NTSB Recommendation re Airbus Rudder Travel Limits

[stepwilk]

Well, you're absolutely right about that: Suggesting that there is a connection between 9/11 and the way that AA flight was flown is indeed pure conjecture. Also pure claptrap.

[vapilot2004]

The rudder channel raw data from the AA587 crash was so inconceivably unrealistic, NTSB investigators had to run it through the filter of a matrix of other A300 aircraft recordings to make any sense of it.

Part of the challenge involved the slow recording rate the A300 FDR box used. Those units were de-certified and replaced.

[411A]

The rudder channel raw data from the AA587 crash was so inconceivably unrealistic...

And, no wonder.
Whom in their right mind would bicycle pedal the rudder on a jet transport airplane at any speed, let alone a higher speed.
Oh, I forgot...it was an American Airlines airplane.

[Zymurgist]

c46r

Quote “Where does AF 447 fit into this????????????"

AF447 was in Alternate Law flight mode. This turns off the rudder limiter and allows the pilot full rudder deflection (~+/-30d) even if travelling at cruise speed. Normally the rudder is limited to about 4d of travel at cruise.

It takes only about a half inch of pedal movement to get the full 4d deflection and there is nothing to stop the pedal from traveling to the floor nor stop the rudder from deflecting the full 30d if the limiter is turned off.

It would be very easy for the pilot to overcorrect for an aircraft upset due to turbulence and or a stall and exceed the maximum design load on the VS with the limiter turned off.

[fox niner]

I don't get it. If a very, very dangerous flight technique can not be trained out of somebody, then that pilot should be grounded. Regardless of the strength of the vertical fin.

[411A]

If a very, very dangerous flight technique can not be trained out of somebody, then that pilot should be grounded.

One might reasonably then ask, if the company training program taught the pilot how to improperly apply the rudder in potential 'upset' situations...what the heck was wrong with the company?

[mm43]

Zymurgist wrote in Post #24

AF447 was in Alternate Law flight mode. This turns off the rudder limiter and allows the pilot full rudder deflection (~+/-30d) even if travelling at cruise speed. Normally the rudder is limited to about 4d of travel at cruise.

AF447 was an A330-206, and contrary to what you have indicated, when the control mode changes from Normal to Alternate Law, the Rudder Travel Limiter Unit locks the maximum available amount of rudder travel at the last valid angle based on CAS and barometric altitude. At FL350 and M0.80 this was +/-7.9°, and is the value found by Airbus on investigation of the RTLU after the recovery of the Vertical Stabilizer. The A330 series also has a PTLU fitted as standard.

mm43

[Flight Safety]

RatheBeFlying said

The present design is an ergonomic trap. The ratio between breakout force and full deflection at high speeds makes rudder next to impossible to modulate.

He's absolutely correct.

During the NTSB investigation of AA587, they had several pilots (including airline and test prilots) run through a series of tests at various airspeeds to see how well the pilots could modulate rudder deflection. Rudder deflection on this type is limited with increasing airspeed by progressively limiting rudder pedal travel. At higher speeds there's only a small difference (forget the exact figures now) between the breakout force and full deflection. The tests showed that beyond a certain airspeed it was no longer possible to modulate rudder travel, as the pedals basically became on-off switches providing either no deflection at all or full deflection. Even the test pilots could not produce a modulated value like 25% rudder deflection at higher airspeeds.

This is not the fault of the pilots, but rather a fault in the design of the rudder travel limiting system on the A310 and A300-600. This is the issue that NTSB wants to address.

[fox niner]

one might reasonably then ask, if the company training program taught the pilot how to improperly apply the rudder

Well 411A, true, true.....But: If a company can teach a pilot a wrong flight technique, it should be possible to unteach him/her. Otherwise I would be very interested in those teaching methods. Sounds like Aldous Huxley's Brave New World brainwash sessions!

[deSitter]

Have to agree with Flight Safety here - no airplane should be breakable from control input. Why not put strain gauges in the structure so that rudder max would be controlled by accumulated strain? Could be overridden in a dire emergency I guess..

-drl

[BEagle]

.....no airplane should be breakable from control input.....

You really expect designers to design aeroplanes for total idiots?

I'm with 411A on this issue! What on earth was that company doing, proposing such a stupidly aggressive way of flying the aeroplane?

[Flight Safety]

Beagle, the pilot was not stupidly aggressive. He was taught to use rudder inputs as part of upset recovery (questionable training I know). What he probably didn't know (while doing this) was there was no way to control the amount of rudder deflection he could input. At that airspeed, the rudder travel limiting system provided either full deflection or no deflection. So the "aggressiveness" you mention is part of the design flaw of the rudder travel limiter, not the fault of the pilot. Full rudder deflections just added to the upset, which the pilot continued to try and correct. This is the PIO situation NTSB is trying to address.

What you don't seem to want to accept, is there is a design flaw here the NTSB wants corrected. If this fault were absent, I doubt this accident would have occurred, even with the questionable training. I don't think properly modulated rudder inputs would have overloaded the vertical stab, or added to an upset that seemed (to the pilot) to require additional rudder inputs.

[misd-agin]

Huck,

Spoke with a party to the investigation. He seemed to disagree with the direction the investigation was heading until he got into a full motion sim that could duplicate the lateral loads, etc. That seat of the pants sensation is completely missing from just looking at the data. He said the control inputs made a lot more sense when you acutally felt similar lateral loads.

[bugg smasher]

The A330 series also has a PTLU fitted as standard.

Is this also a standard feature on the A319/20/21?

[misd-agin]

Fligth Safety - it's not impossible to modulate the rudder throws. It's not just zero or 100%. However, it is very very tough to modulate even after undergoing additional training that was implemented after the accident. Been there done that. However this was in training AFTER learning how sensitive the rudder was. Perhaps a key difference in being able to modulate the rudder travel was the lack of 'real time' pressure/time constraints.

Training included rudder displacements at <165 kts(?), 210, and 250 kts. The change from <165 to 210 kts was "hmmm, ok". The change from 210 kts to 250 kts was "whoa". It's just a different experience at the higher speeds. And the sim couldn't accurately duplicate the side loads being generated which might have changed it to "WHOA!".

I think the rudder travel is 1.3 inches above 250(?) kts. DC-9/S80 is 1.2(?) inches so it's not just the rudder movement that is the problem. It's rudder pedal travel, break out force, force after break out is achieved, degrees of rudder travel per pound of force, etc, etc.

google came up with this -

http://www.alliedpilots.org/Public/t...submission.pdf

Seems to me that the A320 incident shows the need for all airline pilots to have training in rudder use at high speeds to understand any limitations their particular a/c might have.

[Flight Safety]

Misd-agin, I agree with all you said. However can you imagine in the adrenalin rush of an upset at 250kts, being able to modulate the rudder on the airplane with the current design? I think not.

[PJ2]

bugg smasher;

Yes, the A319/A320 series is fitted with an RTLU. It is under the authority of the FAC, Flight Augmentation Computers. The AMM states that the limitation remains lower than that which would induce load limits on the structure, throughout the flight envelope.

[grumpyoldgeek]

I don't get it. If a very, very dangerous flight technique can not be trained out of somebody, then that pilot should be grounded. Regardless of the strength of the vertical fin.

The pilot is not completely to blame here. Look carefully at the breakout force issue and compare it to a Boeing. Or compare it to a trainer. Responsive rudder action is required on some aircraft and many trainers. There's also the issue of the Va specification. The aircraft in question was flying below Va, which by definition allows lock-to-lock control throws without breaking the aircraft. After the accident Airbus qualified that to mean one excursion to lock and then back, not multiple cycles. I'd appreciate some input on this issue from the more experienced pilots on the forum.

[misd-agin]

grumpyoldgeek - "The aircraft in question was flying below Va, which by definition allows lock-to-lock control throws without breaking the aircraft. After the accident Airbus qualified that to mean one excursion to lock and then back, not multiple cycles."

I think the problem is line pilots thought Va for light a/c applied to large transport jets. The problem is the engineers or test pilots of transport a/c had different interpretations of what Va meant vs. what the line pilots thought it meant.

Boeing issued clarification also.

[411A]

I think the problem is line pilots thought Va for light a/c applied to large transport jets. The problem is the engineers or test pilots of transport a/c had different interpretations of what Va meant vs. what the line pilots thought it meant.

The problem here is...'thought it meant'.
Versus...what the actual certification criteria are in CAR4B/14CFR25.
If we look back more than thirty years with jet transport training, the trainers then knew damn well what the problems were, and yet...the 'new breed' that came after, did not, and further, would reject any input from older/more experienced guys.
The result?
A large number of dead bodies.
American Airlines (as a company) has a lot to answer for, with their misguided 'upset training' and will pay dearly, in the end.
Airboos, with their poor design, will also pay.

Meanwhile, dead bodies cannot speak from the grave....but their relatives can, through their attornies.
It will take at least 15 years to settle the legal action.