Self-Sustaining Street Light Final Report Part B

Self-Sustaining Street Light

L.C. Smith College of Engineering and Computer Science

Syracuse University

Reid Berdanier

Karen Hernandez

Christopher Horvath

Laura Graham

Chuck Raye

MEE 472

Spring 2010

Table of Contents Self-Sustaining Street Light

2

Abstract ......................................................................................................................................4

Acknowledgements.....................................................................................................................5

Chapter 1: Introduction ...............................................................................................................6

Chapter 2: Ethical Implications ...................................................................................................8

Chapter 3: Stress Analysis.........................................................................................................12

Chapter 4: Wind Turbine...........................................................................................................23

Chapter 5: Off-The-Shelf Components ......................................................................................35

Chapter 6: Manufacturing..........................................................................................................38

Chapter 7: Conclusions and Recommendations .........................................................................45

Appendices

Appendix A: Gantt Charts

Figure A-1. Original Gantt Chart .................................................................................47

Figure A-2. Final Gantt Chart ......................................................................................48

Appendix B: Graphs

Figure B-1. Rotor performance using published data for CP. ........................................49

Figure B-2. Predicted power performance utilizing CFD data. .....................................49

Figure B-3. Predicted power performance utilizing CFD data with extrapolated

generator data...........................................................................................50

Appendix C: Engineering Drawings

Figure C-1. McMaster-Carr parts list order ..................................................................51

Figure C-2. Long truss rod...........................................................................................52

Figure C-3. Short truss rod...........................................................................................52

Figure C-4. Truss plate ................................................................................................53

Figure C-5. Short solar panel bracket ...........................................................................53

Figure C-6. Long solar panel bracket ...........................................................................54

Figure C-7. Rotor.........................................................................................................54

Figure C-8. Top and bottom housing sections ..............................................................55

Figure C-9. Middle housing sections............................................................................55

Appendix D: CFD Results

Figure D-1. Sample calculation grid for 15 degree converging angle............................56

Figure D-2. Torque fluctuations for λ=1.0....................................................................56

Figure D-3. CFD results for Cm as a function of time showing velocity independence

of turbulence model..................................................................................57

Table of Contents Self-Sustaining Street Light

3

Figure D-4. CFD results for CP as a function of tip-speed ratio showing velocity

independence of turbulence model ...........................................................57

Figure D-5. CFD results for Cm as a function of tip-speed ratio for various

converging angles ....................................................................................58

Figure D-6. CFD results for CP as a function of tip-speed ratio for various

converging angles ....................................................................................58

Appendix E: Wind Speed Data Analysis

Figure E-1. Histogram for Skytop area anemometer data (11/13/2009 – 4/22/2010).....59

Figure E-2. Histogram for Standart lot anemometer data (11/13/2009 – 4/22/2010) .....59

Appendix F: MATLAB Code ..............................................................................................60

Abstract Self-Sustaining Street Light

4

ABSTRACT

In continuation of the project to design a self-sustaining street light, the main efforts of

this project were focused on the manufacturing and construction of a prototype. This included

acquiring the individual components, designing appropriate interconnections between parts, and

drafting engineering drawings for various parts. In order to complete the prototype in a feasible

and timely manner, the design was scaled down by a factor which would ease the assembly

requirements. Through cooperation of the team members and faculty advisors, the problems

encountered in the design for manufacturing were creatively approached and solved. In the end,

the prototype was assembled and verified to function correctly. Within this report, ethical

implications are also discussed.

Note to the reader: This report serves as a continuation of the companion report presented for fall

semester 2009.

Acknowledgements Self-Sustaining Street Light

5

ACKNOWLEDGEMENTS

Our deepest gratitude goes to everyone who continued to support our project this

semester. We would like to thank Professor Thong Dang and Professor Michael Pelken once

again for their help and weekly advice. We would also like to thank Professor Frederick Carranti,

for his continued support and encouragement. A special thank goes to MAE graduate students

Nhan Huu Phan and Premkumar Siddarth for performing further finite element analysis on our

lightpost to analyze wind speed amplification effect and stress applied on the street light

assembly. We also sincerely thank Debbie Brown and the MAE purchasing department for

taking care of all purchasing orders for our prototype. We express a deep thanks to Ryan Dygert,

for assisting us in the setup of our generator testing, and advice on manufacturing procedure of

lightpost housing. To Jeff Ellison and the SU Physical plant team for helping us install

anemometers and collect data, as well as manufacture components for our prototype. To the

Engineering Machine Shop for constructing components for our prototype and Physics

Department machine shop for assembling the rotor. Finally, special thanks go to Tim Hatlee

(team member from fall 2009), Evan Beckerman, Denis Pradhan, Jakub Walczak, and all other

undergraduate students who took time away from their own projects to assist in the production of

this project.

Chapter 1: Introduction Self-Sustaining Street Light

6

CHAPTER 1: INTRODUCTION

Last semester, our senior design group designed and developed a “Self-sustaining Street

Light” device powered by a vertical-axis wind turbine and solar photovoltaic panels. The rotor

design, energy storage system, and sizing calculations were a few of the aspects analyzed. Our

goal this semester was to assemble a working prototype of the “Self-sustaining Street Light,”

utilizing off-the-shelf components, to further investigate its performance. This semester, our

team improved previous semester’s Computer Aided Design (CAD) drawings to give to machine

shops. In this report, ethical implications for patent pending design of Self-sustaining Street

Light are also taken into account, and discussed.

Our team collaborated with following multi-disciplinary teams: an advertising team,

MAE graduate students, and an industrial design team (COLAB). The advertising team

investigated a target market and came up with an advertising campaign for our product. This

campaign was presented to our team by the end of the semester.

A simulation of wind velocity amplification effect and stress analysis was carried out by

two MAE graduate students. Computational Fluid Dynamics (CFD) was used to analyze the

wind amplification effect and the optimal converging angle for the housing section. ANSYS

software was utilized by a graduate student to perform stress analysis on the street light and pole

assembly. Results obtained were then compared with basic handmade calculations for stress

applied to the pole. In addition, the industrial design group, COLAB, implemented our design to

their plans of an environmentally friendly service station.

In addition to multi-disciplinary collaborations, and manufacturing of working prototype,

our group submitted an abstract and paper to participate in the 7th international conference on

Indoor Air Quality, Ventilation and Energy Conservation in Buildings (IAQVEC) 2010 at

Chapter 1: Introduction Self-Sustaining Street Light

7

Syracuse University to be held August 15-18, 2010. This conference is hosted by Syracuse

University, Syracuse Center of Excellence, National Research Council Canada, and the United

States Environmental Protection Agency (EPA).

Chapter 2: Ethical Implications Self-Sustaining Street Light

8

CHAPTER 2: ETHICAL IMPLICATIONS

This section will seek to explain the important points of this design that deal with more

qualitative issues of the project, such as liability risks, manufacturing methods, environmental

concerns and responsibility, and the handling of intellectual property. From the beginning, the

original concept for the design came from the collaboration of Professors Michael Pelken and

Thong Q. Dang. While the technical aspects of the converging section involved Dr. Dang, the

aesthetic properties of the design originated from Professor Pelken. The original patent filed by

these two is owned by the University. Recently, rights for sole production of this patent has

been extended to a company with whom they will continue collaboration. This exciting

collaboration and sharing of intellectual property will provide a means of funding and further

development of the idea and prototype into a manufactured product. This means that the project

started here by this interdisciplinary team has the potential to eventually expand into a profitable

market, making the research and prototype development a worthwhile investment.

As with any engineering product developed, there is a certain amount of risk of injury

associated; this is why safety factors are usually included and why stress analyses are done.

Hopefully, the stress analysis will provide a useful understanding of the limitations of this

technology and the recommendations will be followed in further development and manufacturing

of the product to ensure that civilians and users will be safe from harm. With many products,

services, and machines, there is a way in which the consumer could be hurt in its use. However,

with this particular product, it is particularly static in that it is not intended to be interactive, and

should not cause harm in this way. In maintenance, the technician will need to interact, and

therefore certain design elements should be included to ensure his safety. This would include a


9

shutoff of electricity from the base, as well as a shutoff for the rotor, to exclude the possibility of

injury from moving parts or from electricity.

When originally considering the possibility to proceed with the production of this project,

the motivation was to provide a solution for lighting in several situations. The idea was to

produce a product that was sustainable, environmentally responsible, and applicable for regions

without established electric infrastructure. As such, utilizing renewable sources of energy was

an ideal solution. By taking advantage of solar and wind resources, this lightpost achieves these

goals; it produces free energy as necessary, produces no greenhouse gases, and can be put into a

place that require lighting without the need for expensive trenching or access to an existing

electric grid. Since this product could be put into place in a variety of locations, the payback

period would vary depending on how far the location is from standard infrastructure.

Nonetheless, it will provide a service that has more value than its monetary component. Without

contributing to the ever-increasing global warming trend, this product will also serve as a

consistent reminder to those who see it in use, to change their individual habits towards a

sustainable and responsible lifestyle.

With the future manufacturing of the product, it is desirable to employ methods which are

sustainable and environmentally friendly, keeping in conjunction with the goals of this project.

This would include manufacturing methods which use machines that produce the least amount of

waste materials. Accordingly, it may be best to use methods of casting to produce the rotor and

housing parts rather than using costly machining methods.

It is possible that in the future the rotor will be made out of a non-metal material which is

lighter and non-reflective, so as to provide for non-distracting, easier rotation. It also may be


10

more cost-effective to use a cheaper material such as a plastic through a type of mold casting.

The casting method would also imply a greater production, and much less material waste.

For greater efficiency of the system, the individual electrical components would be

designed and manufactured by the company, rather than bought off the shelf to closely match the

product’s purposes. The generator and the LED drivers when custom designed would especially

provide great increases in the overall efficiency. It would allow the generator to meet the exact

range for which the rotor will function to avoid losses in production. It would also eliminate the

losses in converting DC to AC to DC for the electricity leaving the batteries to power the LED

lights at the necessary current and voltage.

If the poles would like to be custom built, it is likely that the most cost-effective method

would be to make these within the company as well. This will provide the ability to create a

method of manufacturing the quantity necessary for production. Producing all of these

components and pieces of the product also means that there is no need to ship items – a process

which adds both cost and a carbon footprint to the product.

In conjunction with the responsibility in manufacturing of the product, it is just as

important to consider the end of the product’s life. While the lifespan is currently uncertain, it is

estimated that it should last as long as typical VAWTs with the occasional maintenance as

deemed necessary. This should ensure at least 10-15 years of safe usage, and can be better

determined as the product is further developed and tested for durability and function. The cost

of the life-span will be built into the overhead considered for the cost analysis of the company.

This includes the dismantling of the product at the end of its life, and the recycling and

processing of the materials. The metals and plastics used in the rotor and housing will be

salvaged within the company, melted down for reuse in the manufacturing processes. The


11

batteries will be forwarded to a recycling plant where they will be properly disposed of, rather

than adding to a typical landfill where they could leak heavy metals and corrosive acids into the

earth. The electrical components will be assessed and inspected carefully for damage and

reused where appropriate. It will obviously be more costly to do this processing, but it will most

definitely provide for a more responsible product, also making it a more legitimately green and

marketable technology. In such a way, it will easily end up paying for itself.

As a sustainable solution, this project aims to meet three important criteria that make it

exactly that – sustainable: to develop a product that is environmentally responsible, meets a need

of society, and is financially plausible. In view of these ethical implications, this project meets

the criteria for sustainability, and hopefully as the product continues in development, it will be

financially feasible while meeting this important criteria.

Chapter 3: Stress Analysis Self-Sustaining Street Light

12

CHAPTER 3: STRESS ANALYSIS

1.1 Static Stress Analysis

As the design of the rotor and housing have changed, so has the way that the load will be

distributed upon the pole which supports the weight and other forces involved with the

lightpost’s function. Therefore, it is important to analyze these forces and determine whether a

typical pole would be suitable for this function, or whether it would be prudent to have a custom

built pole that would suit the particular needs of this project.

Figure 3.1. Diagram demonstrating weight distribution for determination of compressive stress.

The first stress that was investigated was the compressive force due to the structure’s

weight. The total weight, estimated at 140 pounds, was divided into the rotor and the housing,

which were estimated to be 45 and 90 pounds respectively. With the weight of the rotor applied

C

D

WRotor

WHousing

Supports

weight of

Rotor

Supports

weight of

Housing

Support

Pole


13

at point C, and the weight of the housing at point D, the total compressive force was found to be

59 KPa, with more detail given below in Table 3.2. These important geometries of the pole and

rotor are displayed in Figure 3.1.

The value of the compressive stress is well within the acceptable loads for both Steel and

Aluminum, whose important material properties are summarized in Table 3.3. The given strain

values due to this compression are also given below.

Measurement Metric English

Units

Rotor Height (Total) 0.889 M 35 in

Rotor Diameter 0.610 M 24 in

Rotor Radius 0.305 M 12 in

Rotor W 200.2 N 45 lbs

Housing W 400.3 N 90 lbs

a 0.076 M 3 in

b 0.076 M 3 in

c 0.266 M 10.5 in

d 0.596 M 23.5 in

x 0.254 M 10 in

Pole Outer Radius 0.076 M 3 in

Pole Inner Radius 0.051 M 2 in

CS Area of Pole 0.010 M^2

J 0.005 M^2

I 2.12E-05 M^2

L (pole height) 7.62 M 25 ft Table 3.1. Important geometrical values of rotor, housing, and pole.

][/)(/ PaAreaWhWrAP +==σ (3.1)

][/ mAEPL Dc δδδ +== (3.2)

Table 3.2. Determination of compressive stress due to weight.

Value Steel Pole Aluminum Pole

σ [Pa] 59255.8 59255.8

δc [m] 0.000753 0.002184

δd [m] 0.001455 0.004223

δtotal [m] 0.002207 0.006408


14

After finding the compressive stress, the torsional stress due to the wind on the rotor was

also investigated. The force due to the wind was estimated by assuming a static situation of the

rotor. The analysis is as follows, and results in a moment around the z-axis of the rotor, with

values that increase with the change in the wind speed. The values of the moment and the stress

are given in Table 3.4 found later in this section.

Figure 3.2. Determination of torsion moment and stress.

heightDiameterPW VY **5.*= (3.3)

][**5. NmRadiusWM YZ = (3.4)

]/[* 2mN

J

cM

Pole

Z=τ (3.5)

)(*5.0;22

ioPolerrJRadiusc −== π (3.6)

In a similar way, the forces and stresses on the pole were found by means of a similar

assumption; a static situation was considered and the rotor was assumed as a flat plate, with the

two rotors parallel to each other, rather than orthogonal. Such an analysis results in an

MZ

[Nm] WY [N]

PV [Pa]

D [m]

H

[m]


15

over-estimation of the forces and, in a way, provides a factor of safety in the analysis. The

forces are then translated to the pole as displayed in Figure 3.3 and Figure 3.4.

Figure 3.3. Dynamic pressure to linear force along axis of rotor and housing.

DiameterPWV

*= (3.7)

heightWWX

*= (3.8)

Figure 3.4. Determination of forces on rotor bearings.

W [N/m]

WX [N]

PV [Pa]

D [m]

H

[m]

W [N/m]

Ax [N]

WX [N/m]

Bx [N]

Bearings

a [m]

b [m]


16

;)(*)2

(*0;;0 hbaAbh

WMWBAFXxBxxxX

++−+==Σ=+=Σ

;)(/)2

(* hbabh

WAxx

+++= (3.8)

xxX

AWB −= (3.9)

There are two bearings which translate the forces from the rotor to the rest of the

structure. In accordance with Figure 3.1, the bearing at point A translates part of the force to the

housing, while the bearing at point B within the pipe below the rotor translates the force to the

coupling that is attached to the pole. These are shown in the diagram, with the eventual result as

shown in the diagram of the pole. The main points of force are given at points C and D, with

moments at the same points due to the translation from the point at which the original force is

applied.

Figure 3.5. Determination of forces on housing collar and on rotor/generator collar.

xxX

AWB −= (3.9)

;)(*;;0 dhaAMDAF XDXXX ++===Σ (3.10)

DX

AX

MD

d

Bearing Pipe

Pole Collar

Generator

BX

CX

c

MC


17

;*;;0 cBMCBF XDXXX ===Σ (3.11)

The bottom of the pole, point G, shows the connection of the pole to the ground, and

shows the point of greatest shear stress and torque. This was determined by the method of

moment and shear diagrams. As a result this is the point that was investigated for the point of

failure of the pole. The value of stress due to the forces and moments at the top of the pole are

indicated in Table 3.4.

;cbdX −−=

XXXXDCGF +==Σ ;0

DCXXGMMxLDLCMM −−−+== )(**;0 (3.8)

StressShearPaI

cMG

−= ][;*

σ (3.9)

)(*25.44

iorrI −= π (3.10)

Figure 3.6. Determination of forces on housing collar and on rotor/generator collar.

It is important to see the values of this stress and find where it reaches the shearing yield

point for a given material. For the Steel investigated – ASTM–A36 – with dimensions of

diameter 6 inches, and thickness of 2 inches, even at high wind speeds of 60 m/s or 134 mph the

shearing stress at the bottom only reaches about 1.8 MPa; this is well below the expected

x

GX MG

L

CX

MC

DX

MD


18

yielding point of Steel due to shearing, which is around 145 MPa. Even the aluminum pole, for

6063-T1, it is equally well below its yielding point (about 97 MPa). It seems that the structure

would fare quite well on a typical pole, of either Aluminum or Steel, at the given height.

Apparently, it seems to require even more of a force to cause the pole to yield, although perhaps

with cyclical loading, this may occur easier. Property values for these materials are shown below

in Table 3.3, while the values of stress as they vary with wind speed are shown in Table 3.4.

Table 3.3. Material properties.

Wind Speed (m/s)

Dynamic Pressure (Pa)

W - (N/m)

Wx - (N)

Mz (Nm)

Torque σ [Pa]

Ax [N]

Bx [N]

Mc (Nm)

Md (Nm)

Mg (Nm)

Gx [N]

Shear σ [Pa]

1 0.6 0.4 0.3 0.03 0.38 0.2 0.2 0.0 0.1 0.1 0.3 460.6664

4 9.8 6.0 5.3 0.41 6.11 2.7 2.7 0.7 2.0 2.1 5.3 7370.663

7 30.1 18.4 16.3 1.24 18.71 8.2 8.2 2.2 6.1 6.3 16.3 22572.65

10 61.5 37.5 33.3 2.54 38.19 16.7 16.7 4.4 12.5 12.8 33.3 46066.64

13 103.9 63.4 56.3 4.29 64.54 28.2 28.2 7.5 21.1 21.7 56.3 77852.62

16 157.4 96.0 85.3 6.50 97.77 42.7 42.7 11.4 31.9 32.9 85.3 117930.6

19 222.0 135.3 120.3 9.17 137.87 60.2 60.2 16.0 45.0 46.4 120.3 166300.6

22 297.7 181.5 161.3 12.29 184.85 80.7 80.7 21.5 60.4 62.2 161.3 222962.5

25 384.4 234.3 208.3 15.87 238.70 104.2 104.2 27.7 78.0 80.3 208.3 287916.5

28 482.2 293.9 261.3 19.91 299.43 130.6 130.6 34.8 97.8 100.7 261.3 361162.5

31 591.0 360.3 320.3 24.41 367.03 160.1 160.1 42.6 119.9 123.5 320.3 442700.4

34 710.9 433.4 385.3 29.36 441.50 192.6 192.6 51.3 144.2 148.5 385.3 532530.4

37 841.9 513.2 456.3 34.77 522.85 228.1 228.1 60.7 170.8 175.9 456.3 630652.3

40 984.0 599.8 533.3 40.63 611.07 266.6 266.6 70.9 199.6 205.5 533.3 737066.3

43 1137.1 693.2 616.3 46.96 706.17 308.1 308.1 82.0 230.7 237.5 616.3 851772.2

46 1301.3 793.3 705.2 53.74 808.15 352.6 352.6 93.8 264.0 271.8 705.2 974770.1

49 1476.6 900.1 800.2 60.98 916.99 400.1 400.1 106.5 299.6 308.4 800.2 1106060

52 1663.0 1013.7 901.2 68.67 1032.72 450.6 450.6 119.9 337.4 347.4 901.2 1245642

55 1860.4 1134.1 1008.2 76.82 1155.31 504.1 504.1 134.1 377.4 388.6 1008.2 1393516

58 2068.9 1261.2 1121.2 85.43 1284.78 560.6 560.6 149.2 419.7 432.1 1121.2 1549682

61 2288.4 1395.0 1240.2 94.50 1421.13 620.1 620.1 165.0 464.2 478.0 1240.2 1714140

Table 3.4. Resultant forces and stresses.

Property Steel Aluminum Unit

E (modulus of Elasticity) 200 68.9 Gpa

Compressive Yield 152 152 MPa

Ultimate Tensile Stress 400 400 MPa

Yield Shear 145 96.5 MPa


19

1.2 Conclusions and Recommendations

Considering these findings, a few recommendations can be made. While this analysis

makes a few assumptions that results in a lower wind speed for the material’s yield point, this

can be seen as an appropriate safety factor for the design point. Therefore, these values should

be interpreted appropriately. If there is still worry regarding the strength of the pole, then a

custom pole could be built from a stronger Steel alloy, with a wider radius, or with a thicker

cross section. Such a procedure would work to strengthen the pole and would likely result in a

smaller value of strain at the top of the pole.

Another method of reinforced restraint would involve cables that anchor the housing of

the lightpost, and help to secure it from swaying in the wind while dispersing the stress from the

pole. If these two procedures were used in conjunction, it would ensure that the prototype is safe

for use while in testing. Additionally, the lightpost could be secured by wrapping some sort of

netting or webbing to the housing, and attached to the anchor cables, catching the housing in case

it falls from the pole. This is very important, because if the proper precautions are not followed,

there is the possibility that the lightpost could fall due to failure and possibly injure a bystander.

In order to avoid cyclical loading on the structure from high winds, a better method of

safety precautions could be implemented in future design of the lightpost when manufactured for

actual production. The best thing to do would be to include a control system that senses when

the wind speed is at a point that is too great for energy harvesting, and would automatically

disconnect the shaft from the rotor to the generator. This would allow for a free-spinning

situation of the rotor. This is different than the safety procedure for Horizontal-Axis Wind

Turbines in which they move to stall position in high winds and storms, but should provide an

appropriate. The free spinning situation would allow the wind to pass through the structure


20

without translating its full force to the structure and the pole. This would, however, result in a

high rpm of the structure, and it should be investigated as to whether it could eventually reach a

rotational speed that creates resonance for the structure. While this implementation is beyond

the scope of this project, it could easily be developed for future manufacturing purposes.

The way it currently stands, when the battery is full or the wind speed is too great, the

generator simply applies a load so great that the rotor is unable to spin, resulting in a static

situation as was investigated above. That is why this additional recommendation would provide

better safety than the current prototype. Another possible safety recommendation involves a

method that would result in a stall position, while dispersing the force of the wind on the

structure. Such a method could involve a similar control system that would cover or enclose the

rotor within a cylindrical or other smooth surface that would move the air around the housing

and not move the rotor – it could be a system that has moving doors. However, this has several

negatives. Notably, it introduces additional moving parts and also adds additional weight to the

structure. The additional weight could be easily supported, as the compression forces were well

within a reasonable range, but the additional moving parts means that there would be a greater

need for maintenance of the structure.

1.3 Dynamic/Harmonic Analysis

In view of the function of this lightpost, it was desirable to do a harmonic analysis in

order to determine the safety of functioning within the desired design range of wind speeds. In

cooperation with a graduate student, Siddharth Premkumar, the program ANSYS was utilized to

analyze a simple model of the structure, using the diagrams below which estimate the function of

the structure.


21

Figure 3.7. Simplified lightpost housing.

Figure 3.8. Simplified lightpost rotor.

17.5”

41.25” 6.25”

24”

0.125”

50.2”

42”

4.8”

2” .5”

18”

24

6


22

The first attempt was to build the rotor and housing separately in ANSYS, and then

analyze them together in a dynamic analysis. However, this method was unable to produce

results, because the number of nodes produced from the meshing exceeded the allowable number

for analysis in the program. The meshing was not able to be simplified and still keep the desired

shape of the structure. Instead, a second method was to analyze the parts separately, to find the

critical wind speed and corresponding rotational speed for the rotor, and then find the associated

stresses on the housing. Unfortunately, because of the low rotational speed and the complexity

of the model, it was explained that the critical speed was not able to be found from the analysis.

It was qualitatively explained by Siddharth that the structure would not reach an

rotational speed at which it would fail, because of the design range. Due to time constraints and

responsibilities outside of this project, Siddharth was unable to provide additional services in

pursuit of quantitative dynamic analysis results. Therefore, the dynamic analysis of this lightpost

structure will need to be investigated further at a later time in order to determine the range within

which the lightpost can safely operate.

Chapter 4: Wind Turbine Self-Sustaining Street Light

23

CHAPTER 4: WIND TURBINE

4.1 Generator Selection:

Significant research was conducted last semester to determine an appropriate permanent

magnet generator/alternator (PMG/PMA – both terms will be used interchangeably in this report)

for the vertical-axis wind turbine (VAWT) in this project. Those investigations showed that the

Ginlong GL-PMG-500A PMG was an appropriate match for this light source design. Upon

beginning preparations for prototyping, however, it was determined that the current size was far

too large to produce effectively, inexpensively, and quickly – all necessities for this specific

prototype. As a result, the power production capabilities of the rotor were cut in half; this led to

a reduction in all sizing parameters by a factor of 2 . Understandably, this would, in turn, also

effectively reduce the light-on time of the LED luminaire by approximately a factor of two.

Nonetheless, this handicap was accepted by the project team after determining that the primary

goal of the prototyping phase is to prove the efficacy of the design – a proof of concept. After

deciding that a size reduction was appropriate for the prototype production, though, a new

generator was also necessary.

New investigations into available PMG products showed that few, if any, off-the-shelf

products were available to appropriately match the rotor size and, accordingly, power output of

this light source. Specifically, all products from last semester were readdressed for plausible

inclusion, including: Ginlong GL-PMG-500A, Ginlong GL-PMG-1000, WindBlue DC-520,

WindBlue DC-540. In addition, new products were also considered: Wind Turbine

Technologies’ Cat 3 and Cat 4 PMAs. However, new power curves exposed the unfortunate fact

that the same issue from last semester – the requirement of low power at relatively high


24

rotational speeds – was amplified by the prototype size reduction. This observation can be

explained by the relation for tip-speed ratio, λ:

U

d

2

ωλ = (4.1)

defined for diameter, d, rotational speed, ω, and effective wind speed, U. From (4.1), it can be

deduced that, for a given wind speed, the tip-speed ratio (the independent variable when

comparing non-dimensional relations to power coefficient, CP) shows an inverse relationship

between diameter and rotational speed. As a result, the reduction in diameter discussed earlier

appropriately led to an increase in rotational speed by the same factor of 2 at all wind speeds.

From this increase in rotational speed, none of the generators of interest showed adequate rotor

matching behavior.

Nonetheless, the search for a generator was not abandoned and a product was discovered

from a domestic distributor, Georgia Generator. Specifically, the product was described as a

PMA rated for 100W of output at a rotational speed of 450 rev/min. Unfortunately, though, the

industry-standard power curves showing performance at all rotational speeds provided by most

manufacturers and distributors were not available for this product. Despite this fact, an

approximation for power characteristics according to the observations of all other generators

listed above showed that this Georgia Generator product would, in fact, match nearly perfectly

with this rotor design if it behaved in the same manner as other generators. As the only true

option, the decision was made to purchase the product and physically test its performance at

other rotational speeds.


25

4.2 Generator Testing:

In order to determine the power output characteristics of the purchased generator, the

generator needed to be tested. Various methods for accomplishing this goal were considered,

including the measurement of voltage and output from the generator explicitly. However, the

three-phase AC output of the generator creates considerable issues when considering such an

undertaking and, as a result, the experimental setup in Figure 4.1 was designed implementing a

torque meter (TM) and a DC drive motor. A variable frequency drive (VFD) was connected to

the motor, allowing for rotational speed control output from the motor. The three-phase AC

output from the generator was connected to the appropriate inputs on the charge controller

purchased for use with the prototype and, finally, a 12V car battery was connected to the charge

controller as the sink for the power.

Figure 4.1. Generator testing rig setup.

The benefit of such an arrangement over other recommendations to explicitly measure

voltage and current output from the generator is the simulation of the true system load

experienced by the generator during operation. Therefore, the data created by this test was the

most accurate representation of generator performance with the prototype model; in fact, the

information achieved from this test is more reliable than performance data received from

manufacturers for the other generators which were considered.

DC Motor

TM

Gen

erato

r

VFD


26

By adjusting the VFD, the output from the motor changed, leading to a change in system

torque. By reading real-time data from the torque meter and recording rotational speed of the

system shaft via a laser tachometer, the performance data in Table 4.1 was created.

Unfortunately, though, the load on the system would not permit the motor to operate at rotational

speeds higher than approximately 200 rev/min. As a control for comparison, the same test was

performed while having the battery disconnected. These results are reported in Table 4.2.

VFD Setting N (rev/min) Torque (N-m)

10.0 95 0.20

14.5 115 0.55

18.0 133 0.99

19.0 160 1.46

21.0 191 1.87

VFD Setting N (rev/min) Torque (N-m)

12.0 100 0.18

14.0 165 0.20

15.0 186 0.21

16.0 242 0.22

18.0 322 0.24

Table 4.1. Loaded generator performance. Table 4.2. Unloaded generator performance

From Table 4.1, a torque value of 0.2 N-m was the point at which the generator began to

turn freely – indicative of the startup torque. This observation is appropriately congruent with

the published value for the product of 2.0≤ N-m, making the generator a good match by

creating a low cut-in speed for our wind turbine. Similar to the process for plotting the

performance for the other generators, these results were plotted against the power output from

the rotor at various wind speeds.

4.3 Power Analyses:

By plotting the predicted power output of the rotor with the performance characteristics

of the generator, operating conditions can be determined for varying wind speeds. Last semester,

generator curves were plotted against rotor performance assuming CP characteristics observed by

Modi and Fernando [1]. Due to the similarity of the rotor cross section for this project to the one


27

used by Modi and Fernando, the same assumptions were made at this point in the design stage

and these data were plotted against the generator performance from Table 4.1 – see Figure 4.2.

Figure 4.2. Rotor performance using published data for CP.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

λ

Cp

No Surrounding Disk 0 Degree Disk 5 Degree Disk 15 Degree Disk 22 Degree Disk 30 Degree Disk

Figure 4.3. CFD results for CP as a function of tip-speed ratio.


28

Alternatively, though, further advancements were made utilizing computational fluid

dynamics (CFD), incorporating the converging effect from the outer housing – one of the

fundamental aspects of this design. Figure 4.3 presents the results of these investigations,

showing the relative increases which can be achieved for corresponding increases in the angle of

convergence for the housing.

By processing the results from Figure 4.3 corresponding to the prototype housing angle

of 15 degrees, the performance data presented in Figure 4.4 is acquired. Comparing the results

of Figure 4.4 with those of Figure 4.2, the beneficial contributions of the converging section are

apparent. It should be noted, however, that the data of Figure 4.4 are accounting for a traditional

Savonius rotor cross-section, whereas the data from Figure 4.2 implement a Bach-type cross

section similar to the one implemented for this design. Based on the comparison results between

Savonius- and Bach-type rotor cross sections, it is assumed that the data of Figure 4.4 will

improve further yet, particularly as Reynolds number increases with higher wind speeds. This

effect can likely be attributed to the lift effects which contribute with the Bach-type cross

section.

From Figure 4.4, however, it can be seen that the predicted rotor performance at 6.76 m/s

extends beyond the observed generator performance data. As a result, Figure 4.5 presents the

predicted rotor performance (utilizing CFD data for CP) across the entire range of wind speeds

observed at Skytop. It should be noted that the generator performance beyond the five points

recorded in Table 4.1 is extrapolated by a curve-fit of known data.


29

Figure 4.4. Predicted power performance utilizing CFD data.

Figure 4.5. Predicted power performance utilizing CFD data

with extrapolated generator data.


30

The data of Figures 4.2-4.5 tell an important story. Notably, this information can provide

a good estimation of the cut-in speed of the VAWT. Combining the cut-in information observed

during the generator test with the data in these figures and similar reproductions, it can be

determined that the cut-in speed of the VAWT is approximately 2.50 m/s – a number that is

extremely encouraging for the productivity of this system.

However, further analysis can lead to a calculation of rated power for the wind turbine

system. As with last semester, two anemometers have been collecting and continue to collect

wind speed data for this project. Continued studies of these data through histogram plots has

given an even greater idea of wind speed spreads at the two locations of interest – Standart

parking lot and Skytop parking lot. These histogram results are presented in Figure 4.6 and

Figure 4.7 corresponding to data recorded between November 12, 2009 and April 22, 2010. On

April 22, 2010, the anemometer recording data at Standart parking lot was moved to a location

near the Syracuse University Warehouse in downtown Syracuse, NY to measure wind speeds as

part a potential coordination with the Connective Corridor project.


31

Figure 4.6. Histogram plot of wind data at Standart parking lot.

Figure 4.7. Histogram plot of wind data at Skytop parking lot.


32

The results from these histograms can then be applied to calculate the rated power:

∑

∑=

total

i

ii

n

Pn

RP (4.2)

for n time intervals producing power, P, for all velocity bins, i. The important observation in

(4.2) comes from the fact that the denominator takes into account the total number of time

intervals, ntotal, thus incorporating the amount of time during which the wind speed is lower than

the cut-in speed of the turbine. Understandably, in the limit, ∞→i , this equation becomes

increasingly accurate.

The operating points utilized for calculating rated power of this wind turbine system from

the Skytop lot histogram of Figure 4.7 are shown in Table 4.3 (calculated over the same data

collection period presented in Figure 4.7). The anemometers used for this project average wind

speeds over 10-minute intervals, which provides the time intervals for calculating n in (4.2).

Assuming the cut-in speed discussed earlier of 2.50 m/s, (4.2) was applied for these data.

Although the anemometers observed wind speeds as high as 11 m/s, the extrapolated data

prediction for generator performance shows the generator staying within its 150W peak limit for

wind speeds up to 9.75 m/s. As a result, the power output of the rotor at any wind speed higher

than 9.75 m/s can be essentially considered to be zero. Ultimately, this calculation process yields

a resulting rated power value of 6.541 Watts.


33

Wind Speed (m/s) RPM Power (W) Torque (N-m) # of 10-min intervals W-(10 min)

2.50 96.0 2.3 0.23 1371 3153.3

2.75 100.0 3.1 0.30 902 2796.2

3.00 103.5 4.0 0.37 1287 5148.0

3.25 108.5 5.1 0.45 847 4319.7

3.50 113.5 6.3 0.53 1163 7326.9

3.75 117.5 7.7 0.63 722 5559.4

4.00 121.5 9.2 0.72 992 9126.4

4.25 126.0 11.0 0.83 610 6710.0

4.50 130.8 12.8 0.93 811 10380.8

4.75 136.3 15.1 1.06 431 6508.1

5.00 142.5 17.5 1.17 580 10150.0

5.25 149.5 20.3 1.30 321 6516.3

5.50 157.0 23.3 1.42 409 9529.7

5.75 165.8 26.8 1.54 219 5869.2

6.00 175.0 30.7 1.68 288 8841.6

6.25 185.0 34.8 1.80 110 3828.0

6.50 195.0 39.2 1.92 175 6860.0

6.75 204.0 44.2 2.07 86 3801.2

7.00 215.0 49.5 2.20 124 6138.0

7.25 225.3 55.2 2.34 58 3201.6

7.50 236.0 61.2 2.48 68 4161.6

7.75 248.0 68.0 2.62 40 2720.0

8.00 259.5 75.0 2.76 61 4575.0

8.25 272.0 82.6 2.90 26 2147.6

8.50 284.3 90.7 3.05 25 2267.5

8.75 297.0 99.2 3.19 23 2281.6

9.00 309.3 107.9 3.33 27 2913.3

9.25 322.1 117.0 3.47 8 936.0

9.50 335.0 126.7 3.61 15 1900.5

9.75 348.2 136.8 3.75 8 1094.4

not running 11241 0.0

sum 23048 150761.9

Rated Power (W) 6.541

Table 4.3. Rated power data.

4.4 Rotor Production:

In partnership with the Syracuse University Physics Machine Shop staff, the two VAWTs

for this project were produced. A cross-sectional drawing of the prototype rotor is shown in

Figure 4.8 with all dimensions in inches. Parallel with the design considerations discussed last

semester, the cross-section maintains a Bach-type shape, leading to increases in power output

and a slight shift of peak power output toward a higher value of tip-speed ratio, as discussed


34

earlier. It is assumed, though, that the effect of shifting the power curves up and to the right will

replace them near peak operating points on the generator curve of Figure 4.4 and Figure 4.5.

The material used to produce the generator was selected to be aluminum alloy 5052,

according to recommendations from the machine shop staff. The blades were produced by a

rolling technique with 0.080” thick sheets and the circular endplates were cut by a computer-

controlled mill from 0.125” thick sheets. Holes were appropriately added (not shown in Figure

4.8) for the middle connection of the two rotor sections and the attachment of shaft collars.

Finally the two blade sections and the two end plates were mechanically fastened for use.

Figure 4.8. Prototype rotor cross-sectional drawing.

References:

[1] Modi, V.J. and M.S.U.K Fernando. “On the performance of the Savonius wind turbine.”

Journal of Solar Energy Engineering. 111 (1989): 71-81.

135º

R4.00

0.080

Chapter 5: Off-The-Shelf Components Self-Sustaining Street Light

35

CHAPTER 5: OFF-THE-SHELF COMPONENTS

For the purpose of the prototype of the design, it was easiest to order some key

components from vendors as opposed to constructing or assembling these components ourselves.

These components consisted of the charge controller, the solar panel, the LED streetlight, the

batteries and the generator.

5.1 Charge Controller

The charge controller is a Windmax Wind Turbine/Wind Generator Charge Controller.

The charge controller was purchased from Applied Magnets in Plano, Texas. The charge

controller was purchased last semester, and arrived timely and with no problems. The charge

controller is used to connect and safely control input and output of electricity to every electrical

component of the light post.

5.2 Solar Panel

The solar panel is a Kyocera KC65T, 65 W 12 V Nominal Solar Panel. This solar panel

was purchased from Affordable Solar Group, LLC in Albuquerque, New Mexico. The solar

panel arrived timely and with no problems, ready to be assembled into our design. The only

assembly required with the solar panel is to hook up wires from the solar panel down to the

charge controller. The solar panel will be held up by small rectangular aluminum rods attached to

the truss of the housing.


36

5.3 LED streetlight

The LED streetlight is a Beta LED streetlight, type 4, 30 LED light fixture. This

streetlight was purchased through a local vendor, Ferrini-Konarski Associates in Baldwinsville,

NY. The vendor had attended a weekly meeting of our group in early October 2009, and brought

a light fixture that was similar to the light we were interested in. After deciding the exact light

we would need, we ordered the light from him in early February. Unfortunately, the vendor did

not seem to be very helpful in ordering the light and seemed a little confused on the process of

ordering through the University. This caused a significant amount of problems and the light was

in the process of being ordered until late April. Debbie Brown in the Mechanical and Aerospace

office was very helpful in this process and was persistent in trying to get the light. The team

contacted the vendor company and contacted the parent company to try to get the situation sorted

out, and the team was about to give up and find a new plan for the light when the light arrived.

The light is very simple to assemble into the design. The light only needs a power source

which will be taken from the batteries connected to the charge controller. The light fixture can be

placed outside of the assembly, or could be directly integrated into the housing of the light post.

The LED light ordered has three panels that contain 10 LEDs that can be easily removed from

the light housing and arranged in any pattern in our own housing. The wires and driver are easily

removed from the housing allowing the placement into our design to be simple.

Figure 5.1. LED light panel of 10 LEDs (three of these total).


37

Figure 5.2. Entire light fixture.

5.4 Batteries

The batteries for this design are picked out, but not yet ordered. The batteries are two 6

volt Sun Xtender deep cycle cell batteries. The reason these batteries have not been purchased

yet is to wait until the light post is fully assembled, when the batteries are completely necessary.

Luckily, the Link Hall Machine shop has been helpful in providing a 12 volt car battery to our

group when performing tests on the generator, and when testing our light post before completion

of the project.

Chapter 6: Manufacturing Self-Sustaining Street Light

38

CHAPTER 6: MANUFACTURING

The wind turbine structure was designed so that our group would be able to assemble it

fairly easily once all of the parts were ordered. Using SolidWorks helped with dimensioning and

the order of assembly for the entire structure. First, the drawings for the disks were sent to a local

metal spinning company by the end of March so that the company would have enough time to

make them. With that order in, a work order was filled out for the physics machine shop to build

three individual rotors. Two rotors would be used for our prototype as shown before (two rotors

offset 90 degrees and approximately three feet tall), while the third rotor will be used in future

wind tunnel testing. The three struts with a T-cross section needed to be rolled and this was given

to the Physical Plant machine shop. Those three components were the main pieces of the puzzle

that needed to be outsourced and could not be made in the engineering machine shop. The rest of

the parts were done with help from the engineering machine shop.

The physics building machine shop finished our three rotors and they came out very

nicely. The team wanted to have the two plates welded to the top and bottom of each rotor but

the aluminum was too thin and it was easier to fasten each rotor together with screws and an

adapter, as seen below in Figure 6.1 and Figure 6.2.

Figure 6.1. Wind tunnel test rotor. Figure 6.2. Mechanical fastener for rotor blades.


39

Once the rotor was finished, it could be put together in the orientation for which it was

designed. Aluminum pipe was used for the spacing between the two rotors and ran four seven-

inch screws through the pipe to hold the rotors together. This also made the pipe spacers

essentially a solid rod connecting the rotors. The collars were attached to the top and bottom of

the rotors as well – see Figure 6.3 and Figure 6.4.

Figure 6.3. Dual rotors with spacers and one shaft collar. Figure 6.4. Close-up of rotor shaft collar.

Figure 6.5. Generator shaft enclosure with pressed thrust bearing and

attached rotor shaft coupling.


40

With the two rotors connected the base of the turbine could be assembled with bearings,

bearing fittings, shafts and shaft collars. The bearing was pressed into an aluminum pipe with a

fitting which was made in the LCS machine shop – see Figure 6.5.

The bearing is a tapered thrust bearing that is rated up to 1000 pounds of axial loading,

which is well over the weight of the turbine. The pipe was screwed into a flange that rests on top

of the generator. The mounting screws for the generator are extended through the flange to the

mounting base, which is a four foot square piece of plywood that the entire structure sits on – see

Figure 6.6. With the rotor completely assembled and the base with the generator assembled, the

two components were connected via the shaft and bottom collar on the rotor – see Figure 6.7.

Figure 6.6. Base plate with generator assembly attached. Figure 6.7. Dual rotors attached to base plate.

Once the rotor was assembled, the outer housing could be assembled and be slid over the

rotor. The funneling disks were made by Hy-Grade Metal in Syracuse, NY and they came out

beautifully. The middle housing disks were each 42 inches in diameter and the top and bottom

disks were 38 inches in diameter. This change in diameters adds an aesthetically pleasing

curvature to the entire structure. One of the spun disks is shown in Figure 6.8.


41

There are 3 curved pieces of steel that line the inner radius of each spun disk to stability.

These rings were cut in the LCS machine shop and they were clamped inside of each ring so that

holes could be drilled through both the spun parts and the rings, so they can attach to the struts.

Once the holes were drilled, screws were placed in them to hold it in place – see Figure 6.9.

Figure 6.8. Spun disks (middle disks). Figure 6.9. Bottom spun disk with support rings temporarily bolted on.

With the metal disks and inner rings all drilled and ready to go, the struts could be

attached. The struts had to be rolled in the Physical Plant machine shop so they could attach to

all of the disks. The struts were bolted to the outside of each disk with the inner race of rings on

all of the disks – see Figure 6.10.

While attaching the struts to the disks and inner rings, the top and bottom trusses were

assembled so they could be attached to the top and bottom housing pieces. On the top, the truss is

used to hold the top bearing and also hold the solar panel. On the bottom the truss would be able

to hold the light fixture that would go onto an actual structure. For the presentation, though, the

light will be held on a small five foot PVC pole and the truss will be there for stability. Both of

the trusses add stability to the entire structure – see Figure 6.10 and Figure 6.11.


42

Figure 6.10. Top housing attached to top truss and struts. Figure 6.11. Bottom truss assembly.

Figure 6.12. Full assembly of rotor and housing attached to wooden base.

Before the rotor was attached to the base, the bottom truss was slid over the base because

the middle part of the truss would not be able to slide over the rotor. With the entire housing

assembled, it was slid over the rotor and rests at the right height with the assistance of some

wooden blocks secured to the wooden base – see Figure 6.12.


43

With the housing secure around the rotor, the alignment of everything was checked and

any necessary adjustments were made. Then, the top bearing was placed on the top shaft and the

corresponding mounting holes were marked on the top bearing plate of the top truss. The bearing

was then attached to the plate and secured to the rotor. This secured the rotor from the top and

the bottom and completed the wind turbine structure.

Figure 6.13. Solar panel attached to full assembly. Figure 6.14. LED luminaire attached to pole.

After everything was secure the solar panel was mounted to the top of the truss with the

solar panel brackets that were made in the LCS machine shop – see Figure 6.13.

As stated before, the light was not mounted to the bottom of the turbine housing for our

presentation. Instead, we made a pole out of PVC pipe and mounted the light to the top of that

with a switch so we can turn the light on/off as we pleased. This pole was attached to a piece of


44

wood and a secured so that it would not fall over. The light will also be angled so that it is

perpendicular to the solar panel and would power the solar panel when on – see Figure 6.14.

All of the electrical components were mounted to the wood base and the wires were all

attached as necessary. The final structure came out very well and the group worked very hard

assembling everything. Everyone working with the team was very pleased with the final product

and is excited to see how it performs and consider some of the improvements that can be done to

create a better product.

Chapter 7: Conclusions and Recommendations Self-Sustaining Street Light

45

CHAPTER 7: CONCLUSIONS AND RECOMMENDATIONS

As a result of this senior design project, our group successfully developed an off-the-grid

lighting solution that will be powered by a vertical axis wind turbine and a solar PV panel, to

power a LED streetlight. This semester, our group built a working prototype for our “Self-

sustaining Street Light”. The size of the product considered last semester was decreased by ½,

with a swept area of 0.545m² and rated power output of 7 Watts in the Syracuse area. However,

the PV solar panel size did not change. Material chosen to build prototype was mostly aluminum,

for its light weight compared to steel. Components that need to resist more strength were made

of steel. A different generator from last semester was obtained and tested to ensure its

performance would match our rotor size.

Finite element analysis and computational fluid dynamics simulations showed that a

converging angle of 15 degrees would provide sufficient improvements upon power output,

although larger converging angles could likely be implemented; future wind tunnel testing will

provide more qualitative results. Stress analysis showed that the device is indeed safe enough to

resist high wind speeds of even 60 m/s, with a maximum shear stress of 1.65 MPa. In future

testing of the prototype, anchor lines and safety net around housing are recommended. In

production, it is recommended to adapt a freely rotating mechanism for safety reasons.

The following is the final cost of our working prototype. Understandably, estimated

production cost of the streetlight should be significantly lower due to the fact that most

components utilized for this working prototype are off-the-shelf components.

Chapter 7: Conclusions and Recommendations Self-Sustaining Street Light

46

Bill of Materials:

Appendix A: Gantt Charts Self-Sustaining Street Light

47

APPENDIX A: GANTT CHARTS

Figure A-1. Original Gantt Chart.

Appendix A: Gantt Charts Self-Sustaining Street Light

48

Figure D-2. Final Gantt Chart.

Appendix B: Graphs Self-Sustaining Street Light

49

APPENDIX B: GRAPHS

Figure B-1. Rotor performance using published data for CP.

Figure B-2. Predicted power performance utilizing CFD data.

Appendix B: Graphs Self-Sustaining Street Light

50

Figure B-3. Predicted power performance utilizing CFD data with extrapolated generator data.

Appendix C: Engineering Drawings Self-Sustaining Street Light

51

APPENDIX C: ENGINEERING DRAWINGS

Figure C-1. McMaster-Carr parts list order.


52

Figure C-2. Long truss rod.

Figure C-3. Short truss rod.


53

Figure C-4. Truss plate.

Figure C-5. Short solar panel bracket.


54

Figure C-6. Long solar panel bracket.

Figure C-7. Rotor.


55

Figure C-8. Top and bottom housing sections.

Figure C-9. Middle housing sections.

Appendix D: CFD Results Self-Sustaining Street Light

56

APPENDIX D: CFD RESULTS

(All CFD results compliments of Nhan Huu Phan)

Figure D-1. Sample calculation grid for 15 degree converging angle.

Figure D-2. Torque fluctuations for λ=1.0.


57

Figure D-3. CFD results for Cm as a function of time showing velocity independence of turbulence model.

Figure D-4. CFD results for CP as a function of tip-speed ratio showing velocity independence of turbulence model.


58

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

λ

Cm


Figure D-5. CFD results for Cm as a function of tip-speed ratio for various converging angles.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

λ

Cp


Figure D-6. CFD results for CP as a function of tip-speed ratio for various converging angles.

Appendix E: Wind Speed Data Self-Sustaining Street Light

59

APPENDIX E: WIND SPEED DATA

Figure E-1. Histogram for Skytop area anemometer data (11/13/2009 – 4/22/2010).

Figure E-2. Histogram for Standart lot anemometer data (11/13/2009 – 4/22/2010).

Appendix F: MATLAB Code Self-Sustaining Street Light

60

APPENDIX F: MATLAB Code

%% MEE 472 Lightpost Project

%% Reid Berdanier

%% Rotor-Generator Matching

clc

clear all

close all

%% Generator Performance %%

N_gen = [95, 115, 133, 160, 191]; %tested rpm of generator (rpm)

T_gen = [0.2, 0.55, 0.99, 1.46, 1.87]; %tested torque of generator at

N_gen rpm (N-m)

P_gen = N_gen .* T_gen .* 2 .* pi ./ 60; %power generated from N_gen &

T_gen (W)

N_gen(6:13) = [200, 225, 250, 275, 300, 325, 350, 360];

P_gen(6:13) = [42.085, 54.9825, 69.13, 84.5275, 101.175, 119.0725, 138.22,

146.229]; %extrapolated power from curve fit of N_gen and T_gen data

%% Designed Rotor Performance %%

lambda = [0.2:0.1:1.4];

Dft = 24.5/12; %rotor diameter, ft

Dm = Dft*12/39.37; %rotor diameter, m

Hft = 17.25/12; %rotor height, ft

Hm = Hft*12/39.37; %rotor height, m

rho = 1.204; %air density at 20 deg C, kg/m^3

Cp = [0.124; 0.183; 0.24; 0.29; 0.333; 0.371; 0.4; 0.423; 0.44; 0.451;

0.4488; 0.4368; 0.42];

U = [4.40 5.70 6.76 7.5 8.5 9.5]; %wind speed, m/s

for j=(1:1:length(U))

omega(:,j) = 2 .* lambda .* U(j) ./ Dm;

end

N = omega .* 60 ./ (2.*pi);


R_power(:,j) = Cp(:,1) .* (1/2) .* rho .* 2 .* (Dm .* Hm) .* U(j) .^3;

%calculated rotor output power, W

end

%% Plot

figure(1);

hold on

plot(N(:,1), R_power(:,1), '-b', 'Linewidth', 1.5)

plot(N(:,2), R_power(:,2), '-g', 'Linewidth', 1.5)

plot(N(:,3), R_power(:,3), '-c', 'Linewidth', 1.5)

plot(N(:,4), R_power(:,4), '-m', 'Linewidth', 1.5)

plot(N(:,5), R_power(:,5), '-y', 'Linewidth', 1.5)

plot(N(:,6), R_power(:,6), '-r', 'Linewidth', 1.5)

plot(N_gen(1:5), P_gen(1:5),'sk-', 'Markersize', 9,

'MarkerEdgeColor','k', 'MarkerFaceColor','y')

plot(N_gen(5:13), P_gen(5:13), 'k:', 'Linewidth', 1.5)

title({['Power vs. Rotational Speed'];['15 Degrees']})

xlabel('Rotational Speed (rpm)')


61

ylabel('Power (W)')

grid on

legend('U = 4.40 m/s','U = 5.70 m/s','U = 6.76 m/s','U = 7.50 m/s', 'U =

8.50 m/s', 'U = 9.50 m/s',...

'Generator','Extrapolated Generator','Location','Northwest')

% axis([50,300,0,40])

hold off

figure(2);

hold on


plot(N(:,j), R_power(:,j),'Linewidth',1.5)

end

plot(N_gen, P_gen,'k-','Linewidth',1.5)

title({['Power vs. Rotational Speed'];['15 Degrees']})

xlabel('Rotational Speed (rpm)')

ylabel('Power (W)')

grid on

% legend('U = 4.40 m/s','U = 5.70 m/s','U = 6.76

m/s','Generator','Location','Northwest')

hold off


62

%% MEE 472 Lightpost Project

%% Reid Berdanier

%% Skytop Wind Speed Analysis

clc

clear all

close all

%% Data Input %%

data = load('C:\Documents and Settings\Reid Berdanier\My Documents\8 - 10 SU

Spring\MEE 472\Anemometer Data\Skytop_Done_20100202_fix.txt');

index= data(:,1); %original data point

indices

V_ms = data(:,2); %wind speed, m/s

V_avg = mean(V_ms); %average wind speed, m/s

points= length(index);

hours = points/6;

hours_round = round(hours);

%% Histogram %%

bins1 = [0:1:11];

n1 = hist(V_ms, bins1);

figure(1);

bar(bins1, n1)

title({['Histogram of Wind Speed at Skytop'];['Average Wind Speed =

',num2str(V_avg),' m/s']})

xlabel('Wind Speed, (m/s)')

ylabel({['Frequency, (# of 10-minute intervals over

',num2str(hours_round),' hours)']})

axis([-1,12,0,3500])

bins2 = [0:0.1:11];


figure(2);

bar(bins2, n2)

title({['Histogram of Wind Speed at Skytop'];['Average Wind Speed =





axis([-0.1,11.1,0,400])

bins3 = [0:0.25:11];


figure(3);

bar(bins3, n3)

title({['Histogram of Wind Speed at Skytop Lot'];['Average Wind Speed =





axis([-0.25,11.5,0,2500])

Self-Sustaining Street Light Final Report Part B

Documents

engineering drawings

appendix c

graphs figure b

gantt charts figure

appendix b

published data

rotor performance

short truss rod