Tycho Brahe (1546-1601) set up an observatory at Uraniborg on the Danish
island of Hven. A stream of young assistants from all over Europe came to Urani-
borg where they carried out experiments in chemistry during the day and observed
the heavens at night. Uraniborg has been described as the first research institution
involved in “big science.” For the full story read the book by John Robert Christian-
son, On Tycho’s Island: Tycho Brahe and his Assistants, 1570-1601, Cambridge
University Press, Cambridge, 2000. At the time Kepler went to visit him, Brahe
had moved to Prague.
320 10. CENTRAL FORCE MOTION: THE KEPLER PROBLEM

Kepler went to visit Brahe to get his data. He was certain that Brahe
would be overwhelmed by his wonderful new theory. Brahe was not
overwhelmed. In fact, at first B