bubbers44

757 was my final airplane, I still do not get the question.

OK465

scary.......

bubbers44

bye maybe someone knows what you mean

Old Carthusian

The yoke/sidestick debate is another of those irrelevancies which crops up from time to time. The sheer number of Aibii whizzing around safely using sidesticks proves this. The question as to whether the use of a yoke would have helped is rather moot. Given the captain's possible befuddled state on entering the cockpit he could have possibly overlooked the yoke being pulled back or misinterpreted the angle so discounting the evidence. The only reliable indicator would have been the instruments (minus airspeed). We do seem to have some indication that the PF's scan had broken down and that the PNF was deriving some information from his instruments but not acting on it. A responsible captain would first of all look at the instruments not the position of a yoke - to do so would be otherwise unprofessional. The instruments would give an accurate indication of the state of the aircraft. As Dozy sensibly put it neither method of input is perfect and both have their disadvantages but the substitution of a yoke in this situation would have made no difference. A different more active PNF would have made a difference but once again human factors are the most important here.

CONF iture

Or he could have possibly realized how the picture was wrong with a yoke in the stomach at FL350.

If I had to enter that flight deck at that time, the first obvious clue would have been that full back yoke at 350 ... now call me unprofessional.

HazelNuts39

Just learned something on another thread:
At first sight, this would seem to be preferable to the BUSS in an UAS situation. What are the disadvantages?

rudderrudderrat

Hi HazelNuts39,
We have that system already. With blocked pitots and your IAS shows say 60 kts, then the superimposed stall speed is uselessly displayed off scale.

control surface deflection:
The crew of A380 VH-OQA were concerned about their roll controllability with a big fuel imbalance. At a lecture I attended, the Captain mentioned that there was no feed back through the side stick as to how much aileron was being applied. They therefore frequently looked at their flight controls system page to see how close to aileron saturation they were getting. (only inboard ailerons were working)

see pages 6-7 of http://www.atsb.gov.au/media/2888854...y%20report.pdf

"Prior to leaving the holding pattern, the crew discussed the controllability of the aircraft and conducted a number of manual handling checks at the holding speed. The crew decided that the aircraft remained controllable...As the crew started to reconfigure the aircraft for the approach by lowering flaps, they conducted further controllability checks at the approach speed and decided that the aircraft remained controllable..."

On Airbus, because the control surfaces are being moved "invisibly" (except on a systems page which may not be displayed) when the ailerons or elevator reach say half deflection, then I think there should be some alert. When the stab trim runs outside its normal range - then there should be a warning.

A33Zab

Not as much weight and complexity as this:

787 Controls




Maybe IF SS feedback/interconnection was available it would not have happened.

But

IF ADR2 hadn't have been the voted median.....
IF they hadn't switch over to ADR3 (display on RH PFD)...
IF captain had returned to cockpit earlier (or never left)....
IF PF didn't overreact......
IF multiple UAS had lasted longer than 10s in the first fase....
IF they had realized what was going on....
IF training,knowledge, CRM, communication, SOPs would have been sufficient.....
IF AoA indicator was fitted (or BUSS was opted)....
IF inertials had been outside the abnormal attitude values.....

The outcome could have been different too!

Maybe the clues are in the modifications which have been adopted after this tragedy.

For systems:
- Inhibition of AP/FD reconnect after UAS.
- More rigid USE MAN PITCH TRIM PFD msg when autotrim is not available.
- F/CTL ALT LAW (PROT LOST) msg accompanied with reason what caused the transition.

IMO the installation of an AoA indicator will be mandated - in the final report - because it is already optional and provisions are available.

Old Carthusian

The uncertainty is the issue but on balance the postion of the yoke will not make any difference to an appraisal of a situation. A337ab summed the situation up perfectly. A yoke will not help in this kind of environment.

HazelNuts39

I don't understand that it would do that. Are you sure? Please explain.

rudderrudderrat

Hi HazelNuts39,

Please see John T's explanation http://www.pprune.org/tech-log/48445...areness-2.html.
If the stall speed is calculated as being X kts and superimposed on the speed display accordingly, but due to iced over pitots at 250kts, your IAS only shows 60 kts, then your stall speed will be off the scale at X kts. All you'll know is that you think you are below the stall speed on the IAS read out - even though you may still be at 250kts. The BUSS system (when displayed) is independent of pitots.

As far as I know, BUSS is only displayed after all 3 ADRs are turned off manually.

Hi A33Zab.

Thanks for diagram of 787 yoke system. It's definitely a lot more complicated, heavy and costly than the Airbus side sticks - but at least the control surface displacement feed back loop to both pilots is complete.

HazelNuts39

Let's assume that X = 200 kts, i.e. 80% of 250. When IAS then drops to 60 kts, my understanding is the stall speed will be indicated as 48 kts (80% of 60 kts). Am I wrong?

rudderrudderrat

Hi HazelNuts39,

I think you are probably correct.
I hadn't thought through the logic of the dynamic stall margin being presented as a %age of what's displayed.

A33Zab

As I recently understand from the 757/767, despite the 'interconnected' yokes, the possibily of splitted RH/LH Elevator surface control doesn't seem to be a good idea either.

May hope that the 787 FBW protects from such a situation.
How would it determine which input to follow?

rudderrudderrat

Hi A33Zab,
I don't know if you are referring to some incident (link?) which I don't know about.
On the TriStar we could split the Capts & FO's load paths to the L & R elevators and L & R ailerons respectively, so it was technically possible to fly each side independently. They were only split if we had a control jam on one side. PFM.
In that case, you'd have to do as Airbus now do - only one pilot at a time flies the plane.

DozyWannabe

EgyptAir 990 possibly. Full forward in the RHS and full back in the LHS led to split elevators - I think Boeing changed the design after that.

Strictly speaking I think all airliner ops are only supposed to have one pilot at a time flying the plane, are they not? You can usually spot documentaries and programmes that have done their homework versus those who haven't because the latter will have both pilots with hands on the yoke in cruise even when nothing is untoward.

I know you and others perceive the feedback loop as essential, but I'm not sure it is as essential as you believe it is when an airliner's flight surfaces are fully-hydraulic. A lot of people give Airbus a hard time on the KISS principle, but in the case of the control setup, theirs is a lot simpler than the yoke systems of other manufacturers. From an engineering standpoint there's a lot more points of failure in a backdriven feedback system than the spring-driven feel of the Airbus sidestick design.

Not to mention some of the disadvantages of the yoke system that have already been mentioned (e.g. dual input becoming a strength contest between pilots, lack of positive override).

The short version is that for every "might have made things better", there's a "might have made things worse" - *if* you're looking at things dispassionately, and all other things being equal.

Is tactile feedback necessary or is the yoke a crutch? I feel sure I've said this before, but notwithstanding the brilliant job they did saving as many people as thay managed to, the Captain and F/O of UA232 kept forcing their yokes forward and left even when it was having no actual effect on the flight surfaces. I suspect the answer is somewhere inbetween.

Just to point out a few indisputable facts:

• Northwest Airlines Flight 6231 (B727) - PNF had yoke in his lap during stall, did nothing.
• Birgenair Flight 301 (757) - PNF had yoke in his lap during stall, did nothing.
• Air Florida Flight 90 (737) - PNF had yoke in his lap during stall, did nothing.

rudderrudderrat

Hi DozyWannabe,
It doesn't matter that the flight surfaces are powered, it's more about being kept in the loop. e.g.
I can see the total fuel on board constantly, but the individual fuel tank gauges out of view, (they are on a systems page) so I get a warning when one tank is low, or an imbalance is building.

With the flight controls, I don't get any warning that I'm approaching maximum deflection with aileron or elevator, (or that the stab trim has run beyond it's normal range automatically). If I don't get that feed back through the side stick - then it would be nice to receive a warning somehow.

If AF 447 had some such warning, then I bet the Capt. would have been better able to diagnose their problem.

DozyWannabe

Understood. All that is fairly easily implemented without having to mess with the fundamental systems design and without having to add much complexity in software terms.

Of course in Alternate and Direct Laws, you know that full lateral stick deflection = full aileron deflection. In stall conditions, ailerons are more-or-less useless. What puzzles me is that the Captain tells them to use the "rudder bar" towards the end, but neither explains why nor does he explicitly state he suspects a stall - this leaves a lot of open questions. All that aside, I still don't understand how he could have seen an aircraft out of positive control with a rapidly unwinding altimeter - information that was clearly available to him - and not said "We're stalled". You shouldn't need to confirm yoke or sidestick position to be able to work that out if you understand aircraft.

Over the last few pages, those that consider yokes and feedback to be a panacea regardless of the circumstances are trying to hammer the known facts about this accident to fit a position they've already taken for reasons of their own, and yet they ignore incidents where an aircraft has stalled, the yoke has been back and the PNF did nothing, or (as in the case of the LH A320) they take the assumption that in a pressure situation a PF will automatically relinquish the controls as soon as a superior PNF touches them as hard fact, when there are plenty of incidents where that has not been the case.

Ian W

Logically your argument also supports removing all aural warnings as there are numerous examples of flight crew disregarding them.

Where your argument is weak is that it misses all the potential incidents that became non-incidents because of the extra information provided.

A33Zab

EgyptAir 990 is indeed I did refer to.

I know it was initiated by FO but still the captain was unable to recover the situation.

Maybe summation (Airbii) isn't too bad at all.


All suggestions to improve safety/recovery should be considered.
but since this is a one time occurance NOT all the suggestions needs to be implemented, first of all they need to be effective.

The A330 is late 80 technology, CRT displays at that time were limited in performance and in symbology displayed at the same time.
Large LCDs in A380(and will A350) / 787 are capable to display a lot more information at the same time e.g. flight controls position.

For the AF447 the F/CTL SD page was automatically called and in view when F/CTL RUD TRV LIM FAULT ECAM was triggered.
In the BEA reports they do not mention any CLEAR action or SD page switching. (if recorded), the F/CTL page could still have been in view all the time.