Let's think about EOLS:
Dave came through the RAF system, which trains you (in a single engine aircraft) various different EOL techniques for dealing with an engine failure, completing the exercise with the 'full blown' Engine Off Landing. Before others leap in I also realise that some ac - particularly the UH-1 and derivatives - have other techniques as well.
Firstly there was the 'standard' technique - the engine fails, flare back to approx 60 -70 kts (depending on ac) and lowering the collective, then manoeuvre the ac to achieve an into wind landing in a suitable field with a flare - check - level and run on. The harder the flare the less the run on speed but more chance of striking the tail. This is the EOL of choice when VMC in daylight and the ground (and suitable site) visible. The Rate of descent of most rotary will be about 1500-2000ft/min - but again AUM, DA etc will give differing figures. Variations of this include both range auto - drooping the RRPM but increasing IAS giving greater range but slightly higher RoD - and Zero Speed - again exactly what it says to enable a vertical (or even backwards) auto in strong wind but with a very high RoD of 3000ft/min or greater. Both of these techniques demand a return to the 'standard' technique by approx 500-800 ft.
Then there was the 'Min Speed' auto. This was to be flown at night/IMC or when the surface was obscured. Immediately flare back to approx 35-40kts (whilst lowering the lever of course) and turn back into wind. At night flick the landing light on. The RoD is now closer to 2-2500ft/min. When the rad alt reads about 30ft or so you rapidly raise the lever and hopefully cushion the landing. There is no attitude change (although a very slight nose up is possible) and the run on speed much more wind dependant. The idea behind this was that if you were to hit something close to the bottom of your EOL, 40 kts might be more forgiving than 70 kts! These were mandatory practice every month as well; to an observer on the ground they look almost like a vertical descent with a very high RoD. Come to Tern Hill to watch on a windy day! PFLs were also practiced at night (although without shutting the engine down) to a flare recovery at an airfield T (an arrangement of lights on the ground).
However, whilst these techniques were practiced on twins they were never flown to the EOL but only to the flare or overshoot. After all, two engines will never fail at the same time (so they said, until I found out they could!), or simulators are/ were available as well.
Now the conjecture bit.
Dave was presented with something catastrophic (gearbox, transmission or whatever) over a heavily built up area and enters autorotation. He might well have gone straight for the min speed technique (old training) and concentrated on making for an apparent dark area away from traffic etc. To observers on the ground it looks almost vertical, and if turning/manoeuvring may appear to be 'tumbling' as well. He might have had time to pull the Fuel shut off levers (hence the 'popping' sounds) and is also watching the rad alt to confirm his check height. The rad alt is reading from river/ground level until just approaching the LS, when it suddenly reads much lower (roof height) and he has to apply a lot more collective than anticipated to try to stop the descent so perhaps they hit the roof harder than planned. However, the rotors will then stop pretty quickly. After that events deteriorate rapidly.
I stress conjecture. I personally cannot see any way a helicopter can have the rotors stop in flight or enter a tumbling descent without some parts of it falling off / being ejected / completely breaking up - but again there may be a way this could happen that someone could explain. The only other anomalies to me are the lack of a radio call and lack of landing light.
My gut feeling is the aircraft was under some control before touchdown, and Dave would have used all his skill to achieve a safe arrival: unfortunately they didn't walk away from this one.
Last edited by Jetdriver; 5th Dec 2013 at 16:13.