Turkish airliner crashes at Schiphol

[DozyWannabe]

@BARKINGMAD:

The salient difference between the two is down to phase of flight - I think what shocks a lot of people about AF447 was that it happened in cruise, and as such despite having 38,000ft of altitude to recover, the crew was unable to. The THY aircraft was on finals, and the margin for error in that situation is considerably less.

As far as I know the THY aircraft was being handled correctly aside from improper monitoring of what the autothrust was doing. Remember that the first hole in the cheese was a faulty RA on the Captain's side feeding bad data to the autothrust system. One of the hypotheses put forward was that because the crew were expecting a reduction in thrust at that point, they simply did not double-check by how much thrust was being reduced. This provides an interesting counterpoint to the "moving thrust levers are always better" argument.

While the accident flight was a check ride (and by extension the flight deck workload was higher than average), the flight crew aside from the FO under training were no novices - at least one was ex-military and a veteran Captain.

The consternation over the increase in reliance on technology and how that reliance affects basic flying skills and experience is understandable (and despite my reputation on here I think it's a valid - if often misunderstood - concern). But we must all be careful not to try to hammer accident analysis into a shape that will fit that perception because it risks obscuring other lessons that can be learned.

[CDRW]

DozyW - what you been smoking man!!!!
"As far as I know the THY aircraft was being handled correctly aside from improper monitoring of what the autothrust was doing"

You have almost trivialized the very act of airspeed monitoring. That pesky act of "improper monitoring" was, well, quite fatal.

The fact the RA was out should have been but a minor inconvenience at best. As the speed approaches Vref most of us expect a power increase, a thrust lever movement, a spool up, the trend arrow to disappear.

[Capn Bloggs]

Originally Posted by Dozy

we must all be careful not to try to hammer accident analysis into a shape that will fit that perception because it risks obscuring other lessons that can be learned.

In this case, what are the "other lessons"?

[fdr]

If the only failure that precipitated the THY accident was the AT inoperative, then there is a problem with achieving the systems risk management goals of paragraph 1309, per MIL STD 882E methodology. Basically, the failure case can occur regularly enough to have to not result in a catastrophic outcome... in fact as the outcome is evidently catastrophic... it is not supposed to happen without being able to be mitigated by the system architecture. The THY event is not the only occasion that the fundamental failure of crew to remain aware of their system performance has occurred, and it won't be the last. Some fairly respectable operations have come undone with the loss of crew SA due to complacency and expectancy of nominal system operation. In a close coupled operation, there is a high cost to such outcomes. We have become "politically correct", and warm, caring crew following the social engineering of CRM programs, however the extent of SA failure, and the atrophy of basic skills that permit a recovery from a simple failure once recognised is an abject industry wide failure.

In respect to the medical profession, it would be hoped that it has the ability to take the hard lessons learned in the aviation experience, and avoid the traps that have occurred in the establishment of protective protocols. IMHO, the single most significant factor is the general disconnect between policy and implementation, both in aviation and the medical community. Until the slip between the stated intent and the perceived attitude of systems is removed, it is not likely that interventions will be more than bandaid solutions.

[BARKINGMAD]

ANGLE OF ATTACK.

Thank you Dozy for pointing out the fact that I've been hibernating too long this winter!

I was under the impression that the PF onboard AF447 maintained too high a nose attitude for the phase of flight, failed to recognise the approach and entry to the stall and maintained said misinterpretation of the aircraft performance and incorrect control inputs.

As opposed to the THY, where the PF maintained (manually or via the a/p) too high a nose attitude for the phase of flight, failed to recognise....................Ooops, I think we've been here before?

Heyho, back to that darkened room suitably chastened!

[DozyWannabe]

Originally Posted by CDRW

You have almost trivialized the very act of airspeed monitoring. That pesky act of "improper monitoring" was, well, quite fatal.

I didn't say anything about airspeed monitoring - which, as you say, is a non-trivial and integral part of the scan, I said they did not monitor the *thrust* setting - which many have claimed is second-nature if an aircraft is fitted with an autothrottle system that physically moves the levers.

The fact the RA was out should have been but a minor inconvenience at best. As the speed approaches Vref most of us expect a power increase, a thrust lever movement, a spool up, the trend arrow to disappear.

Completely agree, but as I said, that's not what I was talking about.

Originally Posted by Capn Bloggs

In this case, what are the "other lessons"?

A few off the top of my head:

• Seemingly trivial equipment failures can have serious consequences if the system is not properly understood
• Closer attention should be paid to repeating patterns of equipment failure on a specific type (in this case RA failures on the NG)
• Specific to the question of experience - the FO missed the erroneous thrust retard command and falling airspeed indication, but so did the very experienced ex-mil training Captain

Originally Posted by BARKINGMAD

I was under the impression that the PF onboard AF447 maintained too high a nose attitude for the phase of flight
...
As opposed to the THY, where the PF maintained (manually or via the a/p) too high a nose attitude for the phase of flight...

Technically correct, but in the case of AF447 the PF deliberately pulled up and kept pulling up despite such inputs being wildly inappropriate for the flight phase. The THY crew as far as I know did not - their stall resulted from failing to monitor thrust and airspeed while holding what would be a normal pitch attitude for approach - your mileage may vary, but I think that's a pretty significant difference.

Repeating my third point above, given that the (experienced, ex-mil) training Captain and monitoring Captain - as well as the FO under training - missed the thrust setting and airspeed cues, I have trouble seeing how the "Children Of The Magenta" problem applies here. Have I missed something?

Originally Posted by fdr

We have become "politically correct", and warm, caring crew following the social engineering of CRM programs...

That's not (or shouldn't be) what CRM is about. Of the books I've read and the pilots I've been privileged to speak to, one thing that crops up regarding CRM is that one of best recorded examples of how to do it is the CVR recording of UA232. The contents are not "politically correct". Capt. Haynes' language is very salty - in fact he swears like a trooper in places, and understandably so. The conversation is generally polite, but far from touchy-feely - overall what comes across is a sense of 100% dedication to solving the problem at hand.

As a result, by the time that aircraft is pointed at the Sioux City runway every person on the flight deck has a clear idea of the job they have to do, and their jobs are assigned such that the best person for the task is in fact the one handling it. *That* is what I'm told CRM is supposed - and was always intended - to be.

[Capn Bloggs]

Repeating my third point above, given that the (experienced, ex-mil) training Captain and monitoring Captain - as well as the FO under training - missed the thrust setting and airspeed cues, I have trouble seeing how the "Children Of The Magenta" problem applies here. Have I missed something?

Yes. Pilots who regularly hand-fly have trained their eyes to scan. When the AP is in (as it was in this case) their eyes still scan. Those who sit there day in day out just watching the AP don't scan as well because they don't have to. Not saying they're lazy; just that it's harder to do soemthing that is unnatural.

Thanks for the lessons learnt. The first two are valid and slices of the cheese, I agree.

Not being a 737 pilot, what is the thrust level difference when at Idle when above the speed target ie trying to slow down as opposed to Idle when landing?

[BOAC]

Not being a 737 pilot, what is the thrust level difference when at Idle when above the speed target ie trying to slow down as opposed to Idle when landing?

- in this case none as they were, I believe, a little 'unstabilised' (fast). T/Ls on the back stop in both cases.

[DozyWannabe]

Originally Posted by Capn Bloggs

Yes. Pilots who regularly hand-fly have trained their eyes to scan. When the AP is in (as it was in this case) their eyes still scan.

Fair point, but is this something you know to be true or something that you would hope is true? The statement seems to be a bit of a sweeping generalisation to me.

Those who sit there day in day out just watching the AP

I haven't seen any indication that such a description could be applied to the check Captains on the THY crew. It seems like a shaky assumption to make.

Thanks for the lessons learnt. The first two are valid and slices of the cheese, I agree.

No problem - I stand by the third though, until I see evidence that indicates otherwise.

[fdr]

CRM?

Quote:
Originally Posted by fdr
We have become "politically correct", and warm, caring crew following the social engineering of CRM programs...

That's not (or shouldn't be) what CRM is about. Of the books I've read and the pilots I've been privileged to speak to, one thing that crops up regarding CRM is that one of best recorded examples of how to do it is the CVR recording of UA232. The contents are not "politically correct". Capt. Haynes' language is very salty - in fact he swears like a trooper in places, and understandably so. The conversation is generally polite, but far from touchy-feely - overall what comes across is a sense of 100% dedication to solving the problem at hand.

As a result, by the time that aircraft is pointed at the Sioux City runway every person on the flight deck has a clear idea of the job they have to do, and their jobs are assigned such that the best person for the task is in fact the one handling it. *That* is what I'm told CRM is supposed - and was always intended - to be.

DoZ

What part of the UA232 outcome was predicated on "CRM"? I am confused... (ref 1).

the following is not a criticism of the crew at all, but it is a statement against the undeserved kudos afforded to a program of questionable gains, on an industrial scale... I believe in enhancement by TEM, Risk Mgt, and particularly SA enhancement, as a adjunct to the core skills of a pilot, not as a cost effective replacement of piloting skills

the guys were still fighting the flight controls to impact. Understandable, hard to relinquish at any time, but they were not connected to anything. Where is the CRM in that?

The Captain is talking to the pilot handling the throttles who is the guy who had fortuitously used free time in a simulator to try to fly the aircraft using secondary effects. The flight in the simulator was not handled the same way, ie with the inherent delay of communication to a 2nd crew member to manipulate the controls... adding a delay to a 2nd order control process dealing with a time dependent short and long period phugoid. I would have been a believer that CRM was at play if the FO had been repositioned to the jump seat, the CC in the RHS so he can fly and manipulate the throttles without the delay in advice or interpretation to the control process. The FO was helpful in communications, but could have been doing that at another station.

Did the crew do a good job? damn right they did. Did "CRM" have much to do with it? Don't see it. The question is what difference would you expect form any other crew faced in the same situation? Even the JAL103 worked in a similar manner, with a far more critical control problem... although it was evident in that case that the crew had difficulty maintaining confidence in the difference their struggles would make to the outcome. They still did a great job to hang in the air as long as they did. Try that one sometime... there is about a 10 second delay to the thrust/pitch couple, and a slightly lesser delay to the asymmetric power application (modelled only with a tail, 103 had no vertical stab...). The use of split flaps was undertaken by that crew to attempt to re-trim the aircraft, which was courageous but also ended up with a change to the attitude that resulted in the aircraft finding the ground.

Assume that the Hudson submarine is also a "CRM" good show? It was competently handled, and the decision making made in a time critical environment. The outcome benefitted from the captains hobby of gliding... don't see any part of that where CRM played a factor to the outcome. In fact, I personally think that attributing the outcome of either 232 or 1549 to CRM is actually offensive, the crew competency was in fundamental skills of flying in both cases, had absolutely nothing to do with CRM versions 1-6.

IMHO, CRM is accepted as a reinforcement of existing skills to those that already innately have such grounding, they happen to be receptive to the information as an enhancement to their existing skill sets... management of tasks, time, workload, communication accuracy, threat recognition and error management etc.

The industry has taken the CRM mantra on for better or worse, as a potential enhancement in the areas related to human performance within a complex and close coupled task. The outcome has been that it has been considered we can take any person and run them through a CRM program and hey presto, achieve a competent crew member, don't have to teach them to even be able to fly, recognise or recover from a stall, or even remember that it is probably healthy to look occasionally at the airspeed indicator. How bad is the issue of the race to the bottom? Flying along one day a while back, over the north pole, asking the FO about his background, and the events that lead him into the profession of aviation... his answer was breathtaking, the airline responded before the taxi company did. I think he did the taxi company a service. The reason this had come up was in the ground taxiing, ATC has asked us to "follow the [TYPE] on our right". The FO asked which of the aircraft out there was the type mentioned, which was a surprise as he had been an FO on that type for over 2.5 years.

Does the industry have issues? sure does. Does CRM have a part to play? yes, it assists those interested in refining their knowledge and skills. Did it save 232 or 1549? Not in the slightest, and to grandiosely attribute the saves to CRM rather than the competent application of flying skills demeans the performance of the crews concerned. At the same time, while we can rail at 447 and similar events as being many things with some CRM context, such as TEM and SA failings, the fundamental failure is that on the day, there were no pilots in the cockpit.

The industry has been believing it's own propaganda, and drinking the Koolaid for so long that it has become an "established fact" that we all are supposed to believe.

"but the king has no clothes..."

PS... add UA811 and AQ243 to the stable of aircraft saves that should be attributed to the crew flight skill not CRM. Both of these, as in 232, 1549 were cases where the guys are dealing with a time critical event, herding cats, and they dealt with the outcome as they would have in any aircraft, as they had the basic skills embedded in the psyche, not a divide by zero error on the FMGCS/CDU as depicted by a magenta line.

Had Denny Fitch done LOFT rather than some old fashioned what if compound malfunctions... perhaps it would have been hard to attribute 232 as a CRM save... as the outcome may not have been so fantastic, although I suspect that Denny and Co had enough skill as a "pilot" to improve the outcome.

Is this just a rant? Not really;

Heres the thing:

The THY B738 was an failure of basic flying skills first and foremost, for which the interventions that could have been used at trained by a competent CRM program, of risk management, SA loss and recovery, and basic workload management, failed. Single pilot in a J3 cub, the crew on the day would have had the same sort of outcome, lights on, nobody minding the shop. [the AT failure is insidious, but what else was more important than flying the plane at the time of intercepting the G/S from above?].

There is a sort of symmetry... that with 3 pilots in the cockpit the plane is in need of a pilot, whereas if you put a non pilot passenger into the copilots seat of a B738 and permit them to undertake the takeoff merely results in a tail strike.

Please don't devalue Denny or Sully's skill as a pilot by calling their outcomes "shining examples of CRM". These guys had the piloting skills before, during and after, (and in spite of) any CRM course they undertook.

In my humble opinion...

“Your true nature is something never lost to you even in moments of delusion, nor is it gained at the moment of Enlightenment. It is the Nature of Suchness.” Huang Po Xiyun 黄檗希運



"there's a difference between a philosophy and a bumper sticker..." Charles Shultz




reference:

1. The Evolution of Crew Resource Management Training in Commercial
Aviation
Helmreich, R. L., Merritt, A.C. & Wilhelm, J.A. Department of Psychology
Aerospace Crew Research Project The University of Texas at Austin

URL:
http://homepage.psy.utexas.edu/homep...les/Pub235.pdf

[Capn Bloggs]

+1 for that, FDR.

[Capn Bloggs]

Originally Posted by Dozy

Fair point, but is this something you know to be true or something that you would hope is true? The statement seems to be a bit of a sweeping generalisation to me.

If you were a pilot experienced in what we're talking about, I suspect you'd understand...

[Mac the Knife]

[Pepperseed]

AtomKraft

Pepperseed.

Hands on the controls 'monitoring' the actions of the AP or hands on the controls flying the aircraft. One or the other!
When the Turkish was no. 1, I was no. 3. (a Transavia was no.2). I was hand-flying my aircraft- mainly because I thought we might get a bit of wake but also through preference.


What,s wrong with having the hands on the controls while the AP is still engaged - particularly at the last stage of the approach ? How does that contradict the scanning ?
Ever heard about trim runaway ?

[captplaystation]

Pepperseed,

I think you are misinterpreting what Atom Kraft said. . . He IS actually agreeing with you, and advocating " Hands on the controls 'monitoring' the actions of the AP or hands on the controls flying the aircraft. One or the other! "

so, 50% of his 2 choices are " the hands on the controls while the AP is still engaged "

Or am I missing something here ?

[AtomKraft]

Capt. Platstation.

You got that right. I think that even if you're going to let the AP fly, keeping your paws on the controls keeps you 'in the loop' a lot better.

Stating the bleeding obvious mind you!

[Pepperseed]

Captplaystation, AtomKraft

Okay
Sorry about that.
Cheers

[fdr]

On the assumption of CRM efficacy

As I have stated previously, I am concerned with the distancing of the flight crew from the control loop, and I have reservations on the use of "CRM" training to be other than an reinforcement of skill sets that should have been a screening factor, rather than a cure for inadequate screening or pilot training.

"if all you have is a hammer, everything looks like a nail" proverb, (variously Maslow/Baruch/Kaplan)

While it appears compelling on first glance that Al Haynes and Sully both commented positively on the benefits of CRM (gen1 for Al, gen1-6 for Sully), the accident reports show that the outcome was a result of the ability to control the energy and flight path of the aircraft in unusual circumstances. If CRM is to be attributed properly with value adding to the successful outcomes of the events, then there needs to be a clear cause-effect displayed, and the reports do not make that case at all. It is improbable that any crew member coming out of a high profile event and wanting to continue service with the same company or possibly same industry is going to say anything that is controversial at the time, or buck the hype that is being sprouted about the event. (remember the hammer/nail proverb?).

There is an apparent perception that before the advent of CRM programs, we pilots were all neanderthals, and unable to string a coherent sentence in polite company. The truth is that the good operations were always exhibiting the attributes that are now enshrined in formal CRM programs, which we can confirm as the CRM program developed from the evaluation of high performance teams coming from the group of neanderthals. Observation of differences in individual characteristics, management techniques and group dynamics provided the basis for the attempts to simplify these characteristics and techniques into sound bites that could be used for reinforcement training.

what is the evidence that exists that in the absence of attending a CRM class of either the gen1 or gen1-6 flavour, the outcomes would have not been successful?. These events are characteristically energy and flight path control problems, with a big smattering of critical (irreversible) decision making in conditions of uncertainty.

CRM is likely effective, but the evidence resides more in the daily successes of the vast majority of flight operations, for which the industry should rightly be proud of. The exceptional events do not necessarily support such assertions.

Playing devils advocate, one assumes if CRM is credited with the saves that occur anyway, then perhaps it should also be credited with the losses such as the CRM trained TP's of the A330-300 at Toulouse reverse parking, while communicating and discussing the issue all the way to impact. Apparently it is acceptable to remove the flightpath management and energy management from the equation... (these being pilot stuff...) which are covered by good team coordination etc. Lots of SA going on, just missed the bit about the planet being about to rise up towards their APU rapidly due to the extreme flight condition achieved. Same for 447 and to an extent 1951; lots of communication going on... not much piloting.

[DozyWannabe]

Originally Posted by fdr

I would have been a believer that CRM was at play if the FO had been repositioned to the jump seat, the CC in the RHS so he can fly and manipulate the throttles without the delay in advice or interpretation to the control process. The FO was helpful in communications, but could have been doing that at another station.

Fair enough, but there are a couple of factors that precluded that arrangement.

Firstly, while I agree that the continued pressure on the PFCs had little to no effect due to total loss of hydraulic pressure, the crew (including Capt. Fitch) had no way of being certain that was the case. They may have been thinking of the Windsor incident, where there was some limited control available due to the fact that not all the hydraulic connections were severed - as far as they were concerned, the PFC input plus using asymmetric thrust was what was keeping them stable. First order of business - don't do anything that might make it worse!

Secondly, both Captains Fitch and Haynes have stated numerous times that because the failure of engine 2 had caused its thrust lever to jam in position, manipulating the thrust levers from either seat was very difficult. In fact it was Fitch watching Haynes struggle to do so that led him to offer to take the throttles initially, as his position behind the console would allow unrestricted access.

Even the JAL123 worked in a similar manner, with a far more critical control problem...

Agreed, however unlike UAL232 all of their thrust levers were working normally.

Assume that the Hudson submarine is also a "CRM" good show?

In the sense that "It was competently handled, and the decision making made in a time critical environment" (which is fundamentally all CRM should be about), yes. It just so happens that in that case the handling pilot had most of the requisite skillset to handle the problem (though he has always maintained that the whole crew deserved credit for the outcome).

The industry has taken the CRM mantra on for better or worse, as a potential enhancement in the areas related to human performance within a complex and close coupled task. The outcome has been that it has been considered we can take any person and run them through a CRM program and hey presto, achieve a competent crew member, don't have to teach them to even be able to fly, recognise or recover from a stall, or even remember that it is probably healthy to look occasionally at the airspeed indicator.

I don't see that, but as some are fond of observing, I'm not a line pilot.

If there are instances of airline management treating CRM as a substitute for handling skills, then it should be reported to the regulators, as that was never the intent.

...to grandiosely attribute the saves to CRM...

Please don't devalue Denny or Sully's skill as a pilot by calling their outcomes "shining examples of CRM".

I did not and would not do any such thing. The late Captain Fitch is a hero of mine - as are the rest of that crew, but you mentioned him specifically - and neither he nor Captain Sullenberger have ever so much as hinted that they deserve credit over and above their colleagues on the days in question.

In fact, I'd be willing to argue that it's more of a disservice to his memory to dispute his long-held and oft-stated opinion that the outcome on that day was a team effort.

These guys had the piloting skills before, during and after, (and in spite of) any CRM course they undertook.

In my humble opinion...

Understood, but that opinion seems to be predicated on the idea that the latter is intended to either obfuscate or otherwise supplant the former, which - in my opinion - is nonsense.

Not all CRM is the same, for sure - but at the end of the day it is only intended to be another tool in the line pilot's arsenal. Endlessly discussing issues while SA goes out the window is just as much bad CRM as the apocryphal dictatorial Captain shutting out the rest of their crew. United's implementation of CRM at the time of UA232 was called CLR and it was a much more hands-on and practical implementation of the idea than some others.

At the risk of sounding ephemeral, whilst "checking your ego at the door" can be considered a central plank of CRM (in fact I'd say it's central to Professionalism 101), it applies to the crew as a whole. If a junior pilot is being overbearing to the extent that it's detrimental to the success of the flight then that pilot is as guilty of violating the concept as would be the case if the roles were reversed.

Now - just to be crystal clear, I did not venture an opinion of my own that the UA232 CVR was "a shining example of CRM". I simply stated that I had read a statement which had been endorsed by others - namely that the UA232 CVR (EDIT : which I know to contain content that is most certainly not "politically correct") contained a lot of what would be considered "good CRM". That statement was originally made by Robert McIntosh of the NTSB.

[Oakape]

I am really enjoying the discussion. It is a pity that there are not more of them like this on PPRuNe.

One of the statements that caught my eye -

.... it should be reported to the regulators

A generalization I know, but sometimes there seems to be as much misplaced faith in the regulators getting it right, as there is in the automatics getting it right.