LAB REPORT WRITING GUIDE, with HINTS AND SHORTCUTS
There is no set format for the lab reports.
The goal is for you to develop your own style and
learn how to communicate experimental results accurately and concisely (using as few pages as
possible to get
across all the relevant information). Your reports should look like professionally
published engineering documents e.g., your textbooks, instruction manuals, SAE Papers.
In ME 308 you do not simply follow a prescribed procedure in lab and in writing the
lab reports.
You must think about what data you need to take to get the answers, and then develop your own
procedures to get that data. I will not tell you what figures and tables to put in your lab reports

you must determine for yourself what data is im
portant and how to present it.
Experimental
results should be critically analyzed. When possible, they should be compared to results from
theory or simulations. When possible, the experimental results should be graphed to show trends.
Objectives For ME 3
08 Fluids Lab Experience:
To validate the theoretical analysis techniques learned in lecture and to illustrate the
physical concepts of fluid mechanics discussed in the lecture and the book, and to
demonstrate the limitations of basic fluids theory.
To dev
elop effective communication of technical information, including proficiency in
Microsoft WORD and EXCEL, and to develop computer skills for acquiring data, data
reduction, error analysis, and plotting.
For the students to become familiar with and able to
use experimental techniques for
analysis of fluid flows
To develop an ability for team

work. (will be critical for your senior design experience)
To learn how to write concisely
–
there is a 6 page limit on lab reports.
To foster the confidence and self

re
liance required for open

ended experiments (i.e.,
solving real

world problems) and reduce dependence on a “cookbook” approach. To
encourage creativity in the use of experimental apparatus and data

acquisition.
Before Coming to Lab:
Read through the Lab
Instructions
Print out an EXCEL spreadsheet for taking data. The sheet should be blank except for
row and column headings.
Decide if you need to bring anything else to lab (camera, video camera, calculator,
laptop)
Start working through calculations to det
ermine what data to take.
If the relevant material has not yet been covered in class lecture, you should take the
initiative to read it in the textbook.
The experiments will make much more sense to you if you are prepared, they will take less time
to co
mplete, and you will get better data (which seems to be particularly a problem in the 8 am
lab section).
You may use any of the equipment in the Fluids Lab to get the data you need. This equipment
includes:
Stopwatches
Tachometer
Laser Level
Flowbench mas
s flow meters
Pitot Tubes (for velocity)
Manometers (for pressure)
Rulers
Tape measure
Mercury barometer
Digital barometer
Thermometer
Graduated cylinders
Electronic scales
Radar gun
Flow visualization tools (powder, wax chips, koolaid, cork floats)
Miscel
laneous supplies (wood blocks, plastic bins, string, washers)
You can also borrow equipment from the CEGT shop (micrometers, calipers, clamps, etc.)
Late lab reports penalized
10% per day late. Lab reports turned in after 1:01 pm are considered a
day lat
e. Lab reports are limited to a maximum of 6 pages. There is a penalty of 5 pts for each
page over 6.
Proper Formatting for Technical Reports
:
A lab report is considered a technical report. As such, it must be professional and neat. The entire
report mus
t be computer generated (so that it could be emailed if necessary). Any drawings made
by hand must be made with a straight

edge and clearly labeled, and then scanned in to the report
(you can use the scanner in the senior design room).
Margins must be just
ified, page numbers
must be included on each page. Your entire report must be computer generated (so that the whole
file could be emailed, if necessary). That means everything must be typed. A scanned

in version
of a hand

drawn figure is acceptable if the
figure is drawn neatly, but figures created with
computer programs (such as AutoCAD) are preferred.
Equations must be numbered and in proper format. An example follows:
The effect of these forces on the flight of the plane can be understood though Newton’
s second
law:
Σ
F
= m
a
(1)
where
F
represents each individual force,
m
the mass of the plane, and
a
its vector acceleration.
Think about the best and most concise way to present your data

this is usually with a figure or
data table. In most cases, you will wan
t to compare your experimental results with theoretical or
empirical predictions. Tables and Figures are usually the best way to do this. The percentage
experimental error should be stated and
error bars
included on your graphs. Error bars represent
the
ex
perimental uncertainty
(discussed below).
Format of Pictures and Tables
Refer to your
ME 101
notes for proper formatting of graphs and figures.
Plots are an integral part
of technical reports and must be professional.
Each axis must be labeled with appro
priate units.
Scales must be readable to the same accuracy
(# of significant digits) as the data obtained during the test. Note that data sets derived from data
must show experimental data points, while curves derived from an equation should contain no
sym
bols, and only a smoothed line. You need to know when it is appropriate to show the points
and when to show the line connecting the points in your plots. When plotting experimental data,
show the data points. You may wish to add a
trendline
, but always sho
w the original data points!
When plotting a functional (analytical) relationship, show only the line connecting the points, not
the points themselves. If the functional relationship is smooth, make sure you generate enough
points so that curve is smooth as
well.
The x and y
axes need to be labeled with title and units. If
more than one line is drawn on a graph, each line is identified using a legend. If only one line is
drawn, no legend is needed. Nothing on a plot is drawn freehand.
The background of the
figure should be white, not dark gray as is the default in EXCEL. The
number of digits used in the scale should reflect the accuracy of the data (not more than 3 digits).
Do not leave large areas of blank space at the bottom of a page after a figure or tab
le. Move text
from a subsequent page to fill in the blank area. Do not cut off tables at the end of a page. Your
reports should look like a professionally produced engineering document.
Note that many of the defaults in EXCEL and MATLAB are not optimal fo
r easy reading. Do
use the default gray background for plots in EXCEL

change it to white/clear. Use symbols that
are easy to differentiate from each other. All text labels should be at least 10

point font. Axes
should be labeled with units, and a legend
with descriptive titles used when more than one line is
plotted. Color
should be used only when necessary to differentiate a large number of points or
lines. Use colors that are easy to see (avoid light yellow in EXCEL)
The final results from the experime
nts are listed in Table 1 below.
Trial #
Time
(units)
Speed
(units)
Distance
(units)
Force
(units)
Power
(units)
1
2
3
Table 1: Final results from several airplane experiments.
Below is an example of a properly formatted Figure. The ti
tle goes beneath the figure and should
be descriptive. The legend can be placed directly on the figure to save space (you have to do this
manually in EXCEL
–
it is not the default). Each axis must be labeled with units specified.
Figure 1:
Force due
to stream of fluid hitting an object.
REFERENCES
Use standard MLA or APA citations for books and technical articles. Your textbook is a book
and should be cited as such and not simply as “the textbook”. For citations of webpages include
the title of the
page and the author, as in the following example:
Brain, Marshall and Adkins, Brian “How Airplanes Work” from
How Stuff Works
,
http://travel.howstuffworks.com/airplane.htm
. 11/13/2000.
APPENDICES
The only use for an appendix should be to list the origi
nal raw data if it is not contained
elsewhere in the report, or to show detailed calculations of data when only the final results are
presented in the report.
In Summary, a neat, well organized report is expected. Spelling and grammatical errors are not
acceptable. I will take off at least 1 point for each and every mistake in your lab report, including
typos, grammatical mistakes, incorrect presentation of data, and using too many digits in your
numbers.
Your reports should look like a professionally pu
blished report. The goal is for you to
learn how to present your work in an effective manner to your peers and superiors, so that you
will be successful on your job.
Word Processing Shortcuts
Microsoft Word Keyboard Shortcuts
Subscript
–
Control =
Sup
erscript
–
Control Shift =
Underline
–
Control U
Bold
–
Control B
Italics
–
Control I
Copy
–
Control C
Paste
–
Control V
Repeat last command
–
Control Y (or F4)
(On a Mac use the Apple key instead of the Control key)
Greek Letters
(Change the font from
regular to symbol to create the corresponding Greek letters)
a b c d e f g h i j k l m n o p q r s t u v w x y z
α β χ δ ε φ γ η ι
ϕ
κ λ μ ν ο π θ ρ σ τ υ
ϖ
ω ξ ψ ζ
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Α Β Χ Δ Ε Φ Γ Η Ι
ϑ
Κ Λ Μ Ν Ο Π Θ Ρ Σ
Τ Υ ς Ω Ξ Ψ Ζ
Miscellaneous WORD hints:
To Add an Equation
:
Insert
Object
Microsoft Equation
Special Symbols
:
Insert
Symbol
Select symbol such as:
≤ ≥ ±
Formatting Pictures:
Right click on the picture and select
Format Picture
fr
om the menu. Go to
Layout
and click on
Advanced
. Use
Lock Anchor
and
Move Object with Text
to get the picture to
move as you type more text. For a Table right click and select
Table Properties
.
Experimental Error and Uncertainty Analysis
:
The proc
edure is to first estimate the uncertainty in each measured quantity (time, distance,
mass, temperature, pressure, etc., and then to analyze the propagation of uncertainty into the
results calculated from the experimentally obtained data.
Uncertainty in
measurements can be estimated by looking at the
precision
of the measuring
device being used. Alternatively, error can be estimated by having multiple people taking a
reading at the same flow condition and looking at the range in values different people me
asure. If
the exact value of a variable is known, the difference between the measured value and the exact
value is the
experimental error
. This can be thought of as
bias error
in the system, and if it is
known, it can be subtracted off. The previous error,
due to differences in measuring from
different users, can be thought of as a
random error
. Errors in measurements can be categorized
as either
systematic
or
random
. The randomness of such errors leads to experimental uncertainty.
A good model for the pr
opagation of errors into a calculated quantity from measured ones is that
the error is equal to the square root of the sum of the effects of the square of each of the
individual errors, using the chain rule:
Where each x
i
represen
ts an originally measured quantity, and
x
i
is the experimental
uncertainty in the measured value of x
i
.
Example: Uncertainty in density calculation using Ideal gas law, for temperature measured with
a thermometer and pressure with a mercury barometer. H
ere the accuracy of the density
measurement will be affected by the accuracy of the pressure and temperature measurements.
We can assume that M and R are “exact”, so that we need only be concerned with uncertainties
in the measur
ed values of P and T. The value of M might not be exact if there are impurities
present in the gas mixture, but we will ignore that for now. If the temperature is measured with a
thermometer to an accuracy of +/

0.5 C, and the pressure is read from the el
ectronic barometer
with an accuracy of +/

10 Pa, then the uncertainty in the calculated value of the density can be
estimated. The formula fro the propagation of errors in this case is:
The ideal gas law must be used to evaluate
the partial derivatives:
Substituting these relationships into the propagation of errors equation yields:
Upon dividing both sides by the density:
S
o for the ideal gas law, which is a linear relationship, the
relative error
in the density
calculation is equal to the square root of the sum of the relative errors of the pressure and
temperature measurements. Plugging in the values above on a day when th
e measured pressure is
100 kPa and the temperature is 22
C = 295 K.
So for the calculated density of 1.18 kg/m
3
, the uncertainty is +/

0.0017*1.18
kg/m
3
= +/

0.002
kg/m
3
. So for these relatively accurate measurements, it is ac
ceptable to carry three significant
digits after the decimal point, but no more than three digits.
Example
:
From Impulse

momentum experiment: The directly measured force is linearly proportional to the
displacement of the balancing weight, so that F
dir
=
C x. The uncertainty in measuring x is +/

1
mm, so the uncertainty in the force is:
So the relative uncertainty in the Force is:
For the indirect force on the flat plate target, the primary uncertainty
comes from the
measurement of the water flow rate.
While in theory the mass flow rate error is due only to uncertainty in the stopwatch timing, in
practice the error is much larger, probably around 5%.
If the uncertainty in the mass flow rate is 5%, then the uncertainty in the indirect force is about
7%.
With the uncertainty now known, these values should be used to create error bars on your graphs.
This is easily done in EX
CEL. Double click on a data series and select error bars and select the
relevant options.
Experimental error
Experimental error is defined as the difference between the measured value and the predicted or
theoretical value.
If t
he experimental error is less than the experimental uncertainty, than the difference between
the predicted and measured values can be explained due to the limited precision of the
instruments.
Data Outliers
Occasionally user or equipment error will result
in a data point that does not seem to agree with
theory or the rest of the data gathered.
There may be times after lab when you are analyzing the data, and you find the numbers come
out looking ridiculous. If every number gives you a ridiculous result,
then you know you did the
experiment incorrectly or there was an equipment failure. At a research lab, you would have to
go back and repeat the experiment. The Jobst fluids lab is used for other courses and research, so
your only option here would be to ho
pe you can get the data from another lab group.
It may be instead that most of your data looks good, but there are one or two points that do not
follow the overall trend. You may feel tempted to eliminate these data points from your report.
You may then f
eel your conscience nagging you that arbitrarily removing points is bad science.
Fortunately there is a scientific way of evaluating data statistically to see if a particular data point
is an outlier from the overall sample. In order to throw out data poin
ts you must
be able to (a)
think of a explanation of how the error occurred (such as user error), (b) define an objective
criteria by which bad data is identified and eliminated.
Chauvenet’s criteria is a statistically valid method for removing data outli
ers from a data set. In
the example below it is believed that the data (y) should follow a linear fit with the independent
variable (x).
Procedure
:
1.
Fit a curve to the (x,y) data set using EXCEL’s
Add Trendline
option. Make sure to print
the formula of th
e trendline to the screen. (This example uses a linear fit, but the method
should work for polynomial or exponential curves, too).
2.
Calculate the deviation, d, of each data point from the trendline. For this example with a
linear fit of the form y = m x +
b, the deviation is:
i
= y
i
–
(m x
i
+ b)
3.
For i = 1, 2, 3, …, N, where N is the number of x,y data points.
4.
Compute the standard deviation of the data set, using the formula:
5.
Compute the
Deviation Ratio
, DR, for each data point.
6.
Then Compare the Deviation Ratio, DR, to the maximum statistically allowable deviation
ratio, DR
0
, based on the normal distribution and the number of data points used. If DR >
DR
0
the data point can be rejected and omitted from the
data set.
7.
After dropping the bad data points, re

compute a new curve fit (EXCEL will do this
automatically if you just erase the bad data point from the cells the chart is using).
An example using a linear fit is contained in an EXCEL file
Chauvenet.xl
s
on the class website.
The allowable Deviation Ratios as a function of number of data points is shown in the table
below.
N (number of data points)
DR
max
(max allowed)
5
1.65
10
1.96
15
2.12
20
2.24
25
2.33
50
2.57
100
2.81
Σχόλια 0
Συνδεθείτε για να κοινοποιήσετε σχόλιο