Dalmd88
Topic Author
Posts: 2412
Joined: Fri Jul 28, 2000 3:19 am

American FLt 587 Findings

Tue Oct 26, 2004 5:24 am

Here's the link
http://biz.yahoo.com/ap/041025/ntsb_flight_587_1.html

The NTSB found the tail broke off becuse the copilot used full rudder to attempt to counteract a roll. AB in it's defense stated, we told them not to train their pilots to do that.

It is amazing any manufacturer could allow a plane to fly with such a known defect. Would you buy a car if the salesman said, "Whatever you do don't move the steering wheel real quickly all the way to the end. The wheels might break off."
 
Scorpio
Posts: 4797
Joined: Thu Oct 04, 2001 3:48 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 5:28 am

Has it ever occurred to you that if full rudder was repeatedly applied in the same way on a Boeing, the tail of the Boeing would quite likely break off as well?
 
airtran737
Posts: 3222
Joined: Mon Apr 05, 2004 3:47 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 5:29 am

Tommorow should be interesting. AA vs. Airbus slugging it out.
Nice Trip Report!!! Great Pics, thanks for posting!!!! B747Forever
 
trident2e
Posts: 1286
Joined: Sun Dec 28, 2003 3:38 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 5:31 am

Before our American friends turn this into an Airbus bashing thread, bear in mind that Boeing has formally notified operators of their aircraft that crews must not use excessive rudder or their tailplanes will fall off too. So Dalmd88 - would you buy an aircraft from Mr Boeing if he told you not to use too much rudder?
 
NIKV69
Posts: 10893
Joined: Wed Jan 28, 2004 4:27 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 5:37 am

Scoripo,

I am sure in one or more Boeing aircraft a piot used full rudder to regain control, no tail ever fell off a Boeing. I think the real problem lays with the fact the rudder pedals on that 300 were very sensitive and I also don't like the composite design which is why I feel the tail came off under that strain. I also feel Airbus didn't inform AA about this potential problem. Great article in my local paper today about AA Flight 903, Reg N90070. On May 12th, 1997 the pilots lost control of the plane on a flight to MIA after it stalled and the pilots used the rudder to regain control and nearly lost the tail the same way 587 did. They landed safely. Same model Airbus.
Hey that guy with the private jet can bail us out! Why? HE CAN AFFORD IT!
 
SATL382G
Posts: 2679
Joined: Sat Aug 14, 2004 12:02 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 5:54 am

I can think of a couple of Boeings, a B-52 and a 707, that lost their vertical stabs. They didn't crash though....

[Edited 2004-10-25 22:58:45]
"There’s nothing quite as exhilarating as being shot at and missed" --Winston Churchill
 
col
Posts: 1692
Joined: Thu Nov 27, 2003 2:11 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 5:59 am

Don't think any manufacturer would advise to use full rudder in Vortex conditions, at those speeds. So lets be real about the integrity of the aeroplane.

The issue seems to be of communication, either Airbus to customers, or American to pilots.
 
AAR90
Posts: 3140
Joined: Fri Jan 21, 2000 11:51 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:01 am

The NTSB found the tail broke off becuse the copilot used full rudder to attempt to counteract a roll. AB in it's defense stated, we told them not to train their pilots to do that.

Wrong on both counts. Recommend reading a bit more closely.

It is amazing any manufacturer could allow a plane to fly with such a known defect.

Your original premis is wrong, so your statement is also wrong.

...bear in mind that Boeing has formally notified operators of their aircraft that crews must not use excessive rudder or their tailplanes will fall off too.

Another inaccurate statement.

Before our American friends turn this into an Airbus bashing thread...

Before anybody goes bashing AA, Airbus or anybody else I recommend y'all do a bit more research. The above postings indicate folks forming opinions without facts or knowledge. A good beginning on this site would be:
http://www.airliners.net/discussions/general_aviation/search.main?search_table=archived_general_aviation&search_field=topic&search_active=1&search_year=&search=587&search_order=id&submit=Search+Forum%21

*NO CARRIER* -- A Naval Aviator's worst nightmare!
 
anstar
Posts: 2871
Joined: Sun Nov 23, 2003 3:49 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:09 am

NIKV69, Are these the same composites the 7e7 will use?

From what I see, Airbus warned about rapid rudder movements.

AA claim that they didn't know the tail would snap off.

So, Why would you still train pilots to use loads of rudder if you were told it was dangerous. Who cares if Airbus didn't say the tail would snap off, thery said it was dangerous, so take heed.

Now, the article also states that the rudder is very sensitive. Ummm, wouldn't A300 pilots already be aware of this when using the rudder?
 
MissedApproach
Posts: 678
Joined: Thu Oct 14, 2004 10:12 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:20 am

United 585 & USAir 427, 737s, both had full rudder deflection without the fin breaking off. Unfortunately the rudder deflection was uncommanded due to a flaw in the yaw damper design & caused crashes in both cases, but at least the structure was sound.
While there are tests for metal fatigue, does anyone have a fatigue test for these composite structures that are proliferating in aircraft design now? I'm sure that by the time you have a visual delamination on these things it's already too late.
Oh boy, another A vs B thread!  Laugh out loud
Can you hear me now?
 
Scorpio
Posts: 4797
Joined: Thu Oct 04, 2001 3:48 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:32 am

I am sure in one or more Boeing aircraft a piot used full rudder to regain control, no tail ever fell off a Boeing.

So can you provide us with an instance where the pilot of a Boeing aircraft repeatedly used FULL rudder to regain control, in the same way as on AA587?

and I also don't like the composite design which is why I feel the tail came off under that strain.

Just because you 'feel' that doesn't make it so. The rudder withstood more than it was designed to withstand before it fell off. And it was designed to withstand just as much strain as any other aircraft. Otherwise it wouldn't have have received its certification.

Great article in my local paper today about AA Flight 903, Reg N90070. On May 12th, 1997 the pilots lost control of the plane on a flight to MIA after it stalled and the pilots used the rudder to regain control and nearly lost the tail the same way 587 did. They landed safely. Same model Airbus.

And, more importantly, same airline. Any stories of other airlines using full rudder on their A300s?
 
NIKV69
Posts: 10893
Joined: Wed Jan 28, 2004 4:27 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:39 am

SATL382G,

I thought an aircraft can not fly without a tail. Are you sure they lost their entire tail?

As for Airbus telling AA about the rudder defect, I strongly doubt they did. I find it hard to believe Captains would use hard rudder in any situation if they knew it could mean death to the entire aircraft.
Hey that guy with the private jet can bail us out! Why? HE CAN AFFORD IT!
 
anstar
Posts: 2871
Joined: Sun Nov 23, 2003 3:49 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:41 am

NIKV69, um AA587 didn't fly... hence it fell to the ground.

As the article says Airbus did warn AA of rudder issue, just not to the extent of the tail snapping off.
 
NIKV69
Posts: 10893
Joined: Wed Jan 28, 2004 4:27 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:48 am

ANstar,

No Kidding, ready reply #5 please. Then you will understand.
Hey that guy with the private jet can bail us out! Why? HE CAN AFFORD IT!
 
FedExDC-10
Posts: 183
Joined: Tue Aug 29, 2000 7:53 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:49 am

The article states that American is the only airline that flies this aircraft, and I'm rather curious about that. I can think of other A306 operators. Am I missing something?

FedExDC-10
 
SPREE34
Posts: 1567
Joined: Tue Jun 29, 2004 6:09 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:54 am

The Boeing examples of tailess flight did not have the added fuel storage, thus weight and balance issue the A300-600 was faced with. If fuel was in the tail at the time of the incident, the aircraft would has become instantly nose heavy, that does not usually lead to a climb.

The article I have read incorrectly states that only American operates "this type". Correct me if wrong, but I recall FedEx, and Monarch also fly/flew this, possibly some Asian carriers.
I don't understand everything I don't know about this.
 
Dalmd88
Topic Author
Posts: 2412
Joined: Fri Jul 28, 2000 3:19 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 6:57 am

Full rudder defection even if repeated should not rip the tail off. This is on ANY airplane. I don't care who makes it. I'm not real keen on composites they tend to fail real ugly when they do fail. Yes I know Boeings use a lot of composite parts now also and we might very well see some similar failures in the future on Boeing aircraft.

My main thought on this thread is the only fix they had was "Don't do that, bad things could happen." Why didn't they fix the vertical stab so this couldn't happen?
 
spacecadet
Posts: 2807
Joined: Thu Sep 20, 2001 3:36 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 7:00 am

Tommorow should be interesting. AA vs. Airbus slugging it out.

No, tomorrow will be the NTSB releasing its findings. AA and Airbus have had their chance to make their feelings known.

United 585 & USAir 427, 737s, both had full rudder deflection without the fin breaking off.

The issue is not full rudder deflection. The issue is full rudder deflection from one direction to another. Neither of the two airplanes you mention had that happen.

It is amazing any manufacturer could allow a plane to fly with such a known defect. Would you buy a car if the salesman said, "Whatever you do don't move the steering wheel real quickly all the way to the end. The wheels might break off."

Would you buy a car if the salesman told you that turning the wheel fully from one direction to the other at 70mph might cause it to flip over and crash? Probably not, but millions of cars and trucks out there fit that description and lots of people buy them...
I'm tired of being a wanna-be league bowler. I wanna be a league bowler!
 
MissedApproach
Posts: 678
Joined: Thu Oct 14, 2004 10:12 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 7:09 am

Correct on other A306 operators, AA was the launch customer for the R model. FedEx, Thai, & Monarch also operate it. The columnist of the article was obviously mistaken.
I wouldn't go as far as to say I don't like composites since their light weight goes a long way to improving aircraft performance, but does anyone know what sort of lifespan a composite structure has on an airplane?
Military planes have been using composites for awhile, have there been any failures in these applications?
Can you hear me now?
 
SATL382G
Posts: 2679
Joined: Sat Aug 14, 2004 12:02 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 7:27 am

I thought an aircraft can not fly without a tail. Are you sure they lost their entire tail?

Aircraft designed without a tail seem to fly well. The B-52 and A300 left some vertical stab behind. I'm not sure about the 707.
"There’s nothing quite as exhilarating as being shot at and missed" --Winston Churchill
 
moman
Posts: 708
Joined: Tue Aug 10, 2004 7:17 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 7:30 am

"Would you buy a car if the salesman told you that turning the wheel fully from one direction to the other at 70mph might cause it to flip over and crash? Probably not, but millions of cars and trucks out there fit that description and lots of people buy them..."

I bet every SUV on the American highways would fail this test....

"The Boeing examples of tailess flight did not have the added fuel storage, thus weight and balance issue the A300-600 was faced with. If fuel was in the tail at the time of the incident, the aircraft would has become instantly nose heavy, that does not usually lead to a climb."

Do Airbus jets store fuel in the tail or did I miss the point of this post?

Moman
AA Platinum Member - American Airlines Forever
 
prebennorholm
Posts: 6430
Joined: Tue Mar 21, 2000 6:25 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 8:21 am

Any aerodynamic surface - including the wings - can be ripped off an airliner if improperly flown. It mostly depends upon speed since aerodynamic forces become four times greater as speed doubles.

FDR data has long time been published telling how much the lateral load on AA587 exceeded design and FAA certification limits. That tail stood up to considerably more than it was certified for.

If the A300-600 vertical stab was unsafe, then all examples would of course have been grounded for modification like the DC-10 in 1980 when the "unsafe" - or at least not "fool proof" - baggage door was so dramatically discovered.

The AA587 is an information exchange issue, and/or an AA crew education issue.

That doesn't make the whole thing less serious. Information exchange and education is fully as important as product quality.

To those who bitch over carbon fibres there is only one thing to say: Stay away from any airliner designed less than 30 years ago. Don't even dream about getting near an A380, much less a 7E7 Dreamliner.

Carbon fibres is nothing new. What is fairly new is payable CNC milling of very large moulds making large carbon fibre structures economically viable compared to aluminium. On military planes and spacecrafts, where economy takes the backseat compared to performance optimization, carbon fibres have long time ago been introduced at a much faster pace.

Carbon fibre is a supperior material compared to aluminium in all respects. It is stronger, lighter, stiffer and much less prone to fatigue. The only major disadvange is that it is more expensive.
Always keep your number of landings equal to your number of take-offs
 
2H4
Posts: 7960
Joined: Tue Oct 19, 2004 11:11 pm

RE: American FLt 587 Findings

Tue Oct 26, 2004 8:46 am

Information exchange and education is fully as important as product quality.

One of the best points made thus far, Preben. Well put.


Carbon fibre is a supperior material compared to aluminium in all respects. It is stronger, lighter, stiffer and much less prone to fatigue. The only major disadvange is that it is more expensive.

I think it's important to remember that carbon material can be engineered poorly as easily as it can be engineered well...grouping all prepregs and layup processes together as one may be misleading. Basically, carbon can be the best thing in the world or a brittle, dangerous material, depending on many factors. I would only add the words "properly engineered" carbon fiber to your statement above.

Thanks for yet another great post, Preben. Always a pleasure to read your stuff.


2H4
Intentionally Left Blank
 
prebennorholm
Posts: 6430
Joined: Tue Mar 21, 2000 6:25 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 9:05 am

The B-52, which lost its vertical stabilizer many years ago, was an entirely different thing. It was cruising at high altitude when it encountered wave winds over high mountains (the Rockies?) which put it way above limiting Mach number airspeed. The vertical tail separated due to flutter.

Unlike AA587 it was not exposed to extreme lateral force, and it was not exposed to extreme wake turbulence.

Furthermore not the whole vertical stabilizer was lost. Some 20% of the surface stayed with the plane and it landed safely.

That B-52 was a long time gone B-52D model. The incidence was one of the reasons for the revised and considerably smaller vertical stabilizer on the B-52G and B-52H planes. Today only the -H model is in use.

The AA587 accident could hardly have happened on a B-52. The A300 - like all other airliners - has a generous lateral control ability mainly to make landing in sidewind possible. The B-52 also lands in sidewind, but it has the ability to turn the main landing gear wheels and land slightly sideways with heading pointing directly against the airflow.

This system was mainly made because the extremely low landing gear and long and drooping wings make any sideslip landing completely impossible.

The B-52 has a very small rudder. Powerful lateral control input can only be imposed using the spoilers on the wings, which also perform roll control.
Always keep your number of landings equal to your number of take-offs
 
SATL382G
Posts: 2679
Joined: Sat Aug 14, 2004 12:02 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 9:12 am

Nobody is saying the B-52 lost it's tail for the same reason as AA587. Someone mentioned that a Boeing had never lost a tail and that wasn't true, which I pointed out by mentioning the B-52 and the 707 incident.

Oh and the B-52 involved was an H model....
"There’s nothing quite as exhilarating as being shot at and missed" --Winston Churchill
 
AAR90
Posts: 3140
Joined: Fri Jan 21, 2000 11:51 am

RE: American FLt 587 Findings

Tue Oct 26, 2004 11:10 am

The AA587 is an information exchange issue, and/or an AA crew education issue.

Obviously this (and other inane postings) shows how little folks following this thread actually understand the details of this mishap. The details show AA and Airbus arguing over information exchange and AA pilot training, but neither can be directly contributable to this mishap.

NTSB conclusion is based upon supposition [there is no conclusive proof the FO input the rudder REVERSALS (full rudder deflection itself is not the issue) since the FDR did not record rudder pedal pressures (the only way to know if pilot input control movement or computer/system caused the reversals) and yes, there have been multiple instances of A300 rudder deflection without pilot input reported by numerous airlines]. That supposition is a good logical assumption based upon the available evidence [FWIW, most aviation mishap investigations require/use deductive reasoning based upon at least some logical assumptions].

AA pilot training never taught rudder REVERSALS so AA "crew education" is not a cause factor! Airbus' poor "information exchange" concerned the rudder pedal sensitivity [not rudder REVERSALS] and is also not a cause factor.

Get a grip folks. These are liability/insurance lawyers' arguments y'all are arguing over and have nothing to do with the cause of the mishap!

AAR90
8+ yrs USN aircraft mishap investigator.
*NO CARRIER* -- A Naval Aviator's worst nightmare!
 
User avatar
Starlionblue
Posts: 17117
Joined: Fri Feb 27, 2004 9:54 pm

RE: American FLt 587 Findings

Tue Oct 26, 2004 10:50 pm

I was going to go on a long rant, but Preben made most of my points for me, and much more eloquently than I would have.

It's not about the airframe, it's about who got what information. And we cannot assume that just because Airbus informed AA (under debate of course, but there was that letter, see next section of this post) this information trickled all the way down to line pilots.


AA pilot training never taught rudder REVERSALS so AA "crew education" is not a cause factor!

Maybe not reversals, but there was concern about rudder usage. Boeing, McD and Airbus sent a joint letter to AA warning about overemphasis on rudder in flight upset training. That should, if anything, indicate that this is not an Airbus issue in regards to airframe.

Quoting http://www.iasa.com.au/folders/Safety_Issues/others/AA587hearings.html

The manufacturer points to American's upset training program, saying that, for a time at least, it overemphasized rudder usage in certain situations, including wake turbulence. AA rolled out its Advanced Aircraft Maneuvering Program (AAMP) in 1996 and used it as the template for similar programs.

FAA, Airbus, Boeing and McDonnell Douglas reviewed such programs as part of an industry initiative to develop generic guidance for training. In an August 1997 letter to AA, FAA and the manufacturers complemented AAMP overall but expressed concern with several specifics, including an "excessive emphasis" on using rudder for roll control in high angle-of-attack (nose up/down) situations. The group admitted that "a simple rule about rudder usage cannot be stated" but said that the "more appropriate standard" is to first use full aileron control, then go to the rudder if the plane "is not responding."



Oh and for those who still think the tail was weak:

Again quoting http://www.iasa.com.au/folders/Safety_Issues/others/AA587hearings.html

In an unusually direct opening statement, NTSB kicked off the American Flight 587 hearing yesterday in Washington by all but ruling out a structural flaw in the plane's all-composite tail as the cause of the crash.

"The extensive loads calculations accomplished by Airbus and independently by the safety board revealed that the aerodynamic and internal loads that the vertical stabilizer experienced were significantly above the ultimate loads required by the French and American certification standards," Investigator-In-Charge Robert Benzon told attendees. "In fact, the sustained loads were near the structural test loads demonstrated during the certification process."

Factual reports released yesterday by the board show Flight 587's all-composite tail broke at about 1.97 limit load, while the tail broken by Airbus during A300-600 initial certification tests ago failed at 1.93 limit load. Loosely, limit load is the highest load a primary structure -- such as the tail -- is expected to experience during a plane's service life. The certification standard is 1.5 limit load, referred to as ultimate load, so Flight 587's tail stood up to loads well beyond the certification standard, and as Benzon noted, appeared to fail near the value predicted by the certification testing


The tail on an airliner is certified to handle full rudder deflection and return to neutral, not repeated full deflections in opposing directions.


To quote another article:
Both Airbus Industrie, which manufactured the jetliner, and American Airlines, which trained Molin, agree that if he had taken his foot off the rudder pedal, the tail wouldn't have broken off, the plane wouldn't have plunged into a New York City neighborhood and 265 people wouldn't have died.

But Molin didn't know he was putting more pressure on the tail than it could bear. Why he didn't — and who's to blame for that — is the subject of a bitter fight between Airbus and American.


And that's the core of the issue IMHO.


[Edited 2004-10-26 15:54:58]
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
AAR90
Posts: 3140
Joined: Fri Jan 21, 2000 11:51 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 12:04 am

Maybe not reversals, but there was concern about rudder usage. Boeing, McD and Airbus sent a joint letter to AA warning about overemphasis on rudder in flight upset training. That should, if anything, indicate that this is not an Airbus issue in regards to airframe.

For anybody who understands the basics of aircraft mishap investigations, the Airbus design was ruled out very early on in the investigation. There were some small side issues (primarily about how to properly inspect the composite structures involved), but not the design and certainly not how that design performed under this specific situation.

As to the "concern" about rudder usage:

The manufacturer points to American's upset training program, saying that, for a time at least, it overemphasized rudder usage in certain situations, including wake turbulence. AA rolled out its Advanced Aircraft Maneuvering Program (AAMP) in 1996 and used it as the template for similar programs.

FAA, Airbus, Boeing and McDonnell Douglas reviewed such programs as part of an industry initiative to develop generic guidance for training.


These folks reviewed AA's AAMP training because AA asked them to! AAMP was the FIRST program of its kind so AA asked for manufacturer and FAA input PRIOR to starting line pilot training.

In an August 1997 letter to AA, FAA and the manufacturers complemented AAMP overall but expressed concern with several specifics, including an "excessive emphasis" on using rudder for roll control in high angle-of-attack (nose up/down) situations. The group admitted that "a simple rule about rudder usage cannot be stated" but said that the "more appropriate standard" is to first use full aileron control, then go to the rudder if the plane "is not responding."

Read the above slowly and carefully. The reviewers "expressed concern" was the originally proposed AAMP training statement for "aggressive use of rudder" for roll control "when necessary." Based upon the reviewers' comments, AAMP training as provided to AA line pilots actually said use of "full rudder deflection in one direction only may be necessary if the ailerons do not arrest an uncommanded roll." AAMP was in FULL AGREEMENT with manufacturer's "concerns" PRIOR to being introduced to AA line pilot training. As a result, AAMP became the FAA standard by which all other FAA approved airline training was judged. AAMP training is not a cause factor in this mishap.

But Molin didn't know he was putting more pressure on the tail than it could bear. Why he didn't — and who's to blame for that — is the subject of a bitter fight between Airbus and American.

And that's the core of the issue IMHO.


Somewhat correct (we will never know what Molin knew or didn't know). Your "core of the issue" is the legal liability dispute between Airbus and AA with both sides' legal teams trying desperately to blame someone else (at least partially) for insurance payment reasons. It has nothing to do with the actual cause of the mishap.
*NO CARRIER* -- A Naval Aviator's worst nightmare!
 
User avatar
Starlionblue
Posts: 17117
Joined: Fri Feb 27, 2004 9:54 pm

RE: American FLt 587 Findings

Wed Oct 27, 2004 12:08 am

I think it is rather central to the cause. If Airbus did not inform AA properly about the sensitivity of the rudder, or if AA did not inform it's pilots properly, bears upon how the pilots were trained and used to reacting to upsets. Since the aircraft seems to have been working as it should, pilot reaction to the wake turbulence becomes the central focus.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
cannikin
Posts: 79
Joined: Sat Oct 16, 2004 1:10 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 12:14 am

"...Benzon also said that the rudder control system on the aircraft is sensitive at higher air speeds, which is potentially hazardous..."


Is 200 knots a high airspeed?
 
User avatar
Starlionblue
Posts: 17117
Joined: Fri Feb 27, 2004 9:54 pm

RE: American FLt 587 Findings

Wed Oct 27, 2004 12:21 am

Most commercial jets don't go over about 350KIAS (IIRC) so 200 is not very high but definitely not low. Low would be 160 and under.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
AA717driver
Posts: 1502
Joined: Tue Feb 12, 2002 8:27 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 12:40 am

Correct me if I'm wrong but did a rudder reversal(unknown cause) happen prior to the full rudder input occur?

If it did, the aircraft would have yawed in the direction of initial rudder displacement. If the rudder were then fully displaced in the opposite direction, a "cupping" of the vertical stabilizer with respect to the airflow would occur.

If that is what happened, you can only hang this on the pilot if you can prove he initially displaced the rudder and immediately reversed it. The cupping would place excessive force on the vertical stabilizer at 200kts.

It's my understanding that Boeing tests for the above situation and Airbus does not. (It is not required for certification.) It's encouraging that the stabilizer failed at a point significantly above the required load.

Does the Airbus have a "rudder load limiter"?TC

P.S.--Having been in similar situations several times, I'm glad I added rudder progressively(the DC9/MD80/717 not being known for excessively light rudder control forces!).
FL450, M.85
 
User avatar
Starlionblue
Posts: 17117
Joined: Fri Feb 27, 2004 9:54 pm

RE: American FLt 587 Findings

Wed Oct 27, 2004 12:58 am

Correct me if I'm wrong but did a rudder reversal(unknown cause) happen prior to the full rudder input occur?

Nope. All rudder movements were due to pedal input. There was no rudder reversal due to unknown causes.

The rudder control system is "different" on the A300-600, even in respect of the other Airbi (and earlier A300s). In fact the pedal forces required to unstick the rudder are very low and do not change with speed. Also the pedal travel required to move the rudder to it's maximum extension is rather short and IIRC does not change with speed. On newer Airbi, the FBW control system changes rudder response to pedal inputs with speed (a rudder limited in the usual parlance).

Ok, I may have bollocksed some of that. Feel free to correct as needed.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
phollingsworth
Posts: 635
Joined: Sat Mar 06, 2004 6:05 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 1:16 am

The rudder control system is "different" on the A300-600, even in respect of the other Airbi (and earlier A300s). In fact the pedal forces required to unstick the rudder are very low and do not change with speed. Also the pedal travel required to move the rudder to it's maximum extension is rather short and IIRC does not change with speed. On newer Airbi, the FBW control system changes rudder response to pedal inputs with speed (a rudder limited in the usual parlance).

All Airbuses have pedal movement limiting systems. This means that as the airspeed increases the amount of movement from center to full stop decreases. This is also the case on the A340-5/600s, which are the first Airbuses to have fully FBW rudder controls.

Newer Boeings have variable moment arm systems. Such that the amount of rudder deflection per pedal input decreases as airspeed increases, on the 777 this is done entirely in software.

The A300-600 has a higher deflection to pedal input force sensitivity; however, I don't believe the centerline breakaway force is lowered. This is why it only takes 10 lbs. more force to reach full stop than it does to breakaway from the center at 200 kts.

This extra sensitivity increased the chances of multiple full control inputs. However, the pilots should have never, I repeat, never been chasing the a/c around the sky with the rudder. The response time constant on modern commercial a/c to rudder inputs is way to high for precision flying.


Full rudder defection even if repeated should not rip the tail off. This is on ANY airplane. I don't care who makes it. I'm not real keen on composites they tend to fail real ugly when they do fail. Yes I know Boeings use a lot of composite parts now also and we might very well see some similar failures in the future on Boeing aircraft.

Aircraft are designed for very simple combinations of control inputs, they are not designed for doublets, triplets, etc. You can easily rip the wings off many a higher performance aircraft by pulling up in a steep turn. Remember as you keep switching directions the total load increases. Further, it takes a finite amount of time for stress to flow; therefore, dynamic loads can easily increase the local stress far beyond the failure point. Designing an aircraft for all 100% of possible input combinations would make it non-viable. Training pilots not to do certain things makes much more sense (in FBW a/c you can prohibit certain combinations of inputs).
 
User avatar
Starlionblue
Posts: 17117
Joined: Fri Feb 27, 2004 9:54 pm

RE: American FLt 587 Findings

Wed Oct 27, 2004 1:22 am

Phollingsworth, as usual your input is much appreciated, at least by this A.nutter.

As you say, you can only do so much to protect the vehicle from the driver. While modern technology (ABS, four wheel drive and traction control in cars, FBW controls in planes) will do much to decrease the risk, the driver can still run an Audi A6 with all the electronics in the world into a ditch, or turn a fully fucntional A320 into a very expensive woodchipper.

Ok that was a long sentence...  Big grin
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
spacecadet
Posts: 2807
Joined: Thu Sep 20, 2001 3:36 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 1:40 am

Is 200 knots a high airspeed?

Accident occured at 240 knots from what I'm hearing now. I'm still watching the NTSB webcast right now (it's still going on).

240 knots is a very high airspeed for rudder usage of any kind. Rudders in airliners are generally not used except in engine out conditions, to correct for crosswinds on landing and takeoff, or in extreme cases to correct upsets when there's no other choice (as happened in AA flight 903 in 1997, which was referenced in this meeting several times).

The infamous Airbus memo was brought up - this memo was written after flight 903. The NTSB says this was not directly applicable to flight 587 because the circumstances of the rudder usage were different. Still, they're concerned about the lack of communication. They're going to add some language to the accident report about it, but say it was not a contributing factor to the flight 587 accident.

Boeing was also brought up, and the structure guy was asked specifically if Boeing does anything different than Airbus does with their vertical stabilizers. Flat answer: "No." He had to clarify a bit because Boeing does have some internal standards that Airbus does not, but he stated that while the FAA regulation requires tails to withstand ultimate loads to a factor of 1.5, Boeing's standards do not, and their additional criteria would fit within the FAA regulation (in other words, the tail would withstand an ultimate load factor of 1.5 to the FAA regulation, and a factor of 1.0 to Boeing's standards - or the same overall strength as Airbus' designs).

The rudder on Flight 587 was reversed not just once, as I had initially assumed - it was reversed twice (or side to side four total times). The rudder pedals went to full stop in both directions four times in rapid succession - each time beyond the mechanical limit. I don't care what sort of training AA has, no pilot would ever do this. I'll bring up my earlier analogy again - would any driver swing the steering wheel of his car hard from one side to another four times in a row at highway speeds? That's similar to what we're talking about here.

They're talking now about strengthening certification requirements for vertical stabilizers, but apparently this is not in the accident report. Structure guy is arguing against it; says "there's an old saying, if you can design it, I can break it". Tail exceeded regulations, implies that it's basically impossible to make an airplane pilot-proof. But says there may be additional things that can be done; I'd guess you'll see some recommendations come out after the report. But the cause of this accident itself was not the result of any weakness in the tail; the tail actually broke at about 20% higher load than the maximum it was designed to withstand (about 60% higher load than the load limit, or the largest load you could expect to occur over the airplane's lifespan).
I'm tired of being a wanna-be league bowler. I wanna be a league bowler!
 
miamiair
Posts: 4249
Joined: Thu Jul 15, 2004 9:42 pm

RE: American FLt 587 Findings

Wed Oct 27, 2004 1:46 am

American's point is that cause was "the onset of a design induced, adverse Aircraft-Pilot Coupling." That the co-pilot's input was part of the process, not the casue. That there have been three other instances of problems with rudder inputs (AA903, AF825 & Interflug A310) and Airbus did not "share" that information with the operators.
Airbus' point is that AA's training procedures were not driven to the aircraft design and operating conditions during Upset Recovery Training. According to Airbus, AA modified their A300 simulator without consulting with the FAA or Airbus.
(the above was referenced from AW&ST 25OCT04)
As a pilot and an operator, I would want to know all there is to know about how my airplane flies, normally and to a point, abnormally. I don't believe the ailerons were used to counter the roll, but immediately the pilot booted the rudder; with results we all saw and read about. I believe both American and Airbus have valid points in their arguments, and both should shoulder the responsibility and liability.
Molon Labe - Proud member of SMASH
 
gigneil
Posts: 14133
Joined: Fri Nov 08, 2002 10:25 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 2:25 am

I also feel Airbus didn't inform AA about this potential problem.

How can you feel anything? They either did, or didn't, and you're not in a position to know one way or the other.

Airbus said they did, and you have no reason to believe otherwise except personal issues.

no pilot would ever do this.

This is really what it comes down to, isn't it. Training or no training, no pilot should ever do what these pilots, according to the NTSB report, did.

N
 
User avatar
Starlionblue
Posts: 17117
Joined: Fri Feb 27, 2004 9:54 pm

RE: American FLt 587 Findings

Wed Oct 27, 2004 2:27 am

There is also the point that "Airbus informing AA" does not really cover very much. After all, AA is pretty large. Even if someone at AA had the information, this does not mean that the flying pilots had it.
"There are no stupid questions, but there are a lot of inquisitive idiots." - John Ringo
 
westindian425
Posts: 729
Joined: Sat May 15, 2004 7:46 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 2:41 am

I'm going to add Preben and Phollongsworth to my respected users list.

I'll add 2 things:

First, it's very comforting to know that the airplane is safe (remember the big issue about AA A300 pilots wanting to switch to another aircraft?) and that American Airlines didn't panic and ground all A300's or some other drastic measure.

Second (a little nit-picking), the article I read mentioned that American Airlines is the only US carrier to operate the Airbus A300-600R for passenger service. This statement is correct. Monarch is not a US carrier, and FedEx doesn't operate any aircraft for passenger service.


View Large View Medium
Click here for bigger photo!

Photo © T Silgrim
View Large View Medium
Click here for bigger photo!

Photo © Yo Tanaka



Neil
God did not create aircraft pilots to be on the ground
 
elwood64151
Posts: 2410
Joined: Fri Feb 08, 2002 10:22 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 2:42 am

The issue seems to be of communication, either Airbus to customers, or American to pilots.

I think we can ignore just about every statement made after this one. Thanks, Col. Simple deduction, but brilliant compared to some others here.
Those who fail to learn history are doomed to repeat it in summer school.
 
N766UA
Posts: 7843
Joined: Thu Jul 29, 1999 3:50 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 3:03 am

An aircraft should not be manufactured in such a way that the entire tail can become detatched should full rudder be applied. That's like saying the wings will break off if full aileron is applied. In a Cessna, for example, you can kick the rudder pedals to the stops back and forth repeatedly and nothing of structural importance will happen. It should be that way with large airliners as well and to allow the airplane to fly with such an "intolerance" is wrong.
This Website Censors Me
 
Scorpio
Posts: 4797
Joined: Thu Oct 04, 2001 3:48 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 3:29 am

Some people just don't seem to get it...

N766UA,

In a Cessna, for example, you can kick the rudder pedals to the stops back and forth repeatedly and nothing of structural importance will happen.

Does a Cessna fly at 240 knots? I mean, seriously, are you saying you're comparing a CESSNA to a commercial airliner??????????

It should be that way with large airliners as well and to allow the airplane to fly with such an "intolerance" is wrong.

So I guess, in your opinion, not a single large airliner has been designed to date that was safe? Not by Boeing, Airbus, McDD, Lockheed, Ilyushin, Tupolev,... Is that what you're saying?
 
commander_rabb
Posts: 723
Joined: Sat Feb 26, 2000 6:59 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 3:47 am

What struck me is that Airbus said high speed rudder deflections could cause seperation of the tail from the aircraft.

Climb speed was under 200 kts.

What is the Airbus definition of high speed? Was this warning given in the POH and was it clearly spelled out? Look for Airbus also to share a small blame in this accident too for not fully making it's only American customer of it's A300-600 product aware of this loss of rudder potential.





 
AAR90
Posts: 3140
Joined: Fri Jan 21, 2000 11:51 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 3:54 am

Correct me if I'm wrong but did a rudder reversal(unknown cause) happen prior to the full rudder input occur?

Yes, but only to a barely readable amount on the FDR. There would have been nothing noticable in the cockpit (no yaw observed on FDR to these very slight rudder movements).

Nope. All rudder movements were due to pedal input. There was no rudder reversal due to unknown causes.

Correction: the assumption is that the rudder moved due to pedal inputs, but there is no proof. This is the logical assumption any decent investigator would use when no physical defect can be found in the recovered parts of the aircraft. There have been multiple documented cases of uncommanded rudder movements in A300-600R aircraft well prior to this mishap. Cause was never determined and in none of those cases was the rudder fully displaced. So again, the logical assumption is that the pilot(s) input the full rudder displacements as well as the full rudder reversals... but it remains an assumption, not a "fact."

I think it is rather central to the cause. If Airbus did not inform AA properly about the sensitivity of the rudder, or if AA did not inform it's pilots properly, bears upon how the pilots were trained and used to reacting to upsets.

Pedal sensitivity is not an issue when the training is to place up to full rudder deflection in one direction only in order to arrest an uncommanded roll that the ailerons won't arrest... and then release the rudder pressure. Sensitivity would pay a role if the pilot was adding pressure in the opposite direction, but that has never been in any AA training.

Since the aircraft seems to have been working as it should, pilot reaction to the wake turbulence becomes the central focus.

Correct! The assumption being the pilot(s) input full rudder reversals in direct violation of AA training! The real question is: WHY WOULD THEY DO THAT?

According to Airbus, AA modified their A300 simulator without consulting with the FAA or Airbus.

That is an outright fabrication! AA did "modify" the A300 simulator (with FAA approval). The "modification" was to install a single instructor pushbutton switch and software so that when the switch was pushed, the simulator would simulate an uncommanded roll to a near inverted position. The pilot(s) must attempt to regain control. No flight control logic or laws were changed. Just a simple situation input very similar to windshear encounters are input.

There is also the point that "Airbus informing AA" does not really cover very much. After all, AA is pretty large. Even if someone at AA had the information, this does not mean that the flying pilots had it.

That would be a legal argument made by insurance lawyers... and an excuse! IF AA had the information, it is AA's responsibility to properly distribute that information to its employees.

An aircraft should not be manufactured in such a way that the entire tail can become detatched should full rudder be applied.

Airliner certification (in very simple terms) is for full deflection of a flight control surface to its maximum and then released with no adverse affect on structure or flight characteristics. A300 rudder very easily met this requirement.

This is really what it comes down to, isn't it. Training or no training, no
pilot should ever do what these pilots, according to the NTSB report, did.


At least somebody is starting to figure out the obvious.  Wink/being sarcastic
*NO CARRIER* -- A Naval Aviator's worst nightmare!
 
Udo
Posts: 4288
Joined: Tue May 18, 1999 5:16 pm

RE: American FLt 587 Findings

Wed Oct 27, 2004 3:55 am

Commander,

The "A300-600" product is also heavily used by FedEx and UPS with a total of 83 units in service.


Regards
Udo
Me & You & a Plane Named Blue...
 
EnviroTO
Posts: 723
Joined: Sat Aug 07, 2004 12:11 pm

RE: American FLt 587 Findings

Wed Oct 27, 2004 4:01 am

Regardless of whether or not the pilot did the wrong thing the aircraft should not rip apart unless it crashes first or the pilot was given some sort of feedback that design tolerances were being reached. Aerospace engineers are better able to understand all the forces and tolerances on an aircraft and through design a limit should never be reached without the pilot purposely disregarding the feedback the pilot received from the aircraft. If a pilot is able to apply full rudder multiple times easily but rip the tail off in the process without warning then the feedback to the pilot is unreasonable. A pilot should be pushing or pulling with significant strength in order to put significant stress on an aircraft. As pilots become more disconnected from stress on the aircraft through hydraulic assist or fly-by-wire we can expect more of these problems to surface unless engineers fill in the gap with software applied limits, force feedback, or other devices that limit controls from being pushed into positions beyond the design limits of the aircraft.

All the analogies regarding SUVs are flawed... even if an SUV rolls over the wheels usually stay attached to the vehicle. This is not about the aircraft stopping the pilot from flying the plane into the ground, inverting, or stalling the aircraft, the problem is that the rudder pedals on the aircraft at that speed are basically a self-destruct buttons... the pilot doesn't use much strength to push the pedals down but the aircraft itself translates this input into a much stronger force that rips the tail off.

Imagine if pilots were told that pulling the control stick back 10 degrees over 200kts would rip the wings off but the pilots receive little or no resistance pulling the control stick back that far and nothing prevents that motion. Would that be a reasonable aircraft design? Pilots should be able to expect significant resistance to a control's motion when applying dangerous forces to the aircraft or be prevented from applying these forces altogether.
 
widebody
Posts: 1107
Joined: Wed Aug 02, 2000 5:08 pm

RE: American FLt 587 Findings

Wed Oct 27, 2004 4:03 am

If you fly any aircraft outside its certificated envelope you can expect trouble no matter what the aircraft. If the aircraft is certified to fly, the aircraft is certified to fly. Egyptair 990 ripped apart in-flight because of excessive inputs. AA587 did the same. Not comparing, just clarifying what excessive and rapid movements can do to an aircraft.

This accident is about communication, within Airbus, within AA, and between the two. And maybe most importantly, with the FAA.
 
fspilot747
Posts: 3455
Joined: Mon Nov 01, 1999 2:58 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 4:03 am

"In a Cessna, for example, you can kick the rudder pedals to the stops back and forth repeatedly and nothing of structural importance will happen"

I sincerely hope you're just having a bad day. You're a pilot, and that was a pretty misinformed statement. Go out flying at Vne and start kicking the rudder hard left right--full swings, and start doing some steep turns. You will have structural damage. All airplanes have their limits.


FSP

[Edited 2004-10-26 21:04:16]
 
AA777
Posts: 2358
Joined: Thu May 20, 1999 7:07 am

RE: American FLt 587 Findings

Wed Oct 27, 2004 4:04 am

Im sorry, but if the plane was travelling at 200 kts, and the pilot applied full rudder, the airplane's rudder shouldnt have SNAPPED off in response. The plane had just taken off minutes before, the airspeed should have been less than 250 kts, and it seems that it was travelling at 200 kts. If that is the case, then Airbus clearly has a flawed design. Rudders shouldnt just snap off, even if there is a lot of force on it.... make it STRONGER, fix it, do what you have to do, to prevent terrible accidents like these.

-AA777