AF 447 Thread no. 4
bad weather, et al
To Old Car. and others.......
There is no doubt that some weather started the sequence that resulted in many lives being lost. The stoopid air data sensors went into la la land. That started the sequence of events and human responses.
Anybody disagree?
So we have air data becoming unreliable, we have a system that keeps reverting to less and less "limits", we have a myriad of cautions and warnings for the crew to accept and act upon, and it is NIGHT!
I will gladly take any non-pilot system engineer or software engineer or..... and place them into the situation that the AF447 folks faced. What would they do?
Well, what's your point, Gums?
The point is to get from point "A" to "B", or drop bombs on the enema, or take recce pictures, or take some SLF's to see the Grand Canyon, or .....
BOTTOMLINE:
If we insist upon human crews in our commercial aircraft, then we have to design and implement systems that help them, but still demand some degree of basic airmanship and training.
I cannot judge the actions of the AF447 crew we see from the sparse reports from BEA, given the data we have now.
I can surely judge the sensors' failure and the control logic that causes warnings and cautions and does not take inton account a basic principle of aero - angle of attack.
I would hope we give the humans just a fighting chance to keep the plane flying when unexpected things happen, and we could point fingers at each other later.
I know that many here have flown thousands of hours and have delivered millions of SLF's. Makes me feel comfortable, to a point.
But how far are we gonna go with automation and systems that will "protect" you regardless of your basic airmanship and training?
respectfully.......... and maybe we need another forum/thread for these thots.
There is no doubt that some weather started the sequence that resulted in many lives being lost. The stoopid air data sensors went into la la land. That started the sequence of events and human responses.
Anybody disagree?
So we have air data becoming unreliable, we have a system that keeps reverting to less and less "limits", we have a myriad of cautions and warnings for the crew to accept and act upon, and it is NIGHT!
I will gladly take any non-pilot system engineer or software engineer or..... and place them into the situation that the AF447 folks faced. What would they do?
Well, what's your point, Gums?
The point is to get from point "A" to "B", or drop bombs on the enema, or take recce pictures, or take some SLF's to see the Grand Canyon, or .....
BOTTOMLINE:
If we insist upon human crews in our commercial aircraft, then we have to design and implement systems that help them, but still demand some degree of basic airmanship and training.
I cannot judge the actions of the AF447 crew we see from the sparse reports from BEA, given the data we have now.
I can surely judge the sensors' failure and the control logic that causes warnings and cautions and does not take inton account a basic principle of aero - angle of attack.
I would hope we give the humans just a fighting chance to keep the plane flying when unexpected things happen, and we could point fingers at each other later.
I know that many here have flown thousands of hours and have delivered millions of SLF's. Makes me feel comfortable, to a point.
But how far are we gonna go with automation and systems that will "protect" you regardless of your basic airmanship and training?
respectfully.......... and maybe we need another forum/thread for these thots.
Join Date: Jul 2002
Location: UK
Posts: 3,093
Likes: 0
Received 0 Likes
on
0 Posts
jcjeant : I have many sources of information, of which that was one (admittedly not a good one, but it stuck in my head). For example I don't have anything concrete on the Boeing 777 systems other than a rough outline, but rest assured, I'm not just a TV junkie getting my kicks from talking to pilots on here.
CogSim : The Colgan pilots pulled their aircraft straight into the ground - I'm not saying that's what happened here, but it does happen.
CogSim : The Colgan pilots pulled their aircraft straight into the ground - I'm not saying that's what happened here, but it does happen.
Join Date: Jul 2009
Location: The land of the Rising Sun
Posts: 187
Likes: 0
Received 0 Likes
on
0 Posts
Gums
I think you fail to understand the nature of the FBW system used on Airbus aircraft. I must admit I don't fully understand it but from what those who know have said it is more about placing limits on what you can do with a very stable aircraft. The comments on the various threads by those who fly the beast do not talk about an unstable monster kept in the air by computers but a stable platform which is easy to fly without the protections. As the aircraft encounters different situations the protections or limits drop away but the aircraft doesn't become suddenly unstable.
It may be that there is a problem with the training regime for these aircraft and that not enough attention is paid to flying without these protections but I am not qualified to comment about this and I haven't seen any comments from Air France pilots on this board about their training and whether there were deficiencies so I just don't know. It may also be that the flight crew were deficient in their knowledge but I would not like to make that judgment yet - I believe that it is far too premature to even form that conclusion. Certainly, the pilots actions need considerable more scrutiny but to go as you seem to for a theory that myriad system faults incapacitated the pilots is too reckless at the moment.
I think you fail to understand the nature of the FBW system used on Airbus aircraft. I must admit I don't fully understand it but from what those who know have said it is more about placing limits on what you can do with a very stable aircraft. The comments on the various threads by those who fly the beast do not talk about an unstable monster kept in the air by computers but a stable platform which is easy to fly without the protections. As the aircraft encounters different situations the protections or limits drop away but the aircraft doesn't become suddenly unstable.
It may be that there is a problem with the training regime for these aircraft and that not enough attention is paid to flying without these protections but I am not qualified to comment about this and I haven't seen any comments from Air France pilots on this board about their training and whether there were deficiencies so I just don't know. It may also be that the flight crew were deficient in their knowledge but I would not like to make that judgment yet - I believe that it is far too premature to even form that conclusion. Certainly, the pilots actions need considerable more scrutiny but to go as you seem to for a theory that myriad system faults incapacitated the pilots is too reckless at the moment.
Per Ardua ad Astraeus
Join Date: Mar 2000
Location: UK
Posts: 18,579
Likes: 0
Received 0 Likes
on
0 Posts
Far too much noise and assumption here to comment on it all, but
gums #436:
- please join the thread on the 'Safety' forum?
All this comment on 'deviation'/lack of is based on NO KNOWLEDGE whatsoever! Cells do move around, flight paths differ, mostly we only seek around 20 miles clearance in normal ops. We do not know what they saw on the radar nor what they discussed between themselves apart from going a 'bit left'.
I still remain surprised that the 2001 near accident to the 340 which entered an un-commanded zoom climb to a very low speed remains dismissed by many? No-one has yet explained (simply please) why the system there decided it needed to pull up so violently. That sure as hell would have confused me in the dark of the Atlantic in the ITCZ - and I would have mentioned it to the other pilot........
Edited for my mis-read of Vls.
gums #436:
But how far are we gonna go with automation and systems that will "protect" you regardless of your basic airmanship and training?
respectfully.......... and maybe we need another forum/thread for these thoughts.
respectfully.......... and maybe we need another forum/thread for these thoughts.
All this comment on 'deviation'/lack of is based on NO KNOWLEDGE whatsoever! Cells do move around, flight paths differ, mostly we only seek around 20 miles clearance in normal ops. We do not know what they saw on the radar nor what they discussed between themselves apart from going a 'bit left'.
I still remain surprised that the 2001 near accident to the 340 which entered an un-commanded zoom climb to a very low speed remains dismissed by many? No-one has yet explained (simply please) why the system there decided it needed to pull up so violently. That sure as hell would have confused me in the dark of the Atlantic in the ITCZ - and I would have mentioned it to the other pilot........
Edited for my mis-read of Vls.
Last edited by BOAC; 27th Jun 2011 at 10:10.
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by BOAC
... the 2001 near accident to the 340 which entered an un-commanded zoom climb to below stall speed ...
The indicated airspeed dropped below VLS (the lowest selectable) as the aircraft climbed and the commander took manual control
Join Date: Jul 2009
Location: UK
Posts: 134
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by BOAC
below stall speed
Are you sure?
In cruise, VLS is a speed that provides a certain margin (0,3 g?) to buffet onset. The zoom climb occurred because the FCS entered into alpha-prot law and then maintains AoA=alphaprot until the pilot moves the sidestick. Maintaining alphaprot stick-free prevents the airplane from stalling...
In cruise, VLS is a speed that provides a certain margin (0,3 g?) to buffet onset. The zoom climb occurred because the FCS entered into alpha-prot law and then maintains AoA=alphaprot until the pilot moves the sidestick. Maintaining alphaprot stick-free prevents the airplane from stalling...
Last edited by sensor_validation; 27th Jun 2011 at 13:14. Reason: Fix quote: apologies to HN39
Per Ardua ad Astraeus
Join Date: Mar 2000
Location: UK
Posts: 18,579
Likes: 0
Received 0 Likes
on
0 Posts
No - you are right, HN, my mistake and I have edited my post - not stall speed, but pretty slow whatever you call it.
- please explain the logic here - the a/c was now here near 'stalling' when it all started and was then placed near it by the software??? So why? This defeats me. Suppose for a moment that this happened to 447. Desirable - no. Helpful - no. On a dark night in the ITCZ, confusing - yes. We know that the alpha-prot 'inhibit' would probably have been disabled by the IAS readings.
In cruise, VLS is a speed that provides a certain margin (0,3 g?) to buffet onset. The zoom climb occurred because the FCS entered into alpha-prot law and then maintains AoA=alphaprot until the pilot moves the sidestick. Alphaprot is less than alphamax and maintaining it stick-free prevents the airplane from stalling.
Join Date: Oct 2009
Location: UK
Posts: 1,270
Likes: 0
Received 0 Likes
on
0 Posts
Hi DozyWannabe,
True, but that is not in any QRH / FCOM procedure. Additionally, in all laws (except Direct Law) the elevators will attempt to maintain the last requested pitch attitude (1g value) even when stick free. So there would still be no "feel" in pitch that the aircraft was slowing down. Deprived of reliable airspeed indications and pitch feel, I believe the crew would have had a more difficult task to avoid a stall than if they had been in a conventional aircraft.
That said, if the pilot doesn't like what autotrim is doing all he or she has to do is keep their hand on that trim wheel.
Last edited by rudderrudderrat; 27th Jun 2011 at 11:51.
Join Date: Nov 2010
Location: FR
Posts: 477
Likes: 0
Received 0 Likes
on
0 Posts
Hi wallybird7
Never did I say that one can recover an out of control plane with trim alone.
And I do agree that the (auto) trim seems weird, based on the may 27th note : why did it go (almost) full NU, but didn't go less NU (or ND) when the crew pushed the sidestick ? Many answers come to mind, I choose to wait for more data on that.
That said, and it's all about semantics :
- the alpha prot/alpha floor protection in normal law does restrict the crew's control
- the bank angle protection (67° max) in normal law does restrict the crew's control
- the overspeed protection (inducing a NU command) does restrict the crew's control
- OTOH, the Auto Trim doesn't restrict the crew's control (until proven there was a mechanical and/or logical/software failure). It is designed AFAIK to follow the crew's commands (and to maintain 1g without crew's command).
And as DW stated : if the pilot doesn't like what autotrim is doing all he or she has to do is keep their hand on that trim wheel.
That's the difference I make between restricted and not restricted. And it certainly doesn't mean that I try to say the plane is "perfect" and the blame must "go to the crew". That's not what I think
And I do agree that the (auto) trim seems weird, based on the may 27th note : why did it go (almost) full NU, but didn't go less NU (or ND) when the crew pushed the sidestick ? Many answers come to mind, I choose to wait for more data on that.
That said, and it's all about semantics :
- the alpha prot/alpha floor protection in normal law does restrict the crew's control
- the bank angle protection (67° max) in normal law does restrict the crew's control
- the overspeed protection (inducing a NU command) does restrict the crew's control
- OTOH, the Auto Trim doesn't restrict the crew's control (until proven there was a mechanical and/or logical/software failure). It is designed AFAIK to follow the crew's commands (and to maintain 1g without crew's command).
And as DW stated : if the pilot doesn't like what autotrim is doing all he or she has to do is keep their hand on that trim wheel.
That's the difference I make between restricted and not restricted. And it certainly doesn't mean that I try to say the plane is "perfect" and the blame must "go to the crew". That's not what I think
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by BOAC
... - please explain the logic here - the a/c was now here near 'stalling' when it all started ...
Why are we not steering our cars with a knurl?
Originally Posted by DozyWannabe
The people that designed these systems weren't stupid, and I'm sure there are very logical reasons for the system being designed the way it is. Remember that this was designed to be the next generation of flight controls - doing things a certain way because that was the way they'd been done since WW2 (or even beforehand!) wasn't a valid design input. Being as friendly and logical to the pilot as possible certainly was, the only problem with that being that some pilots prefer different things.
It is my opinion that, after going FBW was decided, the governing specification was to extract the highest possible performance and reliability benefits off the new technology, a respectable challenge for designers, I must say. You can easily trace the major changes that took place and see where the focus was.
In the team that overview specs, the engineer/pilot balance favored the former. Of course, the engineers made a very good job. They reached their goals. Then, they managed to make the design acceptable by certification authorities and by line pilots, a not-so-easy task.
The Airbus FBW provides an elaborate high AoA protection in normal law. Is it just because it was not required by regulations that unless you trip the auto-trim it does not even provide classical low speed longitudinal stability in ALT law?
Last edited by Jetdriver; 27th Jun 2011 at 14:30.
Join Date: Aug 2009
Location: Germany
Age: 67
Posts: 1,777
Likes: 0
Received 0 Likes
on
0 Posts
Hi,
BOAC
And I still remain surprised that the AF445 event remain dismissed by many
Even the BEA was forced to remain dismissed
Incident: Air France A332 over Atlantic on Nov 30th 2009, Mayday call due to severe turbulence
BOAC
I still remain surprised that the 2001 near accident to the 340 which entered an un-commanded zoom climb to a very low speed remains dismissed by many?
Even the BEA was forced to remain dismissed
However, investigators have recently received a new report, more substantial than the first. "He has no objective value, says one close to the investigation. The company has just said it was a non-event and that the crew may have overreacted. " The side of the BEA, the investigators' regret not having the AF445 flight data that could shed new light on what occurred on Flight 447. "
On Dec 17th the French BEA reported, that according to the air safety report (ASR) filed by the captain immediately after arrival in Paris the airplane was at FL380 about 60nm ahead of DEKON on airway UN866, when weather forced the crew to divert from the airway and descend to FL360 employing oceanic contingency procedures after being unable to obtain clearance from ATC. Air France however did not forward the ASR to the BEA within 72 hours as required by law. Instead, the airplane was dispatched again without downloading the flight data recorder, according to the airline the airplane flew to Bangalore (India) and back (flight AF-192 and AF-191), after which both cockpit voice recorder and flight data recorder were overwritten. The quick access recorders had a formatting error and were not useable according to the airline.
Per Ardua ad Astraeus
Join Date: Mar 2000
Location: UK
Posts: 18,579
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by HN39
Turbulence caused a momentary excursion of "phase advanced" AoA beyond alpha-prot, and that put the FCS into alphaprot law (High AoA protection). That protection is lost in Alternate Law 2. I believe I have read on this thread that this particular logic has been changed as a result of the A340 AirProx incident.
JC - on the face of it, 445 looks quite different. The primary need for investigation is in the handling by AF of the report and data, I think.
Join Date: Jul 2002
Location: UK
Posts: 3,093
Likes: 0
Received 0 Likes
on
0 Posts
I still believe Airbus missed a trick by not instituting a pilot outreach programme at the same time it was courting airlines with the economy and safety aspects. All they had to do was say to pilots "Look, our aircraft fly like spacecraft - want to know how Neil Armstrong felt in the Eagle?", and I suspect it may have been looked on a lot more favourably. Much is made by Airbus detractors of the involvement of Bernard Ziegler in the early days of the Airbus FBW programme, which is fair enough - he's a controversial individual who said some less-than-clever things. However not so much is made of the presence of Gordon Corps at the same time - I don't think he could be described as anything other than a "pilot's pilot", and he was *very* comfortable with the design. There's a thread kicking round here somewhere from someone who was suspicious of the A320's ability to get out of extreme situations and Captain Corps took him up on the challenge - in every case the A320 in Normal Law fared better than the conventional aircraft, albeit in the simulator.
When I first read HTBJ, I was struck by D.P. Davies mentioning that some in the piloting fraternity kicked up an almighty stink about the presence of the "stick pusher" in CAA-certified aircraft - the arguments used then were much the same as those made against the Airbus FBW system today - claims of "encroachment on a pilot's authority" etc. The "stick pusher", as many will know, was a simple hydraulic ram system designed to force an aircraft which did not have good stall characteristics (particularly rear-engined T-tail designs, but a "stick nudger" was fitted to all G-registered B707s as well) into a nose-down attitude, and yet pilots of the time still claimed that it was a technological step too far. In other words, the bunfight that always occurs when the subject of automation comes up is nothing new.
The truth is that while Airbus marketed their FBW designs as a quantum leap forward, in fact it was very much an evolutionary rather than revolutionary step in terms of aeronautics - the only thing radical about it was that it was the first time it had been applied to an airliner. Boeing knew this too, and that's why the B777, when it arrived, was basically a FBW airliner with a computer-controlled force-feedback system, kind of like how a nicotine inhaler is compared to a cigarette - the old feelings are all there, but it's artificial - under the hood it's a very similar computer-driven system to that of the FBW Airbii.
Is it just because it was not required by regulations that unless you trip the auto-trim it does not even provide classical low speed longitudinal stability in ALT law?
Last edited by DozyWannabe; 27th Jun 2011 at 14:26.
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by DJ77
The way we flew airplanes before FBW had been refined over decades and was satisfying almost everyone. Why change? Why did Airbus needed to invent a new way of flying airplanes? Until now, I fail to be convinced the reason was for the pilot’s better.
I believe that envelope protection enhances safety, even if it turns out to have features that can be improved in the light of service experience.
The question of when auto-trim should be on or off is not limited to FBW in my view, and I must admit that I don't understand the drama that some make of it. As I view it, when it operates, autotrim obeys the pitch-orders from sidestick or AP by following the elevator. If the order is nose-up, THS trims nose-up. If the order is nose-down, THS trims nose-down. At FL350 there was plenty of time for that to take effect. Maybe I got that all wrong.
Join Date: Jun 2009
Location: somewhere
Posts: 451
Likes: 0
Received 0 Likes
on
0 Posts
HN39:
A33Zab;
Your "Chronology of events @ 5 sec interval." has a column headed "STALL AoA". I presume that is 'alpha-max', which is not stall AoA but somewhat less. Alphamax varies with Mach number as shown on a graph posted a few days ago. For the Mach numbers in your table:
Mach ... alphamax
0.8 ......... 5.2
0.68 ........ 7.1
0.64 ........ 7.7
0.59 ........ 8.1
Your "Chronology of events @ 5 sec interval." has a column headed "STALL AoA". I presume that is 'alpha-max', which is not stall AoA but somewhat less. Alphamax varies with Mach number as shown on a graph posted a few days ago. For the Mach numbers in your table:
Mach ... alphamax
0.8 ......... 5.2
0.68 ........ 7.1
0.64 ........ 7.7
0.59 ........ 8.1
Mach threshold ... AOAsw (I did edit this in the chronlogy diagram)
0.82 .................... 3.8
0.75 .................... 5.2
0.53 .................... 7.6
0.35 .................... 9.9
<=0.28............... 10.8
BEA Report 2:
In alternate or direct law, the angle-of-attack protections are no longer available
but a stall warning is triggered when the greatest of the valid angle-of-attack values exceeds a certain threshold.
In clean configuration, this threshold depends,in particular, on the Mach
value in such a way that it decreases when the Mach increases.
It is the highest of the valid Mach values that is used to determine the threshold.
If none of the three Mach values is valid, a Mach value close to zero is
used.but a stall warning is triggered when the greatest of the valid angle-of-attack values exceeds a certain threshold.
In clean configuration, this threshold depends,in particular, on the Mach
value in such a way that it decreases when the Mach increases.
It is the highest of the valid Mach values that is used to determine the threshold.
If none of the three Mach values is valid, a Mach value close to zero is
For example, it is of the order of 10° at Mach 0.3 and of 4° at Mach 0.8.
I've added your values in a separate Alpha max column.
As I view it, when it operates, autotrim obeys the pitch-orders from sidestick or AP by following the elevator. If the order is nose-up, THS trims nose-up. If the order is nose-down, THS trims nose-down.
Under many conditions, there are times in both Alternate and Normal laws when the auto-trim will briefly operate opposite the pilot SS input when the SS input causes the aircraft to experience "G" noticeably above or below 1 "G" as displayed in amber on the lower display.
The same thing occurs in terms of elevator surfaces movement (SD F/CTL page) if the manual trim wheel is rolled briskly in one direction or the other with hands off the SS, i.e. the elevator will briefly deflect opposite during the manual trim wheel input.
OK, what you describe strikes me as a design "feature" that is aimed at mitigating over control, or enhancing a stability function.
But I may misunderstand you.
Is the "opposite" input driven after the initial response following SS input, or it is made if the SS input exceeds a certain rate?
But I may misunderstand you.
Is the "opposite" input driven after the initial response following SS input, or it is made if the SS input exceeds a certain rate?
LW 50:
My impression is that it's rate driven. It is most noticeable with more aggressive SS inputs creating "G" excursions, and does not occur while making the slower smooth SS inputs associated with normal flying technique.
Only my impression though, I don't know what all the computers are "discussing" at this point. I do know it has no impact on aircraft control and doesn't bother me.
My impression is that it's rate driven. It is most noticeable with more aggressive SS inputs creating "G" excursions, and does not occur while making the slower smooth SS inputs associated with normal flying technique.
Only my impression though, I don't know what all the computers are "discussing" at this point. I do know it has no impact on aircraft control and doesn't bother me.