Menu Sign In Contact FAQ
Banner
Welcome to our forums

AF447

“…they could see the mismatch between thrust lever position and engine instruments.”

No. They saw their IAS dropping. Then started to react.
On the Bus it would be the same.
You don’t asses parameters of the approach by thr lev position because it varies greatly during configuration stages. Speed drops below vapp minus 5 – unstable – go around – thr levers toga.

“ will you tell me you have never witnessed an unsafe condition which would have been averted…”

No. Never from the autopilot side. The famous “..what is it doing now???” comes from the pilot error only. On a data input stage or wrong monitoring of FMA (flight mode annunciator). You can always take over.
Pilot error has to be corrected by positive and clear control transfer and using takeover button.
Works every time. 17 years and more than 13k hrs. on this thing.

Poland

Raven wrote:

No. They saw their IAS dropping. Then started to react.

This is basic airmanship, not type-related. My point is that the reaction after IAS decrease is different if :

777: you have already seen the throttles advancing and the mismatch vs engine parameters (ie throttles went up but engines stayed at idle), then you know the issue is not related to autoflight mode and you have a dual engine problem
A330: first indication being the engine parameters, then you need to check the autoflight mode on the FMA to see whether some fault (see note below) resulted in the correct mode not being engaged. Once you decide the correct flight mode is engaged AND verified correct speed selection, then you know you have an engine problem, but not before.

“Some crucial seconds elapsed”, to quote Chesley Sullenberger…

The same could be said of the control transfer SOP: crucial time is wasted until you realize you want to take over and the button and callout are activated. No button pushing on the 777, simply grabbing the yoke…while making the callout.

(note) Of course PROB90 when the correct autoflight mode is not engaged it is the pilot’s fault, but at that point in time, on short finals, would you start discussing whether someone in the cockpit made a mistake or it was an autopilot fault? Is that the issue in hand?

In the case of AF447, no doubt the pilots would have acted differently had they known the copilot was pulling all the way back. A “dual input” aural and visual warning would have added to the chaos, and such warning will not tell you the amount and direction of input from the other pilot, whereas a mechanically linked yoke will, without any brain power bandwidth usage required. Yes it is easy to blame the pilots, but if you are in PIC on a flight, you are in charge of a good outcome whether your copilot is the best or the worst (both questionable criteria) in the business. A standard position feedback yoke helps. Also, some what is it doing moments would be averted when you know for sure it is the autopilot that is providing a certain command rather than a fault (extremely rare) or a wind gust (more often).

To answer one of the q’s above unless the control takeover button is pushed, both sticks’ input is summed. Then when the button is pushed, the summing is suddenly eliminated resulting in a sudden control input depending on how hard the no-longer-PF was moving the stick.

It is clearly explained here



SOP’s are designed to cover 90% of events. My question is whether you need a thinking pilot in the cockpit to cover most of those. The real issue is the other 10% , as this accident shows. I, for one, like to have a properly thinking pilot in the cockpit, and an aircraft designed to allow him to be in control without needing to be an aerospace engineer.

Last Edited by Antonio at 12 Oct 15:15
Antonio
LESB, Spain

„… This is basic airmanship, not type-related. My point is that the reaction after IAS decrease is different if :…”

Antonio, no. Reaction to dropping speed is exactly the same:
Thr levers TOGA and execute a goaround immediately. Solving problems with modes comes much later. First fly.
So when you push your levers you will find out that you have dual eng flameout on Triple and Bus. Same time.

Poland

Raven wrote:

First fly.

It’s very difficult to ‘first fly’ in any situation if you first have to ‘think Airbus’ to try and work out whether the aircraft is actually responding to your control inputs or doing something else because (a) the guy in the other seat is commanding something else, or (b) the aircraft thinks it knows better.

What is the logic of summing/averaging the two control inputs if they are conflicting? In what possible situation might that be desirable? Surely if you absolutely refuse to link them then the logical thing would be to always give priority to the left – that being the commander’s stick.

There is no reason whatsoever, other than a perceived purity of design philosophy, not to link the sticks. No reason at all why it might be desirable to let the two sticks doing different things.

EGLM & EGTN

Surely if you absolutely refuse to link them then the logical thing would be to always give priority to the left – that being the commander’s stick.

I understand the idea behind your post but it doesn’t work that way, multi pilot ops have clear division of controls (PF/PM) and one simply doesn’t grab the wheel and overrides the other one. I’m sure there’s many reasons why sidestick priority is obtained through a button and a corresponding aural alert. There’s probably lots of research in some archive why airbus went for individual non-linked sidesticks.

always learning
LO__, Austria

Reaction to dropping speed is exactly the same:

Please recall the aircraft suffered a loss of IAS close to service ceiling (high ISA ITCZ) in convective high turbulence conditions. Tops in the ITCZ reach FL600. The airplane maintained FL370 for some minutes while the FO attempted to maintain wings level, while increasing pitch. No stick shaker in FBW, and as the airplane started an uncommanded descent the AoA in a deep stall resulted in the ADC thinking the aircraft was on the ground and for a time IAS was zero, even as the probes started to function again. You then have the effect of the senior FO reducing AoA bringing the stall warner back on as IAS started to function again, which @Graham has pointed out earlier, added to confusion.

Oxford (EGTK), United Kingdom

Snoopy wrote:

multi pilot ops have clear division of controls (PF/PM)

Which means they shouldn’t both be trying to fly the aeroplane at the same time. If they are, that whole system has completely broken down and we are in a non-normal (if not actual emergency) situation. In those circumstances why would it not be appropriate that the aircraft defaults to taking orders from the PIC?

Or just link them – like every other aircraft – and the whole issue goes away. I’m sure Airbus have all sorts of philosophical justification for doing it that way, but my bet is they never considered two pilots in an emergency situation who were both slightly baffled by what the aircraft was doing and who were giving conflicting control inputs.

In the dual controlled light aircraft I fly, I am hard-pressed to think of a single more dangerous and stupid hypothetical modification that de-linking the primary flight controls and replacing that link with a combination of a priority button and aural alert.

EGLM & EGTN

Or just link them – like every other aircraft – and the whole issue goes away.

Fair point.

always learning
LO__, Austria

Yes, fair point.

We cannot think pilots will act like robots. SOP’s have been clearly described by @Raven. Most SOP’s could be acted upon by the machine without pilot intervention. Furthermore, with the advent of CPDLC even ATC interaction could be dealt with by the machine.

I do not know BA SOP’s but in BA38 , the pilots were aware of something amiss before total loss of thrust control by observing throttle lever angle. They never initiated a go around (what was the point?) since they had seen the thrust levers at full with the engines being at or around idle thrust. They only tried to retard and then re-advance the throttles with no result, no time for more. A/P was disconnected by direct action of the PF on the yoke in reaction to shaker. Flap was retracted by airmanship criteria of PNF. What was the SOP in all of those actions? All lives onboard were saved by them though.

Likewise, how departed were the AF 447 crew actions from their SOP? No lives onboard were spared by those actions though.

Think also QF32 …how much did SOP’s help those people?

So most of the times SOP’s will save you . I am not sure we need a PIC for those. It is the rest that we need a PIC for.

Raven wrote:

So when you push your levers you will find out that you have dual eng flameout on Triple and Bus. Same time.

No, the BA38 were ahead of the game by aircraft design (see report linked above). You cannot expect a robot pilot to act the same SOP regardless of his sensory inputs . We should use that for the benefit of flight safety, not the other way around.

The Airbus design is that of aerospace engineering MSc’s (I happen to be one of those) wanting to make a difference vs Boeing so they could at last overtake in sales, and they did. Habsheim was a bitter warning, though, that human pride and ingenuity alone, like the proverbial Tower of Babel, do not make the best product. Add some humbleness, common sense and listening capacity to the equation and the result will likely be better.



Last Edited by Antonio at 12 Oct 19:46
Antonio
LESB, Spain

To be fair, , Airbus have fostered the enhancement of SOP’s and design changes to avoid the re-occurrence of those accidents, and the safety records is overall oustanding.
What will be the next situation that nobody ever thought of? I for one would like a pilot in control to take care of that one.

Antonio
LESB, Spain
Sign in to add your message

Back to Top