Test stalled sure - but at or close to service ceiling? With a simulated full load, an imperfect trim with rearward CoG and perhaps some loss of situational awareness? (ie in a thunderstorm at night, rather than a clear day).


I fully appreciate the economics and the "tombstone imperative", however there are a growing number of the latter and what is clear from this thread is that there are few if any with any clear idea of what happens when a FBW aircraft enters coffin corner and then finds the corner. The sim is no help if its working with duff data. Passing a sim check or even playing with the sim in some rare free sim time is no bloody good if the real thing is different.

If, in high altitude stall test, for real with a test pilot(s) on board, they cannot get the AOA reduced to recover from the stall there is another option as a very last resort. Just to save the test crew.

Why not have a drone parachute attached to the rear of the fuselage? On deployment, the drag from the chute would certainly lower the nose, by raising the tail to a sufficient degree for recovery and could be cut away when or if it has done it's job.

One would think this has been thought of by much smarter people than I. However upon learning these aircraft haven't been (intentionally) stalled, leave alone deep stalled at high altitude, one can only wonder at the mentality. The uncrashable/unstallable mindset, does give it a way a bit.

@captains log - reports due
 

As far as I can read and understand it ...

Yes it is already out at the end of the 1 month period set out by ICAO ... and sent to ICAO and parties involved ... but certainly not made public ... in spite of the expectation of some pprune posters.

KNKT/NTSC have literally stated (look at previous posts - including a translation of the KNKT interview video) that they 'have the key' to the accident.

KNKT suggested that they would have a press conference later divulging more information. Later that changed to - it would be adressed during an already scheduled annual KNKT report presentation - but it was not (as far as I could see - even digging deep in local Indonesian press reporting).

Most public information is from a parliamentary commission presentation of which fragments can be found on-line.

Quote:

---

Due to the fact they were keen(maybe not he best word) to cancel the majority of the salvage operations, surely the investigation team have a clear picture with the aid of FDR/CVR of what happened?

---

The salvage as such already faltered before they had had time to read the flight recorders. So I find that cause-effect highly unlikely.

Quote:

---

If so what's with the delay?

---

I asked that recently. Nobody knows apparently. And most surprising, nobody appears to be upset about that.

Interesting is that originally KNKT stated they expected the final report in 1 year. Recently statements have reduced that to 8 months.

For an innocent bystander it is quite confusing. The state quite fast that they have 'all' the answers. But will publish quite late ... not having a Prelim ... we have to wait at least 8 months for a Final ... which by definition will have a Public version. It would be nice to get some public KNKT explanation of course.

I don’t think that’s as important as people make out. The detailed response of an airframe to “edge of the envelope” events in real life is often highly dependant on things like CG, yaw/roll/pitch angles, asymmetry, entry rate, wing loading, air density and many other factors. Every one is different, sometimes remarkably so.

What needs to be trained is:

1. High-altitude stalls are BAD, don’t go there.
2. If you do end up there, it will most likely need control deflections and pitch angles that you’ve never used/seen before to get out of it.

If the sim is up to replicating that, then it’s good enough. We are trying for a generic recovery procedure, not a specific finessed one.

From an interview I saw with an airbus test pilot, he said that they only test a stall close to the edge of the envelope with high energy. The AF447 and air asia stall are far more lethal because the a/c has lost nearly all energy and have the engines at idle. This requires a lot more time to recover the lost energy. Clearly there is a huge difference between a best case stall and a worst case stall. I think it reasonable to want the ejector seat for a Chuck Yeager willing to take an a/c so close to unrecoverable.

Don't know. Will it rip the rear pressure bulkhead off, or just pull the tail away when it's unfurled?

He was only talking about doing this to allow test flights to be carried out safely, so it wouldn't be a problem - the test aircraft's structure would be reinforced as necessary. This method is routinely used for fast jet spin testing (although obviously you do also have the benefit of an ejection seat if things don't work out in those circumstances). An alternative would be something along the lines that Airbus have already used for A400M stall testing, a downward-firing rocket in the tail.

I make no comment on whether any of this is necessary, merely observing that it's readily available and deployed technology within the flight test community.

Surprised by ASC releasing DFDR readout quickly - yes. Surprised by NTSC - well, not exactly, bearing in mid the outcome of MI185 investigation. Regarding the country of registration and nationality of the captain, stakes are far higher now but DFDR and CVR are read out so evasion will be more difficult.

It is possible that recorders' readouts prompted NTSC to jump the gun.

Whut?
 

xcitation
Exactly what is a "stall close to the edge of the envelope with high energy"? I'm familiar with accelerated stalls, power-on stalls, power-off stalls, PA and CR configuration stalls, etc. but that's a new one to me.

Energy Management
 

My understanding is that kinetic energy is the work done to accelerate the mass of the a/c. This is increased by engine thrust or trading potential energy (losing altitude and claiming the acceleration from gravity). If you have low PE and low KE then total energy is low which is very dangerous.

Update: Boeing have a good description here...Aerodynamic Principles of Large-Airplane Upsets (Aero Magazine, 1998 Q3 edition)

It's a while since I did A Level Physics but here goes:-

So the higher you are (above earth) the more the potential energy?

KE = 0.5 m (V)squared - where m is mass and V is speed. Double the speed and you quadruple the energy.

X citation and others

Essentially what is being inferred here is that stall testing is done with a lot of potential energy. ie at a high altitude and not close to the ground.
In this context the comments (some of which have been allowed to stay) make sense. One does not practise stalls close to the ground even in a c152!

However your later comments re selecting flight idle and equating that with low energy are incorrect. For stall recovery the sequence has to be reduce the AofA then a short while later increase power. Selecting toga or equivalent merely causes a pitch up which makes reducing the AofA almost impossible.
Increasing the power does not increase'the energy state of the aircraft'.
In a stall situation you do not go faster nor do you increase altitude, you merely descend in a stable stall like AF.

Are they? Scale effects are expressed in the Reynolds number, which considers viscosity. Combine that with Mach number, which considers compressibility. Then you'll see the need for full-scale testing.

To clarify my comment about idle detent; N1 take time to spool up when changing from prolonged idle to TOGA. This is another consideration when a/c is falling out of the sky and flight control is urgently needed e.g. Habsheim incident - low total energy, low N1. The impression I got from the airbus test pilot was that stall testing is done in a very controlled manner changing flight controls as little as possible. This protocol allows the a/c to return to the flight envelope very quickly with only small stall excursions. It's not cutting engines, full stick back, see what happens.

From an image of wreckage, it seems that the screw was snatched during the impact. It remained attached to THS
http://www.mediafire.com/view/1hh6qg.../jackscrew.jpg

Avoid stalling your swept wing jet aircraft
 

Risk avoidance will minimise the number of pprune threads like this one. The multiple factors that conspire to place a crew in risky situations need to be considered by all pilots. By the look of the number of accidents like this one, not all of us spend some quiet time at home thinking about the various scenarios and the choices that might be available to minimise risk.
It is OK to consider and practice in the sim for recovery procedures under difficult circumstances, but wouldn't it be heaps better to recognise the risk? Finally it is up to the captain, but the F/O is not in a comfortable position as the risk escalates. Some captains will have their own reasons why staying at F390 in increasing moderate turbulance is acceptable in an A320. A lower time F/O might accept an insufficient reason, like ATC clearance, reduced endurance, possibility of diversion or increased turbulence at lower levels.
I see no indication on these pages that companies are pressing ahead with avoidance training. Rather there seems to be so much discussion on recovery.

Xcitation
I take your point re testing in a very controlled manner, however when a real and unexpected stall occurs on the line, those involved have moved very far away from the contrlled environment of testing.
In the Bournemouth incident both crew were pushing full forward on the control column, but were unable to prevent the pitch increasing to arround 45 degrees. My point being that large control input is likely to be very necessary in order to be able to reduce aofa.
Your example proves how far removed real life can be from testing. No one flying on the line wants to go near the stall, but if they do it will be at the most inconvenient momoment and completely unexpected. Having a clear understanding of what to do is vital. Avoidance is the first step then recovery the second.

I must caution you here: Discussion on PPRuNe bears no resemblance to what companies are pressing ahead with.

Professional pilots are fascinated by other's mistakes (and so they should be) because that's how they learn not to repeat them. The reason threads like these go for hundreds of pages is that professional pilots see a colleague suffer a disaster and they want to know why.

In the impatience of awaiting the final report, they suggest solutions, but most of these solutions come from their creative thinking, not from research of what "those in charge of policy/training/manufacture" are actually doing about it. It's really annoying because those who are actually involved in improving policy/training/manufacture don't jump on PPRuNe every day with an update. Don't they understand the frustration this causes?

However, in direct answer to your question, "avoidance training" is not really something that can be taught in a simulator, or via newsletters/memos, or via flight manuals. This is learned on the line with experience. The only way to learn the best way to fly around thunderstorms is to spend years flying around them under the guidance of someone who has flown around more of them than you have.

Ideally, you fly in the RHS (or in some airlines as a relief pilot/cruise F/O) for many years and learn the avoidance strategies by being exposed to them over many years, with on-the-job guidance and advice from a more experienced pilot in the LHS.

Unfortunately, it doesn't seem to always pan out this way these days. For example, cadetships leading to rapid command promotion in rapidly expanding LCCs, by definition, jump this stage.

The old adage: you start with a full bag of luck and an empty bag of experience. Try to fill the bag of experience before your bag of luck runs out.

But that does not appear to be a factor in this particular accident - the F/O was inexperienced but mature, and the Captain was very experienced. We still don't know why Air Asia pitched up, stalled and stayed stalled. (If indeed that is what happened, we don't know yet.)

The investigators say they know why, but they aren't telling. Again, very frustrating.

In the case of Air France, experience/training was possibly very much a factor, as unfortunately the only chap capable of even recognising the problem (stall) was asleep. The situation may have been unavoidable, the recovery procedure was possibly known to the crew, but they failed to recognise the problem (experience? training?) so did not implement the recovery.

On the contrary, it is companies that have only been doing avoidance training that has led to the current state.

* Anyone flying manually outside the immediate landing or takeoff gets snitched by FOQA and told to desist: the automatics are better than you are and this avoids the risk of something going wrong.
* If this (alert/instrument fail/behavior) happens then memorize this list of precise instructions and follow them. Get the PNF to read through and check that you have done them and nothing else.
* Sit in this reasonable simulation with critical unrealities every month and improve your by rote response to specific (alert/instrument fail/behavior)

These are all risk avoidance by the aircraft operators. To some extent one can understand their point. However, when the real LOC happens it is nothing like the SOPs/FCOM say and usually multiple things happen at once. Unlike the simulator there is negative and positive g, manuals, drinks, grit, EFBs flying everywhere about your head - you cannot even read the instruments and the ones that you have trusted for thousands and thousands of hours are not working. The training to reduce risk may have gone through the motions but the visceral reactions to unexpected g, airstream noises, perhaps even shouts screams and bangs from the passenger cabin. This sudden and extreme stress puts you right at the wrong end of the 'inverted U' ( Yerkes?Dodson law - Wikipedia, the free encyclopedia ) and leads to cognitive/attentional tunneling usually onto an incorrect area. The training in other words has not prepared the crews for what things could get like. But its difficult and expensive and 'how many of our aircraft will have that problem?' So it is not done. This is unfortunate because as people have said here if you actually _have_ experienced negative g, and recovering from an incipient spin _multiple_ times _and_ been scared a few times then you are likely to perform better. The psychologists call this 'stress inoculation', you are put in the situation and shown how to get out of it. Then when it occurs you are not pushed into cognitive decline by the stress and arousal levels.

The current system is built around avoiding nasty problems around the edges of the flight envelope and training people not to go there. The training is done in comfortable shirt sleeve environments without any physical stresses at all. The training and systems have no doubt increased safety but they have had the unfortunate side effect that when an aircraft is in a sudden upset there is a very much higher chance that the crew will not recover the situation. The very risk avoidance approach has increased the risk of crew poor performance with attentional tunneling due to stress and automation surprise at the very time that they become the sole people capable of recovering the situation.

So now take a many thousand hour fighter pilot and put that pilot in a shiny airliner with two underwing engines and go through all the risk averse training ticking all the boxes. That will NOT include stall or spin recoveries "don't even go there!". Then a few thousand hours into flying the airliner go from bumpy but normal flight in early daylight IMC to all the bells alerts cavalry charges and whistles 2 g zoom climb all inside 20 seconds followed (speculation alert) by a port engine compressor stall as the aircraft approaches apogee with full power on starboard - and the edge of the stall turns into a fast rotation into a flat spin driven by the 'good engine'. Hind brain training kicks in and what were the by rote procedures in the fighter aircraft - and they are implemented. However, the procedures for a fighter with roll/yaw diversion due to a weight concentrated in the fuselage is not quite the same as for a twin jet airliner with a dead engine but that is the practiced response and nobody bothered to erase that response with what to do in _this_ aircraft because we "don't even go there!"

So cautious stay out of trouble training needs to be looked at very carefully. As does the reality of what 'trouble' feels like. If crews felt that once - they may be 'inoculated' against the cognitive stress of its occurrence when it happens.

Loud & Very Clear
 

Ian W's post below says it all.


"On the contrary, it is companies that have only been doing avoidance training that has led to the current state.

* Anyone flying manually outside the immediate landing or takeoff gets snitched by FOQA and told to desist: the automatics are better than you are and this avoids the risk of something going wrong.
* If this (alert/instrument fail/behavior) happens then memorize this list of precise instructions and follow them. Get the PNF to read through and check that you have done them and nothing else.
* Sit in this reasonable simulation with critical unrealities every month and improve your by rote response to specific (alert/instrument fail/behavior)

These are all risk avoidance by the aircraft operators. To some extent one can understand their point. However, when the real LOC happens it is nothing like the SOPs/FCOM say and usually multiple things happen at once. Unlike the simulator there is negative and positive g, manuals, drinks, grit, EFBs flying everywhere about your head - you cannot even read the instruments and the ones that you have trusted for thousands and thousands of hours are not working. The training to reduce risk may have gone through the motions but the visceral reactions to unexpected g, airstream noises, perhaps even shouts screams and bangs from the passenger cabin. This sudden and extreme stress puts you right at the wrong end of the 'inverted U' ( Yerkes?Dodson law - Wikipedia, the free encyclopedia ) and leads to cognitive/attentional tunneling usually onto an incorrect area. The training in other words has not prepared the crews for what things could get like. But its difficult and expensive and 'how many of our aircraft will have that problem?' So it is not done. This is unfortunate because as people have said here if you actually _have_ experienced negative g, and recovering from an incipient spin _multiple_ times _and_ been scared a few times then you are likely to perform better. The psychologists call this 'stress inoculation', you are put in the situation and shown how to get out of it. Then when it occurs you are not pushed into cognitive decline by the stress and arousal levels.

The current system is built around avoiding nasty problems around the edges of the flight envelope and training people not to go there. The training is done in comfortable shirt sleeve environments without any physical stresses at all. The training and systems have no doubt increased safety but they have had the unfortunate side effect that when an aircraft is in a sudden upset there is a very much higher chance that the crew will not recover the situation. The very risk avoidance approach has increased the risk of crew poor performance with attentional tunneling due to stress and automation surprise at the very time that they become the sole people capable of recovering the situation.

So now take a many thousand hour fighter pilot and put that pilot in a shiny airliner with two underwing engines and go through all the risk averse training ticking all the boxes. That will NOT include stall or spin recoveries "don't even go there!". Then a few thousand hours into flying the airliner go from bumpy but normal flight in early daylight IMC to all the bells alerts cavalry charges and whistles 2 g zoom climb all inside 20 seconds followed (speculation alert) by a port engine compressor stall as the aircraft approaches apogee with full power on starboard - and the edge of the stall turns into a fast rotation into a flat spin driven by the 'good engine'. Hind brain training kicks in and what were the by rote procedures in the fighter aircraft - and they are implemented. However, the procedures for a fighter with roll/yaw diversion due to a weight concentrated in the fuselage is not quite the same as for a twin jet airliner with a dead engine but that is the practiced response and nobody bothered to erase that response with what to do in _this_ aircraft because we "don't even go there!"

So cautious stay out of trouble training needs to be looked at very carefully. As does the reality of what 'trouble' feels like. If crews felt that once - they may be 'inoculated' against the cognitive stress of its occurrence when it happens."
Ian W is online now Report Post