1 Laboratory Guide - Physics

G. Pearson, 2021

1.1 Lab Objectives

Experiment topics are chosen to explore interesting examples of physics systems.

Labs are meant to be experiential. They're an opportunity for students to develop hands-on
familiarity with equipment and tools, measurement methods and techniques, and to practice
report-writing.

Typically experiments require the investigator to:

Measure parameters of a system. An independent control variable is adjusted in order to

observe its effect on one or more dependent variables.
Record and tabulate the measured values.
Present results graphically in such a way that important dependencies between
parameters are visualized.
Document realistic uncertainties in measured values, and calculate how these
uncertainties affect confidence in any results produced.
Draw a conclusion.

That is to say, labs provide valuable experience in designing experiments, in collecting and
analyzing data, in the presentation and visualization of results, and help develop communication
skills critical in writing clear and readable reports.

Experiments cover topics from course content, however the content may not have been reviewed
in class prior to performing the lab. Some independent preparation is therefore required in advance
of each lab.

Sample Report. An example report from an experiment investigating the oscillatory behaviour
of a rigid rod "physical pendulum" is included with this guide.

1.1.1 Laboratory Process

At their core, physics labs are meant to explore simple relationships. However, performing
experiments, analyzing results, and report writing can be detailed, time-consuming activities.

To develop proficiency, students are encouraged to take a simple, systematic approach.

 Review lab manuals in advance to identify the main relationship being studied.
You might be guided through the steps of an experiment required to collect data. Or you
might be asked to consider a relationship and design your own experiment to measure the
co-dependence of two or more variables.

Identify the key variables you'll study. Usually you'll control one variable and measure
another.
You should be able to anticipate and prepare your data tables in advance of the lab
session.
Organize your results clearly in a table, including values calculated from measurements.
Wherever possible, multiple trial measurements should be taken to ensure
repeatablility.
Average values and standard deviation data can then be included in results tables
within the body of the report.
Raw data tables are often best included in an Appendix rather than within the main
report.

Prepare a graph (or graphs) illustrating the dependency between variables measured.
Graphs should include "error bars" indicating measurement uncertainties.
"Best-fits" to the data should go through most error bars.
The slope of a line, or a fit parameter, often contains interesting information. Compare to
the theory to see if interesting results can be extracted.
Where appropriate, plot multiple sets of data in the same graph, and include a legend so
that they're identifiable.
Graphs should always include a clear title and axis labels.

Examples of all calculations should be provided, including examples of uncertainty

propagation.
This is not to say "all" calculations need to be shown. Examples of each type should be
included with sufficient detail to show your instructor the steps of your analysis work.
𝑐
Where a specific value is determined, like for example the speed of light, , or gravitational
𝑔
acceleration, , include a comparison to the accepted scientific value. This is often
provided as a "percentage difference" calculation.

Provide analysis and discussion of your main results.

Summarize your main results.
Discuss how well the results agree with theoretical expectations.
Be sure to comment on sources of uncertainty, and whether they fully account for any
discrepancy between results and expectations.
Consider, how might the experiment be refined to improve agreement?

By the end of your lab session you should aim to collect your data, complete rough calculations,
and plot rough versions of any graphs in order to confirm that your results align with your
predictions.

After your lab session plan to spend a few hours writing the final version of your report. This will be
your "good" version of tables, calculations, graphs, as well as a few sentences or paragraphs to
discuss your findings and your main conclusions.
 Simplicity is best. When writing your report, keep paragraphs focused. Students sometimes
include rambling, repetitive sentences in their work to add length, however this can seriously
detract from the readability of their analysis. Keeping sentences clear, concise, and to-the-
point is always the best way to communicate that you've fully understood the purpose and
outcomes of your your experiment.

1.1.2 Report Writing

Reports should include:

Title Page
Include: Report Title, Your Name, Course, Section Number, Instructor, TA Name, and Date
of Submission.
Introduction
What is the experiment's objective?
Theory
You may be able to show a derivation of the physics you're investigating, or you may want
to reference a source that provides a description/equation representing the physics you're
investigating.
You may want to provide diagrams and/or graphs that illustrate/predict how you expect
the system under study to behave.
Procedure
Briefly explain the systematic steps required to take any measurements.
Results and Calculations
Tabulate your measurements in an organized manner.
Based on your procedure, you should know what your tables will look like before you take
your measurements.
Graph your main results.
Provide examples of any calculations.
Discussion and Conclusions
Discuss the main observations and outcomes of your experiment.
Summarize any significant conclusions.
References
Appendices

Style. Use the past tense, and write in the passive voice or use the royal "we." (I.e. Avoid use
of personal pronouns like "I").

Academic Misconduct. In accordance with Policy 60, it is considered academic misconduct

to misrepresent the work of others as your own. Plagiarism risks serious academic penalties.

1.2 Measurements and Calculations

1. Notation

𝑎
Where a measured value, , has an associated uncertainty, Δ𝑎 , then the measurement can be
written as

𝑎 ± Δ𝑎.
This notation should be used in calculations and written reporting.

2. Measurement Uncertainty

Measurement uncertainty is often situational. Certain "rules of thumb" are useful, but may not
apply to all experiments.

Instrument-Limited Uncertainty. For precisely-defined measurements use "half the

smallest increment" of the measuring tool.
Sample-Limited Uncertainty. When a measurement is naturally "fuzzy," the experimenter
should choose an appropriate uncertainty value that estimates the variation in the sample
or sample population.
Statistical Uncertainty. See Mean and Standard Deviation.

3. Mean and Standard Deviation

When possible, repeat and record several trials of a measurement. This should verify that
results are repeatable and consistent.

𝜇 𝑁 measurements of 𝑥 is given by:

Recall that the mean value, , of

𝜇 = ∑𝑥 𝑁
𝑖

and the standard deviation, 𝜎, of the measurements is given by:

⎯∑(
⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
𝑥 − 𝜇 ) 2⎯
𝜎=√ 𝑁 . 𝑖

The mean and standard deviation of these trials can be used as 𝜇 ± 𝜎 = 𝑎 ± Δ𝑎 in any
additional calculations.

4. Uncertainty Propagation

Addition and Subtraction:

Recall that for simple measurements, in the case of addition or subtraction,

(𝑎 ± Δ𝑎) + (𝑏 ± Δ𝑏) = 𝑐 ± Δ𝑐
and

(𝑎 ± Δ𝑎) − (𝑏 ± Δ𝑏) = 𝑐 ± Δ𝑐
where

Δ𝑐 = √⎯Δ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
𝑎2 + Δ𝑏2⎯.
 Multiplication and Division:

In the case of multiplication or division,

(𝑎 ± Δ𝑎)(𝑏 ± Δ𝑏) = 𝑐 ± Δ𝑐
and

(𝑎 ± Δ𝑎) = 𝑐 ± Δ𝑐
(𝑏 ± Δ𝑏)
where

Δ𝑐 = ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Δ𝑎 2 + Δ𝑏 2⎯.
𝑐 √( 𝑎 ) ( 𝑏 )
Other functions:

Taking the derivative of𝑦 = 𝑓(𝑥) , and recognizing that Δ𝑥Δ𝑦 → 𝑑𝑦𝑑𝑥 ,
Δ𝑦 = 𝑓 ′ (𝑥)Δ𝑥.
For example, if y=cosx, then for 𝑥 ± Δ𝑥, Δ𝑦 = | − 𝑠𝑖𝑛𝑥Δ𝑥| .

Note, don't forget to correctly account for significant figures in final calculation
results.

5. Percentage Difference

In many experiments a result can be compared to a known or "accepted" constant (e.g. , , 𝑔𝑘

𝜇𝑜 , etc.).

In these cases, include a percentage difference calculation:

% 𝐷𝑖𝑓𝑓 = |𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 − 𝑎𝑐𝑐𝑒𝑝𝑡𝑒𝑑| × 100%

𝑎𝑐𝑐𝑒𝑝𝑡𝑒𝑑
Reference J.R. Taylor, An Introduction to Error Analysis (University Science Books, Mill Valley,
California, 1982).

1.2.1 Tables
Well-organized data is critical when plotting graphs. Tables should include a descriptive title and
clear column headings.

Example Table 1: Physical pendulum measurments of period of oscillation based on pendulum

length.

1/L (1/m) Δ(1/L) 2 2

ω (rad/s) 2
Δ(ω ) 2
Theory (rad/s)
 1/L (1/m) Δ(1/L) 2 2
ω (rad/s) 2
Δ(ω ) 2
Theory (rad/s)

5.41 0.29 85.6 1.4 79.5

6.49 0.42 102.7 2.2 95.5

2.76 0.38 39.8 0.6 40.6

5.62 1.58 84.5 1.6 82.6

If all values in a table column have identical uncertainty then it's usually best to include the
uncertainty in the column heading. If the uncertainty value varies for each data point, a
separate uncertainty column can be included.
If table values are used to make additional calculations (e.g. calculating density from volume
and mass measurements), any calculated results can be (1) added as additional columns, or (2)
organized into a separate table.
Tables of processed data are often included in the Results and Calculations section of a report.
Trial data (i.e. raw data) is often best included in an appendix.

1.3 Graphical Analysis

Graphs provide a powerful way to visualize experimental results. They're often the most effective
and efficient way to summarize an experiment.

Example Figure 1: Provide a brief description of the graph.

Figure 1 shows an example including important features: title, axis labels, legend, error bars, fitted
trendlines, descriptive figure caption.

1.3.1 Slope and Uncertainty

Various fitting approaches (linear regression, polynomial fitting, etc.) are available for calculating the
equations of lines and curves that optimally conform to sets of measured data. Many plotting tools
provide features for calculating such trendlines automatically.

Ideally, best-fit trendlines should overlap any error bars in plotted data. Few fitting tools acutally
provide an uncertainty value to accompany their best-fit slope calculations. However, as illustrated
in Figure 2, when plotted data contains uncertainty then there can exist alternative fitting lines that
 g , p y g
still fall within the error bars associated with the data, and this range in slope values represents
slope uncertainty.

Figure 2: Example of slope uncertainty in cases where data includes uncertainty (shown via error
bars).

Slope Uncertainty

Recall that the "rise over run" slope of a straight line is,

𝑚 = 𝑥𝑦22 −− 𝑥𝑦11 .
As such, an approximate calculation of uncertainty can be performed recognizing that the slope
equation is a simple subtraction and divsion equation.

Using the uncertainty in (𝑦2 − 𝑦1 ) ± Δ𝑦 and (𝑥2 − 𝑥1 ) ± Δ𝑥 values, Δ𝑚 can be calculated as,
Δ𝑚 = ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Δ𝑦 2 + Δ𝑥 2⎯.
𝑚 √( 𝑦2 − 𝑦1 ) ( 𝑥2 − 𝑥1 )
1.3.2 Graphing Tools
Many tools for preparing graphs are available to choose from. Spreadsheet applications like Google
Sheets or Microsoft Excel are commonly used. Programming languages like Matlab and Python are
also powerful options that offer great flexibility. Certain data acquisition programs like Vernier
Logger Pro also include useful plotting capabilities.

When using any tool, it's important to learn the basic functionality required to present x and y data
and build additional skills as needed.

1.3.3 Spreadsheets
Spreadsheets like Excel and Google Sheets offer many powerful tools for organizing, analyzing, and
visualizing data. They're particularly useful for relatively small data sets.
Example Figure: Google Sheet example of xy scatter plot with best fit trendline.

Best Practices. Screenshots of graphs captured from data analysis software sometimes
suffer from issues with poor resolution, pixelation, and background noise.

As a better alternative for reports, export images of plots from your graphing software as a
"png" or similar image file.

1.3.4 Other Software

Programming applications like Matlab, Python, Mathematica, and other languages are extremely
powerful and flexible for data analysis and visualization. They can handle small data set
computations easily, and can often process larger data sets much more efficiently than
spreadsheets.

Programming applications can be more complex than spreadsheet tools, and therefore may take
longer to learn to use proficiently. However, many resources are available to assist users (e.g.
interactive web tutorials, discussion forums, etc.).

1.3.5 Beyond Basics - Advanced Features

It's also worthwhile to go beyond the basics and learn to use additional features these tools offer.
This might include:

Merging multiple data sets by using both left and right axes.
 Using inset plots to magnify regions showing greater detail.

Using multiple panels to visualize groups of plots.

Displaying 2D or 3D data sets.

1.4 Significant Figures and Scientific Notation

Researchers need to be conscientious in correctly carrying significant figures through calculations
during data analysis. Several standardized conventions apply.

Definition: In any given number, the significant figures include all digits that are considered reliable,
as well as the first digit considered doubtful (i.e. the digit defining our uncertainty). Furthermore, we
ignore zeros acting as placeholders.

Examples – Determining Significant Figures

480491 kg →
Six significant figures
480500 kg →
Four significant figures (last two zeros are placeholders)
400000 kg →
One significant figure (last five zeros considered placeholders)
480.491 g →
Six significant figures
480.500 g →
Six significant figures (including last two zero decimals implies they are
considered significant)
0.0003489 kg →
Four significant figures (all zeros before 3 are placeholders)
0.00034 kg →
Two significant figures
0.034 kg → Two significant figures
1.23 g →Three significant figures
0.12 g →Two significant figures (leading zero in this case is a placeholder, included by
convention)
0.120 g →Three significant figures (including the final zero in this case implies that it is
considered significant)

Calculations, Taking into Consideration Significant Digits

It is particularly important to properly account for significant figures when performing calculations,
so that results are not stated with too much or too little precision.

Rule 1: Significant Figures when Adding or Subtracting

When adding or subtracting numbers, the position of the least significant digit in the result is the
same as the position of the last significant digit in the least precise number being added or
subtracted.

e.g. 2.51 + 4.⎯⎯8 = 7.⎯⎯3 (not 7.31!)

Rule 2: Significant Figures when Multiplying and Dividing

When multiplying or dividing a set of numbers, the final result should include only as many
significant digits as the number in the set with the fewest significant digits.

e.g.
⎯5.65
⎯⎯⎯⎯⎯⎯ × 4.875 = ⎯27.5
⎯⎯⎯⎯⎯⎯ (not 27.54!)
1.4.1 Scientific Notation
Scientific notation makes use of powers of ten to adjust the position of the decimal point. It has two
advantages:

Convenience and compactness.

 The mass of an electron is 9.11×10 −31 kg. Scientific notation improves readability significantly
over writing 0.000000000000000000000000000000911 kg.
Elimination of ambiguity regarding precision.

Suppose we see the statement: “The average distance between Earth and Sun is 150,000,000
km”. Are the zeros significant digits or are some of them used only as “placeholders”? By
8
writing the value as 1.50×10 km, we clearly communicate a precision of three significant
figures.

2 Example Laboratory - The Physical

Pendulum
Here we provide an example of an undergraduate physics laboratory, including a typical report.
Some commentary (shown in text boxes) is included to discuss points a student should consider as
they proceed to perform their own experiments and write their reports.

As a case-study example the motion of a "physical pendulum" is investigated. A physical pendulum

is a uniform rigid body that oscillates about a point at one end.

This example report was prepared using a Python-based Jupyter Notebook

(https://www.jupyter.org), "an open-source web application that allows you to create and
share documents that contain live code, equations, visualizations and narrative text."
[jupyter.org]

An alternative Google Sheet spreadsheet providing similar data analysis and graphs can be
viewed here:
https://docs.google.com/spreadsheets/d/11akP22Zte13FjsJzp7npxsN1FibBvMigZrb7f9xLQ_Q/edi
usp=sharing
(https://docs.google.com/spreadsheets/d/11akP22Zte13FjsJzp7npxsN1FibBvMigZrb7f9xLQ_Q/ed
usp=sharing)

As with any tool, ongoing practice is required to build proficiency and develop more advanced
expertise.

2.1 The Experiment

Consider a physical pendulum, as shown in Figure 1. The pendulum's period of oscillation depends
on its physical length. Design and perform an experiment to identify and confirm the nature of this
dependency.
Figure: A physical pendulum, consisting of a rigid rod pivoting about one end point.

Guided Experiments Some lab manuals will take a "step-by-step" approach and provide the
student instructions to guide them through a set of specific measurements using particular
equipment. This method doesn't require quite the same creative thinking as goes into
designing the entire experiment, but does give good hands-on experience performing and
reporting an experiment.

Designed Experiments The simple introduction above does provide enough of a seed to
proceed independently and devise your own experiment. This approach puts the onus on the
student to really understand and be able to work with the physics they're studying. It expects
the student to learn experimental design, in addition to taking measurements and writing and
analytical report.

Regardless of whether you're guided along or you're asked to devise your own procedure, the
conceptual experimental steps and report writing will be similar.

2.2 The Report

How to proceed?

As a student experimenter given the task above, the first questions to ask should probably be
something like, "Does the oscillation of a physical pendulum really depend on its length? And
if so, how?"

A good place to start would be to think about a force diagram and any theory we might know
should apply.

Having done this preliminary review to understand the necessary physics, (1) write a brief
introduction to state the goal of the experiment, and (2) write a concise theory section that
walks a reader through the important physics. (This is usually a good place to include a
derivation explaining the physics.)
2.2.1 Title Page
Title: Investigation of Oscillatory Behaviour of a Physical Pendulum
Name: J. Doe
Course: PCS 130
Section: 1
Instructor: Dr. Smith
TA Name: D. Green
Date Submitted: January 20, 2021

2.2.2 Introduction
The purpose of this experiment is to investigate the dependence of a rigid-rod "physical
pendulum's" frequency of oscillation on the pendulum's length, where the square of the angular
frequency is predicted to exhibit a proportional dependency on length.

2.2.3 Theory
Consider a free body diagram for the physical pendulum as shown in Figure 2.

Figure 1: A physical pendulum free body diagram.

Taking the forces about the rod's centre of mass, a tensile force is directed from the centre of mass
toward the pivot point. Gravity can also be shown as a force acting at the centre of mass. And if the
rod is displaced from vertical, a a non-equilibrium component of the force of gravity, perpendicular
to the rod, pulls the rod back toward vertical equilibrium.

In rotational frame of reference, F=ma can be expressed as 𝐼𝛼 = Σ𝜏 where 𝐼 is moment of inertia,

𝛼 is angular acceleration, andΣ𝜏 sums torques in the system. Thus,

𝐼𝛼 = −(𝑚𝑔𝑠𝑖𝑛𝜃)(𝑑)
𝐼 𝑑𝑑𝑡22𝜃
= −𝑚𝑔𝑑𝑠𝑖𝑛𝜃
Using the small angle approximation, 𝑠𝑖𝑛𝜃 ≈ 𝜃 . (This approximation is accurate to within ~1% for
∘
displacement up to ~15 .)
Then,

𝐼 𝑑𝑑𝑡22𝜃 = −𝑑𝑚𝑔𝜃
This result has the form of a second order differential equation. Because we've encountered similar
systems previously (e.g. spring-mass, etc.), we know that the equation has solutions of the form
𝜃(𝑡) = 𝜃∘ 𝑐𝑜𝑠(𝜔𝑡 + 𝜙) 𝜃(𝑡)
. Substituting into the differential equation, we find that,

−𝐼 𝜔2 𝜃 = −𝑑𝑚𝑔𝜃
such that the pendulum oscillation frequency is
⎯𝑑𝑚𝑔
⎯⎯⎯⎯⎯⎯⎯
𝜔=√ 𝐼
This is pretty interesting. 𝜔 depends on 𝑑, which is just the distance from the pendulum's
pivot point to its centre of mass. This is likely something we can measure pretty easily
(assuming the CM is half the length of the pedulum). It also depends on mass, 𝑚 , which we
can measure, and 𝑔 which we probably already know. The last parameter is 𝐼 , and we might
recall from our study of rotational motion that for a rod rotating about one of its ends
𝐼 = 13 𝑚𝐿2 = 43 𝑚𝑑2 , which again depends on the measurable parameters 𝑚 and 𝑑. So we
substitute this into our theory and further simplify 𝜔...

𝐼 = 13 𝑚𝐿2 = 43 𝑚𝑑2 , then the pendulum's angular

Since for a rigid rod pivoting about one end,
frequency is:
⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
𝜔 = 4𝑑𝑚𝑔
√ 3 𝑚𝑑2
⎯⎯⎯⎯⎯
3𝑔 ⎯
= √ 2𝐿 .
Therefore we expect that 𝜔 is dependent only on 𝑔 and 𝐿 , which we will attempt to measure and
verify experimentally.

Excellent. This is a relatively simple relationship that we might not have anticipated until
𝜔 𝐿
working through the derivation. Since depends only on variable , we should be able to
𝜔
devise a procedure to test experimentally.

𝜔 𝑔 𝐿
Considering this expression a little further, as interesting as it is that depends on and ,
it's perhaps equally interesting that𝑤 doesn't depend on other parameters we might have
expected to be influential. We might even go so far as to perform simple experiments to verify
𝜔
that certain parameters like amplitude, mass, etc., truly do not affect .

2.2.4 Procedure
Four physical pendulums were constructed using pencils of various lengths. (In one case two
pencils were taped together to create a longer rod.)

To create a pivot point, a sewing needle was pressed through the pencil's eraser. The pin was
balanced between two books so the pencil could oscillate freely. The pendulum was then offset 5-
10 degrees from vertical equilibrium and released.

Using a stopwatch, the time required for 10 oscillations was measured. The measurement process
was repeated 5 to 10 times for each pencil.

The averaged oscillation times were tabulated for each pencil and used to determine angular
frequency. A plot of the square of the angular frequency versus the inverse of the pencil length was
compared to theoretical predictions.

We've briefly provided enough additional detail that a reader should understand the method
used and could even repeat the experiment themselves.

Note, we've also thought somewhat ahead and written the procedure in a descriptive way
appropriate for a formal report we'll eventually submit (or even publish). We frame the
procedure in the past tense. And for formality usually we also avoid personal pronouns like "I"
and "we."

2.2.5 Results and Calculations

Here we should take a moment and think about how to tabulate our measurements. Before we
begin taking measurements we should already know what our tables will look like to let us
collect our raw data.

Note, for our report, after data collection we may want to add columns to our table to
calculate new values (e.g. converting the time for 10 oscillations to a frequency or angular
frequency). We might also want to add additional tables that summarize or reorganize our raw
data and results of any calculations.

If our raw data is too large we may want to move it to an appendix instead of including it in the
body of the report. And if there is too much raw data to practically include in the report, as is
sometimes the case, it may even be necessary to only include in graphed form.

Table 1: Summary of physical pendulum period measurements for four pendulums, including
calculations of frequency, angular frequency,𝐿−1
, and 𝜔2
. [See Appendix A for trial data.]
Out[58]:
ΔL (m) T (s) ΔT f (Hz) Δf 𝜔 (rad/s) Δ𝜔 𝐿−1 (1/m) Δ( −1) 𝐿
2 (rad/s)2
𝜔
2
Δ𝜔

L (m)

0.185 0.010 0.679 0.007 1.470 0.020 9.26 0.13 5.41 0.29 85.7 1.3

0.154 0.010 0.620 0.008 1.612 0.021 10.13 0.13 6.49 0.42 102.6 1.9

0.362 0.005 0.996 0.009 1.004 0.009 6.31 0.06 2.76 0.04 39.8 0.5

0.178 0.005 0.684 0.007 1.462 0.015 9.19 0.09 5.62 0.16 84.4 1.2

Significant figures. During example calculations it may be useful to retain an extra decimal
place as it allows us to round our final result, but inclusion of too many excess digits is messy,
unhelpful, and should be avoided. In summary tables, always try to round results to the
correct number of significant figures.

Uncertainty. Here, since the uncertainty changes for each value in the columns, we've
included uncertainty as a separate column. Alternatively, you might include the uncertainty
±
within each column (e.g. 0.679 0.007s). If the same uncertainty applies to all rows of a
column, the uncertainty could also be included in the colum label (e.g. L ( Δ ±0.010
m)).

Example Calculations (Pendulum 1):

Frequency ( ):𝑓
1
𝑓 = 𝑇𝑎𝑣𝑔
1
= 0.679𝑠
= 1.473𝐻𝑧
Uncertainty ( Δ𝑓 ):
Δ𝑓 = ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Δ𝑇𝑎𝑣𝑔 2⎯
𝑓 √( 𝑇𝑎𝑣𝑔 )
⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
0.007𝑠 2⎯
Δ𝑓 = (1.473𝐻𝑧)√( 0.679𝑠 )
= 0.015𝐻𝑧
±
Therefore f = 1.47 0.02 Hz.

Angular frequency ( ): 𝜔
𝜔 = 2𝜋𝑓
= 2 ∗ 𝑝𝑖(1.473𝐻𝑧)
= 9.26𝑟𝑎𝑑/𝑠
 Uncertainty ( Δ𝜔):
Δ𝜔 = ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Δ𝑓 2⎯
𝜔 √( 𝑓 )
⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
0.015𝐻𝑧 2⎯
Δ𝜔 = (9.26𝑟𝑎𝑑/𝑠)√( 1.473𝐻𝑧 )
= 0.13𝑟𝑎𝑑/𝑠
Therefore 𝜔 = 2𝜋𝑓 = 9.26 ± 0.13𝑟𝑎𝑑/𝑠 .

Example Calculations of 𝐿−1 , 𝜔2 (Pendulum 1):

Inverse Length (𝐿−1 ):

𝐿−1 = 1/𝐿 = 1/0.185𝑚 = 5.41𝑚−1

Uncertainty (Δ(𝐿−1 )):

Δ(𝐿−1) = (𝐿−1)(√⎯(Δ𝐿/𝐿)⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯2⎯) = 5.41𝑚−1√⎯(0.010/0.185)

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯2⎯ = 0.29𝑚−1
Angular frequency squared (𝜔2 ):

𝜔2 = (2𝜋𝑓 )2 = (2𝜋𝑇 −1)2 = (2𝜋(1/0.679𝑠))2 = 85.7𝑟𝑎𝑑2 /𝑠2

Uncertainty (Δ(𝜔2 )):

Δ(𝜔2 ) = 𝜔2 (√⎯(Δ𝑇/𝑇
⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
)2 + (Δ𝑇/𝑇 )2⎯) = 𝜔2 √⎯2(Δ𝑇/𝑇)
⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯2⎯ = 85.7𝑟𝑎𝑑2 /𝑠2 √⎯2(0.007/0.679)
⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯2⎯

Figure 2: 𝜔2 vs. 𝐿−1 for four physical pendulums.

 Hopefully our measurements align with our theoretical derivations. Because real data rarely
aligns perfectly with theory though, we should also include error bars that capture any
"uncertainty" in our measurements and calculations. If we've done our work well, our error
bars should overlap our theoretically predicted line. If there's poor agreement between our
results and theory that might be a sign we need to review our experiment for unnoticed
problems or check our analysis for mistakes.

Best Fit. Most graphing software includes tools to calculate and display a "best fit" trendline.
You'll explore fitting alogrithms in future experiments, but for now understand that the best fit
is indeed the line that minimizes the sum of "residuals," which are the differences between
each data point y 𝑑𝑎𝑡𝑎 𝑙𝑖𝑛𝑒
(x) and the line y (x) for each data point x.

Percentage Difference

From theory we expect a slope of 𝑚 = 3𝑔/2 = 14.7𝑟𝑎𝑑2 /𝑠2 .

Comparing to measured results, the percentage difference is:

𝑚(% 𝑑𝑖𝑓𝑓) = |𝑚𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 − 𝑚𝑎𝑐𝑐𝑒𝑝𝑡𝑒𝑑| × 100%

𝑚𝑎𝑐𝑐𝑒𝑝𝑡𝑒𝑑
= |16.64𝑟𝑎𝑑14.7𝑟𝑎
2 /𝑠2 − 14.7𝑟𝑎𝑑2 /𝑠2 |
𝑑2 /𝑠2 × 100%
= 13.2%
2.2.6 Discussion and Conclusions
Figure 2 illustrates measurement of 𝜔 2
versus 1/𝐿
for four physical pendulums. Results exhibited
proportionality between 𝜔2
and 𝐿 −1
as expected. This was in general agreement with theoretical
2 2
prediction, though the measured slope of 16.64 rad /s was greater than expected by 13.2%
percentage difference calculation.

Notably this difference in proportionality is slightly outside of calculated uncertainties, as indicated

by the error bars in the plot which do not quite overlap the theory trendline. This result is probably
due to under-estimation of the uncertainty in the length of the pendulums that were used. Since
some of the pencils used as pendulums were sharpened or included a metal and rubber eraser
attachment, uncertainty in the length-dependent moment of inertia (assumed to be 𝐼 = 13 𝑚𝐿2 )
may be greater than estimated. To achieve better result agreement it would therefore be good to
repeat the experiment with greater attention to the uniformity of the rods used.

Here we have summarized our main results and comment on any significant observations and
unexpected discrepancies.

2.2.7 Appendix A - Oscillation Length and Period Measurement Trial

 Data
Table A1: Measured times for 10 pendulum oscillations. Comma separated value format.

Out[98]: Pen 1 (L=18.5±1.0cm) Pen 2 (L=15.4±1.0cm) Pen 3 (L=36.2±1.0cm) Pen 4 (L=17.8±1.0cm)

[±0.5s] [±0.5s] [±0.5s] [±0.5s]

Trial

1 6.88 6.35 9.93 6.77

2 6.83 6.15 10.05 6.90

3 6.82 6.34 10.06 NaN

4 6.67 6.16 9.80 NaN

5 6.65 6.11 9.96 NaN

6 6.83 6.23 NaN NaN

7 6.87 6.16 NaN NaN

8 6.76 6.12 NaN NaN

9 6.81 NaN NaN NaN

10 6.76 NaN NaN NaN

11 6.86 NaN NaN NaN

Example Calculations (Pendulum 1)

Mean Period:

𝑇𝑎𝑣𝑔 = Σ𝑇𝑁 𝑖
= (6.88𝑠 + 6.83𝑠 + 6.82𝑠+...+6.76𝑠
11
+ 6.81𝑠 + 6.76𝑠 + 6.86𝑠)/10
= 0.679𝑠
Standard Deviation of Period:
⎯Σ(⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
𝑇 𝑖 − 𝑇 𝑎𝑣𝑔 ) 2⎯
𝜎=√ 𝑁
⎯(0.688𝑠
⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
− 0.679𝑠 ) 2 + (0.683𝑠 − 0.679)2 +...+(0.686𝑠 − 0.679𝑠)2⎯
=√ 11
= 0.007𝑠
Therefore for pendulum 1, 𝑇 = 0.679 ± 0.007𝑠.

3 Appendix - Code Examples for Basic

Graphing
3.1 Python
 # Example: Plotting basic XY data with error bars
import matplotlib.pyplot as plt # Import plotting library
import numpy as np

# Data
x_data = [1.00, 2.00, 3.00, 4.00]
x_data_err = [0.10, 0.10, 0.40, 0.10]
y_data = [2.10, 3.20, 3.95, 5.10]
y_data_err = [0.15, 0.15, 0.20, 0.25]

# Calculation of best-fit slope and intercept using Numpy library

m,b = np.polyfit(x_data,y_data,1)
best_fit = [m*x+b for x in x_data]

# Plot configuration
plt.errorbar(x_data,y_data,y_data_err,x_data_err, "o", label="Exper
iment") # Define data to be plotted
plt.plot(x_data,best_fit, label=f"Best fit: y={m:.2f}x+{b:.2f}") #
Add plot of best-fit trendline
plt.title("Title goes here") # Title
plt.xlabel("x-axis label (units)") # Axis labels
plt.ylabel("y-axis label (units)")
plt.legend(loc="upper left")
plt.show()

Google Colaboratory (https://colab.research.google.com/ (https://colab.research.google.com/))

offers an online notebook environment where Python code can be run conveniently.

3.2 Matlab/Octave
% Example: Plotting basic XY data

% Data
x = [1.00 2.00 3.00 4.00];
x_data_err = [0.10 0.10 0.40 0.10]
y = [2.10 3.20 3.95 5.10];
y_data_err = [0.15 0.15 0.20 0.25]

% Plot configuration
title ('Title goes here');
xlabel ('X-axis label');
ylabel ('Y-axis label');
errorbar(x,y,y_data_err);

Matlab is available from https://www.mathworks.com/ (https://www.mathworks.com/). Matlab is

commercial software.

A free, open-source Matlab alternative called Octave is available at

https://www.gnu.org/software/octave (https://www.gnu.org/software/octave).