Citation
Study of wind loads applied to solar panels on flat rooftops

Material Information

Title:
Study of wind loads applied to solar panels on flat rooftops
Creator:
Plas, Rachelle Kay ( author )
Language:
English
Physical Description:
1 electronic file (114 pages) : ;

Subjects

Subjects / Keywords:
Winds -- Speed ( lcsh )
Wind-pressure ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
The goal of the research presented in this thesis is to determine the design wind load on rooftop mounted solar panels by collecting data to be compared with previous research, wind tunnel tests, and the values in current standards. The data collected was measured form structurally suspended, faux solar panels that were installed at the University of Colorado Denver Campus on a flat roof. Strain transducers were used to measure the force on the panels due to wind. The key value in this research is the Coefficient of Force, CF, which was derived from the collected data. Also, the natural frequency of the faux solar panels was determined to investigate if the vibration of the panel impacted the results of the research . ( , )
Review:
The key findings of this research are average CF values of 1.2 for Panel A and 1.8 for Panel B. It was also found that the vibration of the panels did not have an influence on the strain transducer recordings nor the CF values.
Thesis:
Thesis (M.S.) - University of Colorado Denver.
Bibliography:
Includes bibliographic references
System Details:
System requirements: Adobe Reader.
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Rachelle Kay Plas.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
945987074 ( OCLC )
ocn945987074
Classification:
LD1193.E53 2015m P53 ( lcc )

Downloads

This item has the following downloads:


Full Text
STUDY OF WIND LOADS APPLIED TO SOLAR PANELS
ON FLAT ROOFTOPS
by
RACHELLE KAY PLAS
B.S., University of Colorado Denver, 2010
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering
2015


This thesis for the Master of Science degree by
Rachelle Kay Plas
has been approved for the
Civil Engineering Program
by
Frederick R. Rutz, Chair
Peter Marxhausen
Bruce Janson
November 19, 2015


Rachelle Kay Plas (M.S., Civil Engineering)
Study of Wind Loads Applied to Solar Panels on Flat Rooftops
Thesis directed by Assistant Professor Frederick R. Rutz
ABSTRACT
The goal of the research presented in this thesis is to determine the design wind
load on rooftop mounted solar panels by collecting data to be compared with
previous research, wind tunnel tests, and the values in current standards. The
data collected was measured form structurally suspended, faux solar panels that
were installed at the University of Colorado Denver Campus on a flat roof. Strain
transducers were used to measure the force on the panels due to wind. The key
value in this research is the Coefficient of Force, Cf, which was derived from the
collected data. Also, the natural frequency of the faux solar panels was
determined to investigate if the vibration of the panel impacted the results of the
research.
The key findings of this research are average Cf values of 1.2 for Panel A and
1.8 for Panel B. It was also found that the vibration of the panels did not have an
influence on the strain transducer recordings nor the Cf values.
The form and content of this abstract are approved. I recommend its publication.
Approved: Frederick R. Rutz


DEDICATION
I dedicate this work to my parents Rick and RaNae Urso and my husband, Seth
Plas, thank you for your support, encouragement and patience throughout this
journey. I am so very lucky to have such supportive and encouraging people in
my life. I would not be where I am today without you.
IV


ACKNOWLEDGMENTS
There are so many individuals who have helped me throughout this process. I
would like to give a tremendous thank you to my thesis advisor and past
professor, Dr. Frederick R. Rutz. Your passion for engineering and education
has inspired me countless times throughout my theses and coursework with you.
Even in my undergraduate classes with you, your thirst for knowledge and
excitement in teaching others has always stood out.
Erin Andolsek and Jennifer Harris, your knowledge and assistance with this have
been so helpful. You have really set the bar high in your work and have made
such strides in your research in this field. Your creativeness and unique
solutions have made this research possible.
Tom Thuis of the Electronic Calibration and Repair Lab, thanks for manufacturing
the new legs for the panels. Your assistance is very much appreciated.
I would also like to thank the Auraria Higher Education Campus Facilities
Department for allowing us to utilize the roof of the Events Center Building. This
has been a great location for our research.
Thank you Rick Urso for helping me assemble the new panels in extremely cold
weather. I could not have done that without you.
v


And, thank you Fred, Jennifer & Jesse for your help in disassembling and
disposing of the panels after all of the data was collected. It would have been a
daunting task without your help.
Finally, thank you Jared for your help with analyzing the data recorded from the
accelerometers. Your input was very important to our results.
VI


TABLE OF CONTENTS
CHAPTER
I. OVERVIEW......................................................
Introduction...................................................
Goal...........................................................
Procedure......................................................
II. SOLAR ENERGY.................................................
Introduction...................................................
History........................................................
Solar Panels Today.............................................
III. WIND BEHAVIOR...............................................
Introduction...................................................
Wind Tunnel Testing............................................
Current Codes and Standards....................................
IV. PREVIOUS STUDIES.............................................
Introduction...................................................
Study of Wind Loads Applied to Full-Scale Rooftop Solar Panels by
Jennifer Harris................................................
Study of Full Scale Rooftop Solar Panels Subject to Wind Loads by Erin
Andolsek.......................................................
V. PANEL DESIGN, LOCATION AND EQUIPMENT.........................
Location.......................................................
Panels.........................................................
.. 1
.. 1
..2
..3
..4
..4
..4
..5
..7
..7
..7
..9
19
19
24
30
35
35
37
VII


Equipment...................................................41
VI. THEORY....................................................45
VII. VIBRATION OF PANEL.......................................47
Introduction................................................47
Panel Natural Frequency.....................................47
Conclusion..................................................61
VIII. RESULTS.................................................63
Introduction................................................63
Results.....................................................63
Discussion..................................................91
IX. CONCLUSION..............................................94
Conclusion..................................................94
Possible Sources of Error...................................95
Recommendations for Further Research........................96
REFERENCES......................................................97
APPENDIX A......................................................99
Datalogger Program............................................99
viii


LIST OF FIGURES
Figure
3.1 Wind Flow against a Bluff Body......................................11
3.2 External Pressure Coefficients.......................................13
3.3 Net Pressure Coefficients for Monoslope Free Roofs..................15
3.5 Figure 29.9-1 from SEAOC Document...................................18
4.1 Cross section of solar panel experiment.............................20
4.2 Proposed panels with relation to the shear layer....................21
4.3 Diagram for the panel construction.................................22
4.5 Panel Frame Connection Details.at the Diagonal Tension Ties........25
4.8 First Panel Placement...............................................27
4.9 Second Panel Placement..............................................31
4.10 Construction tape Experiment 80 feet from Roof Edge...............32
4.11 Wind Velocities from the three anemometers........................33
5.1 Physical Education Building..........................................35
5.2. Wind Region.........................................................36
5.3. Panel Legs..........................................................38
5.4. Panel Detail Section. (Not to scale)................................39
5.5. Layout. (Not to scale)..............................................40
5.6. Strain Transducer Locations.........................................41
5.7. Study Location and Layout...........................................43
5.8. Study Location and Layout...........................................44
7.1. Accelerometer Test Set Up...........................................48
7.2. Accelerometer Test Set Up...........................................49
IX


7.3. Accelerometer Test Set Up..............................................49
7.4. Heal Drop, all three sensors...........................................50
7.5. Heal Drop, Sensor 0: on panel..........................................51
7.6. Heal Drop, Sensor 1: panel leg, perpendicular to panel.................52
7.7. Heal Drop, Sensor 2: panel leg, parallel to panel......................53
7.8. Leg Tap 1, all three sensors...........................................54
7.9. Leg Tap 1, Sensor 0: on panel..........................................55
7.10. Leg Tap 1, Sensor 1: panel leg, perpendicular to panel.................56
7.11. Leg Tap 1, Sensor 2: panel leg, parallel to panel......................57
7.12. Leg Tap 2, all three sensors..........................................58
7.13. Leg Tap 2, Sensor 0: on panel..........................................59
7.14. Leg Tap 2, Sensor 1: panel leg, perpendicular to panel.................60
8.1. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel A.............64
8.2. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel B.............66
8.3. Cf Values, Strain, and Wind Velocity from 4/27/14, Panel A.............68
8.4. Cf Values, Strain, and Wind Velocity from 4/27/14, Panel B.............70
8.5. Cf Values, Strain, and Wind Velocity from 4/27/14, Panel A.............72
8.6. Cf Values, Strain, and Wind Velocity from 4/27/14, Panel B.............74
8.7. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel A.............76
8.8. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel B.............78
8.9. Cf Values, Strain, and Wind Velocity from 4/27/14, Panel A.............80
8.10. Cf Values, Strain, and Wind Velocity from 4/27/14, Panel B.............82
8.11. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel A.............84
8.12. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel B.............86
x


8.13. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel A..............88
8.14. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel B..............90
XI


LIST OF TABLES
Table
8.1. Cf Values from Figure 8.1.......................................65
8.2. Cf Values from Figure 8.2.......................................67
8.3. Cf Values from Figure 8.3.......................................69
8.4. Cf Values from Figure 8.4.......................................71
8.5. Cf Values from Figure 8.5.......................................73
8.6. Cf Values from Figure 8.6.......................................75
8.7. Cf Values from Figure 8.7.......................................77
8.8. Cf Values from Figure 8.8.......................................79
8.9. Cf Values from Figure 8.9.......................................81
8.10. Cf Values from Figure 8.10.....................................83
8.11. Cf Values from Figure 8.11.....................................85
8.12. Cf Values from Figure 8.12.....................................87
8.13. Cf Values from Figure 8.13.....................................89
8.14. Cf Values from Figure 8.14.....................................91
8.15. Average Peak Cf from Figures 8.1 8.14.......................92
xii


CHAPTER I
OVERVIEW
Introduction
Solar panels are becoming more and more popular as society moves towards the
use of environmental friendly sources of energy. They are now being installed
commonly in residential and commercial areas. Because they are typically
positioned in a way to maximize their exposure to the sunlight, rooftops are an
ideal location. Solar panels are only expected to continue to grow in popularity
as technology progresses and they become more efficient and affordable.
Design professionals typically use engineering standards, such as ASCE 7
(ASCE 7 2010), which does not say what type of wind loads should be applied to
roof mounted solar panels. This leaves the design professionals to use their
judgment to determine the wind loads for these structures. While there is some
guidance available, it is based on small scale wind tunnel testing. Most of the
research that has been done is proprietary, and because of the costs of these
tests, there is not a lot of literature available. While there is the need for full-
scale validation of the results from these wind tunnel tests, there have been very
few studies that have looked at this.
When wind tunnels were first introduced, the original wind tunnel studies were
validated with significant amounts of full scale research (Cochran 2010). It
1


makes since that the next step in determining the design wind loads for solar
panels on flat rooftops is to verify the results from the wind tunnel tests with full
scale models.
Over the past few years, there have been two full scale faux solar panel
experiments at the University of Colorado Denver Campus. These have been
used to collect data to compare to the results of the wind tunnel studies. These
panels were fabricated to record the wind pressure, velocity, and direction as well
as the wind force, derived from strain measurements in the tension ties of the two
panels. These are the same panels that are used in this research; however, the
height has been adjusted to better represent the typical installation of solar
panels on flat rooftops.
Goal
The goal of this research is to use the data recorded from the faux solar panels
to derive a coefficient that can be compared to the current standards and existing
wind tunnel research. Although efforts to determine wind loads on solar panels
have been ongoing for some time, few full-scale experiments have been reported
(Harris 2013). This research will also determine the natural frequency of the
solar panels used in this research to determine if the vibrations of the panel could
have influenced the recorded strains.
2


Procedure
Real time data was collected from the two faux solar panels at the University of
Colorado Denver Campus. Wind velocity and direction was recorded along with
the stain for two faux solar panels. The wind velocity and barometric pressure
were used to find the force on the panels due to the wind. The strain
measurements obtained were used to find the net force acting on the faux solar
panels. The Coefficient of Force, Cf, was then determined from the ratio of these
two values. This can be compared to the Coefficient of Pressure, Cp, which is
used in ASCE 7 (ASCE 7 2010). The comparison of these two coefficients can
show how the measured pressure, Cf, and the pressure form the wind tunnels
that are used in the current standards, Cp, relate to each other.
3


CHAPTER II
SOLAR ENERGY
Introduction
Solar Energy is the most abundant energy resource on earth (US Department
of Energy, 2012). There is no doubt that solar energy will continue to grow in
popularity and be seen in a variety of applications in the future. Today, you can
buy solar powered products anywhere form rooftop panels, to yard and
Christmas lights. There are even several solar do-it-yourself kits available. Solar
technology has been around for a long time, and over the past several years,
there have been great advances in this technology, which is only expected to
continue.
History
Silicon photovoltaic (PV) cells were developed in the 1950s, however, according
to the U.S. Department of Energy, the earliest documented uses of concentrating
the suns energy date back to the 7th Century B.C, when magnifying glasses
were used to start fire (History of Solar 2012). During the next few thousand
years, there have been many different ways that sunlight was used to create fire
and to warm rooms.
In the 1700s, Horace de Suassure used the observation that a room, a carriage,
or any other place is hotter when the rays of the sun pass through glass to build
4


a hot box (Butti, 1980). He started with assembling a box with layers of glass,
similar to a greenhouse. He used this box to cook fruits. He then constructed a
similar box, with insulated wood sides and a layered glass top. This became
known as a hot box. This idea was later used by other scientists and
astrophysicists across the world to create similar boxes to cook food.
In 1873, the photoconductive characteristic of Selenium was discovered by
Willoughby Smith. This discovery showed that light could be converted to
electricity, but it did not produce enough energy to power electrical equipment
(History of Solar 2012). In the early 1900s, there were several developments in
this field both in the United States and in Europe.
The next big advance in solar technology was in 1954 when Bell Labs introduced
the first Photovoltaic Technology (PV) device that generated usable amounts of
electricity. The first cell had only 4% efficiency, but later got up to 11% (History
of Solar 2012). The next year, PV powered products started to emerge on the
market. Since then, the advances in this technology have been remarkable.
Solar Panels Today
As solar panels are becoming more common, the costs are coming down and the
efficiency is increasing. They have also become more cost effective to install as
experience is becoming more common. There are numerous applications for
these panels that convert sunlight into usable energy. This is done by
5


establishing an electric field with PV cells. The PV cell consists of at least two (2)
semiconductor layers, one with a positive charge and the other with a negative
charge. As photons from the sunlight are absorbed by the cells by the negative
layer, this frees up the electrons. The electrons than travel to the positive layer,
and this creates a voltage differential. This flow of electrons creates a current,
which can be used for electricity. One of these cells creates only 1 or 2 watts, so
several of the cells are combined to form a module. Modules are combined to
create an array for the desired amount of energy. A glass surface is on top of the
PV cell arrays for protection. Under the glass, there is an antireflective layer to
increase the amount of light that reaches the PV cells.
The United Stated Federal Government offers tax credits to residential and
commercial properties that have installed solar panels to provide electricity not
only for their use, but also contributing to the local energy grid. This is a great
incentive to counter the high initial cost of these systems. There are also several
avenues to lease solar panels and take advantage of the energy produced.
6


CHAPTER III
WIND BEHAVIOR
Introduction
Since the late 1940s, there has been an increased amount of research into wind
loading on structures throughout the world. Wind and seismic loads are both
considered to determine the controlling lateral load to use in the design of a
structure. This usually depends on the location. While wind events are more
common, earthquakes usually result in more damage. Over the years, wind
storms and earthquakes have on average created the same amount of damage
(Holmes 2007). Due to the frequency of wind events, they impact more people.
Wind Tunnel Testing
The first solar panel test in a wind tunnel was conducted in the 1970s. Since
then, there have been several different tests done with a variety of panels,
placements, geometry, arrays and different environments.
ASCE 7-10 Section 31.2 covers the requirements that wind tunnel test need to
follow, these include scaling, modeling, and instrumentation in the wind tunnel
set up (Maffei, 2014). For wind tunnel tests on roofs of buildings, scaled models
of the buildings are created and tested based on scaled wind load values.
Models used in wind tunnel tests are typically simple buildings that do not
consider several factors that can influence the wind pressure: parapets,
7


landscaping, unique geometry of the buildings, signs, etc. It has been found that
there are significant differences reported among studies (Stathopoulous, et al.
2012). It has also been found that the wind pressures calculated via building
code interpretation were less than the measured pressures from wind tunnel
testing, in part due to the fact that the maximum average pressures measured on
large surface areas are less that the maximum pressures measured on small
surface areas (Tilley, 2012). This is even more of a reason to compare the
results to those from actual size models. The design of these panels could be
consistently under designed if this is the case. There was a study done for solar
panels mounted on a pitched roof looking at both wind tunnel testing and a full
scale test. The results of that study showed that the full scale pressure
coefficients were greater than those from the wind tunnel testing (Stathopolous
2012).
Because of the way that the wind tunnel tests are set up, the pressure tape is
significantly large compared to the scale of the panels. This can easily produce
inaccurate results. Also, these tests have shown that there is some sheltering
occurring the further away the panels are from the outside of the array. The
panels on the edge can experience two to three times greater wind loads than
the interior panels (Banks 2011).
8


Wind tunnel testing is currently the only way to supplement the wind loads in
ASCE7 (Cochran 2010). In fact, most of the written standards are based on the
results of wind tunnel testing.
Current Codes and Standards
The most commonly used standard for engineers in the United States is the
American Society of Civil Engineers (ASCE) standard number seven, Minimum
Design Loads for Buildings and Other Structures ( ASCE 7). This standard is
referenced in the International Building Code (IBC) which is typically adopted by
the local Building departments. For the design of a solar panel, this standard
uses the wind speed to determine the pressure to be used for the design. This is
then used to determine the forces acting on the panel. This will dictate how the
panel is secured to the roof. By attaching a solar panel to the roof, the wind load
acting on the roof surface is not increased, but the structural member will need to
be designed for the panels uplift forces (Banks 2011).
The movement of air can be compared to that of a liquid, as they are both fluids.
That is why the main equation for the design wind pressure is based on
Bernoullis principal.
V=\pV2 (1)
p = density of the fluid
V = velocity of the fluid
The velocity pressure of wind in ASCE7-10 is equation 27.3-1:
9


(2)
qz = 0.00256KzKztKdV2
Kz = velocity pressure coefficient
Kzt = topographic factor
Kd = wind direction factor
V = wind velocity
This is essentially the same as Bernoullis equation, with a few factors to address
the conditions of the site. In this equation, the first value, 0.00256 is to convert
the density of fluid to air at sea level in U.S. customary units. The Kz accounts for
the height above the ground and the exposure of site. This can impact the
surface drag and the mean wind flow. The Kzt takes into consideration the
surrounding topography. Kd is for the angle of the wind. Once the velocity
pressure is determined, the methods in the ASCE7-10 can be used to find the
wind load.
As wind approaches a structure, it is redirected around it: to the sides,
downwards, and upwards. At the top of the wall, there is a separation point and
a shear layer is generated at a slope of 2:1 (SEAOC 2012). Above this, the
streamline flow continues; below is a region of vortices. The area below the
shear layer is the re-circulating region, which produces high uplift force. This
diminishes further away from the edge of the building. The figure below depicts
this behavior in the case of a building with a parapet wall.
10


Figure 3.1 Wind Flow against a Bluff Body.
(Andolsek 2013, used with permission from Erin Andolsek).
ASCE 7 offers guidance on the design wind speeds and pressures for buildings,
rooftop structures, components and cladding, but does not cover the design for
solar panels on rooftops. The IBC does not offer guidance either. Due to this
lack of guidance, the design engineer is left with a few options. One option is to
use results from wind tunnel testing. Another is to use the existing tables in
ASCE7 that are intended for different design purposes, but can be applied to the
design of solar panels on flat roofs. Or, the design engineer can used the set of
tables that SEAOC has provided to assist with this design. In all of these
approaches, the Kz, Kzt, Kd, and the importance factors used are the same as
what would be used for the design of the actual building, as it is not typical that
the components for a building are designed to higher standards than the actual
building.
11


Using the ASCE tables for roofs, but not necessarily solar panels on a roof, the
external pressure coefficient for flush mounted solar arrays is from Figure 30.4-1
of the ASCE7-10, Figure 3.2 below. The intended use of this table in the
standard is for components and cladding for enclosed and partially enclosed
buildings. The external gust pressure coefficients, GCp, for the roof can be used.
This method should have conservative loads (Banks 2011). The pressure
coefficient is then used to calculate the net pressure.
12


Component! and Cladding h£60fl.
| Figure 30.4-I | External Pressure Coefficients, GQ, Walls
| Enclosed, Partially Enclosed Buildings
(0.1) (0.9) (15) (4.6) PIMl (5)(92.91
Effective Wind Area, ft2 (n^}
Notes:
1. Vertical scale denotes GC, to be used with qr-
2. Horizmtal scale denotes effective wind area, in square feet (square meters).
3. Plus and minus signs signi fy pressures acting toward and away from the surfaces, respectively.
4. Each component shall be designed for maximum positive and negative pressures.
5. Values of GCP tor walls shall be reduced by 10% what 9 £ 19.
6. Notation:
a: 10 percent of least horizontal dimension or 0.4h, whichever is smaller, but not less than either 4%
of least horizontal dimension or 3 ft (0.9 m).
A: Mean roof height, in feet (meters), except that cave height shall be used for 9 S UP.
9: Angle of plane Df roof from horizontal, in degrees.
Figure 3.2 External Pressure Coefficients.
Figure 30.4-1 from ASCE 7-10 used to determine external pressure coefficients
on components and cladding of enclosed buildings (ASCE 7-10 2010, used with
permission from ASCE).
13


Figure 27.4.4 from the ASCE7-10, Figure 3.3 below, is used to find the net
pressure coefficient for monopole free roofs on the ground. This can also be
used as guidance in the design of solar panels on flat roofs if they are placed far
enough away from the edge of the roof. The gust factor, G, used for this is
typically 0.85 for rigid structures, but it has been recommended that this be
increased to 1.0 for the design of solar panels (Banks 2011).
14


Main Wind Force Resisting System Pari 1 0.25 £ Wl.£ 1.0
Figure 17.4-4 | Nil Pressure Cucflkitul, CN Monoslope Free Roofs 0£45,Y = O, 180
Open Buildings
Roof Load Wind Direction, y = 0" Wind Direction. '/= 180
Angle Case Clear Wind Flow Obstructed Wind Flow Clear Wind Flow Obstructed Wind Flow
0 Cnw Cni_ Ckw Cnl Cnw Cnc Lnw t-NL
0 A 1.2 0.3 -0.5 -1.2 1.2 0.3 -0.5 -1.2
R -1.1 -0.1 -1.1 -0.6 -1.1 -0.1 -1.1 -0.6
7.5s A -0.6 -1 -1 -1.5 0.9 1.5 -0.2 -1.2
R -1.4 0 1.7 -0.8 1.6 0.3 0.8 -0.3
15 A -0.0 -1.3 -1.1 -1.5 1.3 1.6 0.4 -1.1
R -1.9 0 -2.1 -0.6 1.8 0.6 1.2 -0.3
? A -1.5 -1.6 -1.5 -1.7 1.7 1.8 0.5 -1
B -2.4 -0.3 -2.3 -0.9 2 2 0.7 1.3 0
30 A -l.S 1.8 -1.5 -1.8 2.1 2.1 0.6 -1
R -2.5 -0.5 -2,3 -1 1 2.6 1 1.6 0.1
37.5s' A -1.8 -1.8 -1.5 -1.8 2.1 2.2 0.7 -0.9
R -2.4 -0.6 -2.2 -l.l 2 7 1.1 1.9 0.3
45 A -1.6 -1.8 -1.3 -1.8 2.2 2.5 0.8 -0.9
R -2.3 -0.7 -1.9 -1.2 2.6 1.4 2.1 0.4
Notes:
1. Cnw and Cm denote net pressures iconinbutions from lop And bottom surface*) for uimhvnrd and leeward half of
roof surges, respectively
2. Clear Wind floov denotes relatively uncibstructcd Wind tlov with blockage leu than or equal Lu 50%. Obstructed
wmd ftow denotes objects bduw roof inhibiting wind Ifow <>50% blockage).
3. far values ol B between 7.5* and 45". linear inlcrpstLihon is permitted. For values of 0 less than 7.5, use load
coefficients for O.
4 Plus and minus signs signify pressures riding towards and rwuy from the top roof surface, respectively
5. All load eases shown Icir each nxif angle shall be investigated.
6. Notation.
1. hr>Ti?onMl dimension of roof, measured in the along wind direction, fi (ml
h mean rcwlhcight. IV |m|
y direction of wind, degrees
tl angle uf plane aJ raaf iram horizontal, degrees
Figure 3.3 Net Pressure Coefficients for Monoslope Free Roofs.
Figure 27.4-4 from ASCE7-10 used to determine wind pressure on monoslope
free roofs above ground (ASCE 7-10, used with permission from ASCE).
15


It has become standard practice for design engineers to use standards for other
structures in the design of solar panels of flat roofs. There is a great need for a
standard for this purpose. This need was recognized by the Structural Engineers
Association of California (SEAOC), and they formed a committee for this issue in
2011. They identified many key issues to investigate, and in 2012, SEAOC
published Wind Design for Low-Profile Solar Photovoltaic Arrays on Flat Roofs.
This document contained two figures to guide engineers in determining the
design wind loads for roof mounted solar panel arrays. See Figures 3.4 & 3.5
below. This considers the location of the panel on the roof, the shape of the roof,
and the roof features. This document uses equation 3 below to determine the
pressure.
P Qh(fiCrn) (3)
p = velocity pressure at mean roof height
qh = velocity pressure evaluated at the mean roof height
GCrn = combined net pressure coefficient
ASCE does intend to include a standard similar to this in its next edition to help
address this lack of guidance. Until then, these tables are the best tool that the
design engineers have.
16


Figure 3.4 Design GCp Values Published by SEAOC, August 2012.
(SEAOC 2012, used with permission from SEAOC).
17


Design Wind Loads h < 60 ft, or all heights if h Figure 29.9-1 (com,) | Roof Mourned Solar Panel Arrays Roofs 0 <7
Enclosed, Partially Enclosed Buildings
Notes: 1. (GCJ acts towards and away from the panels top surface.
2. There shall be a minimum air gap around ihc perimeter of each solar module of 0.5 inches or between
rows of panels of 1 inch to allow pressure equalization above and below panels.
3. Alternatively, for a 0*, h, £ 10", and air gap per note 2, use components and cladding procedure per
ASCE 7-10 30.4 (ASCE 7-05 6.5.12.4).
4. Array should not be closer than 2( k< h?) or 4 feet, whichever is greater, from roof edge.
5. Roof structure area covered by solar array need not to be designed for simultaneous app lication of solar
array wind loads and roof components and cladding wind loads. As a separate load case, roof structure
shall also be designed for full roof components and cladding wind loads assuming PV panels are not
present,
6. Notation:
A: Effective wind area for structural clement being designed, in ft.
A.: Normalized wind area, equal to f--------1000 ,
\(max(aIK. 15ft))'
a: 0.5 V feHf, but need not exceed h, in ft.
d,: Horizontal distance measured orthogonal to the panel edges in the north (d), south (ds), east
(dt), and west (dw) direction, from panel being evaluated to adjacent panel or building edge,
whichever is closer, ignoring any rooftop equipment, in ft. For panels in a row, d£ and arc
measured from the end of the row in their respective direction. E£ and Ew apply only to the
panels within 5 ft of each end of the row on their respective side, and panels greater than 5 ft
from both ends of their row shall have de and dr =0.
E: Array edge factor calculated for each panel area in each principle direction at a time, equal to
maximum of £v, Es, Ee, E. If panel area being evaluated is located in zone 2 or 3 and ds
measured to building edge ignoring all other panels is greater than 3 a,,, then Es for that panel
area need not exceed 1.5. If panel area being evaluated is located in zone 2 or 3 and ds, dL, or
dr measured to building edge ignoring all other panels is greater than 3 athen Es, Ee, or Ew for
that panel area in only that respective direction need not exceed 1.0.
(CCm): Net pressure coefficient, equal to yj,£[(GC),,(yc)]
(GC*),,m Nominal net pressure coefficient.
h : Mean roof height above ground, except for monoslope roofs use maximum roof height, in ft.
A,: Solar panel height above roof at low edge, in ft.
ht: Solar panel height above roof at raised edge, in ft.
h,: Characteristic height, equal to min{ h,, 1 ft)+/,sin(w), except when evaluating E toward a
building edge unobstructed by panels, then hc=0. la,, for that panel in that direction, in ft.
hj: Mean parapet height above adjacent roof surface, in ft.
If Chord length of solar panel, in ft.
WL: Width of overall building on longest side, in ft.
Ws: Width of overall building on shortest side, in ft
£ Panel chord length factor, equal to 1.0 for to < 5, equal to 0.6 + 0.06/, for u> >15* but shall not
be less than 0.8. For 5*< ai< 15, apply yr only to 15*- 35* figure values and prior to
interpolating.
£ Parapet height factor, equal to l.Ofor 4 ft, equal to the smaller of 0.25hp and 1.3 for
h>4 ft.
8: Angle of plane of roof from horizontal, in degrees.
m : Angle of plane of panel to roof, in degrees.
Figure 3.5 Figure 29.9-1 from SEAOC Document
Part 2 of SEAOC Document Wind Loads on Low Profile Solar Photovoltaic
Systems on Flat Roofs published in Draft format on March 28, 2012 (SEAOC
2012, used with permission from SEAOC).
18


CHAPTER IV
PREVIOUS STUDIES
Introduction
The construction and installation of the full scale faux solar panel experiment was
first proposed in the Wind load on solar panel experiment by Erin Dowds,
Jennifer Harris, and Frederick Rutz in 2012. These panels were installed at the
University of Colorado Denver Campus on the top of the Physical Education
Building. While there were several other locations investigated, this was
determined to be the most suitable location. The considerations that went into
this were availability of access to the roof, the surrounding structures and
vegetation, shape of the building being similar to those used in wind tunnel tests,
and the features on the roof were flat with a parapet wall and few obstructions in
the area of the panels.
This research was proposed due to the lack of available guidelines for engineers
when designing the installation of solar panels to flat roofs. The goal was to help
fill the gap in the literature by collecting data from the panels installed to compare
to verify the wind tunnel test (Dowds et. al. 2012). The solar panel experiment
proposed can be seen in the figure below.
19


Figure 4.1 Cross section of solar panel experiment.
(Andolsek, 2013, used with permission from Erin Andlosek).
The original proposal for this experiment was to use three panels, all at different
heights. The first panel was intended to be at a height so that the entire panel is
within the recirculation area, Figure 4.2.b. The second panel height was selected
so the panel would be completely in the shear layer, Figure 4.2.c. And the third
panel was constructed to a height so that the panel would be located in both the
shear layer and the recirculation area, Figure 4.2.d.
20


Figure 4.2 Proposed panels with relation to the shear layer.
(Dowds et. al 2012, used with permission from Erin Andlosek).
The construction of the panels is shown in Figure 4.3. The frame was proposed
with pinned connections. Strain gauges were connected to the tension (T) ties
that were on both sides of the panel. The strain in the tension ties will be
measured and recorded to determine the resultant wind force on the panel in the
x-direction.
21


Figure 4.3 Diagram for the panel construction.
(Andolsek 2013, used with permission from Erin Andolsek).
The resultant wind force (FR) is found with the following equation:
Fr = resultant force on the panel
T = tension in tie
p = panel angle
= tension tie angel
(4)
Two locations were studied at the University of Colorado Denver Campus,
Physical Education Building: the first was four feet away from the outside edge of
the parapet, and the second 80 feet from the roof edge. It was anticipated that
22


the location closer to the edge of the roof would experience higher forces than
the location further from the edge of the building. See Figure 4.4 for the two
locations.
Figure 4.4 Aerial View of Events Center Building and Surroundings.
(Andolsek 2013, used with permission from Erin Andolsek).
23


Study of Wind Loads Applied to Full-Scale Rooftop Solar Panels by
Jennifer Harris
The results of the first study were presented by Jennifer Harris in the spring of
2013. The anticipation of this study was that the peak force coefficients would be
greater than the coefficients used in the ASCE (Harris 2013). Two panels were
constructed for this research. Shop drawings were prepared for the construction
of both of the panels. The Electronics Calibration and Repair Lab at the
University of Colorado Denver fabricated the needed members for the panels.
The rest of the materials needed for the assembly of the panels were purchased
from local hardware stores, or were ordered.
Both of the panels consisted of 3/8 inch plywood. Holes were drilled through the
plywood surface on each corner for the legs. The legs were connected by %
diameter bolts in 1x1x1/8 steel angles. The legs had % inch eye bolts that were
inserted through the holes that were drilled in the fabrication lab. The tension
ties were 7/16 inch diameter threaded steel rod that was connected to a strain
transducer through two % inch diameter holes. Nuts were installed on both sides
of the strain transducer in order to leave room for adjustment once assembled. A
coupler nut was used to attach the tension ties to the panel legs. A 7/16 inch
diameter coupler nut was screed onto the treaded rod and was attached to the %
inch eye bolt on both the top and bottom of the panel legs. A 5/16 inch diameter
hole was drilled onto the coupler to connect to the eye bolt. The connection was
24


designed to be a pin-end connection so that the all of the horizontal components
were directed to the tension ties and would be recorded.
Strain gages were adhered to the inside of 2 inch wide, 3 inch diameter steel
rings. For wind blowing in the opposite direction, a slacked cable brace was
installed in the opposite direction of the tension ties.
The legs were connected to the base of the panels with the same connection as
used for the connection to the panels. The bolts were connected to wood 2x6s
that were laid on top of the pavers. See Figures 4.5 through 4.7 for details on the
assemblage of the panels described above.
Figure 4.5 Panel Frame Connection Details.at the Diagonal Tension Ties.
(Harris 2013, used with permission from Jennifer Harris).
25


Figure 4.6 Panel Frame Connection Details in the Short Dimension.
(Harris 2013, used with permission from Jennifer Harris).
Figure 4.7 Close-Up View of Short Direction Panel Frame Connection
Details.
(Harris 2013, used with permission from Jennifer Harris).
26


The two panels constructed for this are shown in Figure 4.8 below. One panel is
at the height of 2 6 above the roof and the other at 5 1.25 above the roof.
Bothe panels were at a 30 degree angle. The tension ties for the shorter panel
were at an angel of 6 degrees and the tension ties for the taller panel were at an
angel of 45 degrees.
Figure 4.8 First Panel Placement.
(Harris 2013, used with permission from Jennifer Harris).
The panels were both assembled and placed on the roof of the Physical
Education Building at the University of Colorado Denver Campus. They were
placed on top of 10 x 10 pavers in order to protect the roof of the building. Sand
27


bags were used to counter the uplift forces instead of connecting the panels to
the pavers or the roof.
Three RM Young 3101 Wind Sentry Anemometers were used to measure the
wind speed. The heights of the anemometers were such that one was at the
assumed shear layer, one above and one below. A RM Young 3301 Wind
Sentry Vane was used to record the wind direction. It was installed above the
three anemometers. To record the temperature, a Campbell Scientific A3537
Type T thermocouple wire was used. These temperatures were all confirmed
with the reported temperatures form the National Climate Data Center records.
A data logger, Campbell Scientific CR5000, was used to record the desired
strains, temperatures, wind speed, and wind directions. This was connected to
an external battery that was powered by a solar panel. A PC card was used to
store the data, which was downloaded when needed. The program for the data
logger was set to collect recordings at 1 second intervals. An SDM-INT8
Campbell Scientific interval timer was used along with the CR5000 data logger to
record the wind speeds form the anemometers. The Campbell Scientific data
logger used RTDAQ 1.1 software for all of the data collected. This can only run
on Microsoft Windows XP, Windows Vista, and Windows 7 operating systems.
To create the program, Short Cut and CRBasic Editor (Campbell Scientific 2001)
were used. The program was written to record the following measurements at 1
28


second intervals: stains from the strain gauges, wind speeds form the three
anemometers, wind direction, temperatures, and the wind direction.
On each of the strain transducers, a 350Q strain gage was attached by soldering
lead wires to the strain gauges. The strain transducers were protected by
silicone and tape. The strain transducers were all calibrated with an MTS
machine that was available at the University of Colorado Denvers Structures
Lab. A section of the treaded rod was used and was representative of how it
would be installed in the field. The calibrated load versus strain curves were
found for each of the strain gauges. In the field, these gauges were all covered
with foil insulation to reduce the impacts that sunlight and temperature would
have on the recordings.
The data recorded for this study was filtered for times when the wind was in the
direction of within 10 degrees of the axis perpendicular to the panel (170 degrees
to 190 degrees). Cf values were only determined for these times. This study
also looked at two other wind directions in order to confirm that the wind in the
direction of the face of the panel was governing. Both 20 and 45 degrees, with a
tolerance of 10 degrees to either direction were also looked at.
The results of the study were presented for the panel that was located within the
shear layer, as the readings from the other panel were inconclusive (Harris
2013). The data was averaged over three-second rolling averages. For the case
29


of wind perpendicular to the panel, and using 3 second averages, the majority of
the Cf values were between 0.6 and 2.2 (Harris 2013). There were often Cf
values at around 10 and a few up to 20. It was assumed that this was due corner
vortices (Holmes 2007). For the other two wind direction cases analyzed, the Cf
values ranged from 3.6 to 7.2 for 20 degrees form perpendicular to the axis of the
panel and for 45 degrees, they ranged from 3.5 to 6.9 (Harris 2013). This
confirms the assumption that the maximum forces occur when the wind is
perpendicular to the panel.
The results of this study were presented at the 2013- 12th Americas Conference
on Wind Engineering in Seattle, Washington (Harris et al. 2013).
Study of Full Scale Rooftop Solar Panels Subject to Wind Loads by Erin
Andolsek
The second study was done in the fall of 2013 by Erin Andolsek. This used the
same two solar panels; however, they were at a different location on the roof of
the Physical Education Building at the University of Colorado Denver. The goal
of this research was to provide baseline data to compare to wind tunnel studies
(Andolsek 2013). The coefficient obtained from this research, Cf, is compared to
Cp used in ASCE 7.
The location of the panels for this study was based on where the shear layer is
expected to reattach to the roof surface (Andolsek 2013). At this location, the
30


wind should be streamlined. With the roof height of 38 feet, the panels were
located about two times the building (80 feet) from the edge of the roof. It was
expected that the wind speeds here would be less than what was recorded in the
first study at the edge of the roof. Figure 4.9 shows the setup for the Andolsek
study.
Figure 4.9 Second Panel Placement.
(Andolsek 2013, used with permission from Erin Andolsek).
At this location, all three of the anemometers should be in the streamlined region,
so it was anticipated that the velocity recordings would be close for all three
elevations. This was confirmed by using a 20 foot section of pipe with orange
construction tape attached at equal spaces, 5 feet apart. On a windy day, this
was brought to the roof and held up at different distances from the edge of the
31


roof. As the pole got further form the edge of the roof, more of the tapes reacted
to the wind in a less turbulent manner. At the location of the panels, all of the
tapes had a similar reaction to the wind, thus confirming that the shear layer had
reattached to the roof surface (Andolsek 2013). See figure 4.10 for the tapes at
the location of this study.
Figure 4.10 Construction tape Experiment 80 feet from Roof Edge.
(Andolsek 2013, used with permission from Erin Andolsek).
For this study, the recording frequency was increased to one tenth of a second.
The data collected was filtered to only look at times when the wind was in the
direction of interest, perpendicular to the panels. This was set at 180 degrees,
and a tolerance of 10 degrees was also considered. After the wind direction was
filtered out, times when the wind speed was less than 17 mph was also
32


eliminated. Cf values were only presented if they were calculated from data with
the wind direction of interest as well as wind speeds greater than 17 mph.
The data collected showed that there was good correlation between the wind
speeds recorded at the three anemometer elevations. Figure 4.11 from this
research shows the coloration of the data.
Wind Velocity Comparison
100 200 300 400 500 600 700 800 900 1000 1100 1200
Time (deciseconds)
Figure 4.11 Wind Velocities from the three anemometers.
(Andolsek 2013, used with permission from Erin Andolsek).
33


The data collected for this study was also averaged over three-second rolling
averages. It was found that there was a clear response in the strain
measurements when the wind velocity changed. The Cp values recorded for
Panel B were on average greater than the values for Panel A. This was what
was anticipated due to the height of Panel B being greater than Panel A.
The presented Cp values reported ranged from 0.1 to 16.3, whit the majority
ranging from 0.1 to 5 (Andolsek 2013). It was found that the vast majority of the
reported Cp values are within reasonable proximity to Cp values presented in
ASCE 7-10 (Andolsek 2013). The Cf values reported form this study were
smaller than when the panels were located at the edge of the roof, as
anticipated.
34


CHAPTER V
PANEL DESIGN, LOCATION AND EQUIPMENT
Location
This study was performed at the University of Colorado Denver Campus. The
roof of the Physical Education Building was selected because it is similar in
shape to the structures used in wind tunnel studies. This building has a flat roof
with a parapet wall and no obstructions. There is also a field in front of the
building, in the direction of the prevailing wind, northwest. See Figure 5.1 for the
site locating.
Figure 5.1 Physical Education Building.
35


The elevation at this location is approximately 5,248 feet above sea level. The
location is also just outside a special wind region, see figure 5.2.
Figure 5.2. Wind Region.
(Andolsek 2013, used with permission from Erin Andolsek).
There is an exposure category of B and a design wind speed of 115 mph for Risk
Category II Buildings, according to the ASCE7.
Both panels were located at about two times the height of the building away for
the edge of the roof, about 80 feet. At this location, it is expected that the shear
layer will have reattached to the roof surface.
36


Panels
Both panels were constructed as a part of previous studies, however, there were
alterations made to them so that the height of both panels were the same and at
a typical height of solar panels installed on flat roofs. David Banks, with CPP
Wind Engineering and Air Quality Consultants in Fort Collins provided that the
lower edge of the panels are typically 3 to 10 inches off the roof. The legs of the
taller panel were removed and brought to the Electronics Calibration and Repair
Lab at the University of Colorado Denver to be shortened to a length of 2 feet, 8
5/8 inch and 11 7/8 inches. See Figure 5.3 below for details.
37


N*0
Figure 5.3. Panel Legs.
New legs were purchased for the shorter panel and were also brought to the Lab
to be fabricated affording to the provided shop drawings above. The new legs
were then brought up to the panels and assembled for the desired height. All of
the connections remained the same as well as the 30 degree angles of the
panels. Also, both of the tension ties needed to be modified or replaced for both
panels. Figure 5.4 is the detail for both panels.


(E) PAPAPET
2x6
PAVERS
(E) ROOF
2i
Figure 5.4. Panel Detail Section. (Not to scale)
2h = approximately 80 feet
The rest of the equipment used for the previous two studies stayed at the
location and were used for this study as well. The anemometers, Campbell
Scientific data logger, and the solar panel all remained in the same locations.
The recording frequency for this study was also increased form the ones used for
the previous studies. A frequency of 30 Hz was recorded, resulting in very large
amounts of data. The program for the data logger needed to be updated for this
frequency, the program can be found in Appendix A. The layout for this study is
shown in figure 5.5.
39


Figure 5.5. Layout. (Not to scale)
2h = approximately 80 feet
(Andlosek 2013, used with permission from Erin Andolsek).
Because the same strain transducers were used for this study, there was no
need to recalibrate them from the last study. The calibration curves for the strain
transducers used are (Harris 2013):
Strain Transducer A: y = 0.8527x + 299.32
Strain Transducer B: y = 0.8358x 31.747
Strain Transducer C: y = 0.8722x + 291.29
Strain Transducer E: y = 0.8809x 76.335
Strain transducers A and B were installed on Panel A and strain transducers C
and E were installed on Panel B. The transducers were all wrapped in insulated
40


foil in order to reduce any impacts that the sunlight and temperature would have
on the recorded data. The layout of the Transducers is in Figure 5.6 below.
r
n

HB
STRAIN C-STRAIN_2(3)
33? a m-

STRAIN ESTRAIN_2(4)
PANEL B
STRAIN F-STRAIN (3)
FOR TEMPERATURE MEASUREMENTS

n HB
STRAIN A-STRAIN (2)
33? a a

STRAIN B-STRAIN
PANEL A
L
J
Figure 5.6. Strain Transducer Locations.
(Andolsek, 2013, used with permission from Erin Andolsek).
Equipment
The Campbell Scientific CR5000 Datalogger was used to record and store all of
the data collected. This was placed in a metal box behind the panels in order to
protect it from being damaged. This was powered by an external battery that
41


was charged by a Campbell Scientific SP20 Solar Panel. All of the collected data
was stored on a Campbell Scientific CFMC2G Flash card. Due to the high
recording frequency, this needed to be downloaded every 2 to 3 days. The
support software used for the datalogger was Campbell Scientific RTDAQ
Version 1.1. To create the program for this, Short Cur and CRBasic Editor were
used. The program for this study can be found in Appendix A.
Data was collected at a 30 Flz frequency for the time, wind direction, wind speed
at all 3 anemometers elevations, the air temperature, and the strain data. The
data was stored in the flash card until it was downloaded. To download the data,
Trendnet, a driver software is need on the computer you are downloading to. As
a part of the downloading process, all of the data needed to be converted to a
format that was usable by Microsoft Excel. Card Convert, a part of the RTDAQ
program was used for this. Figures 5.7 and 5.8 show the setup for the study.
42


I
t
*
4
43


Figure 5.8. Study Location and Layout.
44


CHAPTER VI
THEORY
Individual solar panel manufactures have conducted numerous wind tunnel
studies to determine the appropriate design and installation for their products.
They usually use pressure tapes to measure the pressure distribution for the
panel. The net pressure on the top and bottom surfaces of the panel can be
used to determine the pressure coefficient, Cp. For this study, the force
coefficient, Cf, was calculated to be compared to the pressure coefficient from
the available wind tunnel study results.
The two panels described earlier were constructed and installed to represent how
solar panels would typically be installed on a flat roof. As the wind hits the panel,
the pressure is assumed to be uniformly distributed on the face of the panel. The
resultant force is what is needed, so because the wind does not actually behave
this way, this assumption does not influence the results.
As the wind creates the upward force on the panel, there is tension in the tie
bars. This is measured with the strain transducers on the tension ties. At the
same time, the wind velocity and direction is recorded as well. All of the data
collected was averaged over a period of 3 seconds to eliminate any influence
that thermal effects would have on the recordings. The barometric pressure was
found from the National Climatic Data Center website.
45


The recorded wind velocity was used to determine the force on the panel, Fvp.
Fvp = force on panel
p = air density (not corrected for altitude)
V = measured wind velocity (averaged over 3 seconds)
A = area of the panel surface
The value Cf, the coefficient of force, is found from equation 5 below.
CF = Fr
Fvel press
Cf = net force coefficient
Fr = resultant force on panel
FVeipress = force on the panel form dynamic pressure form wind


CHAPTER VII
VIBRATION OF PANEL
Introduction
Both previous studies by Jennifer Harris (Harris 2013) and Erin Andolsek
(Andolsek 2013) had recommended that the natural frequency of the panels be
investigated to determine if the vibration of the panel produced any internal
forces that could influence the readings from the strain transducers. As a part of
this study, the natural frequency of the panels was determined.
Panel Natural Frequency
In order to determine the natural frequency of the panels, strain gauges were first
explored. These were placed on the bottom center of the panels; however, they
would not stay on the panel surface. Next, accelerometers we brought up to the
panels to record the vibration of the panel.
For this, three sensors were used. One was attached on the top center of the
panel, and the other two were attached to the front legs to measure the
accelerations parallel and perpendicular to the face of the panel. See Figures
7.1 through 7.3 for the locations of the sensors.
47


Figure 7.1. Accelerometer Test Set Up.
48


Figure 7.2. Accelerometer Test Set Up.
Figure 7.3. Accelerometer Test Set Up.
49


Once all of the sensors were fastened to the panel, they were connected to a
separate data logger to record the movement of the panel and legs. Each sensor
had its own channel that was recorded: Chanel 0 was on the top of the panel,
Chanel 1 is on the right leg facing the panel, and Chanel 2 is on the left leg
(looking in the direction of Figure 7.3).
A few tests were performed to initiate the panel vibrations. Heal Drop tests were
done, as well as both legs of the panel were tapped. First, the heal drop test was
conducted. The recorded acceleration and frequency from this can be seen in
Figures 7.4 7.7 below.
18:26:17.610
18:26:19.000
18:26:20.000
Time (s)
18:26:21.000
18:26:22.609
Figure 7.4. Heal Drop, all three sensors.
Results for Individual sensors are shown in Figures 7.5 7.7.
50


^Waveform Chart
Frequency (Hz)
Figure 7.5. Heal Drop, Sensor 0: on panel.
Natural Frequency =11 Hz.
51


Waveform Chart
i i i i i i i i i i i i i i i i i i i i i i i i i i
12 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (Hz)
Figure 7.6. Heal Drop, Sensor 1: panel leg, perpendicular to panel.
Natural Frequency = 13.5 Hz.
52


Waveform Chart
m -requency Graph
Frequency (Hz)
Figure 7.7. Heal Drop, Sensor 2: panel leg, parallel to panel.
Predominate Mode Natural Frequency = approximately 35 Hz.
These graphs show that the natural frequency of the panel is close to 12 Hz.
Both the legs natural frequencies were around 14 Hz. This test was performed
several times, and similar results were found.
The results of the right leg tap (sensor 1) and the left leg tap (sensor 2) are
shown in Figures 7.8 through 7.15 below.
53


hS Waveform Chart
800m-
600m-
x 400m-
-600m -
-800m -
18:35:01.469
18:35:03,000
18:35:04,000
Time (s)
18:35:05,000
i i
18:35:06,468
|fc] Frequency Graph
£ Channel View
Figure 7.8. Leg Tap 1, all three sensors.
Results for Individual sensors are shown in Figures 7.9 7.11.
54


[^Waveform Chart
L\, Frequency Graph
Figure 7.9. Leg Tap 1, Sensor 0: on panel.
Natural Frequency =11 Hz.
55


\f-z Waveform Chart
800m
600m
^ 400m
Qi
'c 200m
-800m
18:35:01.469
18:35:03.000
18:35:04.000
Time (s)
18:35:05.000
18:35:06.468

Frequency Graph
3m-i
2.5m-
e
cn
2m-
1.5m-
c
Qi
Q
lm-
500u -
i i i i i i i i i i i i i i i i i i i i i i
8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (Hz)
Figure 7.10. Leg Tap 1, Sensor 1: panel leg, perpendicular to panel.
Natural Frequency = 13.5 Hz.
56


m Frequency Graph
12 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency(Hz)
Figure 7.11. Leg Tap 1, Sensor 2: panel leg, parallel to panel.
Predominate Mode Natural Frequency = approximately 39.5 Hz.
57


bS
O
%
>n
m
u
i_i
<
Waveform Chart
Figure 7.12. Leg Tap 2, all three sensors.
Results for Individual sensors are shown in Figures 7.13-7.15.
58


Waveform Chart
Figure 7.13. Leg Tap 2, Sensor 0: on panel.
Natural Frequency =11 Hz.


Figure 7.14. Leg Tap 2, Sensor 1: panel leg, perpendicular to panel.
Natural Frequency = 12 Hz.
60


m .Frequency Graph
lm-
_300u-
1/1
£
3 600u-
(D
-o
| 400u -
cn
E 200u -
0-
1 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Frequency (Hz)
Figure 7.15. Leg Tap 2, Sensor 2: panel leg, parallel to panel.
Predominate Mode Natural Frequency = approximately 39.5 Hz.
This test was also performed several times, and similar results were found.
This test resulted in natural frequencies between about 10 Hz and 40 Hz. These
tests were also performed several times, and similar results were found.
Conclusion
The recorded natural frequencies for the panel were quite a bit greater than 1.0
Hz. While the recording frequency for the solar panel itself was not high enough
to look for this in the recorded data, it can be determined that the vibrations of the
61


panel had no significant influence on recordings from the strain transducers.
This is due to the fact that 1.0 Hz is the max frequency to consider vibration per
ASCE7-10. The range of natural frequencies found for the panels were well
above this value so may be considered a rigid structure.
62


CHAPTER VIII
RESULTS
Introduction
The results from this study are presented below. The Cf values, wind velocities,
and strains are shown in the graphs below for both panels and were recorded at
30 Hz. The recordings were processed using three-second rolling averages.
Results
This study resulted in a very large amount of data, most of which was not useful
for this analysis. Only times when the majority of the wind velocities were greater
than 17 mph and in the direction perpendicular to the panels were used. A
tolerance of 10 degrees to either direction was allowed for the wind direction. To
determine the Cf values, the data was filtered even more. Both the wind velocity
and the strain values had to be increasing or decreasing together during the tine
period of interest. This eliminated any Cf values that were not a result from the
change in strain. This process eliminated the vast majority of the Cf values. The
results can be found in Figures 8.1 8.14 And Tables 8.1 8.14. The strain is
graphed on an arbitrary scale because only the differential strain is of interest.
63


Panel A 4/28/14 3:41PM
3s Strain Intervals & 3s Average Wind Speed
45.00
40.00
35.00
30.00
LL
u
o3
a. 25.00
£_
>
'u
O
£ 20.00
~o
c
§
15.00
10.00
5.00
0.00

i___i L. i I i____,________l
oomtNCTi^Dmor^^^HoomtNcri^Dmor^^^H
oo^o^oooooooor^r^r^^D^D^DLOLnLnLn^i-^i-^i-
HNfO^inwisCOOlOHNrO^inWINCOOl
-----Wind Velocity
Strain A
Strain B
-----CF
Time (1/30 sec)
Figure 8.1. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel A.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values. The upward trend in strain is
attributed to thermal drift. The strain output varies with temperature.
64


Table 8.1. Cf Values from Figure 8.1.
CF Values
0.1 0.2 0.2 0.7
0.1 0.2 0.3 0.7
0.1 0.2 0.3 0.7
0.1 0.2 0.3 0.9
0.1 0.2 0.3 0.9
0.1 0.2 0.3 0.9
0.1 0.2 0.3 1.1
0.1 0.2 0.3 1.1
0.1 0.2 0.4 1.5
0.2 0.2 0.4 3.8
0.2 0.2 0.4 3.9
0.2 0.2 0.5 6.7
0.2 0.2 0.6
65


Panel B 4/28/14 3:41 PM
3s Strain Intervals & 3s Average Wind Speed
50.0
45.0
ioo^-LniDr-'OOcnO'H(N
LnOLnOLnOLnoiD'-HiD cncnoooor-'r-'iDU3LnLn^-^-mmrvirM ^HrMm^-LniDr-'OOcnO'Hrvioo^-LniDr-'OOcncn
Time (1/30 sec)
Figure 8.2. Cp Values, Strain, and Wind Velocity from 4/28/14, Panel B.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
66


Table 8.2. Cf Values from Figure 8.2.
CF Values
0.1 0.2 0.3 0.7
0.1 0.2 0.3 0.7
0.1 0.2 0.4 0.7
0.1 0.2 0.4 0.7
0.1 0.2 0.4 0.8
0.1 0.2 0.4 0.8
0.1 0.2 0.4 0.8
0.1 0.2 0.4 0.8
0.2 0.2 0.4 0.9
0.2 0.2 0.4 1.0
0.2 0.2 0.4 1.0
0.2 0.2 0.4 1.1
0.2 0.2 0.4 1.2
0.2 0.3 0.4 1.3
0.2 0.3 0.4 1.4
0.2 0.3 0.4 1.5
0.2 0.3 0.4 1.5
0.2 0.3 0.5 1.6
0.2 0.3 0.5 2.5
0.2 0.3 0.5 3.5
0.2 0.3 0.5 3.8
0.2 0.3 0.5 4.1
0.2 0.3 0.5 5.0
0.2 0.3 0.5 5.0
0.2 0.3 0.6
0.2 0.3 0.6
0.2 0.3 0.7
0.2 0.3 0.7
67


Panel A 4/27/14 1:34 PM
3s Strain Intervals & 3s Average Wind Speed
40.00
35.00
30.00
J- 25.00
08
Q.
E
£ 20.00
u
o
ai
>
T3
§ 15.00
10.00
5.00
0.00
IaIL. idil-1 LL.lL m. U i
^Hrvjm^-Ln^r^ooa^o^Hrvjm^-Ln^r^ooa^o^HrvJ
'rHrvjro'^-LnuDr^-ooO'rHrvjm'^-LnuDr^-ooa^'rHrvjm
^OOrvl^O^OOPNllvHLnO^fOINHinO'lMOOfMtD
fN^pvWfN^WWH^lOCOHfOlOCOOfOLnCOO
rHrHHHfMfMfMfMrOfOfnm^^-^^Lr)
Time (1/30 sec)
Wind Velocity
Strain A
Strain B
CF
Figure 8.3. Cf Values, Strain, and Wind Velocity from 4/27/14, Panel A.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
68


Table 8.3. Cf Values from Figure 8.3.
CF Values
0.1 0.2 0.4 0.9
0.1 0.2 0.4 1.4
0.1 0.2 0.4 4.2
0.1 0.2 0.4
0.1 0.3 0.5
0.2 0.4 0.5
69


Panel B 4/27/14 1:34 PM
3s Strain Intervals & 3s Average Wind Speed
40.00 -------------------------------------------
35.00
rvi^-r'-~cnrvi^-iDcn*-i^-iDoo*-imiDooomLnooo
Time (1/30 sec)
Figure 8.4. Cf Values, Strain, and Wind Velocity from 4/27/14, Panel B.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values. No data collected for Strain D.
70


Table 8.4. Cf Values from Figure 8.4.
CF Values
0.1 0.2 0.2 1.1
0.1 0.2 0.2 1.2
0.1 0.2 0.2 1.3
0.1 0.2 0.3 1.3
0.1 0.2 0.3 1.4
0.1 0.2 0.4 1.5
0.1 0.2 0.4 1.9
0.1 0.2 0.5 3.8
0.1 0.2 0.5
0.1 0.2 0.5
0.1 0.2 0.5
0.1 0.2 0.6
0.2 0.2 0.6
0.2 0.2 0.7
0.2 0.2 0.8
0.2 0.2 0.8
0.2 0.2 1.0
71


Panel A 4/27/14 10:51PM
3s Strain Intervals & 3s Average Wind Speed
35.00
30.00
25.00
LL
u
o3
S' 20.00
Q.
£
>
*u
O
0)
> 15.00
~o
c
5
10.00
5.00
0.00
_i__I_______I ____.__ I
(NinrNO(Ninr>'0(Ninis'0(Nior>'0(Ninis'0(Ninis'
r^^iHCTi^omooomrMCTir^^tHoo^omor^mrviCTi^o
HfriiniflooofNfnirrvooofN^irrsaiHfM^^rso^
HHrlrlHHfMINfMMfMINfOfOmmfOfO
Time (1/30 sec)
Wind Velocity
Strain A
Strain B
CF
Figure 8.5. Cf Values, Strain, and Wind Velocity from 4/27/14, Panel A.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
72


Table 8.5. Cp Values from Figure 8.5.
CF Values
0.1 0.1 0.2
0.1 0.1 0.2
0.1 0.2 0.3
0.1 0.2 0.3
0.1 0.2 0.3
0.1 0.2 0.4
0.1 0.2 0.5
0.1 0.2 0.5
3.4
73


Panel B 4/27/14 10:51PM
3s Strain Intervals & 3s Average Wind Speed
40.00
ooiD^-rvi oor-'iDLn^-<~vi*-iocr>oor'-'Ln^-mrvi*Hcnoor'-'iDLn
*hti*i*Iti(N(N(N(N(Nrommcomm
Time (1/30 sec)
Figure 8.6. Cp Values, Strain, and Wind Velocity from 4/27/14, Panel B.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
74


Table 8.6. Cf Values from Figure 8.6.
CF Values
0.1 0.2 0.2 0.5
0.1 0.2 0.2 0.6
0.1 0.2 0.2 0.6
0.1 0.2 0.2 0.7
0.1 0.2 0.2 0.7
0.1 0.2 0.3 0.7
0.1 0.2 0.3 1.0
0.1 0.2 0.4 1.6
0.1 0.2 0.4 1.9
0.1 0.2 0.5 4.0
75


Panel A 4/28/14 7:24 PM
3s Strain Intervals & 3s Average Wind Speed
T-icrihvLomT-icrihvLoroT-icrihvLoroT-icrir^LnroT-icrihv
inH|\fr)ff)^O^lNOOfOO^inHNfNOO^O0Hrs
rvjLnr^orvjLnoooroLnooomuDoO'rHmuDCJ^rH'^-uD
rlHrlrlfMfMrslfMfOfOfOfO^^^^ininm
Time (1/30 sec)
Wind Velocity
Strain A
Strain B
CF
Figure 8.7. Cp Values, Strain, and Wind Velocity from 4/28/14, Panel A.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
76


Table 8.7. Cf Values from Figure 8.7.
CF Values
0.1 0.1 0.3 0.6
0.1 0.2 0.3 1.1
0.1 0.2 0.4 1.6
0.1 0.2 0.4 2.2
0.1 0.2 0.5 2.4
0.2
77


Panel B 4/28/14 7:24 PM
3s Strain Intervals & 3s Average Wind Speed
40.0
35.0
30.0
U. 25.0
u
o3
S'
Q.
E
£ 20.0
o
ai
>
a
^ 15.0
10.0
5.0

MAIL.
W*d\Uo'WS,AjUJ*^sKA*
o.o
r-'^- rMLnooomiDoo Time (1/30 sec)
-----CF
-----Wind Velocity
Strain C
-----Strain E
Figure 8.8. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel B.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
78


Table 8.8. Cp Values from Figure 8.8.
CF Values
0.1 0.2 0.3 0.6
0.1 0.2 0.3 0.7
0.1 0.2 0.3 0.9
0.1 0.2 0.4 1.4
0.2 0.2 0.4 1.8
0.2 0.2 0.4 2.3
0.2 0.2 0.5 2.6
0.2 0.2 0.5 10.0
0.2
79


Panel A 4/27/14 11:02 AM
3s Strain Intervals & 3s Average Wind Speed
40.00
35.00
30.00
u. 25.00
u
o3
S'
Q.
E
£ 20.00
U
O
0)
>
~o
^ 15.00
10.00
5.00
0.00
h co ld in oi id ro
m IN H ^ 00 (N
ID (N CTl LD H CO
(Y) N O ^ CO H
rl rl rl (N
O 'st1 H CO LD (N
0 0 fO N O ^ 00
o m o LO O (N (N fO fO ^ ^ ^
miOfOOI^^HOO
t-HLOCTirO^DO'vfhv
O^OOHin^MW
Time (1/30 sec)
Wind Velocity
Strain A
Strain B
CF
Figure 8.9. Cp Values, Strain, and Wind Velocity from 4/27/14, Panel A.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
80


Table 8.9. Cf Values from Figure 8.9.
CF Values
0.1 0.2 0.2 0.6
0.1 0.2 0.2 0.7
0.1 0.2 0.4 2.6
0.1 0.2 0.4 2.6
0.1 0.2 0.4 4.1
0.1 0.2 0.5
0.1 0.2 0.6
81


Panel B 4/27/14 11:02 AM
3s Strain Intervals & 3s Average Wind Speed
45.00
40.00
35.00
30.00
Ll_
U
08
£ 25.00
Q.
£
>
*U
O
> 20.00
~o
c
§
15.00
10.00
5.00
0.00
Hoo^fNoiiDfoorN^HoomfMo^iflmoi^^Hoo
fniN'H^OOfM0ClMNO^OOHinaifr)0O^ls'
uDrvjcj^Ln'rHoO'^-or^mouDrvja^Ln'rHoO'vi-'rHr^m
mr^-O'^oO'rHLna^rvjuDOmr^-O'^oO'rHLna^rvjuD
HHH(N(NfMfnm^i,^,si,inininiDiDmrs'rs'
Time (1/30 sec)
-----CF
-----Wind Velocity
Strain C
-----Strain E
Figure 8.10. Cp Values, Strain, and Wind Velocity from 4/27/14, Panel B.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
82


Table 8.10. Cp Values from Figure 8.10.
CF Values
0.1 0.2 0.3 0.4
0.1 0.2 0.3 0.4
0.1 0.2 0.3 0.5
0.1 0.2 0.3 0.5
0.1 0.2 0.3 0.6
0.1 0.2 0.3 0.7
0.2 0.2 0.3 0.8
0.2 0.2 0.3 1.1
0.2 0.2 0.3 1.1
0.2 0.2 0.3 1.1
0.2 0.2 0.4 1.4
0.2 0.2 0.4 2.1
0.2 0.2 0.4 3.2
0.2 0.2 0.4 3.9
0.2 0.2 0.4 4.5
0.2 0.3
0.2 0.3
83


Panel A 4/28/14 4:39 PM
3s Strain Intervals & 3s Average Wind Speed
40.00
35.00
30.00
^ 25.00
u
08
S'
Q.
£_
20.00
O
ai
>
T3
15.00
10.00
5.00
0.00

Wind Velocity
Strain A
Strain B
*CF
OOlD^fNOOOlO^(NOCO0^fNOCO0^fN
IflfflOh'^ON^HOO^HOOinfMOOinfMO^
HfNfNrO^^miOlDP^eOCOOlOOHNN
O 00
UO m
Time (sec)
Figure 8.11. Cf Values, Strain, and Wind Velocity from 4/28/14, Panel A.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
84


Table 8.11. Cp Values from Figure 8.11.
CF Values
0.1 0.2 0.2 0.6
0.1 0.2 0.2 0.7
0.1 0.2 0.3 0.9
0.1 0.2 0.3 0.9
0.1 0.2 0.3 1.2
0.1 0.2 0.4 2.3
0.1 0.2 0.4 3.2
0.2 0.2 0.5
85


Panel B 4/28/14 4:39 PM
3s Strain Intervals & 3s Average Wind Speed
45.0
40.0
35.0
30.0


u
o3
I 25.0
E
LnOLDOLnOLnOLnOLnOLnOLnOLnOLnOLDO
iDmcniDrvicnLnrviooLn^Hoo^-^Hr-'^-or-'moiDm
Time (sec)
CF
Wind Velocity
Strain C
Strain E
Figure 8.12. CF Values, Strain, and Wind Velocity from 4/28/14, Panel B.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
86


Table 8.12. Cf Values from Figure 8.12.
CF Values
0.1 0.2 0.2 0.5
0.1 0.2 0.3 0.5
0.1 0.2 0.3 0.6
0.1 0.2 0.3 1.0
0.1 0.2 0.3 1.4
0.1 0.2 0.4 2.6
0.1 0.2 0.5 2.7
0.2 0.2 0.5 4.3
0.2
87


Panel A 4/28/14 12:00 AM
3s Strain Intervals & 3s Average Wind Speed
35.00
30.00

25.00
u
03
S' 20.00
Q.
E
u
O
0)
> 15.00
10.00
5.00
Wind Velocity
Strain A
Strain B
CF
0.00 -----------------1 I--------1
^HooLnrvicniDmor-'^-rHooLnrviCTitDmor-'^-rHoo
ooiD^-rvioooiD^-rvioooiD^-rvi^Hcnr-'LnmrHcn
HrMni^^i/liDSiOWOiOHrviNro^i/iiiiiii
Time (1/30 sec)
Figure 8.13. CF Values, Strain, and Wind Velocity from 4/28/14, Panel A.
The data is form a selected portion of the study using 3 second rolling averages.
An arbitrary scale is used for the stain values.
88


Full Text

PAGE 1

STUDY OF WIND LOADS APPLIED TO SOLAR PANELS ON FLAT ROOFTOPS by RACHELLE KAY PLAS B.S., University of Colorado Denver 2010 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Civil Engineering 201 5

PAGE 2

ii This thesis for the Master of Science degree by Rachelle Kay Plas has been approved for the Civil Engineering Program by Frederick R. Rutz Chair Peter Marxhausen Bruce Janson November 19 2015

PAGE 3

iii Rachelle Kay Plas ( M.S., Civil Engineering) Study of Wind Loads Applied to Solar Panels on Flat Rooftops Thesis directed by Assistant Professor Frederick R. Rutz ABSTRACT The goal of the research presented in this thesis is to determine the design wind load on rooftop mounted solar panels by collect ing data to b e compared with previous research, wind tunnel tests, and the values in current standards. The data collected was measured form structurally suspended, faux solar panels that were installed at the University of Colorado Denver Campus on a flat roof. Strain transducers were used to measure the force on the panels due to wind. The key value in this research is the Coefficient of Force, C F which was der ive d from the collected data. Also, t he natural frequency of the faux solar panels was determined to investigate if the vibration of the panel impacted the results of the research. The key findings of this research are average C F values of 1.2 for Panel A and 1.8 for Panel B. It was also found that the vibration of the panels did not have an influence on the strain transducer recordings nor the C F values. The form and content of this abstract are approved. I recommend its publication. Approved: Frede rick R. Rutz

PAGE 4

iv DEDICATION I dedicate this work to my parents Rick and RaNae Urso and my husband, Seth Plas, thank you for your support, encouragement and patience throughout this journey I am so very lucky to have such supportive and encouraging people in my life. I would not be where I am today without you.

PAGE 5

v ACKNOWLEDGMENTS Th ere are so many individuals who have helped me throughout this process. I would like to give a tremendous thank you to my thesis advisor and past professor Dr. Frederick R. Rutz Your passion for engineering and education has inspired me countless times throughout my theses and coursework with you. Even in my undergraduate classes with you, your thirst for knowledge and excitement in tea ching others has always stood out. Erin Andol sek and Jennifer Harris, your knowledge and assistance with this have been so helpful. You have really set the bar high in your work and have made such strides in your research in this field. Your creativen ess and unique solutions have made this research possible. Tom Thuis of the Electronic Calibration and Repair Lab thanks for manufacturing the new legs for the panels. Your assistance is very much appreciated. I would also like to thank the Auraria High er Education Campus Facilities Department for allowing us to utilize the roof of the Events Center Building. This has been a great location for our research. Thank you Rick Urso for helping me assemble the new panels in extremely cold weather. I could not have done that without you.

PAGE 6

vi And, thank you Fred, Jennifer & Jess e for your help in disassembling and disposing of the panels after all of the data was collected It would have been a daunting task without your help. F inally, thank you Jared for your help with analyzing the data recorded from the accelerometers. Your input was very important to our results

PAGE 7

vii TABLE OF CONTENTS CHAPTER I. OVERVIEW ................................ ................................ ................................ 1 Introduction ................................ ................................ ................................ 1 Goal ................................ ................................ ................................ ........... 2 Procedure ................................ ................................ ................................ .. 3 II. SOLAR ENERGY ................................ ................................ ...................... 4 Introduction ................................ ................................ ................................ 4 History ................................ ................................ ................................ ........ 4 Solar Panels Today ................................ ................................ ................... 5 III. WIND BEHAVIOR ................................ ................................ .................... 7 Introduction ................................ ................................ ................................ 7 Wind Tunnel Test ing ................................ ................................ .................. 7 Current Codes and Standards ................................ ................................ ... 9 I V. PREVIOUS STUDIES ................................ ................................ ............ 19 Introduction ................................ ................................ .............................. 19 Study of Wind Loads Applied to Full Scale Rooftop Solar Panels by Jennifer Harris ................................ ................................ ......................... 24 Study of Full Scale Rooftop Solar Panels Subject to Wind Loads by Erin Andolsek ................................ ................................ ................................ .. 30 V. PANEL DESIGN, LOCATION AND EQUIPMENT ................................ .. 35 Location ................................ ................................ ................................ ... 35 Panels ................................ ................................ ................................ ...... 37

PAGE 8

viii Equipment ................................ ................................ ................................ 41 VI. THEORY ................................ ................................ ................................ 45 VII. VIBRATION OF PANEL ................................ ................................ ........ 47 Introduction ................................ ................................ .............................. 47 Panel Natural Frequency ................................ ................................ ......... 47 Conclusion ................................ ................................ ............................... 61 VIII. RESULTS ................................ ................................ ............................ 63 Introduction ................................ ................................ .............................. 63 Results ................................ ................................ ................................ ..... 63 Discussion ................................ ................................ ............................... 91 IX. CONCLUSION ................................ ................................ ....................... 94 Conclusion ................................ ................................ ............................... 94 Possible Sources of Error ................................ ................................ ........ 95 Re commendations for Further Research ................................ ................. 96 REFERENCES ................................ ................................ ............................... 97 APPENDIX A ................................ ................................ ................................ .. 99 Datalogger Program ................................ ................................ .................... 99

PAGE 9

ix LIST OF FIGURES Figure 3.1 Wind Flow against a Bluff Body. ................................ ................................ 11 3.2 External Pressure Coefficients. ................................ ................................ .. 13 3.3 Net Pressure Coefficients for Monoslope Free Roofs. ................................ 15 3.5 Figure 29.9 1 from S EAOC Document ................................ ....................... 18 4.1 Cross section of solar panel experiment. ................................ ..................... 20 4.2 Proposed panels with relation to the shear layer. ................................ ......... 21 4.3 Diagram for the panel construction. ................................ ............................. 22 4.5 Panel Frame Connection Details.at the Diagonal Tension Ties. ................. 25 4.8 First Panel Placement. ................................ ................................ ................ 27 4.9 Second Panel Placement. ................................ ................................ ........... 31 4.10 Construction tape Experiment 80 feet from Roof Edge. ........................... 32 4.11 Wind Velocities from the three anemometers. ................................ ............ 33 5.1 Physical Education Building. ................................ ................................ ......... 35 5.2. Wind Region. ................................ ................................ ............................... 36 5.3. Panel Legs. ................................ ................................ ................................ .. 38 5.4. Panel Detail Section. (Not to scale) ................................ ............................. 39 5.5. Layout. (Not to scale) ................................ ................................ .................. 40 5.6. Strain Transducer Locations. ................................ ................................ ....... 41 5.7. Study Location and Layout. ................................ ................................ ......... 43 5.8. Study Location and Layout. ................................ ................................ ......... 44 7.1. Accelerometer Test Set Up. ................................ ................................ ......... 48 7.2. Accelerometer Test Set Up. ................................ ................................ ......... 49

PAGE 10

x 7.3. Accelerometer Test Set Up. ................................ ................................ ......... 49 7.4. Heal Drop, all three sensors. ................................ ................................ ....... 50 7.5. Heal Drop, Sensor 0: on panel. ................................ ................................ ... 51 7.6. Heal Drop, Sensor 1: panel leg, perpendicular to panel. ............................. 52 7.7. Heal Drop, Sensor 2: panel leg, parallel to panel. ................................ ....... 53 7.8. Leg Tap 1, all three sensors. ................................ ................................ ....... 54 7.9. Leg Tap 1, Sensor 0: on panel. ................................ ................................ ... 55 7.10. Leg Tap 1, Sensor 1: panel leg, perpendicular to panel. ........................... 56 7.11. Leg Tap 1, Sensor 2: panel leg, parallel to panel. ................................ ..... 57 7.12. Leg Tap 2, all three sensors. ................................ ................................ ..... 58 7.13. Leg Tap 2, Sensor 0: on panel. ................................ ................................ 59 7.14. Leg Tap 2, Sensor 1: panel leg, perpendicular to panel. ........................... 60 8.1. C F Values, Strain, and Wind Velocity from 4/28/14, Panel A. ..................... 64 8.2 C F Values, Strain, and Wind Velocity from 4/28/14, Panel B. ..................... 66 8.3. C F Values, Strain, and Wind Velocity from 4/27/14, Panel A. ..................... 68 8.4. C F Values, Strain, and Wind Velocity from 4/27/14, Panel B. ..................... 70 8.5. C F Values, Strain, and Wind Velocity from 4/27/14, Panel A. ..................... 72 8.6. C F Values, Strain, and Wind Velocity from 4/27/14, Panel B. ..................... 74 8.7. C F Values, Strain, and Wind Velocity from 4/28/14, Panel A. ..................... 76 8.8. C F Values, Strain, and Wind Velocity from 4/28/14, Panel B. ..................... 78 8.9. C F Values, Strain, and W ind Velocity from 4/27/14, Panel A. ..................... 80 8.10. C F Values, Strain, and Wind Velocity from 4/27/14, Panel B. ................... 82 8.11. C F Values, Strain, and Wind Velocity from 4/28/14, Panel A. ................... 84 8.12. C F Values, Strain, and Wind Velocity from 4/28/14, Panel B. ................... 86

PAGE 11

xi 8.13. C F Values, Strain, and Wind Velocity from 4/28/14, Panel A. .................. 88 8.14. C F Values, Strain, and Wind Velocity from 4/28/14, Panel B. .................. 90

PAGE 12

xii LIST OF TABLES Table 8.1. C F Values from Figure 8.1. ................................ ................................ ......... 65 8.2. C F Values from Figure 8.2. ................................ ................................ ......... 67 8.3. C F Values from Figure 8.3. ................................ ................................ ......... 69 8.4. C F Values from Figure 8.4. ................................ ................................ ......... 71 8.5. C F Values from Figure 8.5. ................................ ................................ ......... 73 8.6. C F Values from Figure 8.6. ................................ ................................ ......... 75 8.7. C F Values from Figure 8.7. ................................ ................................ ......... 77 8.8. C F Values from Figure 8.8. ................................ ................................ ......... 79 8.9. C F Values from Figure 8.9. ................................ ................................ ......... 81 8.10. C F Values from Figure 8.10. ................................ ................................ ..... 83 8.11. C F Values from Figure 8.11. ................................ ................................ ..... 85 8.12. C F Values from Figure 8.12. ................................ ................................ ..... 87 8.13. C F Values from Figure 8.13. ................................ ................................ ..... 89 8.14. C F Values from Figure 8.14. ................................ ................................ ..... 91 8.15. Average Peak C F from Figures 8.1 8.14. ................................ ................ 92

PAGE 13

1 CHAPTER I OVERVIEW Introduction Solar panels are becoming more and more popular as society moves towards the use of environmental friendly sources of energy. They are now being installed commonly in residential an d commercial areas. Because they are typically positioned in a way to maximize their exposure to the sunlight, rooftops are an ideal location. Solar panels are only expected to continue to grow in popularity as technology progresses and they become more efficient and affordable. Design professionals typically use engineering standards, such as ASCE 7 (ASCE 7 201 0 ), which does not say what type of wind loads should be applied to roof mounted solar panels. This leaves the design professional s to use the ir judgment to determine the wind loads for these structures. While there is some gu idance available it is based on small scale wind tunnel testing. Most of the research that has been done is proprietary, and because of the costs of these tests, there is not a lot of literature available. While t here is the need for full scale validation of the results from these wind tunnel tests, t here have been very few studies that have looked at t his When wind tunnels were first introduced, the original wind tunnel studies were validated with significant amounts of full scale research (Cochran 2010). It

PAGE 14

2 makes since that the next step in determining the design wind loads for solar panels on fla t rooftops is to verify the results from the wind tunnel tests with full scale models. Over the past few years, there have been two full scale faux solar panel experiments at the University of Colorado Denver Campus. These have been used to collect dat a to compare to the results of the wind tunnel studies. These panels were fabricated to record the wind pressure, velocity, and direction as well as the wind force, derived from strain measurements in the tension ties of the two panels. These are the sam e panels that are used in this research; however, the height has been adjusted to better represent the typical installation of solar panels on flat rooftops. Goal The goal of this research is to use the data recorded from the faux solar panels to derive a coefficient that can be compared to the current standards and existing wind tunnel research. Although efforts to determine wind loads on solar panels have been ongoing for some time few full scale experiments have been reported (Harris 2013 ) This res earch will also determine the natural frequency of the solar panels used in this research to determine if the vibrations of the panel could have influenced the recorded strains.

PAGE 15

3 Procedure Real time data was collected from the two faux solar panels at the University of Colorado Denver Campus. Wind velocity and direction was recorded along with the stain for two faux solar panels. The wind velocity and baro metric pressu re were used to find the force on the panels due to the wind. The strain measurements obtained were used to find the net force acting on the faux solar panel s The Coefficient of Force, C F was then determined from the ratio o f these two values This can be compared to the Coefficient of Pressure, C P which is used in ASCE 7 (ASCE 7 2010) The comparison of these two coefficients can show how the measured pressure, C F and the pressure form the wind tunnels that are used in the current standards, C P relate to each other.

PAGE 16

4 CHAPTER II SOLAR ENERGY Introduction of Energy 2012 ). There is no doubt that solar energy will continue to grow in popularity and be seen in a variety of applications in the future Today, you can buy solar powered products anywhere form rooftop panels, to yard and Christmas lights. There are even several solar do it yourself kits available. Solar technology has been around for a long time, and over the past several years, there have been great advances in this technology, which is only expected to continue. History S ilicon photovoltaic (PV) cells were developed ccording to the U.S. Department of Energy, the earliest documented uses of concentrating th Century B.C, when magnifying glasses were used to start fire ( History of Solar 2012 ). During the next few thousand years there have been many different ways that sunlight was used to create fire and to warm rooms. Horace de Suassure used the observation that room, a carriage, or any other place is hotter when the ray s of the sun pass through glass build

PAGE 17

5 a hot box (Butti, 1980). He started with assembling a box with layers of gla ss, similar to a greenhouse. He used this box to cook fruits. He then constructed a similar box, with insulated wood sides and a layered glass top. This became known as a hot box. This idea was later used by other scientists and astrophysicists across the world to create similar boxes to cook food. In 1873, the photoconductive characteristic of Selenium was discovered by Willoughby Smith. This discovery showed that light could be converted to electricity, but it did not produce enough energy to power electrical equipment this field both in the United States and in Europe. The next big advance in solar technology was in 1954 when Bell Labs introduced the first Photovoltaic Technology (PV) device that generated usable amounts of electricity. The first cell had only 4% efficiency, but later got up to 11% (History of Solar 2012). The next year, PV powered products started to emerge on the market. Since then, the advances in this technology have been remarkable. Solar Panels Today As solar panels are becoming more common, the costs are coming down and the efficiency is increasing. They have also become more cost e ffective to install as experience is becoming more common. There are numerous applications for these panels that convert sunlight into usable energy. This is done by

PAGE 18

6 establishing an electric field with PV cells. The PV cell consists of at least two (2) semiconductor layers, one with a positive charge and the other with a negative charge. As photons from the sunlight are absorbed by the cells by the negative layer, this frees up the electrons. The electrons than travel to the positive layer, and this c reates a voltage differential. This flow of electrons creates a current which can be used for electricity One of these cells creates only 1 or 2 watts, so several of the cells are combined to form a module. Modules are combined to create an array for t he desired amount of energy. A glass surface is on top of the PV cell arrays for protection. Under the glass, there is an antireflective layer to increase the amount of light that reaches the PV cells. The United Stated Federal Government offers tax credits to residential and commercial properties that have installed solar panels to provide electricity not only for their use, but also contributing to the local energy grid. This is a great incentive to counter the high initial cost of these systems. There are also several avenues to lease solar panels and take advantage of the energy produced.

PAGE 19

7 CHAPTER III WIND BEHAVIOR Introduction there has been an increased amount of research into wind loading on structures throughout the wo rld. Wind and seismic loads are both considered to determine the controlling lateral load to use in the design of a structure. This usually depends on the location While wind events are more common, earthquakes usually result in more damage. Over the years, wind storms and earthquakes have on average created the same amount of da mage (Holmes 2007 ). Due to the frequency of wind events, they impact more people. Wind Tunnel Testing then, there have been several different tests done with a variety of panels, placements, geometry, arrays and different environments. ASCE 7 10 S ection 31.2 covers the requir ements that wind tunnel test need to follow these include scaling, modeling, and instrumentation in the w ind tunnel set up (Maffei, 2014 ). For wind tunnel test s on roofs of buildings, scaled models of the buildings are created and tested based on scaled wind load values Models used in w ind tunnel tests are typically simple bui ldings that do not consider sever al factors that can influence the wind pressure: parapets,

PAGE 20

8 landscaping, unique geometry of the buildings, signs, etc. It has been found that the re are s ignificant differences reported among studies (Stathopoulou s, et a l 2012) It has also been found that the wind pressures calculated via building code interpretation were less than the measured pressures from wind tunnel testing, in part due to t he fact that the maximum average pressures measured on large surface areas are less that the maximum pressures measured on small surface areas (Tilley, 2012). This is even more of a reason to compare the results to those from actual size models. The desi gn of these panels could be consistently under designed if this is the case. There was a study done for solar panels mounted on a pitched roof looking at both wind tunnel testing and a full scale test. The results of that study showed that the full scale pressure coefficients were greater than those from the wind tunnel testing (Stathopolous 2012). Because of the way that the wind tunnel test s are set up, the pressure tape is significantly large compared to the scale of the panels. This can easily pr oduce inaccurate results. Also, these test s have shown that there is some sheltering occurring the further away the panels are from the outside of the array. The panels on the edge can experience two to three times greater wind loads than the interior panels (Banks 2011).

PAGE 21

9 Wind tunnel testing is currently the only way to supplement the wind loads in ASCE7 (Cochra n 2010). In fact, most of the written standards are based on the results of wind tunnel testing. Current Codes and Standards The most commonly used standard for engineers in the United States is the American Society of Civil Engineers (ASCE) standard number seven, Minimum Design Loads for Buildings and Other Structures ( ASCE 7). This standard is referenced in the Internation al Building Code (IBC) which is typically adopted by the local Building departments For the design of a solar panel, t his standard uses the wind speed to determine the pressure to be used for the design. This is then used to determine the forces acting on the panel. This will dictate how the panel is secured to the roof. By attaching a solar panel to the roof, the wind load acting on the roof surface is not increased but the structural member will need to be designed for the panels uplift forces (Bank s 2011). The movement of air can be compared to that of a liquid, as they are both fluids. That is why the main equation for the design wind pressure is based on (1) = density of the fluid V = vel ocity of the fluid The velocity pressure of wind in ASCE7 10 is equation 27.3 1:

PAGE 22

10 (2) K z = velocity pressure coefficient K zt = topographic factor K d = wind direction factor V = wind velocity the conditions of the site. In this equation, the first value, 0.00256 is to convert the density of fluid to air at sea level in U.S. customary units The K z accounts for the height above the ground and the exposure of site. This can impact the surface drag and the mean wind flow. The K zt takes into consideration the surrounding topography. K d is for the angle of the wind. Once the velocity pressure is determined, the methods in the ASCE7 10 can be used to find the wind load. As wind approaches a structure, it is r edirected around it: to the sides downwards, and upwards At the top of the wall, there is a separation point and a shear layer is generated at a slope of 2:1 (SEAOC 2012). Above this, the streamline flow continues ; below is a region of vortices The area below the shear layer is the re circulating region, which produces high uplift force. This diminishes further away from the edge of the building. The figure below depicts this behavior in the case of a building with a parapet wall.

PAGE 23

11 Figure 3 1 W ind F low against a B luff Body (Andols ek 2013, used with permission fro m Erin Andolsek) ASCE 7 offers guidance on the design wind speeds and pressures for buildings, rooftop structures, components and cladding but does not cover the design for solar panels on rooftops. The IBC does not offer guidance either. Due to this lack of guidance, the design engineer is left with a few options. One o ption is to use results from wind tunnel testing Another is to use the existing tables in ASCE7 that are intended for different design purposes, but can be applied to the design of solar panels on flat roofs. Or, the design engineer can used the set of tables that SEAOC has provided to assist with this design. In all of these approaches, the K z K zt K d and the importance factors used are the same as what would be used for the design of the actual building, as i t is not typical that the componen ts for a building are designed to higher standards than the actual building.

PAGE 24

12 Using the ASCE tables for roofs, but not necessarily solar panels on a roof the external pressure coefficient for flush mounted solar arrays is from Figure 30.4 1 of the ASCE7 10, Figure 3.2 below. The intended use of this table i n the standard is for component s and cladding for enclosed and partially enclosed buildings. The external gust pressure coefficients, GCp for the roof can be used. This method should have conserva tive loads (Banks 2011). The pressure coefficient is then used to calculate the net pressure.

PAGE 25

13 Figure 3 2 External Pressure Coefficients. Figure 30.4 1 from ASCE 7 10 used to determine external pressure coefficients on components and cladding of en closed buildings ( ASCE 7 10 2010, used with permission from ASCE).

PAGE 26

14 Figure 27.4.4 from the ASCE7 10, Figure 3.3 below, is used to find the net pressure coefficient for monopole free roofs on the ground. This can also be used as guidance in the design of solar panels on flat roofs i f they are placed far enough away from the edge of the roof. The gust factor, G, used for this is typically 0.85 for ri gid structures, but it has been recommended that this be increased to 1 .0 for the design of solar panels (Ba nks 2011).

PAGE 27

15 Figure 3 3 Net Pressure Coefficients for Monoslope Free Roofs. Figure 27.4 4 from ASCE7 10 used to determine wind pressure on monoslope free roofs above ground ( ASCE 7 10 used with permission from ASCE).

PAGE 28

16 It has become standard practic e for design engineers to use standards for other structures in the design of solar panels of flat roofs. There is a great need for a standard for this purpose. This need was recognized by t he Structural Engineers Association of California (SEAOC) and t hey formed a committee for this issue in 2011. They identified many key issues to investigate, and in 2 0 12, SEAOC published Wind Design for Low Profile Solar Photovoltaic Arrays on Flat Roofs This document contained two figures to guide engineers in det ermining the design wind loads for roof mounted solar panel arrays. See Figures 3.4 & 3.5 below. This considers the location of the panel on the roof, the shape of the roof, and the roof features. This document uses equation 3 below to determine the pre ssure. (3) = velocity pressure at mean roof height q h = velocity pressure evaluated at the mean roof height GC r n = combined net pressure coefficient ASCE does intend to include a standard similar to this in its next edition to help address this lack of guidance. Until then, these tables are the best tool that the design engineers have.

PAGE 29

17 Figure 3 4 Design GCp Values Published by SEAOC, August 2012 ( S EAOC 2012, used with permission fro m SEAOC )

PAGE 30

18 Figure 3. 5 Figure 29.9 1 from SEAOC Document Systems on Flat Roofs published in Draft format on March 28, 2012 (SEAOC 2012 used with permission fro m SEAOC ).

PAGE 31

19 CHAPTER IV PREVIOUS STUDIES Introd uction The construction and installation of the full scale faux solar panel experiment was first proposed in the Wind load on solar panel experiment by Erin Dowds, Jennifer Harris, and Frederick Rutz in 2012. These panels were installed at the University of Colorado Denver Campus on the top of the Physical Education Building. While there were several other locations investigated, this was determined to be the most suitable location The considerations that went into this were availability of access to th e roof, the surrounding structures and vegetation, shape of the building being similar to those used in wind tunnel tests, and the features on the roof were flat with a parapet wall and few obstructions in the area of the panels This research was proposed due to the lack of available guidelines for engineers when designing the installation of solar panels to flat roofs. The goal was to help fill the gap in the literature by collecting data from the panels installed to compare to verify the wind tu nnel test (Dowds et. a l. 2012). The solar panel experiment proposed can be seen in the figure below.

PAGE 32

20 Figure 4. 1 Cross section of solar panel experiment (Andolsek, 2013, used with permission from Erin Andlosek). The original proposal for this exp eriment was to use three panels, all at different heights. The first panel was intended to be at a height so that the entire panel is within the recirculation area, Figure 4. 2 .b. The second panel height was selected so the panel would be completely in th e shear layer, Figure 4. 2 .c. And the third panel was constructed to a height so that the panel would be located in both the shear layer and the recirculation area, Figure 4. 2 .d.

PAGE 33

21 Figure 4.2 Proposed panels with relation to the shear layer. (Dowds et. al 2012, used with permission from Erin Andlosek). The construction of the panels is shown in Figure 4.3. The frame was proposed with pinned connections. Strain gauges were connected to the tension (T ) ties that were on both sides of the panel. The strain in the tension ties will be measured and recorded to determine the resultant wind force on the panel in the x direction.

PAGE 34

22 Figure 4.3 Diagram for the panel construction (Andolsek 2013, u sed with permission fro m Erin Andolsek ) The resultant wind force ( F R ) is found with the following equation: (4) F R = resultant force on the panel T = tension in tie p = panel angle = tension tie angel Two locat ions were studied at the University of Colorado Denver Campus, Physical Education Building: the first was four feet away from the outside edge of the parapet, and the second 80 feet from the roof edge. It was anticipated that

PAGE 35

23 the location closer to the ed ge of the roof would experience higher forces than the location further from the edge of the building. See Figure 4. 4 for the two locations. Figure 4. 4 Aerial View of Events Center Building and Surroundings. ( Andolse k 2013, used with permission fro m Erin Andolsek ).

PAGE 36

24 Study of Wind Loads Applied to Full Scale Rooftop Solar Panels by Jennifer Harris The results of the first study were presented by Jennifer Harris in the spring of 2013. The anticipation of this study was that the peak force coefficient s would be greater than the coefficients used in the ASCE (Harris 2013). Two panels were constructed for this research. Shop drawings were prepared for the construction of both of the panels. The Electronics Calibration and Repair Lab at the University of Colorado Denver fabricated the needed members for the panels. The rest of the materials needed for the assembly of the panels were purchased from local hardware stores, or were ordered. Both of the panels consisted of 3/8 inch plywood. Holes were drilled through the plywood surface on each corner for the legs. The legs had inch eye bolts that were inserted through the holes that were drilled in the fabrication lab. The tension ties were 7/16 inch diameter threaded steel rod that was connected to a strain transducer through two inch diameter holes. Nuts were installed on both sides of the strain transducer in or der to leave room for adjustment once assembled. A coupler nut was used to attach the tension ties to the panel legs. A 7/16 inch diameter coupler nut was screed onto the treaded rod and was attached to the inch eye bolt on both the top and bottom of t he panel legs. A 5/16 inch diameter hole was drilled onto the coupler to connect to the eye bolt. The connection was

PAGE 37

25 designed to be a pin end connection so that the all of the horizontal components were directed to the tension ties and would be recorded. Strain gages were adhered to the inside of 2 inch wide, 3 inch diameter steel rings. For wind blowing in the opposite direction a slacked cable brace was installed in the opposite direction of the tension ties. The legs were connected to the base of the panels with the same connection as used for the connection to the panels. that were laid on top of the pavers. See Figures 4.5 through 4.7 for details on the assemblage of the panels described above. Fig ure 4.5 Panel Frame Connection Details.at the Diagonal Tension Ties. ( Harris 2013, used with permission from Jennifer Harris ).

PAGE 38

26 Figure 4. 6 Panel Frame Connection Details in the Short Dimension. ( Harris 2013, used with permission from Jennifer Harris ). Figure 4. 7 Close Up View of Short Direction Panel Frame Connection Details. ( Harris 2013, used with permission from Jennifer Harris ).

PAGE 39

27 The two panels constructed for this are shown in Figure 4. 8 below. One panel is Bothe panels were at a 30 degree angle. The tension ties for the shorter panel were at an angel of 6 degrees and the tension ties for the taller panel were at an angel of 45 degrees. Figu re 4. 8 First Panel Placement. ( Harris 2013, used with permission from Jennifer Harris ). The panels were both assembled and placed on the roof of the Physical Education Building at the University of Colorado Denver Campus. They were

PAGE 40

28 bags were used to counter the uplift forces instead of connecting the panels to the pavers or the roof Three RM Young 3101 Wind Sentry A nemometers were used to measure the wind speed. The heights of the anemometers were such that one was at the assumed shear layer, one above and one below. A RM Young 3301 Wind Sentry Vane was used to record the wind direction. It was installed above the three anemometers. To record the temperature, a Campbell Scientific A3537 Type T thermocouple wire was used. These temperatures were all c onfirmed with the reported temperatures form the National Climate Data Center records. A data logger, Campbell Scientific CR5000, was used to record the desired strains, temperatures, wind speed, and wind directions. This was connected to an external battery that was powered by a solar panel. A PC card was used to store the data, which was downloaded when needed. The program for the data logger was set to collect recordings at 1 sec ond intervals. An SDM INT8 Campbell Scientific interval timer was used along with the CR5000 data logger to record the wind speeds form the anemometers. The Campbell Scientific data logger used RTDAQ 1.1 software for all of the data collected. This can only run on Microsoft Windows XP, Windows Vista, and Windows 7 operating systems. To create the program, Short Cut and CRBasic Editor (Campbell Scientific 2001) were used. The program was written to record the following measurements at 1

PAGE 41

29 second intervals : stains from the strain gauges, wind speeds form the three anemometers, wind direction, temperatures, and the wind direction. lead wires to the strain gauges. The strain transducers were protected by silicone and tape. The strain transducers were all calibrated with an MTS Lab. A section of the treaded rod was used and was representative of how it would be installed in the field. The calibrated load versus strain curves were found for each of the strain gauges. In the field, these gauges were all covered with foil insulation to reduce the impacts that sunlight and temperature would have on the recordings. The data recorded for this study was filtered for times when the wind was in the direction of within 10 degrees of the axis perpendicular to the panel (170 degrees to 190 degrees). C F values were only determined for these times. This study also looked at two other wind directions in order to confirm that the wind in the direction of the face of the panel was governing. Both 20 and 45 degrees, with a tolerance of 10 degrees to either direction were also looked at. The results of the stud y were presented for the panel that was located within the s h ear layer, as the readings from the other panel were inconclusive (Harris 2013). The data was averaged over three second rolling averages. For the case

PAGE 42

30 of wind perpendicular to the panel, and u sing 3 second averages, the majority of the C F values were between 0.6 and 2.2 (Harris 2013). There were often C F values at around 10 and a few up to 20. It was assumed that this was due corner vortices (Ho l mes 2007). For the other two wind direction c ases analyzed, the C F values ranged from 3.6 to 7.2 for 20 degrees form perpendicular to the axis of the panel and for 45 degrees, they ranged from 3.5 to 6.9 (Harris 2013). This confirms the assumption that the maximum forces occur when the wind is perpendicular to the panel. The results of this study were presented at the 2013 12 th Americas Conference on Wind Engineering in Seattle, Washington ( Harris et al. 2013). Study of Full Scale Rooftop Solar Panels Subject to Wind Loads by Erin Andolsek The second study was done in the fall of 2013 by Erin Andolsek. This used the s ame two solar panels; however, they were at a different location on the roof of the Physical Educatio n Building at the University of Colorado Denver. The goal of this research was to provide baseline data to compare to wind tunnel studies (Andolsek 2013). The coefficient obtained from this research, C F is compared to C P used in ASCE 7. The location of the panels for this study was based on where the shear layer is expected to reattach to the roof surface (Andolsek 2013). At this location, the

PAGE 43

31 wind should be streamlined. With the roof height of 38 feet, the panels were located about two times the bu ilding (80 feet) from the edge of the roof. It was expected that the wind speeds here would be less than what was recorded in the first study at the edge of the roof. Figure 4. 9 shows the setup for the Andolsek study. Figure 4. 9 Second Panel Placement. (Andolse k 2013 used with permission fro m Erin Andolsek ). At this location, all three of the anemometers should be in the streamlined region, so it was anticipated that the velocity recordings would be close for all three elevations. This was confirmed by using a 20 foot section of pipe with orange construction tape attached at equal spaces, 5 feet apart. On a windy day, this was brought to the roof and held up at different distances from the edge of the

PAGE 44

32 ro of. As the pole got further form the edge of the roof, more of the tapes reacted to the wind in a less turbulent manner At the location of the panels, all of the tapes had a similar reaction to the wind, thus confirming that the shear layer had reattach ed to the roof surface (Andolsek 2013). See figure 4. 10 for the tapes at the location of this study. Figure 4. 10 Construction tape Experiment 80 feet from Roof Edge. (Andolsek 2013, used with permission fro m Erin Andolsek ) For this study, the re cording frequency was increased to one tenth of a second. The data collected was filtered to only look at times when the wind was in the direction of interest, perpendicular to the panels. This was set at 180 degrees, and a tolerance of 10 degrees was al so considered. After the wind direction was filtered out, times when the wind speed was less than 17 mph was also

PAGE 45

33 eliminated. C F values were only pres ented if they were calculated fro m data with the wind direction of interest as well as wind speeds great er than 17 mph. The data collected showed that there was good correlation between the wind speeds recorded at the three anemometer elevations. Figure 4. 11 from this research shows the coloration of the data. Figure 4. 11 Wind Velocities fro m the three anemometers (Andolsek 2013, used with permission from Erin Andolsek )

PAGE 46

34 The data collected for this study was also averaged over three second rolling averages. It was found that there was a clear response in the strain measurements when the wind vel ocity changed. The C F values recorded for Panel B were on average greater than the values for Panel A. This was what was anticipated due to the height of Panel B being greater than Panel A. The presented C F values reported ranged from 0.1 to 16.3, whit t he majority ranging from 0.1 to 5 (Andolsek 2013). It was found that the vast majority of the reported C F values are within reasonable proximity to C p values presented in ASCE 7 10 (Andolsek 2013). The C F values reported form this study were smaller than when the panels were located at the edge of the roof, as anticipated.

PAGE 47

35 CHAPTER V PANEL DESIGN, LOCATION AND EQUIPMENT Location This study was performed at the University of Colorado Denver Campus. The roof of the Physical Education Building was selected because it is similar in shape to the structures used in wind tunnel studies. This building has a flat roof with a parapet wall and no obstructions. There is also a field in front of the building, in the direction of the prevailing wind, northwest. See Figure 5.1 for the site locating. Figure 5.1 Physical Education Building.

PAGE 48

36 The elevation at this location is approximately 5, 248 feet above sea level. The location is also just outside a special wind region, see figure 5.2. Figure 5.2. Wind Region. (Andolsek 2013, u sed with permission from Erin Andolsek ) There is an exposure category of B and a design wind speed of 115 mph for Risk Category II Buildings, according to the ASCE7. Both panels were located at about two times the height of the building away for the edge of the roof, about 80 feet. At this location, it is expected that the shear layer will have reattached to the roof surface.

PAGE 49

37 Panels Both panels were constructed as a part of previous studies, however, there were alterations made to them so that the height of both panels were the same and at a typical height of solar panels installed on flat roofs. David Banks, with CPP Wind Engineering and Air Quality Consultants in Fort Collins provided that the lower edge of the panels are typically 3 to 10 inches off the roof. The legs of the taller panel were removed and brought to the Electronics Calibration and Re pair Lab at the University of Colorado Denver to be shortened to a length of 2 feet, 8 5/8 inch and 11 7/8 inches See Figure 5.3 below for details.

PAGE 50

38 Figure 5.3. Panel Legs. New legs were purchased for the shorter panel and were also brought to the Lab to be fabricated affording to the provided shop drawings above The new legs were then brought up to the panels and assembled for the desired height. All of the connections remained the same as well as the 30 degree angles of the panels. Also, both of the tension ties needed to be modified or replaced for both panels. Figure 5.4 is the detail for both panels.

PAGE 51

39 Figure 5.4. Panel Detail Section. (Not to scale) 2h = approximately 80 feet The rest of the equipment used for the previous two stu dies stayed at the location and were used for this study as well. The anemometers, Campbell Scientific data logger, and the solar panel all remained in the same locations. The recording frequency for this study was also increased form the ones used for the previous studies. A frequency of 30 Hz was recorded, resulting in very large amounts of data. The program for the data logger needed to be updated for this frequency, the program can be found in Appendix A. The layout for this study is shown in figu re 5. 5.

PAGE 52

40 Figure 5.5. Layout. (Not to scale) 2h = approximately 80 feet (Andlosek 2013, used with permission from Erin Andolsek). Because the same strain transducers were used for this study, there was no need to recalibrate them from the last study. The calibration curves for the strain transducers used are (Harris 2013) : Strain Transducer A: Strain Transducer B: Strain Transducer C: Strain Transducer E: Strain transducers A and B were installed on Panel A and strain transducers C and E were installed on Panel B. The transducers were all wrapped in insulated

PAGE 53

41 foil in order to reduce any impacts that the sunlight and temperature would have on the recorded d ata. The layout of the Transducers is in Figure 5. 6 below. F igure 5.6. Strain Transducer Locations. (Andolsek, 2013, used with permission from Erin Andolsek). Equipment The Campbell Scientific CR5000 Datalogger was used to record and store all of the data collected. This was placed in a metal box behind the panels in order to protect it from being damaged. This was powered by an external battery that

PAGE 54

42 was charged by a Campbell Scientific SP20 Solar Panel. All of the collected data was stored on a Ca mpbell Scientific CFMC2G Flash card. Due to the high recording frequency, this needed to be downloaded every 2 to 3 days. The support software used for the datalogger was Campbell Scientific RTDAQ Version 1.1. To create the program for this, Short Cur a nd CRBasic Editor were used. The program for this study can be found in Appendix A. Data was collected at a 30 Hz frequency for the time, wind direction, wind speed at all 3 anemometers elevations, the air temperature, and the strain data. The data was stored in the flash card until it was downloaded. To download the data, Trendnet, a driver software is need on the computer you are downloading to. As a part of the downloading process, all of the data needed to be converted to a format that wa s usable by Microsoft Excel. Card Convert, a part of the RTDAQ program was used for this. Figures 5.7 and 5.8 show the setup for the study.

PAGE 55

43 Figure 5. 7. Study Location and Layout.

PAGE 56

44 Figure 5. 8. Study Location and Layout.

PAGE 57

45 CHAPTER V I THEORY Individu al solar panel manufactures have conducted numerous wind tunnel studies to determine the appropriate design and installation for their products. They usually use pressure tap e s to measure the pressure distribution for the panel. The net pressure on the t op and bottom surfaces of the panel can be used to determine the pressure coefficient, C P For this study, the force coefficient, C F was calculated to be compared to the pressure coefficient from the available wind tunnel study results. The two panels described earlier were constructed and installed to represent how solar panels would typically be installed on a flat roof. As the wind hits the panel, the pressure is assumed to be uniformly distributed on the face of the panel. The resultant force is w hat is needed, so because the wind does not actually behave this way, this assumption does not influence the results. As the wind creates the upward force on the panel, there is tension in the tie bars. This is measured with the strain transducers on the tension ties. At the same time, the wind velocity and direction is recorded as well. All of the data collected was av eraged over a period of 3 seconds to eliminate any influence that thermal effects would have on the recordings. The barometric pressure was found from the National Climatic Data Center website.

PAGE 58

46 The recorded wind velocity was used to determine the force on the panel, F VP (4) F VP = force on panel = air density (not corrected for altitude) V = measured wind velocity (averaged over 3 seconds) A = area of the panel surface The value C F the coefficient of force, is found from equation 5 below. (5) C F = net force coefficient F R = resultant force on panel F vel press = force on the panel form dynamic pressure form wind

PAGE 59

47 CHAPTER VI I VIBRATION OF PANEL Introduction Both previous studies by Jennifer Harris (Harris 2013) and Erin Andolsek ( Andolsek 2013) had recommended that the natural frequency of the panels be investigated to determine if the vibration of the panel produced any interna l forces that could influence t he readings from the strain transducers. As a part of this study, the natural frequency of the panels was determined. Panel Natural Frequency In order to determine the natural frequency of the panels, strain gauges were first explored. These were placed on the bottom center of the panels; however, they would not stay on the panel surface. Next, accelerometers we brought up to the panels to record the vibration of the panel. For this, three sensors were used. One was attached on the top center of the panel, and the other two were attached to the front legs to measure the accelerations parallel and perpendicular to the face of the panel. See Figures 7.1 through 7.3 for the locations of the sensors

PAGE 60

48 Figure 7.1 Accelerometer Test Set Up Sensor 0 Sensor 2 Sensor 1

PAGE 61

49 Figure 7.2 Accelerometer Test Set Up Figure 7.3 Accelerometer Test Set Up

PAGE 62

50 Once all of the sensors were fastened to the panel, they were connected to a separate data logger to record the movement of the panel and legs. Each sensor had its own channel that was recorded: Chanel 0 was on the top of the panel, Chanel 1 is on the right leg facing the panel, and Chanel 2 is on the left leg (looking in the direction o f Figure 7.3) A few tests were performed to initiate the panel vibrations. Heal Drop tests were done, as well as both legs of the panel were tapped. First, the heal drop test was conducted. The recorded acceleration and frequency from this can be se en in Figures 7.4 7.7 below. Figure 7.4. Heal Drop, all three sensors. Results for Individual sensors are shown in Figures 7.5 7.7.

PAGE 63

51 Figure 7.5. Heal Drop, Sensor 0 : on panel Natural Frequency = 11 Hz.

PAGE 64

52 Figure 7.6. Heal Drop, Sensor 1 : panel leg, perpendicular to panel Natural Frequency = 13.5 Hz.

PAGE 65

53 Figure 7.7. Heal Drop, Sensor 2 : panel leg, parallel to panel Predominate Mode Natural Freque ncy = approximately 35 Hz. These graphs show that the natural frequency of the panel is cl ose to 12 Hz. Both the legs natural frequencies were around 14 Hz. This test was performed several times, and similar results were found. The results of the right leg tap (sensor 1) and the left leg tap (sensor 2) are shown in Figures 7.8 through 7.1 5 below.

PAGE 66

54 Figure 7.8. Leg Tap 1, all three sensors. Results for Individual sensors are shown in Figures 7.9 7.11.

PAGE 67

55 Figure 7.9. Leg Tap 1, Sensor 0: on panel. Natural Frequency = 11 Hz.

PAGE 68

56 Figure 7.10. Leg Tap 1, Sensor 1 : panel leg, perpendicular to panel. Natural Frequency = 13.5 Hz.

PAGE 69

57 Figure 7.11. Leg Tap 1, Sensor 2 : panel leg, parallel to panel. Predominate Mode Natural Frequency = approximately 39.5 Hz.

PAGE 70

58 Figure 7.12. Leg Tap 2, all three sensors. Results for Individual sensors are shown in Figures 7.13 7.15.

PAGE 71

59 Figure 7.13. Leg Tap 2, Sensor 0 : on panel. Natural Frequency = 11 Hz.

PAGE 72

60 Figure 7.14. Leg Tap 2, Sensor 1 : panel leg, perpendicular to panel. Natural Frequency = 12 Hz.

PAGE 73

61 Figure 7.1 5. Leg Tap 2, Sensor 2 : panel leg, parallel to panel. Predominate Mode Natural Frequency = approximately 39.5 Hz. This test was also performed several times, and similar results were found. This test resulted in natural frequencies between about 10 Hz and 40 Hz. These tests were also performed several times, and similar results were found. Conclusion The recorded natural frequencies for the panel were quite a bit greater than 1.0 Hz While the recording frequency for the solar panel itself was not high enough to look for this in the recorded data it can be determined that the vibrations of the

PAGE 74

62 panel had no significant influence on recordings fro m the strain transducers. This is due to the fact that 1.0 Hz is the max frequency to consider vibratio n per ASCE7 10. The range of natural frequencies found for the pane ls were well above this value so may be considered a rigid structure.

PAGE 75

63 CHAPTER VII I RESULTS Introduction The results from this study are presented below. The C F values, wind velocities, and strains are shown in the graphs below for both panels and were recorded at 30 Hz. The recordings were processed using three second rolling averages. Results This study resulted in a very large amount of data, most of whic h was not useful for this analysis. Only times when the majority of the wind velocities were greater than 17 mph and in the direction perpendicular to the panels were used. A tolerance of 10 degrees to either direction was allowed for the wind direction. To determine the C F values, the data was filtered even more. B oth the wind velocity and the strain values had to be increasing or decreasing together during the tine period of interest This eliminated any C F values that were not a result from the chan ge in strain. This process eliminated the vast majority of the C F values. The results c an be found in Figures 8.1 8.14 And Tables 8.1 8. 14 The strain is graphed on an arbit rary scale because only the differential strain is of interest.

PAGE 76

64 Figure 8.1. C F Values, Strain, and Wind Velocity from 4/28/14, Panel A. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. The upward trend in strain is attributed to thermal drift The strain output varies with temperature. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 1 998 1995 2992 3989 4986 5983 6980 7977 8974 9971 10968 11965 12962 13959 14956 15953 16950 17947 18944 19941 Wind Velocity (mph) & C F Time (1/30 sec) Panel A 4/28/14 3:41 PM 3s Strain Intervals & 3s Average Wind Speed Wind Velocity Strain A Strain B CF

PAGE 77

65 Table 8.1. C F Values from Figure 8.1. C F Values 0.1 0.2 0.2 0.7 0.1 0.2 0.3 0.7 0.1 0.2 0.3 0.7 0.1 0.2 0.3 0.9 0.1 0.2 0.3 0.9 0.1 0.2 0.3 0.9 0.1 0.2 0.3 1.1 0.1 0.2 0.3 1.1 0.1 0.2 0.4 1.5 0.2 0.2 0.4 3.8 0.2 0.2 0.4 3.9 0.2 0.2 0.5 6.7 0.2 0.2 0.6

PAGE 78

66 Figure 8.2 C F Values, Strain, and Wind Velocity from 4/28/14, Panel B. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0 1 952 1903 2854 3805 4756 5707 6658 7609 8560 9511 10462 11413 12364 13315 14266 15217 16168 17119 18070 19021 19972 Wind Velocity (mph) & CF Time ( 1/30 sec) Panel B 4/28/14 3:41 PM 3s Strain Intervals & 3s Average Wind Speed CF Wind Velocity Strain C Strain E

PAGE 79

67 Table 8.2. C F Values from Figure 8.2. C F Values 0.1 0.2 0.3 0.7 0.1 0.2 0.3 0.7 0.1 0.2 0.4 0.7 0.1 0.2 0.4 0.7 0.1 0.2 0.4 0.8 0.1 0.2 0.4 0.8 0.1 0.2 0.4 0.8 0.1 0.2 0.4 0.8 0.2 0.2 0.4 0.9 0.2 0.2 0.4 1.0 0.2 0.2 0.4 1.0 0.2 0.2 0.4 1.1 0.2 0.2 0.4 1.2 0.2 0.3 0.4 1.3 0.2 0.3 0.4 1.4 0.2 0.3 0.4 1.5 0.2 0.3 0.4 1.5 0.2 0.3 0.5 1.6 0.2 0.3 0.5 2.5 0.2 0.3 0.5 3.5 0.2 0.3 0.5 3.8 0.2 0.3 0.5 4.1 0.2 0.3 0.5 5.0 0.2 0.3 0.5 5.0 0.2 0.3 0.6 0.2 0.3 0.6 0.2 0.3 0.7 0.2 0.3 0.7

PAGE 80

68 Figure 8.3. C F Values, Strain, and Wind Velocity from 4/27/14, Panel A. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 1 2412 4823 7234 9645 12056 14467 16878 19289 21700 24111 26522 28933 31344 33755 36166 38577 40988 43399 45810 48221 50632 Wind Velocity (mph) & C F Time ( 1/30 sec) Panel A 4/27/14 1:34 PM 3s Strain Intervals & 3s Average Wind Speed Wind Velocity Strain A Strain B CF

PAGE 81

69 Table 8.3. C F Values from Figure 8.3. C F Values 0.1 0.2 0.4 0.9 0.1 0.2 0.4 1.4 0.1 0.2 0.4 4.2 0.1 0.2 0.4 0.1 0.3 0.5 0.2 0.4 0.5

PAGE 82

70 Figure 8.4. C F Values, Strain, and Wind Velocity from 4/27/14, Panel B. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. No data collected for Strain D. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 1 2412 4823 7234 9645 12056 14467 16878 19289 21700 24111 26522 28933 31344 33755 36166 38577 40988 43399 45810 48221 50632 Wind Velocity (mph) & CF Time ( 1/30 sec) Panel B 4/27/14 1:34 PM 3s Strain Intervals & 3s Average Wind Speed CF Wind Velocity Strain C

PAGE 83

71 Table 8.4. C F Values from Fig ure 8.4. C F Values 0.1 0.2 0.2 1.1 0.1 0.2 0.2 1.2 0.1 0.2 0.2 1.3 0.1 0.2 0.3 1.3 0.1 0.2 0.3 1.4 0.1 0.2 0.4 1.5 0.1 0.2 0.4 1.9 0.1 0.2 0.5 3.8 0.1 0.2 0.5 0.1 0.2 0.5 0.1 0.2 0.5 0.1 0.2 0.6 0.2 0.2 0.6 0.2 0.2 0.7 0.2 0.2 0.8 0.2 0.2 0.8 0.2 0.2 1.0

PAGE 84

72 Figure 8.5. C F Values, Strain, and Wind Velocity from 4/27/14, Panel A. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 1 1726 3451 5176 6901 8626 10351 12076 13801 15526 17251 18976 20701 22426 24151 25876 27601 29326 31051 32776 34501 36226 37951 39676 Wind Velocity (mph) & C F Time ( 1/30 sec) Panel A 4/27/14 10:51 PM 3s Strain Intervals & 3s Average Wind Speed Wind Velocity Strain A Strain B CF

PAGE 85

73 Table 8.5. C F Values from Figure 8.5. C F Values 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.3 0.3 0.3 0.4 0.5 0.5 3.4

PAGE 86

74 Figure 8.6. C F Values, Strain, and Wind Velocity from 4/27/14, Panel B. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 1 1883 3765 5647 7529 9411 11293 13175 15057 16939 18821 20703 22585 24467 26349 28231 30113 31995 33877 35759 37641 39523 Wind Velocity (mph) & CF Time ( 1/30 sec) Panel B 4/27/14 10:51 PM 3s Strain Intervals & 3s Average Wind Speed CF Wind Velocity Strain C Strain E

PAGE 87

75 Table 8.6. C F Values from Figure 8.6. C F Values 0.1 0.2 0.2 0.5 0.1 0.2 0.2 0.6 0.1 0.2 0.2 0.6 0.1 0.2 0.2 0.7 0.1 0.2 0.2 0.7 0.1 0.2 0.3 0.7 0.1 0.2 0.3 1.0 0.1 0.2 0.4 1.6 0.1 0.2 0.4 1.9 0.1 0.2 0.5 4.0

PAGE 88

76 Figure 8.7. C F Values, Strain, and Wind Velocity from 4/28/14, Panel A. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0 5 10 15 20 25 30 35 1 259 517 775 1033 1291 1549 1807 2065 2323 2581 2839 3097 3355 3613 3871 4129 4387 4645 4903 5161 5419 5677 Wind Velocity (mph) & C F Time ( 1/30 sec) Panel A 4/28/14 7:24 PM 3s Strain Intervals & 3s Average Wind Speed Wind Velocity Strain A Strain B CF

PAGE 89

77 Table 8.7. C F Values from Figure 8.7. C F Values 0.1 0.1 0.3 0.6 0.1 0.2 0.3 1.1 0.1 0.2 0.4 1.6 0.1 0.2 0.4 2.2 0.1 0.2 0.5 2.4 0.2

PAGE 90

78 Figure 8.8. C F Values, Strain, and Wind Velocity from 4/28/14, Panel B. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 1 271 541 811 1081 1351 1621 1891 2161 2431 2701 2971 3241 3511 3781 4051 4321 4591 4861 5131 5401 5671 Wind Velocity (mph) & CF Time ( 1/30 sec) Panel B 4/28/14 7:24 PM 3s Strain Intervals & 3s Average Wind Speed CF Wind Velocity Strain C Strain E

PAGE 91

79 Table 8.8. C F Values from Figure 8.8. C F Values 0.1 0.2 0.3 0.6 0.1 0.2 0.3 0.7 0.1 0.2 0.3 0.9 0.1 0.2 0.4 1.4 0.2 0.2 0.4 1.8 0.2 0.2 0.4 2.3 0.2 0.2 0.5 2.6 0.2 0.2 0.5 10.0 0.2

PAGE 92

80 Figure 8.9. C F Values, Strain, and Wind Velocity from 4/27/14, Panel A. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 1 3638 7275 10912 14549 18186 21823 25460 29097 32734 36371 40008 43645 47282 50919 54556 58193 61830 65467 69104 72741 76378 Wind Velocity (mph) & C F Time ( 1/30 sec) Panel A 4/27/14 11:02 AM 3s Strain Intervals & 3s Average Wind Speed Wind Velocity Strain A Strain B CF

PAGE 93

81 Table 8.9. C F Values from Figure 8.9. C F Values 0.1 0.2 0.2 0.6 0.1 0.2 0.2 0.7 0.1 0.2 0.4 2.6 0.1 0.2 0.4 2.6 0.1 0.2 0.4 4.1 0.1 0.2 0.5 0.1 0.2 0.6

PAGE 94

82 Figure 8.10. C F Values, Strain, and Wind Velocity from 4/27/14, Panel B. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 1 3638 7275 10912 14549 18186 21823 25460 29097 32734 36371 40008 43645 47282 50919 54556 58193 61830 65467 69104 72741 76378 Wind Velocity (mph) & CF Time ( 1/30 sec) Panel B 4/27/14 11:02 AM 3s Strain Intervals & 3s Average Wind Speed CF Wind Velocity Strain C Strain E

PAGE 95

83 Table 8.10. C F Values from Figure 8.10. C F Values 0.1 0.2 0.3 0.4 0.1 0.2 0.3 0.4 0.1 0.2 0.3 0.5 0.1 0.2 0.3 0.5 0.1 0.2 0.3 0.6 0.1 0.2 0.3 0.7 0.2 0.2 0.3 0.8 0.2 0.2 0.3 1.1 0.2 0.2 0.3 1.1 0.2 0.2 0.3 1.1 0.2 0.2 0.4 1.4 0.2 0.2 0.4 2.1 0.2 0.2 0.4 3.2 0.2 0.2 0.4 3.9 0.2 0.2 0.4 4.5 0.2 0.3 0.2 0.3

PAGE 96

84 Figure 8.11. C F Values, Strain, and Wind Velocity from 4/28/14, Panel A. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 1 681 1361 2041 2721 3401 4081 4761 5441 6121 6801 7481 8161 8841 9521 10201 10881 11561 12241 12921 13601 14281 Wind Velocity (mph) & C F Time (sec) Panel A 4/28/14 4:39 PM 3s Strain Intervals & 3s Average Wind Speed Wind Velocity Strain A Strain B CF

PAGE 97

85 Table 8.11. C F Values from Figure 8.11. C F Values 0.1 0.2 0.2 0.6 0.1 0.2 0.2 0.7 0.1 0.2 0.3 0.9 0.1 0.2 0.3 0.9 0.1 0.2 0.3 1.2 0.1 0.2 0.4 2.3 0.1 0.2 0.4 3.2 0.2 0.2 0.5

PAGE 98

86 Figure 8.12. C F Values, Strain, and Wind Velocity from 4/28/14, Panel B. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 1 651 1301 1951 2601 3251 3901 4551 5201 5851 6501 7151 7801 8451 9101 9751 10401 11051 11701 12351 13001 13651 14301 Wind Velocity (mph) & CF Time (sec) Panel B 4/28/14 4:39 PM 3s Strain Intervals & 3s Average Wind Speed CF Wind Velocity Strain C Strain E

PAGE 99

87 Table 8.12. C F Values from Figure 8.12. C F Values 0.1 0.2 0.2 0.5 0.1 0.2 0.3 0.5 0.1 0.2 0.3 0.6 0.1 0.2 0.3 1.0 0.1 0.2 0.3 1.4 0.1 0.2 0.4 2.6 0.1 0.2 0.5 2.7 0.2 0.2 0.5 4.3 0.2

PAGE 100

88 Figure 8.13. C F Values, Strain, and Wind Velocity from 4/28/14, Panel A. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 1 808 1615 2422 3229 4036 4843 5650 6457 7264 8071 8878 9685 10492 11299 12106 12913 13720 14527 15334 16141 16948 Wind Velocity (mph) & C F Time ( 1/30 sec) Panel A 4/28/14 12:00 AM 3s Strain Intervals & 3s Average Wind Speed Wind Velocity Strain A Strain B CF

PAGE 101

89 Table 8.13. C F Values from Figure 8.13. C F Values 0.1 0.2 0.1 0.3 0.1 0.4 0.1 0.4 0.1 0.6 0.1 0.9 0.1 7.4

PAGE 102

90 Figure 8.14. C F Values, Strain, and Wind Velocity from 4/28/14, Panel B. The data is form a selected portion of the study using 3 second rolling averages. An arbitrary scale is used for the stain values. 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 1 808 1615 2422 3229 4036 4843 5650 6457 7264 8071 8878 9685 10492 11299 12106 12913 13720 14527 15334 16141 16948 Wind Velocity (mph) & CF Time ( 1/30 sec) Panel B 4/28/14 12:00 AM 3s Strain Intervals & 3s Average Wind Speed CF Wind Velocity Strain C Strain E

PAGE 103

91 Table 8.14. C F Values from Figure 8.14. C F Values 0.1 0.1 0.3 0.1 0.1 0.3 0.1 0.1 0.4 0.1 0.2 0.5 0.1 0.2 0.5 0.1 0.2 0.6 0.1 0.3 0.9 7.1 Discussion The periods of time represented in the graphs were chosen based on the criteria of velocity and direction discussed above. Both the strains and the wind velocities were averaged over 3 seconds due to the high variability in these values. The results from each panel are represented on separate graphs, but the time intervals are the same. With both panels being the same height, it is expected that similar values would result from both. The C F Values that resulted from this study range from 0.1 to 10 S imilar to the previous studies, the majority of the C F Values found were less than 1. The C F values are the ratio of F R the measured force, over the F VP the force due to the velocity pressure. So, if there was a large change in the strain, than there wo uld be a large C F value. Smaller change in strains would result in small C F values. Along with the minimum wind velocity and wind direction requirements, the slopes of the strain and velocity also had to be in the same direction to qualify the

PAGE 104

92 correspondi ng C F In addition to these requirements, if there was not a peak in the wind velocity, but there was a large C F value recorded, it was discarded. The C F Values were also discarded if there was not a similar value from the other panel. To validate these results, a repeatability study was done. This study was conducted in order to verify that the peak force coefficients were consistent during different times. The top 10 coefficients for each time period pleated above were averaged (Xypnitou 2012). The results of this stud y are shown below in Table 8.15. Table 8.15. Average Peak C F from Figures 8.1 8.14. C F Average Figure Table Panel 2.15 8.1 8.1 A 2.99 8.2 8.2 B 0.95 8.3 8.3 A 1.53 8.4 8.4 B 0.63 8.5 8.5 A 1.23 8.6 8.6 B 0.98 8.7 8.7 A 2.13 8.8 8.8 B 1.29 8.9 8.9 A 1.99 8.10 8.10 B 1.11 8.11 8.11 A 1.46 8.12 8.12 B 1.05 8.13 8.13 A 1.11 8.14 8.14 B

PAGE 105

93 The top average coefficients are close in values, therefore, confirming the validity of this study. T he majority of the C F values were relatively close to the C P values in the ASCE7 10. Using figure 27.4 4 of the ASCE7 10, it ca n be assumed that this location on the roof (2h, or about 80 feet from the edge of the roof) will behave similar to those on the earth. The panels have the following characteristics: = 30 degrees and = 0 degrees. This results in the clear wind flow, C p ranging from 0.5 to 2.5 using this figure. The majority of the C F values from this study do fall within this range.

PAGE 106

94 CHAPTER IX CONCLUSION Conclusion The Force Coefficients resulting from this study were found to be less than the coefficients from in the Study of Wind Loads on Rooftop Mounted Solar Panels by Jennifer Harris. This was expected; as the coefficients are expected to increase the closer the panels are located to the edge of the roof. The average C F value from this study for Panel B was 4.9. When the results of this study are compared to those from the Study of Full Scale Rooftop Solar Panels Subject to Wind Loads by Erin Andolsek, the coefficients were also lower. The average C F value from study for Panel B was 3.7 and 2.4 for Panel A. The average C F values from this study were found to be 1.2 for Panel A and 1.8 for Panel B There were less cases of wind in the direction of interest as well as with speeds greater than 17 mph than the study done by Erin Andolsek. This would result in lower averages overall for this study. It was found that the vibration of the panel did not have an influence on the strain transducer recordings, so did not influence the coefficients for this study.

PAGE 107

95 Figure 27.4 4 from the ASCE7 10 is currently used to determine the loads on solar panels, as discussed above. Us ing the characteristics of the faux solar panels used for this study, the C F values determined do for the most part fall within the range of C P values from figure 27.4.4 of the ASCE7 10. Possible Sources of Error Some factors that could have resulted in e rror for this study are: 1. The frames were not attached directly to the roof surface. In order to prevent any damage to the roof, the panels were pavers on pavers that sat on the roof. The panels were held down by sand bags and scrap metal. These obstruc tions could have affected the wind around the panels. 2. Another factor that could affect the results is that the measurements taken for this study were based on wind speeds less than the design wind speed. 3. The C F values for Panel B were consistently great er than the values for Panel A. This could be due to physical differences between the panel assemblies despite efforts to make them identical as they were designed to be identical. It could also be due to the locations of the gusts on the roof, or the placement of the equipment (sandbags and rod iron pieces) used to hold down the panels.

PAGE 108

96 Recommendations for Fu rther Research Below are a few items that could be considered for any further research on wind loads on solar panels installed on flat roofs: 1. Ac tual solar panels could be used instead of the faux panels constructed for this study. This would be a good verification of the C F values from this study as well as the previous studies. 2. A similar study can be done at a different location with similar site characteristics. This would be a good study to corroborate the recordings at this location. 3. A similar study could also be done at a different location with different site characteristi cs This would be a good verification of the results from this study as well as the previous two. 4. A similar study could also be done at the same location, but at a different place on the roof. Studying the effects of the wind loads when the panels are l ocated at the corner of the roof would be useful to see the impacts of corner vortices. 5. A study with a larger array would be good to see the sheltering affects. 6. Recording at a higher sampling rate to verify that the vibrations of the panels do not influen ce the strain gauge recordings.

PAGE 109

97 REFERENCES American Society of Civil Engineers (2013 ). Minimum Design Loads for Buildings and Other Structures ASCE 7 10. Andolsek, E. K. Colorado Denver, De nver, Colorado Butti, K. Perlin, J (1980). A Golden Thread: 2500 Years of Solar Architecture and Technology 1st Edition, Cheshire Books. Campbell Scientific. (2007). 03001 R.M. Young Wind SentrySset Instruction Manual. Logan, UT: Campbell Scientific, Inc. Campbell Scientific. (2011). SDM INT8 8 Channel Interval Timer Instruction Manual. Logan, UT: Campbell Scientific, Inc. Campbell Scientific. (2001). CR5000 Measurement and Control System Cochran, L., Derickson, R., (2010). A Wind Engineering, The Fifth Symposium on Computational Wind Engineering, Capitol Hill, North Carolina, May 23 27, 2010. Dowds, E.K., Harris, J.S., Rutz, F.R Engineers Workshop, Hyannis, MA, Aug. 12 14, 2012, AAWE, Ft. Collins, CO. Wind Engineering, Seattle, WA, June 16 20, 2013, AAWE, Ft. Collins, CO. Harris, J. D. Dept. of Civil Engine ering. Univ. of Colorado Denver, Denver, Colorado, United States. Holmes, John D., (2001). Wind Loading of Structures 2 nd Edition, Taylor and Francis, New York, N.Y.

PAGE 110

98 Maffei, J. Telleen, K. Ward, R. Kopp, G. and Schellenberg, A. (2014). Design Practice and Recommendations for Solar Arrays on Low Slope J. Struct. Eng. 140(2), 04013040. U.S. Department of Energy. The History of Solar. ( accessed August 23 2013). SEAOC Solar Ph otovoltaic Systems Committee, (2012). Wind Design for Low Profile Solar Photovoltaic Arrays on Flat Roofs SEAOC PV2 2012, Structural Engineers Association of California, Sacramento, August, 2012. Stathopolous, T., Zisis, I., Xypnitou, E. (2012). Proceedings of Structures Congress 2012 March 29 31, Chicago, Ill., Structural Engineering Institute. Based Mounting Structur e Magazine July 2012. 10 12. U.S. Department of Energy. The History of Solar. (accessed September 24 201 5 ). < http://www.energy.gov/articles/top 6 things you didnt know about solar energy > (accessed September 23, 2015) Xypnitou Eleni, (2012). Wind Loads on Solar Panel Systems Attached to Building Roofs, M.S. Thesis, Concordia University, Montreal, Quebec, Canada.

PAGE 111

99 APPENDIX A Datalogger Program

PAGE 112

' C R 5 0 0 0 C r e a t e d b y S h o r t C u t ( 2 9 ) D e c l a r e V a r i a b l e s a n d U n i t s P u b l i c B a t t V P u b l i c F C L o a d e d P u b l i c P T e m p C P u b l i c C R e p s P u b l i c Z M o d e P u b l i c Q B S S M o d e P u b l i c C I n d e x P u b l i c C A v g P u b l i c L C o u n t P u b l i c S t r a i n ( 3 ) P u b l i c V r 1 0 0 0 ( 3 ) P u b l i c G F A d j ( 3 ) P u b l i c B r Z e r o ( 3 ) P u b l i c C K n o w n ( 3 ) P u b l i c C R e p s 2 P u b l i c Z M o d e 2 P u b l i c Q B S S M o d e 2 P u b l i c C I n d e x 2 P u b l i c C A v g 2 P u b l i c L C o u n t 2 P u b l i c S t r a i n 2 ( 4 ) P u b l i c V r 1 0 0 0 2 ( 4 ) P u b l i c G F A d j 2 ( 4 ) P u b l i c B r Z e r o 2 ( 4 ) P u b l i c C K n o w n 2 ( 4 ) P u b l i c C R e p s 3 P u b l i c Z M o d e 3 P u b l i c Q B S S M o d e 3 P u b l i c C I n d e x 3 P u b l i c C A v g 3 P u b l i c L C o u n t 3 P u b l i c S t r a i n 3 ( 3 ) P u b l i c V r 1 0 0 0 3 ( 3 ) P u b l i c G F A d j 3 ( 3 ) P u b l i c B r Z e r o 3 ( 3 ) P u b l i c C K n o w n 3 ( 3 ) P u b l i c W S m p h P u b l i c W i n d D i r P u b l i c W S m p h 2 P u b l i c T e m p F P u b l i c G F s R a w ( 3 ) = { 2 1 1 5 2 1 1 5 2 1 1 5 } P u b l i c G F s R a w 2 ( 4 ) = { 2 1 1 5 2 1 1 5 2 1 1 5 2 1 1 5 } P u b l i c G F s R a w 3 ( 3 ) = { 2 1 1 5 2 1 1 5 2 1 1 5 } P u b l i c I n t 8 ( 5 ) P u b l i c P u l s e C h ( 2 ) D i m I U n i t s B a t t V = V o l t s U n i t s P T e m p C = D e g C U n i t s S t r a i n = m i c r o s t r a i n U n i t s V r 1 0 0 0 = m V / V U n i t s G F A d j = u n i t l e s s U n i t s B r Z e r o = m V / V U n i t s S t r a i n 2 = m i c r o s t r a i n U n i t s V r 1 0 0 0 2 = m V / V U n i t s G F A d j 2 = u n i t l e s s U n i t s B r Z e r o 2 = m V / V Page 1 of 3 Program: Wind7_4new2.CR5

PAGE 113

U n i t s S t r a i n 3 = m i c r o s t r a i n U n i t s V r 1 0 0 0 3 = m V / V U n i t s G F A d j 3 = u n i t l e s s U n i t s B r Z e r o 3 = m V / V U n i t s W S m p h = m i l e s / h o u r U n i t s W i n d D i r = d e g r e e s U n i t s W S m p h 2 = m i l e s / h o u r U n i t s T e m p F = D e g F D e f i n e D a t a T a b l e s D a t a T a b l e ( W i n d 7 T r u e 1 ) D a t a I n t e r v a l ( 0 3 0 m S e c 1 0 ) C a r d O u t ( 0 1 ) S a m p l e ( 1 P T e m p C F P 2 ) S a m p l e ( 1 S t r a i n ( 1 ) I E E E 4 ) S a m p l e ( 1 S t r a i n ( 2 ) I E E E 4 ) S a m p l e ( 1 S t r a i n ( 3 ) I E E E 4 ) S a m p l e ( 1 S t r a i n 2 ( 3 ) I E E E 4 ) S a m p l e ( 1 S t r a i n 2 ( 4 ) I E E E 4 ) S a m p l e ( 1 S t r a i n 3 ( 1 ) I E E E 4 ) S a m p l e ( 1 S t r a i n 3 ( 2 ) I E E E 4 ) S a m p l e ( 1 S t r a i n 3 ( 3 ) I E E E 4 ) S a m p l e ( 1 W i n d D i r F P 2 ) S a m p l e ( 1 T e m p F F P 2 ) S a m p l e ( 1 I n t 8 ( 3 ) F P 2 ) S a m p l e ( 1 I n t 8 ( 4 ) F P 2 ) S a m p l e ( 1 I n t 8 ( 5 ) F P 2 ) E n d T a b l e D a t a T a b l e ( T a b l e 2 T r u e 1 ) D a t a I n t e r v a l ( 0 1 4 4 0 M i n 1 0 ) M i n i m u m ( 1 B a t t V F P 2 F a l s e F a l s e ) E n d T a b l e C a l i b r a t i o n h i s t o r y t a b l e D a t a T a b l e ( C a l H i s t N e w F i e l d C a l 1 0 ) C a r d O u t ( 0 1 0 ) S a m p l e F i e l d C a l E n d T a b l e M a i n P r o g r a m B e g i n P r o g I n i t i a l i z e c a l i b r a t i o n v a r i a b l e s f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 ( ) C I n d e x = 1 : C A v g = 1 : C R e p s = 3 F o r L C o u n t = 1 T o 3 G F A d j ( L C o u n t ) = G F s R a w ( L C o u n t ) N e x t I n i t i a l i z e c a l i b r a t i o n v a r i a b l e s f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 2 ( ) C I n d e x 2 = 1 : C A v g 2 = 1 : C R e p s 2 = 4 F o r L C o u n t 2 = 1 T o 4 G F A d j 2 ( L C o u n t 2 ) = G F s R a w 2 ( L C o u n t 2 ) N e x t I n i t i a l i z e c a l i b r a t i o n v a r i a b l e s f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 3 ( ) C I n d e x 3 = 1 : C A v g 3 = 1 : C R e p s 3 = 3 F o r L C o u n t 3 = 1 T o 3 G F A d j 3 ( L C o u n t 3 ) = G F s R a w 3 ( L C o u n t 3 ) N e x t L o a d t h e m o s t r e c e n t c a l i b r a t i o n v a l u e s f r o m t h e C a l H i s t t a b l e Page 2 of 3 Program: Wind7_4new2.CR5

PAGE 114

F C L o a d e d = L o a d F i e l d C a l ( T r u e ) M a i n S c a n S c a n ( 3 0 m S e c 1 0 ) D e f a u l t D a t a l o g g e r B a t t e r y V o l t a g e m e a s u r e m e n t B a t t V B a t t e r y ( B a t t V ) D e f a u l t W i r i n g P a n e l T e m p e r a t u r e m e a s u r e m e n t P T e m p C P a n e l T e m p ( P T e m p C 2 5 0 ) Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 ( ) B r F u l l ( V r 1 0 0 0 ( ) 3 m V 2 0 1 V x 1 3 5 0 0 0 T r u e T r u e 0 2 5 0 1 0 ) C a l c u l a t e d s t r a i n r e s u l t S t r a i n ( ) f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 ( ) S t r a i n C a l c ( S t r a i n ( ) 3 V r 1 0 0 0 ( ) B r Z e r o ( ) 1 G F A d j ( ) 0 ) Q u a r t e r b r i d g e s t r a i n s h u n t c a l i b r a t i o n f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 ( ) F i e l d C a l S t r a i n ( 1 3 S t r a i n ( ) 1 G F A d j ( ) 0 Q B S S M o d e C K n o w n ( ) C I n d e x C A v g G F s R a w ( ) 0 ) Z e r o i n g c a l i b r a t i o n f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 ( ) F i e l d C a l S t r a i n ( 1 0 V r 1 0 0 0 ( ) C R e p s 0 B r Z e r o ( ) Z M o d e 0 C I n d e x C A v g 0 S t r a i n ( ) ) Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 2 ( ) B r F u l l ( V r 1 0 0 0 2 ( ) 4 m V 2 0 4 V x 2 4 5 0 0 0 T r u e T r u e 0 2 5 0 1 0 ) C a l c u l a t e d s t r a i n r e s u l t S t r a i n 2 ( ) f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 2 ( ) S t r a i n C a l c ( S t r a i n 2 ( ) 4 V r 1 0 0 0 2 ( ) B r Z e r o 2 ( ) 1 G F A d j 2 ( ) 0 ) Q u a r t e r b r i d g e s t r a i n s h u n t c a l i b r a t i o n f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 2 ( ) F i e l d C a l S t r a i n ( 1 3 S t r a i n 2 ( ) 1 G F A d j 2 ( ) 0 Q B S S M o d e 2 C K n o w n 2 ( ) C I n d e x 2 C A v g 2 G F s R a w 2 ( ) 0 ) Z e r o i n g c a l i b r a t i o n f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 2 ( ) F i e l d C a l S t r a i n ( 1 0 V r 1 0 0 0 2 ( ) C R e p s 2 0 B r Z e r o 2 ( ) Z M o d e 2 0 C I n d e x 2 C A v g 2 0 S t r a i n 2 ( ) ) 0 3 0 0 1 W i n d S p e e d & D i r e c t i o n S e n s o r m e a s u r e m e n t s W S m p h a n d W i n d D i r B r F u l l ( V r 1 0 0 0 3 ( ) 3 m V 2 0 1 6 V x 3 3 5 0 0 0 T r u e T r u e 0 2 5 0 1 0 ) C a l c u l a t e d s t r a i n r e s u l t S t r a i n 3 ( ) f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 3 ( ) S t r a i n C a l c ( S t r a i n 3 ( ) 3 V r 1 0 0 0 3 ( ) B r Z e r o 3 ( ) 1 G F A d j 3 ( ) 0 ) Q u a r t e r b r i d g e s t r a i n s h u n t c a l i b r a t i o n f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 3 ( ) F i e l d C a l S t r a i n ( 1 3 S t r a i n 3 ( ) 3 G F A d j 3 ( ) 0 Q B S S M o d e 3 C K n o w n 3 ( ) C I n d e x 3 C A v g 3 G F s R a w 3 ( ) 0 ) Z e r o i n g c a l i b r a t i o n f o r Q u a r t e r B r i d g e S t r a i n 3 w i r e 3 5 0 o h m w i t h 4 W F B S 3 5 0 T I M m e a s u r e m e n t V r 1 0 0 0 3 ( ) F i e l d C a l S t r a i n ( 1 0 V r 1 0 0 0 3 ( ) C R e p s 3 0 B r Z e r o 3 ( ) Z M o d e 3 0 C I n d e x 3 C A v g 3 0 S t r a i n 3 ( ) ) 0 3 0 0 1 W i n d S p e e d & D i r e c t i o n S e n s o r m e a s u r e m e n t s W S m p h a n d W i n d D i r P u l s e C o u n t ( W S m p h 1 1 1 1 1 6 7 7 0 4 ) I f W S m p h < 0 4 1 T h e n W S m p h = 0 B r H a l f ( W i n d D i r 1 m V 5 0 0 0 1 7 3 1 5 0 0 0 T r u e 0 2 5 0 3 5 5 0 ) I f W i n d D i r > = 3 6 0 T h e n W i n d D i r = 0 0 3 1 0 1 W i n d S p e e d S e n s o r m e a s u r e m e n t W S m p h 2 P u l s e C o u n t ( W S m p h 2 1 2 1 1 1 6 7 7 0 4 ) I f W S m p h 2 < 0 4 1 T h e n W S m p h 2 = 0 T y p e T ( c o p p e r c o n s t a n t a n ) T h e r m o c o u p l e m e a s u r e m e n t s T e m p F T C D i f f ( T e m p F 1 m V 2 0 C 8 T y p e T P T e m p C T r u e 0 2 5 0 1 8 3 2 ) m e a s u r e 0 3 1 0 1 o n S D M I N T 8 c h a n n e l 1 t h r o u g h c h a n n e l 5 S D M I N T 8 ( I n t 8 ( ) 0 0 0 0 2 2 2 2 2 0 0 0 2 2 2 2 2 3 2 7 6 8 1 1 6 6 7 0 4 ) F o r I = 1 t o 5 I f I n t 8 ( I ) < 0 2 1 T h e n I N T 8 ( I ) = 0 n e x t I C a l l D a t a T a b l e s a n d S t o r e D a t a C a l l T a b l e ( W i n d 7 ) C a l l T a b l e ( T a b l e 2 ) C a l l T a b l e ( C a l H i s t ) N e x t S c a n E n d P r o g Page 3 of 3 Program: Wind7_4new2.CR5