Dickinson College

Dickinson Scholar
VPython for Introductory Mechanics

2019

VPython for Introductory Mechanics: Complete

Version
Windsor A. Morgan
Dickinson College

Lars Q. English
Dickinson College

Follow this and additional works at: https://scholar.dickinson.edu/vpythonphysics

Part of the Astrophysics and Astronomy Commons, Curriculum and Instruction Commons,
Numerical Analysis and Computation Commons, Physics Commons, and the Science and
Mathematics Education Commons

Recommended Citation
Morgan, Windsor A. and English, Lars Q., "VPython for Introductory Mechanics: Complete Version" (2019). VPython for Introductory
Mechanics. 1.
https://scholar.dickinson.edu/vpythonphysics/1

This Book is brought to you for free and open access by Dickinson Scholar. It has been accepted for inclusion in VPython for Introductory Mechanics
by an authorized administrator of Dickinson Scholar. For more information, please contact [email protected].
VPython for Introductory
Mechanics
Incorporating numerical simulations
in the introductory physics
curriculum

W.A. Morgan and L.Q. English

Department of Physics and Astronomy
Dickinson College
 © 2019 W.A. Morgan and L.Q. English, Carlisle, PA 17013
All Rights Reserved.
https://scholar.dickinson.edu/vpythonphysics/

For more information about permission to reproduce selections

from this book, write to [email protected].

On the front cover: the large photograph is of Melba Roy Mouton.

She calculated artificial satellite orbits as a Head Computer
Programmer for NASA.

Left Cover Image source: NASA on the Commons

Right Cover Image source: https://pxhere.com/en/photo/1199802
Contents

1 Introduction 1

2 Getting Started with VPython 5

2.1 “Hello.” . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Box – Your First Shape . . . . . . . . . . . . . . . . 8
2.3 Another shape . . . . . . . . . . . . . . . . . . . . . 11
2.4 Comments . . . . . . . . . . . . . . . . . . . . . . . 11

3 Moving Objects Using Formulas 13

3.1 Motion of a Ball with Constant Velocity . . . . . . . 13
3.2 Motion of a ball with constant velocity with a plot
of position vs. time . . . . . . . . . . . . . . . . . . 14
3.3 Motion of a Ball Being Dropped from Rest Near the
Surface of the Earth . . . . . . . . . . . . . . . . . . 16
3.4 Stop and Go . . . . . . . . . . . . . . . . . . . . . . 17
3.5 Motion of a Planet . . . . . . . . . . . . . . . . . . . 18

4 A First Look at Simulating Motion 21

4.1 The Physics . . . . . . . . . . . . . . . . . . . . . . . 21
4.2 The Basic Code . . . . . . . . . . . . . . . . . . . . . 23
4.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . 25

5 Visualizing Projectile Motion 27

5.1 The Physics . . . . . . . . . . . . . . . . . . . . . . . 27
5.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . 28
5.2.1 Projectile over Flat Ground . . . . . . . . . 28
5.2.2 Projectile over Sloping Ground . . . . . . . 30
5.2.3 Projectile Striking a Target . . . . . . . . . . 30
5.2.4 Mystery Code . . . . . . . . . . . . . . . . . 31

6 Simulating Central-Force Problems 33

6.1 Stating the general problem and a possible line of
attack . . . . . . . . . . . . . . . . . . . . . . . . . . 33
6.2 A first example: The mass-spring problem . . . . . 36
 6.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . 37
6.4 A second example: Projectile motion with air drag . 38
6.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . 42
6.6 Third Example: Trajectories of Planets and Comets 43
6.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . 45

7 Conservation of Momentum and Energy 47

7.1 A binary-star system . . . . . . . . . . . . . . . . . 47
7.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . 51
7.3 Energy in the Spring-Mass System . . . . . . . . . . 52
7.4 Simulating the Rutherford experiment . . . . . . . . 53
7.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . 56

8 Rotational Motion, Torque, and Angular Momentum 59

8.1 The Physics . . . . . . . . . . . . . . . . . . . . . . . 59
8.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . 61

9 Two Capstone Projects 63

9.1 The Gravitational “Slingshot” . . . . . . . . . . . . . 63
9.1.1 The Physics . . . . . . . . . . . . . . . . . . 64
9.1.2 The Code . . . . . . . . . . . . . . . . . . . 67
9.1.3 Exercises . . . . . . . . . . . . . . . . . . . . 69
9.2 Asteroid near a Binary Star System - Chaotic orbits,
or the “Three-Body Problem” . . . . . . . . . . . . . 70
9.2.1 Modifying a previous code . . . . . . . . . . 70
9.2.2 Exercises . . . . . . . . . . . . . . . . . . . . 73

10 Conclusion 75

Bibliography 77

About Authors 79
Preface

Many physics departments nationally and internationally have

moved to incorporate more computation into their curriculum. In
many cases this is accomplished by offering an upper-level elective
in computational physics. Frequently, computational exercises are
also incorporated in upper-level course work. Less often do we see
substantial integration of coding in the first semester of physics.
One exception is the Matter and Interactions curriculum developed
at North Carolina State University.1

In this book, we describe a series of group exercises that we in-

corporated into our first course for majors. The calculus-based in-
troductory sequence at Dickinson College is taught in the Work-
shop Physics mode which emphasizes inquiry-based exploration in
groups. Students work through exercises in groups of four, and
these mostly involve guided observations, measurements and exper-
iments. In the fall semester of 2018, we started supplementing these
with numerical projects, and this book grew out of that experience.

In the following chapters, we describe a series of computational exer-

cises that can be easily woven into a project-centered course format.
Alternatively, these exercises could also be included into the labora-
tory part in a more traditional course. The sequence of topics was
meant to follow a typical calculus-based intro-mechanics course, as
well as an Advanced Placement Physics high school class. (It does
not, however, mesh well with alternative paradigms, like the energy-
first approach.) For instance, we start with graphing motion using
the kinematics equations, then move on to projectile motion, before
delving into dynamics and the simulation of motion using physical
law. The next-to-last chapter is an exploration of rotational motion,
torque and angular momentum conservation, topics usually at the
end of a typical course. The last chapter considers the “real-world”
1
R. Chabay, B. Sherwood, Matter and Interactions (John Wiley & Sons, 2015)
example of gravitational assist in space travel.

One important precondition for such an endeavor is that the pro-

gramming language not get in the way of the physics. For us, it was
not very important for students to become proficient in coding. In
fact, we chose to provide the students with a scaffold by giving them
an initial template for almost all exercises, and students were then
asked to modify the code to complete the activities. In the same
spirit, we tried to keep the code we gave students as short as possi-
ble, while encouraging them to embellish as they saw fit.

We chose VPython for a number of reasons. First, it is convenient

that it lives online. Nothing has to be installed locally. Secondly, its
well-known strengths in 3D visual rendering is obviously a major
benefit in a first course in classical mechanics. Finally, its basis in
the Python programming language allows students to get some early
exposure to a language that is becoming ever more widely used in
scientific computation.

L.Q. English and W.A. Morgan

Carlisle, Pennsylvania, August 2019
Introduction

This book is intended to supplement the course materials of a first-

semester introductory course in physics by adding a computational
dimension into the course content through the use of VPython - a
“visual” extension of the Python programming language that incor-
porates 3D animation and vector math. The book is not designed to
replace the regular course textbook. Since it is intended to be used in
conjunction with other material presented in the course, one of the
goals was to make its integration into the course flow as seamless
as possible. To this end, we have organized the topics and exercises
so that they should more or less follow the traditional sequencing
of topics encountered in most introductory mechanics courses.

The exact order of topics was chosen to enable its easy integration
into our Workshop Physics curriculum.1 While the pedagogical ap-
proach, format and presentation of Workshop Physics is very differ-
ent from the more standard lecture-based course, its content selec-
tion is fairly conventional. Thus, we believe that this presentation
should also work well in conjunction with a more traditional intro-
ductory mechanics course (either at the senior high-school or first-
year college level). Furthermore, swapping neighboring chapters or
omitting certain sections should create only minimal disruption.

Having said that, introducing computation in the first semester does

allow for some curricular innovation, such as in the form of inclu-
sion of types of problems that are not typically encountered other-
wise. In particular, it allows students to examine problems that have
no closed-form analytical solutions, to engage in more creative and
open-ended problems,2 and to practice physical modeling. One ex-
ample we include in this book has to do with the law of universal
1
Workshop Physics was developed at Dickinson College, principally by P. Laws, as an
inquiry-based, active-learning curriculum for introductory physics
2
E.F. Redish and J.M. Wilson, “Student programming in the introductory physics course”,
Am. J. Phys. 61, 222 (1993).
2 Chapter 1 Introduction

gravitation. After a sequence of exercises where students discover

Keplerian orbits, we ask them to explore what would happen to plan-
etary orbits if gravity were instead a “1/r” kind of force. Such a
question would be quite difficult to answer analytically, and would
require an upper-level arsenal of mathematical tools, but intro stu-
dents can tweak the code very easily to explore this scenario.

Thus, the use of VPython has the pedagogical virtue of empowering

students to engage in more physical modeling, and it enables them
to better visualize processes, both of which in turn strengthen stu-
dents’ conceptual understanding.3 Finally, a very straightforward
benefit for introductory students is that they are exposed to a lan-
guage, Python, that is fast becoming the premier tool for scientific
computation.

The structure of the book is as follows. We start in Chapter 2

with some preliminary exercises to get students familiar with the
interface and basic coding in VPython. In Chapter 3, we then
turn our attention to one-dimensional motion and motion diagrams.
Quite a lot of emphasis is placed in Workshop Physics, as well as
RealTime Physics,4 on describing basic motion in multiple represen-
tations: verbally, in graphical form, and with mathematical equa-
tions. Physics education research has shown that real conceptual
and problem-solving progress occurs when a student can move flu-
idly between these various representations of motion.5 The idea in
this chapter is to bolster students’ comprehension of position- and
velocity-graphs, which are one layer removed in abstraction from
the literal description of the motion, by having them program an
object performing a simple kind of motion on the screen and then
3
R. Chabay, B. Sherwood, “Computational Physics in the introductory calculus-based
course”, Am. J. Phys. 76, 307 (2008).
4
D. Sokoloff, R. Thornton, P. Laws, RealTime Physics - Active Learning Laboratories (John
Wiley & Sons, 2012).
5
A. Van Heuvelen, “Learning to think like a Physicist: A review of research-based instruc-
tional strategies,” Am. J. Phys. 59, 891 (1991).
L.C. McDermott, “A View from Physics” in Toward a Scientific Practice of Science Education
(Lawrence Erlbaum Assoc., 1990).
 3

having VPython generate the corresponding motion graphs. In a

sense, it is the computational analogue to the common laboratory
exercise of using a motion sensor and walking in front of it.

In Chapter 4 we assume the course has by now advanced to a discus-

sion of average and instantaneous velocity and acceleration. (In a
traditional intro course, this may come very quickly.) We use some
of these ideas to introduce the basic Euler method at the heart of all
numerical simulation encountered here.

In Chapter 5, we return briefly to the realm of plotting formulas, as

the course is now at the point of exploring projectile motion, i.e.,
two-dimensional motion under constant acceleration. This topic is
naturally suitable for adding a computational facet. Exact analytical
results (such as the range formula) can be easily tested numerically,
for instance, as outlined in the exercise section of this chapter. Then,
additional complications can be added (such as a sloped ground). We
end by having students code a “hit-the-target” game.

Chapter 6 in many ways strikes at the heart of mechanics. We are

finally ready to simulate Newton’s laws, and the full power of com-
putational physics is revealed for the first time. We can see the kinds
of motion that Newton’s laws actually give rise to, given the forces
acting on a particle and the particle’s initial state. We specifically
highlight three important problems in order of increasing difficulty:
the 1D mass-spring problem, projectiles traveling through air, and
the orbits of planets and comets. What makes this approach so po-
tent in the context of an introductory course is that the analytical
solutions are too mathematically challenging to derive and therefore
remain beyond reach for this audience. Yet, the numerical pathway
is accessible.

For Chapter 7, it is assumed that the course has now reached a dis-
cussion of momentum and energy. We chose the binary star system
4 Chapter 1 Introduction

as a vehicle for introducing the concept of center-of-mass and to tie

it to momentum conservation. We then return to the mass-spring
problem to illustrate the conservation of mechanical energy in this
system. The Rutherford experiment can be skipped, but it provides
a nice preview to electrostatics.

Maybe unsurprisingly, we now follow up with rotational motion

and angular momentum, which often represents the last material
covered in the semester. We end with what we believe to be two
engaging final examples - the gravitational “slingshot” and chaotic
orbits - as a kind of capstone project for students to explore.

Lastly, before we get started, all the VPython codes featured in this
book can be accessed online at
https://scholar.dickinson.edu/vpythonphysics/ .
Getting Started with VPython

2.1 “Hello.”

VPython is a language that is pretty easy to use, once you get used to
it. Python is becoming widely used by scientists around the world,
and VPython is a visual implementation of it. It will help you visu-
alize the physics you will be learning.1

To learn the basics of programming in VPython, the first thing we

will do is one of the simplest – having the program print the phrase
“Hello, world!” – a traditional first program.

To do so, we will be in the Glowscript-VPython programming envi-

ronment – an environment that is easy to use and understand. Here
is what you do:

1. Launch a web browser (try Google Chrome), and go to the site

www.glowscript.org . The webpage will look like this:

2. Sign in with a Google account.

3. Follow the link where it says “your programs are here.”

4. Click on the “Create New Program” link.

5. In the dialog box that appears, type the title “MakingShapes”,

then click the “Create” button.

1
See also D. Schroeder, “Physics Simulations in Python” (2018).
6 Chapter 2 Getting Started with VPython

Figure 2.1: The Glowscript webpage.

2.1 “Hello.” 7

6. You will now see an editing space that is blank except for the
line “GlowScript 2.8 VPython” (possibly with a different ver-
sion number). Click in the white space below that line and
type the following, verbatim:

print("Hello, world!")

7. Then click the “Run this program” link at the top of the page.
Your editing space should then vanish, replaced by a new text
area with the words “Hello, world!”

8. Click on ”Edit this program” and replace “Hello, world!” with

something else. Run the program. You should get similar re-
sults.

If the program didn’t work, check the spelling and capitalization and
punctuation. Make sure that you typed the line precisely as shown
above.

GlowScript-VPython is a programming environment that stores

your programs in the cloud. The interface is very spare – however,
there should be no difficulty just starting it up, making new pro-
grams, saving them, and managing them.

If you click “Edit this program”, you can see the program you just
typed. The first line, which is the same for all GlowScript programs,
tells the system the language (VPython) and version (2.8) that is to
be used. You must have this line, and you should always just leave
it alone.

The next line tells the computer to call the print function and passes
to it a string of characters. Don’t worry that you don’t know what
the italicized words mean – we’ll explain them here:
8 Chapter 2 Getting Started with VPython

• The function is a pre-defined task that does something – in this

case, print to the screen. Just as you don’t need to know how
a square-root key on a calculator works, a function does its
job on the information passed to it. This information is called
the argument or the parameter (they are slightly different, but
there is not much of a difference). Here the argument is the
string.
• The string is text that is either single- or double-quoted.
• To tell the program what to do it and when, the function is
called, which is done in the act of stating the function (here,
“print”).

2.2 Box – Your First Shape

Now, let’s do something a bit more exciting.

On a new line below your print command, type the following com-
mand:

box()

Here you’re calling a function called box, and passing it no argu-

ments at all (but notice that you still need the parentheses). Run the
program again, and you should see a black rectangular area (called a
canvas) containing a light gray square (the box), such as one shown
in Figure 2.2. Again, this function is doing a lot of work “under the
hood”, but you do not need to concern yourself about how.

The box that you see on the screen is actually three-dimensional on

the screen. You are viewing it closely from one side, so it appears as
2.2 Box – Your First Shape 9

Figure 2.2: A Simple Box

a square. To change the perspective, you can do three things:

• Rotate: Right-click button on your mouse and drag the box

one way or another, or, if you don’t have a right mouse button
(such as on a Macintosh), press the control key and drag using
your mouse or trackpad.
• Zoom: Use the scroll wheel on your mouse, or, if there isn’t
one, press the alt or option key and drag using your mouse or
trackpad.
• Pan: Hold down the shift key while you drag using the mouse
or trackpad.

You can change the attributes of your box by passing some differ-
ent parameters to the box function. Try typing this to replace the
original box line:
10 Chapter 2 Getting Started with VPython

box(pos=vector(1,0,0),size=vector(.5,.3,.2),
color=color.green)

Here you’re providing three parameters, separated by commas, and

you’re identifying them by their names, which means you can pro-
vide them in any order. The pos parameter specifies the position
of the center of the box in Cartesian coordinates; the size parame-
ter specifies its dimensions (so here, it is not a cube; and the color
parameter is self-explanatory. In the first two cases, the parameter
values are three-dimensional vectors, which you create using the
vector function. We will be learning more about this powerful func-
tion soon. This function in turn takes three parameters, x, y, and z,
which are usually given in that order. When first presented and be-
fore any rotation, the x direction points to the right; the y direction
points up; and the z direction points directly outward, toward you
(or “out of the screen”).

You can learn more about other colors by clicking the Help link
at the upper-right corner of the window. It might be beneficial to
open the help page in a new browser tab. Then, from the second
drop-down menu in the left sidebar, choose “Color/Opacity”.

There you will find a list of pre-defined colors, and also see how you
can use the vector function to create arbitrary colors.

Exercise: Create at least two more boxes, so you’ll have a total of at

least three, each with different positions, shapes, and colors. Keep
their positions within the range -5 to 5 in each dimension, and keep
their sizes small enough to leave plenty of room for more shapes
with that range. Make sure to “Grab” (a MacOS program) a screen-
shot of what you did.
2.3 Another shape 11

2.3 Another shape

Fortunately, VPython provides functions for creating many different

shapes, but in this course you’ll need just two others: spheres and
cylinders. Try this instruction to create a sphere:

sphere(radius=0.25)

The sphere function can also accept the pos and color parame-
ters, so use those now to change the defaults according to your taste.
Don’t forget to separate the parameters by commas!

2.4 Comments

Sometimes you want to have some comments to remind you what a

certain line or area of a program is doing. All you need to do is put a
# sign in front of the comment. Then VPython simply ignores every-
thing following the # sign on that line. (It is very easy to “comment
out” a line of code you want to not run, sometimes as a diagnostic.
Then you can remove the # sign if you want the program to execute
that command).

Exercise: Put a multi-line comment at the top o your program right

after the line “GlowScript 2.8 VPython”, to indicate the name of your
program, your own name, and the day when you created it, and
to give a one-sentence description of what it does. (From now on,
please include a similar comment at the top of every program that
you write.)
Moving Objects Using
Formulas

In this activity, you will be doing four simple exercises. In three

of them, a ball will move with either constant velocity or in in the
presence of a constant force. In the last, you will move a planet (just
a large ball!) around the Sun.

3.1 Motion of a Ball with Constant Velocity

Examine the program provided for you in Figure 3.1. Type it in.
Please note that the indentations are important (as is capitalization,
usually).

Figure 3.1: Motion of a ball with constant velocity

What is each line doing? You can probably figure it out using the
commented-out lines. Now run the program. What is happening?

You probably found that this program is having a sphere move hor-
izontally at a constant velocity. The command in line 14:
14 Chapter 3 Moving Objects Using Formulas

x = x0 +v*t

is telling the program to take the original value of x, x0, and to add
a constant velocity times the time elapsed t. This works, but we
will see in the next chapter that it is not the most elegant way of
updating the value of x.

3.2 Motion of a ball with constant velocity

with a plot of position vs. time

Looking at the motion of the ball is fine, but plotting the values of
position, velocity, or acceleration versus time is a way to visualize
what is happening.

Take the first program, copy it with a new name, and add commands
to plot the ball’s position with respect to time. To do this, you must
set up a graph and declare its width and height:

Your program should look something like the one in Figure 3.2:
3.2 Motion of a ball with constant velocity with a plot of
15
position vs. time

Usar em MRU

Figure 3.2: Program Illustrating Graphing Commands

The command graph in line 3 sets up a graph with width 400 pixels,
height 200 pixels, with an x-axis labeled “time” and a y-axis labeled
“position”. The set-up graph is called “g1” for ease of use. Line 4
uses the command gdots to say how the data is to be plotted, and
call it “xDots”. Here, it says that green dots should be used in the
graph called “g1” that was set up in the line before. Finally, line 21
says that “xDots” should be plotted with the values of “t” (for time)
for the x-coordinates, and the values of “x” (for position) for the y-
coordinates.

Your plot should like what is shown in Figure 3.3. Is this what you
would expect? Why?
16 Chapter 3 Moving Objects Using Formulas

Figure 3.3: Plot of position versus time for ball with constant
velocity.

3.3 Motion of a Ball Being Dropped from

Rest Near the Surface of the Earth

Now, let’s simulate the motion of a ball being dropped from a height
of two meters.

Questions: What has to be done to modify the second program to

mimic the ball being dropped? What is different about this ball’s
velocity compared to the ball’s velocity in the second program?

What does the position-time plot look like now? How is it different
from the previous plot? Why?

Plot the velocity versus time. How does this compare to the current
position versus time plot and the previous position versus time plot?
3.4 Stop and Go 17

3.4 Stop and Go

In the previous section you probably discovered that instead of using

a constant speed v0, you needed to increment the speed with every
pass within the WHILE loop.

Now let’s use a similar idea to make the ball move to the right at a
constant speed for 2 seconds, then to have it stop and sit there for 2
second, and then to have it move to the left at a constant speed for
2 seconds.

Discuss a possible strategy with your partners. You may find that
there are multiple ways in which you could accomplish this objec-
tive. One way would be to have three different WHILE loops - one
after the other. However, perhaps a more elegant approach uses the
if-then-else syntax - a core syntax structure of any programming
language. In VPython, it gets implemented in a very intuitive way:

If (“Condition to be tested”):
“Execute these commands”
Else:
“Execute those commands”

Here “Condition to be tested” should be something like t<2, t>2,

or t==2, for example. Notice the double equal signs that are needed
because the line should be interpreted as a test of a conditional state-
ment (and not an assignment). Note that the indentations are es-
sential - horizontal alignments signify and enforce structure in any
Python program.

Can you incorporate if-statements to make the ball move to the right,
stop, and then move to the left without any discontinuities? It may
18 Chapter 3 Moving Objects Using Formulas

take you a while to get the third part right.

3.5 Motion of a Planet

In Figure 3.5 examine the program simulating the motion of a planet

(the Earth) around the Sun. This is a very simplified depiction; it
does not take into account gravity or any of the laws of planetary
motion. It merely assumes the Earth’s orbit is a perfect circle (which
in reality is not too far from the truth). In the program, note that
there are cosine and sine terms, to define the x and y positions.

Figure 3.4: Simplified Motion of a Planet Around the Sun

The term omega, written as ω, is called the angular velocity (or the
frequency). You will see this term later in the course when we dis-
cuss rotational motion. Its value in the program will help you to see
3.5 Motion of a Planet 19

how it is defined.

Exercise: Change the parameters of the provided program. How

does the executed program change?

Exercise (optional): Define a third sphere and call it “Moon”. Use

the same approach to make the Moon revolve around the Earth,
while the Earth still simultaneously revolves around the Sun. Hint:
Vector addition will be very useful here. You have to add something
to the position vector of the Earth.
A First Look at Simulating
Motion

In the last activity, we managed to move objects on the screen by

repeatedly evaluating the position function, x(t), for successively
larger time values. This iterative evaluation of the function was done
in the WHILE loop.

We were also able to graph the values in a position-time graph.

This approach was quite powerful already, but it also limits what
we can do. For example, it would already be quite difficult to make
an object move forward, stop for a while and then move backward
using this approach, as you probably discovered in the last chapter.
Furthermore, this approach also imposes solutions rather than nu-
merically finding them, thus sort of bypassing the underlying phys-
ical laws, as we will see later.

For these reasons, we would like to move from a mere evaluation of

a function to an actual simulation of a physical process in real time.
Today, we will take our first step in the direction of a true numerical
simulation of physics.

4.1 The Physics

Let’s start with a piece of physics that you already know – the for-
mula for average velocity in one dimension – and then translate it
to a statement a computer can understand. So, average velocity is
defined as,
22 Chapter 4 A First Look at Simulating Motion

∆x x2 − x1
vavg = = (4.1)
∆t ∆t

In words, the average velocity over a time interval of motion is de-

fined as the displacement divided by the elapsed time.

As a general rule, computers cannot handle continuous motion.

Therefore, in order to make any simulation of real physical motion
look realistic, we will have to “chop up” the true continuous motion
into many, many frames separated by very small time intervals. The
situation is not unlike watching a movie where we perceive continu-
ous motion when in reality we are being exposed to a rapid sequence
of still images.

The upshot is that we will want to make our time interval ∆t really
small in Equation (4.1). Over the course of this small time interval
we would not expect the velocity to change very much, so a pretty
good assumption is that it is simply constant over this interval. This
also means that the average velocity is just the same as this con-
stant velocity and we can drop the subscript “avg”. If we then solve
Equation (4.1) for x2 , we get:

x2 = x1 + v∆t (4.2)

Here we assumed that ∆t was sufficiently small, which allowed us

to omit the subscript avg.

Let’s describe Equation (4.2) in words: Between two frames sepa-

rated in time by ∆t, we can compute the position of the object in
the new frame, x2 , based on the position in the current frame, x1 ,
and the velocity of the object, v. We simply have to add v∆t to x1 .

Now all we have to do is repeat the step in Equation (4.2) over and
over again. Every time we advance forward in time by the small
4.2 The Basic Code 23

amount ∆t. And every time, we replace the starting position x1 by

the ending position x2 of the previous step. In other words, we want
to use Equation (4.1) recursively in order to evolve the position of
our object forward in time, step by step.

So far, we have assumed that v in Equation (4.2) would remain con-

stant in each successive step, but this does not have to be the case.
In general, we can allow the velocity v to get updated as well with
each new iteration. We will see some straightforward examples in
the section 4.3. The procedure we have described here is also known
in the numerical analysis community as the Euler Method.1

4.2 The Basic Code

Now that we a taste of the physics involved, how do we translate

that idea into code a computer can interpret? Let’s take a look at
the following few lines of VPython code - see Figure 4.2.

We see that the core of the program is contained in the WHILE loop,
as before. Let’s start our discussion with the central line, namely
LINE 22:
x = x + v ∗ dt (4.3)

As a mathematical statement this is, of course, nonsense. But we

should not interpret the equal symbol in the mathematical sense. It
is not an equality. It is an ASSIGNMENT. “x” is a variable that holds
a certain number, and whenever you see “x = something” this signals
that the value stored in the variable “x” is about to be updated.

So Equation (4.3) should be read in two parts. The expression to

1 Named after the mathematician and physicist Leonhard Euler.
24 Chapter 4 A First Look at Simulating Motion

Figure 4.1: The basic code

the right of the equal sign is computed first based on the current
value of x. Secondly, the “x = ” part then assigns the result of that
computation to the variable x. The upshot is that x is updated and
now holds the new value.

Thus, if we had to translate the statement in Equation (4.3) into a

mathematical equation, we would say:

x2 = x1 + v∆t,

where x2 is the new value and x1 is the old value of position. Note
also that “dt” in Eq. 4.3 should not be interpreted as a differential in
the Calculus sense but represents a small time interval, defined in
LINE 12 as 0.05.

We are now in a position to discuss the entire code. Line 4 and 5

set up a position graph (time on the horizontal axis, position on the
4.3 Exercises 25

vertical axis). LINE 8 defines the object we want to move around in

space – a red sphere of radius 0.1. Lines 11-13 set the initial condi-
tions as well as the size of the small time interval, dt. Within the
WHILE look, we keep updating the variable x and then, using this,
to update the position of the object using the command in LINE 19:
obj.pos=vector(x,0,0). Here obj was defined as the sphere, and pos
is an attribute of the sphere, namely the position (i.e. location) of the
sphere’s center. This position is a three-dimensional vector which
we create using the vector command. Finally, the xDots.plot(t,x)
command in LINE 20 appends the newest data point to our position
graph.

The last line, Line 23, updates the time variable. This is necessary in
order to test the conditional of the WHILE loop (t<3), as well as for
the purpose of plotting the position (and velocity) graph. It does not,
however, come into play in calculating the position of the object – an
important difference from the previous approach of simply plotting
the function x(t).

4.3 Exercises

1. • Run this program to make sure it works. Describe what

you see.
• Modify the code so that during the first second, the ball is
moving to the right as before, but for the second second
it stays still, and during the third second, it returns to
the original starting point. Also have VPython generate
the corresponding position-time graph.
Hint: Think about how you could use if-then statements
to accomplish this task.
26 Chapter 4 A First Look at Simulating Motion

2. • Discuss with your group how you would have to modify

the code printed above to make the object accelerate to
the right at constant acceleration. Hint: Think about the
role of “v”!
• Once you verify that the code really does produce an
accelerating object on the screen, let’s have VPython
generate the position-time graph, as well as the velocity-
time graph. Ideally, the data would be displayed in two
separate graphs. You may have to duplicate and slightly
modify some of the lines already appearing in the sam-
ple code.
Take a screenshot of the code, as well as the two graphs
to include in your lab notebook.
• Now that you have the position-time graph, verify that
it agrees with the first kinematics equation: x(t) = x0 +
v0 t + 12 at2 . You can do this, for instance, by evaluating
the formula at a particular time (for the initial velocity
and the acceleration that appears in your code). Does
this point lie on the the graph?
Visualizing Projectile Motion

In this activity, you will extend what you learned in Chapter 3 about
the kinematics equations into two dimensions. These equations de-
scribe the motion of a projectile undergoing a constant acceleration.

Usually a projectile, such as a thrown ball or a launched rocket, is

propelled at an angle to the ground θ, usually measured from the
horizontal. One must take into account the components in the hori-
zontal (usually “x”) and vertical (usually “y”) directions.

5.1 The Physics

The kinematic equations are as follows:

1
x = x0 + vx0 t + ax t2 ; (5.1)
2
vx = vx0 + ax t; (5.2)
and
vx2 = vx0
2
+ 2ax (x − x0 ). (5.3)

While “x” is used above in Equations (5.1), (5.2), (5.3), they are simply
placeholders; one can easily replace them with “y”:

1
y = y0 + vy0 t + ay t2 ; (5.4)
2
vy = vy0 + ay t; (5.5)
and
vt2 = vy0
2
+ 2ay (y − y0 ). (5.6)
28 Chapter 5 Visualizing Projectile Motion

Therefore if we have a projectile launched at an angle θ to the

ground, then the horizontal component of the velocity is
vx = v cos θ (5.7)
and the vertical component of the velocity is
vy = v sin θ. (5.8)

So, if we throw a ball at angle θ = 43◦ , with a velocity of 10 m

s , then
the initial x- and y-components of the velocity are (cf. Equations
(5.7) and (5.8))
m
vx0 = 10 cos 43◦
s
and
m
vy0 = 10 sin 43◦ .
s

As for acceleration, if we are talking about locations near the surface

of the Earth and ideal conditions (no atmosphere, so therefore no
wind), there is no acceleration in the horizontal (x) direction. The
only acceleration is the familiar ay = −g = −9.8 m s.

5.2 Exercises

5.2.1 Projectile over Flat Ground

Use Equations (5.1), (5.2), (5.4), (5.5), (5.7), and (5.8) to write VPython
code that simulates the flight of a projectile. Make the projectile a
sphere of radius 0.1. Assume that there is no air resistance. Assume
that we launch the projectile at the origin (0,0), and stop the pro-
jectile motion once it hits the ground at y = 0. This means that the
terrain is flat. Draw a line (or horizontal plane) that indicates this.
5.2 Exercises 29

Figure 5.1: A Projectile Trajectory Over a Flat Slope

To accomplish this last condition, use the following while loop:

while(y>-0.001)

Your program should have an output such as depicted in Figure 5.1.

One thing to keep in mind is the built-in unit for angles in VPython
is always radians. This means that you will need to convert any
angle in the argument of a trig-function from degrees to radians.

Now play around with launch angle at constant launch speed. Make
a table of the range for each launch angle. For what angle θ do
we seem to get the largest range? Is this range consistent with the
theoretical formula for the range of a projectile, given by
30 Chapter 5 Visualizing Projectile Motion

v02 sin 2θ
R= . (5.9)
g

To answer this question, open up Excel, have it compute the theo-

retical value from Equation (5.9) at regular angle intervals, and then
compare the values with your numerical results from VPython sim-
ulations.

5.2.2 Projectile over Sloping Ground

Now let’s make the ground not level (or flat). Let’s have a steady
downward slope. Redraw the line (or plane) so that it tilts down
with a slope of rise/run = -0.1.

The way we can simulate that in the code is by modifying the WHILE
statement. How do you have to do that? Once your group finds a
solution, implement it in the code.

What angle now produces the largest range?

Repeat the simulation, but with a terrain that gently slopes up, with
a slope of +0.1. What is the best angle now?

5.2.3 Projectile Striking a Target

In addition to the projectile and the (flat) ground, draw in a station-

ary sphere of radius 0.5. You decide the coordinates of this stationary
sphere.
5.2 Exercises 31

Make the code stop if or when the projectile hits the object, which
we will define as anywhere within the volume of the object. If we
do not hit the sphere, then make the code stop as before when the
ground it hit. Because we are only thinking in two dimensions, in
order to hit the sphere we need to just have the projectile get within
0.5 units of the center. How can we do that?

Hint: there is a kind of brute-force approach using the Pythagorean

theorem, but there is also a much more elegant method using the
two position vectors associated with the projectile and the target
sphere. Can you see how you could use the difference of these two
vectors?

One thing that will probably need here is the following command:

BREAK

If this command is encountered within a WHILE-loop, we im-

mediately leave the loop and continue with statements outside of it.
Oftentimes, you will find the break command wrapped inside an
IF-statement all inside a while-loop.

Projectile motion is a fairly easy thing to simulate in VPython. In

the next chapter, you will see how to make the simulation more
realistic by the incorporation of air resistance in the code, with the
use of vectors.

5.2.4 Mystery Code

Scan the following VPython code and annotate each line. Discuss
what you think this code will do. Make a prediction as to the kind
of motion that would be highlighted.
32 Chapter 5 Visualizing Projectile Motion

Note that in the VPython environment, to raise a quantity to some

power, you do NOT use the carrot-symbol, but instead you use two
star-symbols “∗∗”. Thus, x2 would be represented in the code as
x ∗ ∗2.
Simulating Central-Force
Problems

6.1 Stating the general problem and a

possible line of attack

In Chapter 4, we learned how we could use an iterative procedure

(called the Euler method) to advance the position of an object in
small increments based on its current velocity. This allowed us to
numerically obtain x(t) given v(t). In other words, we performed
a numerical integration: given the known velocity-time graph, we
generated the position-time graph.

The task in many physics problems is slightly different, however.

Very often, we know the force acting on an object at all points in
space, and we would like to somehow calculate the object’s trajec-
tory in response to the forces it encounters. How do we do that?

So, let’s assume that we know the force that is acting on an object as
a function of the object’s location. What this means is that no matter
where the object happens to be, we can calculate the force that it
experiences there. Mathematically speaking, what we are given is
the force-function, F⃗ (⃗r). This notation communicates that we can
evaluate a force vector, F⃗ , by inserting a position vector, ⃗r, into a
function. Mathematicians would call such a function a map from R3
to R3 . A good example are the so-called central-force problems, such
as the gravitational force a comet feels in the vicinity of the sun.

We also know that the force on the object is related to the accelera-
tion of the object via Newton’s second law:

F⃗ = m⃗a (6.1)
34 Chapter 6 Simulating Central-Force Problems

This means that simply dividing the force-function by the object’s

mass gives us the acceleration function ⃗a(⃗r).

Now that we have the acceleration, how do we get the trajectory?

Recall that we faced a somewhat similar task in Chapter 4. There we
knew the velocity and were able to compute the position. But now
we are one step further removed. We know the acceleration, not
velocity. What’s more, we know the acceleration not as a function
of time, but as a function of position.

You may know that in these types of problems, in order to compute

the object’s trajectory, we must know how the object was initialized.
At what position was it released and what velocity did it have at that
moment? These two things must be known to us if we are to find
the unique trajectory - they are called the initial conditions.

So here is the basic idea. Let start from this position and velocity in
our code, call them: ⃗x1 and ⃗v1 . From ⃗x1 we can compute the force
at that location and thus the acceleration ⃗a1 . Using this ⃗a1 , we now
update the velocity. Here we make use of the formula for average
acceleration,
∆⃗v
⃗aavg = . (6.2)
∆t
As the time interval gets very small (and, in the limit, infinitesi-
mal), the average acceleration becomes the instantaneous acceler-
ation, and so,
⃗v2 = ⃗v1 + ⃗a ∗ ∆t. (6.3)

This completes the second step.

In the third and final step, we now use this new velocity to update
the position, using the vector-verson of Equation (4.2), namely:

⃗x2 = ⃗x1 + ⃗v2 ∗ ∆t. (6.4)

6.1 Stating the general problem and a possible line of attack35

Now that we have computed the new position vector ⃗x2 , we can
start the whole procedure over again by first computing the new
force and acceleration (F⃗2 , ⃗a2 ), then the new velocity (⃗v3 ), and then
finally the new position (⃗x3 ). Schematically, we are proceeding in
steps given by the following flow chart:

In summary, we basically apply the Euler method twice, once for

the velocity and then for the position. Notice also that we update
the velocity first and then use the new velocity to update position
second. It turns out that this order of operation is often superior to
the reverse order in terms of computational accuracy. It is called the
Euler-Cromer method 1 .

1 A.
Cromer, Stable solutions using the Euler Approximation, American Journal of Physics,
49, 455 (1981).
36 Chapter 6 Simulating Central-Force Problems

6.2 A first example: The mass-spring problem

To illustrate the method, let’s get our feet wet with a relatively sim-
ple problem - a mass on a horizontal spring. All we have to remem-
ber is Hooke’s law for ideal springs,

F (x) = −kx, (6.5)

where k is called the spring constant which corresponds to the stiff-

ness of the spring, and x represents the amount of stretch (if posi-
tive) or compression (if negative) of the spring. We should appreci-
ate this formula as the one-dimensional version of F⃗ (⃗r) from before.
Dividing Equation (6.5) by the mass of the object that we attach to
the end of the spring yields the object’s acceleration.

Let’s look at the basic VPython code, shown in Figure 6.1, to see
how the Euler-Cromer method gets implemented in practice here.
The basic steps (outlined in the flowchart in the previous section) are
contained in lines 24 through 27. Line 28 is not absolutely necessary,
but it is included for the purpose of making position and velocity-
time graphs.

Notice also how easy it is in VPython to draw a spring-mass setup

- we basically select a sphere for the end-mass and a helix for the
6.3 Exercises 37

Figure 6.1: The basic code simulating a mass at the end of a spring

spring. The only thing we have to take care of is to deform the helix
in dependence upon the position of the end-mass.

6.3 Exercises

• Run the program in Figure 6.1 and examine the position-time

graph. What does the graph look like? What math function
does it seem to follow? What about the velocity-time graph?

• What can you say about the relationship between the

position-time graph and the velocity-time graph. Is there
38 Chapter 6 Simulating Central-Force Problems

a phase difference between them, and if so, approximately

what is it?

• Find the place in the code where the initial conditions (x0
and v0 ) are specified. Try running the program with a
few different sets of initial conditions. Does the period of
oscillation seem to depend on this choice?

• Find the place in the code where system’s physical parameters

are specified, namely the spring constant and mass. Change
one of these parameters at a time, and observe qualitatively
how it changes the period of oscillation.

• What if you double (or triple) both the mass and the spring
constant? Do the graphs change?

• Are your observations above consistent with the well-known

formula for the period of oscillation, T , given below?

√
m
T = 2π
k

6.4 A second example: Projectile motion with

air drag

As a second example - one that is slightly more difficult while also

showing off some of VPython’s strengths - let us consider a pro-
jectile launched through the atmosphere. Here we don’t want to
neglect air drag, as is customary in introductory physics, and as we
did in Chapter 5. Instead, we recall that a good formula for the drag
6.4 A second example: Projectile motion with air drag 39

force (under certain assumptions) is given by,

1
F⃗D = ρACD v 2 (−v̂), (6.6)
2
where ρ is the density the medium (here, air), A the cross-sectional
area of the projectile, and CD the coefficient of drag. We see that
the magnitude of the drag force is proportional to the square of the
speed, and that its direction is opposite to the motion. This latter
point is mathematically represented by the last term in Equation
(6.6), where v̂ is the unit vector in the direction of instantaneous
motion, or v̂ = ⃗v /|v|.

You might be thinking that the changing direction of the drag force
would be difficult to “handle”, and ordinarily you would be right.
In fact, this facet of the problem coupled with the quadratic depen-
dence on speed makes this problem impossible to solve in closed
form with pencil and paper. So here we now encounter our first in-
stance of a problem that has no closed-form analytical solution and
where we rely on a computer to give us the solution numerically.

We said “ordinarily” because within the VPython environment there

exist high-level commands that will make this problem substantially
easier. Particularly nice are the built-in vector operations that in-
clude easy evaluations of dot and cross-products between vectors, as
well as evaluations of a vector’s magnitude and direction. For our
purposes here, we want to highlight two operations we can perform
on a vector A:

• mag(A) and mag2(A), where the output yields the length and
length-squared of that vector, respectively.
• A.hat, which produces from the vector A its unit vector (by
dividing it by its length). This syntax treats the directionality
(given by the unit vector) as a property of the vector. It is just
like any other property of a vector, such as A.x, which yields
the x-component of the vector A.
40 Chapter 6 Simulating Central-Force Problems

Armed with these commands, implementing Equation (6.6) should

seem a lot more straightforward. Let’s look at the basic code first
(see Figure 6.2).

Figure 6.2: The basic code simulating a the trajectory of a projectile

with air drag

One thing to point out before we delve in is that everything here is

in SI-uits. This means that whenever you encounter a number (as-
signed to a variable), that number should be considered to have the
appropriate SI-unit. So, for example, Line 10 states: speed=10. This
means that we set the speed of the projectile to 10 m/s. Similarly,
Line 12 defines the density, rho = 1.2, as 1.2 kg/m3 (the density of
air). Since the SI-unit system is closed, any calculations that the
code performs will automatically be in the appropriate SI-unit for
that variable.

We can start by looking at Lines 13 through 15. Here we first com-

pute the x- and y- components of the velocity vector from the speed
6.4 A second example: Projectile motion with air drag 41

and launch angle. Line 15 is interesting: ball.v = vector(vx,vy,0).

How should we interpret this statement? First, it is an assignment.
The velocity vector on the right of the equal sign is assigned to a
quantity called ball.v. The notation on the left side of the equal sign
suggests that v is a property of ball. In fact, with this statement we
define the velocity to be a property of the object we introduced ear-
lier in the code and called ball. So now in addition to the properties
pos, size, and color by virtue of the ball being defined as a sphere,
we have added the property v.

What is interesting is that ball.v has itself all the properties that a
vector can have, and so, for instance ball.v.x would refer to the x-
component of the ball’s velocity vector. Similarly, ball.v.hat would
give the unit vector in the direction of motion.

Finally, let’s examine Line 23:

F=vector(0,-m*g,0) - 0.5*C*rho*Area*mag2(ball.v)*ball.v.hat.

You should recognize this as the net force acting on the projectile.
The first part is the gravitational force near the surface of the earth
and the second is the drag force. Notice that VPython automatically
adds these two parts as vectors.

When you run the program you should get something like the fol-
lowing:
42 Chapter 6 Simulating Central-Force Problems

Notice that the code returns the approximate range of the projectile
in the print statement. Also observe the shape of the trajectory. It
clearly deviates from the parabola - the highest point in the trajec-
tory is not reached at the horizontal half-way point, and the descent
is steeper then the ascent.

6.5 Exercises

• Run the program with different launch angles and determine

the angle that gives you the largest range. Without air-drag,
it is fairly easy to prove that that angle is 45 degrees. What is
it now?

• Let’s play a game. The object is to make the range as close to

100 meters as possible. You are only allowed to change the
initial launch angle and speed. What combination gets you
closest to 100 meters? There may be more than one solution.
6.6 Third Example: Trajectories of Planets and Comets 43

• Instead of modifying the initial conditions, as in the previous

exercise, now examine the role of the projectile properties.
The shape of the projectile enters the problem via the drag
coefficient CD ; it can vary from as low as 0.05 (for a very
aerodynamic shape) to about 1.0 (for a cube). Adjust first the
mass and then the drag coefficient, and describe how these
two parameters affect the trajectory.

• Add to the code given above so that you can get two trajecto-
ries on screen simultaneously. These two trajectories should
correspond to two different projectiles (differentiated either
by mass or drag coefficient).

6.6 Third Example: Trajectories of Planets

and Comets

Our final example is also the most famous in the sense of historical
significance - the Keplerian orbits of planets and comets around the
sun. The perfectly circular orbit is one special solution to Newton’s
second law. This is usually demonstrated in introductory physics
by setting the formula for the centripetal force equal to the gravita-
tional force,
mv 2 GM m
= , (6.7)
r r2
We set them equal because gravitation actually provides us with the
centripetal force necessary for moving in a circle. Solving Equation
(6.7) for v yields the orbital speed as a function of the orbital radius,
√
GM
v= (6.8)
r
44 Chapter 6 Simulating Central-Force Problems

Proving that elliptical orbits also satisfy the governing equations is

much more difficult and usually reserved for an junior-level course
in classical dynamics. The same goes for parabolic and hyperbolic
orbits - the remaining cone sections. However, we can explore those
orbits numerically using VPython without any advanced knowledge
of physics.

Imagine for a moment that we had the power to launch a comet at a

particular distance from the sun, call it r0 , as well as with a certain
initial velocity. It is not hard to see that if we chose a velocity of
zero, the comet would head straight for the sun; it would accelerate
in a straight line toward the sun and be swallowed up by it. In fact,
many people unfamiliar with physics think that this scenario is the
only one possible and that circular orbits are only feasible due to
other planets or stars in the picture, or by some other magic (which
is, of course, ludicrous).

But what if we chose an initial velocity at right angles to the line

connecting the comet to the sun? We would at least have a chance
of obtaining a circular orbit, but only if the speed in this direction
matched Equation (6.8). Indeed, this set of initial conditions would
produce a circular orbit.

Now imagine what would happen if the speed that we impart to

the comet at right angles did not match that speed. What if it were
much smaller or larger than what Equation (6.8) demands? It stands
to reason that we would then not recover a circular orbit, but what
do we get instead? Let’s find out by running a numerical simulation!

The basic code is surprisingly short. In this version we decided for

simplicity to set the universal gravitational constant to 1, but you
can change it to the actual value in SI-units. In that case, however,
you should also make the masses and distances involved realistically
large. The basic code, then, is shown in Figure 6.6. Feel free to make
6.7 Exercises 45

the parameters more realistic; the mass ratio that appears here is
only 1:100, for instance. Nonetheless, we can use this simplified
code to explore the possible orbits.

Figure 6.3: The basic code simulating gravitational orbits.

6.7 Exercises

• Examine the code provided, make sure that you understand

what each line does, and annotate.

• Use Equation (6.8) with the parameters given in the code to

find the orbital speed for a circular orbit. Enter this speed as
the starting speed in the code at the appropriate place. Do
you get a circular orbit?

• If you doubled the initial distance of the comet from the sun,
what speed would be necessary now for a circular orbit? Try
it in the code.
46 Chapter 6 Simulating Central-Force Problems

• Now select a starting speed that is smaller than the one

you chose above. Observe the kind of orbit you obtain now.
Does the orbit trace out the same path after each revolution?
Astronomers call this scenario a closed orbit.

• What can you say about the orbital speed? Is it constant or

does the comet appear to speed up at certain points? Explain!

• If you make the initial speed too small, you will see that
the simulation eventually breaks down and returns non-
sensical results. Why is that and what can you adjust to
remedy the situation? [Hint: When the comet gets very
close to the sun, the acceleration it experiences becomes very
large. What line in the code is going to be adversely affected?]

• Let’s be even more creative. Instead of the normal law of uni-

versal gravitation that decreases with the square of distance,
what would happen in a universe ruled by a gravitational
force proportional to 1/r? Make this change in the code, and
start with a circular (or near circular) orbit by using Equation
(6.7) but with the new gravity to solve for orbital speed.

• Now lower the speed and observe the orbits that result. What
is the most obvious qualitative difference about the orbits that
we get now in this alternative universe?
Conservation of Momentum
and Energy

7.1 A binary-star system

In the previous chapter we considered the motion of a planet subject

to a central force that was always directed from the planet towards
the origin where a star was sitting motionless. This is, of course,
not quite correct, since the star also experiences a gravitational pull
towards the planet and must therefore also move. This motion is
usually very small since stars tend to be much heavier than planets,
but it is not zero.

Let’s consider now the full two-body problem. To give us maxi-

mal flexibility in parameters, we can consider the two bodies both
stars in a binary-star system. We will creatively call the heavier
star “bigstar” and the lighter star “smallstar”. We could, of course,
proceed exactly like before, modifying the code in Chapter 6 to in-
clude the dynamics of the second object (i.e., the sun). But to illus-
trate another important physical concept, namely that of momen-
tum and momentum conservation, let us incorporate this quantity
in the code.

To review briefly, momentum is defined by p⃗ = m⃗v . An object

has momentum by virtue of having mass and velocity. We can also
talk of the total momentum of a system of objects. In that case we
simply add the individual momenta of all the objects that make up
the system together (as vectors). Now, we can prove that when this
system is isolated, in other words when nothing outside this system
exists that would push or pull on the objects within the system, then
the total momentum of the system doesn’t change. We say that the
total momentum is conserved.
48 Chapter 7 Conservation of Momentum and Energy

Another useful concept in this context is the center of mass of a sys-

tem comprised of discrete objects/particles. We can compute the
coordinates of the center of mass using the well-known formula,

m1 x1 + m2 x2 + ...
xcom = ,
m1 + m2 + ...
and similarly for the y-coordinate. Since VPython deals well with
vector quantities, we can also dispense with the individual compo-
nents and refer only the position vectors, ⃗ri , of the objects them-
selves: ∑
mi⃗ri
⃗rcom = ∑ . (7.1)
mi

You may have learned that whenever the total momentum of a sys-
tem is conserved, i.e., when there is no net external force acting on
the system, then the center of mass cannot experience any accelera-
tion. In our notion, we can write,

d2
⃗acom = (⃗rcom ) = 0. (7.2)
dt2

The proof of this statement is actually not difficult and well within
the reach of an introductory physics student. See if you can work it
out yourself.

One of our objectives will now be to “prove” this property about

the center of mass numerically by simply computing its location for
each time-step and then to observe its motion across the screen. An-
other will be to show that in this binary-star system, total momen-
tum will be conserved. It should be conserved because the way we
have set up the problem, the system is clearly isolated - there are no
other objects around at all.
7.1 A binary-star system 49

Figure 7.1: The code for the binary-star system.

Let us again start with a code template - see Figure 7.1. One thing
you will notice immediately is that we are now not working in ar-
tificial units (as before) and instead use the true value for the gravi-
tational constant. That also forces us to use astronomically realistic
values for all other quantities in the problem, such as for the masses
and distances involved. Again, as long as all the inputs are in SI-
units, so will all the computed outputs.

As you can see, in the first six command lines, we basically set up
the properties about the two stars. For instance, the bigger star has
a mass of 3 × 1030 kg - three times that of the smaller star. For
50 Chapter 7 Conservation of Momentum and Energy

comparison, our sun’s mass is about 2 × 1030 kg. One of the things
you will have a chance to play around with is the mass ratio of the
two stars.

The next line (Line 17) defines the computational time step. This
might strike you as an incredibly large time step, especially com-
pared to the values we have used before, but remember - everything
has to be astronomical, including time. The time step of 105 sec-
onds translates to a duration of slightly longer than an earth day.
Compared to the period of revolution this is still quite small.

Lines 19 and 20 are recognized from Equation (7.1) as nothing other

than ⃗rcom . The next two lines are there just so we can visualize the
location of center of mass on the screen, here in the aspect of a green
cone whose trajectory we will also keep track of via the “make_trail”
command.

The iterative part of the program starts with Line 25. We define the
vector, r, that runs from the small star to the big star. We could have
also reversed it, but it is important to be consistent. The way it is
defined, it will be parallel to the force on the small star and anti-
parallel to the force on the big star. By Newton’s third law, these
two interaction forces must be equal and opposite.

Lines 30 and 31 implement Newton’s second law,

d⃗
p
F⃗ = . (7.3)
dt
If we solve this for the small change in momentum, we get: dp =
F dt; in other words, it takes a force to change the momentum. This
is also sometimes referred to as the impulse-momentum theorem.
Once we have calculated the new momentum, we can also update
the position of the corresponding star, since momentum, of course,
is intimately related to velocity - we get velocity by dividing the mo-
mentum by mass. This way, we do not have to explicitly refer to
7.2 Exercises 51

velocity at all in the WHILE loop.

7.2 Exercises

• Run the program - does the computed total momentum

change over time? Is physics correct? Also examine the code
- is the result at all surprising?

• You will also notice that the center of mass does not stay
stationary, but that it moves on the screen. We can make that
motion cease if we give the two stars initial momenta that
add to zero. Show that right now (i.e., in Figure 7.1) they two
initial momenta do not add to zero.

• Now change the initial conditions such that the total mo-
mentum is in fact zero initially. What do you notice about
the center of mass motion on the screen? Also change
the viewer’s perspective to see the two-body motion from
different angles.

• Make an additional change in initial conditions (positions

and velocities) and watch the orbits of the two stars around
each other. Consider turning on the “trails” on the orbits for
better visualization.

• VPython allows you to view the motion from the perspective

of different observers (or reference frames). The command is
scene.camera.follow(bigstar). In the parenthesis appears
the name of the object that you want to make as the reference
frame. In the command above we take the big star as the
observer’s reference frame. To make things less confusing,
52 Chapter 7 Conservation of Momentum and Energy

turn off all the trails again.

• Now that you have explored the role of initial conditions a

bit, let’s turn to the mass ratio. Try one case where the ratio
is much larger than 3, and then one case for the ratio is close
to 1. What do you conclude from those two scenarios.

7.3 Energy in the Spring-Mass System

In Section 6.2, we explored the motion of a mass at the end of a

spring. We saw the sinusoidal motion that resulted from the com-
bination of Newton’s second law and Hooke’s law. Now we can
revisit this problem and analyze it through the lens of potential and
kinetic energy.

Remember that the potential energy stored in a stretched or com-

pressed spring, also called elastic potential energy, is given by U =
1
2 kx . Furthermore, the kinetic energy of the end-mass is given by
2

K = 12 mv 2 . Finally, the total mechanical energy is simply the sum

of the two energies, Emech = U + K.

Modify the code, as given in Figure 6.1, in the following manner.

• Add some lines that will compute the three energies, U , K,

and Emech .

• Then, instead of plotting position and velocity as a function

of time, instead plot the three energies that are now being
computed at each time step. What do you observe? Is the
total mechanical energy conserved over time? Explain!
7.4 Simulating the Rutherford experiment 53

7.4 Simulating the Rutherford experiment

Another example that we will see is actually mathematically very

similar to the binary-star system is the famous Rutherford scattering
experiment. We know that two like charges repel via the Coulomb
force. This force has the same structure as the law of universal grav-
itation. It is given by,
q1 q2
F⃗c = k 2 r̂, (7.4)
r
where the Coulomb constant k = 9.0 × 109 N m2 /C2 plays the role
of the gravitational constant, G. Notice how this force also depends
inversely on the square of the distance between the two objects, and
that it is again a central force (indicated by the direction vector r̂).

Ernest Rutherford’s experiment was to shoot alpha particles, which

are fast Helium nuclei, at a thin gold foil. The idea was to see if
and how the alpha particles would be deflected upon hitting the foil.
What Rutherford discovered was that every once in a while the de-
flection angle was very large so that the alpha particle was scattered
back to the source. The explanation for these large deflections is that
the positive charge in the gold atoms is actually highly concentrated
in the nucleus (and not smeared out). In fact, while the size of the
gold atom is on the order of 10−10 meters, the nucleus is 100,000
times smaller, i.e., on the order of 10−15 meters.

So if the positively charged alpha particle hits a gold nucleus close

to head-on, since the two repel it should be deflected back. At the
same time, the Gold nucleus should also experiences some recoil. (If
it didn’t, momentum could not be conserved during the collision.)
Let us now use VPython to explore the different scenarios that can
54 Chapter 7 Conservation of Momentum and Energy

occur when we shoot an alpha particle at a gold nucleus.

While Equation (7.4) is very similar in form to Equation (6.7), the

length scales of the nuclear-collision problem could not be more dif-
ferent from that of the binary-star problem. In the nuclear prob-
lem, a typical distance (or characteristic length scale) is on the or-
der of 10−15 m, whereas in the gravitational context it was 10+11
m - an incredible difference of 26 orders of magnitude. With this
change in spatial scale come other difference, such as in characteris-
tic mass and time. In short, we are now dealing with vastly smaller
lengths, vastly tinier masses, and vastly shorter times. Luckily for
us, VPython can handle all of these changes very well.

Figure 7.2 shows the new setup for the nuclear context.1 The first
thing we notice is the new scene.range which catapults us into the
microcosm. Next we define the nucleus and its properties of mass,
charge and momentum.

In Line 13, we introduce the so-called impact parameter, b. This will

be an important control parameter for us to vary from run to run.
This parameter tells us, roughly speaking, by how much the alpha-
particle would miss the nucleus if it felt no repulsive force. Thus,
b = 0 corresponds to a perfectly head-on collision.

Next, we define the alpha particle and its properties of mass, charge
and momentum. Notice that we give the alpha particle some ini-
tial speed to the right. In fact, this initial speed is quite large,
v0 = 5 × 107 m/s, about 17 percent of the speed of light. We need
large speeds for the alpha particle to get close to the gold nucleus.
This initial speed is another parameter we will be able to adjust later.

Line 22-25 set up arrows for the momenta of the two particles (alpha
1 This code is loosely based on the code by R. Chabay, “03-particle collision”, available in the

Matter-and-Interactions’s Glowscript-programs folder.

7.4 Simulating the Rutherford experiment 55

Figure 7.2: Setting up the Rutherford-scattering code

and Gold nucleus) and the total momentum. In order to see these
arrows on the screen, they have to be stretched by the scale-factor,
scale.

Finally, we have added here for the first time a start-pause button,
which will allow us to start the code and pause it at any time. The
details of this section of code need not concern us here.

The main part of the code, namely the WHILE loop, is shown in Fig-
ure 7.3. We should recognize large chunks of it from before. Lines
52 to 55 are new. They will update the momentum vectors through-
out the interaction of the two particles. The lengths and direction of
the three momenta (alpha, nucleus, and total) are set via the .axis
assignment. Additionally, we have displaced the tail of the momen-
tum vector associated with the nucleus to coincide with the head of
56 Chapter 7 Conservation of Momentum and Energy

the alpha’s momentum vector. This way we can verify visually that
the two individual momentum vectors do indeed add to the same to-
tal momentum vector (via the head-to tail method of adding vectors).

Figure 7.3: The core of the Rutherford-scattering code

7.5 Exercises

• Run the code and observe the trajectories of the two par-
ticles. Also observe the particle’s momentum vectors in
the upper-right corner. You should see something like this:
7.5 Exercises 57

• Verify that the momentum vectors are in fact tangentially

parallel to the trajectory at any given instant of time. Why is
the red arrow so large if the gold nucleus seems to move so
slowly?

• Given the vector diagrams rendered in the upper-right

corner of the screen, is momentum conserved throughout the
scattering process? How do you know?

• Adjust the impact parameter, b. Make it progressively

smaller (all the way to zero) and observe what happens to the
deflection angle of the alpha particle.

• Now make b incrementally bigger than the value in Figure

7.2. What do you see now? Is momentum still conserved?

• You can now explore the role of the initial speed of the alpha
particle. How does the angle of deflection seem to depend on
this initial speed?
58 Chapter 7 Conservation of Momentum and Energy
Rotational Motion, Torque, and
Angular Momentum

In the previous chapters you saw how a binary star system or a

comet can be simulated in VPython. In this chapter, we will review
the motion of a comet, and explore torque and angular momentum.
The physics of a binary star system and a Sun-comet system are
exactly alike. As a matter of fact, the same physics governs any
gravitational interaction (in a classical mechanics sense): a cluster
of stars, a moon and its planet, or a cluster of galaxies.

8.1 The Physics

Recall Newton’s Law of Universal Gravitation:

GM m
F⃗ = − r̂ (8.1)
r2

(Note this looks similar to Equation (7.4), the Coulomb’s Law force
between two charges).

Note that the direction of the position unit vector r̂ is always point-
ing in the exact opposite direction (because of the minus sign) of the
force vector.

Angular momentum (symbol L), ⃗ is the physical quality that is the

rotational equivalent of linear momentum. Remember that linear
momentum p⃗ = m⃗v . Angular momentum L ⃗ = I⃗ω , where I is the
moment of inertia, and ω⃗ we encountered way back in Chapter 3 - it
is called the angular velocity. Torque (symbol ⃗τ ) can be thought of
as the rotational equivalent of force. It is determined from the cross
 Chapter 8 Rotational Motion, Torque, and Angular
60
Momentum

product ⃗r and F⃗
⃗τ = ⃗r × F⃗ (8.2)
Equation (8.2) shows that the torque is the cross product of the
position vector ⃗r and the force applied F⃗ . The cross product is a
vector. Remember that the magnitude of the cross product here is

|⃗r||F⃗ | sin θ,

where θ is the angle between ⃗r and F⃗ . Its direction is perpendicular

to both of the two vectors in the cross product. Here, therefore, the
torque ⃗τ is perpendicular to both ⃗r and F⃗ .

It is very easy to program the cross product in VPython. To calculate

the cross product of A ⃗ and B,
⃗ one uses the command

cross(A,B)

Remember that A and B must have been declared to be vectors in

VPython. VPython will know automatically that the value assigned
to cross(A,B) is a vector implicitly.

Just as force is the change of linear momentum with respect to time,

see Equation (7.3), torque is the change of angular momentum with
respect to time:
dL⃗
⃗τ = (8.3)
dt

Furthermore, just as linear momentum is conserved in the absence

of an outside force, angular momentum is conserved in the absence
of an outside torque. Note that this follows from Equation (8.3).

Look back at Equation (8.1). Recalling that the unit vector r̂ is ro-
tated 180◦ from F⃗ , what is the value of ⃗τ ?
8.2 Exercises 61

Figure 8.1: Program for Comet Orbiting the Sun

⃗ must be conserved.
Because τ = 0, then L

See the sample program and output in Figure 8.1.

8.2 Exercises

• Add to the code in Figure 8.1 to calculate the torque and plot
it (remember it is a vector and you can print the components
of a vector!).
 Chapter 8 Rotational Motion, Torque, and Angular
62
Momentum

Is it pointing in the correct direction?

Show that the program is calculating a torque that is close
to the correct answer (it may not, due to limitations in how
precisely VPython does calculations).

• Write code to calculate the angular momentum and plot

it. Is it pointing in the correct direction? Print the angular
momentum with every iteration. What do you notice? Does
it make sense?

• Why is the Sun not appearing to move?

• Change the parameters of the program to illustrate another

scenario (for example, the Earth revolving around the Sun).
Discuss what you see.
Two Capstone Projects

9.1 The Gravitational “Slingshot”

In the previous chapters of this book you have gone from investigat-
ing the simple motions of an object with constant velocity to explor-
ing the physics of a comet about the Sun. While these phenomena
are “real life”, we want to end with a consideration and discussion of
a truly “Space Age” phenomenon that is used regularly in sending
space probes to the outer solar system – gravitational assist, or
the “slingshot” maneuver.

The outer planets – Jupiter, Saturn, Uranus, and Neptune – as well

as the objects in the Kuiper Belt – including Pluto, Makemake, and
2014 MU69 – are extremely far away. Jupiter is just over 5 astro-
nomical units1 away from the Sun, Neptune is 30 a.u. away from
the Sun, and Pluto is 40 a.u. away from the Sun. Traveling to any of
these objects will take several years to get there.

But how to get there? Traveling directly away from the Sun to ar-
rive at a distant planet would take an extraordinary amount of fuel
to travel – and that extra fuel means more mass, and the more mass,
the more fuel needed. It would be extremely costly, and at this writ-
ing we do not have the technology to develop rockets to transport
a space probe those distances. The Hohmann orbit (least energy
method) saves on fuel, but takes quite a while to complete the jour-
ney.

Gravitational assist allows a space probe to take some of the energy

from a nearby planet and use it for itself. The energy is “free”!

1 An astronomical unit is the average distance of the Earth to the Sun. It is equal to 1.50×108
kilometers.
64 Chapter 9 Two Capstone Projects

9.1.1 The Physics

The slingshot maneuver involves a planet and a space probe. Both

objects have a velocity (therefore, both a speed and a direction).
When the probe nears the planet and is appreciably affected by the
planet’s gravity, it accelerates (both in speed and direction). The
interaction is a “collision”, but generally nothing collides. Further-
more, the collision is elastic, so therefore there is no loss of energy
(and, of course, no change of total momentum) in the planet-probe
system.

So, if the kinetic energy of the probe increases, then the kinetic en-
ergy of the planet decreases. The planet slows down! However, the
amount of decrease of kinetic energy is negligible to the planet, be-
cause the mass of a planet is huge compared to the mass of a space
probe. For example, the mass of Uranus or Neptune is approximately
1026 kg, compared to the mass of a space probe, which is typically
on the order of 103 kg.

In the rest frame of the Sun, the Sun “sees” the planet moving with
⃗i . The space probe is moving with initial velocity
initial velocity U
v⃗i . The two objects are not necessarily going to meet head-on; that
would certainly result in the loss of the spacecraft. Instead, their
velocity vectors are separated by a distance b, called the impact
parameter. If b ≤ the radius of the planet Rp , then there would
be a head-on collision.

Consider a planet and a space probe. In the Sun’s frame of reference,

the probe has velocity ⃗v and the planet has velocity U ⃗ (see Figure
1
9.1). Therefore the probe’s kinetic energy is 2 mv . In the planet’s
2

frame of reference, the planet sees the probe approach with speed
v + U . As it gets closer, the speed of the probe increases, and after
the encounter, the probe’s speed decreases until it appears to the
planet to have a speed v + U again, as shown in Figure 9.2. (We are
9.1 The Gravitational “Slingshot” 65

Figure 9.1: Space Probe Approaching Planet

66 Chapter 9 Two Capstone Projects

Figure 9.2: Space Probe After Experiencing Gravitational Assist

From Planet

assuming an ideal encounter that causes the probe to make a 180◦

change of direction). Therefore, in the Sun’s frame of reference, it
appears as if the probe has speed v + 2U .

In the Sun’s frame of reference, then, the kinetic energy of the probe
is 21 m(v + 2U )2 . The increase of energy to the probe is 2mU (v+U ).
(By the same token, because the energy is conserved in this elastic
collision, the planet’s kinetic energy goes down by the same amount,
but the change is negligible for the planet).

Several probes to the outer Solar System have used gravitational

assist. They include the Cassini mission to Saturn (see Figure 9.3),
the New Horizons mission to Pluto and 2014 MU69 , and the famous
9.1 The Gravitational “Slingshot” 67

Figure 9.3: Trajectory of the Cassini Mission to Saturn (source:

NASA)

probes Voyager 1 and 2, which are now in interstellar space.2

9.1.2 The Code

In Figure 9.4 the kinetic energies of the probe and the planet are
p2
calculated (e.g., in line 11) as 2m , which is the same as 21 mv 2 . In line
27, the gravitational force is calculated, and lines 34 and 35 update
the planet’s and the probe’s positions at each instant.

2 Voyager
1, launched in 1977, is now almost 150 a.u. from the Sun and is the farthest human-
made object in existence.
68 Chapter 9 Two Capstone Projects

Figure 9.4: Sample Code for Gravitational Assist

9.1 The Gravitational “Slingshot” 69

9.1.3 Exercises

1. • Run this program to make sure it works. Describe what

you see.

• Modify the code to incorporate an impact parameter, b.

• What happens when you change the impact parameter,
to make it larger or smaller?

• What happens when you change the velocities of

the probe or the planet? In particular, what happens
to the speed of the probe when the planet’s velocity is 0?

• Similarly, what happens when the planet’s velocity

changes sign? In that scenario, the probe and the planet
are initially moving in the same general direction, with
the probe catching up to the planet.

2. • Now that you have explored the b-dependence a bit,

let’s make this a bit more precise. Choose five different
impact parameters and keep track of the angle of
deflection of the probe for each. Enter the values for
b and deflection angle into a spreadsheet and plot the
relationship between those variables. If the graph is not
clear, go back and add a few more b-values into your
table.

• For the same b-values that appear in your list, write

down the the change in kinetic energy of the probe, ex-
pressed as a fraction of the initial kinetic energy, based
on the code output. Then enter these values for frac-
tional change in kinetic energy into your spreadsheet
and plot its dependence on impact parameter.
70 Chapter 9 Two Capstone Projects

Based on this plot, if we want a large increase in kinetic

energy for the probe, what do we have to do?

9.2 Asteroid near a Binary Star System -

Chaotic orbits, or the “Three-Body
Problem”

In the previous chapters you have gone from investigating the sim-
ple motions of an object with constant velocity to exploring the
physics of a comet about the Sun. In all of these examples, the
motion did not exhibit what physicists call deterministic chaos. In
the gravitational problem of two bodies attracting one another, solu-
tions can, in fact, be found by pencil and paper. Interestingly, when
we add a third body all bets are off. The orbits can no longer be pre-
dicted with a mathematical formula, but worse than that, the orbits
exhibit what is known as extreme sensitivity to initial conditions - a
hallmark of chaos.

9.2.1 Modifying a previous code

To keep things as simple as possible, let us revisit an earlier problem

- the binary-star system in Section 7.1. The question we want to
ask now is: if we introduced a small asteroid into this system how
would it move under the influence of the gravitational attraction to
each star? To make the problem even simpler, we assume that the
asteroid is so small that it does not affect the motion of the two stars.
Since it is 8 orders of magnitude (or 100 million times) less massive
than the stars, this is a very good approximation. In that case, the
mass of the asteroid really doesn’t enter into the picture. All we need
to do is to calculate the acceleration experienced by the asteroid at
9.2 Asteroid near a Binary Star System - Chaotic orbits, or the
71
“Three-Body Problem”

every time step, and this is given by the universal law of gravitation
as,
m1 m2
⃗a = −G 2 r̂1 − G 2 r̂2 , (9.1)
r1 r2
where ⃗r1 is the vector from first star to the asteroid, and ⃗r2 is the
vector from the second star to the asteroid, and the m’s in the nu-
merator are the star masses.

Once we have computed the accelerations, we can update the aster-

oid velocity, and then we use the updated velocity to find the next
position of the asteroid, in a recursive procedure outlined in Chapter
6.

So here is the idea. Let us start with the code in Figure 7.1 which
will take care of the motion of the two stars. To give us a little more
“room”, let’s start these two stars out along the x-axis at -5e11 and
+5e11 (instead of ±2e11. Also, to keep the center of mass of the
system fixed, let’s adjust the y-component of the small start from
-1e4 to -9e3. Finally, turn off the trails on these two stars, and delete
all calculations of the center of mass (we won’t need that anymore).
This takes care of the binary-star system.

Now let’s introduce the asteroid. Add a line defining this object
near the top that defines the asteroid as a sphere. Give it an initial
position of vector(1.50e12,0,0). Then define a asteroid ve-
locity as one of its attribute as
asteroid.v = vector(0,-7.5e3,0).

Now that we have defined the asteroid and its initial position and
velocity, we need to turn our attention to its motion. At the end of
the while-loop, add two lines that compute the vectors ⃗r1 and ⃗r2 ,
appearing in Equation 9.1. Each of them are just the differences in
the position vectors of the asteroid and the respective star. With
these two vectors in hand, we can compute the asteroid’s accelera-
tion, asteroid.a, using Equation 9.1 directly.
72 Chapter 9 Two Capstone Projects

Figure 9.5: A rudimentary adaptive time step

This program would now run, but it is a good idea to add two more
things. The first one is straightforward. If the asteroid comes too
close it burns up in the stellar atmosphere. So we should include an
if-statement to terminate the program if the asteroid comes within,
say, 2e9 (or 2 million kilometers) of either star.

The second refinement has to do with the computational integration

technique itself. Right now we are using a simple Euler-Cromer al-
gorithm, as explained before. But you may have noticed that when
the asteroid gets really close to either star, the dynamics is no longer
smooth. The asteroid picks up so much speed that it covers a signifi-
cant amount of space between successive time steps. In other words,
the computational time interval, dt, seems to be too large, and the
result is no longer accurate.

In a more advanced course, you would now probably be learning

about the Runge-Kutta algorithm with an adaptive time-step. But
that would go beyond the scope of this course. So, instead, let’s im-
plement a most rudimentary version of an adaptive time-step. The
basic idea is that when the asteroid comes within a certain radius of
either star, then we need to compute more finely by reducing the dt.

Figure 9.5 shows one simple way we can accomplish this. Keep in
9.2 Asteroid near a Binary Star System - Chaotic orbits, or the
73
“Three-Body Problem”

mind, however, that this is not a highly accurate algorithm either,

and that there are much more sophisticated ways to do this out there.
Nonetheless, it improves on our previous method.

9.2.2 Exercises

Now that you have completed the coding, let us have fun and run
the program with a few initial conditions. You should get output
like the following:

One thing that is immediately clear is that the orbits are highly ir-
regular and non-periodic. Furthermore, what you can do is to vary
the initial position of the asteroid slightly. For instance, you can
check the orbits that result from a initial x-coordinate of 1.49, 1.50,
and 1.51 × 1012 meters.

Find initial conditions that lead to the three possible long-term out-
comes for the asteroid: (a) it crashes into one of the stars, (b) it
74 Chapter 9 Two Capstone Projects

escapes the binary-star system and drifts further and further into
space, and (c) it settles into a semi-stable orbit around one of the
stars.
Conclusion

We hope that the previous chapters helped you learn the material in
your introductory physics textbook a little better by giving you the
opportunity to explore and play with the ideas in a concrete way.
We also hope that these exercises made you appreciate the power of
scientific computation. It can be an invaluable tool in finding and
visualizing solutions and also in helping us build up our physical
intuition and our conceptual understanding of physics.

We encourage you to use the sample codes provided in this book

as a springboard for developing your own codes. These programs
are pretty barebones and could be embellished in numerous ways.
More importantly, we hope that you use the tools you have acquired
to simulate your own favorite physical processes. Play around with
the code! Tinker with it, and see what happens!

Keep VPython in mind also for the second semester of your intro-
ductory physics sequence. There should be plenty of chances to use
it for, say, visualizing electric and magnetic fields and so on. In
fact, there are a number of excellent sample programs already on
the glowscript.org platform. Additional programs and resources can
be found at multiple online repositories.1

Finally, if you think you want to pursue the physics major, remem-
ber VPython can come in handy for your upper-level physics course-
work. At that time, it is likely that you have become more familiar
with scientific computing and may be comfortable using the Python
programming language. You can choose from a number of excellent

1 see,
for instance:
https://vpython.org/contents/contributed.html
https://bphilhour.trinket.io
https://phys221.wordpress.com/vpython-tutorials/
76 Chapter 10 Conclusion

recent textbooks on computational physics using Python.2 In fact,

we consider VPython an excellent gateway into the broader world
of computational physics and scientific computing.

2 For instance:
R.H. Landau and M.J. Páez, Computational Problems for Physics: With Guided Solutions
Using Python (CRC Press, 2018).
J.M. Kinder and P. Nelson, A Student’s Guide to Python for Physical Modeling (Princeton
U. Press, 2018).
M. Newman, Computational Physics (2013).
Bibliography

R. Chabay, B. Sherwood, Matter and Interactions (John Wiley & Sons,

2015).

R. Chabay, B. Sherwood, “Computational Physics in the introduc-

tory calculus-based course,” Am. J. Phys. 76, 307 (2008).

A. Cromer, “Stable solutions using the Euler Approximation,” Amer-

ican Journal of Physics, 49, 455 (1981).

J.M. Kinder and P. Nelson, A Student’s Guide to Python for Physical

Modeling (Princeton U. Press, 2018).

R.H. Landau and M.J. Páez, Computational Problems for Physics:

With Guided Solutions Using Python (CRC Press, 2018).

P. Laws, Workshop Physics Activity Guide, Mechanics, Module 1 and

2 (Wiley, 2004).

L.C. McDermott, “A View from Physics” in Toward a Scientific Prac-

tice of Science Education (Lawrence Erlbaum Assoc., 1990).

M. Newman, Computational Physics (2013).

E.F. Redish and J.M. Wilson, “Student programming in the introduc-

tory physics course,” Am. J. Phys. 61, 222 (1993).

D. Schroeder, Physics Simulations in Python (2018).

D. Sokoloff, R. Thornton, P. Laws, RealTime Physics - Active Learning

78 Chapter 10 Conclusion

Laboratories (John Wiley & Sons, 2012).

A. Van Heuvelen, “Learning to think like a Physicist: A review of

research-based instructional strategies,” Am. J. Phys. 59, 891 (1991).
Authors

Windsor A. Morgan, Jr. is an Associate Professor of Physics &

Astronomy at Dickinson College. He specializes in astronomy edu-
cation research, and has also done research in X-ray-emitting active
galactic nuclei, X-ray binary star systems, new statistical methods of
studying astronomical surveys, and the formation of hydrocarbons
in the early solar system.

Lars Q. English is a Professor of Physics at Dickinson College.

His research explores “the nonlinear dynamics of complex systems”
where both nonlinearity and lattice/network geometry play impor-
tant roles in enabling and guiding processes of spontaneous patterns
formation. Other interests include the Calculus of Variations, mag-
netism and spin resonance, microwave spectroscopy, medical imag-
ining techniques, and issues within the philosophy of science. Visit:
http://www.larsenglish.com