Does anyone use a tug tow hook that releases automatically whenglider kites?

On Tuesday, February 18, 2014 1:48:05 AM UTC-5, wrote:

Here is the link to the active thread on the AvCom site regarding this incident including the eye witness account. http://www.avcom.co.za/phpBB3/viewto...p?f=9&t=130916

Based on what was posted on that link, the pilot of the glider was expert, current, and flying a modern glider.

Chris has hit the nail on the head with his testing and reporting. When it goes wrong, it goes wrong fast. Faster than you can possibly react. We can try all we want to teach how to not do it, but it will sometimes happen. Why not come up with a system that will minimize the risk to the towpilot? I think we have concluded that for now, we cannot eliminate it without creating other significant problems.

If we are considering an automatic system, why wait to full nose up elevator and nose down 20 degrees? As Chris said, if you are there, you are probably doomed. If the nose is down 10 degrees (or maybe even 5?) and the elevator is half way to full up, something is wrong. Would you agree? On tow, elevator is up, nose is up. On descent, elevator is down, nose is down.

Why not develop and test a parallel releasing system that has inputs from a gyro for pitch attitude and a simple sensor for elevator position? Test at safe altitude, as Chris and company did. I know this is not going to prevent kiting, but if we can reduce the risk when it happens and maybe not even add bad failure modes, it should be well worth the effort.

Just my thoughts.

Steve Leonard

On Wednesday, February 19, 2014 7:36:56 PM UTC-5, Steve Leonard wrote:
Chris has hit the nail on the head with his testing and reporting. When it goes wrong, it goes wrong fast. Faster than you can possibly react. We can try all we want to teach how to not do it, but it will sometimes happen. Why not come up with a system that will minimize the risk to the towpilot? I think we have concluded that for now, we cannot eliminate it without creating other significant problems. If we are considering an automatic system, why wait to full nose up elevator and nose down 20 degrees? As Chris said, if you are there, you are probably doomed. If the nose is down 10 degrees (or maybe even 5?) and the elevator is half way to full up, something is wrong. Would you agree? On tow, elevator is up, nose is up. On descent, elevator is down, nose is down. Why not develop and test a parallel releasing system that has inputs from a gyro for pitch attitude and a simple sensor for elevator position? Test at safe altitude, as Chris and company did. I know this is not going to prevent kiting, but if we can reduce the risk when it happens and maybe not even add bad failure modes, it should be well worth the effort. Just my thoughts. Steve Leonard

The least expensive improvement that many operations can make immediately is to adopt the use of longer tow ropes.The effect of pilot error is drastically reduced by this one simple change. Many places I fly as guest use ropes that are marginally short as matter of standard practice.
We try not to use ropes shorter than 200 ft.
UH

On Wednesday, February 19, 2014 8:08:24 PM UTC-6, wrote:
The least expensive improvement that many operations can make immediately is to adopt the use of longer tow ropes. The effect of pilot error is drastically reduced by this one simple change. Many places I fly as guest use ropes that are marginally short as matter of standard practice. We try not to use ropes shorter than 200 ft. UH

Same practice at Sunflower Gliderport. (my home base)

Steve

How about mounting a camera on the towplane that's keeping an eye on the glider. If the vertical position exceeds some predefined rate of change, the release is pulled.

Got this idea from this story:
http://www.gizmag.com/air-hockey-tab...r-parts/30840/

A camera is mounted above an air hockey table and software is able to track the puck and use that information to drive a paddle to return pretty much all incoming shots.

To test this out one only needs to mount a camera on a busy towplane to collect lots of "normal" operations as well as some simulated kiting scenarios.. Then some smart person could analyze the video to see if it's possible write some code to effectively track the glider.

The big question is whether a kite is detectable early enough, and different from normal operations.

-Tom

At 02:08 20 February 2014, wrote:
On Wednesday, February 19, 2014 7:36:56 PM UTC-5, Steve Leonard wrote:
Chris has hit the nail on the head with his testing and reporting. When
i=
t goes wrong, it goes wrong fast. Faster than you can possibly react. We
ca=
n try all we want to teach how to not do it, but it will sometimes
happen.
=
Why not come up with a system that will minimize the risk to the
towpilot?
=
I think we have concluded that for now, we cannot eliminate it without
crea=
ting other significant problems. If we are considering an automatic
system,=
why wait to full nose up elevator and nose down 20 degrees? As Chris
said,=
if you are there, you are probably doomed. If the nose is down 10
degrees
=
(or maybe even 5?) and the elevator is half way to full up, something is
wr=
ong. Would you agree? On tow, elevator is up, nose is up. On descent,
eleva=
tor is down, nose is down. Why not develop and test a parallel releasing
sy=
stem that has inputs from a gyro for pitch attitude and a simple sensor
for=
elevator position? Test at safe altitude, as Chris and company did. I
know=
this is not going to prevent kiting, but if we can reduce the risk when
it=
happens and maybe not even add bad failure modes, it should be well
worth
=
the effort. Just my thoughts. Steve Leonard

The least expensive improvement that many operations can make immediately
i=
s to adopt the use of longer tow ropes.The effect of pilot error is
drastic=
ally reduced by this one simple change. Many places I fly as guest use
rop=
es that are marginally short as matter of standard practice.
We try not to use ropes shorter than 200 ft.
UH


The tests I described were done with ropes about 180 feet long, I think it
would probably need ropes of 400 feet or more to give sufficient time for
there to be even a chance of releasing in time if a kiting event started.
There are practical reasons for not using ropes that long at most sites.

I had a quick look at BGA statistics for 1997 (probably a fairly
representative year in the long term), approx 104,000 aerotows. I think we
have only had about four kiting fatalities in the UK, in the last 40 years
(I'd be pleased if someone with access to the stats could check that, I'm
relying on fallible memory). That makes the kiting fatality approx a one
in a million event.

Of the various automatic releases proposed, I can forsee two possible
problems, one is a glider mounted automatic release being unintentionally
activated for a winch launch and giving the glider a launch failure at the
moment it starts to rotate into the full climb (might be pilot error or
switching failure that caused it to be activated). The second is for a
tow-plane or glider mounted one dumping an innocent glider at an early
stage of the tow, this is potentially very dangerous as the land ahead
option often doesn't exist between 50 feet and 300 feet, which necessitates
turning at very low altitude, the commonest cause of spin-ins.

It would take an exceptionally reliable automatic release to be sure that
the launch failure fatal accidents caused by its malfunction came out as
less than one in a million launches.

Please don't misunderstand, I am not arguing against attempts to develop
such a mechanism, just trying to make sure there is understanding of the
degree of reliability needed.

At 02:50 20 February 2014, wrote:
How about mounting a camera on the towplane that's keeping an eye on the
gl=
ider. If the vertical position exceeds some predefined rate of change,
the=
release is pulled.

Got this idea from this story:
http://www.gizmag.com/air-hockey-tab...r-parts/30840/

A camera is mounted above an air hockey table and software is able to
track=
the puck and use that information to drive a paddle to return pretty
much
=
all incoming shots.

To test this out one only needs to mount a camera on a busy towplane to
col=
lect lots of "normal" operations as well as some simulated kiting
scenarios=
.. Then some smart person could analyze the video to see if it's
possible
w=
rite some code to effectively track the glider.

The big question is whether a kite is detectable early enough, and
differen=
t from normal operations.

-Tom

Hi Guys
Just read through the last 66posts on this page. After what happened here
in South Africa this week I though something needs to be done. I was
thinking along the lines of a tilt switch. Whilst noodeling I cam across a
UK company see...
http://www.leveldevelopments.com/pro...tilt-switches/
and what they did for another company.
http://www.leveldevelopments.com/cas...ghting-system/
This really stood out....
**The final sensor was a custom solution comprising a three axis
accelerometer and 3 axis tilt sensor. The tilt sensor used a customized
filter algorithm to ensure the dynamics of the vehicle in use did not cause
false alarms.**

This system can be tested a low initial cost with a light on the tug pilots
dashboard every time it is triggered to iron out false alarms.

Only thing you need is a tilt switch. With a safety ON and OFF.
Basically before the tug the pilot turns "ON" the tilt switch switch then
proceeds with the tug. If the tug is pulled into a **30 degree dive the
tilt switch is activated and drops the rope automatically through a yet to
be worked out system.(Could be explosive, pneumatic or some other method)
Once the glider pilot has released (under normal gliding conditions) the
switch is turned "OFF" by the tug pilot so it will not be deposited over
some farm field when the tuggie dives back to the field at **30 degrees.

**30 degrees is an example could be 25, 35, 40 Tests would have to be done.

At 14:14 20 February 2014, Hans Heydra wrote:
Hi Guys
Just read through the last 66posts on this page. After what happened here
in South Africa this week I though something needs to be done. I was
thinking along the lines of a tilt switch. Whilst noodeling I cam across
a
UK company see...
http://www.leveldevelopments.com/pro...tilt-switches/
and what they did for another company.
http://www.leveldevelopments.com/cas...ghting-system/
This really stood out....
**The final sensor was a custom solution comprising a three axis
accelerometer and 3 axis tilt sensor. The tilt sensor used a customized
filter algorithm to ensure the dynamics of the vehicle in use did not
cause
false alarms.**

This system can be tested a low initial cost with a light on the tug
pilots
dashboard every time it is triggered to iron out false alarms.

Only thing you need is a tilt switch. With a safety ON and OFF.
Basically before the tug the pilot turns "ON" the tilt switch switch then
proceeds with the tug. If the tug is pulled into a **30 degree dive the
tilt switch is activated and drops the rope automatically through a yet
to
be worked out system.(Could be explosive, pneumatic or some other method)
Once the glider pilot has released (under normal gliding conditions) the
switch is turned "OFF" by the tug pilot so it will not be deposited over
some farm field when the tuggie dives back to the field at **30 degrees.

**30 degrees is an example could be 25, 35, 40 Tests would have to be
done.



To repeat one point, our tests left us convinced that the tail-plane on the
tow-plane stalled quite early on in the kiting event, it would need some
tests to check exactly how early. It might be that the nose down angle
would have to be quite small, something that could happen in turbulence
perhaps.

On Thursday, February 20, 2014 9:36:03 AM UTC-5, Chris Rollings wrote:


To repeat one point, our tests left us convinced that the tail-plane on the

tow-plane stalled quite early on in the kiting event, it would need some

tests to check exactly how early. It might be that the nose down angle

would have to be quite small, something that could happen in turbulence

perhaps.


That's a pretty interesting (and gut wrenching to this tow pilot) point Chris.

Stall warning flipper on the horizontal? Feasible?

-Evan Ludeman / T8