FDMII,

An excellent summary.

:D:D:D

We also can trace flight envelope protection back to technology such as Q-feel systems, cracking struts and stick pushers; all features of pre-FBW flight control systems that limited a pilot's 'authority' to break an aircraft. Some were controversial when introduced, but are old hat now...

@FDMII, neither me intend to be instructive or to initiate a polemic.
The hydraulic fluid does not interpret the pilot inputs as flight computers determine the actuators movement

Maybe the active sidestick
If the pilot releases the stick, the aircraft keeps the pitch until the stick is moved again, then the flight computer moves the actuators without pilot input whatsoever. Also pitch control provides automatic elevator to compensate in turns "You just roll the aircraft and leave it alone"
Airbus said long time ago: "Flying the fly-by wire is similar to driving a stagecoach. Where the coach driver gives a command and the horses take care of the road, the A320 pilot makes a control input and the aircraft takes care of the flight-path. Having been educated in all the permissible speeds, attitudes, and manoeuvres, the A320 will never overspeed, overload, or stall"

Unfortunately you have not been able to quote a single document to state :
"Never pull full back stick in alternate"

Why applying full back stick in Alternate law which has the same pitch control than Normal law ?
Because flight experience has revealed that an immediate 2.5 g reaction provides larger obstacle clearance, than a hesitant and delayed high G Load maneuver.

_Phoenix;

Yes, agree - for FBW systems the computers determine actuator movement through both FBW & Airbus protections.

The action/result of the flight controls is transparent to flight crews in the sense that, (hopefully without conflating the description!), something like CWS, control-wheel-steering in Boeings and the Lockheed, the airplane stays in the last set position, albeit, for the Airbus, without the control surface movement to maintain 1g. Minor corrections to pitch & roll attitudes during manoeuvering are natural pilot inputs.

The feedback/feel of the SS would be the same for both sticks as a function of the design of the SS mechanism itself, which provides the required amount of pull to maintain altitude when bank-angle is > 33°.

To your third, again agreed, (same reasons as your first). But I would hasten to observe that informal, (marketing) remarks made by Airbus a "long time ago" are not in, or part of the AOM and FCTM, nor are they part of formal training curriculae. I would emphasize that understanding the EFCS of the airplane through the usual training & checking regimes is key, just as it is with any type and design.

The AOM and FCTM cautions when in Alternate or Direct law are abundant and clear: protections are degraded or lost: in all ways, fly the airplane as one would a conventional transport aircraft, respecting all airframe limitations, cautions and warnings.

FBW
 

Surely the argument of computers being between pilot and control surfaces has long vanished from relevance. Many modern aircraft cannot be directly hand flown as they are aerodynically unstable by design e.g. f117. The days of real direct control are gone, transport pilots no longer require bulging biceps to execute emergency flight control inputs.

@xcitation;

In my view the discussion regarding computers-flying-airplanes was over decades ago.

However, the subject keeps re-surfacing perhaps from those who don't fly transports or who do but haven't been trained on and haven't flown the Airbus.

Understanding is important and a collegial discussion on this forum may provide a way for all to gain an insight into FBW/automation issues.

Equally important is the reading and study of accidents, beginning with the reports. I think that makes a well-rounded airline pilot at a time when such well-roundedness, including airmanship is, in my perception, actively discouraged in favour of a box-ticking-for-audit-purposes culture and jammed-full simulator scripting.

From excite:

I must take exception, as I flew the first operational jet intentionally designed to have negative static stability belore 0.95M, and was fully FBW with zero mechanical anything connected to the control surfaces. You can see my bio on the info for we posters, unlike 99% of those who post here.

The Concorde prolly led us, but still had a few cables or whatever, maybe tabs controlled by cables.

Even the 'bus is not unstable, tho it does allow a cee gee back further than would be approved 40 years ago.

I cannot find a single commercial jet or more than one ot two military designs that cannot be flown with out cosmic computers, protections and such.

Although the computers or "boxes" shape the signals to the control surfaces, being hydraulic pressure or electronic waveforms, most jets fly today exactly as they did 60 years ago. So my jets 50 years ago used hydraulics, but the valves had shapes that did not provide a 1:1 pressure change for each pound/degree of displacement. Those early jets also had neat dampers that used gyros and/or rate sensors to dampen things. So dutch roll was not a biggie for we clueless jocks.

We even had "protections"!! No kidding. My Voodoo back in the 60's had an AoA limiter and then a 28 pound pusher on the stick if we exceeded 'X" Aoa or a certain pitch rate closse to the AoA limit. Sucker worked like a charm, although that beast provided awesome warning before the stall/pitchup - wing rock, hard buffet, etc. No roll protections, but what the hell. On other jets we had stick shakers or duddr vibrators or......

So the "manual" advocvate folks here have it close, but I am not convinced many of the nuggets have their basic skills burned-in by hours of practice and exposure to the dark side of things in their planes.

The practice of "engage otto" at 300 feet with gear up and the reverse when landing bugs me beyond anything. I don't feel all should be Chuck Yeager, but if we can't fly the profile without using otto, then we should not be flying at all. I used and expounded otto for a few thousand hours in single-seaters to reduce workload when wx or other problems arose. But I never flew a profile that I had to use otto versus my own skills/eperience, and remember that otto has ZERO experience.

out

Indeed!

Designers have used the best available means to make aircraft as easy as possible to handle, ever since the earliest days, whilst also allowing progressively higher performance. That's led successively to refinements in reversible (purely manual) control system with spring tabs, balance tabs, bob weights and the like; servo controls; and then the introduction of powered flight controls and associated artificial feel.

The nirvana is surely something such as described in the second quote above (although it's typically the flight path that's kept, not the pitch attitude as such), in which the pilot directs the flight path with the stick and the flight control system removes external and unwanted disturbances such as turbulence. This is what modern FBW seeks to achieve, but clearly it has to move the waggly bits on the wings and tail to do so, without a direct input from the pilot. This is nothing new; we've had yaw dampers moving the rudder without pilot command for over half a century!

FBW gives the opportunity to remove secondary effects such as (for example) roll due to sideslip, thus making tasks such as crosswind landings easier. Designers can now get much closer to providing pilots with handling qualities that they should have had decades ago, had the ability to provide them existed.

Turning to degraded modes; well, yes... If a Comet or 707 lost yaw damping, then the pilot had a very significantly greater problem on his/her hands than being in Alternate Law in an Airbus. And you'd better be careful not to reach the stall AoA in a VC-10, 1-11 or the like with a failed stick pusher.

FBW is still relatively new technology in the big scheme of things. There are rough edges still, to be sure, and it's being refined just as previous generations of flight control systems were. But the fundamental principles are right.

P.S. Sorry - some duplication with Gums' post, with which I agree completely.

With three Airbii in the water after something resembling a "full aft stick stall" (T-38 stuff) and a Colgan Q400 crashing after a similar stunt; I am bugles by the pilot reaction to yank BACK. Where are they learning this?

GF

It's called "normal human reaction".

Well, the first (XL) crash was more of a mouse trap where defective AOA data allowed the aircraft to slow to a stall (trimming as it went) and then at the stall, dropped into a mode that required manual trim (probably without adequate warning to the crew) so that as the aircraft accelerated under TOGA thrust, the nose climbed uncontrollably, stalled, recovered only to fly into the water due to inadequate recovery altitude. The crew was actually pushing as hard as they could. I wouldn't count that one.

AF447 was the first piloting failure where the guy flying lost control of pitch and spent a lot of time with too much aft stick.

QZ8501 is now the second of these strange aft stick events where the pilot flying lost the bubble and ended up holding in aft stick where it wasn't needed.

In both of these accidents, they started with a roll PIO which is an extremely concerning situation to the pilot. I cannot state that I know a reason that this would cause a pilot to pull full back stick, but we now have two Airbus instances of unexplained inappropriate aft stick plus the Colgan accident.

It could be something as simple as an inadvertent reaction by a pilot who's scan has broken completely and who is trying to hold his posterior in the seat in a manner to which he is accustomed.

Somebody had better be doing some basic research PDQ on human response to better understand the reasons for this "Pull the Stick" phenomena. Then we can actually fix the problem.

I have no gripe with Airbus Normal Law. What I have real problems with is the transition to Roll Direct as part of Alternate Law.

Letting an unsuspecting pilot start flying in Alternate law with Roll Direct before he understands fully the implications of that configuration, is asking for trouble. If a pilot is startled into making a large initial correction or if he naturally pilots with large stick motions, he is very likely to set up a roll PIO. Roll Direct requires gentle control motions for success, particularly at cruise altitude where damping is low. Roll PIO destroys the trust that a pilot has in his flight control system and (empirically) can fully saturate a weak pilot's scan in seconds.

There is a lot that can be done to make the transition to Alternate Law easier. I would hope that Airbus will give that area some study.

peekay4, reaction to what?
Break it down to its elements.
What is the pilot reacting to that causes the "pull back" reflex?

In normal flight, pull back stick == plane climbs up.

When plane plunges down, instinctive reaction == pull back stick (to make plane go back up).

All pilots do this almost sub-consciously. Feel plane sink a little? Pull back the stick a little to compensate. In 99.99999% of the cases this works out for the pilot, i.e., in normal flight with sufficient power and nominal AoA. (And especially with a working autothrottle).

Of course we all "know" that this "doesn't work" during a stall -- or even near a stall (area of reverse command).

Thus the problem with surprise stalls is that the pilot must overcome his/her instinctive reaction (which until this point has worked 100% of the time) and do the complete opposite: push the stick down when the plane is already going down.

To summarize:

• Pushing the stick down is an "intellectual" reaction. It requires knowledge that the plane is in a stall, and recall of the correct stall recovery procedure.
• Pulling the stick back is a "instinctive" reaction. It is based on "muscle memory" from thousands of hours of normal operations.

There is a human-factor safety theory that says that if you have to "remember" to do something different (opposite) during an emergency vs. normal ops, then you will always have a high risk of failures.

Training does help, but only to a certain extent. Training happens at most a few times a year. Yet pilots "pull the stick back to climb" on every single flight. That's "training" as well (from a neuroscience perspective).

I'm sure every single pilot familiar with AF447 and QZ8501 has concluded that "this can NEVER happen to me!" but the truth is, when you are startled and confused, it's hard to fight instinct.

Imagine it's year 2100. AirBoeingBus 78320-ER somehow gets into a stall and starts losing altitude. The Pilot, last of her kind, pulls back the stick, to go back up. Autopilot-G (for George) recognizes the stall, automatically lowers the nose to reduce AoA, recovers from the stall, and then pitches up until the plane climbs -- as commanded. In this case, the pilot does nothing different between normal and stall condition. From a pure human-factors perspective, this is the preferable system behavior.

Edit: to add, even in the non-stall case, when a plane is rolled with nose down attitude, the instinctive (and incorrect) upset recovery action is for the pilot to simultaneously apply opposite roll and pull back on the stick. This is in part why UPRT training is being mandated.

From Gums:... but I am not convinced many of the nuggets have their basic skills burned-in by hours of practice and exposure to the dark side of things in their planes.
The practice of "engage otto" at 300 feet with gear up and the reverse when landing bugs me beyond anything. I don't feel all should be Chuck Yeager, but if we can't fly the profile without using otto, then we should not be flying at all. But I never flew a profile that I had to use otto versus my own skills/eperience, and remember that otto has ZERO experience.

I find it amusing that many interviews still include a sim check where it is common to fly a basic IR profile without FD or A/T. Then, in that same airline you are, near as damn it, forbidden do such such a dangerous thing again. Same goes for base training: a manually flown circuit with only Mk1 eyeball guidance, but don't let them catch you doing such a heinous thing on line. If you can't fly a procedural ILS raw data, including a hold and reversal turn to finals ILS and GA, you should not be there. Computers fail and must be able to maintain SA and cope with achieving a safe arrival.

Peekay4: some interesting observations and comments about human reactions. It is an old chestnut that to fly an a/c you "push on stick = houses get bigger: pull back on stick = houses get smaller: keep pulling back = houses get bigger again."
We all know that, but you went on to say that to do the opposite of what is instinctive you need to know e.g. you are in a stall. There has been much discussion about this on this thread and Tech Log (FD in a stall) thread. Is the stall warning on AB, or any a/c, sufficient to make your training kick in and combat the panic? I know nothing of AB warnings, but the Boeing Stick Shaker is unmissable.
In old tech a/c, at a stall there was often a stick shaker and a pusher/nudger. If you/A.P hadn't trimmed too much you could even just let go and the a/c would possibly lower the nose slightly by itself. What happens with FBW at the stall if you let go? Does it hold the attitude à la CWS?
From the human factors point of view what I think needs considering: in AF & QZ they had an enormous amount of air beneath them; perhaps time to let go and start analysing, but they didn't. (in my world of paragliders it is often the pilot, in a panic, who makes UAR worse. Often letting go allows the aerodynamics to start the recovery process.) In 'close to the ground stalls' the training has a much greater difficulty combating the panic of rapidly rising ground. Similarly it would be difficult, without extreme training and practice, to know when in a car incident it is better to use power rather than brakes.
However, one would hope, that an unmissable STALL visual AND audible warning + stick shaker (tactile) would not be ignored and training would kick in. That training involves use of stick, thrust levers, attention to instruments. We know that under stress your ears shut down, hence the addition of a powerful visual warning. Designers could also include that the a/c will attempt to recover if you just let go.
Those with more knowledge of current FBW actions and future designs could share it with us.

… ‘zero experience autopilots’.
One view of the problems being discussed is that the level of human experience is decreasing and the opportunities for gaining experience have significantly reduced.
Many accidents suggest ‘zero experience pilots’ in the specific conditions.
More manual flight might aid the physical skills, but without encountering demanding, rare situations, little or no cognitive experience is gained – that which helps understand a situation before acting.
A more beneficial approach to safety might be to consider the precursor situations to ‘those which have never been experienced’, where appropriate awareness and action might avoid the upsets.
Are the recent accidents also an indication that pilots’ experience levels of the precursor situations are also weak?


“What is the pilot reacting to that causes the "pull back" reflex?”
‘It’ would be based on whatever the pilot perceives, which with the experiences from memory, provide an awareness of the situation. Thus the actions would be ‘normal’ for what was perceived – doing what is always done or what is believed to be required in a situation which has never been encountered before

Many people present arguments for changing 'the aircraft', but these are based on probable and inconclusive evidence; the result is a weak argument particularly as it is impossible to determine what the pilots perceived.
The alternative argument – no change – based on what happens in every day operations, might be similarly hampered by not knowing. However, it is strengthened by the overwhelming numerical success of safe operation and some ability to enquire what pilots perceive. This should not be seen as a winning argument – complacency, but an opportunity to consider what is not known in both normal and accident scenarios.

We should not conclude that if every flight ends safely we are safe, but equally we should not label all accidents as being ‘unsafe’ because of … , because we just don’t know.
‘Unsafe’ is our judgement of the outcome; for the accident crews it’s quite possible that they believed what they were doing was ‘safe’, but … we don’t know.
The various arguments are presented on the basis of not knowing which is immensely frustrating both for debaters and audience.

RAT 5;
It would take the knowledge of an aerodynamicist who knows C* laws to discuss the "letting go" scenario.

My thoughts as a former transport pilot are that in a fully-developed stall, (35°+ AoA, vertical speed > 10,000fpm), just releasing the controls to neutral would not recover the aircraft, FBW or no.

Bear in mind that all aircraft would have different responses; whether such differences make a difference in handling during a full-stall is one of many unanswered questions.

A full nose-down stick command, wings level, thrust at idle is almost certainly a recover strategy, but no one's done it in flight tests and the sim data, post-stall, is extrapolation from best-knowledge as there is no basis in testing - yet.

The flight control computers for the Airbus attempt to maintain 1g, (Nz, in relation to the longitudinal axis in Normal and Alternate I Laws) and the trim horizontal stabilizer, (THS) autotrim function attempts to unload the elevator, but the follow-up is not instantaneous.

In AF447, the THS was driven full NU and remained there because of full-up stick inputs. If the stick had been placed full-forward, the THS would follow-up and return close to a neutral setting.

From discussions with others here and elsewhere, it is likely the elevator retained sufficient effectiveness to get the nose down, unload the wing and recover.

Peekay,

The absolute last thought I'd have with a stall warning is to pull back! Unload the wing FIRST. This is pretty basic flying. Also, basic training is not panicking.

Then again, I have a couple of thousand hours of fast jet time.

GF

Peekay, If we have pilots flying who when stressed, revert to that simplistic mode of thinking, then we have unqualified pilots flying aircraft! Period.

I graduated from the USN fast jet school of training where I learned to overcome the sensation of sitting in a waste basket while I flew instruments and to fly intellectually-because your physical sensations were not to be trusted. Later, as an instructor, I would have washed out any pilot that could not measure up to this basic standard. (They would only have gone on to kill themselves and others.)

If we cannot properly qualify pilots to actually take direct control of an aircraft, then we had darned well better not be putting those guys occupying the seats with the best view into that position. The airframe folks are going to have to ensure that their aircraft never gives up to the point where it has to pass highly degraded control to the "pilots". There is a body of evidence building to suggest that an improper transition to Airbus Alternate Law-Roll Direct can become a highly degraded condition. Do we need more accidents to prove the thesis?

The thesis requires proving, not that way of course but through the usual examination, data, studies then statements, etc., and, as part of the process, needs challenging by those who have the knowledge and experience to contribute to the thesis.

At present there is almost no research available regarding human factors and the actual transition from 'normal', (which is a protected airplane), to Alternate I/II, and possibly to Direct Law depending upon system losses, (a conventional airplane in theory and, I have discovered, in practise).

Such absence in and of itself may speak to the question - it doesn't appear to constitute a primary problem, particularly for those experienced in manual flight. I've flown the aircraft at cruise altitudes many times - the A320, A330 & A340 fly just like any other transport, and as you'd know, you must absolutely be gentle with the controls.

However, inexperience, need-to-know knowledge levels, and low or absence of exposure (the case in both these accidents) to abnormal attitudes does not leave much 'resourcefulness-in-reserve' when faced with such circumstances, especially for those entering the profession now. I think there may be something to the apparent sensitivity of the controls which speaks to training issues regarding handling the stick.

Nor is there anything informal regarding the transition. Those who have trained-on / flown the Airbus do not seem to find the transition (from Normal to Alternate I/II / Direct) problematic, nor are two samples in millions of flights, statistically significant.

I think what alf5071h has to say under "zero-experience autopilots", above, would be a more rewarding area of examination.

I graduated from the USN fast jet school of training where I learned to overcome the sensation of sitting in a waste basket while I flew instruments and to fly intellectually-because your physical sensations were not to be trusted

Did you earn to fly limited panel IMC? Do students still at CPL still learn that skill? It certainly gave one an appreciation of what was going on and how to control it. It also taught you not to panic when you lost what had previously been perceived as a vital parameter. You calmly used other information to maintain control. If it has been removed from basic flying school IMC training then IMHO it should be re-introduced.
Has anyone attempted this in a modern jet? The worst I was ever given in recent recurrency was an FMC failure or flight on SBY (EFIS) display. Neither a big deal. An ILS with total unreliable airspeed is a good one, but rarely trained. As a trainer my dictated syllabus was very basic when displaying unreliable airspeed scenarios. A nice tick in the box for 3.4 system failures, but as a training exercise to 'save the day' when it happened for real, not so useful.
One of the best I had in a B757 sim was to fly an ILS on full old fashioned SBY displays. It became even better when the RMI went awol and we had to use the old fashioned compass. Again no big deal if you used the basics.

The absolute last thought I'd have with a stall warning is to pull back! Unload the wing FIRST. This is pretty basic flying.

I have to confess that when teaching stick shaker stall recovery I questioned the FCTM. It said "apply power & reduce attitude." It didn't say anything about minimum height loss. That was an exercise 'if ground contact was a factor'. I stressed to reverse the wording might be better, especially in underslung engines, to 'reduce attitude and increase thrust.' Split second, even both at the same time, but elevator leads the way. To me aerodynamics had not changed so why pretend it had. Beforehand I saw guys apply the thrust and then go closer to a full stall than they were. i.e. they made it worse before they made it better. After THY at AMS guess what happened. Aerodynamics and common sense won the day and FCTM was changed to UNSTALL the wing first.

One view of the problems being discussed is that the level of human experience is decreasing and the opportunities for gaining experience have significantly reduced.

Great comment, and then ask why & how? Let's get to the root cause of this demise and start to solve the problem there, not massage it and put a few sticking plasters on a gaping wound. It may need an operation or stitches, but there needs to be a serious reasoned debate from all parties. There appears to be a lack of leadership on that. Many interested parties are spouting words of wisdom, and some even showing disinterest. Will it be EASA, an XAA, ECA, IALPA, Boeing, AB, a government, an airline group, anybody, but it needs somebody to lead. There is a real problem and we are perhaps seeing the tip of an iceberg. A/C will evolve into more computerised sophisticated beasts; MPL cadets will be less experienced and thrust into RHS of said jets; companies will expand and give commands with less experience to meet that expansion; cadets will pay for their training and grab any job with any T's & C's they can find; rigid SOP's will evolve to reduce piloting skills. So how can the industry protect itself from a declining spiral of skills that could hurt it? We already know that there are airlines from certain regions that certain pax avoid due to suspicion of standards. We want to reverse that trend not allow it expand.