thoughts on stalls
 

And here we go...

Round and round we go....

Ailerons an spoilerons do a much better job at high AOA. Many pilots would be better off engaging the gust lock and drinking some coffee instead of trying to fight an upset situation.

Ailerons (or spoilers for that matter) for the transport category aircraft that stall near the tips first and may be experiencing slideslip from yaw to further decay the AOA? During the extended envelope training a key point was made that reducing the AOA and unloading the aircraft had been "washed out" of training over the years, to the point where when wings drop people were trying to "ride the shaker" and force the wings back up with rudder an ailerons, never correcting the AOA issue in the first place, either at all, or to the extent necessary, given the very gradual C/L curve of transport category wings (not shaped like your GA Cessna). I don't agree about doing nothing, that gradual curve where the slope is nearly flat has caused a lot of "not sure what is going on?" situations due to not actively reducing AOA. A lot of those "wild ride" stall/yaw incidents were due to never (significantly) reducing the AOA.

I'm not a fan of the engineering behind the stick shaker activation. I would much rather have something along the lines of the Q-Alpha energy state indicator or similar. https://www.kansas.com/news/business...132377769.html
It also indicates how much float you will have at round out, possibly preventing runway overruns.

That does depend on the aircraft though.

Just the idea that someone would slam in a bunch of rudder while stalling is...appalling :)

True. I was specifically thinking of my former aircraft where at HIGH AOA (30+) you intentionally used the rudders for control. Influence with a tad of aileron; but control the turn/bank with the rudders. Once you were actually out-of-control then the second step of the immediate action was 'feet off the rudders' :)

Umm no, just the opposite.

I was thinking of American 587. Its amazing how airliners can go from flying to stalling and into full pancake mode without dropping a wing.

That comes down to simple airmanship. You don't need anything to tell you about energy state when landing to prevent an overrun. Fly the appropriate speed and AoA, and land, don't hold it off until midfield, and calculate the landing performance in advance. Now you know, before you go.

Simple airmanship.

It’s very aircraft dependent. In many if not a majority of aircraft ailerons introduce adverse yaw when deflected at high AOA. The quickest way to end up on your back in many high performance aircraft was using aileron to try and pick up a wing at high AOA.

Unfortunately, some DO need these warning systems. The Colgan accident for example.

And knowing...
http://www.abilityministry.com/wp-co...12/Knowing.jpg

:confused: Because the warning system was used and heeded so effectively by the Colgan Captain? I have no problem with a stall warning system, but it's presence in no way guarantees a positive outcome as evidenced by the very accident you cite. A stall warning system doesn't do much good if the user ignores it and pulls harder.
Like JB said, it's simple airmanship. The ones who truly "need" such a warning system to be safe are probably in the wrong line of work.

So you're saying you had alpha?

https://youtu.be/335GdTqtyLs?t=59

Yeah giving the problem children more info won't usually solve the problem.

And to take it an extreme case, IIRC, the AK C17 crash.

No. Not remotely so.

For starters, that mishap had nothing to do with landing long or an overshoot, so the original assertion that additional displays indicating potential float at roundout" and overshoots, is irrelevant.

The colgan mishap had ample warning information, none of which was needed, and the reaction to that warning information was incorrect...all the way to impact.

This is about stalls, not about float, that's just a secondary aspect of a significantly better system.

Engineers are constantly looking to improve systems. A better system would have prevented the Colgan accident. Whether captain Renslow was qualified is a completely different topic.

Renslow's qualification and ability is not a different topic at all; it's key to the loss of the aircraft and the mishap, and the outcome.

A better system would absolutely not have prevented the flight. The crew had ample warning and indication and in fact upon receiving a stick shaker continued to increase AoA and progressed deeper into the stall, failed to provide more than partial power, and maintained back pressure and a high AoA all the way to impact.

Neither crew member recognized the stall, and their inputs were consistent with confusion, and sudden reaction. This actually happens more than it should in this business. The Q-Alpha system clearly warns pilots in spoken language of an approaching low energy state. The aircraft would have been barking at them to increase power several seconds before the shaker activation. The shaker is engineered to prevent false activation, every pilot needs to understand they can be in a stall, or parts of the wing can be stalling before the shaker activates.

Again, ample warning was given. In fact, because of the selectable stall speed input on the Dash 8, early warning was given, with a greater-than-normal cushion of AoA and airspeed prior to the stall; the crew had adequate notice: enough so that crew who were doing their jobs could have and would have reacted with plenty of time to not only avert a stall but entirely avoid a departure from controlled flight.

This was NOT a lack of cockpit data. This was NOT a lack of warning. This was NOT a lack of reaction time. This was NOT an insufficient margin of energy, AoA, or any other metric regarding lift reserve or capability. This was simply a matter of incompetence.

The aircraft manufacturer could have had flashing windows the color of the rainbow with peach, kiwi bubblegum, and Fantastic Fatfinger Freddy Pink thrown in for good measure, and it wouldn't have mattered. And it didn't matter. It didn't matter that aerodynamic buffeting occurred, that the stick shaker occurred, that the pusher occurred, that cockpit instrument indication warned of the stall, or the fact that the crew had calculated the speeds and knew them and had briefed them, because despite the plethora of warning in the cockpit from visual to tactile to airframe buffet, the crew still failed to act, still failed to increase power adequately, still failed to reduce AoA, still failed to do the most basic requirement and tenet of airmanship:

Fly the damn airplane.

Adding gee-which gizmos, light, reserve lift indicators, PLI's, and all the other wonder gimmicks for the modern cockpit wouldn't have made one iota of difference when the crew didn't do their most basic, core job. Fly the damn airplane.

Q-Alpha or similar system is NOT a gee-whiz gizmo. It takes the same philosophical approach to warning the crew in the same fashion as GPWS, TCAS, depressurization, etc. This would have absolutely reduced the element of confusion in this accident. If you think the crash would have still happened, with both pilot knowing the aircraft state, then we agree to disagree.

I think, that had the FO or the airplane warned the captain to check airspeed, this crash would have been prevented. I mean, what do you do as PM when you see the other pilot making a mistake, say nothing because they should know better?

I agree with you. Q-alpha would have made a big difference. Those who are arguing against you in this thread are indicating a lot of hindsight bias. It is easy to see what YOU would have done differently with the benefit of hindsight, and it makes us feel good to think that a particular accident would not have happened to us, but it is almost always wrong. The reason is that such thinking ignores the actual contextual factors occurring at the time. If you cherry pick what you are perceiving or should be perceiving in hindsight it is obvious. In real time there are all sorts of other aspects. Put yourself, as Dr. Sidney Dekker says, "in the tunnel", knowing ONLY what they knew at the time, and having ONLY the training and experience that they had then, plus with any other fatigue, experience, etc.

It would have made no difference whatsoever.

Your dismissive viewpoint abandons decades of experience that the commenters here have, individually, and the centuries, collectively. Call nearly four decades of doing it, teaching it, observing it, studying it, etc, mere bias (and an objective view of a mishap that's been picked apart every which way from sunday) a "feel good" biased view if you like (and you do), but it's wrong, and such dismissiveness is hardly beneficial.

Yes, we can view the mishap based on hindsight, because very clearly we cannot view it with foresight. It existed in time, and exists now as a detailed mass of data. We know that the participants were minimally experienced, the F/O woefully so, and the captain one who bought his job, had numerous failures in his training and past, and who bragged on the transcribed cockpit conversation of shortcutting his career by buying his job with Gulfstream. The F/O made uncommanded changes in configuration.

The response to the events, which had ample margin from the stall and were not close to the critical AoA at the time of initial warning, was wholly inadequate, despite considerable warning and data providing information about the aircraft state. Upon recognition of the problem, the crew's response was similar to a tailplane icing encounter, though the airline taught no such response, and the procedure didn't exist for the airframe. It was the opposite of what was required, both increasing AoA to the stall and use of partial power, and far worse, was continued from an altitude that allowed far more than enough time for recovery, right to impact.

Numerous cues were available, active, and presenting, from shaker to pusher to airframe buffeting to visual cockpit indications and stall warnings, airspeed indications, power indications, configuration indications, and the event is one trained for at every initial, every recurrent, and one emphasized on every checkride from the time of a student pilot onward. None of this was new, none of this was deceptive or subtle. Further, the crew had misset the stall warning switch, allowing an additional 20 knot margin above the usual alerts; the crew received this warning and continued through that gate all the way to a stall by using partial power and an increasing AoA and then held it.

To suggest that it might or might not have happened to us is a straw-man argument which holds no merit. It's irrelevant. It DID happen to them. We know exactly what happened. We know their state of mind, their background, their personal and professional history, their rest history, their family situation, everything, including training records and cockpit conversation leading up to and throughout the event, as we do with most transport category mishaps. We have all the CVR transcript, the report, and the FDR information to provide us excruciating detail on control inputs, responses, and aircraft behavior, second by second, through the entire event.

This is not a matter of making ourselves feel good. This crew unquestionably had an ample plethora of data giving good, solid indications of the aircraft state, airspeed, and flight condition throughout the entire event, and far more than enough warning that it was coming, impending, happening, and continuing to happen. Physics are the same before, during, and after. The correct responses before, during, and after remain the same, and the crew simply followed their standardized training, they'd have been okay. There was nothing at any time wrong with the aircraft. There was no unrecoverable situation. This was not a challenging event. This was not hard to decipher, nor were the clues vague or easy to misinterpret.

The same clues, in fact the same hard data, the same warnings and progression that the crew in question received are handled and used by tens of thousands of pilots working in the field today; the same that we all train with and use on a daily basis and which still work very effecitvely. Stall recognition is not a murky, mysterious, unfathomable phenomena, particularly a straight ahead, unaccelerated stall as the Colgan crew not only experienced, but caused.

To suggest that the crew suffered from a lack of warning, a lack of flight data, a lack of useful, easily interpreted information, is ridiculous and far misses the mark. There are many things that can be taken from that event, but insufficient cockpit data is NOT one of them.

Certainly the issue of proper stall recovery has been addressed as a result of this mishap: the long-taught, ill-conceived concept of no altitude loss and powering out of a high AoA situation might have merit to prevent ground contact in a windshear situation at extremely low altitudes, but a reduction in AoA has always been and remains the only way to preclude the stall, and the drawn-out process of holding pitch, holding altitude and using power for recovery (especially in this case of partial power) was a poor and inappropriate concept propagated in the industry for far too long. That, fortunately, has been mitigated somewhat by alleviating the hard line for minimal altitude loss in the recovery in training.

The crew might be in part absolved by the problematic training standard of minimal altitude loss and no manual AoA reduction (beyond pusher) in training, except that despite all the warning received, the crew did NOT follow their training.

Be dismissive all you want, but gimmicks and fresh display formats wouldn't have fixed this. This was a human error, not the result of the lack of data.

Quite frankly, John, I think you are the one being dismissive here. You are ignoring that some of us also have many decades of experience both as a pilot, as well as investigating major accidents. Your experience is certainly valuable, but you are also not looking past your bias here.

I am not sure you understand what q-alpha does at all either. If you do it is not evident from your post. While I am only an acquaintance with the creator of it, Andy Skow, he has done something of real value. If you ever knew Pete Reynolds before he passed (google the name and "learjet") you should know that Pete strongly agreed with what I am writing here. For the record, I have no connection to the system or company aside from being aware of what it does.

Your view, and your response, clearly indicate hindsight bias, whether or not you want to admit it. It is right in your words. I would suggest you attend MIT's workshop coming up. It might be of interest to you. Partnership for Systems Approaches to Safety and Security (PSASS)

Best regards-

I'm familiar with it. It may have some merit for the lowest hanging fruit that can't fly an aircraft or maintain situational awareness or do their job at the most basic level...but such really ought not be in the cockpit in the first place. Renslow, for example. Given his wholly inappropriate response to impact, there is no reason to believe an annoying voice or additional flashing light would have made any difference. Everything he needed was already in front of him.

[QUOTE=ptarmigan;2777113]
I would suggest you attend MIT's workshop coming up. It might be of interest to you. [/url]

Not remotely of interest, thanks.

Is it too much to ask professional pilots now to intuitively understand how a wing works, how to manage energy and anticipate changes in energy states? Is it too much to be trained not to panic in an entirely recoverable attitude?

Apparently it is. Sad, too.

GF

[QUOTE=JohnBurke;2777150]I'm familiar with it. It may have some merit for the lowest hanging fruit that can't fly an aircraft or maintain situational awareness or do their job at the most basic level...but such really ought not be in the cockpit in the first place. Renslow, for example. Given his wholly inappropriate response to impact, there is no reason to believe an annoying voice or additional flashing light would have made any difference. Everything he needed was already in front of him.

If you are advocating that we entirely change how we are recruiting and training pilots, as you can see from my multiple writings on the topic, I'm 100% onboard with that. If you are saying that the design of our aircraft is "good enough" and it is the fault of pilots that accidents are happening, I will STRONGLY disagree! Between our system design and training we are completely failing pilots (and the traveling public!).

As for the MIT workshop, that is your loss, and reflects more on you than anything else, sadly. I suggest you read my friend's article on the topic for a primer: https://aviationweek.com/business-av...-out-accidents

I don't see how this is any different from other warning systems, especially GPWS.

Quite a bit different. EGPWS providing earnings of a unknown and unseen threat—terrain obscured from view by dark or cloud, especially in times where navigation was more black art—no GPS, no Nav Display laying the map out.

Airspeed is right in front of you. Stick shaker is unmistakable. We’re taught stalls from about the second lesson.

GF

I would add that many professional pilots *think* they would recognize a stall. Unfortunately the are mistaken, hence my article on the topic.

I didn't say anything remotely close to an opinion regarding recruiting of pilot, so no, I'm not saying that at all. You're saying that. Speak for yourself.

I don't care if you're on board with it. You've drifted from gimmicks to altering the recruiting universe. Good luck with that.

There are those who advocate ballistic parachutes in lieu of skill, judgement, and thought, and we have an industry full of children of the magenta line. Why bother training pilots to recognize a stall, or approach to a stall, when we can have circles of lights flashing in their face, and a loud voice yelling at them? No need to calculate or pre-plan the descent; just wait for the FMC to calculate it. What else can we dumb down?

Good god. Get off your high horse.

In the same time frame, I'm also skipping painting camp, baseball spring training, and an outstanding seminar on gastric distress. I'll spend some of that time at recurrent, some working, shoot a couple of steel matches, and write a chapter if there's time. If not doing what you'd prefer I'd be doing reflects badly on me because I choose not to go to a MIT workshop, that's too god damned bad.

The thing to remember with Colgan was that they were nowhere near a stall when the stick shaker went off. They were in a plane with no ice, but they had selected the vref ice speeds, so the stick shaker went off at higher speeds than their configuration/energy state warranted.

With the complete lack of awareness of what was happening, I think they would have ignored any other clues, no matter how verbal or obvious they would've been, in their desperate panic reaction to a situation they did not expect.

Well, the good news is that if you ever have an accident, I'll be able to look at the events in context to see why the actions you did made sense to you *at the time*, rather than throw stones and talk about how you were just "weak". Pilots throw each other under the bus quicker than just about any other group, and that is sad. You see that quite a bit less with engineering test pilots, but most pilots are pretty bad about it.

As for the last part, this section of the board is about "suggestions for improving safety", and the workshop is about improving safety.

This is the common perception of this accident, and one I used to share as well. However, many pilots will simply not recognize a real stall when it happens to them outside of the training environment, and that is a sad fact. Reading it again, maybe we are not saying something so different. Once a person has created a mental model of the aircraft state confirmation bias kicks in so to that extent it true that they will ignore cues.

What I like about q-alpha is that it prevents the errant mental model in the first place. Ideally you have that an an AoA gauge that you are trained on.

Maybe 1500 hours of fighter time colors my attitude, but we did stalls in the sim every 6 months, how would that not be drilled in? I only saw two inadvertent stalls, both at altitude—unmistakeable. The first, I was riding in a C-5, crew was put in hold, heavy-ish, slowed to typical holding speed at training weights at 4,000’. You feel the buffet coming up thru the upper deck floor, I asked the load what was on the inter phone, “stall”. Recovered immediately. It was dark, in cloud, off circadian time, too.

Second time was on the north coast of South America, entered the top of those equatorial clouds that don’t paint but hold a lot of bumps and warm air. Mach rapidly went away, the “eyebrow” starting coming down on the PFD and the slightest tickle of the shaker. Pushed over, as trained, to maintain some Mach and lost about 1200’ before we popped out of the cloud. We were at or just above optimum.

Both cases, if you missed the warnings, you’d have to be dead.

GF