Interesting note about AA Airbus crash in NYC

[misd-agin]

Originally Posted by Mac the Knife

Forgive me for enquiring of the elite, but I'd like to ask...

"By the time the rudder surface position display blanks out, there were 7 control wheel reversals and 5 rudder pedal reversals, all in about 7 seconds."

This is a large aircraft, not an aerobatic biplane - surely it could hardly have even begun to respond to one control input before another was initiated (and so why the rapid reversal?)?

Just asking, if you don't mind.

The A300 has the lightest control forces, and perhaps the quickest response(havn't timed them) of any large jet a/c I've flown or jumpseated in. Flown would be 5 large jets, j/s'd would add three additional types.

I've been with folks that have over controlled roll inputs. All the hydraulically powered jets with roll spoilers react very quickly. Review the NW 757 @ SEA video and watch how quickly the a/c responds once the roll spoilers extend to a signifcant angle.

[misd-agin]

Originally Posted by bomarc

IN the USA, there are about 25 passenger Airbus 300 aircraft? Isn't that about right?

34 in passenger service with more at the freight carriers.

[misd-agin]

Originally Posted by bomarc

few cloudy

I think you make an important point. none of us as airline pilots would, in the course of smooth manual flight, "flick" the rudder around...or any controls really.

(recall how over controlling the 757 at KSEA was shown during a crosswind landing...see other thread(hairy xwind landing)...it didn't come apart.

the copilot in this crash probably wouldn't either, especially after the admonishment some years earlier while on the 727 (which did seem to survive the rudder movement).

all of this makes me think that there is more than meets the eye to the crash in question. that something happened in a different sequence than we are lead to believe.

I am sure that If I danced on the rudders at 200+knots I would make myself quite air sick. that should be a limiting factor in and of itself.

On a scale of 1-100 abrupt rudder inputs at 160 kts is similar to taking off in gusty conditions. Call it a 10. 210 kts rudder inputs would be 20-30, and 250 kt rudder inputs are 100.

The magnitude of the lateral movement/load increases exponentially. So much so that the AA pilots union had expert testimony about the sensitivity being close to a POI response rate.

[AirRabbit]

Originally Posted by Mac the Knife

Forgive me for enquiring of the elite, but I'd like to ask...
"By the time the rudder surface position display blanks out, there were 7 control wheel reversals and 5 rudder pedal reversals, all in about 7 seconds."
This is a large aircraft, not an aerobatic biplane - surely it could hardly have even begun to respond to one control input before another was initiated (and so why the rapid reversal?)?
Just asking, if you don't mind.

Originally Posted by misd-agin

The A300 has the lightest control forces, and perhaps the quickest response(havn't timed them) of any large jet a/c I've flown or jumpseated in. Flown would be 5 large jets, j/s'd would add three additional types.
I've been with folks that have over controlled roll inputs. All the hydraulically powered jets with roll spoilers react very quickly. Review the NW 757 @ SEA video and watch how quickly the a/c responds once the roll spoilers extend to a signifcant angle.

Originally Posted by misd-agin

On a scale of 1-100 abrupt rudder inputs at 160 kts is similar to taking off in gusty conditions. Call it a 10. 210 kts rudder inputs would be 20-30, and 250 kt rudder inputs are 100.
The magnitude of the lateral movement/load increases exponentially. So much so that the AA pilots union had expert testimony about the sensitivity being close to a POI response rate.

Hey Mac the Knife … I’m not sure about the “elite” remark, but … your question is an excellent one; “why the rapid reversal?” THAT is the $64 question! I think if you watch the animation of the FDR provided by the NTSB, you will see that the aircraft did begin to move in the direction mandated by the control inputs. It IS a large aircraft … and the momentum it would build up would be substantial … and to reverse that momentum would take A LOT of force. But, THAT is precisely the point. Of all the control reversals noted on the FDR in that 9 seconds, I believe only one was held long enough to potentially allow the aircraft to get close to stabilizing with THAT particular control input. In each of the other cases, the momentum was established … that is, the aircraft began movement in the commanded direction … and then, because of control reversal, that momentum was stopped and reversed … multiple times … in a VERY short period of time. As any structures expert will confirm, while rapid control movement will stress the structure, a significantly larger stress, by far, is generated from applying full control input opposite the movement of the structure.

I don’t know that we’ll ever know the “why” behind this particular accident. It would appear that the F/O was pre-disposed to the aggressive application of rudder, and that he was also very quick to apply what he thought was necessary control input to prevent what he thought was happening, or was about to happen, to his airplane. Again, the FDR traces of the cockpit controls (i.e. column, wheel, and pedal) indicate that those controls were displaced – and since they are not “back-driven,” someone in the cockpit had to displace them.

I hope that the readers here don’t think I’m saying that the complete responsibility of the accident is the fault of the F/O. However, he WAS “on” the controls … so he bears at least part of the responsibility. But, as I’ve said, I think he had significant “help” in getting to the scene of the accident. The control sensitivity of the airplane helped. (I’m not rated on the A-300 so I appreciate misd-agin’s comments about light control forces and quick responses.) The training to which he was exposed helped. The understanding of “design maneuvering” speed (or more likely, the “mis-understanding” of that term) helped. His own control application strategy helped. Not being able to recognize and escape from a “pilot induced oscillation” (PIO) also helped.

With these “helpful” pieces all fitting together, along with the comments offered by misd-agin regarding the exponential increase in structural “load” with airspeed increase, I think that the result of the sequence of these particular events is rather straight-forward. My hope is that pilots will take away from this event several important pieces of information … an understanding of the way aircraft are certificated; an understanding of what control inputs due to the airplane; an understanding of what a PIO really is, how to recognize one, and, most importantly, how to get OUT of one if one does develop.

Let me add one additional and very important point here. I think it important to say, and to say clearly, the development of a PIO is NOT necessarily an indication of poor piloting technique. It’s somewhat like a squad of soldiers marching across a bridge … such an event is ALWAYS done “out-of-cadence” as a precautionary measure to avoid setting up a resonant frequency vibration. A PIO is similar. A PIO is almost always initiated through some triggering event – that may, or may not, have a pilot input involved. But once the trigger has occurred, the pilot’s response is critical to either adding to or damping out that PIO. It’s hard to dampen out a PIO unless you know that you are in one, AND you know HOW to get out of one. Maybe these things will be included in future pilot training programs.

[Rainboe]

Bomarc

if the copilot was dancing on the rudder pedals in an excessive fashion, why wouldn't the captain say, "I'VE GOT IT" and take over control? Why nothing like, "what the hell are you doing to the rudder? on the CVR...if this is on the CVR please let us all know.

you are apparently convincing yourself, as many people do in many incidents, that 'there is more to this than meets the eye!'. That 'we are not being told the full story!'. The Captain was not flying, his feet would not have been on the pedals. When the vortex encounter hit, it was obviously very powerful. He would have been more concerned about watching the situation. If he felt the copilot was handling it and responding correctly, I believe he would not have been looking down at the pedals (more important to monitor what was happening), or placing his own feet on them as this would interfere with the copilot's handling. Why should he take control immediately? Seeing the aileron control moved in front of his nose, he would be confident the copilot was responding. I am sure he was probably unaware of large rudder inputs in the buffeting, turbulence and strong roll being experienced. So please don't lead yourself up a wrong track!

[Mac the Knife]

AirRabbit, thank you for your courteous reply. It helps a bit, but I still don't quite see what he was trying to achieve by this.

I don't fly (in my next life!), but I do sail small boats, so for me it's as if he'd put the tiller hard over and then, just as the bows started to swing, slammed the tiller hard over to the other extreme. And then repeated this 4 more times.

This won't get you anywhere in a small boat and I can't imagine a situation when one might do this, even in a blow and a running sea. I know aircraft are not the same as small boats but it still sounds rather odd.

I'm sorry if it's a silly analogy but I'm trying to understand.

PS: On reflection, you might well break the rudder off in a dinghy if you tried this, so what happened is not that surprising. But why do it?

[bomarc]

dear rainboe:

I've been in the left and right seat of jet transports...and I have never been out of touch with what the other pilot was doing with all the flight controls...

I suppose we shall just disagree courteously on this concept.

---

I do ask the following question for all the pilots on the forum:

Given the following:

pay the same
schedule and destinations the same

equality of good looking flight attendents the same

all the same on everything except the plane:


would the airbus 300 series be your first choice in planes to fly? you may choose from any plane currently flying, comparing it to the above plane?

hypothetical of course.

[bubbers44]

I was scratching my head on the captain having his feet on the floor too. I thought all pilots monitored all flight controls when the other pilot was flying especially during critical phases of flight or when something unusual was happening.

I have talked to the pilot flying the A300 trying to land at MIA when he got continuing uncommanded rudder inputs forcing him to go around. He questioned whether they would be able to control it enough to get on the runway at all. No, I wouldn't fly an airbus 300 if I had any other choices no matter how good looking the FA's were or how well it paid.

[Desert Dingo]

No, I wouldn't fly an airbus 300 if I had any other choices

Yeah. All airbusses are crap, aren't they?
The newer ones are fly-by-wire, and the software won't allow pilot input to stall or over stress the airframe.
The older ones, however, will allow a ham-fisted (footed?) pilot to over stress the airframe.
Just like the Boeings that real men fly.
Errr...

[Rainboe]

Well you two (bomarc and blubbers) surprise me. In the normal course of events, a Captain would be aware of what the Pilot Flying the aircraft (in this case the co-pilot) was doing on the controls, but throw in a violent wake encounter and I think he would be forgiven for just concentrating on what was going on rather than monitoring the controls, as evidenced by his lack of comment or overcontrol on the rudders.
For myself, I would be just as happy to fly the A300 as I am to fly the 737 now. Like almost all pilots, I am happy with its certification just as the A300 deserves its certification. You can overcontrol any aeroplane in all phases of flight. Many of us have always been aware of the dangers of overcontrol- I do not understand why so many pilots now seem to find it a shock that it can occur. I started 35 years ago on the VC10 and recalled being hit with a rolled up newspaper for being too 'expressive' on the rudder during dedicated training flights.

[AirRabbit]

Originally Posted by Mac the Knife

AirRabbit, thank you for your courteous reply. It helps a bit, but I still don't quite see what he was trying to achieve by this.
I don't fly (in my next life!), but I do sail small boats, so for me it's as if he'd put the tiller hard over and then, just as the bows started to swing, slammed the tiller hard over to the other extreme. And then repeated this 4 more times.
This won't get you anywhere in a small boat and I can't imagine a situation when one might do this, even in a blow and a running sea. I know aircraft are not the same as small boats but it still sounds rather odd.
I'm sorry if it's a silly analogy but I'm trying to understand.
PS: On reflection, you might well break the rudder off in a dinghy if you tried this, so what happened is not that surprising. But why do it?

Well, I don’t think it’s a silly analogy, and, as I said, the “why” is the big question here. The “what” of this accident is not terribly difficult to understand – even though some here would want to argue with the findings of the NTSB. It appears that, in addition to improperly using the rudder to prevent a roll from occurring, this pilot actually developed an oscillation similar to what is called a “Dutch roll” – where the yaw-roll coupling is excited. He then managed to exacerbate this oscillation and continued to apply, and then reverse apply, maximum (or nearly maximum) roll and yaw control inputs until the tail structure failed.

You are not alone in trying, but failing, to know “why” this pilot did what he did. Given that this pilot was certainly not a novice pilot, I would presume that the “why” included the fact that he thought he was doing the right thing to prevent what he thought was happening, or was about to happen, to his aircraft. He had just encountered wing-tip vortices from the preceding aircraft (the JAL B-747), and I presume he was responding to a second encounter. Further, it appears that he thought the proper control application strategy included the aggressive use of rudder application. Normally, a pilot is trained to use the rudder to keep the airplane in a “coordinated” flight condition. The instrument used to verify the need for such control, as well as indicating the direction and the magnitude of that control input, is the inclinometer, commonly called the “slip/skid” ball. The idea is to “keep the ball in the center,” and by doing so, keep the airplane in “coordinated” flight; that is, not skidding and not slipping. I’m not familiar with the A-300 so I don’t know if there is an electronic version of the inclinometer that was recorded as part of the FDR, or if there is only a simple, mechanical inclinometer available for pilot reference. I only know that the animation included vertical and lateral acceleration indications, which are not necessarily the same.

There was some testimony/evidence that was presented during the investigation of this accident that indicated this pilot may have been trained to use the rudder during other situations – like the recovery from “other than normal” attitudes. However, the evidence and testimony also shows that this training always included references to “minimal use” or “smooth, coordinated use” of any rudder input. Contrary evidence indicated that at least once, earlier in this pilot’s career, he may have used essentially the same control inputs (i.e., full or almost full rudder application) to, as he was reported to have said, “level the wings due to wake turbulence.” Of course, rudder deflection will cause a rolling moment, due to the sideslip that will develop (in fact, in some airplanes, particularly swept-wing airplanes, a competent aerobatic trained pilot may be able to “roll” the aircraft completely through 360 degrees, using only the rudder), but the rudder is not a roll-control device; and no one seems to know where this pilot may have picked up this particular control application impression.

I know very little about boats of any size – other than the fact that they are supposed to float – but your analogy and your conclusions seem to be pretty well on target. As you say, “you might well break the rudder off in a dinghy if you tried this.” And, it would seem that this situation verifies your concerns.

[bomarc]

dirty dingo

the question was only about the a300(a310) series of planes and not the a320 and later. that is a whole other discussion.

some people feel that this is a boeing vs airbus thing. it is not intended to be. I have no great fondness for the b 737 for example.
---
rainboe: I think that during a violent wake turbulence encounter the captain would be even more in tune with what the other pilot was doing...and if it all in disagreement with the copilots actions would use the time honored phrase: "I've got it".

As far as being hit with a newspaper...it is not one of my preferred instructional methods.

There does seem to be a valid point you make about knowing about over controlling a plane. No where in my 30 years of flying has anyone mentioned breaking apart an airplane by over controlling (referencing safe operating speeds) unless a specific limitation printed in the aircraft flight manual (sometimes called POH).

I have searched the flight manuals of 4 jets that I have flown and nothing about breaking the thing apart.

Now, perhaps you were taught something some of us weren't. Time to share the knowledge.
Perhaps we were taught something you don't know.

Learning to fly is still a bit of an art and not just a pure science.

==
bubbers 44, I have to agree with you.
-----
as far as using the rudder to "roll" the plane...I have used the rudder in a very gentle manner to initiate a slight bank at cruising speeds in a jet. It seems to work fine...but of course the ailerons/spoilers are the intended control for roll. rudder is the intended control for yaw...but yaw a plane and it might just bank.
;-)

[Rainboe]

From before I started passenger flying, I was taught in BOAC to 'go easy on the rudder', it is for engine failure, not for flying the plane. I was taught it at flying school- always be ginger on the rudder, it's not stressed to take misuse. Frankly it is so ingrained I cannot conceive of people using it like in this example. I have been caught flying along a 747 vortex after Canarsie on a Canarsie approach in a VC10. It took most aileron to hold it- I didn't touch the rudder. A pilot who thinks he can apply rudder control as in this case has controlled way outside design parameters and I consider has had a significant part of his flying education and experience missed or ignored. I'm sorry to be brutal, but no pilot flying a passenger jet should have missed this point. I know all my colleagues at a big British Airline have been brought up with this. We have had significant jet upsets for various reasons. All have apparently flown and recovered correctly and 'let it go' rather than bully the thing with overcontrol to get it to do what they desire. I think constantly trying to defend the indefensible and plead ignorance will not instill confidence in the flying public.

[cornwallis]

I have flown almost all the airbus line apart from the 340.I have also flown the 737/757 and the nicest aeroplane I have ever hand flown is the 300.It has a good roll rate and is light but I have never used the rudder in normal flight apart from engine failures.The roll spoilers and aileron are more than enough to pick up a dropped wing.I always felt that the 320 family could be a bit slow in this.
Most instruction I have seen about recovering from upsets advised to roll to the blue-it never advocated rudder inputs.

[bomarc]

dp davies admonishes us not to use the rudder in certain turbulence conditions.

however we must ask ourselves about being taught to use rudder in certain situations as part of a recovery. indeed, since the tragic crashes in pittsburgh and colorado springs involving the 737 hardover, new methods have been introduced...these started about 1995 to my memory.

if one limits the use of rudder to engine out and crosswind landings, leaving it on yaw damper for the express purpose of countering "dutch roll" fine.

but it is not being taught that way...and no specific mention of NOT using the rudder is being made at many large and respected airlines.

indeed in certain wake turbulence encounters, rudder and aileron/spoilers are part of the equation, at least according to some schools of thought at some major airlines in the us.


may I suggest, what has been suggested before, that since American's primary aircraft is the MD80 (super 80) a mindset might have evolved that what was good for the MD80 was good for everything else. And perhaps it is not so.


Indeed, for the record, since 1935 in the US, Rudder has been a vital part of the controls. It is said that a CAB test pilot took off in a plane that had the ailerons rigged backwards and used the rudder to land and since that time the rudder was part of the equation.

in 1982 my ailerons failed inflight on a small aircraft and rudder is what I used to land the plane safely.

We have all learned in different ways how to fly, how our decision making process is influenced by how and what we have learned in the past.


good luck to us all

[Casper]

Has it been officially proved (and beyond doubt) that rudder hardover was a factor in Colorado Springs and Pittsburgh?

[JustAnothrWindScreen]

Originally Posted by bomarc

dear rainboe:
would the airbus 300 series be your first choice in planes to fly? you may choose from any plane currently flying, comparing it to the above plane?
hypothetical of course.

Bomarc,

Hypothetically, I could have bid the B777/A300/B767/B757. Before this accident, I really had no problem with any of them. After the accident and talking with numerous A300 drivers, my choice went to the B767/B777 and bypassed the A300, hypothetically speaking of course. Airbus/Boeing was never one of my criteria, only rosters and QOL. I simply dropped the A300 from consideration.

[bubbers44]

Rainboe, We will never know for sure if Sten put the control inputs in that caused the failure or if it was already happening and he was just trying to keep it flying straight. I still don't think the captain would have allowed him to overcontrol to the point of structural damage for wake turbulence. I'm done.

[AirRabbit]

Originally Posted by bubbers44

Rainboe, We will never know for sure if Sten put the control inputs in that caused the failure or if it was already happening and he was just trying to keep it flying straight. I still don't think the captain would have allowed him to overcontrol to the point of structural damage for wake turbulence. I'm done.

Two points - once again.

1) According to the FDR, the cockpit controls were not displaced in an effort to counter a control surface movement or to counter a vertical or lateral acceleration. The A-300 cockpit controllers (column, wheel, pedals) are not reversable controls. If the cockpit controllers were moved (and according to the FDR all three were moved - rather substantially and rather briskly), someone in the cockpit with access to the respective control moved it.

2) The entire episode, from the first controller movement to the end of the data feed was just under 9 seconds in duration. To say that the Captain, or anyone else for that matter, should have been able to observe what was happening, evaluate the cause, determine what was being responded to and how, recognize the adequacy or inadequacy of what was being done, decide on a course of action, and initiate that course of action - and do all that cold, with zero warning, and do it all successfully - is a mighty tall order for a mere human to grasp. I might point out that it probably took you 7 or 8 seconds to read that last sentence.

[misd-agin]

The A300 is a nice flying a/c. The accident made us aware of how sensitive the rudder is.

My comments using numbers (1-100) is just a subjective opinion to perhaps expand the understanding of what increased speed has on rudder inputs and is not based on engineering data.

To fly, or not fly, the A300 for my personally has nothing to do with the accident. My experience is that Boeing products(757/767/777) are more reliable than the A300 but if my schedule was better on the A300 I'd fly it.

I believe the AAMP program developer was on the 757/767 when the program was developed. Numerous profiles were flown, by qualified pilots in the different a/c AA flew, so the AAMP wasn't developed using the S-80 as the baseline.