The theoretical analyses of this sort of thing I've read point to
somewhat different maneuvers.

First, you don't really want to slow down. The only reason for slowing
down is to expoit a vertical wind shear. If the shear were discrete,
i.e. 30 kts above 1000' and 0 below, then you wouldn't slow down much
at all: you'd do the into wind turn at 1001 feet, 100 kts, and the
downwind turn at 999 feet, 99 kts. You would not pull way back and go
up to 1500 feet at 40 kts.

You gain more energy the faster you're going. Think of a tennis ball
gaining energy by being swatted back and forth between two rackets.
That's what we're doing here. You gain more energy the more massive the
ball, and the faster it goes.

In "real life" the shear is not so discrete. So there's a tradeoff: the
higher you go, the greater the wind difference, but the less energy you
gain since you're going slower. The optimum is not, though, to rise up
to near stall and then dive down. The upwind turn should still be a
high speed, high bank, high g affair.

Second, circles are not optimal. The model airplanes we've seen do this
because they need to stay near the guy with the transmitter. The
optimal path changes course as little as possible. It's more efficient
to do S turns, ideally in a heavily ballasted glider going pretty fast.


Where might we actually use this in real life? My winter day dreaming
has me trying to make an upwind transition on a ridge day, realizing
I'm not going to make it, and then turning sideways, dynamic soaring in
the lee of the upwind ridge, doing S turns parallel to the ridge but
slowly moving upwind, until I gain enough energy to clear the top of
the ridge. We know model airplanes can dynamic soar in the lee of
ridges; we need to find out if the shear is strong enough and well
enough defined a mile or so downwind of the ridge to make this useful
for sailplanes.

If someone gets motivated to try it this winter, let us know. Matt: how
about programming something like this in the simulator, and let's see
if you can pick up enough energy in a 15-20 kt shear over say 500
vertical feet to make progress upwind?

John Cochrane BB