OK,

So I tried Dynamic Soaring using a flight simulator and it works! I
used X-Plane on a mac, and a Fox glider as it has a large working range
for speed. I set up a wind gradiant that was stil air from the ground
to 5000 feet, and then 30 knots/1000 feet above that. So at 7000 feet
there was a 60 kt wind from the North. Not very realistic, but I just
wanted to see if it would work.

Starting at 7000 ft I dove steeply strait downwind until I hit about
200 kts at 5500 ft. then I banked sharply into the wind and was doing
about 220 kts at 5000 (where there was no wind). I then immediately
climbed steeply until I was down to 100 kts at which point I did a
climbing turn to the right and ended up pointed downwind, leveled off
at 80 knots. But this time I was at 7500 feet! I had gained 500 ft in
one hairly circle. However I had drifted downwind from my original
starting point. I repeated this routine and quickly climbed to 11000
feet, always diving about 2000 feet below my starting altitude. As I
climbed my average headwind was increasing, and so I was drifting
downwind faster and faster.

In the next test I wanted to see if I could maintain my upwind position
rather than drifting downwind. I started the same way, diving
downwind from 7000 ft and banking sharply to end up at 5000 ft (where
the headwind was zero) heading upwind. This time I held it level for a
while to make some upwind progress without a headwind. When my speed
dropped to 200 kts, I pulled up sharply and climbed, made my climbing
turn at 100 kts and ended up at about the same altitude and location as
I had started doing 80. I know this because X-Plane can leave a wire
flightpath trail in 3D. I was able to repeat this circuit many times,
maintaining my altitude and holding my position into an average 30 kt
headwind. X-Plane can also display wind vectors in the air, handy for
orienting myself in flight. One can imagine if I turned left at the
bottom instead of right I could make crosswind progress without loosing
altitude or drifting downwind as well.

In the next test, I tried dolphining into the wind without turning. I
dove steeply, and climbed more slowly. While I obviously made progres
upwind, I was unable to maintain my altitude and soon droped out of the
bottom of the shear at 5000 ft.

Finally, I tried looping into the wind. I wasn't able to make this
work, but I suspect it was because my flying sucked and I usually ended
up stalling at the top of the loop. I imagine you would need a fairly
sharp pullover when vertical, and then a sharp pullout when heading
strait down, but I haven't proven it yet.

The next test will be to see how weak I can make the shear before I
can't maintain my altitude (while drifting downwind). I would also
like to try a horizontal shear with the headwind being stronger to the
right and weaker to the left. Plain old flat circles between these
areas would probably work, and you might even gain altitude if you
pulled up on one side of the turn. this type of shear might arrise
downwind, and at the edge of a ridge, etc. Unfortunately X-Plane can't
model this. It would also be interesting to fly behind the vertical
shear created by an obsticle, but X-Plane can't model this either.

I would be interested to see if anyone else could confirm these
experiments, or try them with a different simulator. I will also see
if I can take a movie of the sim and post it.

Cheers,

Matt