Two docs copied from USDA server-
ACCESSION NO: 1006331 SUBFILE: CRIS
PROJ NO: OREW-2015-00340 AGENCY: NIFA OREK
PROJ TYPE: SMALL BUSINESS GRANT PROJ STATUS: TERMINATED
CONTRACT/GRANT/AGREEMENT NO: 2015-33610-23555 PROPOSAL NO: 2015-00340
START: 15 JUN 2015 TERM: 14 FEB 2016
GRANT AMT: $100,000 GRANT YR: 2015
AWARD TOTAL: $100,000
INITIAL AWARD YEAR: 2015
INVESTIGATOR: Schaefer, D. B.
PERFORMING INSTITUTION:
EWINDSOLUTIONS LLC
30678 SW ORCHARD DR
WILSONVILLE, OREGON 97070
NEXT GENERATION WIND ENERGY SYSTEMS FOR CASH-STRAPPED FARMERS AND COMMUNITIES
NON-TECHNICAL SUMMARY: eWind is
proposing a highly efficient and low-cost wind-energy system for small
and mid-sized farms that will produce approximately four times the
electricity per year as comparably priced "conventional" wind turbines.
This project addresses the USDA priorities of Energy Efficiency and
Alternative and Renewable Energy and Agriculturally-related
Manufacturing Technology. We directly support Energy Efficiency and
Alternative and Renewable Energy since we will enable farms and rural
communities to produce electricity from a clean wind source. Our
project will also positively impact the fledgling unmanned aerial
vehicle (UAV) industry, a manufacturing field that is increasingly used
by farmers and agriculture. We also expect that most of the components
of the airborne wind-energy system will be made and assembled in the
U.S.--a combination that supports Agriculturally-related Manufacturing
Technology. Finally, the project addresses the NIFA Societal Challenge
Area 2: Climate Change. Adoption of our technology will reduce the
overall carbon footprint of farms in fossil fuel-intensive electricity
markets by producing clean wind energy.As Oregon farmer and Organically
Grown Company (OGC) executive director Andy Westlund confesses, being
one of the 800,000 small commercial farmers in America is challenging.
Small farms are highly exposed to market risks, tend to focus on
commodities, and have profit margins of just 3-4% (Hoppe, MacDonald,
& Korb, 2010, p. 6). As such, small price changes of a single crop
can have a substantial impact on their financial status(Page, 2011, p.
11). Thus, variable input costs, such as energy, can play havoc with
financing annual operating loans(Page, 2011, p. 11). Therefore,
reasonable methods of reducing the uncertainty of operating costs and
diversifying the income of small farms are desirable--not just to the
farmers themselves, but also for maintaining this sector of our
agricultural industry.As Andy Westlund and many other small farmers
have discovered, the desire to reduce operating costs is the primary
factor that motivates farmers to produce their own electricity, usually
through solar or wind (Page, 2011). Andy uses recycled solar panels and
"primitive" wind turbines to reduce his energy bills. For small
farmers, the use of alternative energy simultaneously lowers the
operating cost of the farm by reducing the usage of grid electricity
and reduces exposure to risk from fluctuating energy prices (Sands
& Westcott, 2011; United States Department of Agriculture, 2011).
Finally, farmers recognize that generating renewable energy and using
fewer fossil fuels reduces dependence on foreign oil, providing greater
local and national energy security while reducing the risk of climate
change (Sustainable Agriculture Research & Education, 2008, p.
1).However, there are distinct challenges that have kept many small
farms from implementing solar and wind renewable energy. The top three
barriers that Oregon farmers identified are: 1) up-front project costs,
2) permitting, and 3) troublesome paperwork for the incentive
programs(Page, 2011, p. 31). Andy Westlund, for example, had to secure
county permits for his twenty-foot wind tower, which is sited just a
few steps from his thirty-foot-tall house. Additionally, he admits that
the electricity produced doesn't cover the up-front cost of the
system.In addition to the general renewable energy challenges Andy has
encountered, traditional wind turbines have difficulties specific to
farmers that has limited their adoption. The construction of the wind
tower can disrupt farming activities and cause soil compaction issues
(Linowes, 2013). The tower and its blades can also pose an operating
and safety problem for agriculture aerial work and can significantly
hamper their access to cropland, in turn detrimentally affecting
agricultural production(National Agricultural Aviation Association,
2014). Finally, while the permit and incentive program paperwork
problems are important, the up-front costs of a large wind turbine can
remove any chance of adoption by small farmers. For example, wind farms
often used the Vestas V82-1.65 turbine (a 1.6 MW unit) which had an
installed cost of approximately $3.3 million ($2,000 per kW capacity)
(National Renewable Energy Laboratory, 2014). This is well outside the
financial range of any small farm. While wind turbines are available in
smaller sizes and prices, their efficiency drops quickly with the
shorter tower while maintaining a similar level of product complexity.
As a result, the system install cost doubles to $4,000 per kW capacity.
Additionally, they actually produce only 10-15% of their stated
generating capacity, about half the efficiency of utility scale systems
(National Renewable Energy Laboratory, 2014). Thus, a wind tower system
that may be affordable to the average small farmer--e.g., tens of
thousands of dollars--does not produce enough electricity to make it
financially sensible.eWind Solutions
sees these obstacles as both a problem and an opportunity. Our
intention is to remove these barriers and create affordable wind-energy
generation systems tailored to small farms and rural communities. These
systems will produce four times the power per year as comparably priced
traditional wind turbines and are compliant with current federal
regulations. We propose to do this by using a novel method for
generating electricity from wind. Instead of a large, vertical wind
tower with equally large blades, eWind Solutions
uses airborne wind-energy technology. The system consists of a novel
flying craft that is tethered to a power-generating ground station.
Figure 1 shows the basic layout of the system. As the wind blows the
flying craft downwind (step 1), the tether spins an electrical
generator on the ground. When the tether is fully extended (step 2),
the airborne craft glides back to the start of its power cycle (step
3), while the ground station simultaneously winds up the tether and the
process is repeated (step 4).
OBJECTIVES: Technical ObjectivesThe primary technical objective to enable implementation of the eWind airborne
wind energy system is the design and construction of the flying wing
craft. As such, we have broken its research and development into a set
of requirements and four Tasks. Task #1: Determine Design Variables
Task #2: Design Flying Craft Task #3: Build Flying Craft Prototype Task
#4: Test Prototype, Validate Models and Satisfy RequirementsThe
requirements, listed in Table 4, come from a mixture of addressing the
small farmers' needs, complying with FAA regulations, and our initial
research. 1: Determine Design VariablesWhile the requirements form the
general backbone of the research effort, they inform numerous
aerodynamic design variables with various trade-offs. Thus, the first
task of the proposal's technical objective is to determine the
aerodynamic design variables that satisfy the requirements. First, the
required tether tensions and wind speeds represent two points on
the eWind power
curve shown in Figure 3 and were chosen such that the nameplate
capacity, when combined with the calculated capacity factor, will
produce approximately the amount of electricity used each year by a
small farm (see Table 3). To achieve these forces, the flying craft
must have the appropriate balance of aerodynamic parameters. Most
importantly, but not exclusively, in this case is the lift-to-drag
ratio, number and arrangement of wings, control surfaces, and wing
aspect ratio of the craft and a lightweight design.Second, we have
determined that we must utilize at least 75% of the available tether
length within the power generation stage of the flight pattern. Since
the ground and FAA sets hard limits on our vertical operating range, we
must use the tether length that can fit in that space efficiently. This
directly affects the maneuverability of the craft. If the craft takes
too long to turn, the top and bottom of the figure-8 flight patterns
will start to impede into the vertical limits. Figure 5 demonstrates
how a larger turn radius of the flying craft would reduce the length of
the power-generating portion of our cycle. If the craft turns in a
small radius, such as Example #1, it can fit more turns and pull out
more tether in the available space. If, however, it has a larger
turning radius such as Example #2, it quickly hits both the lower and
upper vertical altitude limits before it has time to pull out as much
tether. In this case, the flying craft would spend too much time
resetting its position and not enough time generating power to meet our
necessary capacity factor.While the 75% tether utilization requirement
is necessary for the finished eWind system
to achieve, we will, however, not be able to measure that value using
the Phase I flying craft prototype. Thus, we have determined that a
turn radius of 20 m or less will demonstrate that the flying craft will
be able meet that requirement when it can be directly measured during a
Phase II prototype of the full power-generating system. The advantage
of this modified requirement is that we can measure it directly during
Phase I testing. Working with Dr. Roberto Albertani of Oregon State
University, the result of Task #1 will be a computer model that allows
us to determine the combination of aerodynamic variables (e.g. number
of wings, wing area, wing length, chord, etc.) that will best meet our
requirements.Once we have determined the aerodynamic properties of the
flying craft, we will move on to the task of physically designing it.
This will include the incorporation of the aerodynamic properties with
material selection, structural design, and mechanical control
mechanisms. Using a 3-D structural model, we will evaluate the dynamic
forces on the craft, as well as a finite element stress analysis. In
combination, these will allows us to work out any significant design
problems before we progress to the construction of the craft
prototype.Once the flying craft is fully designed, our research will
transition to the construction of a full-sized prototype. Initial
estimations indicate that the prototype will be approximately eight
feet across. Using existing tools developed by eWind,
we have the space and expertise to create custom wing molds of this
size. These, in turn, will create the actual wings while retaining the
ability to be easily modified. Since it is reasonable to assume that
the design will be tweaked upon actually constructing a prototype, the
mold's potential to be easily adjusted is crucial. Once the prototype
is fully constructed, we must ensure that it actually conforms to our
expectations and requirements. Thus, we will design and implement a
series of experiments to test the values detailed by the requirements
(i.e. the tether forces and the turning radius detailed in Table 4).
Outdoor tests at various wind speeds will measure the force the craft
exerts on a tether using a strain gauge. The turning radius of the
flying craft will be measured by the placement of GPS sensors in the
body and wing tips. These will capture the spatial position of each
point on the craft and will allow evaluation of the maneuverability.
The final objective of the research effort is to prove the feasibility
of designing, developing, and flying the envisioned craft that will
enable highly efficient and low-cost power generation within the
defined settings. We will produce a final report summarizing our work:
the design choices of the craft, the construction and testing of the
prototype, and the success of meeting our tension and turning radius
requirements.
APPROACH: Usage
of general engineering tools and principles, such as
design-of-experiments, failure modes and effects analysis, statistical
principles, finite element analysis, and computational fluid dynamics.
PROGRESS: 2015/06 TO 2016/02
Target
Audience: Nothing Reported Changes/Problems: Nothing Reported What
opportunities for training and professional development has the project
provided? Nothing Reported How have the results been disseminated to
communities of interest? Nothing Reported What do you plan to do during
the next reporting period to accomplish the goals? Nothing Reported
IMPACT: 2015/06 TO 2016/02
What
was accomplished under these goals? Impact: The goal of eWind Solutions
and this research is to create an airborne wind energy system that will
provide cost effective renewable energy to small and mid-sized farms.
Not only will the cost of the electricity produced be comparable to the
price of the local utility grid (depending on wind quality, local
utility prices, etc.), but the system will have several advantages over
existing solutions (traditional wind towers, solar, other airborne wind
energy companies). eWind is scaling the system to produce approximately
the amount of electricity used each year by a small farm. When combined
with a local utility net-metering arrangement, this can alleviate most
of the farm's annual electric bill. The system is four times more
efficient than existing traditional wind turbines of comparable cost.
In addition, our system requires no heavy construction equipment or
large concrete pad, negating the soil compaction issues that have
soured some farmers to wind energy. Land space use is minimal land
space, allowing more room for cultivation, a clear advantage over wind
towers and solar panels. Finally, we are following existing FAA
regulations. Therefore, we do not need special permission or waivers
that other airborne wind energy companies are seeking in order to fly
in the United States. When completed, the eWind airborne wind energy
system will provide most of the electricity for a small farm for a
fraction of the cost of a modern tractor (about $50,000). This will
replace carbon intensive power generation and increase the financial
resilience of our farming sector. This will also bring jobs to rural
communities through installation and maintenance work. The goal of this
grant was the development of the flying kite that will power the
electrical generator. The entire system is complex, but the flying kite
is one of the most technically challenging aspects of the development.
It must be rugged, fast and highly maneuverable while producing enough
tension on the attaching tether to turn the electrical generator. This
grant focused on beginning that development by meeting the initial
physical requirements of the kite, namely turning radius and pull
strength. Results: eWind accomplished each goal laid out in the
proposal objectives and goals. The primary means of measurement on our
progress was the three physically measurable properties of the latest
version of the kite. Specifically we wished to have a turning radius of
less than 20m, a pull strength of 2.1kN at 7 m/s of wind and a pull
strength of 3.8kN at 9 m/s of wind. These measured objectives required
the completion of numerous other tasks that were harder to quantify,
but no less important to the success of the project. Specifically, in
collaboration with Oregon State University, we developed a computer
model that predicts the performance of a given kite design (with
numerous other variables: flight path, tether properties, generator
efficiency, etc.). We combined this with other Computer Aided Drafting
(CAD) and aerodynamic modeling software to create a pipeline from
initial idea to design to build to test. In addition, we streamlined
the manufacturing of prototypes using the CAD software and 3D printers,
allowing us to go from improvement idea to built prototype in just a
few days. While there were minor changes to our work plan which could
be expected of any technology development effort, we followed the
proposed steps closely. They showed we had a realistic and
appropriateplan before beginning this project. While the building
ofthis software and construction pipeline took a large portion of the
grant period, we were still able to meet our quantifiable goals. There
were three criteria we measured (as mentioned above): 1) Turning radius
of less than 20m: This was accomplished during one of our earlier
prototypes and has been the expectation and standard for all following
kites. It was successfully achieved and measured during our latest
test, where we had a consistent turning radius of approximately 10 m. 2
and 3) A pull strength of 2.1kN at 7 m/s of wind and a pull strength of
3.8kN at 9 m/s of wind: These two goals were not technically measured,
but we are considering them accomplished. The final tests were
performed in approximately 6.2 m/s winds. Since we are currently unable
to force nature to provide the exact winds we desire on command, our
best option is to use basic aerodynamic scaling equations to determine
what our forces would be at faster wind speeds. In this case we
measured 1.65 kN at 6.2 m/s wind. Because the force scales at the
square of the wind speed, this equates to an expected force of 2.11 kN
at 7 m/s and 3.40 kN at 9 m/s. While these estimations are based on
rough, simple calculations, they have proven to be accurate to
approximately 10%. Therefore, even though the higher wind speed
estimate is slightly below the grant goal, it is close enough that it
is likely we will meet it with the current prototype or, failing that,
with the next iteration that will continue our tension improvements. We
anticipate being able to directly measure the force at these faster
wind speeds the next time the weather provides them to us. It should
also be noted that all these measurements are collected by a custom
built electronics system. Elements of this system on the kite are:
location, speed and orientation. Ground elements record the tension on
the tether and wind speed. A nearby laptop computer collects both the
ground and kite signals and stores everything at a 5 Hz rate on a
nearby laptop. The electronics also have the capability to send signals
from the laptop back up to the kite. While the kite is currently human
controlled, this capability is the first step in transferring control
to the computer and beginning the work of creating an autonomous flying
system. Finally, because most of our demonstrative progress is best
shown in pictures and plots, we have created a final report outside of
this web form. This report will combine pictures of our prototypes and
data collected from their testing. In addition, we will include the
software results created for us by Oregon State University that
describes their efforts and conclusions.
PUBLICATIONS (not previously reported): 2015/06 TO 2016/02
No publications reported this period.
Item No. 1 of 1
ACCESSION NO: 1010100 SUBFILE: CRIS
PROJ NO: OREW-2016-03858 AGENCY: NIFA OREK
PROJ TYPE: SMALL BUSINESS GRANT PROJ STATUS: NEW
CONTRACT/GRANT/AGREEMENT NO: 2016-33610-25696 PROPOSAL NO: 2016-03858
START: 01 SEP 2016 TERM: 31 AUG 2018
GRANT AMT: $600,000 GRANT YR: 2016
AWARD TOTAL: $600,000
INITIAL AWARD YEAR: 2016
INVESTIGATOR: Schaefer, D. B.
PERFORMING INSTITUTION:
EWINDSOLUTIONS LLC
30678 SW ORCHARD DR
WILSONVILLE, OREGON 97070
CONTINUED DEVELOPMENT OF A NOVEL NEXT GENERATION AIRBORNE WIND ENERGY SYSTEM FOR SMALL AND MID SIZE FARMS
NON-TECHNICAL SUMMARY: As
Oregon farmer and Organically Grown Company (OGC) executive director
Andy Westlund confesses, being one of the 800,000 small commercial
farmers in America is challenging (Hoppe, MacDonald, & Korb 2010).
Small farms are highly exposed to market risks, tend to focus on
commodities, and have profit margins of just 3-4% (Hoppe, MacDonald,
& Korb 2010). As such, small price changes of a single crop can
have a substantial impact on a farm's financial status (Page 2011).
Variable input costs, such as energy, can play havoc with financing
annual operating loans (Page 2011). Therefore, reasonable methods of
reducing the uncertainty of operating costs and diversifying the income
of small farms are desirable--not just for the farmers themselves, but
also for maintaining this sector of our agricultural industry.In our
market research, we conducted sit down interviews with more than 45
small farmers such as Andy Westlund. Not surprisingly, their desire to
reduce operating costs is the primary factor that motivates farmers to
produce their own electricity, usually through solar or wind. Andy uses
recycled solar panels and primitive wind turbines to reduce his energy
bills. For small farmers, the use of alternative energy simultaneously
lowers the operating cost of the farm by reducing the usage of grid
electricity and reduces exposure to risk from fluctuating energy prices
(Sands & Westcott 2011; USDA 2011). Finally, farmers recognize that
generating renewable energy and using fewer fossil fuels reduces
dependence on foreign oil, providing greater local and national energy
security while reducing the risk of climate change (Sustainable
Agriculture Research & Education 2008).However, there are distinct
challenges that have kept many small farms from implementing solar and
wind energy. The top three barriers that Oregon farmers identified are
(based on our own and external research): 1) up-front project costs, 2)
permitting, and 3) troublesome paperwork for the incentive programs
(Page 2011). Andy Westlund, for example, had to secure county permits
for his twenty-foot wind tower, which is sited just a few steps from
his thirty-foot-tall house. Additionally, he admits that the
electricity produced will never cover the up-front cost of the
system.In addition to the general renewable energy challenges Andy has
encountered, traditional wind turbines have difficulties specific to
farmers that have limited their adoption. The construction of the wind
tower can disrupt farming activities and cause soil compaction issues
(Linowes 2013). The tower and its blades can also pose an operating and
safety problem for agriculture aerial work and can significantly hamper
access to cropland, in turn detrimentally affecting agricultural
production (National Agricultural Aviation Association 2014). Finally,
while the permit and incentive program paperwork problems are
important, the up-front costs of a large wind turbine can remove any
chance of adoption by small farmers. For example, wind farms often use
the Vestas V82-1.65 turbine (a 1.6 MW unit), which has an installed
cost of approximately $3.3 million ($2,000 per kW capacity) (NREL
2014). This is well outside the financial range of any small farm.
Although wind turbines are available in smaller sizes and prices, their
efficiency drops quickly with the shorter tower while maintaining a
similar level of product complexity. As a result, the system install
cost doubles to $4,000 per kW capacity. Additionally, they actually
produce only 10-15% of their stated generating capacity, about half the
efficiency of utility scale systems (NREL 2014). Thus, a wind tower
system that may be affordable to the average small farmer (e.g., tens
of thousands of dollars) does not produce enough electricity to make it
financially sensible.eWind Solutions
sees these obstacles as both a problem and an opportunity. Our
intention is to remove these barriers and create affordable wind energy
generation systems tailored to small farms and rural communities. These
systems will produce four times the power annually as comparably priced
traditional wind turbines and are compliant with current federal
regulations. We propose to do this by using a novel method for
generating electricity from wind. Instead of a large, vertical wind
tower with equally large blades, eWind Solutions uses airborne wind energy technology.The eWind system
consists of three main components: a novel flying craft (1) that is
tethered by a rope (2) to a power-generating ground station (3). The
flying craft will be approximately 8 feet across and weigh about 15
pounds. It will fly between 200-500 feet above the ground and will be
tethered to the ground station by use of a nylon or steel rope. At the
ground station, the rope will be wrapped around a steel drum/cylinder
with an electrical generator attached to the drum axle. Figure 1 shows
the basic flying motion of the system and how it generates electricity.
The flying craft does figure-8s as it is blown downwind (step 1),
pulling out the tether and spinning the drum and electrical generator.
When the tether is fully extended (step 2), the airborne craft glides
back to the start of its power cycle (step 3), while the ground station
simultaneously winds up the tether and the process is repeated (step
4). Additionally, this entire system will be completely autonomous,
requiring no attention or effort from the farmer.Although there are
other companies pursuing airborne wind energy, most are focused on
utility-scale electricity generation that will not be feasible for
small farmers. At eWind,
we focus on smaller systems that are specifically tailored in cost and
capacity to the needs of small farmers. We are also compliant with
current Federal Aviation Administration (FAA) regulations, which
require tether flags, lighting, and an altitude limit. Other airborne
wind energy companies are attempting to receive complicated waivers or
rule modifications.
OBJECTIVES: The goal of eWind Solutions
is to create a wind energy system for small farmers that is efficient,
affordable and easy to use. We accomplish this by using a novel way to
collect the stronger, faster, more reliable winds found at higher
altitudes that are inaccessible to wind towers that small farmers can
currently afford. Because we intend to directly compete with
traditional wind turbines, it is instructive to highlight the technical
background of wind energy in general and to then discuss how our
approach differs from existing wind-power technologies.It is important
to remember that the specifications of the flying craft are actually
secondary to the overall goal of the entire system. Thus, our
overriding goal is to provide 40,000 kWh to the small farmer over the
course of a year because that is what actually creates value for our
customer. As the research into the tether diameter showed, when a
particular attribute is (nominally) beneficially improved, it can
actually reduce the power of the overall system. Therefore, we will be
listing specific tensions and specifications we currently believe will
be sufficient to achieve our overall goal. However, we will modify
those specifications if necessary, while maintaining a focus on our
power production goals. This Phase II proposal has technical objectives
that should take the flying craft development to a state that is
commercially viable. This will require the final physical design of the
flying craft and near completion of the autonomous flight control
system and software. We propose technical objectives for Phase II:Group
I:Generate enough tension while the tether is reeled out to generate
40,000 kWh annually (at a good wind resource location).1) A peak force
of 3.8 kN on a static tether at 9 m/s wind speed2) A peak force of 2.1
kN on a static tether at 7 m/s wind speed3) Maintain existing
maneuverability accomplishments (<20 m turn radius)4) Continue to
improve our software analysis capability to include more advanced CFD
(computational fluid dynamics) packagesGroup II:Transition the control
of the flying craft from mechanical means to fully electronic
ground-based signals (fly-by-wire), to computer controlled and then
autonomous crosswind flight.1) Stepwise progression from human
fly-by-wire control to computer control (using a static tether length)
a) Fully human controlled via electronics only (fly-by-wire) b)
Computer controlled stable stationary hover c) Computer controlled
side-to-side drift motion (slow and with small angle deviations from
the vertical) d) Computer controlled side-to-side motion (directed and
with turns) e) Automated guidance for figure-8 crosswind turns2) Fully
automated crosswind flight: follow power production path put out by
computer3) Repeat objectives II:1 and II:2 while actively reeling out
tether (at increasing speeds)4) Stretch goals: Time of flight under
autonomous control a) 5, 10, 30, 60 minutesGroup III:Determine
reliability requirements and incorporate safety aerodynamic
characteristics.1) Use reliability growth curve analysis to track
progress and use research progress and financial considerations to
determine reliability thresholds/goals.2) Introduce aerodynamic, weight
and structure characteristics into the flying craft design that when
control is lost, the craft enters a safer flat spin and settles to the
ground at a reduced speed.The first group of objectives remain similar
to Phase I, and as noted before, the work on them is on-going. Now that
they have been validated with our mathematical model, they remain
acceptable tension goals. Again, if we do modify them, they will
continue to support our power production goals, the cornerstone of our
commercial viability. We will also continue to upgrade our aerodynamics
design software pipeline. Notably, this will include replacing our
current XFLR5 package with a more advanced CFD suite. The second group
of objectives is based around the transition from human controlled
flight to autonomous flight. Based on our research, current progress
and discussions with experts, we believe these are good markers for a
continuous progression to this goal. It is worth noting that these
goals represent a progression of increasingly complex control
objectives. Essentially, we need fully autonomous flight and the
progression to that goal will be a spectrum of improvement. Initially,
these objectives will be met with a static tether length (neither
reeling out nor pulling in). Once completed, we will increase the
difficulty by allowing the tether to reel out in a manner similar to
the power generating cycle. It is expected that the faster the tether
reels out, the harder the autonomous control will become due to the
increased likelihood and severity of perturbations of the flight path
from wind gusts and tether tension oscillations. These tests will
gradually increase the tether reel out speed until it exceeds the
expected speed necessary to maximize the power generation of the system
(approximately 1/3 the wind speed). Finally, we must address the time
duration of flight. In many respects, we actually expect this to be
both easy and difficult simultaneously. To clarify, to achieve the
objectives in Group II we must maintain flight for at least several
minutes to demonstrate the success of each milestone repeatedly. If
wind and weather conditions are stable during that time, longer flight
times are relatively easy. However, the longer the flight lasts, the
more likely it is that something will change with the wind and/or
weather. It is those transitions or perturbations that will cause the
greatest difficulty in extending our autonomous flight time. It is also
the hardest to test and plan for because you have to find a stable wind
condition, start flying and then have the wind be courteous enough to
get mildly unpredictable so that you can test the responses of the
flying craft. (It should also be noted that a parallel SBIR Phase I
submission would develop the tension control system that would help
mitigate these effects during the testing process.) That difficultly
will likely lead to an extended time that we will have to work on
'fringe' flight conditions. Thus, we have decided to place flight time
duration as a "stretch goal" for this grant. As part of Group III, we
must determine how reliable the flying craft must be. To that end, we
will employ reliability growth curve analytics. Our prototype testing
will provide the necessary reliability/failure data. Additionally, we
must determine the failure rate goal. While we obviously do not want
the commercial flying craft to ever fail, it is inevitable that it will
at some point. Assuming that these failures damage the craft and
require a repair visit, we need to balance the economics of that with
the time and money in research and development necessary to increase
the reliability, at least in the short-to-medium term. This reliability
growth curve analytics study will help us manage both the cost of the
craft and its field reliability in preparations for commercial use.
Thus, one of the more important goals of the Phase II proposal is
determining the necessary reliability of the flying craft.
Additionally, we will incorporate design elements into the flying craft
that will minimize the impact of these failures. For example, we are
currently shaping the system to enter a "flat spin" when tether
tension/control is lost. This state puts the flying craft horizontal to
the ground and allows air drag the maximum amount of time to bleed
velocity and kinetic energy before it lands. Currently, this helps us
preserve prototypes, but when commercialized it will increase safety
and reduce damage during flying craft control failures.
APPROACH: Task
#1: Design, analyze, optimize, build and test a flying craft that can
match the tension and maneuverability requirements of 40,000 kWh (Group
I objectives) (Months 1-24). As previously described, the focus of the
entire research and development effort within this SBIR proposal is to
create a flying craft capable of creating enough tension in a tether
that can be converted into 40,000 kWh of electricity. This task will be
carried out by founders P.I. David Schaefer and Dr. Brennan Gantner,
with the help of the expected two new employee hires. The research and
development process will be similar to the work of Phase I. We are
continuing to refine the university collaboration mathematical model
(using an Oregon BEST grant, see Commercialization Plan). The lessons
and refinements of that knowledge are then folded into new designs
built in our aerodynamic software package (currently XFLR5). These
designs are imported into our CAD (computer aided drafting, Autodesk)
software, split into parts and individual pieces are created via 3D
printers, machined by a CNC router and foam hot wire molds. The
prototype is then built/assembled. Tests consist of flying the
prototype in a relatively steady wind for as long as possible/needed.
During that time, our custom equipment is continuously measuring the
tension generated on the tether and the Pixhawk/Labview
software/electronics is tracking velocity, orientation and location (as
well as temperature, air pressure, wind speed/direction, etc.).
Currently, we typically conduct these tests on the Oregon coast. Not
only does this area have reliable winds, but the loose sand provides
reasonable cushioning during control failures and minimizes damage to
the prototype. We are investigating the possibility of also using the
FAA designated drone testing area at the nearby Warm Spring
reservation. While we do not legally need FAA exceptions/permission
this area provides, it does allow us to forgo some of the visibility
requirements for higher/longer tests on a temporary basis.Task #2:
Transition the control of the flying craft from mechanical means to
fully electronic ground signals (fly-by-wire), to computer controlled
and then autonomous crosswind flight (Group II objectives) (Months
1-24) Founder Sean Mish will lead the effort to automate the flight of
the flying craft. After optimizing the CFD software system, Dr. Gantner
will transition to contributing to the automation task. As discussed,
this will be a spectrum of improvement that will slowly transition
control away from direct hand/mechanical links to a flying craft
controlled completely by a software/electronic servo system. While this
is technically a separate task from the aerodynamic design, Mr. Mish
and Dr. Gantner will continually contribute and add requirements to the
flying craft design. Because the computer control aspect of the flying
craft is mandatory, we have determined that it is most efficient to not
only advance the automation concurrently with the aerodynamic design,
but to immediately incorporate its needs and lessons into the current
prototype. For example, if the computer response time is slow to
correct for an unanticipated turn than a human, it may need larger or
more aggressive turning control surfaces to compensate. This
requirement needs to be known to the aerodynamic design team so that
these enhanced control surfaces can be incorporated into the latest
design. This task will begin with the attempt to have the computer hold
the flying craft stationary on a static length tether. Our initial work
suggests that the Pixhawk controller and its corresponding open-source
control software is capable of completing these tasks. At its core, it
is designed to take a radio-controlled hobby plane and turn it into a
computer controlled drone. We essentially want to do something similar,
but with a different default orientation (perpendicular to the tether
instead of gravity) and different physical responses to the standard
commands (it likely won't have airplane standard ailerons, rudder,
etc.).Therefore, a large portion of this task will be the modification
of this open-source software to our specific needs. To this end, we are
initiating a collaboration with Dr. Christopher Lum, who runs the
Autonomous Flight Systems Laboratory at the University of Washington.
His lab specializes in drone control systems and several of his
students have worked on very similar control system modifications of
Pixhawk software. Once we had modified the Pixhawk software to our
flying craft, we will begin to transfer control to it. As stated, the
first task will be to have it hold the flying craft stationary in the
sky. Once it can reliably do that, we will have it create side-to-side
motion, likely by having it make small orientation changes that creates
a slow drift to one side, followed by the opposite motion in the other
direction. The key to this progression will be the ability of the
system to return to a stationary or 'resting' position as quickly as
possible. To succeed at these steps, the Pixhawk will have to
demonstrate that it can move the flying craft in the desired direction
and stop that motion as needed. These tests will increase in difficulty
by increasing the orientation angle from vertical (see Figure 13) that
the Pixhawk is allowed to turn the flying craft, thus, increasing the
speed of the side-to-side motion. Essentially, true side-to-side or
figure-8 flight will be achieved when this angle is increased to (or
above) 90 degrees; when it is pointing horizontally.Task #3: Determine
reliability requirements and incorporate safety aerodynamic
characteristics (Group III) (Months 1-24). Task #3 will be run by P.I.
Schaefer, who has decades of experience in these types of product
development cycles. Reliability growth curves track the improvement of
reliability of the system. One of the key ingredients of properly using
them is having a meaningful reliability goal to achieve. In this case,
that goal will be a balance of costs based on the value of the flying
craft and the cost of replacement (time, travel, materials, etc.)
versus the rate of flight time improvement per dollar spent in research
and development. This task is a combination of financial considerations
balanced with hard data of reliability gathered from prototype testing.
To accomplish this, each failure mode will be tracked and rated on
severity, likelihood, detectability and consequences. These measures
will then be combined (Risk Priority Number (RPN)) to determine the
most severe failures so that efforts can be directed to mitigate those
cases. This process is a standard approach to product development.
While this work will cover the full grant duration, its efforts will
ramp up significantly during the second half.
