These are three things that I've spent an inordinate amount of time buggering around with to not much benefit: mine, the world's, or even for the local supermarket (consequence of our family therefore spending years in the lower decile for grocery purchases).
Except traction kites maybe, a bit, - but they're too easy so don't really count.
Stirling engines, from the eponymous Robert Stirling (1790 to 1878), extract mechanical power from heat energy by alternately heating and cooling a bit of air- or a bit of some more perfect gas, pressurised to a hundred bar or so as well, if you're being serious.
Very many engineers have put their all into courting Stirling engines without any very satisfying consummations as yet ( so seductive they are, but ice maidens). But what is 184 years (since Presbyterian minister Robert S took time out from administering to his congregation for this mechanical diversion) in the 14 billion or so years life of our universe so far; let's not bury his idea untimely. And I've been chasing this maiden since 1987 with a series of 13 prototype engines so far. Initially there was serious world changing intent in the mix somewhere, but they seem to have become more whimsical in recent years.
Number 11 (wood fired with the occasional encyclopaedia thrown in to keep the fire hot) now powers a yard cart entrant in our yearly "Great Race", Number 12 is fitted to an 8m replica great lakes gentleman's fan tailed launch (burns handy sized Mills and Boons bodice rippers) and Number 13 is intended to drive an exercycle (for virtuous feelings without having to labour unnecessarily), and heated by burning fitness books of course. Number 11 Prototype Stirling Engine 2009
The process of Stirling engine development has considerable parallels with kite development- for neophyte wanabee inventors they exercise the same fatal attraction it seems.
At the Seattle "Ancient Interface" conference in 1985 that looked at the future for kite sailing, a significant proportion of contributors listed the arcane field of Stirling engine research in their CV's - and I would have as well, but I wasn't there (not invited).
Number 12 Protype Stirling Engine First Start Up
Like for traction kites, Stirling engines require multiple variables to be simultaneously optimised. Far too many for any analytical technique to encompass, but not too many for evolutionary trail and error processes to eventually deal to.
Nothing about traction kite development is now particularly mysterious- sure it's difficult, (and there will never be the ultimate traction kite, because there can be no single metric for 'ultimate'), but cause and effect are clear enough, qualitatively at least, and successful designs are about balancing compromises.
Stirling engines also yield to iterative non-analytical approaches, but their development has been slower on account of (I think) that it takes much longer to build and test a Stirling engine variant than it does ditto for a traction kite - and that many more experimenters have a sewing machine and fabric, than have the engineering equipment, skills (and spare cash) required to make engines. And it's not entirely the case that analytical techniques aren't useful in these evolutionary processes. For Stirling engines and for kite aerodynamics, numerical simulation programs now usefully reduce the number of prototypes required in any development sequence. Number 12 Prototype Stirling Engine 2011
Single line kites are a different game though:
Of all possible shapes, the very small number that will fly when tethered by just one line share just a few characteristics, but vary from each other in ways that just don't seem to make sense. In essence the problem is that the stability of single line kites is not only a function of their interaction with the wind at any particular instant, but also of their responses prior to this instant, Stability (or lack of) is a time dependent relationship. Kite Ice skating Lake Cw Peel 1993
I could easily write a book or two on what I've learnt so far about Stirling engines and traction kites, but after 40 something years of reasonably focussed effort, I can write everything I know with reasonable certainty about single line kites in just two sentences:
"Kites are pendulums; they point themselves upwards by having the lifting force from the wind acting above where their weight acts to hold them down".
"Hunting/looping, the most common form of kite instability, is driven by lift and restrained by drag , so changes that decrease the ratio of lift/drag make H/L susceptible kites more stable."
Definitive statements summarising a lifetime of work, such as these are supposed to be, are usually exquisitely economical of word use (and therefore unintelligible to other than the cognoscenti). I'm working on this.
And as for all the other theories about single line kites I've acquired along the way; sometimes they work, sometimes they have the opposite effect to what's predicted, and sometimes they have no effect at all- in roughly equal thirds.
But this is what makes single line kites so interesting! Mathematicians have an appropriate ditty to describe such fatal attractions: "Any problem worth attack will prove itself by fighting back."
I just wish that the problem of single line kite stability were as simple, as for example, Fermat's last theorem.
And here's another relevant ditty: "Big fleas have little fleas upon their backs to bite them, and little fleas have lesser fleas, and so ad infinitum".
Appropriate because it describes a solution to the inverting problem that the 4 bridle Pilot kites I now mainly use, have in turbulent winds and when flown too low. This inverting seems to be an inherent side effect of their de-power (they have more pull in light winds than similar sized Pilots, but much less in stronger winds, a characteristic worth hanging on to even if there's occasional downside), so a solution that doesn't also cancel out this design's greatest virtue is unlikely to be easy. Big Fleas Have Little Fleas
I expect there will eventually be an elegant answer though, and cutting the flares back from the kite's leading edge by 220mm as has been done for the purple 10sq.m 4 bridle Pilot in the accompanying photo is definitely a step in the right direction.
But here's an inelegant solution that, so far in 50 or so hours of turbulent wind flying, has completely eliminated inversions:
For now, a development of the PLK 2m Pilot fixes everything; it has a wind range from too little to too much, is inexpensive, and when flown above any of the 10 sq.m and 13 sq.m 4 bridle series Pilots (which have stackability), does the launching for you and makes them crash proof.
And just in case anyone out there then asks; "if the 2m Pilot is so good , why not just scale it up and avoid all this drama", the answer is that kites, and especially ram air inflated kites don't scale. One of the theories for why this is so, is that fabric stiffness has a significant effect: For a small ram air inflated kite, the leading edge openings tend to hold their form in zero wind, but for larger versions they flop closed. Another is that the mass of enclosed air increases with the cube of dimension while lift and drag forces increase with the square of dimension. Larger soft kites therefore tend to respond much more slowly after developing a lean to either side- with the consequence that they use up more of the available sky before correcting, which has profound effects for stability. (and yes, of course, these two theories are subject to the 1/3rds rule from above).
Ashburton, 31 July 2011
PS. that's Amos our first grandchild in the Number 12 Stirling engine prototype photo.