Another what would you do question.

[Lu Zuckerman]

Another post on this forum asked what would you do if you suffered a control failure after V-1 well, here is another question for Airbus jocks.

What would you do if during take off your flaps started to retract or, if you were flying at cruise speed your flaps started to extend?

An Air Canada A-320 suffered a flap retraction on take off so don’t say it cant happen.

:confused: :eek:

[411A]

Lu,
You elaborated on this before in one of your postings, but suspect many of the younger guys missed same...

How about a repeat...'tis important for all to know.:eek:

[Lu Zuckerman]

If an uncommanded signal is sent to retract or extend the flaps it could be from an external source or it could come from the computer. In the former condition it is assumed that the command sensor unit is in the off position and in this case the computer is off and therefore can’t correct for the erroneous signal. In the latter if the signal is generated by the computer it is assumed to be a valid signal and therefore the computer will not respond. Another point to consider is that not all signals are electrical in nature.

:confused:

[Lump Jockey]

Did anyone, (and more importantly, know the cause of), see the "caught on camera" footage of an A320? going VERTICAL then dropping back down and skipping around for a bit!! What was that about!?:eek:

[Onan the Clumsy]

skipping? Must've been a girl A320 :cool:

[Saint-Ex]

And what would you have done on a 747 400 that commanded a downward signal to the master elevator when the u/c was retracted?:confused:

[Lu Zuckerman]

My original post was directed towards Airbus drivers and not 747 operators.

The questions are very relevant and the answer is equally important to the flight safety of Airbus (A-310 / A-320) operators.

:eek:

[RadarContact]

First case: uncommanded flap retraction.

Before V1: abort (well, yeah. Obvious)

After V1: TOGA, depending on RWY length, selected VR and terrain situation: rotate at Vmu (clean) and do a shallow climb, slowly accelerating (my personal opinion)


Second case: uncommanded flap extension.

Most likely at cruise speed you'll lose your flaps... reduce speed to min clean anyway and plan for flaps up landing...

[Lu Zuckerman]

Most likely at cruise speed you'll lose your flaps... reduce speed to min clean anyway and plan for flaps up landing...

Am I to assume that you are saying the flaps have been ripped off and the wing structure has not been compromised? If so, the A-310 I believe has a supercritical airfoil that is aft loaded and as such the retracted flaps make up a large part of the lift. Without the flaps the aerodynamics have been seriously compromised and the aircraft is uncontrollable. You have to be able to halt the extension or retraction before it becomes dangerous. The question is what do you do to stop the movement?

QUOTE: After V1: TOGA, depending on RWY length, selected VR and terrain situation: rotate at Vmu (clean) and do a shallow climb, slowly accelerating (my personal opinion)

In the case of the Air Canada A-320 that suffered a flap retraction on takeoff the pilots had to use every bit of available power and they still almost lost it.



.
:eek:

[Max Angle]

You can formulate a million and one situations like this, some of which could never happen, some of which could and a few that have. I don't really see the point of your question, many aircraft types have suffered uncommanded movement of various flight controls and surfaces, it is not a problem peculiar to Airbus.

If fly the A320/321 so if it happened to me (assuming one of us noticed in time) above V1 then I guess I would select the flap handle to 0 then back to the t/o setting. If that didn't work then I would just do my best to salvage the situation as best I could with power and pitch. If the a/c weight, runway length and obstructions etc. etc. are on my side then we would get away with it and if not then we are going to have a big accident and that would that I suppose. Excrement happens!.

[Lu Zuckerman]

To: Max Angle

I agree that uncommanded movement of both primary and secondary flight controls has happened in the past and it will continue to happen in the future. In this particular case I am addressing the secondary flight controls on the A-310 and the A-320. Although I specifically addressed the flap systems on both aircraft it applies to the slat system on the A-310. I am not familiar with the A-320 whether it has leading edge slats. So that this does not degenerate into a P*****g contest I will explain what I think should be done to counter the uncommanded movement of the secondary flight controls.

Axiom 1. If the computer fails and generates an uncommanded signal then the computer can not stop it. Axiom 2:If the uncommanded signal stems from an internal failure of the PCU then the computer can not stop it if the signal is to extend the flaps. If the flaps are extended and there is an internal failure of the PCU causing the flaps to retract the computer can sense this and stop the movement. This will provide sufficient time to make the necessary corrections.

In either case the movement is generated by the PCU. Both the flap and slat system have two different hydraulic modules connected to a hydraulic motor driving through a gearbox. If there is an internal failure of the hydraulic module then only one of the two motors will be commanded to rotate which means that any movement will be at half speed but could in fact be slower depending on the extent of the failure.

If the failure is in the computer the movement will be at half speed.

The secondary system is powered by three different hydraulic systems. The Green and the Yellow system power the Flap system and the Green and the Blue system power the Slat system.

If an uncommanded signal is generated on the flap system it is necessary to turn off both the Green and Yellow system. This will leave you operating on only one hydraulic system. Once both systems are off, turn one system on. If the movement continues then turn that system off and the other one on. Now you are operating on two systems. A similar action should be taken on the Slat system.

As I had previously stated the Air Canada A-320 suffered a flap retraction during takeoff. I do not know if it was an electronic failure or a hydraulic failure. The firm that developed the Slat drive system suffered an uncommanded extension of the Slats on their iron bird. It could have easily been an uncommanded retraction. In either case the computer could not identify the movement or even to stop it. Since the PCUs are almost identical between the Flap and Slat systems the same problem could happen on the Flap system.

:eek:

[rottenlungs]

Hi there

I`m curious about the Air Canada a320 incident. Do you know where I might get some more information about it?

Thanks

Lungs

[Max Angle]

Well I see the logic of what you are saying, the A320 system has the same layout in terms of hydraulic power split and reaction (half movement speed etc.) to failure. In theory what you say would work, in practice however realising you have a problem, working out what has gone wrong, working out what to do about as well as trying to make the aircraft fly and keep it flying is going to be impossible in most cases. Sods law means that it will happen on a short runway in foul weather etc. etc.

The fact is that there are some failures on aeroplanes that are more often than not going to result in a fatal accident despite the best efforts of the crew. The certification process recognises that fact and should ensure that failures of that nature occur very infrequently.

[Lu Zuckerman]

To: Max Angle

The fact is that there are some failures on aeroplanes that are more often than not going to result in a fatal accident despite the best efforts of the crew. The certification process recognises that fact and should ensure that failures of that nature occur very infrequently.


What you have stated is correct but in this case it is only theory. That is how it is supposed to work but in this case it did not. The FAA, CAA and the DGCA were made aware of the problem but did nothing about it. When the runaway condition was first discovered there were 17 A-310 aircraft in service. The failure was an internal crack in one of the two hydraulic modules causing the slat iron bird to drive in the extended direction. The firm determined that the crack developed as a result of a faulty machining operation. The location of the crack did not allow direct visual or manipulative access to the damaged area. The firm manually tried to work out the area using sand paper and crocus cloth on 17 ship sets. They contacted the two operators (Lufthansa & Swissair) telling them that they had developed a more reliable power control unit and when they had a plane on the ground the firm would install a new pair of power control units at no charge. The contract required that any failure or incident during development that effected Reliability, Safety or Maintainability had to be reported to Airbus. No information regarding the failure was ever reported. Another design defect that was never reported to Airbus was the fact that the flaps and slats were not electrically bonded to the airframe. This was also reported to the FAA and the design was never changed. The internal crack developed at or about 1800 duty cycles and the in service aircraft were quickly approaching that number of duty cycles. Regarding the electrical bonding Airbus had a design directive that defined the design requirements to achieve electrical bonding and lightning protection. In that same directive were the procedures to be used to verify the electrical bonding and this would be performed on the production line by Airbus. Either the procedure was faulty or Airbus never performed the check because the lack of bonding was never discovered.

It has often been said that Line pilots are also test pilots.

To: Lungs

The information regarding this incident was never published nor, was it reported to the Canadian MOT. I found out about it when I was riding up front on a flight from Toronto to Chicago. I was discussing the problems above regarding a runaway system when the Co Pilot who was a four-striper and chief safety pilot handed me a fax. The fax stated that two days prior to that day an Air Canada A-320 experienced an uncommanded flap retraction on take off and the pilots had to use all available thrust to keep the aircraft flying until they built up air speed. It happened two days before Christmas in the year 2000 if memory serves me correctly.



:eek:

[rottenlungs]

Hi Lu

Thanks for the info.

There seems to be quite a large number of these type events, 747 elevators, 737 rudders, A320 flaps.

What is the certification standpoint, bearing in mind that there is real potential for accidents with this type of incident?

Is it just a case of testing to ensure that it is so unlikely that its not worth considering or are procedures written for these events in the operating handbooks?

Thanks again

Lungs

[411A]

Oddly enough, it doesn't happen with old iron....B707 and TriStar come to mind.
Stout birds....and reliable too:D

[Lu Zuckerman]

To: Lungs

Quote: "What is the certification standpoint, bearing in mind that there is real potential for accidents with this type of incident"?

"Is it just a case of testing to ensure that it is so unlikely that its not worth considering or are procedures written for these events in the operating handbooks"?


From a certification standpoint the FAA requires a detailed FMECA and a Safety report. From an operational standpoint the manufacturer must demonstrate the operational capability of the flight control system using what is termed an Iron Bird. The Iron Bird cycles the flight control system under controlled flight loads to simulate the air loads during various phases of flight. The Iron Bird will simulate a very high number of cycles to demonstrate both operability and reliability and the iron Bird will in most cases be still operating after the aircraft is certified and in revenue service.


The FMECA defines the relationship of a piece parts failure to its’ effect on the next higher level all the way to the aircraft level. The FMECA also includes the failure rate of the part, which is determined in the Reliability analysis. This is where the system gets skewed due to the methodology of determining the failure rates. Sometimes the analyst will select failure rates that are not representative of the part under analysis in order to meet the failure rate for the unit under analysis. In most cases, the failure rates are determined through the use of K factors that allow the analyst to determine the ultimate failure rates by adjusting the failure rates through the use of these k factors. As such, the final number in no way relates to the actual part in the unit.

These numbers are then inserted in the safety hazard analysis and are then mathematically manipulated to determine the safety level of the unit under analysis. To boil it down to the lowest denominator it is GIGO (Garbage In Garbage Out).

All of this is then turned over to the certification authority and if it meets their requirement certification is approved.

Even though the Safety report indicates that the flaps will never move uncommanded at a frequency no more frequent that 1 10 9 it has happened at a frequency much higher than the requirement and the FAA does nothing about it. The NTSB may require a modification but the FAA will overrule that requirement after conducting a cost benefit analysis.

It’s just the manipulation of numbers and that is what safety is all about.

:eek:

[RadarContact]

@ Lu:

You would willingly switch off two hydraulic systems to stop the retracting flaps? Wow... :eek:

I'd think this would leave you with quite a bigger problem.
But even if you tried, how long would it take (in terms of time and number of buttons/levers to move) to really switch off both required HYD SYS? Wouldn't it be easier in this case to quickliy locate the WTB CBs and pull them? Ok, you'd have a frozen, non-resettable configuration. But it would be faster and you'd still have your HYD power for primary flight control.

Just a thought...

[Lu Zuckerman]

To: Radar Contact

You bring up a good point about turning off two hydraulic systems and then troubleshooting the problem but being unfamiliar with the electrical set-up at the aircraft level I ask the following. By pulling the WTB CBs it would seem to me that you would disable the WTBs and not turn them on.

In my estimation the most dangerous situation is to have the flaps extend at cruise speed which could result in the loss of the flaps and serious damage to the wing structure. At this time the Command sensor unit is in the “Off” condition and as such the computer is sleeping and can not command a corrective action. If in fact the wing tip brakes are set by the computer the PCU is still driving and the torque limiters will lock up the system. However when this same situation occurred on the slat iron bird the wing tip brakes did not lock up. If the slats were commanded out during cruise flight the oncoming air loads would keep them from extending. However the flaps have the assistance of the oncoming air flow which would assist the flaps in extending adding to the effort created by the internal leak or computer malfunction.

The manufacturer of the Flap/Slat computer did not perform a correct FMECA as required by contract and as such there was no examination of the potential failures in the monitoring and control circuits. This may have contributed to the inability of the computer to counter the runaway slats on the iron bird. Also on the first revenue flight of an A-310 (Lufthansa) from Frankfurt to Cairo the pilot could not retract the flaps and had to return to Frankfurt with the flaps extended. The computer was totally ignorant of the failure and could not provide any corrective action on the “dolls eyes”. Naturally he had to fly at a very slow speed as compared to cruise flight.


:eek:

[Ignition Override]

The Airbus which steeply climbed and dived was supposedly somewhere near or over Paris, and my impression is that it was a Romanian (airline) A-320.

I've never had any Airbus training (but might, someday).

Veni, vidi, vici pizza.:}

[2daddies]

Aircraft over Paris was a Tarom A310.

Problem developed when the PF hit TOGA by mistake and instead of disconnecting the autopilot the crew fought against it. The trim ran full travel before the A/P turned itself off leaving a horribly out of trim Airbus which only just stopped short of ploughing into the suburbs.

Just like the two Air China A300-600R crashes (Nagoya/ Taipei).