Winch Launch - Fatal

At 12:09 14 December 2016, Don Johnstone wrote:
At 10:30 14 December 2016, Terry Walsh wrote:
At 09:19 14 December 2016, Jim White wrote:
At 01:59 14 December 2016,
wrote:

Don, you brought up the old pitch-up myth so I'll go there.

Pitch-up on acceleration can only happen with the tailwheel in
the ai
and
=
that can only happen if the glider is slowly dragged into the air
with
feeb=
le acceleration and THEN strong acceleration applied. It can't
happe
if
s=
trong acceleration is applied at the very start of the ground roll
and
main=
tained until the glider reaches the normal airspeed for rotating
int
the
c=
limb.

I could go on at length about this but I wrote a long paper on it
available=
here
https://dl.dropboxusercontent.com/u/...cceleration.pd
Rather
=
than take my word for it, I very strongly suggest the BGA and
other
nationa=
l clubs do their own investigation using tension loggers. What
you will
se=
e may terrify you.


Hi Bill. If you follow URAS you will see that Don, whilst old and
wise
doesn't speak for the UK gliding community.

Hi Bill,

The theory may be correct but I can assure you that a K18
ALWAYS needs th
application of almost full forward stck after lift off to prevent ove
rotation. This is using a powerful V8 winch with about 2 seconds
of groun
roll. Yes the accleration is still happening after lift off. With a
narro
steel drum throttle response is never instantaneous due to inertia
of th
drum, even with the powerful (300 hp+) V8.

I believe there is a solution to this, one which no one will like, but
which would reduce the frequency of the accidents, unfortunately it
may also very expensive and difficult to achieve on existing aircraft
and would result in a reduction in winch launch height. Move the
launch hook forward.



I've picked this one up rather late, and probably like Don, I have many
thousands of winch launches, in vintage gliders, modern gliders, gliders
with compromise hooks, and with C of G hooks, initially on Tost, not
exactly underpowered, and for the last 20 years on Skylaunch, certainly not
underpowered.
To comment on something Bill said, I have analysed quite a few, from 1 sec
logger traces (perhaps not the most accurate way, but all I've got), and
find initial acceleration to be nearer .6g/.7g, rather than 1g. Tost weak
links are approximately 1.3 times auw, so should take 1g , at least when
new.
The problem with the wing going down is I think, that the pilot can be
unaware. A long time ago I had a 19m Jantar, with a compromise well forward
hook, and a loop of cord attached to the release toggle, because you could
not reach it. One day I took a launch, when I got back a few hours later, a
fellow instructor asked if I knew that the wing had gone down, so hard that
everyone at the launch point turned to look. I probably had 1500 hours on
that glider alone, and quite a few hundred launches in it, but hadn't a
clue anything had happened. To me it was a normal launch, and I got away
with it.
I totally agree though that high acceleration is good, the opposite is not.
Again, years ago, probably in the Jantar again, I had a very slow ground
run, but got away. When I got back, I found that a subsequent launch in an
ASW27, had had a wing down accident, substantial damage, fortunately no
injury. Though not there to see it, I always put that down to a slow ground
run like mine.
Most of the early Schleicher gliders, K8, K18, K6, do need a forward stick
position at the start, unlike most others, when neutral is best. The glider
will aviate when ready, it doesn't need encouraging by the pilot.
Winch launching is not inherently dangerous, any more than walking down the
street is dangerous, unless something goes wrong. Not sure without checking
how many are done in the UK each year without incident, but it runs into
hundreds of thousands. Just a very few go wrong.
What I would like to see is a re-think on launch speeds. High speed is
safe, low speed is not. Yet we have ridiculous situations like the Puchacz,
which I think is otherwise the best training glider available, with a
maximum of 59 kts, yet not really safe on a winch launch below 55 kts. This
is a fully aerobatic, inverted flight glider!!!
With the Skylaunch, perhaps others, the throttle stop is set to suit the
wind and the glider. The driver then advances the throttle lever to that
stop over a second or two. I would like to see a completely computerised
system, whereby the wind profile and the glider are selected, and the
driver just presses a button to launch, and all the parameters are built
in, including initial acceleration, surely not that difficult with modern
technology. Obviously there must be a dead man's handle.
Dave

On Tuesday, December 6, 2016 at 6:15:05 AM UTC-8, Jock Proudfoot wrote:


When I started flying a 19meter Kestrel years ago (on aerotow) I quickly found I could not reliably keep my wings level at the start of the aerotow launch if I looked straight ahead, because the ailerons were so inneffective at low speed. So when we started to move, I looked directly left or right towards one wing, for the first few seconds. That made it much easier to keep the wings level, although it usually needed full aileron inputs initially.. My tailwheel Kestrel would initially roll straight ahead wth the tailwheel on the ground, so there was no need to look ahead. It just needed a few seconds to gain reasonable aileron effectiveness, then I could look ahead again.

On Thursday, December 15, 2016 at 10:30:28 AM UTC-7, wrote:
On Tuesday, December 6, 2016 at 6:15:05 AM UTC-8, Jock Proudfoot wrote:


When I started flying a 19meter Kestrel years ago (on aerotow) I quickly found I could not reliably keep my wings level at the start of the aerotow launch if I looked straight ahead, because the ailerons were so inneffective at low speed. So when we started to move, I looked directly left or right towards one wing, for the first few seconds. That made it much easier to keep the wings level, although it usually needed full aileron inputs initially. My tailwheel Kestrel would initially roll straight ahead wth the tailwheel on the ground, so there was no need to look ahead. It just needed a few seconds to gain reasonable aileron effectiveness, then I could look ahead again.

OK, now that we're off on tangents, here's another.

I once flew with a pilot checking out on winch launch who always applied full-right aileron as soon as the acceleration set in - often to the point the right tip nearly dragged. I asked why he was doing that. He replied that he "felt" the left wing was dropping. Video of the takeoff showed the wings level until he slammed the stick over to the right. He was baffled still insisting he "felt" a strong left wing drop.

Eventually, an AME found this individual had a left-right asymmetry in his vestibular organs which, apparently, caused him to experience a sensation of roll under linear acceleration.

This story shows why one must use visual cues and not "feel" to keep the wings level. It also shows the value of making videos of takeoffs.

At 02:30 15 December 2016, wrote:
Having the tailwheel rolling along the ground with=20
the mainwheel several feet in the air is not, and never should be,=20
considered a normal winch launch.

Absolutely true, but counter-intuitively, the only way this could happen
is with low acceleration. High acceleration means high rope tension
which
acts to pull the nose down if the tailwheel is still rolling.


Not really. Where you go wrong is assuming that the tailwheel is the centre
of rotation -
it isn't, it's an arm hitting a limit.

If you look at videos of winch launch takeoffs frame-by-frame you'll find
that it's not unusual to see the main wheel a few centimeters off the
ground while the tailwheel is still rolling on the runway. The ASK-21
photo in
the paper is typical of this. The duration of a "spur roll" is usually
onlya few frames of video so it's very hard to catch with a still camera.

One can get these videos with a YouTube downloader. Then, the file can
be
played back frame-by-frame.

On Sunday, December 18, 2016 at 9:45:10 PM UTC-7, B4soaring wrote:

Not really. Where you go wrong is assuming that the tailwheel is the centre
of rotation - it isn't, it's an arm hitting a limit.

Same difference. When an arm hits a limit, the center of rotation shifts to the contact point.

On Mon, 19 Dec 2016 12:11:35 -0800, Bill.Daniels wrote:

On Sunday, December 18, 2016 at 9:45:10 PM UTC-7, B4soaring wrote:

Not really. Where you go wrong is assuming that the tailwheel is the
centre of rotation - it isn't, it's an arm hitting a limit.

Same difference. When an arm hits a limit, the center of rotation
shifts to the contact point.

Here's a different approach to winch training.

I learnt to fly on a winch (Supacat, with a 240 hp high-torque diesel V8
providing the grunt) initially in an ASK-21 and a G103 with an Puchacz
added to the mix on my 20th launch. Pre-solo, the only aero tows I had
were flight zero (the trial flight in Front Royale's ASK-21 that set the
gliding hook) and flight 25 in the Puchacz for spin demos. The rest of my
pre-solo spin training was off the winch followed by thermal climbs to a
safe height. It was a good, thermally summer.

But I digress. The way I was taught to launch nosewheel gliders on a
winch was to start with enough back stick to raise the nose-wheel after a
few metres and, once it was off, to ease the stick forward to balance on
the main wheel until the glider lifts off. Putting the tail wheel down
was frowned on as being an unnecessarily violent rotation. The same style
works just as well for those gliders on the rather more powerful Skylaunch
that replaced the Supacat.

Similarly, in dry taildraggers I aim to lift off in the ground roll
attitude and maintain that attitude until the airspeed passes 50 kts with
definite acceleration. After that the glider is rotated into the full
climb attitude: no or poor acceleration means don't rotate.


--
martin@ | Martin Gregorie
gregorie. | Essex, UK
org |