Originally posted by pegasus
View Post
Announcement
Collapse
No announcement yet.
Air France Off the Hook on AF447
Collapse
X
-
Originally posted by Evan View PostNo. I mean they are unreliable and will give misleading indications, specifically, that you are now in a descent and must command some load factor to regain your previous flight level. This will result in increased pitch, a departure from level flight and a speed reduction. This is where it begins to go wrong.
What you want is to maintain your current flight path with as little deviation as possible. That requires you to focus on the artificial horizon, not the altimeter, not the VSI.
If he attempted to correct a 300 ft deviation with such a zoom climb, he lacked basic airmanship
Now, if he instead attempted to climb to 37000 ft with that zoom climb, he lacked basic airmanship.
(when I say "lacked" I mean at that specific moment, which can be the result of reacting under terror, shutting down the rational mind, or just lacking that airmanship in the first place)
--- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
--- Defend what you say with arguments, not by imposing your credentials ---
Comment
-
Originally posted by Gabriel View PostAccording to yourself, he was going to pull up anyway because he had decided to climb to 37000 ft.
If he attempted to correct a 300 ft deviation with such a zoom climb, he lacked basic airmanship
Now, if he instead attempted to climb to 37000 ft with that zoom climb, he lacked basic airmanship.
(when I say "lacked" I mean at that specific moment, which can be the result of reacting under terror, shutting down the rational mind, or just lacking that airmanship in the first place)
If he had it installed in his mind that, when the auropilot suddenly disconnects during cruise, and a quick crosscheck of airspeeds shows a disagree, to maintain level flight by pitch attitude and only pitch attitude and to do nothing else (aside from the FD and thrust lock issues) for 1-2 minutes, this wouldn't have happened. Without such a memorised procedure, with only 'airmanship', that leaves the door wide open for stealth, deception and poor judgment, and, as you well know, there are pilots up there vulnerable to deception with potentially poor judgement under fire.
Gabriel. do you disagree with either of these statements:
1) The altimeters are dependent on airspeed data and are thus compromised by a pitot failure. The error is on the order of approximately 300ft. The vertical speed indicator is also erronous. The only 'good' basic flight instrument on the PFD is the artificial horizon. That is why the procedure must stress the need to fly by pitch, not by any of the other instruments.
2) During the critical first minute, the stick was never pulled back to the stop. That is a myth that perpetuates on internet forums, including this one. This only occurred after the plane was fully stalled. Also, the stick inputs were very active, ranging from backward (nose up) to forward (nose down). Watching the stick movements would not have told the story. Watching the artificial horizon would have however.
Comment
-
Originally posted by Evan View PostGabriel. do you disagree with either of these statements:
--- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
--- Defend what you say with arguments, not by imposing your credentials ---
Comment
-
Let me clarify a bit. Your statements are quite black and white, and capricious at some points.
The only "perfect" instruments in the PFD were the AI. The altimeters and vertical speed indicators were "good enough". And you can't miss 1.5G. Oh, and the AIs were perfect too, right?. Not doing anything crazy when the AP disconnects is basic airmanship. Even if you want to climb or something, first stabilize and get your wits. There was more than enough information on those 3 PFD to get feedback and not do what the pilot did. The FDR plot as I posted it in my previous post is a testament of that.
During the 1st minute? What's magical in a minute? You want to only focus in what happened before the stall because, you claim, that by then the situational awareness that we could not expect that the pilot would pull aggressively up when the stall warning sounded shouting "stall" in plain English. Then I offer you the first 20 seconds where the pilot actively entered that zoom climb and let it happen (because the story could have been that he unintentionally pulled up too much and when he saw the pitch, altitude and vertical speed going up as crazy he corrected, but that's not what happened). Please integrate the stick input during the first 20 seconds. How do you explain that? How do you explain that he keeps a mostely nose-up input for 20 seconds while the parameters are showing a climb that is off the charts? So I propose that the situational awareness was out of the window already by 20 seconds. Or by 2 seconds. Sure, he did timidly let the nose go down after the PM told him to do so. But by when he stabilized the energy situation had horribly deteriorated that even in a mild descent in slightly less than 1G the stall warning AoA was exceeded (by the way, the plane stopped climbing but the pitch attitude was still excessive, ESPECIALLY if the plane is not climbing). They were behind the power curve by then and the only way out was to trade potential energy for kinetic energy, i.e. trade altitude for speed. It was an unsustainable climb (even if you say it wasn't) and the only way to recover from such is to undo the climb. It would have taken quite e bit of situational awareness to void the stall warning at that point, it would have NOT taken barely any situational awareness to react to the stall warning by pushing the nose down. But I want to focus in the first few seconds again. And I said this a gazillions of times. I don't trust this pilot to apply any memory item or procedure under stress. The way he reacted was irrational. And he reacted irrationally since the beginning of the event. His situational awareness did not deteriorate during the event. It was never there since the AP disconnected. Why? I don't know. It could be that he just didn't have the tools in his toolbox. Or it may be that the tools feel out the toolbox the minute that the fit hit the shan.
Why I didn't say all this in my previous reply? Because I've already said it many times. That's why I've said a few days ago, just like you did now, "I give up". But I am a fool.
--- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
--- Defend what you say with arguments, not by imposing your credentials ---
Comment
-
Originally posted by Gabriel View PostLet me clarify a bit. Your statements are quite black and white, and capricious at some points.
The only "perfect" instruments in the PFD were the AI. The altimeters and vertical speed indicators were "good enough". And you can't miss 1.5G. Oh, and the AIs were perfect too, right?. Not doing anything crazy when the AP disconnects is basic airmanship. Even if you want to climb or something, first stabilize and get your wits. There was more than enough information on those 3 PFD to get feedback and not do what the pilot did. The FDR plot as I posted it in my previous post is a testament of that.
During the 1st minute? What's magical in a minute? You want to only focus in what happened before the stall because, you claim, that by then the situational awareness that we could not expect that the pilot would pull aggressively up when the stall warning sounded shouting "stall" in plain English. Then I offer you the first 20 seconds where the pilot actively entered that zoom climb and let it happen (because the story could have been that he unintentionally pulled up too much and when he saw the pitch, altitude and vertical speed going up as crazy he corrected, but that's not what happened). Please integrate the stick input during the first 20 seconds. How do you explain that? How do you explain that he keeps a mostely nose-up input for 20 seconds while the parameters are showing a climb that is off the charts? So I propose that the situational awareness was out of the window already by 20 seconds. Or by 2 seconds. Sure, he did timidly let the nose go down after the PM told him to do so. But by when he stabilized the energy situation had horribly deteriorated that even in a mild descent in slightly less than 1G the stall warning AoA was exceeded (by the way, the plane stopped climbing but the pitch attitude was still excessive, ESPECIALLY if the plane is not climbing). They were behind the power curve by then and the only way out was to trade potential energy for kinetic energy, i.e. trade altitude for speed. It was an unsustainable climb (even if you say it wasn't) and the only way to recover from such is to undo the climb. It would have taken quite e bit of situational awareness to void the stall warning at that point, it would have NOT taken barely any situational awareness to react to the stall warning by pushing the nose down. But I want to focus in the first few seconds again. And I said this a gazillions of times. I don't trust this pilot to apply any memory item or procedure under stress. The way he reacted was irrational. And he reacted irrationally since the beginning of the event. His situational awareness did not deteriorate during the event. It was never there since the AP disconnected. Why? I don't know. It could be that he just didn't have the tools in his toolbox. Or it may be that the tools feel out the toolbox the minute that the fit hit the shan.
Why I didn't say all this in my previous reply? Because I've already said it many times. That's why I've said a few days ago, just like you did now, "I give up". But I am a fool.
So, one the first purpose, I'm stated it as clearly as I can. Yes, we all know that he zoom climbed. I think that was his intention. The climb was likely exaggerated by unpracticed skills and/or panic but WAS sustainable. He succeeded. I see this as the opposite of a rapid descent instinct during a decompression event. The instinctive need to get to a safer flight level. It was the wrong thing to do. It was poor airmanship. You don't have to keep making that point. I've never disputed that. But perhaps it can be explained...
On the second purpose, you say the altimeters and VSI's were "good enough". Good enough for what? To maintain flight level without slowing or accelerating? No. Not if you fly by them. They will compel you to increase your pitch and thus decrease your speed. If your hand isn't practiced at this flight level (and most aren't) you might over-control. Will that result in a 7000fpm climb to a higher flight level? No. That would require a different intent. Will it cause you to lose your previous 'known' place in the speed envelope with no airspeed reference. Yes.
Seriously Gabriel, of all the pilots interviewed on their UAS incidents prior to AF447, how many got out of thrust lock? How many switched off the FD's? How many maintained level flight and their 'known' place in the speed envelope? Perhaps a procedure is needed here...
Comment
-
Originally posted by Evan View PostGabriel, of all the pilots interviewed on their UAS incidents prior to AF447, how many got out of thrust lock? How many switched off the FD's? How many maintained level flight and their 'known' place in the speed envelope? Perhaps a procedure is needed here...
Can you maybe give me a simple black or white, yes or no?
I'm with you on that much.
Where I lose you (and maybe join Gabe), is that super mega uber ultimate basic airmanship is that the overly aggressive climb should have PROMPTLY been adjusted to a FDnH climb with more than adequate (albeit imperfect) instrumentation.
The attitude indicator, the altimeter, the DGINS were extremely adequate to discern an aggressive-great-way-to-stall climb from a FDnH climb.
Just can't give your "thought" much credence past 5 seconds unless it was total freak out loss of ALL RATIONAL THOUGHT for pretty much the final couple of minutes of flight.Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.
Comment
-
Originally posted by 3WE View PostEvan: Do I hear your right?...Bonin wanted to do a reasonable climb, but since he was out of practice, he mistakenly initiated a climb that was a bit to aggressive???
Can you maybe give me a simple black or white, yes or no?
Where I lose you (and maybe join Gabe), is that super mega uber ultimate basic airmanship is that the overly aggressive climb should have PROMPTLY been adjusted to a FDnH climb with more than adequate (albeit imperfect) instrumentation.
The attitude indicator, the altimeter, the DGINS were extremely adequate to discern an aggressive-great-way-to-stall climb from a FDnH climb.
Just can't give your "thought" much credence past 5 seconds unless it was total freak out loss of ALL RATIONAL THOUGHT for pretty much the final couple of minutes of flight.
The "total freakout" seems to occur about a minute after the loss of autopilot. This, to me, is more mysterious. He had climbed to 37,500. He had agreed to "go back down". He had reduced his pitch commands and pitch had reduced back to around 6° ANU. He had 216kts airspeed and AoA was around 5°. Then, for some reason, he pulled back up into a 3000fpm climb, reaching almost 18° ANU pitch. His inputs were still never full stop and many were reductions including three forward pushes well past neutral, so it doesn't seem to be a Renslow-type freakout. My only guess is that he was following the FD's which did suddenly reappear at that time and were giving instructions similar to his stick inputs. Had they followed a procedure and switched them off...............
Comment
-
Originally posted by Evan View PostSo, one the first purpose, I'm stated it as clearly as I can. Yes, we all know that he zoom climbed. I think that was his intention.
The climb was likely exaggerated
by unpracticed skills and/or panic but WAS sustainable.
He succeeded.
The instinctive need to get to a safer flight level. It was the wrong thing to do. It was poor airmanship. You don't have to keep making that point. I've never disputed that. But perhaps it can be explained...
On the second purpose, you say the altimeters and VSI's were "good enough". Good enough for what?
At that time, all 3 altimeters were indicating 35350ft and winding up quickly, all 2 vertical speed indicators were showing a climb of 5500 ft per minutes and increasing, and all 3 attitude indicators were showing 10 degrees nose up. These numbers are already ridiculous for a cruise climb.
And at that moment, at 2:10:19, all 3 altimeters, all 2 vertical speed indicators, and all 3 attitude indicators were PERFECTLY ADEQUATE for the pilot to know better than to keep pulling up.
What does the pilot do next? He keeps making ONLY NOSE UP inputs for the next 5 seconds, bringing the vertical speed to 7000 fpm and the pitch to 12 degrees nose-up.
That was not an exaggerated control input because he was not skillful at high altitude manual flight. His inputs were OPPOSITE to what NEEDED to be done at that time.
Why, in your theory, did the pilot do that? What was he doing while actively commanding a zoom climb? What were his eyes looking at? What was he thinking? Was he thinking? Was he making DECISIONS? RATIONAL decisions? I very much doubt it.
Seriously Gabriel, of all the pilots interviewed on their UAS incidents prior to AF447, how many got out of thrust lock? How many switched off the FD's?
How many maintained level flight and their 'known' place in the speed envelope?
The BEA has studied thirteen unreliable indicated airspeed events in the A330/A340.
With regard to the crews’ reactions, the following points are notable:
The aeroplanes remained within the flight envelope during these relatively
short events.
The variations in altitude were contained within about one thousand feet. There
were five cases of deliberate descent, including one of 3,500 feet. These descents
followed a stall warning.Nine cases of the activation of the stall warning were noted.
All the warning activations can be explained by the fact that the aircraft was in
alternate law at cruise Mach and in turbulent zones. Only
one warning triggering event was caused by a distinct input on the controls.
None of them crashed.
Perhaps a procedure is needed here...
Now, would this crew have managed this event correctly had there been a procedure in place?
I have zero confidence in that. A pilot that, upon AP disconnect, pulls up a 1.5G, 2500 ft, 7000 fpm, 12 degrees nose-up climb, I don't trust him to be able to follow a procedure that involves keeping a given pitch. He was incapable to keep ANY pitch in this incident.
--- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
--- Defend what you say with arguments, not by imposing your credentials ---
Comment
-
Originally posted by Gabriel View PostAnd I say it wasn't. I don't know what is sustainable for you, but I will not call sustainable a climb from which the only way out is a descent because if you don;t descend you will stall and then descend.
2:10:19. It's been 14 seconds since the AP disconnected and 12 seconds since he started playing with the sidestick. The inputs were ONLY nose up and up to 3/4 travel for the first 9 seconds, but for the last 3 seconds his inputs, while still clearly nose-up in average, were slightly more balanced with some instants of very slight nose-down inputs.
It is not that he climbed but how he climbed what removes any confidence in me that a better procedure would have saved the day with this pilots at the controls.
At that time, all 3 altimeters were indicating 35350ft and winding up quickly, all 2 vertical speed indicators were showing a climb of 5500 ft per minutes and increasing, and all 3 attitude indicators were showing 10 degrees nose up. These numbers are already ridiculous for a cruise climb.
And at that moment, at 2:10:19, all 3 altimeters, all 2 vertical speed indicators, and all 3 attitude indicators were PERFECTLY ADEQUATE for the pilot to know better than to keep pulling up.
What does the pilot do next? He keeps making ONLY NOSE UP inputs for the next 5 seconds, bringing the vertical speed to 7000 fpm and the pitch to 12 degrees nose-up.
I'm theorizing (not declaring) that the expedited climb was intentional, though probably more excessive than intended due to unpracticed skills. The only role the instrument inaccuracies play is in encouraging a departure from level flight in the first place.
None. I wonder why. After all the procedure did say AP/AT/FD off, thrust levers move. So it surely wasn't the lack of a procedure.
I have zero confidence in that. A pilot that, upon AP disconnect, pulls up a 1.5G, 2500 ft, 7000 fpm, 12 degrees nose-up climb, I don't trust him to be able to follow a procedure that involves keeping a given pitch. He was incapable to keep ANY pitch in this incident.
Comment
-
Originally posted by Evan View Post2hr 10min 49: The aircraft has successfully climbed to 37,500. It has not stalled. The pitch is 6 deg nose up. The vertical speed is 1100fpm. The AoA is around 5 deg (just below stall warning activation). If, at that point, autopilot had been restored for that flight level (and it may have been available at that point), the pitch would have reduced and the aircraft would not have stalled. It would have simply continued on its merry way at a higher flight level.
They were doing 215 knots
In a fully loaded A330
Yes, the vertical speed was 1100 fpm, but it was "decreasing"
Which means that they were at less that 1G (and hence lift was less than weight)
And the stall warning activated 2 frigging seconds after that.
The moment that they increased the load factor to barely 1G.
How can the word "sustainable" even cross your mind in relation to the condition at 2:10:49?
What are you looking at?! The input at 2:10:19 was nose-down to neutral. The inputs over the following five seconds were all slightly nose up (none exceeding 1/4 nose up)
Gabriel, are you kidding me?! It absolutely is the fact that he climbed! You don't climb out of cruise level flight without airspeed data! You maintain a safe pitch and power setting! (that little dot is an exclamation point).
I will show you how lack of manual flight skills at altitude look like, WHEN YOU HAVE A GOAL IN MIND (and if you say that the goal was just CLIMB, then I will say that the pilot lacked -at that moment at least- any reasonable airmanship)
From the BEA:
Case of TAM flight on 12 November 2003
This case, which happened to an A330-200, was not one of the thirteen events studiedabove because no crew report was available. However, in the light of the data fromflight AF 447, it seems useful to mention it. In fact, following icing of at least twoPitot probes at FL360, the crew made some high amplitude flight control inputs (tothe stop), sometimes simultaneously. When the AP disengaged, both pilots madepitch-up inputs (one went to the stop) that resulted in an increase in pitch of 8°. Onseveral occasions, the stall warning was triggered due to the nose-up inputs, and thecrew reacted with strong pitch-down inputs. During the 4 minutes that the sequencelasted, the load factor varied between 1.96 g and -0.26 g, the pitch attitude reached13° nose-up and the angle of attack reached 10°. Altitude variations, however, wereless than 600 ft.
I'm theorizing (not declaring) that the expedited climb was intentional, though probably more excessive than intended due to unpracticed skills.
- He was in a mental state that he reacted irrationally, or
- He lacked any minimum reasonable level of airmanship (meaning that he didn't understand AoA, stalls, and performance compromises at high altitude)
And in any of these 2 situations, the moment that the AP disconnected, he lacked said airmanship. Either because he didn't bring it with him or because he lost it in that instance due to the mental state.
It was a lack of proficiency on procedure. Memory procedures are meaningless if not committed to memory and practiced recurrently. Such procedures defend against erroneous improvisation.
And I believe it was the lack of a simple, solid procedure that led him to improvise in this way. At that critical moment of surprise, his mind asked 'what do I do?' and, not finding an answer there, made one up based on prior concerns. It needed to find the correct answer there, which most likely would have been there if he had memory-procedure training for this.
I support your campaign for a better procedure, but I am not buying that the existence of the procedure, by itself, would have avoided AF447 (in the same way that the existence of a stall procedure didn't avoid that they stalled all the way from cruise altitude to the ocean)
--- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
--- Defend what you say with arguments, not by imposing your credentials ---
Comment
-
Originally posted by Gabriel View PostHello? 2:10:49????
They were doing 121 knots
In a fully loaded A330
Yes, the vertical speed was 1100 fpm, but it was "decreasing"
Which means that they were at less that 1G (and hence lift was less that weight)
And the stalled warning activated 2 frigging seconds after that.
The moment that they increased the load factor to barely 1G.
You are looking at the invalid ISIS speed however:
Originally posted by BEA ReportAt around 2 h 10 min 34, the speed displayed on the left side became valid again and was then 215 kt; the speed on the ISIS was still incorrect.
But if you do climb, for whatever reason, you don't make a sustained unsustainable climb of 12 deg nose up and 7000 fpm. You have a goal in mind and my over-react, overshoot and oscillate towards that goal.
Originally posted by BEA ReportThe “Unreliable IAS” emergency manoeuvre requires as a first step to disconnect the automatic flight controls and disengage the Flight Directors. The two copilots had only been trained for the emergency manoeuvre at lower levels, in the course of which the pitch attitude to adopt is 10° or 15°.
At 1 h 52, the PF said to the captain “look, the REC MAX has changed to three seven five”
Comment
-
Originally posted by Gabriel View PostFrom the BEA:
Case of TAM flight on 12 November 2003
This case, which happened to an A330-200, was not one of the thirteen events studiedabove because no crew report was available. However, in the light of the data fromflight AF 447, it seems useful to mention it. In fact, following icing of at least twoPitot probes at FL360, the crew made some high amplitude flight control inputs (tothe stop), sometimes simultaneously. When the AP disengaged, both pilots madepitch-up inputs (one went to the stop) that resulted in an increase in pitch of 8°. Onseveral occasions, the stall warning was triggered due to the nose-up inputs, and thecrew reacted with strong pitch-down inputs. During the 4 minutes that the sequencelasted, the load factor varied between 1.96 g and -0.26 g, the pitch attitude reached13° nose-up and the angle of attack reached 10°. Altitude variations, however, wereless than 600 ft.
Comment
-
Originally posted by Evan View PostYou are looking at the invalid ISIS speed however:
Regarding my theorized intention to climb to REC MAX, it does line up:At 1 h 52, the PF said to the captain “look, the REC MAX has changed to three seven five”
well the little bit of turbulence that you just saw we should find the same ahead we’re in the cloud layer unfortunately we can’t climb much for the moment because the temperature is falling more slowly than forecast So what we have is some REC MAX a little too low to get to three seven
--- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
--- Defend what you say with arguments, not by imposing your credentials ---
Comment
Comment