|
Methodology
Extended Dimensionless-Analysis- Reynolds
Numbers are dimensionless, allowing confidence in scaling predictions
between viscosity and structural limits. In KiteLab's work, it
became clear that dimensionless analysis also applies to time, funding,
and staffing of research.
Obliged to work at small scales, KiteLab found its design and experimental
cycles were highly accelerated--compared to, say, Makani Power, Inc. [a
major player in the commercial kite power field], which
boasted it could e-mail a CAD file to China and get a prototype expressed
back in 2-4 days, whereas KiteLab often does a novel kite concept in
minutes or an hour or two. Several experiments a day is the norm.
Makani had 10 million bucks; KiteLab had 10 thousand (rounded
#s)--three orders-of-magnitude difference. More crudely, Makani
kites spanned the 100 sq. meter range; KiteLab ruled the 1 sq. meter world
(two orders of magnitude difference). Makani had 30 employees;
KiteLab was one guy helped by many great kite masters/volunteers/advisers;
so in dimensionless-staff terms KiteLab was surprisingly
advantaged.
A key to rapid progress was developing good experimental method. A leading
kitemaster was aghast at my filthy habit of making numerous
simultaneous changes to a design, and then expecting results to sift
out--by some inner voodoo Fourier Transform, which served me well enough
before. His method was one change at a time, which gradually gave certain
results.
Heeding the leading kitemaster's critique, but not wanting to slow down, KiteLab developed a
protocol of using matched kite elements, making a change to one, and
directly comparing performance in a fly-off format (later I would even
change one side of a symmetrical kite; and its lopsided response indicated
a result).
Direct fly-off was a major tool to compare disparate complex
solutions, leading to the development of a large kite-arch with
halyards where varied components are to be run up, compared, and run down
in record time. Debugging goes very fast. A dozen or so ground-based
work-cells compare electrical generation, pumping air/water, traction, and
other applications.
KiteLab's location at the mouth of the
Columbia River (Ilwaco, WA) is a decisive advantage, a virtual
Fermi Lab of fluidic environments--perhaps the best such place in the
world. It is paradise compared to other kite research areas, with
exceptional quality-of-life. Nearby is the
World Kite Museum, whose collection and library is a major asset
no other development project has had (much kite knowledge is not online).
The use of salvage to make prototypes was important--COTS
from birth.
These advantages and methods led to rapid progress; KiteLab was able to
offer an advanced first-to-market kite energy product in '08 (Sputnik
Flipper-Wings) while Makani was predicting five years of
wait-&-see (dimensionless-time). KiteLab is scaling-up with confidence,
almost overnight.
The world's-first Flying Wind-Farm continues as an open research
tool to compare and validate major published concepts. In the rush to
create open-source knowledge, KiteLab cuts corners in documentation, and
left many alleys unexplored. It is still not clear which major
kite-energy concept is the ultimate winner; they all work, only that
the flipper wing was so elegant (~six parts), while the superb
airborne-turbine was much more complex and expensive (~200 parts).
Dave Santos
KiteLab
Ilwaco, WA
Austin, TX
================================================
|