Chapter 0 Introduction

Section 0.1: Why Calculus?

After one young moviestar’s first day at college, she wrote a friend: Time(s) Voltage(mv)
’My Calculus book is three inches thick. I can’t live through three .001 22
inches of calculus.’ .002 25
.003 31
Not a lot of mediastars in my class, but some students wonder: ’I’m .... ...
going to be a doctor. Why do I need calculus?’ Others ask, ’I’m .... ...
12.602 -4
gonna be an engineer: we use computers for everything. Why do I
Table 1: Raw Data
need this class?" We’ll start with the doctor, then the engineer, and Data, measured in volts, straight from
then get them working together. our secret experiment.

Table 1 is data, straight from a recording of . . . well, let’s wait on

that for now. It’s the stuff of modern science: raw numbers. 12,602
mind-numbing numbers, more data than the human mind can ’get’
without some kind of help.
Even people who don’t like graphs might like Figure 1. Those 12,000+
data points were graphed, and now we see it’s a recording of a heart-
beat. The change from ’ick’ to ’oh, I get it’ happens because humans
are good at understanding pictures, so a graph helps us ’get’ what Figure 1: Old-Style EKG Recording
The graph of the data in Table 1. Even
numbers mean. As Stephen Few, in the book Now You See It, remarks today, you can go to your doctor and
get a strip recording just like this. I
’When we represent information in visual form, our ability to think about hope you’re not having what this guy is
it is dramatically enhanced. Visual representations extend the capacity having: diagnosis, Doctor?

of our memory, making available what we couldn’t otherwise hold all at

once in our minds. Visualization helps us to think.’

Figure 2: Fun With Graphs

Back to the doctor: the diagnosis. The patient is suffering from fib-
What we can do with graphs:
rillation, that is, an erratic heart beat, and you can tell because the i) Compare
distance between peaks varies dramatically. In Figure 1 each little ii) Zoom in: focus; zoom out: overview
ii) Discover patterns and trends.
box is .04 sec wide, so what you’d do – if you were an old-fashioned
MD – is count squares between the peaks, which would tell you how
many beats per minute you get. Fibrillation would show up as a
wildly varying number of beats per minute.
2 kathy davis

That’s what you do if you’re a doctor in the 1920’s. Here’s the origi-
nal set-up, in the picture below. The technique of EKG recording was
invented by a Dutch MD, Willem Einthoven, who received the 1924
Nobel Prize in Medicine or Physiology for his work.

Fortunately, you’re not in the 1920’s. Today you can buy a little cell-
phone style EKG unit, like the one in Figure 3. But besides the size,
and not-sticking your arm in salt water, something else happened.
Instead of drawing the graph on paper with little squares, and count-
ing squares, a microchip draws the graph on a LCD screen . . . and it
counts the squares, and it computes the beats-per-minute.
Figure 3: Modern EKG Recording
So . . . so say you’re the engineer programming the chip. You’re
EKG monitor about the size of a 2012-
working with the MD who also didn’t need calculus, and before era smartphone. Quantities like beats-
you can get the chip to count squares, it has to somehow ’see’ the per-minute are calculated by a mi-
crochip in the unit.
peaks. How does that happen? The first engineers to write software
to recognize heartbeats with a microchip were Jiapu Pan, with the
Department of Biophysics, Shanghai Second Medical College, and
Willis Tompkins, Department of Electrical and Computer Engineer-
ing, University of Wisconsin – our MD and our Engineer. They used
a technique called a derivative-based algorithm. Derivatives are calcu-
lus.
 calculus sin frontera 3

Pan-Tompkins was programmable on 1974 chips and it could reach

98% accuracy on hospital patients. Now let’s bring it up to today. In
early 2005 the American military began a project to recognize heart-
beats of soldiers in a warzone (actually you’re supposed to call them
’warfighters’ not soldiers). If we want to start with the Pan-Tompkins
technique, we have to know it likes nice, regular heartbeats:

The Army needed to deal with heartbeats of soldiers who might be

running, scared, dehydrated. More like

For these, Pan-Tompkins has only a 25% accuracy! To deal with these
much more complex EKGs, the Army brought together a group: one
physiologist, from a famous medical school; one hot-shot program-
mer to write the software, two experienced electrical engineers to
build the recording devices and design the chip, and one out-of-place
mathematician.
Four people, with different backgrounds. The language they all
spoke was math, starting with calculus. All the team understood
the math, discussed it, argued about it: a strong understanding of
math was required for the job. MD, programmer, engineers and
mathematician.
In this story, the moral is that people who want to do serious work,
even if they’re "only" engineers or doctors, might want that extra
mathematics background. Now a different story.
4 kathy davis

In 2009, Professor Ivo Babuska, Figure 6, along with Professor Barna

Szabo, a mechanics professor emeritus at Washington University and
Juhani Pitkaranta, a mathematics professor emeritus at Aalto Univer-
sity, issued a challenge to professional engineers: use professional-
level engineering software to determine the forces the walls of the
Sydney Opera House exert. The answers had to be accurate to 5%.
Of the fifteen replies, half were wrong: the error was more than the
Figure 4: Modern Architecture
required 5%. In practical terms, the walls might collapse under un-
The famous Sydney Opera House,
expected extra forces: a moderate earthquake, very strong winds. A the defining structure of the modern
collapse might kill many people. city. Formally opened by the Queen of
England.
Some of the submitted answers were off by as much as 50%. This
could lead to almost immediate collapse of the building during con-
struction, killing many workers. Babuska summarized the results like
this:

’How is it possible that this happened is a good question. There could be

many various reasons. Nevertheless, in this case the reason was only one.
Some of the analysts did not have sufficient engineering intuition and
mathematical and engineering knowledge and possibly used the software
incorrectly.’
Figure 5: Modern Engineering
It’s called the Girkmann Problem: the
This story has a moral too. Babuska describes the issue as ’signing walls of the Opera House, Figure 4,
the blueprints.’ In any significant engineering operation there has to are depicted by the curved line; they’re
held up by the small blocks shaded in
be a person who signs the blueprints: that is, a person who takes le- grey here. The problem is to find the
gal responsibility for the design, safety and cost. If there is a problem forces the wall exerts on the block.
as small as a cost-overrun, or as large as a structural failure causing
the loss of human life, the person who signs the blueprints is legally
and financially responsible.
The signature on the blueprints has to be a human, not computer.
In the world of engineering and construction, there’s no ’we use
computers for everything so I don’t have to understand it.’ The alter-
native to understanding could be bankruptcy or even prison.

Figure 6: A Modern Engineer

Professor Ivo Babuska, Robert B. Trull
Chair in Engineering at the University
of Texas at Austin.
 calculus sin frontera 5

Section 0.2: What Calculus is About

The word calculus derives from the same stem as the word calculate.
Interesting, but . . . what is calculus calculating?
Calculus was born from a Renaissance meeting of world cultures,
when Western thinkers learned of Hindu-Persian arithmetic and tech-
niques of computing; of Arabic algebra for working with variables.
Another influence was Descartes’ invention of analytic geometry,
translating Greek geometry into algebra. This confluence marked the Figure 7: Rennaisance Physics
Cannonballs always hit you on the top
age of computation.
of your head.
Here’s a concrete example, from mid 1500’s Italy. Gunpowder had
recently been brought from China, and Italian city-states found it
useful in resolving trade and territory disputes. Build cannon, apply
gunpowder, end dispute.
Figure 7: shows Renaissance physics describing the motion of a can-
nonball. According to Aristotle, close to the surface of the earth,
objects move in a straight line. When they’re high enough, though,
they move under heavenly laws: in perfect circles. So the cannonball
turns down to the earth and back to moving in a straight line again.
This was the dominant model of physical motion at the beginning of
the Renaissance, and it is wrong. You only have to follow the path of
water from a fountain as in Figure 8 to see Figure 7 is wrong.
In contrast, Figure 9 is a page from the lab books of Galileo. He knew
Greek geometry, he knew algebra and if you look very closely at Fig-
ure 9, you can even read Hindu-Arabic numerals. He used algebra, Figure 8: Renaissance Fountain
geometry, and he computed. He described his work like this: The fountain at Villa Lante, mid 1500’s.
Compare the path of the water jet with
’Philosophy is written in this grand book, the universe, which stands that of the cannonball in Figure 7.

continually open to our gaze. But the book cannot be understood unless
one first learns to comprehend the language and read the characters in
which it is written. It is written in the language of mathematics, and
its characters are triangles, circles, and other geometric figures without
which it is humanly impossible to understand a single word of it; without
these one is wandering in a dark labyrinth.’

Both Aristotle’s and Galileo’s description of the path of a cannonball

are applications of geometry and algebra to problems of motion. All
applications of calculus are examples of mathematical modeling, as Figure 9: Galileo’s Notes
was the discussion of heartbeats in the previous section. We’ll talk Galileo illustrates the path taken by a
about many different kinds of mathematical models, beginning next falling body.

section.
6 kathy davis

Section 0.3: Models and Mathematics

When we looked at Aristotle and Galileo in Section 0.2, we were

looking at models. Let’s talk about that, starting with the kinds of
models most everyone knows.
Figure 10 shows a model of the world historical site, the Golden
Temple in Kyoto; Figure 11 shows a a view of the temple itself. Some
differences: the model is sitting on a desk, not on a lake with forest in
the background, and it isn’t covered in real gold.
Why build a model? Partly aesthetic reasons: you can see beauty in
the model itself; you an admire the craftsmanship, you can enjoy the
model when you’re far from Kyoto.
Figure 10: Model
But a model could also be useful, to architects. They could explore
A model of Rokuon-ji Golden Pavilion
space and proportion, whether the effect will be calming or majestic. Temple in Kyoto.
With a lamp to mimic the sun, they could determine when court-
yards will be shady or gardens sunny. They could use the model as
a tool, to help design buildings and to make predictions about real
temples.
Check out Figure 12: two more models of buildings. We have a sus-
picion that these models aren’t quite as good as the one in Figure 10;
we have a sense that some models are better than others.
This sense are related to the idea of abstraction; the word derives from
an old meaning: to take something away (abstract, n. Classical Latin:
abstrahere to drag away. In post-classical Latin also to summarize).
In modeling a real object, you eliminate the parts you don’t need to
Figure 11: Reality
use; you keep the parts that are important to you. In the example
Photo of the actual temple in Kyoto.
of the temple, we kept the design and relative proportions; we left The temple, lake and gardens were
out information about the weight of the wood and stone, about the built by a Buddhist Pure Land sect, and
the golden pavillion itself meant to be
strength of the beams and joints. So: the very physical model that we viewed from a distance, across the lake.
can pick up, turn in our hands, that model is an abstraction!!
Because it is an abstraction, the model doesn’t answer questions
about resistance to typhoons, or how the building will sink into the
ground. It has the limitations we built in when we threw away in-
formation about weight and strength. The model reflects both our
understandings and our lack of understandings.
There’s another point, though: if we knew the weights of the wood,
stones, pillars, we could compute what kind of foundation the tem-
ple would need, so as to not sink. This is important: a good model
allows modifications, allows the exploration of new questions.
Figure 12: A Different Model
Models of buildings, built by kids.
Technical competence, the quality of
materials and tools and craft, count too.
 calculus sin frontera 7

Let’s revisit the cannon models in Section 0.2, thinking about abstrac- Figure 13: Abstraction:
Not Just For Breakfast
tion. First, Aristotle. His model was geometric; it tells the shape of
"An abstraction is a general model
the path of a cannonball: the path is composed of straight line pieces of something. It is a definition that
and circle pieces. It won’t tell you where the ball will land, or how includes only the general characteristics
hard it will hit. The only real information you get from the model is of an object without the details that
characterize the specific instances of the
that a cannonball will go up in the air and then go down again. Aris- object."
totle developed this model to explain not only how motion happens Gaddis et. al. C++ Early Objects
but why; it was part of a philosophy of all change. For the cannon-
ball, the model is too abstract to give detailed information.
Galileo’s model is both geometric and numerical: it tells us the path
of the ball is a parabola, but is very specific on how to get the exact
shape of the parabola: you can use it to discover where the ball hits
and how fast and how long it will take to get there. People of the
time even used it to find the best angle to shoot the cannon to make
the ball would travel furthest.
Back to our temple model. Yoko Kawaguchi, in Japanese Zen Gardens,
writes: "From the tenth century on, this longing for paradise had
been expressed in the form of temple gardens designed in the image
designed in the image of Amida Buddha’s paradise." That is to say, Figure 14: A Different Temple
The great Temple to Yahweh in
the actual Golden Pavilion is meant to be a model of a spiritual par-
Jerusalem, before its destruction by
adise. This is a common motif throughout across the world. Compare the Roman armies.
Exodus 25; Yahweh is instructing Moses: "Let them make me a sanc-
tuary that I may dwell among them. According to all that I show thee
... shall you make it"
Mircea Eliade, in History of Religious Ideas, remarks that this motif
f the temple as a reflection of the eternal temples of the gods long
pre-dates Platonic ideas that the world is a mere shadow of the world
of eternal forms. There’s a good deal of scholarly debate whether
Galileo was a Neoplatonist, but it certainly would help explain his
confidence in writing

’. . . the universe . . . is written in the language of mathematics, and its

characters are triangles, circles, and other geometric figures . . . without
these one is wandering in a dark labyrinth.’

Figure 15: A Different World

The ’characters’ Galileo mentions are all taken from the ideal forms
Half a world away, a completely dif-
of classical Greek geometry and philosophy. Galileo’s certainty may ferent concept of temple, yet with the
well have been that the world in which we live is a mere reflection of same underlying idea: the building
reflects the design of god.
the ideal world which god creates.
A last thought: our temple model was a physical object; the temple
itself a model of paradise; Aristotle’s model was a philosophical
system, and Galileo’s a set of equations. Quite a range of concepts,
covered under ’model.’
 calculus sin frontera 1

Section 2.13: The Calculus of Force

We all have an intuitive sense of force, even from early childhood:

push a toy, it moves away; pull, it moves towards us. Children get
this very early. Mathematicians and physicists, not so much.
Right away, we have two kinds of motion: towards and away; we
also have two kinds of force: pull and push. It seems unlikely in the
that we could make this muscle-sense we have into mathematics, yet
humans were thinking about issues like this – well, probably before
written history. We do have a record of these ideas from the ancient
Mideast, Mesopotamia around 1000 BCE, in Figure 1. The idea was,
you record when and where the moon, the planets, the constellations
rose and set, and keep records, then you can predict when and where
they’ll be in the future.
There’s the first key idea: where and when, position and time. Our
toy moves away: it changes position. But instead of a toy, we’re going Figure 1: Early Astronomy
to think about Figure 2: a Karateka (kah-rah-tay-kah: karate practi- Mesopotamian tablet (MUL.APIN)
recording star positions. A typical line
tioner) using his hand to break through a board. A whomping thick
from the tablet would read, "The Steer
board. We’ll record the height of his hand, above the board, over of Heaven GU4.AN.NA [Taurus] rises
time. And we’ll think of towards and away like this: the board is and SHU.PA [Arcturus] sets".

gonna be at height zero. If the hand is above the board, we’ll record
the height of the hand as a positive length. Once the hand breaks
through the board, it’s below the board, and we’ll record that height
as a negative number.
So that’s not bad: we’ve got position, time, positive, negative. Begin-
ning in the Renaissance, these recordings get displayed as a graph, as
in Figure 3.
What you see is height plotted on the vertical axis, and time plot-
ted on the horizontal axis. As you move to the right, time goes on,
and the height decreases. Surprise: the hand is heading towards the Figure 2: Karateka Breaking Board
How do they do that? Does it hurt?
board. Eventually, the hand hits the board; the height is zero, and
you can see the curve wiggling around zero. The board bends and
moves; the hand gets very weird, and all this bouncing results in the
board breaking and the hand moving though.
What you see in Figure 3 is exactly what you’d see, if you had very
fast eyes, and you were watching the Karateka. Nothing new, just
what you see: not helping.

Figure 3: Karate: Position

Remember the hand starts above the
board, and moves downward.
2 kathy davis

The next stage is also very old: star charts like MUL.APIN were used
to predict future positions of stars and planets – hence, events like
eclipses. The ancient astronomers noticed that at one time of year, it
might be 20 days between the rising and setting of a constellation; at
another, 15 days. Rising and setting is a change in position, ∆x; the Figure 4: Karate: Position
Compare this with Figure 8 below.
time of rising to time of setting is a change in time, ∆t. This was, at
least the beginning of, speed of motion: ∆x/∆t
We notice this too, with our toys. We push: the toy wasn’t moving
but now it is: we’ve given the toy a speed. Again, we’ll use positive
and negative numbers: motion away from us has positive speed;
motion towards us has negative speed. But, there’s a better way to
say it: if something is moving away from us, its position from us is
Figure 5: Karate: Velocity
increasing; if it moves towards us, its position is deceasing. So the
Compare this with Figure 4 above.
rule is:
increasing ↔ positive velocity,
decreasing ↔ negative velocity.
Figure 8 shows velocity for our Karate-guy; compare it with position
in Figure 4. The hand starts by moving down: that’s a decreasing
height, so in Figure 8, the velocity starts as negative. Next, the hand
bounces around on the board, and the board bounces, and so the
speeds wobble about zero.
There’s another way to think about the velocity, though: it’s the slope
of the position graph. In Figure 6, I’ve glued the position and velocity
graphs together, and I’ve marked three places in red. Figure 6: Velocity as Slope
On the top left, the red line shows that the position is almost a The velocity in the bottom graph is just
the slope of the top graph.
straight line moving down. The bottom left graph shows this as a
slope: constant, and negative. In the top middle, the hand moves
above and below the board, with an average position of zero, echoed
on the bottom by an average zero slope. Finally, on the top right,
the hand has broken the board, and again you see the hand moving
downward; the position is almost exactly a straight line, with con-
stant negative slope. And again you see that constant negative slope
as a fixed negative line on the bottom right.
 calculus sin frontera 3

This was about as far as one could get, before Newton. A group of
philosopher-scientists such as Descartes tried to explain all physics
as we are explaining it: forces cause motion, by one object pushing
on another. But, try this: pick up a toy and drop it. It falls; I don’t
need to throw it down, or push it, or anything. By the way, if you’ve
ever watched a child sitting in a shopping cart, happily dropping
oranges out of the cart, you’re watching the little bu-bu try to figure
out exactly the same thing: how come it falls when I let go? It’s OK,
little one: Descartes couldn’t figure it out, either.
Figure 7: Mechanistic Universe
And then came Newton. Etching by Albrect Durer, showing the
mechanistic universe, with the planets
First idea: If an object experiences no force, it will move in a straight and stars on rotating wheels.
line, with constant velocity.
Second idea: If an object moves away from a straight line, or changes
velocity, it is experiencing a force.
Third idea: The force an object experiences can be measured: it is
proportional to the change in velocity (or, in geometric language, the
slope of the velocity curve).
So what you had was like a force detector. This annoyed some
philosopher-scientists a great deal, because Newton didn’t describe
why things moved; he only said how they moved. Not the Philosophi-
cal Way! Aristotle told us planets moved in circles because circles are
the perfect object. Newton said that if a planet is moving in a circle,
there must be a force shaping it into that circle.
Worse, forces were now these invisible – things – we can’t see them
with our eyes, only their effects. Newton was describing forces as
some kind of ghostly essence that made things move. Do the plan-
ets move in a circle because angels push them that way, like a child
pushing a toy? Why did it happen?
Newton, and Galileo before him, marked the end of the era of sci-
entist as philosopher. From that time on, it was clear that to really
understand the world, you needed to observe it, and compute with it.
And leave the ’why’ to the Supreme Being.
4 kathy davis

Following Newton, let’s look at our Karateka. Figure 8 shows the

velocity; Figure 9 shows the force as the slope of the velocity curve.
Once again, we’ve glued the two together in Figure 10. The constant
velocity curves used to be constant negative curves, but their slopes
are up at zero. The force occurs exactly when the velocity jumps from
Figure 8: Karate: Velocity
about -12 meters/second to almost zero as the hand hits the board Compare this with Figure 9 below.
and is stopped cold. I used a little computer tool to measure the rise
over the run; the ratio was about 3.3. Divide that by the 10−3 seconds
the graph paper uses to measure time, and you get your force of
about 3300 meters/second2 , close to what’s shown in the graph.
Voila, viva Newton, etc, blah, blah.
Remember the toy we started with? Push one way, the toy moves
away. The speed goes up from zero, so the speed is increasing. The Figure 9: Karate: Force
Compare this with Figure 8 above.
derivative of the speed is positive, and we have a positive force. If
we used our hand to slow the toy down, we’d be exerting a negative
force:
velocity increases ↔ positive force,
velocity decreaes ↔ negative force.
We now know a bit more than just this: now that we know how
much force the Karateka uses on the board„ we can do other com-
putations. For example, experiments show that a bone like those in
the hand can withstand forces of up to 25, 000 meters/second2 , so
breaking the board is a piece of cake.
Knowing the forces empowers me; I can do things, know things,
Figure 10: Force as Slope
detect things. Predict. The force in the bottom graph is just the
slope of the top graph.
 calculus sin frontera 1

Section 2.15: Modeling I: The Differential Equations

We talked about forces and velocities, and their relation to position

and derivatives; now we’ll link these, to understand a simple prob-
lem: check out Figure 1, showing a boxy-looking object attached to a
spring. We’ll move the box then let go: the spring will make the box Figure 1: Equilibrium
move. What happens? A weight attached to a spring, not
stretched, not compressed.
Now, we’re starting the box with no velocity and we expect it to
move, that is, it’ll have some velocity. Remember Newton said that
change in velocity is caused by force; to figure out how our box
moves, we have to figure out the force. The first person to give a
modern version of spring forces was Robert Hooke: he did exten-
sive experiments for many years (see Figure 2), and in 1679 wrote
ut tensio, sic vis: the force is as the tension, that is, the amount of
force the spring exerts is proportional to the amount the spring is
stretched. He was also aware that his work applied to any kind of
material being compressed or stretched: he wrote . . . it is observable
in metal, wood, stones baked earth, hair, horns, silk, bones, sinews, glass . . .
. Hooke was writing when bow and arrows were still used; he knew
his spring theory applied. He knew his theory applied to piling one
stone block ontop of another: the design of buildings, bridges. He
included ’bone, sinew:’ he knew his theory applied to the body. A
modern biomedical engineer would add ’blood vessels’ and ’mus- Figure 2: Springs
cle’ but also things Hooke could never have dreamed of, like the hip Robert Hooke’s experiments on springs.
replacement in Figure 3).
Thi isn’t just about springs; it’s about metals and bricks and bones
and all the materials of the world.
But we’ll just do springs, and we’ll think through Hooke’s Law. We’ll
think about force, and velocity, and position, and how to relate the
three of those. Let’s start. In Figure 1, the spring isn’t stretched or
compressed: it’s neutral. We say the spring-mass combination is at
equilibrium.
What we want to know is, if we move the box somewhere away from
equilibrium, then let go, how will the box move? We say ’the box will
move’ which means we see the box in different places at different
times. So we need notation for places and times. We’ll let t denote
time, measured from when we let the box go. We’ll let x (t) be posi-
tion, measured from . . . equilibrium, why not? . We want to find a
formula for x (t).
Figure 3: Hip Replacement
Metal hip replacement: what are the
forces, as the patient walks?
2 kathy davis

We’ll start by talking about positions.

In Figure 3, equilibrium, the spring isn’t stretched or compressed wnen x = 0.

In Figure 4, we push the box left, compressing the spring. Now, x < 0.
In Figure 5, the box is moved right, stretching the spring. This makes x > 0.

Now a little thought about forces.

Figure 6: The compressed spring pushes; the force is in the direction of the right arrow.
Figure 7: The stretched spring pulls; the force is in the direction of the left arrow.
 calculus sin frontera 3

The words ’push, pull’, are force words. We mentioned that force
cause changes in velocity; more precisely, force is proportional to
the derivative of velocity. So: let’s go back, think about velocity and
changes in velocity.

Figure 8: Equilibrium, no motion, no velocity, no force.

Figure 9: Compression, move right, positive velocity, positive force.
Figure 10: Stretching, move left, negative velocity, negative force.

We can summarize:

 negative if x > 0
force = positive if x < 0

zero if x = 0
But we can actually say more: if you’ve ever used a rubber band to
launch a paperclip or a wad of paper, you know that the more you
stretch the rubber band, the faster longer higher your projectile goes.
The technical way to say this is that force is proportional to − g( x );
here g has to get bigger as x gets bigger (or, to sound less silly: g is
an increasing function). Also, seeing as the force is zero when x = 0,
you’d have to have g(0) = 0.
There are many functions like this, but the simplest is a straight line.
This is pretty simple: for one thing, if you pull the rubber band out
five feet, it breaks instead of giving a huge force. So the line is really
only an approximation – actually, a tangent line approximation – and
it’s only good for small x.
4 kathy davis

This tangent line approximation is a theme in Calculus, so I’ll spend

a second on it. The tangent line approximation to g at zero gives us
g( x ) ≈ f (0) + g0 (0)( x − 0).
We already said the force of an unstretched, uncompressed spring
is zero, so we need g(0) = 0. We also said the more you stretch or
compress, the greater the force is, so g has to be increasing: we need
g0 (0) > 0. Together this gives us g( x ) ≈ g(0) + g0 (0)( x − 0) = cx
where c > 0.
We’re almost done, except. Except, I am going to get really tired of
carrying around that ’where c > 0’ where ever I go. Mathematicians
have a trick for that: we write c = k2 , because everyone is supposed
to know that squares aren’t negative. So: g( x ) ≈ k2 x.
Now, to put it all together:
i) force is the negative of some increasing function, g
ii) g ≈ k2 x
iii) force ≈ −k2 x
iv) force, according to Newton, is mass m times acceleration, d2 x/dt2
v) and finally,
d2 x
m 2 = − k2 x
dt
It probably is a good idea to say something about k. Remember that
the force f is proportional to −k2 x. So a small k means the force is
small; a large k means the force is big. And by ’the force’, here. we
mean the force the spring uses to push or pull the box. The constant
k2 is called the spring constant; it tells you how hard the spring pulls.
Actually k2 is a very physical quantity you can feel: k2 measures the
stiffness of the spring: whether it’s easy to stretch it, or hard to stretch
it.
Nice equation. How ’bout getting around to solving it? Next Section.
 calculus sin frontera 1

Section 2.16: Modeling III: The Solutions

Remember from Section 2.15: we were trying to understand the mo-

tion of a box pulled by a spring. Figure 1 shows the box at equilib-
rium: the spring isn’t pushing or pulling it. The position of the box
at time t was called x (t). We also set up our measuring system so
Figure 1: Equilibrium
that equilibrium corresponds to x = 0. When the box moves right,
A weight attached to a spring, not
x (t) > 0; if it moves left, x (t) < 0. stretched, not compressed.

Say you do move the box then let it go and do its spring thing. We
did a lot of work figuring out an equation to describe how the box
would move after you let go: if m is the mass of the box, and k is a
constant describing how hard the spring pushes or pulls, then

d2 x d2 x
m = − k2 x ↔ m + k2 x = 0
dt2 dt2

What we’re going to see is that whenever we have an equation like

d2 x dx
a 2
+ b + cx = 0
dt dt

then we should try for solutions x = ert . So try it now:

dx d2 x
x = ert ; = rert ; = r2 ert
dt dt2
Now the equation
d2 x dx
a 2
+ b + cx = 0
dt dt
becomes
a · r2 ert + b · rert + c · ert = 0

If we factor out the ert , we get a quadratic equation for r, which be-
comes ar2 + br + c = 0. Let’s try it with

d2 x
m + k2 x = 0
dt2

Then a = m; b = 0; c = k2 . The equation is mr2 + k2 = 0, or,

√
r = ± −k2 /m. I can take the negative outside the square√ root using

i, the square root of negative one. This gives me r = ± k2 /m i.
2 kathy davis

Two solutions? Solutions that are complex numbers? Actually the

complex bit is not an issue: I get 21 eit + e−it = cos(t), while
 
1 −it = sin( t ). Fine: no more complex numbers, but . . .
 it 
2i e − e
why do we get to add solutions ?
These equations are called linear homogeneous equations, and they have
a special property: if s1 (t), s2 (t) are two solutions (as we have here),
then for any constants c1 , c2 c1 s1 (t) + c2 s2 (t) is also a solution. We’re Figure 2: Cosine
thinking here of c1 = + 2i1 , c2 = − 2i1 , which gave me the sine. Try it: y = cos(2πx ), graphed for one period.

d2 d
a 2
[c1 s1 (t) + c2 s2 (t)] + b [c1 s1 (t) + c2 s2 (t)] + c [c1 s1 (t) + c2 s2 (t)]
dt dt

d2 s d2 s
   
ds ds
= c1 a 21 + c1 b 1 + c1 cs1 + c2 a 22 + c2 b 2 + c2 cs2
dt dt dt dt
 2   2 
d s ds d s ds
= c1 a 21 + b 1 + cs1 + c2 a 22 + b 2 + cs2
dt dt dt dt
= c1 [0] + c2 [0] = 0
Figure 3: Sine
The last equation holds because s1 (t), s2 (t) are two solutions, so they y = sin(2πx ), also for one period.
give zero when you put them in the equation.
We’re not quite done: so sine and cosine are solutions? Then, by
what we just said, so are A cos + B sin. Or, more properly,
r ! r !
k2 k2
x (t) = A cos t + B sin t
m m

give me all the solutions of my equation. Whew!

Figure 4: Hperbolic Cosine

If the complex
√ numbers
 weren’t there, I would have combined just y = cosh( x ). It doesn’t repeat.
k2 /m without the i to get, for example, 12 et + e−t =
 
r = ±
cosh(t), while 2i1 et − e−t = sinh(t). But the cosine and cosh are
 

very different functions: I’ve graphed cosine and sine in Figures 2

and 3, and hyperbolic cosine, hyperbolic sine in Figures 4 and 5.
What I want to focus on here is that cosine and sine both oscillate
while the hyperbolic functions both go to infinity. While I don’t
know from oscillate, I do have a pretty good idea that if you pull
on a spring, the box doesn’t fly off to infinity. For one thing, the
spring would break. So the complex numbers prevent that, keeping
Figure 5: Hperbolic Sine
the solutions finite. And causing "oscillation’. y = sinh( x ). It doesn’t repeat either.
 calculus sin frontera 3

What is this ’oscillation’? Someone who oscillates goes back and

forth, never settling on one thing. Sine and cosine do the same:
cos(0) = 1; cos(π ) = −1; cos(2π ) = 1; cos(3π ) = −1, etc.
That’s oscillation alright, but why it it happening here with my box-
and-spring? We’re gonna follow it through, thinking about what
happens, and we’ll be looking at Figures 6, 7, and 8. The top row
shows us the position of the box as it moves. A red dot shows that
there’s no force on the box; a red arrow shows the direction of any
force acting. The second row shows where the box is, using the graph
of x (t) = cos(2πt). A blue dot locates the position of the blue box on
the cosine curve.

So, we start at time zero, Figure 6: t0 = 0. Then the position is

x (t0 ) = cos(2π0) = 1, since we dragged the box one unit right.
The velocity is x 0 (t0 ) = −2π sin(2π0) = 0, since we said we just let
go of he box and didn’t push it or pull it in any direction. Finally, the
force acting on the box is x 0 (t0 ) = −4π 2 cos(2π0) = −4π 2 . The force
is negative, which means the spring is pulling the box back towards
the origin.
Now we move up a quarter unit of time, Figure 7: t1 = 14 . Then
the position is x (t1 ) = cos(2π 14 ) = 0, the box has moved left and
arrived at the origin. The velocity is x 0 (t1 ) = −2π sin(2π 14 ) = −2π.
This means the box is still moving left; which is certainly correct;
there’s nothing stopping the box. Finally, the force acting on the box
is x 0 (t1 ) = −4π 2 cos(2π 41 ) = 0. The force is zero; as we said: we’re
here at equilibrium, and there’s no force acting to stop the box from
moving. It continues on moving left.
 calculus sin frontera 5

Moving up another quarter unit of time, Figure 8: t2 = 12 . Position is

x (t2 ) = cos(2π 12 ) = −1, the box has moved all the way to the left of
the origin. The velocity is x 0 (t2 ) = −2π sin(2π 12 ) = −0. Hey! Where
did the velocity go? It went into compressing the spring, and the box
stops when it has given up all its velocity to the spring. Finally, the
force acting on the box is x 0 (t2 ) = −4π 2 cos(2π 12 ) = 4π 2 . The force
is positive; because the spring is compressed and it’s pushing back,
pushing the box right.
And the cycle continues, pushing out, pulling in. Oscillating. This
is why I must have sine and cosines in the solution, and so complex
numbers for r.
Can I understand why there have to be two solutions? Yes. One way
to figure out what everything is doing in the equation is to isolate
out one piece at a time. What we’re doing is called analysis, breaking
up a problem into small parts, to understand what each part does (cf
Figure 14).
q
2
Let’s make life very simple by taking km = 2π; later, after we’ve
q
2
sorted out A, B we’ll put km back in the equation and sort it out.
We’ll figure out the roles of A, B first taking A = 1, B = 0. Then
x (t) = cos(2πt); we know this. It just tells me that I started the box
to the right of equilibrium, and let it go without pushing or pulling it
any further. Been there.
It isn’t too hard to guess that if I’d taken A = 2, B = 0, that would
mean stretching the spring twice as far before letting it go. If I’d
taken A = −1, B = 0, that would mean that I’d pushed the spring
6 kathy davis

left, compressing it, instead of stretching it by pulling it right.

So actually the solutions to the equations gave me a lot more than I
put in! What I ’put in’ are called initial conditions: where did I start?
Did pull right or push left? The constants A, B are determined by the
initial conditions.
What about B?
We’ll eliminate A this time, and just examine B, by taking A = 0, B =
Figure 9: The A = −1, B = 0 Solution
1. This time x (t) = sin(2πt); we know sin(0) = 0, which tells me that
If I push left instead or puling right,
x (t0 ) = 0 and our little box starts at zero. We kind of knew that A I get a negative cosine function, so
controlled where the box started, so it’s not much of a surprise that x (t0 ) = −1 is left of equilibrium.
A = 0 makes the box start at zero. What’s weird is that zero is the
equilibrium point: if I start at zero, there’s no force acting to move
me away from zero! I shouldn’t be moving at all.
This isn’t helping, is it?
This is the tricky part: B isn’t about position; B is about velocity.
d
If x (t) = sin(2πt), then the velocity v is v = dt (sin(t)) = 2π cos(2πt).
The velocity at t0 is 2π cos(0) = 2π > 0, a positive velocity. This
Figure 10: The A = 0, B = 1 Solution
means my box started at zero but it started by moving right. It’s as
I get a sine function.
though we had a box just sitting there at equilibrium, and we flicked
it with a finger, sending it right. Because the box moves right, the
position x (t) is increasing. That’s what we see in Figure 10, the sine
starts at zero, but increases. So it all checks out, and now I know
what the B term is doing: it’s there to tell me what the starting veloc-
ity is.
So the initial conditions here are starting position and starting ve-
locity. The equations reminded me that I need to know both, before
I can tell how the box will move. The equations also tell me I don’t
need to know any more than that; the spring takes care of everything
after that. I’m actually getting more out of these equations than I put
in.
 calculus sin frontera 7

There is one thing, though: we said "The box used up all its velocity,
compressing the spring as far as it can." That’s a metaphor, as though
our little friend boxy pushed as hard as it could but just couldn’t
push any more. Like I’ve said, we’re scientists and engineers; we
want computations, not metaphors.
Physicists think about problems like this by by talking about energy,
kinetic energy and potential energy. Kinetic energy is energy associ-
Figure 11: Large k
ated with motion and is proportional to v2 (the word ’kinetic’ comes
Cosine, but this time doing four oscilla-
from the Greek, meaning ’to move’). Potential energy is associated tions.
with compression and stretching of the spring. When I said "The
box used up all it’s velocity, compressing the spring as far as it can",
what I mean is that all the kinetic energy has been transferred to the
spring. When kinetic energy is zero, v2 = 0, so there’s no longer
any velocity. That’s physics: as the box moves back and forth, the
two kinds of energy trade off, and the oscillations record that trading
off. No energy is lost, so the oscillation continues – forever, I guess.
That’s another fact I get out, that I never thought I put in. q
k2
q
Figure 12: Small: m = 2π
2 A cosine, going through one oscillation.
At last, it’s time to come to terms with the km . Again, we’ll keep it
q 
k2
simple by taking A = 1, B = 0, so that I get x (t) = cos m t.

Let’s fiddle with k. We called k the spring constant, and said it tells
us how springy our spring is. More precisely, k tells us how strongly
the spring pushes or pulls when q its compressed or stretched. We’ll
2
start as we did before, by taking km = 2π, or 8π. The motions
are compared in Figures 12 and 13. How would you describe the
difference between the two? If you keep k constant but increase m
to go from 8π to 2π, what would you expect to happen? What does q
k2
Figure 13: Large: m = 8π
happen? If you keep m constant but increase k to go from 2π to 8π,
Same cosine, but this time doing four
what would you expect to happen? What does happen? oscillations.

Figure 14: Analysis

Oxford English Dictionary: "A detailed
examination or study of something so
as to determine its nature, structure, or
essential features. From Latin, analysis:
act of resolving (something) into its
elements."