• 1
  • 2
  • 3
  • 4
  • 5
  • 8
 
tommytoyz
Topic Author
Posts: 1195
Joined: Thu Jan 18, 2007 9:08 am

BEA Recommendations - AF447

Thu Jul 05, 2012 6:32 pm

Consequently, the BEA recommends (beginning on page 205):

€€1. that EASA review the content of check and training programmes and
make mandatory, in particular, the setting up of specific and regular
exercises dedicated to manual aircraft handling of approach to stall and
stall recovery, including at high altitude.

2. that EASA and the FAA evaluate the relevance of requiring the presence
of an angle of attack indicator directly accessible to pilots on board
aeroplanes.

I know many on this board have not agreed in the past when these exact ideas were brought up. In particular, I think there is a lack of understanding about high altitude flying. The change in critical AoA (low speed) with altitude, for instance, small as it is (only a few degrees on most planes), is not widely known. Debate about that fact by pilots and engineers is the canary in the coal mine, IMHO.

The AoA indicator is an idea that the military has implemented since the 1950s, especially the US NAVY, after which they saw the stall accident rate fall in half. While some airlines do have the AoA indicator, I think that most do not. Certainly most civilian pilots are not trained on AoA. Out of curiosity, did any of the AF447 pilots have military flying experience?

In these recommendations, BEA departs from the philosophy of only relying on stall avoidance training for pilots. Rather, recommending training in actual stall recovery, especially at altitude. I totally agree with the BEAs recommendations and commend them on a job very well done.

[Edited 2012-07-05 11:36:51]

[Edited 2012-07-05 11:43:17]
 
User avatar
zeke
Posts: 9854
Joined: Thu Dec 14, 2006 1:42 pm

RE: BEA Recommendations - AF447

Thu Jul 05, 2012 9:58 pm

Quoting tommytoyz (Thread starter):
In particular, I think there is a lack of understanding about high altitude flying.

Like to explain how you arrived at this opinion ? What is the basis for you saying this ?

It would not be an informed opinion which any pilot of a jet transport would not agree with.

Quoting tommytoyz (Thread starter):
In these recommendations, BEA departs from the philosophy of only relying on stall avoidance training for pilots. Rather, recommending training in actual stall recovery, especially at altitude. I totally agree with the BEAs recommendations and commend them on a job very well done.

Not being rude, by going by your profile, I do not think you would be able to understand the report in its entirety, the target audience of the report is not the general public. I also do not think you know much about airline initial and recurrent pilot training and checking. I therefore find it difficult then for you to have an informed opinion either way.

It is a pitty when reports like AF447 and you get people coming out from all areas making fantastic observations, many are made by people who have no qualification or experience to base informed opinions with.

It is a thick document, I am yet to learn anything from it.
We are addicted to our thoughts. We cannot change anything if we cannot change our thinking – Santosh Kalwar
 
User avatar
Starlionblue
Posts: 17097
Joined: Fri Feb 27, 2004 9:54 pm

RE: BEA Recommendations - AF447

Thu Jul 05, 2012 11:18 pm

Quoting zeke (Reply 1):
Quoting tommytoyz (Thread starter):
In particular, I think there is a lack of understanding about high altitude flying.

Like to explain how you arrived at this opinion ? What is the basis for you saying this ?

It would not be an informed opinion which any pilot of a jet transport would not agree with.

Quite. You hardly have to scratch the surface of the swept wing, high altitude, high speed jet literature without reading plenty about high altitude flying and its effects.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
tdscanuck
Posts: 8572
Joined: Wed Jan 11, 2006 7:25 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 12:45 am

Quoting tommytoyz (Thread starter):
The change in critical AoA (low speed) with altitude, for instance, small as it is (only a few degrees on most planes), is not widely known.

It's actually a change with Mach, not a change with altitude. The practical impact is the same for most conditions (higher Mach when you're flying high altitude) due to the temperature changes but, crucially, not the same for lower speed maneuvers like a high altitude stall.

This is part of the challenge of pilot training; do you want to go way down into the weeds of very specific and relatively esoteric aerodynamic phenomenon, or do you want to stick with simple, practical rules that can be trained and remembered and not get you in trouble even when it all hits the fan?

Tom.
 
tommytoyz
Topic Author
Posts: 1195
Joined: Thu Jan 18, 2007 9:08 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 12:50 am

Quoting zeke (Reply 1):
It would not be an informed opinion which any pilot of a jet transport would not agree with.

I guess you don't count BEA as credible, which is OK. It's OK to be at odds with what the BEA. But ask yourself - Do you think there is a specific reason they are making these recommendations? Once you read the report you'll see why they made them. Perhaps you don't give any credibility to the BEA. That's the only way to square what you say.

Quoting Starlionblue (Reply 2):
Quite. You hardly have to scratch the surface of the swept wing, high altitude, high speed jet literature without reading plenty about high altitude flying and its effects.

That's not enough. One AF captain here on A.Net said he knew about high altitude because of pamphlets in training. BEA agrees. Why else would they make the recommendations to train more, differently and expressly including high altitude recovery? Unless you give the BEA no credibility, then fine.

Quoting zeke (Reply 1):
Not being rude, by going by your profile, I do not think you would be able to understand the report in its entirety,

We are talking facts. However, it's good that you say it is your belief, because that's all it is.
 
tommytoyz
Topic Author
Posts: 1195
Joined: Thu Jan 18, 2007 9:08 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 1:04 am

Quoting zeke (Reply 1):
It would not be an informed opinion which any pilot of a jet transport would not agree with.

Unless you exclude those at the BEA and Sullenbrger and my sister who is an ex USAF pilot. Zeke, you only represent yourself and your own opinions.

Training also needs improvement. "Currently, to my knowledge, air transport pilots practice approaches to stalls,
never actually stalling the aircraft.
(discussing training)
- "Sully" Sullenberger

I suggest we discuss the facts in the report and the recommendations contained therein. You can agree or disagree, but don't shoot the messenger.
 
tommytoyz
Topic Author
Posts: 1195
Joined: Thu Jan 18, 2007 9:08 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 1:22 am

Quoting tdscanuck (Reply 3):
This is part of the challenge of pilot training; do you want to go way down into the weeds of very specific and relatively esoteric aerodynamic phenomenon, or do you want to stick with simple, practical rules that can be trained and remembered and not get you in trouble even when it all hits the fan?

It's not a hard concept to understand and helps explain a lot of things that change with altitude. For all practical purposes it doesn't really matter if this is known or not. But it shows training has been simplified in certain areas.

I would argue that if the pendulum swings too far in the direction of dumbing everything down and oversimplifying things, for fear of confusing pilots, then that can be dangerous too. I have had two discussion with experienced pilots on this and they have little good to say about how new pilots are trained today.

I don't think the pilots of AF447 were incompetent or dumb. They responded in a manner that their training compelled them to respond. It is certainly not trained or written anywhere to do what they did. However, the totality of their training, resulted in them responding the way they did and were never able to figure it out.
 
User avatar
zeke
Posts: 9854
Joined: Thu Dec 14, 2006 1:42 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 2:58 am

Quoting tommytoyz (Reply 4):
But ask yourself - Do you think there is a specific reason they are making these recommendations?

The two points that you posted in the OP are asking the regulators review and evaluate, they do not require any changes to be made.

Where I work, we do approach to the stall, and stall training, also LOFT style simulator sessions are required under the LOSA system, which we are a member of. We also receive false stall warning training, which is actually more relevant to this incident.

Quoting tommytoyz (Reply 4):
We are talking facts. However, it's good that you say it is your belief, because that's all it is.

I asked a very simple question, what is the basis for your opinions ? i.e. what is your technical competency ? are you rated on the A330 ? are you in an airline check and training system ? are you a pilot ? are you an aeronautical engineer ? do you work for a regulator ?

The report is not designed for the general public to understand, it is written with an assumed level of knowledge by the target audience. The regulators will read the report with their technical competency, they will then evaluate the contents, and compare it to the existing design and operating regulations. They may decide that no changes are required.

Quoting tommytoyz (Reply 5):
Unless you exclude those at the BEA and Sullenbrger and my sister who is an ex USAF pilot. Zeke, you only represent yourself and your own opinions.

I am A330 rated, an airline pilot, and design engineer. I think that makes my opinion more relevant in looking at the report contents than just someone off the street, or paraphrasing other people and taking their opinions out of context. Sullenberger is correct, regular line pilots do not take the real airliners up and actually stall the aircraft, we do in the simulator, both false stall warnings, and stall recoveries. We have all stalled real aircraft before, during training.

Quoting tommytoyz (Reply 5):

I suggest we discuss the facts in the report and the recommendations contained therein. You can agree or disagree, but don't shoot the messenger.

I do not mind discussing the facts, if they are the actual ones in the report.
We are addicted to our thoughts. We cannot change anything if we cannot change our thinking – Santosh Kalwar
 
User avatar
Starlionblue
Posts: 17097
Joined: Fri Feb 27, 2004 9:54 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 3:00 am

Tommytoyz, let's keep it friendly please. We can have a discussion but please try to keep it factual, without posturing or rudeness. If you disagree, just say so and say why. Saying things like "Unless you give the BEA no credibility, then fine." is not very helpful.

Quoting tommytoyz (Reply 4):
Quoting Starlionblue (Reply 2):
Quite. You hardly have to scratch the surface of the swept wing, high altitude, high speed jet literature without reading plenty about high altitude flying and its effects.

That's not enough. One AF captain here on A.Net said he knew about high altitude because of pamphlets in training. BEA agrees. Why else would they make the recommendations to train more, differently and expressly including high altitude recovery? Unless you give the BEA no credibility, then fine.

Of course the literature is not enough. You also have to have training. But I would think most airliner pilots have more knowledge than just pamphlets. Pick up any well-known text like "Handling the Big Jets" and effects of high speed and high altitude with swept wing aircraft are discussed in great detail.

And in any case: Even if the pilots had had no idea, there are defined standard procedures to be followed in case of invalid speed indication. Set known pitch and power and so forth. The AF447 pilots did not follow these procedures.

As Tom says:

Quoting tdscanuck (Reply 3):
This is part of the challenge of pilot training; do you want to go way down into the weeds of very specific and relatively esoteric aerodynamic phenomenon, or do you want to stick with simple, practical rules that can be trained and remembered and not get you in trouble even when it all hits the fan?

***

Quoting tommytoyz (Thread starter):
The change in critical AoA (low speed) with altitude, for instance, small as it is (only a few degrees on most planes), is not widely known.

Correct me if I'm wrong but don't wings stall at the same AoA regardless of speed? Last place I read this was "Pilot's Handbook of Aeronautical Knowledge", an FAA publication. Would be happy to learn more about the subject.

[Edited 2012-07-05 20:09:03]
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
rfields5421
Posts: 5565
Joined: Thu Jul 19, 2007 12:45 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 3:01 am

I expect that an AOA indicator will become standard on newly delivered transport catagory aircraft within the next couple years.

It doesn't matter that it very likely would have had no impact upon this accident.

It will be a CYA item by the manufacturers and the airlines.

I find the discussion about developing better flight simulator fidelity so that training can encompass recovery from actual stalls interesting. Also, that such development and training was part of the NTSB recommendations after the Colgan Air Dash-8 crash.

The very depressing thing in the report for me (besides the obvious lack of comprehension by the crew) was how many of the crews in similar UAS incidents did not recognize them as UAS, and did not apply the correct checklist procedure.
 
User avatar
Starlionblue
Posts: 17097
Joined: Fri Feb 27, 2004 9:54 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 3:05 am

Quoting rfields5421 (Reply 9):
The very depressing thing in the report for me (besides the obvious lack of comprehension by the crew) was how many of the crews in similar UAS incidents did not recognize them as UAS, and did not apply the correct checklist procedure.

Note, the following is my layman's understanding:
Following such procedures (known pitch and power) would do no harm even if the plane is flying fine and and allow the pilots time to troubleshoot the issue in a stable situation as opposed to a potentially rapidly deteriorating one.

Even if they didn't believe they had unreliable airspeed, it shouldn't matter. You only have the instruments and if you discount them you're in a world of trouble.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
rfields5421
Posts: 5565
Joined: Thu Jul 19, 2007 12:45 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 3:14 am

Also from a laymans perspective (since I've never flown anything above a small GA plane, and current medical keeps me for doing that without a safety pilot)

From June 3, 2009 we were discussing UAS on these thread from the ACARS messages. After the first interim report, we were discussing UAS procedures being properly applied by about three dozen crews. Now the details of those few incidents with enough information show those crews also failed to properly apply UAS checklists.

Yes, proper pitch and power would likely have made this a non-event.

I did find in the BEA report something which we already knew from Mandala499's theory - that the PF may have been chasing 12 degrees pitch up - the correct pitch power for a low altitude UAS.

If the PNF had run the checklist - the proper pitch/ power would have been identified.
 
User avatar
zeke
Posts: 9854
Joined: Thu Dec 14, 2006 1:42 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 4:18 am

Quoting rfields5421 (Reply 9):
I expect that an AOA indicator will become standard on newly delivered transport catagory aircraft within the next couple years.

I know of no airliner that actually displays AoA, all that I am aware of have a vane that goes via a computer to give a calibrated value. Their usefulness outside their tested range is not known.

A better solution would be for a laser based system which needs no mechanical part to move, or free vents to operate.
We are addicted to our thoughts. We cannot change anything if we cannot change our thinking – Santosh Kalwar
 
tommytoyz
Topic Author
Posts: 1195
Joined: Thu Jan 18, 2007 9:08 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 6:36 am

Quoting zeke (Reply 7):
We also receive false stall warning training, which is actually more relevant to this incident.

I disagree, the AF447 crew did not receive any false stall warnings. Why then, do you think false stall warnings are relevant to AF447? The report does not mention that as a factor as far as I know.

Quoting zeke (Reply 7):
Sullenberger is correct,

Yes, he is correct that training is deficient, and I agree with that. That was his entire point and where you are at odds with Sully there and it refutes your entire "no airline pilot would agree"...

Quoting Starlionblue (Reply 8):
Correct me if I'm wrong but don't wings stall at the same AoA regardless of speed?

Speed yes, but not at different altitudes on the low end of the speed spectrum (watch many here jump on this) everything else being equal. There are two types of stalls, if you will. High speed, and low speed. The high speed stall is Mach dependent, the low speed is not. Talking strictly about the low speed stall and the critical AoA there, the critical AoA varies slightly with altitude approaching the low speed spectrum. Most people erroneously think low speed critical AoA remains constant at all altitudes, and it is often taught that way, as it varies little, but it does vary.


Quoting rfields5421 (Reply 9):
I expect that an AOA indicator will become standard on newly delivered transport catagory aircraft within the next couple years.

This is a suggestion I have been making for quite some time now. The military has show their usefulness and the improved safety after implementing them. Almost all military fixed wing aircraft display it. I concur with the BEA report and I think they are starting to think along the same lines and learning teh same lessons the military learned a long time ago. Does anyone know if any Af447 pilots had military pilot training?

Quoting zeke (Reply 12):
I know of no airliner that actually displays AoA, all that I am aware of have a vane that goes via a computer to give a calibrated value

That's plenty good enough. The calibrated values are pretty right on as measured in testing, representing the wing AoA well enough to full fill it's purpose. I have never heard anyone saying it's not good enough. Matter of fact the entire Airbus FBW system uses these calibrated AoA values, including the AoA values we know of that AF447 experienced.

Quoting zeke (Reply 7):
Where I work, we do approach to the stall, and stall training,

Does that mean you actually train in stall recovery, post stall, in the SIM - at high altitudes over 20,000 feet? Yes? What type of simulators do you use?

Quoting zeke (Reply 7):
what is your technical competency ? are you rated on the A330 ? are you in an airline check and training system ? are you a pilot ? are you an aeronautical engineer ? do you work for a regulator ?
Quoting zeke (Reply 7):
I am A330 rated, an airline pilot, and design engineer. I think that makes my opinion more relevant in looking at the report contents than just someone off the street,

So what is your opinion about the two recommendations?

[Edited 2012-07-05 23:39:50]
 
airtechy
Posts: 544
Joined: Sun Dec 24, 2006 7:35 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 6:54 am

Quoting tommytoyz (Reply 13):
If the pilots had all gotten up and left the cockpit unattended for 5 minutes, autopilot off and all - they would all probably still be alive today, IMHO.

Actually, one minute would have been enough.

Quoting zeke (Reply 7):
Where I work, we do approach to the stall, and stall training, also LOFT style simulator sessions are required under the LOSA system, which we are a member of. We also receive false stall warning training, which is actually more relevant to this incident.

With all due respect, I don't think you work for Air France so the training you receive has no bearing on the training, or the lack thereof, that the pilots at Air France receive.

Quoting zeke (Reply 7):
I asked a very simple question, what is the basis for your opinions ? i.e. what is your technical competency ? are you rated on the A330 ? are you in an airline check and training system ? are you a pilot ? are you an aeronautical engineer ? do you work for a regulator ?

That is a very condescending attitude. Some of us have experience beyond what might be indicating by posting on this forum. Although I am three of the "are you's" that you posted, the lack of the other three should not prevent me or others from offering ideas or opinions especially with the detailed background and analysis provided by this report.

Jim
 
User avatar
autothrust
Posts: 1458
Joined: Sun Jun 11, 2006 8:54 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 9:01 am

Quoting zeke (Reply 12):
A better solution would be for a laser based system which needs no mechanical part to move, or free vents to operate

Can you please explain me how exactly such a system would work? Just curious.
“Faliure is not an option.”
 
rfields5421
Posts: 5565
Joined: Thu Jul 19, 2007 12:45 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 1:35 pm

Quoting tommytoyz (Reply 13):
Does anyone know if any Af447 pilots had military pilot training?

The summary of the career progression of the three pilots does not indicate any military experience.

The captain - PPL in 74, FA with Air France from 76 to 82 - and completed his CP, IFR, CFI and ATP during that period. Started his professional flying career Inter Avia Service company in 83, got on with Air Inter 88, Air Inter and Air France merged in 97

The PNF Co-Pilot - PPL in 92, apparently was on the Air France training program, but delayed by the early 90s economic downturn, started Air France type rating in mid-1998 and began flying as an A320 co-pilot that fall.

The PF Co-Pilot - PPL in 2000, trained by AF in 2003, A320 in 2004

Quoting airtechy (Reply 14):
Quoting tommytoyz (Reply 13):If the pilots had all gotten up and left the cockpit unattended for 5 minutes, autopilot off and all - they would all probably still be alive today, IMHO.

Actually, one minute would have been enough.

I disagree with both timelines.

Unless you mean the crew should have not touched the controls when the AP disconnected.

Just before the one minute mark after A/P disconnect - the aircraft was stalled. And fully stalled past the point it could fly itself out.

Even during the inital disconnect and during the first minute - the aircraft wanted to roll right. Leaving his hands off the controls would likely have allowed the aircraft to roll sufficiently right to become unable to fly.
 
tdscanuck
Posts: 8572
Joined: Wed Jan 11, 2006 7:25 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 4:00 pm

Quoting Starlionblue (Reply 8):
Correct me if I'm wrong but don't wings stall at the same AoA regardless of speed? Last place I read this was "Pilot's Handbook of Aeronautical Knowledge", an FAA publication.

There is some Mach dependance (modern airfoils have a shockwave on the upper aft surface, this shock can trigger seperation). At low speeds, despite Tommytoyz insistance, there's no altitude variation (we have a longstanding disagreement on this point that's not going to get resolved here). Unless you get Mach (or, equivalently, Reynolds number) effects into play, stall AoA is constant.

Quoting Starlionblue (Reply 10):
Following such procedures (known pitch and power) would do no harm even if the plane is flying fine and and allow the pilots time to troubleshoot the issue in a stable situation as opposed to a potentially rapidly deteriorating one.

Correct.

Quoting zeke (Reply 12):
I know of no airliner that actually displays AoA, all that I am aware of have a vane that goes via a computer to give a calibrated value.

It's an option on all current Boeings; relatively few airlines take the option.

Quoting rfields5421 (Reply 16):
Just before the one minute mark after A/P disconnect - the aircraft was stalled. And fully stalled past the point it could fly itself out.

Stalled past the point of recover is a deep stall; AF447 was never in a deep stall. They had nose-down pitch authority the entire time. They could have flown out of it at any time as long as they had enough altitude.

Quoting rfields5421 (Reply 16):
Even during the inital disconnect and during the first minute - the aircraft wanted to roll right. Leaving his hands off the controls would likely have allowed the aircraft to roll sufficiently right to become unable to fly.

There's no such thing as a roll attitude at which you can't fly. Although highly not recommended, modern airliners are perfectly capable of rolling and recovering.

Tom.
 
rfields5421
Posts: 5565
Joined: Thu Jul 19, 2007 12:45 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 4:18 pm

Quoting tdscanuck (Reply 17):
They could have flown out of it at any time as long as they had enough altitude.

The report list 31,500 feet as unrecoverable except by a crew trained for that specific flight event and expecting to execute the maneuvers. The aircraft was passing through 31,500 feet a little before the 1 minute mark.

There must be some simulator exercises or math or other data to support their conclusion - I only reported their conclusion.

[Edited 2012-07-06 09:19:29]
 
canoecarrier
Posts: 2569
Joined: Sat Feb 14, 2004 1:20 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 4:39 pm

Quoting tdscanuck (Reply 17):

There's no such thing as a roll attitude at which you can't fly. Although highly not recommended, modern airliners are perfectly capable of rolling and recovering.

Haven't we discussed that a rather significant roll would have been the only way to recover from the stall once the plane passed through 31,500'? I thought at some point just pointing the nose down would not have allowed the plane to recover. Although, I may be remembering a pre-Interim Report discussion where we hypothesized they could have entered a deep stall.

Quoting Starlionblue (Reply 8):
And in any case: Even if the pilots had had no idea, there are defined standard procedures to be followed in case of invalid speed indication. Set known pitch and power and so forth. The AF447 pilots did not follow these procedures.

Sort of. As Fields mentions it is possible they were using the values for the wrong phase of flight.

Quoting rfields5421 (Reply 11):
I did find in the BEA report something which we already knew from Mandala499's theory - that the PF may have been chasing 12 degrees pitch up - the correct pitch power for a low altitude UAS.

Right. The report discusses:

"the flight director displays could have prompted him to command a positive pitch angle, of about 12.5°. This value appears in the stall warning procedure for the take-off phase. It is possible that, even though he did not call it out, the PF had recalled this memorised value and then had clung to this reference without remembering that it was intended for a different flight phase. The conjunction of this remembered value and the flight director displays may have constituted one of the few (and maybe even the only) points of consistency in his general incomprehension of the situation"

[Edited 2012-07-06 09:40:36]
The beatings will continue until morale improves
 
tommytoyz
Topic Author
Posts: 1195
Joined: Thu Jan 18, 2007 9:08 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 4:49 pm

Quoting rfields5421 (Reply 16):
Just before the one minute mark after A/P disconnect - the aircraft was stalled. And fully stalled past the point it could fly itself out.

Even during the inital disconnect and during the first minute - the aircraft wanted to roll right. Leaving his hands off the controls would likely have allowed the aircraft to roll sufficiently right to become unable to fly.

That is actually one recovery technique from a stall. The 727, while in flight testing, it was decided to stall it. Well, it is a T-Tail and it can get into a deep stall, like most T-Tails. That is the horizontal stabilizer is in th wake of the main wing and is ineffective and normal revocery is impossible. The Boeng test pilot, anticipating this might happen from previous Bac-111 and DC-9 crashes due to this, was able to recover from the deep stall, by rolling steeply and letting the nose fall out.

I don't think the A330 would have stalled if left to fly without anyone touching the controls and left to it's inherent aerodynamic stability, despite turbulence. Might have been a wild ride at worst, but that's about it.

Quoting tdscanuck (Reply 17):
It's an option on all current Boeings; relatively few airlines take the option.

The same for A330s, though not on AF A330s.

[Edited 2012-07-06 09:52:09]
 
airtechy
Posts: 544
Joined: Sun Dec 24, 2006 7:35 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 6:22 pm

Actually, as we have discussed before, they would have been better off if the auto-thrust part of the autopilot had disconnected but the pitch and roll part had remained active. In a minute or so the correct airspeed would have been restored by the heaters and the auto-thrust would have/could have re-engaged.

As Tom has pointed out, this would have worked in this case but maybe not others.

I believe the auto-trim cranked in max up trim due to limits being removed in the ALT 2 LAW mode which, even with full down stick command, would have taken a while to crank out. Still they should have had enough altitude to recover.

Although it's obvious the pilots screwed up royally here, I think the report was light on criticism of Airbus regards the way information was being displayed to the pilots, the audible alarms adding to the confusion, the way the airplane handles in alternate law modes, and other things.

We have not had a high profile accident involving a large Boeing aircraft, thank goodness, so we don't have the type detailed analysis that this report provides us about the way the Airbus works. Who knows, Boeings may have similar issues.

Jim
 
canoecarrier
Posts: 2569
Joined: Sat Feb 14, 2004 1:20 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 7:37 pm

If the purpose of this thread is to discuss the Final Report..which I'm not exactly sure it is, I'd add that the BEA did have some interesting observations about search and rescue as well as ATC in these remote areas of the world.

Particularly, the recommendation that a permanent and reliable link be made to between the aircraft and the ground in all areas where HF remains the only means of communication. And, that the SAR organizations in both Brazil and France could have operated more efficiently. Leading to a recommendation that coordination plans be made in the event something like this occurs again.

I'm not sure either recommendation would have made any difference here. It's unlikely that even with a permanent air to ground voice communication system that the pilots would have called out their position. Certainly the only thing that may have been improved with better coordination is that the crash site might have been not taken 3 tries to find if SAR aircraft were on scene earlier.

They do recommend the mandatory utilization of ADS-C in these areas. I thought Tom mentioned that all new aircraft are being fitted with this technology?
The beatings will continue until morale improves
 
rfields5421
Posts: 5565
Joined: Thu Jul 19, 2007 12:45 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 8:05 pm

Quoting airtechy (Reply 21):
I think the report was light on criticism of Airbus regards the way information was being displayed to the pilots, the audible alarms adding to the confusion, the way the airplane handles in alternate law modes, and other things.

I find the report to be a strong negative of the way all aircraft manufacturers do such things - the presentation of data, the heavy reliance on audible alarms when they are known to be less effective than visual alarms, etc.

While the report does not specifically mention Boeing aircraft - being outside the scope of the report - the implication is clear that Boeing, Bombardier, Embraer, etc - all provide sub-optimal information to pilots when things go to heck.

I find the report to says there are serious problems in the entire current systems. This accident should not have been possible in these modern aircraft.

Quoting canoecarrier (Reply 22):
I'm not sure either recommendation would have made any difference here.

The SAR recommendations would not have done anything for this crew/ passengers. However, if the crew at taken actions sufficent to make this a survivable crash - in my opinion - the SAR system would have failed to get to the survivors in time to rescue them.

Reading the chronology of the confusion, lack of understanding of areas of responsiblity, not even having the correct phone numbers to talk to other centers watching the area - scary stuff if you could end up in a raft waiting for rescue.

BEA is recommending some fundamental changes in the way airlines do business, the way authorities do business, etc.
 
canoecarrier
Posts: 2569
Joined: Sat Feb 14, 2004 1:20 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 8:33 pm

Quoting rfields5421 (Reply 23):
While the report does not specifically mention Boeing aircraft - being outside the scope of the report - the implication is clear that Boeing, Bombardier, Embraer, etc - all provide sub-optimal information to pilots when things go to heck.

BEA has put together a number of working groups on other systems (e.g., transmission of DFDR/DCVR data, locator beacons) that are not aircraft manufacturer specific. Therefore, there is the likelihood that some of the information gleaned from the development of this report may later be shared with other aviation regulatory bodies that could contribute to an overall improvement in flight safety regardless of the manufacturer (e.g., CRM).

Quoting rfields5421 (Reply 23):
I find the report to says there are serious problems in the entire current systems. This accident should not have been possible in these modern aircraft.

It's interesting that the report does mention that the PF may have embraced the common belief that this (modern) airplane could not stall.
The beatings will continue until morale improves
 
tommytoyz
Topic Author
Posts: 1195
Joined: Thu Jan 18, 2007 9:08 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 10:18 pm

Quoting canoecarrier (Reply 22):
If the purpose of this thread is to discuss the Final Report..which I'm not exactly sure it is,

That was the intent. Especially the two recommendations I copied from the report.

I would like to explore further, First: Why, as the report says (page 182), regarding the flight director, that "...the flight director was advising a nose-up attitude...." Why would the flight director indicate that in a fully stalled condition at cruise altitude? That could really confuse any pilot, me thinks into pulling.

Secondly - and this may answer the 1st - An article in Flight Global made a good point, were it is explained that while "Power then attitude" had been taught in jets for stall recovery for a long time - the aircraft themselves were never certified that way at altitude. Test pilots used "attitude then power" at altitude.

So a seemingly experimental recovery technique was being institutionalized for high altitude stall recovery and avoidance. The effects of under wing engines were perhaps not examined closely enough in establishing this, for instance. Maybe the Flight Director was programmed with this philosophy, without it ever having been proven in certification flights for the A330?

[Edited 2012-07-06 15:23:09]
 
airtechy
Posts: 544
Joined: Sun Dec 24, 2006 7:35 am

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 10:18 pm

Quoting canoecarrier (Reply 24):
It's interesting that the report does mention that the PF may have embraced the common belief that this (modern) airplane could not stall

Which would indicate that the pilot either through lack of training or confusion did not understand how the various "LAWS" affected the control of the airplane.

Jim
 
canoecarrier
Posts: 2569
Joined: Sat Feb 14, 2004 1:20 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 10:51 pm

Quoting airtechy (Reply 26):
Which would indicate that the pilot either through lack of training or confusion did not understand how the various "LAWS" affected the control of the airplane.

The report says there was a lack of training. And confusion. It doesn't say that had anything to do with which "LAW" the plane was in.

edit: I think any Airbus pilot would or should know what protections are gone when you enter Alternate Law.

[Edited 2012-07-06 15:53:09]
The beatings will continue until morale improves
 
User avatar
Starlionblue
Posts: 17097
Joined: Fri Feb 27, 2004 9:54 pm

RE: BEA Recommendations - AF447

Fri Jul 06, 2012 11:54 pm

Quoting rfields5421 (Reply 23):
I find the report to says there are serious problems in the entire current systems. This accident should not have been possible in these modern aircraft.

Well... That would mean the aircraft would have to have the means to recover by itself. And that means taking away those means from the pilots (since they can apparently make mistakes). That seems counterproductive. Pilots must retain the authority to recover the situation. It therefore follows they will have the authority to crash the plane.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
rfields5421
Posts: 5565
Joined: Thu Jul 19, 2007 12:45 am

RE: BEA Recommendations - AF447

Sat Jul 07, 2012 12:16 am

Quoting Starlionblue (Reply 28):
Pilots must retain the authority to recover the situation.

I'm not saying pilots should not have the ability to turn off protections, or that aircraft should not drop protections when the data sources necessary to provide those protections fail as in this instance.

The aircraft should be able to provide the pilots with clearer indications of the problem, both the initial problem and the later fatal problem.

The training system should have made these pilots better prepared to deal with the initial problem, and not create the stall.

Part of the problem in my mind after reading the report is the reliance on aural alarms.

We've heard - "They had 54 seconds of continuous stall warning - how much more did the need?" over and over. Apparently they do need something more in a very high stress situations.

We now know there were two other aural warnings occurring, and from the BEA Human Factors group that aural warnings are secondary and easy to ignore - especially when conflicting visual information is presented to the pilots - such as the FD appearing and indicating for the pilots to pitch up.
 
User avatar
Starlionblue
Posts: 17097
Joined: Fri Feb 27, 2004 9:54 pm

RE: BEA Recommendations - AF447

Sat Jul 07, 2012 3:20 am

Quoting rfields5421 (Reply 29):
The aircraft should be able to provide the pilots with clearer indications of the problem, both the initial problem and the later fatal problem.

I don't think the aircraft was unclear. It told them speed data was invalid and the pilots confirmed it on the CVR recording. It told them the autopilot disconnected and the pilots confirmed it. It told them they reverted to alternate law and the pilots confirmed it. It told them the aircraft was stalled. In fact it warned them of an impending stall 1m36s before the actual stall and multiple times after that.

Short of banging the pilots over the head with a cartoon mallet I don't know what exactly the instrumentation could have done to be more obvious.

Quoting rfields5421 (Reply 29):
Part of the problem in my mind after reading the report is the reliance on aural alarms.

We've heard - "They had 54 seconds of continuous stall warning - how much more did the need?" over and over. Apparently they do need something more in a very high stress situations.

We now know there were two other aural warnings occurring, and from the BEA Human Factors group that aural warnings are secondary and easy to ignore - especially when conflicting visual information is presented to the pilots - such as the FD appearing and indicating for the pilots to pitch up.

We've discussed this in the past. Tunnel vision is part of the problem. Once you've decided you don't trust the instruments, and your mind is taking you down a course of action, it is easy to discard conflicting inputs, especially under stress.

Stall warning should lead to pilots pitching down in all cases, right? But it didn't. I don't think visually presenting the information would have made a difference in this case.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
Slcpilot
Posts: 565
Joined: Thu Aug 07, 2003 3:32 am

RE: BEA Recommendations - AF447

Sat Jul 07, 2012 12:19 pm

While this post is a little late, I think some the original concerns about the nature of the OPs post are unfounded.

To me, tommytoyz sounds like a screen name a very young person might choose.

That being said, his posts are always articulate, educated, and thought provoking (unlike mine).

Cheers!

SLCPilot
I don't like to be fueled by anger, I don't like to be fooled by lust...
 
OldAeroGuy
Posts: 3188
Joined: Sun Dec 05, 2004 6:50 am

RE: BEA Recommendations - AF447

Sat Jul 07, 2012 2:49 pm

Quoting Starlionblue (Reply 8):
Correct me if I'm wrong but don't wings stall at the same AoA regardless of speed?
Quoting tommytoyz (Reply 13):
There are two types of stalls, if you will. High speed, and low speed. The high speed stall is Mach dependent, the low speed is not.

All stalls are Mach related. You need to correct the stall AoA (and stall CL) for Mach effects across the flight envelope. Below 0.7M or so, the corrections are small. At higher Machs, the effects are more pronounced with significant reductions in stall AoA and stall CL.

Without a source of reliable airspeed (ie Mach), it's not possible to provide a totally accurate stall AoA. However, I agree that a flight deck indication of AoA might have helped the AF447 situation.
Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
 
tommytoyz
Topic Author
Posts: 1195
Joined: Thu Jan 18, 2007 9:08 am

RE: BEA Recommendations - AF447

Sat Jul 07, 2012 5:20 pm

Quoting OldAeroGuy (Reply 32):
Without a source of reliable airspeed (ie Mach), it's not possible to provide a totally accurate stall AoA. However, I agree that a flight deck indication of AoA might have helped the AF447 situation.

An AoA indicator would have shown the trajectory of their AoA and that their inputs were making the AoA worse, not better, even with inaccuracies due to no speed.

Another question in my mind is why GPS speeds can't be used in a pinch such as this? Winds aloft would be know by the pilots, so within a few knots, the GPS calculated speed should help pilots avoid catastrophe. No?
 
Mir
Posts: 19093
Joined: Mon Jan 05, 2004 3:55 am

RE: BEA Recommendations - AF447

Sat Jul 07, 2012 5:47 pm

Quoting tommytoyz (Reply 13):
I disagree, the AF447 crew did not receive any false stall warnings.

They absolutely did, right after the problems started occurring.

Quoting tommytoyz (Reply 33):
Another question in my mind is why GPS speeds can't be used in a pinch such as this? Winds aloft would be know by the pilots, so within a few knots, the GPS calculated speed should help pilots avoid catastrophe. No?

You can't fly a groundspeed, which is what GPS gives you - the difference between that and IAS at altitude is too great. Winds aloft are only known because the aircraft's systems can use air data combined with GPS or inertial information to determine them - take the air data away and you've lost that ability.

But what people who ask about GPS information always seem to forget is that the crew had GPS information readily available to them, and could have easily looked at their groundspeed if they wanted to. But putting it up on the airspeed tape is not a good idea.

-Mir
7 billion, one nation, imagination...it's a beautiful day
 
rfields5421
Posts: 5565
Joined: Thu Jul 19, 2007 12:45 am

RE: BEA Recommendations - AF447

Sat Jul 07, 2012 6:58 pm

Quoting tommytoyz (Reply 33):
Winds aloft would be know by the pilots

Actually no - pilots don't know the winds aloft.

They know what is predicted and over the period of a half-hour or so, can calculate the average winds.

But at cruise level for a passenger jet - the margin between flying, stalling and over-speed is relatively small - average winds over the past half hour often would not be good enough to rely upon.

Traveling through disturbed air - like the ITCZ where this aircraft was flying - the winds aloft can change directions around the compass and speeds by 50 or 100 kts in a few seconds.

Because of the way aircraft work in the air as a fluid - this doesn't produce a dramatic change in the Indicated Air Speed - but can produce huge dramatic changes in the GPS calculated ground speed.

Quoting Mir (Reply 34):
the crew had GPS information readily available to them

As Mir said - they had the GPS information available to them. One item on the UAS checklist is to display the GPS altitude on the MCDU - because UAS also means the Vertical Speed Indicator is likely not working properly.

This crew never ran the UAS checklist.

This crew never indicated at any point that they thought their airspeed was low. Only that they "lost the speeds". They might have though they were too fast based on some of their actions and words. They appear to have though the computers/ instruments were malfunctioning.

Also note, of the other crews with documented UAS incidents - they didn't run the checklist either. But their basic airmanship procedures did not place those aircraft in mortal danger.
 
User avatar
zeke
Posts: 9854
Joined: Thu Dec 14, 2006 1:42 pm

RE: BEA Recommendations - AF447

Sat Jul 07, 2012 8:37 pm

Quoting tommytoyz (Reply 13):
Why then, do you think false stall warnings are relevant to AF447?

What confirms the difference between a real stall warning and a false warning ?

A bit like having a barometer at home indicating 28.5" and saying it will rain with blue skies outside.

CONFIRMATION, means processing and understanding the information being presented and making a decision to establish if it is sensible.

Quoting tommytoyz (Reply 13):

The military has show their usefulness and the improved safety after implementing them. Almost all military fixed wing aircraft display it.

What you failed to mention is those military aircraft are usually relatively short skinny aircraft, and the do not have large differences between the sides of the fuselage, or the distance from the nose to the wing, and have low aspect ratio wings. The may even have the AoA on the wing itself.

It is very possible for AoA vanes to indicate that an airliner is not stalled, whilst one wing is stalled and he other is not, AoA probes on airliners are located normally on the forward fuselage, they do not measure the wing AoA, and can provide the incorrect information to a pilot. They are not placed on the wing as they have leading edge devices.

Boeing has implementation an AoA indicator on the PFD for some operators. With unreliable airspeed the red stalled band can actually show the aircraft is stalled, when not stalled. It is just another source of potentially confusing information to the pilots. It is not a complete fix for the problem.

Quoting tommytoyz (Reply 13):
Yes, he is correct that training is deficient, and I agree with that. That was his entire point and where you are at odds with Sully there and it refutes your entire "no airline pilot would agree"...

I think you have taken his comments out of context. I am not a fan of making the next simulator session as a " let's redo xyz accident", I am a fan of the basics, like pitch and power, I.e. preventing incidents, rather than spring loading people to react and not think.

Quoting tommytoyz (Reply 13):

So what is your opinion about the two recommendations?

I think the regulators should evaluate all the recommendations. My personal view is that technology currently exists in the form of laser based airspeed computation which requires no probes or vanes> I think the main reason why they are not installed on airliners is due to accountants, not due to the technology being unavailable, and not because it has no safety benefit.

I am not a fan on cockpit videos, I am for the potential safety benefit, however in this part of the world they would not be used as intended. Crews already are automatically arrested following incidents. With a video, my fear is that they will get processed, and the families billed for the $0.20 for the bullet that was used to carry out the sentence following the short trial the next day. The principle of a fair trial does not exist worldwide. The family will naturally be able to appeal the sentence, they may get a refund on the $0.20.

Quoting tdscanuck (Reply 17):
It's an option on all current Boeings; relatively few airlines take the option.

The AoA displayed on a Boeing airliner is not the wing AoA, it is as I said, a calibrated value which goes via a computer. It requires valid airspeed data to function correctly, with unreliable airspeed it can also provide misleading information.

Quoting airtechy (Reply 14):
With all due respect, I don't think you work for Air France so the training you receive has no bearing on the training, or the lack thereof, that the pilots at Air France receive.

I am under the same regulatory standards as AF (EU OPS), the training listed in the report is very similar to what I have received.

Quoting autothrust (Reply 15):

Can you please explain me how exactly such a system would work? Just curious.

Basically the existing systems use bouncing light off air molecules and analyzing the changes in reflection caused by the air%u2019s motion relative to the aircraft, no reason why something like a Microsoft Kinnect could not be used either. The principle relies on something being in-font of the aircraft, that could the air, cloud, ice, snow, dust, sand etc to produce change in time of flight of the light source sent out. They are currently used in a few aviation applications, like for helicopters where at low speed pitots are next to useless, and visual contact with the ground can be obscured with sand or snow.

Quoting tdscanuck (Reply 17):
There is some Mach dependance (modern airfoils have a shockwave on the upper aft surface, this shock can trigger seperation). At low speeds, despite Tommytoyz insistance, there's no altitude variation (we have a longstanding disagreement on this point that's not going to get resolved here). Unless you get Mach (or, equivalently, Reynolds number) effects into play, stall AoA is constant.

They did decelerate from their cruise speed, in my view mach effects were negligible.

Quoting rfields5421 (Reply 18):
The report list 31,500 feet as unrecoverable except by a crew trained for that specific flight event and expecting to execute the maneuvers.

That is not exactly what the report said.

Quoting canoecarrier (Reply 19):

Haven't we discussed that a rather significant roll would have been the only way to recover from the stall once the plane passed through 31,500'?

I have suggested this before, the aircraft in my view was flown into an attitude where I think a more appropriate technique would be one for nose high jet upset, the reason being the aircraft was flown into a condition were airspeed was inappropriate for the conditions. An aircraft still has a lot of potential energy at FL315, and the ailerons and roll spoilers would still have been effective.

The tests normally applied for the application of a jet upset recovery are as follows :

Pitch attitude more than 25 degrees nose up.
Pitch attitude more than 10 degrees nose down.
Bank angle more than 45 degrees.
Flight within these parameters at airspeeds inappropriate for the conditions.

They should have tried a recovery with full nose down elevator and manually applied some nose down stabilizer trim and reduced thrust, rolling the aircraft to around 45 degrees would have assisted in starting the nose down. That should then change the pitch attitude to allow airspeed to increase, thereby improving elevator and aileron control effectiveness.

The crew had other options as well, including gear, spoilers, and flaps, I would not have given up.

Quoting tommytoyz (Reply 20):
I don't think the A330 would have stalled if left to fly without anyone touching the controls and left to it's inherent aerodynamic stability, despite turbulence. Might have been a wild ride at worst, but that's about it.

Probably the most sensible statement you made all thread. Why provide pilots with even more information to entice them to touch something, where the best course of action in the cruise is to do nothing until it has been confirmed what is going on.

Quoting airtechy (Reply 21):
We have not had a high profile accident involving a large Boeing aircraft, thank goodness, so we don't have the type detailed analysis that this report provides us about the way the Airbus works. Who knows, Boeings may have similar issues.

The reports lists a number of similar accidents in it with Boeing/McD aircraft, I get the impression you have not read it.

Quoting canoecarrier (Reply 22):

They do recommend the mandatory utilization of ADS-C in these areas. I thought Tom mentioned that all new aircraft are being fitted with this technology?

That is FANS B, however it needs a ground VHF datalink network. A simple, cheap. commercially available option to get position reporting is a low earth satellite system, like the SPOT tracker ( http://www.findmespot.com/en/index.php?cid=101 ), one would not think it would be too hard to get something like that into an airliner and to transmit its GPS position/speed/alt every couple of minutes. They also have automatic messaging to the GEOS Rescue Coordination Center as well as part of the product.

Quoting tommytoyz (Reply 25):
Why would the flight director indicate that in a fully stalled condition at cruise altitude?

That is exactly how it is supposed to work, flight directors only do what the FMA is saying it is doing. When the flight directors came back on, they do so in the basic HDG V/S mode, and the V/S would correspond to the climb they were in. This is common for TCAS, EGPWS, stall etc. The choice is, turn them off, or set the FMA to what you want and follow them. This is basic stuff familiar to any pilot.

Quoting tommytoyz (Reply 25):
An article in Flight Global made a good point, were it is explained that while "Power then attitude" had been taught in jets for stall recovery for a long time - the aircraft themselves were never certified that way at altitude. Test pilots used "attitude then power" at altitude.

Have you got a link to that ? I suspect you have not understood what is being said. recovery from a low speed scenario involves thrust first, stall recovery is not the same.

Quoting airtechy (Reply 26):
Which would indicate that the pilot either through lack of training or confusion did not understand how the various "LAWS" affected the control of the airplane.

The aircraft are controlled the same way, regardless of what law the aircraft is in.

Quoting OldAeroGuy (Reply 32):
Below 0.7M or so, the corrections are small. At higher Machs, the effects are more pronounced with significant reductions in stall AoA and stall CL.

They had actually slowed down, as you know they would have need to increase the LF to stall at the higher mach number.

Quoting tommytoyz (Reply 33):
Another question in my mind is why GPS speeds can't be used in a pinch such as this?

You can, just depends on what you mean by "use". These pitot blockages only tend to last short period of time, it is unlikely that any large change wind speed would occur in a few minutes. No reason why you cannot have a look at it to see if there is any significant trend developing. The ground speed should not change much at all if you maintain the pitch and power.
We are addicted to our thoughts. We cannot change anything if we cannot change our thinking – Santosh Kalwar
 
canoecarrier
Posts: 2569
Joined: Sat Feb 14, 2004 1:20 pm

RE: BEA Recommendations - AF447

Sat Jul 07, 2012 9:41 pm

Great post Zeke.

Quoting zeke (Reply 36):
I am not a fan on cockpit videos, I am for the potential safety benefit, however in this part of the world they would not be used as intended.

What if the video only focused on the cockpit instruments and didn't show the pilots themselves? I'd have to go back to some of the technical reports BEA put out related to triggered transmission of flight data, but I don't see a (technical or professional) problem (in an emergency) of using an automatic uplink of DFDR/DCVR data along with a visual image of what the pilots see in front of them. It took 2-3 years to find the crash site in this case, then recover the recording devises before we had a reasonable idea what happened that evening.
The beatings will continue until morale improves
 
Mir
Posts: 19093
Joined: Mon Jan 05, 2004 3:55 am

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 1:57 am

Quoting canoecarrier (Reply 37):
What if the video only focused on the cockpit instruments and didn't show the pilots themselves?

It would defeat the point of the video. You might as well just record what the avionics computers are putting up onto the screens (which should be done, IMO, for both pilots).

-Mir
7 billion, one nation, imagination...it's a beautiful day
 
airtechy
Posts: 544
Joined: Sun Dec 24, 2006 7:35 am

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 4:17 am

Quoting zeke (Reply 36):

Quoting airtechy (Reply 26):
Which would indicate that the pilot either through lack of training or confusion did not understand how the various "LAWS" affected the control of the airplane.

The aircraft are controlled the same way, regardless of what law the aircraft is in.

I understand that, but evidently the transfer function of the control loop is different....higher gain in ALT B and DIRECT LAW vs NORMAL LAW. This could lead to overcontrolling the airplane especially in one that is "neutrally" stable.

Quoting zeke (Reply 36):
That is FANS B, however it needs a ground VHF datalink network. A simple, cheap. commercially available option to get position reporting is a low earth satellite system, like the SPOT tracker ( http://www.findmespot.com/en/index.php?cid=101 )

The SPOT tracker is a one-way device that transmits only. This would probably be OK in an airplane due to the lack of obstructions to the satellite, but there would be no confirmation that the data was received. It works great on my motorcycle though.

Quoting zeke (Reply 36):

Quoting airtechy (Reply 21):
We have not had a high profile accident involving a large Boeing aircraft, thank goodness, so we don't have the type detailed analysis that this report provides us about the way the Airbus works. Who knows, Boeings may have similar issues.

The reports lists a number of similar accidents in it with Boeing/McD aircraft, I get the impression you have not read it.

I have read the report and the sections that describe the operation....and consequences of operation....several times. My comment was not specifically regards accidents, just the lack of detailed info on the Boeing's.

Quoting zeke (Reply 36):
Quoting airtechy (Reply 14):
With all due respect, I don't think you work for Air France so the training you receive has no bearing on the training, or the lack thereof, that the pilots at Air France receive.

I am under the same regulatory standards as AF (EU OPS), the training listed in the report is very similar to what I have received.

I probably should have said "specific training" as my understanding is each airline tailors training to there requirements and aircraft. In this case, the subject of "high altitude stall training" or the lack thereof is a key part of the report. Some airlines may have actually trained for this and after this accident I'm sure all will at least touch on it now that we know even the most automated airplane can be stalled.....under the right conditions. In this case it was the lack of airspeed killing the protections.

Am I reading correctly from the report that even in the presence of the stall alarm that the command bars were saying to increase the angle of attach? That would certainly be confusing. Which brings up the question....how much of the data displayed on the FD is accurate with incorrect airspeed? I would assume attitude and altitude at least.

Jim
 
User avatar
zeke
Posts: 9854
Joined: Thu Dec 14, 2006 1:42 pm

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 5:01 am

Quoting airtechy (Reply 39):

I understand that, but evidently the transfer function of the control loop is different....higher gain in ALT B and DIRECT LAW vs NORMAL LAW. This could lead to overcontrolling the airplane especially in one that is "neutrally" stable.

I say again, they are controlled the same way, regardless of the law. With direct you have the addition of manual pitch trim. None of the laws requires super human piloting skills, nor are the significantly different. That is a certification requirement.

Quoting airtechy (Reply 39):

I have read the report and the sections that describe the operation....and consequences of operation....several times. My comment was not specifically regards accidents, just the lack of detailed info on the Boeing's.

You are not the target audience of the report, the sort of questions you are asking makes that clear. The regulators have those other reports, those reports also had recommendations which the regulators have already evaluated. They will go back and have a look at those reports, as well as this one, and their previous evaluations and make a reassessment on the previous reports as well as an evaluation on this report. This report will not be looked at in isolation or previous events.

The regulator needs to have certification requirements that cover all situations, not just this isolated event.

Quoting airtechy (Reply 39):
I probably should have said "specific training" as my understanding is each airline tailors training to there requirements and aircraft. In this case, the subject of "high altitude stall training" or the lack thereof is a key part of the report

Perhaps you should have asked that question earlier rather than saying "I don't think you work for Air France so the training you receive has no bearing on the training, or the lack thereof, that the pilots at Air France receive".

The regulatory training and checking schedule items are the same, operators will change the departure and arrival airports, location in the world where the simulator events take place, the schedule items that need to be completed should all be the same regardless of the operator, what they use to achieve them is up to them.

The schedule outlines that there needs to be things like electrical malfunctions, engine failures, instrument problems, hydraulic problems, low visibility operations, cold weather operations, downgrades etc etc etc, the list is long. How these are achieved are left to the specific operator, they may try to have similar profiles between fleets, or may have fleet specific profiles. In any case, the schedule items are covered.

These profiles are done a minimum of twice a year, the malfunctions are different, so are the airports. In addition to the actual simulator events, we have pre-study modules, they are normally associated with with the events in the training session, the point of the training session is to reinforce the pre-study material, the briefing before the session that expands on that material, and then to see it in action. These simulators cost a lot to operate, and the people who develop the training matrix of events to be covered I think do a good job. These sessions are not about teaching you to fly, it is about expanding your knowledge, and giving exposure.

Quoting airtechy (Reply 39):
Am I reading correctly from the report that even in the presence of the stall alarm that the command bars were saying to increase the angle of attach?

Only to someone who does not fly them, they are doing exactly as the FMA is saying, they are behaving as I would have expected.
We are addicted to our thoughts. We cannot change anything if we cannot change our thinking – Santosh Kalwar
 
airtechy
Posts: 544
Joined: Sun Dec 24, 2006 7:35 am

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 5:34 am

Quoting zeke (Reply 40):
Quoting airtechy (Reply 39):

I understand that, but evidently the transfer function of the control loop is different....higher gain in ALT B and DIRECT LAW vs NORMAL LAW. This could lead to overcontrolling the airplane especially in one that is "neutrally" stable.

I say again, they are controlled the same way, regardless of the law. With direct you have the addition of manual pitch trim. None of the laws requires super human piloting skills, nor are the significantly different. That is a certification requirement.

So are you saying that the response of the airplane is identical with regards to sidestick movement in any of the LAWS as long as we don't exceed what would have been a protection limit in NORMAL LAW.....ignoring the trim difference? Same rates with the same deflection with the same filtering? I ask because this guy seemed to be overcorrecting... especially in roll...if you look at the FDR trace oscillations.

Quoting zeke (Reply 40):
Quoting airtechy (Reply 39):
Am I reading correctly from the report that even in the presence of the stall alarm that the command bars were saying to increase the angle of attach?

Only to someone who does not fly them, they are doing exactly as the FMA is saying, they are behaving as I would have expected.

The report seems to indicate that the FD was providing bad command bars intermittently because of the LOA...ie UP command. Are you saying that this is what you would expect? I believe this was in reference to the fact that one of the checklists that was not run said to turn the FD off.....probably for this reason.

I'll have to ask the Delta 330 crew on my next NRT-BKK flight if that is the case....plus a few other questions.  

Jim
 
User avatar
zeke
Posts: 9854
Joined: Thu Dec 14, 2006 1:42 pm

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 6:34 am

Quoting airtechy (Reply 41):
So are you saying that the response of the airplane is identical with regards to sidestick movement in any of the LAWS as long as we don't exceed what would have been a protection limit in NORMAL LAW.....ignoring the trim difference?

Flying 101 is the operation and effects of controls. The control input response is not identical with variation of speed or flight control system/law. How a pilot controls the aircraft however is identical, they make the necessary inputs to generate their desired response.

I have flown and landed the A330 in all 3 laws, in the emergency electrical configuration, high speed dives, and even stalled it at idle thrust, it is very docile across the flight envelope regardless of the law.

Quoting airtechy (Reply 41):
I ask because this guy seemed to be overcorrecting... especially in roll...if you look at the FDR trace oscillations.

What speed were they at ? What can you tell me about how effective the ailerons are at slow speed near a stall compared to normal S&L even in a small GA aircraft. This is basic flying 101 stuff.

Now look up a text book and tell let us know about where the airflow first tends to separate in a stall on large swept wing aircraft, and if this is a deliberate design feature.

Quoting airtechy (Reply 41):
Are you saying that this is what you would expect?

Yes of course. When you turn the FD on again after TCAS, EGPWS, stall recovery etc, it goes into the HDG V/S mode when turned back on, and the V/S will reflect what you are doing at the time. The aircraft was in a climb, so should the FD. They do the same when the disappear and come back on, in this youtube video of part of an A320 test flight you can see the FD bard disappear and reappear in HDG V/S. http://www.youtube.com/watch?v=5XcWd5WwoA0

Quoting airtechy (Reply 41):
I believe this was in reference to the fact that one of the checklists that was not run said to turn the FD off.....probably for this reason.

Turning the FD off is in a number of checklists, and is recommended whenever you are not planning to follow the FD orders.

Quoting airtechy (Reply 41):

I'll have to ask the Delta 330 crew on my next NRT-BKK flight if that is the case....plus a few other questions.

I wish the DL crew the best of luck and fortitude, they will need it.
We are addicted to our thoughts. We cannot change anything if we cannot change our thinking – Santosh Kalwar
 
User avatar
Starlionblue
Posts: 17097
Joined: Fri Feb 27, 2004 9:54 pm

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 9:05 am

Quoting rfields5421 (Reply 35):
This crew never ran the UAS checklist.

Yepp. There was a checklist that would have made the whole problem go away.

Actually what I like about the reporting is that even the tabloids I have checked out do not blame the scary Airbus controlled by computer gremlins.

Quoting zeke (Reply 36):
The crew had other options as well, including gear, spoilers, and flaps, I would not have given up.

I don't think they gave up. I think they had huge authority and communication issues on the flight deck and this prevented them from making good decisions. The whole discussion on the flight deck reminds me of two teenagers who have done something they knew they shouldn't and are under time pressure to fix the issue and not be discovered. It's as if the PF didn't want to admit he was out of his depth and the PNF was afraid to just tell him to stop fiddling.

Quoting zeke (Reply 36):
Why provide pilots with even more information to entice them to touch something, where the best course of action in the cruise is to do nothing until it has been confirmed what is going on.

I'm just a layman but this has always struck me as one of the central points since the recordings were released.. The plane was flying fine with lots of air under it. Why immediately fiddle? I can understand having to react immediately close to the ground, but it's not like it went from straight and level to plummeting towards the ground just because the A/P and A/T disconnected.

Quoting canoecarrier (Reply 37):
What if the video only focused on the cockpit instruments and didn't show the pilots themselves? I'd have to go back to some of the technical reports BEA put out related to triggered transmission of flight data, but I don't see a (technical or professional) problem (in an emergency) of using an automatic uplink of DFDR/DCVR data along with a visual image of what the pilots see in front of them. It took 2-3 years to find the crash site in this case, then recover the recording devises before we had a reasonable idea what happened that evening.

No point in just showing the instruments. You can get the raw data from the DFDR.

As for automatic uplink of all the DFDR data, it is technically feasible but economically prohibitive, especially over oceans where even normal radio comms are flaky. And an image of what was in front of the camera would have shown blackness and ocean. Unless you see some stars, the horizon, the moon, etc, not very helpful.

[Edited 2012-07-08 02:10:08]
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
rfields5421
Posts: 5565
Joined: Thu Jul 19, 2007 12:45 am

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 1:48 pm

Quoting Starlionblue (Reply 43):
I'm just a layman but this has always struck me as one of the central points since the recordings were released.. The plane was flying fine with lots of air under it. Why immediately fiddle? I

From reading the latest report I can see three reasons the PF may have felt he had to do something immediately.

1) His preoccupation with getting above the cloud tops - climbing to FL360

2) The aircraft started to roll - 8.4 degrees before he moved the stick

3) The pitch changed from 1.8 to 0 - initiating a 600 fpm descent

He's been worried about flying at FL350 which is apparently in the top of the cloud layer, and he wants to get above the clouds. In the previous 30 seconds the roll has been oscillating between 2.8 right and 4.8 left - the AP has been catching and correcting the roll. Suddenly the AP shuts off and the aircraft rolls right farther than the AP was allowing.

He might have recognized the start of the descent.

That could scream "do something" to someone surprised or startled.

Quoting Starlionblue (Reply 43):
You can get the raw data from the DFDR.

But we have NO RAW DATA from the DFDR in this accident about what was displayed on instruments in front of the Pilot Flying in the right hand seat.

Quoting Starlionblue (Reply 43):
As for automatic uplink of all the DFDR data, it is technically feasible but economically prohibitive, especially over oceans where even normal radio comms are flaky.

The automated up-link of DFDR data when an emergency is detected would not be that costly - but might be subject to technical limits. It wouldn't occur on most flights, very few flights actually. And only when 'an emergency' was detected by the systems. Defining 'an emergency' for the computers to start automated uplink would be the most difficult part of the process.

We now know that by 0530 that Air France operations was concerned about the lack of communications from the aircraft - the every 10 minutes position reports stopped after 0210. We also know that no later than 0537 - the maintenance staff at Air France operations had decoded the ACARS messages enough to be concerned about the safety of the flight.

But if the aircraft has started a DFDR emergency data transmission at 0210 or 0211 - that would have immediately raised a huge red flag at the Air France OCC.

It is very true that in this case it would have had no impact upon survival of anyone aboard the aircraft - it would have (1) narrowed the search location, and (2) started the SAR activation process much earlier.

All the communication up until 0815 was "I don't know anything, do you know anything? Does anyone know for certain who last was responsible for the flight?"

At 0815 the Madrid Duty Officer convinces his ARCC to issue an ALERFA-INCERFA message - the first official notice that an aircraft is definitely missing. Even then he has to overcome "Not my responsibility" agruments from his own staff.

Had a DFDR emergency data transmission began, and stopped in mid-transmission - the ALERFA, INCERFA, DETRESFA process would have started about SIX HOURS earlier.

Again, while it would have made no difference to the people on this plane, it might make the difference in the survival of another aircraft passengers/ crew in the future - in the unlikely event the plane were to go down in the ocean where someone could get off alive.

I'm not sure BEA is actually thinking such a transmission might save lives - that is more of a hope. But such a transmission would save a lot time and money on localizing a crash site.
 
Pihero
Posts: 4197
Joined: Mon Jan 31, 2005 5:11 am

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 2:24 pm

The case of the "not -executed unreliablre airspeed check-list " was bound to come up again...
... and again I will fight it.
Let's look at what happened in the cockpit :
- ASI info was lost
- A/P and ATHR were disconnected,
- ... leaving the crew to revert to manual flight.

Now what did the crew have at their disposal ?:
- First and very importantly, a card placed on the bottom left of the central ECAM panel, on which the PNF updat-es every hour or so the parameters for flight in turbulence : Flight Level, Mach / ASI , Engine required N1.
- Secondly, three altitude indications : PFDs, ISIS and IRS. These indications have to be sorted out of the clutter of the different flags, but they're quite exploitable.
- Thirdly, they had just decided to fly the turbulence speed, set up the ATHR, therefore all the parameters were right at or at least very very close to the required values for .8M / FL 350 / 205 T.

So, it was just a question of flying FL 350 on the altimeters, making sure the required N1s were set.

The above prrocedure is one of the reasons the other cited aircrews did not go into the Unreliable IAS check list..

Quoting zeke (Reply 40):

The regulatory training and checking schedule items are the same

  

Quoting airtechy (Reply 41):

The report seems to indicate that the FD was providing bad command bars intermittently because of the LOA...ie UP command.

There is no mention of"bad" or "good" in the report : the flight directors commands were only subject to the required command they'd provide under the required phase : here Heading and Vertical speed. The last picture of the PFD and the command bars on fig 69, page 102 depicts a very sad situation : They flew away right of the initial heading, with a very high rate of descent, both aspects the FD bars advised to correct with a greater pitch attitude !!!

Quoting tommytoyz (Reply 33):
Another question in my mind is why GPS speeds can't be used in a pinch such as this? Winds aloft would be know by the pilots, so within a few knots, the GPS calculated speed should help pilots avoid catastrophe. No?

Again this mistake that there is a direct correlation between aerodynamic speed and the GPS-born ground speed.
What is of importance is the calibrated airspeed or CAS very close to what we call indicated airspeed or IAS
. To illustrate this point, CAS = 272 kt for M= .8 while at the A/P disconnect point, the GPS value was 468 kt. Knowing the wind at ( roughly) 270° / 100 kt would still put you 200 kt adrift of the actual value of the air velocity over our wing !

[Edited 2012-07-08 07:43:55]
Contrail designer
 
User avatar
Starlionblue
Posts: 17097
Joined: Fri Feb 27, 2004 9:54 pm

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 2:28 pm

Quoting rfields5421 (Reply 45):
Quoting Starlionblue (Reply 43):
You can get the raw data from the DFDR.

But we have NO RAW DATA from the DFDR in this accident about what was displayed on instruments in front of the Pilot Flying in the right hand seat.

Fair point. However if the instruments screw up that badly don't we need to ground all 330/340s today?

Quoting rfields5421 (Reply 45):
Quoting Starlionblue (Reply 43):
As for automatic uplink of all the DFDR data, it is technically feasible but economically prohibitive, especially over oceans where even normal radio comms are flaky.

The automated up-link of DFDR data when an emergency is detected would not be that costly - but might be subject to technical limits. It wouldn't occur on most flights, very few flights actually. And only when 'an emergency' was detected by the systems. Defining 'an emergency' for the computers to start automated uplink would be the most difficult part of the process.

I think you answered your own question. How would you know what constitutes an emergency. And after all, the boxes were found in the end right?
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
rfields5421
Posts: 5565
Joined: Thu Jul 19, 2007 12:45 am

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 5:20 pm

On my phone headed to Akansas now. Can't quote

Re instruments / DFDR - BEA has recommend a change in recorder standards to require recording all instruments the pilots see. The presumption that recording the Captain's instruments only has clearly been inadequate. This is an industry wide issue for all manufacturers.

Re defining an emergency - BEA has some recommendations about what type events would trigger an automated DFDR upload. After AF447 a UAS would be one, but also things like fire alarms, engine shutdown, cabin pressure loss, etc. Any such list would of course be evolving over time.

I'm not sure all these changes could be incorporated into existing airframes. But one of the things I really like about how the BEA has handled this whole investigation is the focus on how things can be improved not only in the short term, but also long term. Many of the BEA recommendations will take years to study and determine if they are technically possible, much less economically possible. Some could only be incorporated in new design aircraft.
 
canoecarrier
Posts: 2569
Joined: Sat Feb 14, 2004 1:20 pm

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 5:53 pm

Quoting Starlionblue (Reply 46):
I think you answered your own question. How would you know what constitutes an emergency. And after all, the boxes were found in the end right?

Airbus and BEA already defined that in their technical documents related to AF 447. They developed equations that defined what an "unsafe event" was, including overspeed, excessive pitch, roll, etc. They then tested it on 9,333 long range flights from one airline and on 11 accidents flights from the BEA’s database.

It's on page 12 of this report if you're interested:
http://www.bea.aero/en/enquetes/flig...ed.transmission.of.flight.data.pdf

Quoting Starlionblue (Reply 43):
As for automatic uplink of all the DFDR data, it is technically feasible but economically prohibitive, especially over oceans where even normal radio comms are flaky.

BEA studied the cost of emergency transmittal of DFDR/DCVR data as well. On aircraft equipped with Satcom it reasonably cheep to implement. This is more Mandala's background though. He should be able to speak to the costs better than I.
The beatings will continue until morale improves
 
tdscanuck
Posts: 8572
Joined: Wed Jan 11, 2006 7:25 am

RE: BEA Recommendations - AF447

Sun Jul 08, 2012 7:01 pm

Quoting airtechy (Reply 21):
I believe the auto-trim cranked in max uptrim due to limits being removed in the ALT 2 LAW mode

It went to max uptrim because the average command from the PF was a sustained nose-up. In FBW pitch laws, the elevators do the fast responses and the trim does the long period response. If the commands are majority nose-up then the stab will trim nose up. As the airplane slowed down, more and more uptrim was needed to provide the PF with the flight trajectory he was asking for. They had full nose up and nose down authority all the time.

Quoting airtechy (Reply 21):
I think the report was light on criticism of Airbus regards the way information was being displayed to the pilots, the audible alarms adding to the confusion, the way the airplane handles in alternate law modes, and other things.

You can only assume that the crew will disobey their procedures and training to some reasonable degree; any sufficietly motivated or confused flight crew can crash an airplane regardless of the level of automation or protection.

Quoting tommytoyz (Reply 25):
Why, as the report says (page 182), regarding the flight director, that "...the flight director was advising a nose-up attitude...." Why would the flight director indicate that in a fully stalled condition at cruise altitude?

They were way below original altitude and in a descent...the FD was telling them to go back up because that's where their flight plan was and where the last point that the FD initialized itself was. That's exactly what any trained crew would expect the FD to do. That's why you don't do recovery maneuvers off the flight director.

Quoting tommytoyz (Reply 25):
Maybe the Flight Director was programmed with this philosophy, without it ever having been proven in certification flights for the A330?

The A330 FD is certified, just like all the other airliner FD's.

Quoting Mir (Reply 34):
Quoting tommytoyz (Reply 13):
I disagree, the AF447 crew did not receive any false stall warnings.

They absolutely did, right after the problems started occurring.

The stall *warning* isn't about telling you whether you're stalled or not, it's about telling you you're about to stall and need to take corrective action immediately. Every time it sounded it was correct; the crew needed to take immediate action to recover. They didn't.

Quoting rfields5421 (Reply 35):
Quoting tommytoyz (Reply 33):
Winds aloft would be know by the pilots

Actually no - pilots don't know the winds aloft.

The INS does, and it's pretty accurate at cruise. It takes a few seconds to pick up the wind but it's much better than half an hour.

Quoting rfields5421 (Reply 44):
Quoting Starlionblue (Reply 43):
You can get the raw data from the DFDR.

But we have NO RAW DATA from the DFDR in this accident about what was displayed on instruments in front of the Pilot Flying in the right hand seat

The point is that it's a lot easier to just record both sides on the DFDR as opposed to trying to videotape the instruments (same information, way more bandwidth required).

Quoting rfields5421 (Reply 44):
Quoting Starlionblue (Reply 43):
As for automatic uplink of all the DFDR data, it is technically feasible but economically prohibitive, especially over oceans where even normal radio comms are flaky.

The automated up-link of DFDR data when an emergency is detected would not be that costly

It's costly; you're talking about pushing a huge amount of data very fast. Even if you rarely use it, that's expensive technology.

Tom.
  • 1
  • 2
  • 3
  • 4
  • 5
  • 8

Who is online

Users browsing this forum: EIASO and 18 guests