The picture above is from my accident report.
http://ozreport.com/pub/fingerlakesaccident.shtml
Tucking on the cart
Preface
This is a detailed accident analysis of my cart/aerotowing accident at the Finger Lakes Aerosport Flight Park in Macedon, New York, at around 3 PM on Thursday August 2nd, 2004. I discuss all the factors involved in this accident with the goal of making cart aerotowing safer. I've aerotowed with a cart hundreds of times and not had the slightest problem that I can recall. (I did attempt to launch unhooked once while foot towing behind a Dragonfly in Australia.)
Credits
I want to thank the many pilots and others who helped with this analysis. They include Rob Kells (who gathered information, heard the whack, but missed the action), Steve Prepost, who watched the whole thing especially the wheels, Maureen Grant, who caught almost everything on film, Greg Dewolf, who provided a lot of feedback and got a video of the second accident, Marylin Nichols, who got me the pictures on video, David Cameron, who did a thorough analysis of the castering of the wheels on the cart (especially after the second accident), Marty Beckenbach, who reviewed this and other reports and took quick action, Joel Spano, who engineered the carts, and many many others.
Summary
I was flying a Wills Wing Sport 2 demo glider provided by Rob Kells at the National Fly-In. On my third aerotow this glider fell (was pulled) off the cart asymmetrically with the left corner bracket coming off the cradle first and the glider yawing to the right (clockwise) when the wheels started castering. The left wing drug over the ground for a second or two. Then the left corner bracket/wheel dug into the ground as the glider wasn't lifted and was in fact pulled into the ground by the V-bridle. The nose came over, the weaklink broke, and I plowed into the left side of the undersurface of the glider cutting my lip on the plastic shrink wrap of the nose wires, ripping the undersurface and putting a dent in the left leading edge.
There were many factors that lead to this accident and I will attempt to put all of them in their proper place.
Before the accident
On the previous tow I had a similar (except for the bad parts) experience. The winds are light to nil, sometimes a bit (1 - 3 mph) out of the east or south. We were towing to the southeast from a spot near the glider setup area. The immediate tow area is smooth and we were lining up between orange cones which mark the start point. The tow lane is a gradual up slope to the crown of the drummel which is on the main north/south runway.
About twenty yards into the tow lane there is a rougher area with sparse grass and perhaps some slight north/south ruts from some previous car or perhaps tractor travel to the south at the western edge of the main runway. Most of the ground has been thoroughly rolled with the Bomag. After this area the grass gets thicker and the main runway begins.
On this previous tow, I felt that the glider was somewhat "stuck" to the cart. The cart wheels on this Lookout version of the cart started to wobble/caster near the end of my tow. I did something (or perhaps not) to get myself off the cart, perhaps pushing out a little. I found myself going over the ground at about twenty five mph at about one foot. I then slowly climbed out.
It has been very rare for me to have the wheels on a cart wobble like these did. This cart was designed and built at Lookout and this design is used at a number of flight parks.
I was being towed by Rhett Radford flying a Bailey-Moyes Dragonfly with the 582 model motor which is the less powerful version of the engines used on some Dragonflies. I've towed behind them many times.
I was flying the Wills Wing Sport 2 with a double V- bridle with both a shoulder and keel attachment. The shoulder attachment is my normal Pro V-bridle with a barrel release (found at any flight park). I normally just tow with this attachment. I rarely use a bridle attached to the keel, as I find that I can handle/tow any glider just towing off my shoulders.
The second V bridle goes from the center of my shoulder bridle and to the release attached to the keel. It is released using a bicycle brake handle and cable release that Rob Kells had attached to the keel and right downtube. The release and end of the cable is attached to the keel with a cord and comes out just behind the trailing edge of the under surface. It is quite a ways in front of the carabineer (see first picture below).
This double V-bridle setup is used in order to make the Sport 2 easier to tow by reducing the tow pressures. The idea is that you don't have to pull in as much to go fast enough to stay with the tug. I really didn't need this mechanism and should not have used it as I was more comfortable using my standard V-bridle off my shoulder only.
On my first tow I was hung four inches lower than on my second tow. This discouraged me from pulling forward while on the cart nearly as much as I normally would. I normally let the tug pull me forward when I first start to get the secondary release/barrel in front of the base tube. I usually let myself be pulled forward until the bar is just under my chest.
On this second tow the base tube is just below my chest and I can wait for the keel to come off the cradle in the back as the glider rotates up. I stay right there.
The double V-bridle pulls my shoulder bridle up a bit so that it doesn't rest on the base tube as is normally the case when I start to tow just off my shoulders. I therefore don't have to hold up the bridle off the base tube when I just start, as I normally do to make sure that I don't catch the barrel on the base tube. It also rocks me up a bit as the carabineer attached to the end of the tow rope is now quite a bit higher than I am normally use to when it is attached to just my shoulder bridle.
So let's look at the factors: No head wind, uphill launch, lower power on the tug for a slower start, perhaps uneven ground, wheel starting to caster when the cart gets going at high speed (and thereby dramatically increasing the drag on the cart), double-V bridle that pulls the keel down in certain circumstances (see more on this later), pulling myself forward to match my normal position associated with using only a shoulder bridle (pro tow).
We'll see how these factors combined on the next tow to spell disaster.
The accident
On this third tow the same factors that were at play on the previous tow, that had no really bad consequences, were slightly tweaked and therefore caused a big problem.
I lined up to tow in a cart that David Glover went and got for me at my request. It was one of the new carts built specifically for Finger Lakes Aerosports. I looked at the angle of the glider on the cart and it looked good.
I hooked up to a Dragonfly piloted by a pilot with whom I was not familiar. The start was uneventful. I made sure that I swung forward as the tug pulled on the bridle and I came to my standard position (see below). I was thinking that I should stay on the cart a little longer so that when I came off the cart I would get a little higher off the ground quickly this time.
I was pulled up the up slope as the tug accelerated and everything appeared to be normal. As my speed and the cart's speed increased to about twenty five miles per hour, I was still on the cart. My keel had raised just slightly out of the cradle at the back.
At this point the cart wheels began to caster/wobble. I was going very fast and I was still on the cart. I had been on the cart for a while and I was now in the area where the ground wasn't quite as smooth. Except for the previous tow I had really not had wheels on a cart wobble anywhere near as much as these wheels were wobbling. I couldn't tell if it was because I was stuck on the cart going fast, because I was hitting rough ground, or because the wheel holders weren't dampened enough to slow down the oscillations. Most people felt that if the cart was heavily loaded then the wheels shouldn't be wobbling.
So I was going fast, I was on the cart, and the wheels on the cart were wobbling causing lots of bumps for me on the cradle on top of the cart. I wasn't feeling the glider trying to fly or attempting to get off the cart.
Soon after the wheels began to wobble the left hand side of the glider's control frame came out of the cradle (which is very slick) and moved forward about a foot (Of course, I had been holding onto the safety rope but it is much harder to hold onto the rope when the glider goes forward than when it goes up off the cart). The glider yawed to the right. Here's the shot of this situation by Maureen Grant:
Look closely at the photo. Line up the two forward cart wheels, then line up the control frame. You'll notice that they don't line up. This is a shot where I have just rotated clockwise and the left hand side of the control bar is in front of the cradle. The left corner bracket is about four or five inches below the right corner bracket which is still in the cradle.
The keel is just above the back cradle and behind it (from this view point) as the glider rotates. My elbows are cocked. The bar is just below my chest (I have a much higher resolution version of this and all the other photos if you are interested).
I have been pulled forward as my feet have been pulled off the aluminum bars where I normally place them to hold up the tail of the harness. My 5030 vario has been rotated down by the bumps (this is pretty normal).
The glider is being pulled by the keel/bridle connection which is about a foot in front of the control frame junction at the keel. The glider isn't flying or going up. It is falling on the left side. The bridle is pulling the glider down. The glider is tucking.
Here is Maureen's next shot:
In this shot the left wing is now dragging on the ground. I can definitely feel this. The 5030 is hitting the ground (it survives with no problems). I have been pulled further forward and I am laying on the cart. The left corner bracket is obscured and probably isn't hitting the ground just yet. The glider is being pulled down by the bridle connection to the keel. It would have been nice to release from the tow rope just a bit before this so that I wouldn't be being pulled down.
Rob Kells noticed that the bridles had kicked in and the the glider looks like it does when it is undergoing a negative load test.
Things get worse:
At this point the bridle hooked to the keel is parallel to the under surface and pulling the glider directly into the ground. I've been pulled further forward. The left corner bracket has hit or is about to hit the ground. The cart has a lot of drag because the wheels have been castering to at least 70 degrees. Notice the front wires of the glider and how loose they are. The keel is bent down by the wind on the top of the sail.
The final shot:
The nose goes in, the weaklink breaks, my head goes through the under surface missing the keel because the glider is asymmetrical having yawed clockwise. The shrink wrap ends of the left nose wire rips my lip and I get thirteen stitches later at the emergency room.
The crash was quite loud and many people felt that much worse had happened to me. It easily could have if I had hit the keel with my head. I swung through and didn't really hit the ground with my head. I could easily have had much worse injuries.
How will I avoid this problem in the future? I will tow only from my shoulders. I won't let the tug pull me forward quite as much (just enough to clear the barrel of the release). If I'm towing with a double V-bridle, I will barely pull in at all, in fact I will lock my arms with the bar at around my face and make sure that I am not pulled forward too much (get in the position where the bar would be to be between minimum sink and best glide).
After the crash
The next day a similar crash took place with the pilot tucking the glider while the control bar was still on the cradle. The pilot was pulled right through the control frame.
After these two accidents and a detailed analysis of castering by David Cameron, Marty decided to quit using the new carts. There are five Lookout type carts here and only they are being used for towing. There are a number of changes to the new carts that are being reviewed.
Joel and Marty performed tests on the carts late in the evening on Saturday (I crashed on Thursday) and I happened along to see how it went. Joel Spano was flying a Wills Wing XC and did not pull in. As soon as the cart began to roll it experienced dramatic castering. He was able to launch as the cart continued to caster throughout the tow.
We had been watching tows all day using the Lookout version of the carts and there was no castering on these carts at all.