Pinnacle NTSB Update

[surplus1]

At my 121 carrier, during my initial training, we also did a couple of high-altitude stalls, both a coordinated one, and then an intentionally uncoordinated one, with the yaw damper failed . . . .

Ty,

Not trying to be arrogant or snooty or anything else but your concept here, in my opinion, is endemic to the problems that may have occured on this flight.

The rehersal of stalls in a simulator, whether they be at high altitude or low altitude is not the problem. "Practicing" stalls and how to recover from them is essentially a waste of time in transport category aircraft.

An example of what I'm trying to say is perhaps proven by the "upset/recovery" experiments conducted in Miami some time ago. In that "investigation" some of the most seasoned and experienced line pilots and test pilots were unable to come up with any meaningful "recovery" technique for a jet upset. The only thing they were able to prove conclusively is that upsets must be avoided at all costs. [Some are "recoverable", most are not and the difference is anybody's guess.]

Trying to teach an airline pilot how to recover from a stall or an upset (which usually results from a stall or near-stall) in a simulator is, IMO, a good example of too little too late.

What we need to teach the pilot is how to avoid the conditions that lead to these events, not how to recover from them. In depth understanding of the performance envelope, the aerodynamics, and the effects of wheather plus knowledge of how to consult and use the applicable charts will keep the pilot from getting his aircraft into perilous situations from which there is really only a slim chance of recovery without damage to the aircraft.

Having to deal with a full stall, an upset and a consequential double flameout, decompression, temporary loss of electrical power and God knows what else, in the middle of the night (not to mention the element of total surprise) is a tall order for anyone, regardless of their experience level. When the exprience level is limited to begin with the chance of successful recovery is virtually zero. Chances are they had no idea what happened or why (which might explain the reporting of one engine out initially and the report of the second engine only coming after the loss of more than 25K feet).

There is little chance of learning how to do this in a simulator (let alone a real airplane).

In my opinion the training that was lacking was not in the simulator, it was in the classroom. The comments in this thread about how "easy" it is for this airplane to go to FL410 and why it should be able to do so because it is "certified" to that altitude, clearly reflect that there is little understanding of the principles involved, the resulting limitations or the potential consequences.

This tragedy appears to be a direct result of that lack of understanding. Unless this is corrected, the possibility that it could happen again is very real.

Expecting guidance from the FAA through the POI is yet another unrealistic understanding of what's involved. I've flown for an airline operating jets where the POI had never flown anything larger than a Twin Comanche. That fellow needed to be in charge of a flight school with some Cessnas, not an airline with more than 100 jets. There are some very competent people in the FAA, but there are also some very marginal people that are in the wrong place and in way over their heads. That came about when the ACDO and GADO were combined into the present FSDO. In many cases the supervisory POI's with little experience are assigned to airlines with little experience. That is because they aren't accepted by the experienced carriers and the small outfits (regionals) still have a General Aviation mentality that is afraid to challenge them. That brings us right back to "the blind leading the blind".

When the final report is written on this one we are likely to see the flight crew saddled with the blame for all of this. However, I won't be among the people ready to point the finger at them. In this case I feel they are the "Victims" of a disfunctional system.

Whenever we talk about a pilot who has been killed in a flying accident, we should all keep one thing in mind. He called upon the sum of all his knowledge and made a judgment. He believed in it so strongly that he knowingly bet his life on it. That his judgment was faulty is a tragedy, not stupidity. Every instructor, supervisor, and contemporary who ever spoke to him had an opportunity to influence his judgment, so a little bit of all of us goes with every pilot we lose.

— author unknown​

​

 

[TIS]

Trying to teach an airline pilot how to recover from a stall or an upset (which usually results from a stall or near-stall) in a simulator is, IMO, a good example of too little too late.

Yup, it probably is.

What we need to teach the pilot is how to avoid the conditions that lead to these events, not how to recover from them.

Yup, but that's a ground school issue and we all know that we can't actually be PAYING someone to teach a school that isn't in the air! That isn't the point of air carrier initial and recurrent training!

...There is little chance of learning how to do this in a simulator (let alone a real airplane).

This I need to take some exception to. While I agree that avoidance (just as with thunderstorms, I might add) is clearly the most important underlying approach that must be preached with respect to jet upset/high altitude stalls, I cannot agree that exposure to the realities, even if only simulated, does not provide a beneficial lesson to anyone to which it is administered.

Even if NO ground classroom time is devoted to the subject, a wake turbulence upset that leaves you inverted with the stick shakers rattling in a simulator is a) far better encountered in the sim than in real life and, b) engenders a VERY healthy respect for just how quickly things can get COMPLETELY out of control – particularly when the student is a newbie who has perhaps only ever considered how fast a Baron will roll over in a Vmc situation.

In short, if it does NOTHING else, it promotes respect for the things they now KNOW they don’t know, and makes them believe in the possibility that there are other unknowns that could be just as shocking. Even as a rote lesson, THAT lesson will keep the ill-informed better clear of things that they are not well prepared to cope with. It will also, in time, inspire curiosity and a desire to learn more so that simple passive coping is replaced with active and aggressive command leadership when conditions demand it.

TIS

 

[Ty Webb]

Ty,

Not trying to be arrogant or snooty or anything else but your concept here, in my opinion, is endemic to the problems that may have occured on this flight.

The rehersal of stalls in a simulator, whether they be at high altitude or low altitude is not the problem. "Practicing" stalls and how to recover from them is essentially a waste of time in transport category aircraft.

Although your point is well-taken, I think you are making some assumptions about the training I was describing, which was not a "rehearsed" maneuver. The instructor put the aircraft into a couple of different high-altitude stall, and had us recover from the ensuing upset.

Perfect? No. Better than nothing? Yes. Emphasis should be placed on not getting in that situation in the first place . . . . .. you betcha.

When I learned to fly jets first, in the corporate environment, we did a lot of hand-flying, usually during empty-leg re-positioning flights, and I mean not only down low, but up in cruise, too. When you are hand-flying, it becomes very evident what the airplane can and cannot do, and a temperature change or a little turbulence makes it clear how thin your margins are at the edges of the TAB data. Nothing like doing "brain surgery" at FL410 to give you an appreciation for this, IMO.

 

[surplus1]

I didn't mean to get into an argument guys, just express an opinion. Perhaps I wasn't clear enough. Then again perhaps I'm full of it and dead wrong. My comments are only opinions based on my own experiences.

Let's look at "teaching stall recovery" as an example. In reality, that's stuff and nonsense. Perhaps we can show a pilot how to get out of an approach to a stall, but how to recover? These airplanes have shakers and pushers precisely because their characteristics are so unstable when "stalled" that recovery is an unknown quantity. Noone really knows exactly how the aircraft will respond in a full stall. That is why it has a pusher. On that night apparently the pusher unleashed a bucking bronco that you'd have your hands full trying to put into a syllabus or a simulator's programming.

How do you teach recovery from an upset like that? What is the proper procedure? In fact there is none. So, the instructor would have to invent one. What good is that? Yes, there are ideas on the subject but that's all they are, ideas. There is no proven technique than can be classified as correct or incorrect. That is exactly what the Miami testing proved beyond any doubt. Once you get into that situation you're in the realm of the unknown. You can't teach the unknown.

I'm not against demonstrations of unusual situations, but as soon as they become "instruction" we're treading on highly questionable territory. We know how to avoid so, let's just do it. If we don't get into the condition there is no need to recover from it.

I like your example of a thunderstorm for it validates my point. You can and should teach a pilot how to avoid thunderstorm penetration and what could happen if he does not. However, you can't "teach" him how to fly in a thunderstorm. There just isn't any established fool proof way to do that nor is there anyone who can reliably predict what will happen inside such a storm; everyone of them is different.

As for the idea that I can get into wake turbulence that inverts my airplane on final and learn a procedure of how to recover from that, I see that as hogwash too. Upset a transport at that altitude and you're not going to recover, period.

You can teach me how to avoid CFIT but you can't teach me how to recover from CFIT. It doesn't matter how big or small the moutain may be, hit it and you're history. We aren't flying F-15's and low level aerobatics in airliners will not result in increased longevity.

Do you for instance really think that American can teach its pilots just how much rudder is enough and how much is too much for the AirBus in a simulator? I don't think so. AirBus needed to come clean and tell its customers that the use of rudder in turbulence could cause the vertical fin to separate from the aircraft. Had they done so then yes, AA would have been able to teach its pilots not to use rudder at all in that situation. The people in the AA Training Department are not amateurs. They did not fail to teach this because they were negligent. They didn't teach it because they didn't know it. It's easy to blame the pilot after the fact, but the truth is AirBus failed to provide adequate information about its product. Boeing failed to design a system that would not permit rudder hard-over in the 737 and, just like AirBus, denied it when it did. Could it have been corrected before it happened? Yes, it's being done now isn't it? But, you can hardly blame the pilots who didn't recover or their airline for not teaching them how to.

Who would have thought that one failed tire would bring Concorde to its demise? Some of these things just can't be forseen and we have to live with that. It's not a perfect science. The AirBus, Boeing and Concorde scenarios are all mechanical problems that should have been forseen but were not. Events that you can't train for.

In the accident we are discussing however, there are known quantities and there are effective means for avoiding the ragged edge to which this aircraft appears to have been exposed. The evidence we have to date appears to indicate that the flight crew was unware of the danger into to which they were flying. In my opinion, they could have been taught how to avoid that. Not by demonstrating or practicing what to do after the fact in a simulator, but by learning how not to get there at all. You can teach a pilot how not to forget the landing gear, but practicing gear-up landings in a simulator will not accomplish much.

I just don't believe that they had any idea that they were conducting some sort of flight test on that fatal night. But, in fact it appears they were. One for which it also seems they were completely unprepared. Apart from the aerodynamic problem itself, a dual engine failure was apparently induced, with all of the associated difficulties that we know about.

According to the report of the NTSB, the minimum speed for an airstart was never attained. None of us know why so we can only speculate, but what we do know of the communication with ATC does not appear to indicate an awarness that both engines, not one but both, were inoperative until much later in the sequence of events. We, writing in this forum, do not know what happened between the upset and the report of dual failure made at 13,000 ft, but chances are the investigators do know and will tell us in the final report.

Whatever it may have been, once control of the aircraft had been regained what could and should have happend thereafter can indeed be taught in the simulator. From reading these posts, it appears that perhaps it was not.

Speculation seems to show that what happened before the upset resulted from inadequate training. It also appears that what didn't happen after flight control was regained was also the product of inadequate training.

Somewhere along the way someone is likely to brand this crew with the pilot error label. Technically that may be correct, but based on what I've read so far it will also be unfair. The blame for this scenario was not on board that aircraft that night. If I'm right, it's still on the loose and it won't be fixed with another hour or two in the simulator.

 

[efiscompmon]

When I learned to fly jets first, in the corporate environment, we did a lot of hand-flying, usually during empty-leg re-positioning flights, and I mean not only down low, but up in cruise, too. When you are hand-flying, it becomes very evident what the airplane can and cannot do, and a temperature change or a little turbulence makes it clear how thin your margins are at the edges of the TAB data. Nothing like doing "brain surgery" at FL410 to give you an appreciation for this, IMO.

Sadly, when I was an FO I was strongly discouraged from hand flying the airplane at cruise--flew it by hand MEM-TUL one time and the captain thought it was just amazing. Most wanted "George" to be on above 10,000'. Some wanted it on upon reaching cruise altitude (in the 20's, for the short MEM legs then). Some wanted it on ASAP.

"Speed Mode. Heading Mode. Autopilot on. Whew!"
--someone has/had that on their profile here. Sad thing is, they're not kidding. It's how PCL teaches--fear the wing, don't fly it. Otherwise we might have to teach you why and that costs money.

And then Phil T. couldn't be Regional Self-Stimulator of the Year or whatever his little sash says.....

...and our Central Safety Air Chairman is so disgusted with things he's publicly stated he's looking for another job. Nice.

 

[Lear70]

...and our Central Safety Air Chairman is so disgusted with things he's publicly stated he's looking for another job. Nice.

I thought we were ALL looking for another job.

While I don't agree with Surplus' conviction of stall recovery training being useless (or perhaps I'm taking you out of context), I do believe that there is much that cannot be taught in the sim, not to mention this sim is not programmed to do "high altitude unusual attitude" airwork - who knows whether its responses to handling are even vaguely indicative of what you would see in the real aircraft in that regime? (Not to mention I'm wondering how the h*ll the aircraft got into a FULL stall with the pusher activating, unless it was overridden?)

Bottom line, don't believe everything you see in sim land and keep your wits about you on the line. Prevention has always been the best medicine; training and self-education is simply a way to accomplish that goal.

 

[TIS]

Do you for instance really think that American can teach its pilots just how much rudder is enough and how much is too much for the AirBus in a simulator? I don't think so.

You don't think so? Well, listen here. I KNOW they do now. You just have to know the right people - like an AA Bus CA - who can tell you what it's really all about. It’s actually quite an intense subject, depending on who you talk to.

AirBus needed to come clean and tell its customers that the use of rudder in turbulence could cause the vertical fin to separate from the aircraft.

You're really barking up the wrong tree here. There may have been a lot of stuff swept under the rug on this crash but the deal with the rudder wasn't part of it. I'm no defender of Airbus but on this issue, Airbus was no worse than anyone else.

During the investigation, one of the outstanding questions was why a wake turbulence encounter would cause the tail to fall off. It wasn't until the investigation stumbled across the fact that AA trained ALL its pilots to respond to unusual attitudes with aggressive use of flight controls - particularly rudder for roll control - that it became apparent that perhaps that's where the problem was. They taught ALL their pilots this stuff. Now, do you really think they would have done that if there was so much as one sentence in ANY of the AFMs applicable to their fleet?!?

The design and manufacturing engineers were consulted and they all pretty much unanimously agreed - across the manufacturers' spectrum - that rapid reversal of controls is not what is meant by "full control deflection" with respect to structural integrity guarantees inherent in the engineering of an aircraft. Addenda explaining this were added almost immediately to every AFM out there and we are all the wiser for it.

TIS

 

[100LL... Again!]

By the time someone is behind the airplane enough to put it in such a situation, they are likely too far behind to correct it.

Deficient pilots are mostly likely to put themselves in situations such as this, and likely lack the airmanship to recover.


Complacency and clulessness are tough to remedy.

 

[Ty Webb]

By the time someone is behind the airplane enough to put it in such a situation, they are likely too far behind to correct it.

Deficient pilots are mostly likely to put themselves in situations such as this, and likely lack the airmanship to recover.


Complacency and clulessness are tough to remedy.

I hope you're not talking abut the AA FO. According to his father, (a retired airline CA), he was an accomplished aerobatic pilot, and I, for one, am not convinced that the rudder movement was caused by pedal application.

 

[KeroseneSnorter]

I hope you're not talking abut the AA FO. According to his father, (a retired airline CA), he was an accomplished aerobatic pilot, and I, for one, am not convinced that the rudder movement was caused by pedal application.

I too feel that the report of the cause of the rudder seperation is suspect, but I wasn't there so will have to take the NTSB's word for it.

However, it took them years to admit that there was a design problem with the 737 rudder. Personally I think that economics or other considerations had a factor in their final ruling on the AA Airbus.

 

[TWA Dude]

A few short thoughts:

In the DC9/717/MD80 I almost never climbed in VS mode at high altitude -- the airspeed could drop off very quickly. I did, however, witness a couple of captains who happily did and allowed airspeed to get below my comfort level -- and these were guys with many years of experience. I hand-flew at altitude a lot -- sometimes the whole leg.

Now on the A320 I've found it's common technique to climb in VS mode up high. I do it but I put down the latest edition of Seventeen (I know I'm not the only guy to read it!) and watch closely. If the speed drops I level off as needed though it rarely happens. Also before going to 370 or 390 we check the FMS max alt and also refer to our speed cards and ensure at least a 30 knot spread between high and low airspeed limits. Company philosophy is to let the AP/ATHR do almost all the work. Hand-flying is becoming a dying art in this plane. My understanding is that with RVSM hand-flying at altitude will even be prohibited, but at least that makes sense.

As much as some people would like to argue it PFT has never been a safety issue. Pilots can and do screw up no matter where, how, or when they got trained. In general, training at the regional level is a joke compared to the majors. (That's a jab at the companies and not the pilots!)

 

[surplus1]

While I don't agree with Surplus' conviction of stall recovery training being useless (or perhaps I'm taking you out of context), I do believe that there is much that cannot be taught in the sim, not to mention this sim is not programmed to do "high altitude unusual attitude" airwork - who knows whether its responses to handling are even vaguely indicative of what you would see in the real aircraft in that regime? (Not to mention I'm wondering how the h*ll the aircraft got into a FULL stall with the pusher activating, unless it was overridden?)

Bottom line, don't believe everything you see in sim land and keep your wits about you on the line. Prevention has always been the best medicine; training and self-education is simply a way to accomplish that goal.

You're not really taking me out of context, it's just a little difficult for me to be clear without taking up several pages. You are right on target when you say that the sim is "not programmed" for these conditions. As I'm sure you know, when the sim's "program" is exceeded it "tilts". These boxes have limitations (much like real airplanes) that are often unknow to the "instructors" and always unknown to the "students". It is my opinon that they are not very useful in the types of scenarios that we are discussing. That's why I think the sim training in these things is of little value.

Based on the NTSB report, it appears that the pusher activated several times. If that is accurate, I respectfully submit that the first "stall" was inadvertent and the others were induced by excessive control input (given there was no power available - which propably was not realized at that time). Yes, it's conjecture but it seems logical to me.

I relate this somewhat to the general scenario in basic multi-engine training. When and engine fails the pilot is taught, by rote, "max power, gear up, flaps up, identify, feather" --- all in less than 5 seconds. Many an engine that would deliver partial power has been shut down because of this. Many a turbo-charger failure has resulted in an unnecessary engine shutdown followed by an avoidable accident. People are "taught" to attempt to fly on one engine in airplanes that really aren't capable of single-engine flight. This in turn produces VMC rolls and tragedy. All this sort of thing comes about because of the "blind-leading-the-blind" theory. It's built into our system, i.e., the most inexperienced pilots are generally the "instructors". By the time one gets to an airline that isn't supposed to happen any more but sometimes it does. I'll leave it to you to figure out how that might apply in this case.

I agree that "prevention has always been the best medicine." That's the crux of everything I've said. Teach people how to avoid these situations and there will be no need to recover from them.

It seems to me these pilots weren't taught much of anything, which got them into the stew. What happened next is all guesswork (on their part and ours). From what I've read in this forum the corporate culture over there is the real culprit (and that isn't confined exclusively to management).

 

[FN FAL]

I hope you're not talking abut the AA FO. According to his father, (a retired airline CA), he was an accomplished aerobatic pilot, and I, for one, am not convinced that the rudder movement was caused by pedal application.

But an Airbus isn't an Aerobatic airplane.

Manufacturers of handguns incorporated in their sales lineup, autoloading pistols designed to operate double action only. This was because police and civilians, who had long mastered double action revolvers, had a hard time transitioning to the single action trigger features of modern autoloading handguns.

Maybe you can't make the correlation, but good reflexes and muscle memory are hard to change. Even CFI's know that in an emergency, students and pilots will revert to basic procedures they were taught in infancy, even though they might have felt they may have become rusty on those responses.

We also see this muscle memory and reflex problem, with military people transitioning to civilian skydiving. The fatality reports bear this out...for some people, the axiom "you can't teach an old dog new tricks" rings true.

Certainly, being an aerobatic pilot helps create a well rounded pilot, but the Airbus is not an aerobatic plane and sometimes reflexes get the best of people.

 

[Crossky]

The CRJ doesn't have a sloppy autopilot? What's up with the wing-wagging while tracking in NAV mode?

I never saw that in the -80. Listen, they all have their quirks. The 80 is a better airplane, overall.

No comparison.

Yea, I know it does that. Does it when I hand fly it too . My theory is it's due to the influence of the winglets and the autopilot is unable to dampen out the oscillations.

I remember the -80 doing some pretty crazy control wheel roll oscillations while on an ILS, and the autopilot tracking a VOR course - forget it. We always flew it with HDG to keep it from zig-zagging our way across the country. (We had a few birds which didn't have an FMS yet so it was /A and Jet routes the whole way). The autothrottles worked, yes, but they were trickey, and if you didn't finnesse the flight guidance with VS changes and lead the thrust setting changes, the pax would get pressed forward or back too much or floated or G'd too much for comfort IMO.

The guys transitioning from the right seat 73 to the left seat of the -80 hated it. I head classic comments like "Boeing builds airplanes, Douglas builds character", "the original fly by wire airplane, by 1/8 inch cable, and etc.

I thought we were comparing autopilots, not airplanes anyway. If you want to compare, the packs on the CRJ beats the pants off the -80 all summer long.

All in all I don't hate the -80, wish I was still flying it, would have if I could have. Former company was added to the list of dead airlines 2.5 years ago.

 

[surplus1]

During the investigation, one of the outstanding questions was why a wake turbulence encounter would cause the tail to fall off. It wasn't until the investigation stumbled across the fact that AA trained ALL its pilots to respond to unusual attitudes with aggressive use of flight controls - particularly rudder for roll control - that it became apparent that perhaps that's where the problem was. They taught ALL their pilots this stuff. Now, do you really think they would have done that if there was so much as one sentence in ANY of the AFMs applicable to their fleet?!?

That was exactly my point. Had they known of the potential danger the training would have been modified, before this "event". You can defend AirBus all you want. I would sooner defend the AA pilots. JMO.

Addenda explaining this were added almost immediately to every AFM out there and we are all the wiser for it.
TIS

Again you make my point. This should not have been "addenda". The Warning should have been there from day one. Thanks to AirBus it wasn't.

As I tried to point out there are other airplanes and other manufacturers with similar problems. Boeing is one of them. The folks that make the ATR another and Hamilton Standard re the Brasilia's propeller a third. Problems that result from engineering flaws or lack of information are different from those created only by humans in the cockpit or the Training Departments.