We have been hearing about this since AF447 (which apparently wasn’t any kind of a deep stall anyway). Now that Air Asia is another.
Apparently, every aerofoil has two points where slow stable flight can be achieved (if you have the power). One at around 10-15 deg AOA, and the other at around 45 deg AOA.
This is nothing to do with T-tails getting their airflow obscured by the wing, which is a separate issue and AIUI is not a simple loss of airflow but is caused by the T-tail getting “stuck” in a portion of the wake from which any attempted exit increases the elevator efficiency so much that an exit is impossible using elevator authority alone.
One can supposedly recover by tipping the aircraft on its side, retrimming like crazy, and once speed builds up (in what becomes a spiral dive) you can sort things out. AF447 could have been recovered that way, by a pilot with the knowledge and big balls. But that needs functioning ailerons, or you use the rudder.
The problem with a ‘stable’ flight at very high AoA is not only with having sufficient power; in this zone of the power curve, most aircraft become very unstable laterally and very hard to control in the roll and yaw channels, hence the spin or a falling-leaf stall. Notable exceptions are Socata Rallye and Antonov AN-2. Alternatively, modern fly-by-wire aircraft can be artificially stabilised in the roll channel under these conditions, but it’s only fighters that are designed with that in mind. It’s not only a T-tail that can get into an aerodynamic shadow – in fact, all aerodynamic controls lose efficiency and authority to different extent, so that recovering a large aircraft from a deep stall may either require the skills of a test pilot, or involve a huge altitude loss (e.g. China Airlines’ B747SP lost almost 30,000 feet to recover from a spin). Different aircraft may require different recovery procedures, too.
AFAIK (and from what a LH 330/340 captain told me) it was indeed a deep stall, and the recovery procedure is practiced in the LH SIM regularly. I’ll try to call him today and ask him how exactly they do it.
it was indeed a deep stall, and the recovery procedure is practiced in the LH SIM regularly
That would be very interesting to know – do they really have a realistic simulation of deep stall behaviour? Until very recently, even the most advanced simulators couldn’t do that.
I find this sim scenario interesting. AFAIK there are no data points to program the sim with. This comes up time and again in discussions about the two referenced accidents over on Pprune.
My feeling is that the real issue lies with the design philosophy of Airbus. Not giving any feedback as to the side stick, throttle or trim is, IMHO, madness.
I wouldn’t want to have to fly an airplane that puts itself into some weird configuration and then hands itself back to me screaming without really telling me what it’s up to.
Personally, if I have a choice as pax, I avoid these things.
It was more the lack of the crew working together the right way than anything else. Similar accidents have happened in Boeing aircraft. Either you know how to recover from such an attitude – or you don’t. Also if the captain or copilot presses the priority button on the stick he is in control of the airplane.
It’s maybe a bold statement but my feeling is that THIS would not have happened in a Lufthansa plane
I do not believe most simulators accurately reflect aerodynamics at high altitude, but they do at high AoA. There is of course a limit for all these things, they are not trying to teach you to be a test pilot, so maybe deep stall is not covered. It’s beyond my area anyway, as it’s really more of a swept wing phenomenon. If Flyer59 knows Lufthansa people that do it then it would seem to be worthwhile, LH will not waste time on unrealistic training
I think both the FAA and EASA have documented standards for simulators, I have been there when the FAA came and did a routine test on a sim and it was extensive work.
A very similar crash occurred in 2006 with Aeroflot’s TU-154M – the crew decided to hop over a CB and climbed very sharply from FL360 to FL420, stalled and broke into a flat spin from which they couldn’t recover. After that, several Russian airlines decided to introduce periodic stall/spin training for their crews, but instead of a simulator they used a military jet trainer (a Czech L39, to be specific). Don’t know if they are still doing it, though.
and the other at around 45 deg AOA.
Peter, have not come across this before? Once flow separation and turbulent wake sets in am not sure it can be re-attached by increasing alpha, unless through some theoretical boundary layer control mechanism? or blended wing body theoretical designs?
It would be interesting to know whether climbing in convective weather took these aircraft into buffet onset territory, or the crew managed to stall the aircraft while still being at a safe altitude.
I think the reason a wing produces “lift” at 45deg AOA is because it displaces so much air downwards
If you stick your hand out of a car window at 70mph, at 45 degrees, there will be a lot of lift for sure. A huge drag too but we aren’t talking about that if you have the power
After all, lift is produced by displacement of the air, downwards. A barn door will do that. An aerofoil just does it nicely, with relatively little drag.