Could look like a gust lock at :30 sec. A square plate at the root of the elevator.
https://www.youtube.com/watch?time_continue=1&v=BUtxp4V-jPw

http://i889.photobucket.com/albums/a...psq9g7zhkv.jpg

No discernable elevator movement during the take-off attempt, but one of the post crash photos clearly shows the elevator hanging down . Given that the control gust locks are external this cannot be a result of cables shearing and releasing the elevator to drop under its own weight. Possible that the lock got knocked off during impact, but that area of the plane does not appear to have impacted anything, so unlikely.

They are large wooden blocks, normally painted red and with a long red banner attached. But the banners tend to get frayed and shortened with time, so that's up to the operator. It would be astonishing if the pilot had not checked full-and-free control movement before starting the take-off. Hurried, multi-sector operations can lead to a degree of complacency. So failure to remove locks and pins, followed by failure to perform the pre-take-off check properly, has been done before. But that regime hardly applies to an historic-flight operation.

[Start watching at 11:00]

The key to this accident may lie in that pronounced swing to the left early in the take-off run. That could have been caused by a crosswind gust, but not by a locked elevator.

And that is certainly true because the test flight to determine the stall characteristics would have been in level flight and speed gradually reducing at one knot per second until the stall occurs. A similar discussion arose several years ago on Pprune about DC3 stalls. My experience on type confirmed that a wing drop would occur with startling rapidity if the power was on and in level flight. Test pilot John Farley replied on Pprune that a sharp wing drop would occur if the stall was approached at a faster rate than one second per knot reduction. That may explain why reports of a vicious wing drops during practice stall recoveries were quite common in DC3 days.
That was because pilots would tend to raise the nose to around ten degrees above the horizon to accelerate speed reduction towards stall IAS, rather than gently maintain level flight while reducing IAS at approx. one knot per second which I think is a certification parameter.

One assumes there must have been problems in the very early days for NACA to do that test programme with United Airlines as early as September 1937. Without running the video for a second time, I think the speed decay in each demonstration run is fairly modest. But they do investigate power-on and banked stalls. I also infer that they considered the retrofitting of the sharp leading-edge inboard of the engines solved the problem, so it's interesting to read the comments of you and others.
Yes, I'm sure that 1 kt/s deceleration is used for certification. I guess that covers the approach case quite well, but not the sudden pull-up to avoid an obstruction.

One type of lock:
https://cimg7.ibsrv.net/gimg/www.gmf...7366945904.jpg

In this photo you can clearly see the elevators down, no control lock. Later in the fire, the elevators sag to the ground and move upward to a more neutral position.

https://static-16.sinclairstoryline....?1532222524452

https://static-16.sinclairstoryline....?1532222524452
Yes, that was the photo I was refering to in post no. 42.

In later pictures, the aft fuselage has canted to the right with the right hand stabilizer and elevator tips resting on the ground. That's why the elevators look to be streamlined with the stabilizer in those pictures.

One other thing that will affect the severity of a stall in a DC-3 is the control rigging. If the ailerons are not rigged correctly you can get an abrupt wing drop.

I know of one case of buffet followed by the aircraft rolling almost inverted - told to me me by the Captain doing the training.

To re-state the obvious, none of these discussions of DC-3/C-47 stall characteristics, interesting as they are, helps to explain why the tail-wheel remained more or less on the ground until the a/c became airborne.

It is fairly apparent that it became airborne before it was meant to. Think about an aircraft doing, say 40 knots. Stall speed is about, say 60. Encounters a rapid increase in headwind of say, 25 knots. It will fly until that headwind component drops back to say, 15. Then we see the behaviour that others have discussed. What happened appears obvious - it stalled just after lift-off. Until the pilots are interviewed, why it happened is the mystery.
As an anecdote, I have heard of a DC3 becoming airborne with the elevator gust lock still attached. The crew flew it to a successful landing using trim. Don't know whether fact or fiction, but believe it could be done in the right conditions.

The left stab/elevator hit the ground partway through the sequence. It appeared to be quite a whack and there was a distinct dust cloud generated.

If the pilot lifts the tail of a taildragger, it'll stay on the runway. It is possible to run a taildragger down the runway at a speed much greater than the normal liftoff speed for that weight, just by lowering the nose enough. The same principle applies in landing taildraggers, once the main wheels are on the runway during a wheel landing, lifting the tail a bit will prevent a bounce.

I can imagine a wind gust from 15 to 25 knots in some circumstances, zero to 25 knots is much less common - that would be quite a ride for most types! A 15 to 25 knots gust is easily managed in any certified airplane, with a little piloting skill. A 25 to 15 knot lull next is the reason to lift the tail of a taildragger, and hold it on the runway a little longer, so the margin of speed over the gusts is what you want to maintain control nicely.

I believe one case of a DC-3 taking off with the gust locks on and landing successfully occurred on December 27, 1984, perhaps out of Naples, Florida, with PBA (Provincetown-Boston Airline). The feds had given PBA their operating certificate back earlier in the month after an emergency revocation for shady practices like pencil whipping DC-3 type ratings and flying a YS-11 without a type rating.

Tragically, a PBA EMB-110 crashed in Jacksonville, Florida three weeks earlier with the loss of all 13 people on board so the gust lock incident received little notice outside the local area. The DC-3 gust lock takeoff is mentioned in this 1985 Inc. magazine article about the rise and fall of PBA which at one time was the largest commuter airline:

https://www.inc.com/magazine/19850301/1288.html

That was really ugly, looked like it went off the left side of the runway before they pulled it up and rolled it right. It will be interesting to find out who was on the controls.

It would be interesting to know what it was that had the guy taking the video invoking religion and excrement so early in the take-off run. He appears to have seen something unfolding before it actually did. Let's hope he is a reliable witness.

Long time ago, now, an RAAF C-47 was lost practicing "short field" takeoffs, prior to being used as part of the Australian antarctic flight.
They used a "tail down" takeoff run, got airborne more or less in the "three point" attitude.
What happened was described to me by the only survivor, who worked for Qantas for many years after the RAAF.
At very low speed, just lifting off, one wingtip was hit by a williewillie (dust devil/small whirlwind), a common and completely unpredictable event, and the aeroplane dropped a wing and cartwheeled.
A possible explanation??
Tootle pip!!
PS: Low speed characteristics of the DC-3/C-47, in the real world, were such that we always conducted stalling exercises at a minimum of 10,000 AGL, daylight CAVU only, not just VMC. The wing drop in the stall was quite sudden, with as near as made no difference nil warning, incipient spins were common, developed spins happened on occasions and were, apparently, quite scary, with a height loss of around 5000' being the norm.

I read a comment on a thread on the Facebook page where the video was originally shared from someone who claimed to be there at the time. She (I seem to recall!) confirmed that the concern expressed by the man was due to a violent swing to the left when the tail briefly became airborne, which isn't quite so obvious in the video.

Certainly no gust locks here - thanks!!

Can anyone confirm that the a/c was using the asphalt runway? Very hard to tell from the video - could almost be a grass runway. The asphalt runway is apparently only 75 ft wide - half the standard width. So not much room to recover from a severe swing. Perhaps the first exclamation from the camera operator was a response to the left main-wheel leaving the runway, as I think Airbubba was implying in the above post?

KarlADrage,
I think the tail comes up briefly after the swing has been corrected, but probably with the left main-wheel already on the grass.

More views
 

While riding jump seat in the turbine DC-3 today, I watched carefully what the very experienced pilot did during the takeoff: On the airspeed alive call, control wheel pressed gently forward, and the tail coming up at about 40 knots. Power set and confirmed by 50 knots, and aircraft gently rotated off the runway at the bug speed of 82 knots. My comprehensive stall testing in this aircraft eleven years ago showed a stall speed of 59 to 57 KIAS. The wind was 12G20 right down the runway, and the variability of the headwind seemed very easily compensated by the use of bug speeds from the performance charts. I'm confident that had the winds been more severe, or more variable, the pilot would have allowed a little extra caution speed.

Directional control and pitch attitude control are not really related during takeoff. If the accident aircraft had a directional deviation, it would still have had a better outcome were the tail to be up, and the aircraft at flying speed before being allowed to become airborne. The sudden wing drop visible in the video is exactly what happened to me numerous times during stall testing of a turbine DC-3. Recovery was gentle if initiated before the stall had occurred. If the stall progressed to loss of control, it was demanding to recover with minimum upset. Recovery from a half turn incipient spin in t DC-3 will use up around 2000 feet of altitude for my experience.

I agree with your analysis, Pilot DAR. Whether the a/c had left the side of the paved runway or not, any attempt to use the locked tail-wheel to regain directional control by deliberately lowering the tail again would have been unnecessary and inappropriate. The rudder is very effective with the tail up, although PIO can be a problem.

That said, what puzzles me most is why, if things were going so badly wrong that the tail could not be lifted, the take-off was not simply aborted.

En route to Oshkosh, full of people looking forward enthusiastically to the show, we can deal with this once we're airborne.
Edit: not suggesting lack of professionalism, just that there may have been some unspoken get-there pressure around. An apparently minor problem rapidly growing into something more serious.

In a standard powered DC-3 the forward push to get the tail up can hardly be called "gentle" when heavily loaded. Now this airplane was nowhere near that condition but I have flown them where it takes everything you have to push the tail up. As a matter of fact in those conditions you have to use the trim wheel to get things going. As for directional control, as I mentioned in a separate post, the ailerons are highly effective for directional control and MUST be used, rudder alone cannot overcome a large swing. And we routinely used aggressive differential power when needed to keep the thing straight when aerodynamic control input was not sufficient.

That all sounds very dramatic for a DC3 and may I say a little exaggerated? If you have "used everything" then to get the tail up during the early part of the take off run in a standard DC3 it suggests maybe the Centre of Gravity was a problem. The DC3 is a normal tail wheel aircraft with normal take off and landing characteristics. No unusual control forces are needed and the rudder is effective early. We never had to use ailerons for directional control (apart from appropriate aileron into a crosswind) and certainly never use "aggressive" differential power.

Are you sure you are not talking about flying a DC3 Microsoft simulator? Never heard of using the elevator trim to get the tail up. The elevator trim is useless at low speeds anyway. By what you have described is not an average DC3 handling at all. Of course if you have allowed the aircraft to gather momentum in a large swing, that suggests slow pilot reaction to prevent any initial swing from trying to force the tail up. Once you allow that to happen due poor flying ability, then all bets are off in terms of getting the aircraft to straighten up again and you may even have to resort to dabs of appropriate brake pedal to prevent further yaw on the roll. Not trying to rubbish your description but something not quite right IMHO.

If you don't believe my story above then you probably think that winds in St. Thomas, USVI cannot possibly be reported as "wind 22 knots gusting to 34, all quadrants" when coming in to land and it is a pretty wild ride that DOES take everything you have including near full aileron deflections on takeoff and landing. And yes, in a DC-3 loaded with 30 passengers, 600 lbs baggage aft and 400 lbs baggage forward it IS hard to keep the tail up when you want it up and yes, some guys roll the trim forward as they accelerate and roll it back once the tail is up. When it is gusty you manhandle the controls, there is nothing gentle about it.

Granted, it is calm in this video but that is deceiving, too. A light crosswind is easier to handle than no wind at all since the airplane is pushed to one side and there is no question about which way to correct. In calm conditions it is easier to swing from side to side.

Not Microsoft simulator but 1600 hrs in DC-3s from 1977 to 1979 in San Juan, Puerto Rico. Air Caribbean. 5 airplanes. They all flew different. "No unusual control forces are needed". Hmmm, not really, did you ever fly one at gross wight in gusty conditions? Some had light controls and others had super heavy controls and were pretty tough to handle. Some flew great on one engine with a load in the back, others were pigs and would barely fly with the good engine at METO power and lightly loaded. We had P&W powered and Wright Powered. We had left door DC-3's and right door DC-3's. My memory is pretty good...

Must concur with Judd and Weheka. The aircraft that Bobby G describes above bears little similarity to the C-47s that I flew fifty years ago. I invite Bobby G to explain precisely how - on the aircraft - the ailerons could be used to such high effect in regaining directional control after a swing.

As for differential power, this helps in the early stages before the tail is lifted but, IMHO, could land you in a lot of trouble later on. In our operation the PNF took control of the throttles early on, to ensure that the engines were not over-boosted, and then the PF took charge of them again.

Yes, I've seen that by a rookie on a big jet. Apparently, it had worked in the simulator...

Apologies to Bobby G for doubting his experience in the DC3 aircraft. Still don't understand why the characteristics are not very similar to other tail dragger aircraft but my experience is only in the single engine types. For example in a loaded 185 I would set trim so tail will come up virtually on its own, once six inches or so off the ground, slight back pressure and you are airborne. For you DC3 experienced guys, where is the elevator trim set for take off with a reasonable load?
Sent you PM Bobby but your inbox full?

Thanks Weheka. I'm new here and don't know why my mailbox shows full with 0 messages.
The trim is set for liftoff speed/weight. But really, that's about the same setting for all weights. The heavier and bigger the airplane, the more out of trim it will be in the early takeoff roll until 60-70 kts, that's just the way it is. In some DC-3s we could "fly the tail" during runup when empty, in others there was no way you could. All of the DC-3s handled different, have their own character. Any of the multi-engine airplanes I sent you pictures of would surprise you the first time you fly it. A few years back I started flying the B-25 and although from a distance it looks like a nimble airplane, it flies the same way as the lumpy C-46. It takes muscle. My point with this is that not every DC-3 takeoff is the same. Bluebonnet Belle was loaded for Oshkosh. I wasn't there but the airplane was likely loaded with a lot of fuel and extra gear, different from the usual load. The takeoff would've been much more challenging than normal. As a former CAF pilot I can tell you that it's difficult to maintain proficiency in those airplanes and through a combination of several factors directional control was lost and the airplane yanked off the ground. It is slow to react to control forces at low speeds and heavy weights. But I cannot see aileron movement which is very suspect in my opinion and is the main reason why I reacted to this thread, nobody mentioned the all important ailerons.

The experiences of both Bobby G and ehwatezedoing are very similar to my experiences flying the DC-3.

Ailerons can be used for directional control. I've seen this demonstrated on landing with the tail up and the rudder neutral - the aircraft was moved left and right using only aileron inputs.

The aircraft can be very tail heavy especially with baggage behind the passengers. It wasn't unusual to have to trim to help reduce the forces to get the tail up. Tail not up by 60 knots was an abort.

It's a beautiful aircraft to fly but it needs to be treated with respect - these old aircraft can bite you hard!

Just to put this out there. Could a slow ASI have come into this chain of events, i.e. alive but not yet up to 40 when the aircraft was well past that?

C of G
 

If the pax had repositioned just prior to t/o, moving the C of G aft, could that have initiated the chain of events?