AWE researchers have long struggled to match power kites to machinery directly in all conditions, only skilled touchy flying or hit-and-miss synching worked. Now we have a clear theoretic basis for resonantly tuned kite pumping: Kite Eigen Frequency.
The general design problem was not centering the driven load-motion in a symmetric two-way spring-return pumping-stroke cycle. A one-way return-spring helped, but ideal spring-mass action needs to be smooth sinusoidal two-way, cycling potential-to-kinetic energy both ways.
A power kite wants to see a soft load, to not be perturbed in flight, and then will easily do work swinging between two limits, just like a child’s swing (or pendulum) is easy to start and then gradually gets going higher and higher and faster and faster, at its Eigen Frequency at high efficiency.
An AWES working-load pumping mechanism needs to be spring-mass tunable to the Resonant Frequency, which is the maximal driven Eigen Frequency (Natural Frequency) of the Kite-qua-oscillator (loop or 8). Just so, this is how to properly push a child on a swing. Do it wrong, no “magic”
A hidden prediction is that looping kite does for a time continue to passively loop on its own momentum, even after it’s Gibbs Free Energy (energy above that needed for bare flight) is removed.
Some still struggle to relate classical thermodynamics to “room temperature” kite work, others struggle to understand how kites are passively automated (under active supervisory control), but these are separate topics.
In practice, the design challenge is to consistently match wind, load, and kite in resonant pumping. At constant load, as wind increases, a smaller kite is chosen from a quiver to match, and Eigen Frequency increases. At constant wind, as load increases, a larger kite is chosen to match, and Eigen Frequency drops. Power-depower further refines the tuning. The AWES ground machinery must compensate by varying its Eigen Frequency accordingly. This variable-transmission role can somewhat be done electronically, but its better mostly done mechanically, and refined electronically.
It is theoretically sub-optimal to create an artificially low Eigen Frequency by a downwind reeling cycle containing higher-frequency looping or figure-eight cycles, with parasitically asymmetric return cycles. It is worth the trouble to only work a power kite crosswind, both ways, at its Natural Frequency, with minimal cycle-turning-loss.