Vous êtes sur la page 1sur 45

Buiteweg-Liu 1

Introduction
The military builds huge aircrafts that are extremely hard to miss, but for
some reason, radars are somehow unable to detect them. This is the result of
research for stealth technology. The design of the aircraft can make it practically
invisible to radar waves; it is possible to design the aircraft in a way that it reflects
most of the radar waves directed towards it. Figure 1 shows a visual example of
how radar waves can be reflected off an aircrafts surface because of how it is
designed. Obviously, there are many strategic military applications with this.

discovery.




Figure 1. Radar Waves Reflected Off Aircraft
Figure 1 shows how radar waves can reflect off a carefully designed
aircraft. As an electromagnetic wave, light will demonstrate the same properties
as radar waves and so in place of radar waves in the research light waves will be
used. Four three-dimensional shapes were chosen the researchs objective was
to discover which of the shapes would return the lowest amount of light back to
the source and why. A curved cone, curved cylinder, flat rectangular face, and
the edge of a rectangular prism were chosen. A bright light was shone on the
shapes and a light sensor detected how much light returned to the source. As
Buiteweg-Liu 2

light is similar to the radar waves, the light sensor is similar to a radars antenna
(Brain). The angle at which incident waves of the electromagnetic wave spectrum
approaches is the same angle at which the waves are reflected, as stated in the
Law of Reflection (Giancoli). Smooth and flat surfaces are best at reflecting light
waves. It was hypothesized that the curved cone would have the lowest light
return, or illuminance, measured due to the way the shape is angled and its
curves.
If the more curved shapes produce a lower light return than those with flat
vertical surfaces, this indicates that aircrafts will be made more undetectable to
radar waves if designed similar to the curved shapes than similar to the other
shapes used in the research experiment. Not only can this information be applied
in the design of aircraft, it can be used in the design of other military vehicles like
submarines. On a more practical level, the research can be used to predict the
path of light as it hits a surface. Being undetectable by enemy radar waves can
help in saving military personal and help in the victory of the nation in a war.
In April of 1993 the United States Navy released its new Sea Shadow
stealth surface ship. The ship was a 50 million dollar ship, part of a 195 million
dollar research program, looks like a cross between a catamaran boat and
Lockhead F-117A stealth strike fighter (Ashley). The ship has a layer of
radar-absorbent coating that is resistant to salt water corrosion. The ship is
above water with very little of the actual ship submerged into the water and is
shown in Figure 2. The ships sides angle back from the vertical in order to avoid
reflecting the waves back to the search radars that are usually horizontal. The
Buiteweg-Liu 3

number of parallel lines and right angles are also kept to a minimum. With the
radar-absorbent coating, angles of the ship, and very little of the ship submerged
in water the radar waves respond slower than normal making it possible for the
ship to go almost undetectable.

Figure 2. Sea Shadow Stealth Surface Ship
Figure 2 shows the Sea Shadow Stealth Surface Ship that is designed to
have a low radar signature.











Buiteweg-Liu 4

Review of Literature
Energy waves can come in the form of mechanical or electromagnetic
waves. In the 1800s, James Clerk Maxwell discovered that electric and magnetic
fields together would create electromagnetic waves (Anatomy of an
Electromagnetic Wave). Radio waves were found to be a form of light wave
when Heinrich Hertz conducted research following Maxwells conclusions. He
also discovered how to make electromagnetic waves and his research would
later help the development of radars and the understanding of visible light.
When an object is charged, the object creates a surrounding electric field
while a magnet will produce a surrounding magnetic field. The electromagnetic
waves are formed from the vibrations of both the electric and magnetic fields that
are perpendicular to each other and the direction of wave travel. Any charged
object with a nonzero acceleration produces an electromagnetic wave
(Electromagnetic Waves). Electromagnetic waves, unlike mechanical waves, do
not need to pass through a medium and can pass through a vacuum of space. A
wave can be described by its frequency or wavelength. Electromagnetic waves
can be categorized as radio waves, microwaves, infrared, visible light, ultraviolet,
x-rays, or gamma rays. For a wave to be in the category of visible light, the
frequency of the wave has to be in the range of 3.9x10
14
to 7.5x10
14
Hertz.
Electromagnetic waves, including light waves, have a certain property that
is stated in the Law of Reflection, as shown in Figure 3. It was found that the
angle at which electromagnetic waves hit a surface is the same angle at which
the waves are reflected off the surface (Fellers and Davidson). The
Buiteweg-Liu 5

characteristics of the reflecting surface have a great effect on how much light is
reflected and the way it is reflected and smooth surfaces were found to be the
best at reflecting light waves.





Figure 3. Law of Reflection
Figure 3 demonstrates the Law of Reflection, where the incident angle at
which the electromagnetic wave hit the reflecting surface is the same as the
reflected angle of the reflected wave.
Both light waves and radar signals are in the same electromagnetic
spectrum. Because of this, the effects recorded for the reflection of light waves
would be the same for radar signals. Knowing that angles and the surface
characteristics have a direct relationship with light reflection, the military has
found good use of this knowledge. The military has achieved low radar return for
a stealth ship by angling the ships side from the vertical and keeping parallel
lines and right angles to a minimum.
In an experiment done by the University of Texas of the Austin McDonald
Observatory, they tested reflection of light off of different mirrored angles. This
experiment was performed by first predicting and tracing what the experimenters
thought the path of light was to travel as it was reflected off of a flat, smooth
Buiteweg-Liu 6

mirror. Then in a dark room a light was directed towards the mirror at different
angles and its light path was recorded. The experimenters also chose a curved
inward and outward mirror for testing. At the end of the experiment it was
concluded that the angle of incidence equaled the angle of reflection with the flat
mirror. This experiment demonstrated the Law of Reflection. When the light was
directed into the curved inward mirror, at every angle the light would reflect back
to the center. When the light was directed into the curved outward mirror the path
of light would go directly back to where the light originated from (Reflection).
This experiment done by University of Texas helped in the understanding of how
light is reflected off different shaped surfaces and properties of light.
Electromagnetic waves include both visible light and radar signals whose
reflections are directly related to the reflection surfaces characteristics and the
angle at which the waves hit the surface. The Law of Reflection states that the
angle at which the light is received by the surface is reflected at the same angle.
The angle that will yield the largest light return will be a right angle while curved
surfaces will have lower light return measured. From the conclusion that was
found in the experiment above, by the University of Texas, it can be understand
exactly where the light would be reflected and how the light sensor reads the light
from its different paths. These concepts were applied to hypothesize which
shape will produce the lowest light return of the ones they have chosen; they
have chosen the cone.


Buiteweg-Liu 7

Problem Statement
Problem:
Which shape, the cone, rectangular prism face, rectangular prism edge, or
cylinder, will reflect the least amount of light back to the light sensor?
Hypothesis:
The cone shape will produce the least amount of illuminance.
Data Measured:
The independent variable for the experiment was the type of shape. There
is a cone, rectangular prism, and cylinder. The rectangular prism is used in two
different ways; the prism will be turned to show a face to be considered one of
the shapes and will be turned to only show an edge for a third shape tested in the
research. In total there were four different shapes being used. Each shape has
30 trials, for a total of 120 trials. The dependent variable is the amount of light
reflected back to the light sensor, or illuminance, measured in lux. The amount of
light measured is used to describe how waves scatter off curved surfaces. To
test these variables the researchers have chosen to do a series of three two
sample t-tests, comparing the cone to the other three shapes. This will help the
researchers reject or accept their hypothesis that the cone will produce the
lowest illuminance reading.








Buiteweg-Liu 8

Experimental Design

Materials:

4.5 Volt LED Flashlight TI-Nspire CX Calculator
Vernier Software Light Sensor Black Electrical Tape*change in appendix
Paper Cone LabQuest
Paper Rectangular Prim Cardboard Box 40.5 cm x 30.5 cm x 31.5 cm
Paper Cylinder

Procedure:

1. Set up the cardboard box (Refer to Appendix A).

2. Create the paper shapes (Refer to Appendix B, C, and D).

3. Set up the LabQuest and light sensor (Refer to Appendix E).

4. Use the calculators random integer function to randomize the order of the
120 trials of the four shapes, thirty trials for each shape. Record the shape
to be used for each trial (Refer to Appendix F).

5. Take off the cardboard lid and place the designated trials shape inside so
that it is lined up with the centered long line on the inside bottom of the box.
The shape is also placed right before the flap on the bottom of the box, in
front of the flashlight and light sensor meter. Replace the lid.

6. Press the green arrow in the bottom left hand corner of the LabQuest to
begin collection of data.

7. Turn on the flashlight that is connected to the box.

8. Record the average light return that the light sensor detects by taking the
average of a consistently horizontal interval on the light sensor graph.

9. Remove the shape from the box.

10. Repeat steps 5-10 until all 120 trials are completed.

11. Zero out the every five trials by tapping on the home screen and tapping the
Zero option that pops up.

12. Every ten trials, the light sensor has to be given a three minute break to
produce consistent results again. Unplug the light sensor from the LabQuest
when a break is taken and keep light away from the light sensor during this
time.
Buiteweg-Liu 9

Diagram:


Figure 4. Set-Up of Experiment
Figure 4 shows the set-up of the experiment. The flashlight and light
sensor need to be secured down with tape so that they are not movable,
providing more consistent results. In this experiment, black tape was used to
avoid reflection of light due to its color. The picture on the left shows the set-up
inside the box while the right-hand side shows the set-up from the outside of the
box.









Lid
LED
Flashlight
Black
Box
LabQuest
Target
Shape
Light
Sensor
Black
Tape
Buiteweg-Liu 10


Data and Observations







Figure 5. Experiment in Progress
Figure 5 shows the experiment in process and how data was collected.
Take into account that the light in the room was turned off and that the light was
only turned on in the pictures above to provide clear pictures. Also, in the left-
hand side picture, the cardboard lid was not placed on, unlike during data
collection.







Figure 6. Four Different Experimental Shapes
Figure 6 shows the four different shapes tested in the experiment. The first
on the left is the cylinder shape, then the cone, the rectangular face, and the
rectangular edge.
Buiteweg-Liu 11


Figure 7. LabQuest Graph
Figure 7 shows a graph of the prism face trial. Due to inconsistency of the
light sensor, the LabQuest readings had to be repeated for many of the trials to
obtain an almost constant reading for five or more seconds. The area was then
selected and the average was found.
Table 1
Cone Illuminance

Data
Cone Cone
Trial
Illuminance
(lux)
Trial
Illuminance
(lux)
1 15.6 51 9.8
2 13.8 62 7.5
4 14.5 66 11.6
6 11.7 87 8.1
7 12.9 89 8.9
13 15.5 96 7.6
15 9.5 101 14.2
18 7.5 103 11.4
25 6.4 109 9.8
27 7.5 113 12.5
34 7.9 114 9.2
37 6.6 115 7.5
41 8.2 117 9.6
45 8.0 118 10.3
49 3.9 120 11.3
Table 1 shows the cones data for its 30 trials. The data ranges from 15.6
for trial one to 3.9 lux in trial 49. None of the values are outliers.
Buiteweg-Liu 12

Table 2
Observations for Cone Sample Data Collection
Trial Date Observations
1 5-6-14
Graph was sporadic, but later evened out. The average of the graph's
consistent interval was used as the value for the illuminance.
2 5-6-14 Trial had to be measured several times to obtain a consistent interval.
4 5-6-14 Average of consistent interval was used as the value.
6 5-6-14 Average of consistent interval was used as the value.
7 5-6-14 Average of consistent interval was used as the value.
13 5-6-14 Average of consistent interval was used as the value.
15 5-6-14 Trial had to be measured several times to obtain a consistent interval.
18 5-6-14 Average of consistent interval was used as the value.
25 5-6-14 Average of consistent interval was used as the value.
27 5-6-14 Average of consistent interval was used as the value.
34 5-6-14 Average of consistent interval was used as the value.
37 5-6-14
Researchers notice that the light flickered and a battery was replaced to fix the
problem.
41 5-6-14 Average of consistent interval was used as the value.
45 5-6-14 Average of consistent interval was used as the value.
49 5-6-14 Light sensor was zeroed out for this trial.
51 5-6-14 Average of consistent interval was used as the value.
62 5-6-14 Light sensor was zeroed out for this trial.
66 5-6-14 Average of consistent interval was used as the value.
87 5-6-14 Average of consistent interval was used as the value.
89 5-6-14 Average of consistent interval was used as the value.
96 5-6-14 Average of consistent interval was used as the value.
101 5-6-14 Average of consistent interval was used as the value.
103 5-6-14 Average of consistent interval was used as the value.
109 5-6-14 Average of consistent interval was used as the value.
113 5-6-14 Average of consistent interval was used as the value.
114 5-6-14 Trial had to be measured several times to obtain a consistent interval.
115 5-6-14 Average of consistent interval was used as the value.
117 5-6-14 Light sensor was zeroed out for this trial.
118 5-6-14 Average of consistent interval was used as the value.
120 5-6-14 Zeroed out due to repetition of trials.
Buiteweg-Liu 13

Table 2 shows the observations recorded during data collection for the
cones light return data. The first trials graph was sporadic and so the average of
a consistent interval that showed up on the graph was the value recorded. This
method of finding a value was applied to all sporadic graphs. The 2
nd
, 15
th
, and
114
th
trial had to redone several times because no consistent interval appeared.
The light sensor was zeroed out for every five trials or when it was necessary to
do so and was zeroed out for trials 49, 62, 117, and 120. It was noticed that the
flashlight was flickering during trial 37 and a battery was replaced. After the
battery was replaced, there were no more flickers and the data after was
consistent with previous collection. The rest of the trials went well.
Table 3
Prism Face Data
Prism Face Prism Face
Trial
Illuminance
(lux)
Trial
Illuminance
(lux)
12 52.7 76 48.8
17 48.0 77 49.4
26 47.0 80 51.2
28 51.8 81 49.7
31 47.3 84 50.2
35 47.3 85 48.9
44 55.5 88 51.5
55 49.8 93 51.7
56 51.1 97 51.6
57 47.9 99 51.9
58 48.4 105 50.7
60 50.5 106 53.2
67 49.3 111 47.5
72 51.2 112 49.6
75 51.4 119 49.7

Table 3 shows the data collected for the prisms face. The data ranges
from 55.5 to 47 lux. That is not a big range and there are no outliers.
Buiteweg-Liu 14

Table 4
Observations for Prism Face Data Collection
Prism Face
Trial Date Observations
12 5-6-14 Trial had to be measured numerous times.
17 5-6-14 Average of the consistent interval was used for value.
26 5-7-14 Graph used was very "jumpy" but a consistent interval did appear.
28 5-7-14
Trial had to be measured numerous times, but the graph used was consistent
throughout the 15 seconds measured.
31 5-7-14 Light sensor was zeroed out for this trial.
35 5-7-14 Average of the consistent interval was used for value.
44 5-9-14 Light sensor was zeroed out for this trial.
55 5-9-14 Average of the consistent interval was used for value.
56 5-9-14
Light sensor was zeroed out for this trial and the researchers took a three
minute break to let the Light sensor rest.
57 5-9-14 Average of the consistent interval was used for value.
58 5-9-14 Average of the consistent interval was used for value.
60 5-9-14 Average of the consistent interval was used for value.
67 5-9-14
Light sensor was zeroed out for this trial and the researchers took a three
minute break to let the Light sensor rest.
72 5-9-14 Average of the consistent interval was used for value.
75 5-9-14 Average of the consistent interval was used for value.
76 5-9-14 Average of the consistent interval was used for value.
77 5-9-14 Average of the consistent interval was used for value.
80 5-9-14 Average of the consistent interval was used for value.
81 5-9-14 Average of the consistent interval was used for value.
84 5-9-14 Average of the consistent interval was used for value.
85 5-9-14 Average of the consistent interval was used for value.
88 5-9-14 Average of the consistent interval was used for value.
93 5-9-14
Light sensor was zeroed out for this trial and the researchers took a three
minute break to let the Light sensor rest.
97 5-9-14 Trial had to be repeated several times.
99 5-9-14 Light sensor was zeroed out for this trial.
105 5-9-14
Light sensor was zeroed out for this trial and the researchers took a three
minute break to let the Light sensor rest.
106 5-9-14 Average of the consistent interval was used for value.
111 5-9-14 Light sensor was zeroed out for this trial.
112 5-9-14 Average of the consistent interval was used for value.
119 5-9-14 Average of the consistent interval was used for value.
Buiteweg-Liu 15

Table 4 shows the observations made during the prism face data
collection. The 12
th
, 28
th
, and 97
th
trials had to be measured numerous times
because of a lack of a consistent interval. The 26
th
trial had a very jumpy trial
where the graph oscillated rapidly from a high value to low value, but it later
evened out. The light sensor was zeroed out for trials 31, 44, 56, 67, 93, 99, 105,
and 111. The remaining trials did not have any problems.
Table 5
Prism Edge Data
Prism Edge Prism Edge
Trial
Illuminance
(lux)
Trial
Illuminance
(lux)
3 23.6 48 19.7
9 24.1 50 23.9
11 27.3 52 21.6
14 24.6 59 23.4
16 21.7 61 19.8
19 23.3 63 23.8
21 24.9 64 22.4
23 23.2 65 22.5
29 23.7 69 23.5
32 18.8 71 24.6
38 22.9 78 23.6
40 22.9 95 33.9
43 17.7 98 25.9
46 19.1 108 22.9
47 20.1 116 23.7

Table 5 shows the data collected for the prism edges values. The data
ranged from 17.7 to 33.9 lux, a 16.2 value gap. There were three outliers for the
data and the outliers were the values 17.7, 27.3, and 33.9 lux.



Buiteweg-Liu 16

Table 6
Observations for Prism Edge
Prism Edge
Trial Date Observations
3 5-6-14 The graph showed slight variation from 14.2 to 14.9.
9 5-6-14 Trial showed a perfectly consistent graph.
11 5-6-14
Flashlight and light sensor had to be readjusted because of accidental
movement.
14 5-6-14 Average of the consistent interval was used for value.
16 5-6-14 Average of the consistent interval was used for value.
19 5-6-14 Average of the consistent interval was used for value.
21 5-6-14 Average of the consistent interval was used for value.
23 5-6-14 Slightly sporadic graph was shown for this trial.
29 5-7-14 Average of the consistent interval was used for value.
32 5-7-14 Average of the consistent interval was used for value.
38 5-9-14 Average of the consistent interval was used for value.
40 5-9-14 Average of the consistent interval was used for value.
43 5-9-14 Average of the consistent interval was used for value.
46 5-9-14 Average of the consistent interval was used for value.
47 5-9-14 Average of the consistent interval was used for value.
48 5-9-14 Average of the consistent interval was used for value.
50 5-9-14 Average of the consistent interval was used for value.
52 5-9-14 Average of the consistent interval was used for value.
59 5-9-14 Average of the consistent interval was used for value.
61 5-9-14 Average of the consistent interval was used for value.
63 5-9-14 Average of the consistent interval was used for value.
64 5-9-14 Average of the consistent interval was used for value.
65 5-9-14 Average of the consistent interval was used for value.
69 5-9-14 Average of the consistent interval was used for value.
71 5-9-14 Average of the consistent interval was used for value.
78 5-9-14 Tape to hold down flaps on the bottom of the box started to come off.
95 5-9-14 Average of the consistent interval was used for value.
98 5-9-14 Average of the consistent interval was used for value.
108 5-9-14 Average of the consistent interval was used for value.
116 5-9-14 Average of the consistent interval was used for value.

Buiteweg-Liu 17

Table 6 shows the observations made for the prism face. For the first trial
done for this shape recorded, there was slight variation in the graph shown. Trial
9 turned out to be a perfectly consistent graph. The flashlight and light sensor
were accidentally moved and the instruments had to be readjusted. Trial 23 was
slightly sporadic for the region of the graph an average was taken for, but the
range of the area was small. Trial 78 was when the blackened tape used to hold
down the flaps on the bottom of the box started to peel off. The output remained
in the same range for the rest of the trials.
Table 8
Data for the Cylinder
Cylinder Cylinder
Trial
Illuminance
(lux)
Trial
Illuminance
(lux)
5 42.4 73 33.6
8 37.1 74 37.4
10 34.3 79 36.4
20 34.2 82 33.8
22 33.8 83 34.9
24 34.6 86 35.7
30 35.8 90 32.7
33 33.8 91 31.5
36 34.0 92 33.4
39 37.0 94 30.8
42 36.7 100 36.8
53 35.8 102 29.9
54 34.6 104 33.8
68 36.6 107 34.7
70 33.7 110 35.6

Table 7 shows the data for the cylinder that ranges from 29.9 to 42.4 lux.
The only outlier was the value 42.4 lux for trial 5.


Buiteweg-Liu 18

Table 8
Observation for Cylinder Data
Cylinder
Trial Date Observations
5 5-6-14 Light sensor was zeroed for this trial.
8 5-6-14 Illuminance had to be measured several times.
10 5-6-14 Average of the consistent interval was used for value.
20 5-6-14
Light sensor was zeroed for this trial and the light sensor was given three
minutes to rest.
22 5-6-14 Average of the consistent interval was used for value.
24 5-6-14 Average of the consistent interval was used for value.
30 5-7-14 Average of the consistent interval was used for value.
33 5-7-14
Break was taken before the next trial because of graph inconsistency and flaps
on the bottom of the box were taped down. Tape blackened with marker, but
slight blue color still showed.
36 5-7-14 Average of the consistent interval was used for value.
39 5-9-14
Before the measuring of the trial, Researcher Emily accidentally stepped on the
cylinder, resulting in a slight folding and probably more area directly facing the
flashlight.
42 5-9-14 Average of the consistent interval was used for value.
53 5-9-14 Average of the consistent interval was used for value.
54 5-9-14 Average of the consistent interval was used for value.
68 5-9-14 Average of the consistent interval was used for value.
70 5-9-14 Average of the consistent interval was used for value.
73 5-9-14 Light sensor was zeroed out for this trial.
74 5-9-14 Average of the consistent interval was used for value.
79 5-9-14
Light sensor was zeroed out for this trial and the researchers took a three
minute break to allow the Light sensor to cool down.
82 5-9-14 Average of the consistent interval was used for value.
83 5-9-14 Reading had to be repeated several times.
86 5-9-14 Light sensor was zeroed out for this trial.
90 5-9-14 Average of the consistent interval was used for value.
91 5-9-14 Average of the consistent interval was used for value.
92 5-9-14 Average of the consistent interval was used for value.
94 5-9-14 Average of the consistent interval was used for value.
100 5-9-14 Average of the consistent interval was used for value.
102 5-9-14 Average of the consistent interval was used for value.
104 5-9-14 Average of the consistent interval was used for value.
107 5-9-14 Average of the consistent interval was used for value.
110 5-9-14 Average of the consistent interval was used for value.

Buiteweg-Liu 19

Table 8 shows the observations made during the cylinder data collection.
Trial 5 was zeroed out, along with trial 79, 86 and trial 20, when a three minute
break was taken. Trial 8 had to be measured several times, and the same
problem occurred during trial 83. In trial 33, the flaps on the bottom of the box
started to come up and the flaps were taped down. Blue tape blackened with
marker was used to reduce any light reflection due to the color. However, some
blue coloring still peeked through the layer of marker. In trial 39, the cylinder was
accidentally stepped on and the shape was flattened out slightly. Effort was
made to adjust the shape back to its normal structure.














Buiteweg-Liu 20

Data Analysis and Interpretation
Each of the four shapes tested had 30 trials for a total of 120 trials to fulfill
the requirement of at least 30 trials each for the statistical tests. To reduce
confounding, the shape tested at each trial was randomized. The trials for each
shape were decided using the Ti-Nspires random integer function, first
randomizing which 30 trials of the 120 trials the cone would be tested, then the
trials for which the prism face would be tested, the edge, and then the cylinder
would be tested for the remaining trials. An effort was made to keep the
environment the same for each trial. The shapes were placed in approximately
the same location every trial, the room was kept as dark as possible for each
trial, the light sensor was zeroed out about every five trials, and a three minute
rest was given about every ten trials. It is shown that the experiment can be
replicated from trial to trial if procedures documented are followed. Other
researchers can replicate the experiment and get about the same results.









Buiteweg-Liu 21








Figure 6. Boxplot of Data
Figure 6 shows the distributions of the four shapes data. The cones data
is normally distributed with a mean of 9.96 that is close to the median 9.55, and
no outliers. It doesnt overlap any of the other shapes data and shows that the
cones data is independent of the others. The prism faces data has a mean of
50.16, which is very close to the median of 50, so the data is normally distributed
with no overlap with other populations and no outliers. The prism edges data has
a mean of 23.1033, very close to the mean of 23.35. The graph shows obvious
skewing to the left. This could be due to error during to data collection. But even
though there are three outliers, there were at least 30 trials and so the abnormal
distribution will not affect the statistical tests. The cylinder data contains a mean
of 34.8467, close to the median of 34.6. There is one outlier, but there are at
least 30 trials and the test can be conducted. The four medians are not near
each other and indicate that the populations are independent of each other.
Excluding the outliers, the data from each shape do not overlap and there is an
obvious difference in the amount of light reflected back to the light sensor
Illuminance (lux)
3
.
9

Q
2
:

9
.
5
5


1
5
.
6

4
7

Q
2
:

5
0


5
5
.
5

1
8
.
8

Q
2
:

2
3
.
3
5


2
5
.
9

2
9
.
9

Q
2
:

3
4
.
6


3
7
.
4

Buiteweg-Liu 22

between the shapes. Three two sample t-tests would be conducted to reveal if
the cone provides the lowest illuminance.
A two sample t-test compares two sample means from two independent
populations to see if the difference between the populations is statistically
significant. This particular test was chosen because the researchs objective was
to find which of the four different shapes had the lowest light return and if the
results did not happen by chance alone. Three of the two sample t-tests were
conducted: cone against prism face, cone against prism edge, and cone against
cylinder. According to the hypothesis, the cone was predicted to have the lowest
light return and the statistical tests tested the cone data against the data of the
other shapes. The assumptions that the data was randomized, came from two
distinct populations, and that the samples are independent were fulfilled, as
discussed on page 21 and in figure 6, respectively. There were at least 30 trials
of each sample so any abnormal distribution can be ignored. Population means
and population standard deviations remain unknown. All four assumptions are
met and the results of the two sample t-tests are reliable.
The null hypothesis of the first two sample t-test for the cone against the
prism face states that the mean for the cone is the same as the mean for the
face. The alternate hypothesis states that the cones mean is lower than the
faces mean.
H
o
:
cone
=
face

H
a
:
cone
<
face

Buiteweg-Liu 23







Figure 7. Cone vs. Face Two Sample t-test Results
Figure 7 shows the two sample t-test results testing the cone data against
the prism faces data. The t-value is -62.4151 which led to an extremely small
p-value of about zero (See Appendix G for sample calculation). The mean of the
cone data was 9.96 and the faces was 50.16, which is a huge difference. The
null hypothesis must be rejected because a p-value of zero is much lower than
the alpha level of 0.05. There is very strong evidence to suggest that the mean
illuminance

for the cone is lower than the prism faces. A p-value of zero means
that the sample means obtained would occur 0% of the time by chance alone if
the null hypothesis was assumed true.








Buiteweg-Liu 24







Figure 8. t Distribution with P-Value for Cone vs. Face
Figure 8 shows the t distribution and the p-value for the cone against the
face samples. The p-value cannot even be seen.
The null hypothesis of the second two sample t-test for the cone against
the prism edge states that the mean for the cone is the same as the mean for the
edge. The alternate hypothesis states that the cones mean is lower than the
edges mean.
H
o
:
cone
=
edge

H
a
:
cone
<
edge






Figure 9. Two Sample t-test Results for Cone vs. Edge
Figure 9 shows the statistical test results for the cone vs. edge. The t
value obtained is -17.3092 with a p-value of zero. The mean of the cones data is
Buiteweg-Liu 25

significantly different than that of the edge. The null hypothesis is rejected
because a p-value of zero is lower than the alpha level of 0.05. There is strong
evidence to suggest that the cones mean illuminance

is different than that of the
edges. There is about a 0% chance of getting the sample means found in this
research by chance alone if the null hypothesis was assumed true.






Figure 10. t Distribution with P-Value for Cone vs. Edge
Figure 10 shows the t distribution for the cone against the edge and the p-
value cannot be seen in the distribution.
For the third two sample t-test for the cone against the cylinder, the null
hypothesis states that the mean for the cone is the same as the mean for the
cylinder. The alternate hypothesis states that the cones mean is lower than the
cylinders mean.
H
o
:
cone
=
cylinder

H
a
:
cone
<
cylinder




Buiteweg-Liu 26












Figure 11. Two Sample t -test Results for Cone vs. Cylinder
Figure 11 shows the results of the test for the cone against the cylinder.
There was a t value of -36.4957 for a p-value of about zero. The means for the
cone and cylinder are very different with over a 20 value difference. Reject the
null hypothesis that the means of the populations are the same because the
p-value of zero is below the alpha level of 0.05. There is strong evidence that the
cones mean illuminance is lower than the cylinders. The extreme results
obtained would occur 0% by chance alone if the null hypothesis was true.







Figure 12. t Distribution with P-Value for Cone vs. Cylinder
Figure 12 shows the t-distribution for the cone vs. cylinder samples. The
p-value is so low that it cannot be seen on the graph. There is a very low
Buiteweg-Liu 27

probability of getting such extreme results by chance alone if the null hypothesis
was true.
When examining the distributions of the samples, the cone and face data
distributions were found to be normally distributed. The edges and cylinders
abnormal data distributions can be ignored because the number of data trials
from each population fulfills the Central Limit Theorem. The populations were
also clearly independent. The results of the three two sample t-tests resulted in
p-values of zero, indicating that the cone did have a statistically significant mean
when compared against the other populations. There is strong evidence to
suggest that the cone has a lower mean than the other three shapes.













Buiteweg-Liu 28

Conclusion
The answer to the question of which shape, the cone, rectangular prism
face, rectangular prism edge, or cylinder, will reflect the least amount of light
back to the source was answered during this experiment. Paper models of the
shapes that all had the same height and total surface area were created. The
models were placed into a blackened box in approximately the same location and
when the flashlight was directed on the shape, the light sensor measured the
light return back to the source. It was hypothesized that out of the four shapes,
the cone would have the lowest mean illuminance measured in lux.
The data collected from each shape population was clearly independent
and their means were obviously different from one another. The cone data
ranged from 3.9 to 15.6 lux with a mean of 9.96 lux. The prism face ranged from
47 lux to 55.5 lux with a 50.16 mean lux while the prism edge ranged from 17.7
to 33.9 lux and had a 23.1033 mean lux. The fourth shape, the cylinder, ranged
from 29.9 to 42.4 lux with a mean of 34.8467 lux. It seemed that the cone had the
lowest mean illuminance out of the four shape populations. Conducting three
two-sample t-tests that tested the mean illuminance of the cone against the other
shapes, three p-values of almost zero were calculated. There was strong
evidence to suggest that the cone did indeed have a lower illuminance than the
rest of the shapes and this led to the researchers finally accepting their
hypothesis that the cone has the lowest mean illuminance.
The cones angled and curved surface produced the lowest mean
illuminance compared to the three other shapes because of the angle at which
Buiteweg-Liu 29

the surface reflected the light rays. As the light hits the cone at an angle, the
angle at which it is reflected is the same as the angle at which the light hit the
cone. The angle of the cones surface reflects the light away from the source in
many other directions, little of it actually reflecting back to the source. The prism
face data contained the highest illuminance data because its flat face directly in
the path of the light reflected most of the approaching light back to the source to
be detected. Surfaces that contain angles perpendicular to the ground or to a
radar signal should be avoided when trying to create a shape that has low
illuminance. The cones curved surface did reflect some light away from the
source, but the surface was directly facing the light and did return some of the
light back. The prism edge had the second lowest values of illuminance because
the only surface area to reflect light back to the source would be the edge facing
the light. The rest of the surface in the path of light would reflect light away from
the source.
When compared to the other shapes tested, the cone had the lowest
mean illuminance. If these results were applied to aircraft design, an aircraft with
a curved surface that is angled sharply back from the vertical will have a smaller
radar cross section than an aircraft that is flat and facing perpendicularly to the
radar signal path. It will also have a smaller radar signature than a simply curved
aircraft or an aircraft that has an edge facing the signal. Many studies have
already been completed about the topic of stealth technology involving aircraft
design and other scientists have also confirmed that surfaces angled back from
the vertical would provide lower radar signatures than vertical surfaces; the
Buiteweg-Liu 30

results of this experiment agree with experiments conducted in the scientific
community. Compared to the other studies, this research experiment is a
simplified, more cost-efficient version of experimenting that compared shapes
that may not have been tested for.
The United States Navy in 1993 came out with its Sea Shadow stealth
surface ship. With its new design and technology it was a step forward in
advancing the Navys goal of going undetectable by radar waves. They
concluded that their new stealth ship slowed down the return of enemy radar
waves more than any other navy ship that was tested. The cone in this
experiment produced the lowest amount of illuminance due to the lack of parallel
lines and right angled surfaces.
While the experiment was being conducted a series of problems occurred.
The largest problem that occurred during the experiment was the inconsistency
of the light sensor and flashlight. The light sensor is extremely sensitive, so to
prevent reoccurring misreading the researchers would take a five minute break
for the light sensor to reset its temperature. At trial 62 the flashlight started to dim
and flicker on and off. The batteries were replaced, but data collection was
resumed it was realized that there were significantly higher readings due to the
new brightness of the flashlight. Data collection had to be restarted. At trial 26,
the flashlight began to flicker again and one of the batteries was replaced and the
data remained consistent. The inconsistency of the flashlight led to corrupt data
that had to be redone later throughout the experiment. Also, the flashlight and
light sensor connected to the box were not secured to the point where accidental
Buiteweg-Liu 31

contact would not move them. As the box was moved and the flashlight turned on
and off, both the flashlight and light sensor were slightly moved and had to be
repositioned.
In the beginning of the trials it was noted that the room the experiment
was being conducted in was not completely dark. The experiment was then
moved to a different, darker location with little ambient light. In trial 33 it was
noted that the flaps on the inside bottom of the blacked-out box started to
separate and this could interfere with the data collected. Black tape was used to
secure down the flap. During the experiment, two of the shapes, the cone and
the cylinder, were lightly stepped on and led to these shapes being slightly
folded. The shapes were smoothed back out and there appeared to be no
change in the readings. Lastly, human error occurred during the experiment. With
the lid needing to be held down during data collection there was slight movement
with the box that could have caused a misreading in the light sensor. The shapes
were placed in approximately the same location, but they could have been
placed slightly differently between each trial.
To further conclude this research multiple advances and changes in
techniques could be beneficial to further understand this topic. The effect of the
material the shapes are made out of could be explored; shapes made of wood,
steel, aluminum, and other resources could be tested against each other to see
which would produce the least amount of illuminance. To obtain more consistent
results, a manufactured wood or metal box could be constructed to perfectly hold
the flashlight and light sensor more securely than the box used in the experiment.
Buiteweg-Liu 32

A better, more reliable flashlight could be obtained to avoid the flickering
problem.
This experiment could be beneficial to research and study the government
is paying millions of dollars for towards finding a wing and aircraft design to make
military aircraft almost undetectable by radar. So to make these military aircrafts
almost undetectable by radar the aircraft and its wings need to be at such angles
to where the radar waves can pass right over the aircraft or simply slow down its
return time. Light waves react very much the same as radar waves. This is why
different angled shapes were used. This experiment could possibly save millions
of dollars if taken the proper steps in a laboratory setting.













Buiteweg-Liu 33

Acknowledgements
It would like to be acknowledged for the help and information from Greg
McMillan toward the inconsistency and problems occurring with the light sensor
would like to be acknowledged. Also for the assistance of Caroline Jankowski
would like to be acknowledged for assisting the researchers in conducting their
trial. Finally it would like to be noticed the assistance and help by Rose Cybulski
for helping with the statistical tests done to analyze the data.
















Buiteweg-Liu 34

Appendix A. Cardboard Box Construction
In order to keep the surroundings constant and to minimalize confounding
and external light, the data will be measured in a darkened box.
Materials:
Cardboard Box 40.5 cm x 30.5 cm x 31.5 cm
Black Spray Paint
Box Cutter
Project Board

Procedure:
1. If there are flaps on the cardboard box to close the box, cut off the flaps
with a box cutter. If box cutters are not available, scissors will work.

2. Cover the inside of the box in three layers of black spray paint, allowing
time to dry after each layer. Do the same for the project boards faces that
fold inwards. Add additional layers if the inside of the box is not a solid
black color. Make the box and faces as dark as possible.

3. Cut a hole in which the flashlight fits snugly on the width face of the box
about 12.5 cm from the bottom of the box.

4. Make another hole that fits the light sensor detecting instrument snugly an
inch below it.

5. Fit the flashlight and light sensor through their respective holes and tape
them in place with black tape, making sure that if any accidental pushing
on them will not move them by much. The light sensor will have to be
angled up to be closer to the path of the flashlights light. The ends of the
flashlight and light sensor inside the box should line up so that neither of
the instruments gets in the way of the other.











Buiteweg-Liu 35

Diagram:







Figure 13. Cut-Outs of Box
Figure 13 shows where exactly the two holes were to be cut out in the
box. The flashlight cut-out is larger and above the light sensor cut out. It is to be
noticed that if the hole is made slightly smaller than the actual size of the
flashlight or light sensor, both will fit more securely.











Buiteweg-Liu 36

Appendix B: Cone Construction
One of the shapes tested was the cone shape. The shapes were
self-constructed.
Materials:
Scissors Poster Paper
Tape Ruler
Pencil Protractor

Procedures:
1. Take the ruler and draw a 20.5 cm line on the poster paper.

2. Take the protractor and measure a 78.55 from the line. Draw a 20.5 cm
line there, intersecting the end of the two drawn line segments at a point.

3. Using the protractor and ruler, rotate the ruler between the line segments,
and around the intersecting point between the line segments, thus
creating an arc. As an arc is being made, make six marks at the end of the
ruler not at the intersecting point. Connect the marks in an arc.

4. Cut out the fan-like shape.

5. Curl the paper around so that the two 20.5 cm lengths touch and tape
them together.

Diagram:







Figure 14. Paper Cone
Figure 14 shows the paper cone that was used during the experimental
trials.
Buiteweg-Liu 37

Appendix C: Rectangular Prism Construction
Two of the four shapes being tested involved a rectangular prism.
Materials:
Scissors Poster Paper
Printer Tape
Paper 8.5 in x 11 in Ink

Procedures:
1. Print out Figure 16 in the diagram below on an 8.5 in x 11 in paper from
the printer.

2. Cut along the borders of the print-out.

3. Line the cut-out with the poster paper and use the scissors to cut out a
similar shape on the poster paper as the one on the paper.

4. Fold along the borders of the shape on the poster paper and where the
edges touch, tape them together.

Diagram:







Figure 15. Paper Rectangular Prism
Figure 15 shows the paper rectangular prism that was used in two
different ways during the experimental trials. The left picture shows it used as a
face, and on the right is the shape shown as an edge.
Buiteweg-Liu 38


















Figure 16. Rectangular Prism Cut-Out
Figure 16 above shows the rectangular prism outline used to make the
shape.



Buiteweg-Liu 39

Appendix D: Cylinder Construction
One of the four shapes tested by the researchers was the cylinder.
Materials:
Scissors Poster Paper
Printer Tape
Paper 8.5 in x 11 in Ink

Procedures:
1. Print out Figure 18 in the diagram below on an 8.5 in x 11 in paper from
the printer.

2. Cut along the borders of the print-out.

3. Line the cut-out with the poster paper and use the scissors to cut out a
similar shape on the poster paper as the one on the paper.

4. Curl the cut out poster paper so that the longer edges touch and then tape
them together. The diameter should be 4.98 cm.

Diagram:

Figure 17. Paper Cylinder
Figure 17 shows the paper cylinder that was used during the experimental
trials.
Buiteweg-Liu 40






















Figure 18. Cylinder Cut-Out
Figure 18 shows the outline for the cylinder used to make the shape.
Buiteweg-Liu 41

Appendix E. LabQuest and Light Sensor Set-Up
The researchers used a LabQuest and light sensor to measure the
illuminance for the data.
Materials:
Vernier Software Light sensor
LabQuest

Procedure:
1. Connect the light sensor to the LabQuest in Channel one.

2. Press the power button on the LabQuest to turn it on. Choose New from
the File menu.

3. On the red rectangle that appears on the screen that appears, tap on it
and click Zero.

4. On the Meter Screen, select Sensor at the top of the Screen. Select CH1
and scroll down to Light. Select 600 Lux. Select OK.

5. On the Meter Screen, tap Rate. Change the data-collection length to 15
seconds. Select OK.

6. Click the reading that shows up

7. Data collection can now be started.

8. To start the data collection click the green arrow in the bottom left hand
corner.

9. To stop data collection click the red square that will be in place of the
green arrow when data collection has started or let the time run out.

10. The value used in the data tables is the average of a consistently
horizontal interval that appears on the graph. Find the average by clicking
Analyze, Statistics, and then Illumination. The average appears in the
box to the right.



Buiteweg-Liu 42

Appendix F. Ti-Nspire CX Trial Randomization
The trials needed to be randomized in order to reduce confounding and
bias. There were 30 trials for each four shapes for a total of 120 trials.
Materials:
Ti-Nspire Random Integer Function
Procedure:
1. Open a new Calculator page.

2. Press the menu button.

3. Select option five, or Probability, in the drop down menu.

4. Select option four, or Random, in the drop down section that appears.

5. Select option two, or the Integer function. The words randInt() will
appear on the calculator page. Type (1, 120, 30). For the first 30 non-
repeating numbers that appear, those are the trials assigned to the first
shape, the cone. Record the trial numbers and shape in the data table.

6. If a number is repeated, repeat step five.

7. Repeat steps two through six for the prism face, edge, and cylinder.









Buiteweg-Liu 43

Appendix G. Two Sample t -test Sample Equation
To see whether the null or alternative hypothesis is accepted or rejected, a
p-value needs to be calculated. It can be found once the t-value is found.


Shape one and shape two corresponds with the first shape and second
shape being compared in the test. The means of the shapes and standard
deviations of the shapes are used in the equation, represented by and s,
respectively. The variable n is the number of samples from each population.
Once the t-value is found, it is used to find the p-value. The p-value represents
the probability that the results obtained would occur by chance alone if the null
hypothesis is true. Below is a sample calculation on how the t value was found.



Figure 19. Two Sample t-test Sample Calculation CHANGE TO FIG
Figure 19 is a sample problem of how the t value was found for the cone
against the prism face. The t value obtained was -62.415132. This t-value is the
number of standard deviations away from the mean, and this value is quite a high
value. The p-value was found to be zero.

Buiteweg-Liu 44

Works Cited
2004. MI6 The Home of James Bond 007. Web. 27 May 2014. < http://www.mi6-
hq.com/news/index.php?itemid=1113&t=mi6&s >
2007. HowStuffWorks. Web. 27 May 2014.
<http://science.howstuffworks.com/question69.htm>.
"Anatomy of an Electromagnetic Wave. Mission:Science. 2010. National
Aeronautics and Space Administration. 13 Apr. 2014. Web. 13 Apr. 2014.
<http://missionscience.nasa.govems/02_anatomy.html>.
Ashley, Steven. "Stealth ship undergoes sea trials." Mechanical Engineering-
CIME June 1993: 68+. General OneFile. Web. 13 Apr. 2014. <http://
go.galegroup.com/ps/i.do?id=GALE%7CA13185613&v=2.1&u=lom_acces
smich&it=r&p=ITOF&sw=w&asid=f3bb5e58f27ce49df53b7c7a414aebc0>.
Brain, Marshall. "How Radar Works." HowStuffWorks Inc. 01 Apr. 2000. Web. 14
Apr. 2014. <http://science.howstuffworks.com/radar.htm>.
"Electromagnetic Waves." Fermi National Accleration Labratory. Fermilab, 18
June 1999. Web. 28 Apr. 2014.
<http://home.fnal.gov/~pompos/light/light_page3.html>.
Fellers, Thomas J., and Michael W. Davidson. "Reflection of Light."
Molecular Expressions Optical Microscopy Primer. National High Magnetic
Field Laboratory, Florida State University. 3 June 2004. Web. 13 Apr.
2014. <http://micro.magnet.fsu.edu/primer/lightand
color/reflectionintro.html>.

Buiteweg-Liu 45

Giancoli, Douglas C. Physics-Principles with Applications. Sixth Edition. New
Jersey: Pearson Education Prentice Hall. 2005. 615-628, 632-646. Print
Reflection. University of Texas at Austin McDonald Observatory. 2003. PDF
File. Web. 28 Apr. 2014.
<http://mcdonaldobservatory.org/sites/default/files/reflection.pdf>.

Vous aimerez peut-être aussi