AA Crash Jamaica

[Murexway]

Rate of change of acceleration per unit time (3rd derivative) = jerk (I am not making this up...)

That's what I was looking for..... Jerk

That's why to someone who's hydroplanning on a short, wet runway at night you'll never convince them that they didn't actually accelerate

[protectthehornet]

very interesting...when I was a CFI(CFIIMEI), I had a number of NASA engineers from AMES research center.

they were all better engineers than they were pilots.

Stable approach, hit your spot, stop...and you did your job.

[AirRabbit]

Originally Posted by barit1

… if you hit a slick spot so that acceleration suddenly goes to zero, then that's "positive jerk". Where you were against the tight shoulder straps, your torso now bounces back against the seat. This gives the PERCEPTION of positive acceleration.

I understand the physics and the derivations involved, and, of course you are correct. Additionally, as it is true in one direction, it is just as true in the other direction. For example: accelerating in an automobile … release the accelerator, engage the clutch and shift, release the clutch, re-engage the accelerator – and so on through the number of forward gears in the example automobile. The occupants in that automobile would be subjected to the same (although opposite) accelerations on their body (and sometimes find themselves bouncing their foreheads off the dash board). However, I have yet to hear of an occupant in an automobile indicating that during this change in the acceleration rate (during the gear change) they felt like they were decelerating. By the application of the same, legitimate rationale, they would have been subjected to the identical (although, again, opposite) force changes. Might I suggest that it could be merely a recognition factor on the part of the individuals involved? In one case, knowing what to expect and recognizing it when it occurs – and the other case(es?), not knowing what to expect and being surprised when it occurs?

[PEI_3721]

IGh, thanks for the extracts (# 516); we often underestimate the risks of tailwind landings.

DownIn3Green, 411A. Knowledge is important, but more so, knowing when and how to use that knowledge.
Knowledge of a tailwind approach might include a higher rate of descent (theory), but do pilots look at the approach vertical speed and associate it with a requirement to change the flare time/distance, i.e. knowing when and how to change ‘gear’ for the flare (practice).
If the HUD ‘FD’ flare path is similar to autoland, how many pilots know this, and if known who adds more flare distance to the advisory landing distance calculations.
Automation / technology per se have little to do with this, the problem is with the humans who use these facilities – how crews think about them.

Murexway, et al, – it’s another of those captivating illusions which plague human performance.
Considering a practical aspect, if most training simulators use tilt to create the illusion of acceleration, then they probably cannot create the sudden change of acceleration necessary for an actual illusion of ‘acceleration’ felt on slippery runways, i.e. an illusion cannot create an illusion.
Then how can pilots gain experience of very slippery runways other than by experiencing the real conditions? (Or as with some ‘extreme’ hazards in aviation, you don’t need to have experience of the condition if it can be avoided with reasonable certainty.)

The real illusion of ‘acceleration’ is often encountered when cancelling thrust reverse when using autobrake on a slippery runway; particularly where autobraking is not very effective due to a low setting. This illustrates a problem of routinely using autobrake (depending on it), which removes opportunities for the pilot to acquire a personal calibration of brake pedal foot-force with the achieved deceleration, for a given runway condition.
This lack of calibration (experience) also contributes to the potential for erroneous reports of runway braking action.

AirRabbit, #525. Perhaps it’s the difference between having or not having the relevant experience?

[AirRabbit]

Originally Posted by PEI_3721

AirRabbit, #525. Perhaps it’s the difference between having or not having the relevant experience?

Without a doubt ... but, it would seem that not many who occupy a window seat in the cockpit are devoid of experiencing both accelerations and decelerations - in fact, I would suspect they all have acquired quite a bit of experience, in both areas.

[PEI_3721]

AirRabbit, re “ … they all have acquired quite a bit of experience …”

IMHO you may assume too much.
In post # 526 I suggested that it might not be possible to gain the required experience via simulation; thus, gaining experience would depend on a pilot’s exposure to real conditions, and an appropriate interpretation and association of the event with those conditions.
i.e. pilots may have operated in the conditions, but this has to be turned into experience in memory and prepared for future recall.

Also, I suggest that pilot’s are rarely exposed to really slippery runways in conjunction with a limiting distance, the latter contributing significant meaning to associate risk with the operation, i.e. where did the aircraft stop vs what was planned / expected, what did it feel like (deceleration).

Thus the problem could be the identification of the situation and associating the risks with it. This requires a timely recall of experience in real conditions (know when, know how). Application could be via the classic TEM, avoid, detect/recover, mitigate.

If you haven’t been in those conditions, then you can’t (shouldn’t) consider doing so; thus how can pilots ever be there to gain experience (catch 22)?
What I am implying is that many accidents of this type occur in conditions which either have not been experienced before, or were not anticipated – they were a surprise.
Thus the problem is how we can reduce surprise.
With previous experience of the conditions (training) … or even with imagination?

Fundamental surprise.

Thinking Ahead.

[AirRabbit]

My comment was directed toward the experience gained in the cockpit through the adjustment of brake pedal pressure, that has, presumably, provided sufficient recognition of the change in the deceleration rates with such braking adjustments. This is recognized when adjusting brake applications during landings and is evident even when taxiing – particularly when bringing the airplane to a full stop.

[p51guy]

I agree with Air Rabbit that after thousands of landings you know what is normal acceleration and deceleration. As a brand new captain on an empty 737 I was the first landing at RNO at 6 AM after a snow storm that closed the airport the night before. Everything seemed normal until I came out of reverse at 80 knots when my deceleration stopped with max braking. Going back to max reverse got us stopped but there was no doubt in my mind that the braking action was nil. Only reverse kept us from going off the end of the runway. No, we never got a braking report from the tower.

[411A]

No, we never got a braking report from the tower.

Neither did I at the old Taipeh airport, long ago, with a B707.
Grease/oil etc at the far end (from mostly C119 piston aircraft, I later learned) in addition to dynamic hydroplanning..prevented reasonable braking action.
A close call...yes, however, MAX reverse held to the very end, saved the day.

[protectthehornet]

boys...in this case...max reverse...do you really mean firewall reverse? reverse to mechanical stops? or just whatever EPR is considered max reverse?

I'm guessing firewall?

[PEI_3721]

AirRabbit (#529), thanks. The hub of my point is that there appears to be too greater use of auto brake; even a dependency on it.

Re “[i] … after thousands of landings you know what is normal …[i]”
Perhaps this is an emerging problem; will pilots upgrading to captain be required to have “thousands of landings”, and will all of those be without autobrake to enable experience of foot force vs deceleration? And even with that many landings how many are likely to be on runways which involve reduced friction conditions.
The conditions required might have to be similar to those which were encountered in the example above, however in hindsight did the experience result from error.
Was it an error of judgement – known conditions but an inappropriate choice of action. Or alternatively as the result of poor information or situation assessment where the choice of action appeared reasonable, but was in fact incorrect?
In either case error may have been the root of the experience, i.e. true experience might only come from error.

Whilst experience of deceleration vs foot force is valuable, in marginal conditions it may only be used to detect a poor situation and mitigate the circumstances. This is an after ‘the fact’ activity – after the decision to land. The real defence against overrun accidents must start much earlier, during the approach briefing – situation assessment, and the decision to land.

Perhaps as in the example, many people rely on reverse to save the situation. Therefore it is important to understand how we value thrust reverse. Either it will save an overrun in marginal conditions because it’s been done before (drift of normality), or it may prevent an overrun in the event of any unforeseen circumstance (#531), i.e. it hasn’t been done before, but it still might be required (risk alleviation).
The first view is not as per the certificated distances (although in Europe it may be when operating on contaminated runways – implying that there is not second option).
The second view provides the better safety margin; a means of improving safety.

Thus the assessment of conditions and how resources are to be used are critical items.
How do pilots gain experience of this? Will pilots have to land in poor conditions resulting in less than expected stopping capability to gain the necessary experience to make a decision, or can this be taught? A similar question as in #528.

[pattern_is_full]

I find it funny that pilots - of all people - are debating something that should be obvious and already trained into anyone with more than 20 hours logged.

The generic term is "vertigo" and it means that seat-of-the-pants kinesthetic sensations are a lousy way to judge accelerations in any plane (pun intended), especially without visual references (IMC/night/closed environment).

Nov. marks 40 years since my first flight as pax in an airliner. Yet I would still swear (if I didn't know better) that the nose-wheel is collapsing on every landing. The deceleration (OK, rearwards acceleration) not only makes it feel like gravity is forward and we are tilted down - the aisle even LOOKS like it is sloping downward to the nose.

But I hope any pilot - or anyone the powers-that-be allow to get involved in an accident investigation - has enough understanding of this phenomenon to discount reported kinesthetic sensations in analysizing a flight path. Even that part that takes place on the ground.

[AirRabbit]

Originally Posted by PEI_3721

AirRabbit (#529), thanks. The hub of my point is that there appears to be too greater use of auto brake; even a dependency on it.
Re “ … after thousands of landings you know what is normal …”
Perhaps this is an emerging problem; will pilots upgrading to captain be required to have “thousands of landings”, and will all of those be without autobrake to enable experience of foot force vs deceleration? And even with that many landings how many are likely to be on runways which involve reduced friction conditions.
The conditions required might have to be similar to those which were encountered in the example above, however in hindsight did the experience result from error.
Was it an error of judgement – known conditions but an inappropriate choice of action. Or alternatively as the result of poor information or situation assessment where the choice of action appeared reasonable, but was in fact incorrect?
In either case error may have been the root of the experience, i.e. true experience might only come from error.

First – I did NOT say, and it was not my point, that recognizing what may be “normal” came only after “thousands of landings.” Most training programs today use somewhere between 8 and 12 simulator periods. If each simulator period provides each pilot with something like 4 or 5 takeoffs and landings, each pilot should have had in the neighborhood of 30 to 60 takeoffs and landings before they ever reach the airplane. At least some of those takeoffs and landings should have been using the runway contaminant capabilities of the simulator – and, dare I say it, the capability of that simulator to simulate short runways as well. Of course, the circumstances we are describing may occur on the very first flight – but that is not usually the case. Regardless, I maintain the opinion that there should be very few cases where a pilot should not be able to recognize “abnormal” from “normal” decelerations – particularly if the “abnormal” deceleration crops up during what was, up to that point, a “normal” deceleration – even if it is his/her first flight in the airplane.

I do happen to agree that there is likely too much dependency on the “auto-brake” systems on today’s airplanes … and that also applies to the use of “auto-throttles,” and with that, I mean manually manipulating the flight controls, but letting “George” fly the throttles. Dependency easily creeps up on you … ask any drug addict. It’s the same principle when using other “crutches” (like auto-brakes or auto-throttles).

Originally Posted by PEI_3721

…experience of deceleration vs foot force is valuable, in marginal conditions it may only be used to detect a poor situation and mitigate the circumstances. This is an after ‘the fact’ activity – after the decision to land. The real defence against overrun accidents must start much earlier, during the approach briefing – situation assessment, and the decision to land.

I could not agree with you more!

Originally Posted by PEI_3721

Thus the assessment of conditions and how resources are to be used are critical items.
How do pilots gain experience of this? Will pilots have to land in poor conditions resulting in less than expected stopping capability to gain the necessary experience to make a decision, or can this be taught? A similar question as in #528.

Again, I say the education of the pilot starts with ground school. That is where the landing “additives” (to which I alluded earlier) should be taught – not necessarily as specific numbers. They were intended as approximations to provide additional information that may influence approach and landing decisions. Not only do I believe that such things “can be taught,” I firmly believe that they MUST be taught! And, when necessary, whatever is necessary at the time must be used. As most pilots recognize … there are 4 things that are essentially useless to a pilot: 1) the air above you; 2) the runway behind you; 3) knowledge you do not use; and 4) the time before now.

[PEI_3721]

AirRabbit, apologies, I omitted to include the link (# 530) to p51’s post for the quote ‘Re: …’
At least you have made your view clear on the issue.
However, I believe that the effectiveness of the simulator in providing experience of deceleration remains open. IMHO new pilots under training in simulator tend to judge the condition of the runway surface (braking effectiveness) on the distance taken to stop, particularly when having been told what the runway conditions are, e.g judgement/experience in hindsight. I doubt that this method of judgement is used that often in routine operations where there is more reliance on the perceived deceleration, particularly if the visual cues are poor (rain / night) or limited (cat 2/3). Whilst use of deceleration is ‘good’ in the time sense of ‘now’ vs hindsight, there still remains the need to reflect on what has been experienced in order to form the memory of the experience.

In all other aspects, we agree.
I would add to the training, a need for greater emphasis on the variability of landing distance margins and the things which can erode them. Many pilots only consider the landing mass, distance, wind, and runway conditions, whereas variable or extreme weather conditions (perhaps as in this accident), the standard of runway surface maintenance and texture, and tyre conditions also have significant effects.
Gaining experience of these aspects is much more difficult and thus the ground school training must instil both knowledge and a trigger for recall; – how to avoid your item 3), e.g. red WXR zone over the runway in previous 15 min then the runway may be flooded – consider contaminated landing data, do not land downwind; TEM I think.

[decurion]

European Action Plan for Runway Excursion Prevention

In Europe an action plan for the prevention of runway excursions is currently being developed. This also considers the training of pilots. The following study is used as starting point of this activity:
http://www.nlr-atsi.com/eCache/ATS/14/919.pdf

[Admiral346]

Decurion,

thanks for providing the link.

A report very much worth to be read.

Nic

[AirRabbit]

Originally Posted by PEI_3721

I believe that the effectiveness of the simulator in providing experience of deceleration remains open. IMHO new pilots under training in simulator tend to judge the condition of the runway surface (braking effectiveness) on the distance taken to stop, particularly when having been told what the runway conditions are, e.g judgement/experience in hindsight. I doubt that this method of judgement is used that often in routine operations where there is more reliance on the perceived deceleration, particularly if the visual cues are poor (rain / night) or limited (cat 2/3). Whilst use of deceleration is ‘good’ in the time sense of ‘now’ vs hindsight, there still remains the need to reflect on what has been experienced in order to form the memory of the experience.

I don’t have a huge disagreement with you. Of course it is true that new pilots in training in a simulator will judge the condition of the runway surface based on the distance it takes to stop – or at least reduce forward speed to a recognized speed for taxi. But, I think that is largely true for new pilots in the airplane as well. The only way to gain experience doing anything is to do whatever it is for a while. Therefore, for the first few flights for a new pilot the only “experience” he or she will have with deceleration on landings will largely be what exposure they had in the simulator.

Simulators – at least the more modern ones – are exceptionally capable machines … and their value simply cannot be ignored. When called upon to do so they are quite capable of providing several differing runway surface or surface contaminant conditions that will result in having the simulator respond the way the airplane would have responded. Unfortunately, not every training organization utilizes the capabilities of their equipment to the degree they could – or should.

As an example, I am still one of those who believe that pilots do not have a realistic understanding of just how far forward the rudder pedals must be rotated to achieve “maximum” braking. And, my tendency is to have them at least try to understand by referring their memory to how far they have to depress those same “brake” pedals when setting the parking brake – and advise them that position is just at the threshold for maximum braking effectiveness. I’ve seen many pilots having to adjust the position of their “posterior” in their seat to allow them to achieve sufficient mechanical advantage to get the parking brake engaged. If that “posterior” position was necessary during a max brake stop or an RTO – there’s a good chance those would have been overrun potentials. And, again, if the simulator is built, programmed, and maintained correctly, even the parking brake application will be measured and compared to airplane data for accuracy.

[IGh]

Update: JCAA Release #3 dated 22Dec2010

The latest release from the investigating authority, JCAA, is available:
THE JAMAICA CIVIL AVIATION AUTHORITY

JCAA release #3 is a pdf, 6pgs, 84KB download [click on the news, UPDATE, right side menu].

Still no explicit statement from JCAA to confirm the LANDING CONFIGURATION for this tailwind approach to a wet rwy.

As described in message #304 (this PPRuNe thread) dated 7Jan2010, JCAA's stated Vref implies FLAPS-30 (Lower-Drag) configuration for approach & landing [NOT F40].

Manufacturer's latest advice toward that Lower-Drag Approach config' [reduced-flap, delayed-flap, or low-drag approach procedures] is available here:
AERO - Conservation Strategies: Descent and Approach

Note the manufacturer fails to mention lessons from the FSF's ALAR Task Force (ALAR Briefing Note 4.2 "Energy Management"), nor any mention of the FAA's "engines spooled-up" requirement imposed upon Part 121 pilots: which requires line-pilots to deploy sufficient DRAG early in the approach, so as to attain "engines spooled-up" PRIOR TO the 1000' IMC approach Gate.

If AA331/ 22Dec09 did fly that tailwind-approach with the Lower-Drag FLAPS 30 configuration [not Flaps 40], then both the operator's management advice, and the manufacturer's advice, should be reflected in the investigation findings, & PROBABLE CAUSE statement. Was the lower-cost, Flaps 30 configuration regarded as the company's "standard" for those mishap-pilots?

[PJ2]

IGh;
Re your comment, "Note the manufacturer fails to mention lessons from the FSF's ALAR Task Force (ALAR Briefing Note 4.2 "Energy Management"), nor any mention of the FAA's "engines spooled-up" requirement imposed upon Part 121 pilots: which requires line-pilots to deploy sufficient DRAG early in the approach, so as to attain "engines spooled-up" PRIOR TO the 1000' IMC approach Gate.", and "operator's management advice"

Well stated. However, the industry continues its primary economic focus even as both the flight data and the overrun accident data as well as the regs dictate otherwise.

The current "advice" almost always offered in airline work in terms of approach power is, "Thrust utilized to maintain target airspeed". This advice leaves open the opportunity to use idle thrust in order "to maintain target airspeed". Is this kind of waffling statement for fuel saving, plausible deniability, leaving the crew to manage their airplane or facilitating the risk of an overrun? The answer lies in examining the overrun accidents themselves and not in economically-driven operations procedures. The admonition, seen in some quarters, to maintain thrust "above idle" is not much better, permitting, for example on the CFM56, a setting of 45%N1 +/-, which is arguably just above idle. Still, industry data which continues to show crossing the fence at 25kts above Vref and landing long, all under idle thrust begs for answers, does it not?

So, are the FAA Pt 121 and ALAR Briefing Notes satisfied? Hopefully the report, as you say, will comment on this. Boeing's and Airbus' economic case for saving 60lbs of fuel in some circumstances, has yet to be made. The ATSB's report on the QANTAS B747 overrun at BKK was clear, saying their accident "was in their FOQA data" many months before it occurred.

PJ2

[misd-agin]

They left out other information also - when the moon is full, when Christmas is observed, the price of tea in China, what day of the week Jan 8, 2021 will fall on, etc, etc. The fact is that stablized approach criteria is still required.

And stating that idle is acceptable for stabliized approach criteria???

Man, you guys are tripping all over yourselves twisting words or intentions to arrive at conclusions that are completely at odds with the intent.