PDA

View Full Version : Fin & Rudder strength: facts about what is NOT tested


YouNeverStopLearning
17th Nov 2002, 20:46
A lot of rubbish in the form of conjecture has been written here recently about Fin and Rudder strength and airworthiness requirements. So here is a fact from the UK CAA detailing what IS NOT TESTED for a CofA.


AIC: 86/2002 (Pink 38) 17 October 2002
Title: USE OF RUDDER ON LARGE TRANSPORT AEROPLANES

Paragraph 1.2 contains the most important sentence of this AIC:

ď However, aircraft are NOT designed to withstand application of large rudder angles opposite to that required to sustain the sideslip. This may occur when, for example, full rudder is applied in one direction followed by full application in the opposite direction. Such a manoeuvre MAY RESULT IN STRUCTURAL FAILURE. ď

This AIC contains much good revision about rudder use.

Lu Zuckerman
18th Nov 2002, 02:27
A lot of rubbish in the form of conjecture has been written here recently about Fin and Rudder strength and airworthiness requirements. So here is a fact from the UK CAA detailing what IS NOT TESTED for a CofA.


The FAA and I assume the other certification authorities allow an airframe maker to demonstrate Reliability, Safety and Structural integrity by the use of computer analysis if the actual in-flight testing is either too expensive or too dangerous. This has caused a lot of problems as the actual loads or conditions did not compare with the computer analysis. However in the case of proving structural integrity of the flight surfaces in respect to the fuselage attachments the manufacturers will perform structural bending loads on the wings and tail surfaces to the point of destruction.

If Airbus followed these directives they did run the structural bending loads test on all aerodynamic surfaces.


:cool:

YouNeverStopLearning
18th Nov 2002, 11:43
Re-read my post.

Read the AIC.

This a fact, not a rumour.

Lu Zuckerman
18th Nov 2002, 13:55
Quote:

"AIC: 86/2002 (Pink 38) 17 October 2002
Title: USE OF RUDDER ON LARGE TRANSPORT AEROPLANES

Paragraph 1.2 contains the most important sentence of this AIC:

ď However, aircraft are NOT designed to withstand application of large rudder angles opposite to that required to sustain the sideslip. This may occur when, for example, full rudder is applied in one direction followed by full application in the opposite direction. Such a manoeuvre MAY RESULT IN STRUCTURAL FAILURE. ď

This AIC contains much good revision about rudder use".

First of all I donít have access to the AIC in question. That being said, I believe we are talking about apples and oranges. It would appear to me that your excerpt from the AIC deals with the crash of the A-300 due to loss of the rudder. And, it advises pilots that if they perform certain maneuvers using the rudder to excess they can exceed the design limits of the fin attachment or the fin structure.

What I was addressing was the structural testing that must be performed to determine that there is adequate safety margin in the operation of the aircraft under standard flight conditions and that if the aircraft is flown within those limits there would be no structural failure.

Based on that it can be assumed that on the AA A-300 either the structural safety margins were exceeded or there was a structural defect in the fin.

PickyPerkins
18th Nov 2002, 17:56
YouNeverStopLearning, Lu Zuckerman, you are both right - I can't see any conflict in your positions.

If you keep within the "standard flight conditions" = e.g. FAR 25 = "the envelope", everything's tested and ok with a good safety margin. Outside the envelope you can break the a/c with either the pitch controls or the rudder, and according to a current Boeing document (http://www.ntsb.gov/events/2001/AA587/exhibits/240009.pdf), possibly the roll controls as well.

The above Boeing document also says on p. 10, "Elevators and ailerons are not designed for abrupt reversals fiom a fully displaced position."

Also, on p. 5, ".... control surface inputs are to be in one axis (not in combination), and do not include control input reversal or oscillatory inputs."

So the enveope apparently does not cover abrupt reversals of any controls, nor simultaneous changes in two axes.

Question: Can you deal with wake turbulence without using control reversals or making changes in two axes?

http://home.infi.net/~blueblue/_uimages/pi.gif

YouNeverStopLearning
18th Nov 2002, 23:11
I wasn't "right" about anything. I was just passing on a fact for the benefit of those pilots who may not have known what was not tested. Thanks for the arbitation though!

In answer to your question:
It all depends on factors such as , but not exclusively, skill, experience, IAS, TAS, configuration, Mass, Va, (Vb?), phase of flight (ie proximity to ground!) ...

I have found that quite often it is better to let nature balance itself out for a while before instigating in a control input. NOTE I did not say always.

Little Keepings
19th Nov 2002, 18:27
I have been following all the threads concerning this accident from the beginning. Right at the start John Farley's words of wisdom about the difference between a static as opposed to a dynamic sideslip situation said it all. Since then these perceptive observations seem to have been ignored by all. What a pity it is that everyone seemed to be more interested in promoting their own incomplete theories rather than take notice of a true expert.

arcniz
19th Nov 2002, 21:13
LittleKeepings - so that's the final word, is it? A bit arrogant, perhaps?

Might you provide a pointer to Mr. Farley's wise remarks for those who would care to read it again without a long excavation journey through the archives?

The details of certification and testing - whether true to the issue or not - are but a small part of this complex situation.

Every fool who flies (or does the math) knows that one can harm the structure of an aircraft, large or small, by horsing it around enough to breach some structural limit.

What is so remarkable about the situation with the carbon-tail Airbus is that the margin of safety between normal flight operations and auto-destruct seems to have dwindled to nearly nothing - in the normal, everyday, best case, mind you - as a result of particular design, manufacturing, and control characteristics incorporated into the aircraft.

It is wildly disingenuous to argue that all the current information about the general strength limitations and doubtful long-term maintainability of the carbon fiber tail, the idiosyncracies of the yaw damper / rudder actuators, vagaries of the YD control logic, the high breakout thresholds at the rudder pedals and other as yet unpublic details about the flight controls were visible or widely understood by flight crews, training managers and regulatory officials prior to the crash of AA587.

I find it hard to believe that many (or any) people had previously evaluated those factors correctly in a context similar to the present case.

Some compromise with reality is appropriate in commercial aircraft design. Nobody is arguing for the repeal of gravity or aerodynamics. The issues here have meaning both for he pilots that fly 'em and folks that design 'em. The issues have to do with how the hard physical realities connect through fallible materials to human nature.

The 100-year evolution of aircraft has caused them to be fitted with progressively more sophisticated instrumentation that helps ordinary mortals (with ample flying skills) keep them right side up and pointed in the desired direction under the wildly variable circumstances of commercial aviation. Passenger aircraft have a panopoly of systems that work to increase the safety and health of the passengers and crew. High among these are monitors and indicators that show critical data such as whether the wheels are up, down, or somewhere inbetween, lights and screens that show what is on fire at a given moment, and guidance for a myriad of management details ranging from COFFEE OVERTEMP to GROUND PROXIMITY.

AA587 points to an apparent propensity for the Airbus tail to depart the aircraft during easily initiated and fairly unremarkable manoeuvering.

If tail breaking off is such a readily accessible risk, then surely some effective operational warnings would be appropriate. The fact that warning mechanisms are not part of the A300 design proves that the high risk of engaging this catastrophic failure mode in normal flight conditions was not 'common knowledge', as some argue.

If the situation is really one that a skilled pilot can stumble into on a clear day while launching a flight in a 'perfect' aircraft, then probably the Airbus flight crews need better real-time information about where the aircraft is in the envelope, and maybe also an extra interlock or two to help keep them on the warm side of eternity.

Rather than taking the CYA position that the crew should have known intuitively what was happening 40 meters behind them in those few critical seconds, would it not maybe make sense to talk about adding some better information for the guys up front? Perhaps a little indicator that pulls together the already available information about the flight envelope and says in red: "TAIL CRITICAL".

Sadly we cannot rewind the clock. The noise here is about is how to make sure this never happens again. Right?

Lu Zuckerman
19th Nov 2002, 23:46
Rather than taking the CYA position that the crew should have known intuitively what was happening 40 meters behind them in those few critical seconds, would it not maybe make sense to talk about adding some better information for the guys up front? Perhaps a little indicator that pulls together the already available information about the flight envelope and says in red: "TAIL CRITICAL".

Many aircraft and a few helicopters have a SCAS (Stability Control Augmentation System) installed to prevent the overstressing of the aircraft structure. It is obvious that the A-300 did not have such a system. However if it did it seems from the above posts and many others that I have read about the AA A-300 such a system would have been subject to information overload. The SCAS system is only as good as the control laws that are embedded in the SCAS computer. If the control laws did not reflect or consider the information overload during the entry into a highly volatile and disruptive air stream then the system could not cope with the information at hand allowing the pilots to overstress the structure.



A similar occurrence happened several years ago when a US Navy 707 flying over the Olympic Mountains in Washington lost itsí tail in a dive in which the pilot overstressed the vertical fin during the descent.

:cool:

YouNeverStopLearning
20th Nov 2002, 00:10
The last few posts have missed the point.

The AIC I informed you about is a statement of fact, a limitation that I and many others but not all pilots know about. AA587 DFDR clearly shows a rudder movement from full one way to full the other and as this AIC says, the possible structural failure occurred.

Two questions out of many that remain now are:

1. Was this a pilot input or not;
2. Do pilots know NOT to do this, or is the training system deficient?

I was told many, many years ago not to do this even on Senecaís, and also during training for all my jet type ratings.

The main purpose for my topic was to makes sure that this fact was as widely disseminated as possible INCASE any pilots didnít know about it.

arcniz
20th Nov 2002, 08:24
Lu - I have great respect for the seriousness and content of your posts and the rigorous standards you apply. In this case, however, your argument seems a bit circular. OR maybe I need a nap.

If I understand your post, you say: evidently the a300 did not / does not have an SCAS package, but IF IT DID, then it would have been overloaded by the sudden burst of exceptional data.

As long as we're hypothesizing a monitor / control system to replace the one that isn't there, why not hypothesize one that works?

I know a thing or two about computers, and betcha we can whittle up a design for an ultrareliable one of them that will process anything you can throw at it, speed and complexity-wise, in the way of airdata and derivative rules.

Possibly - as Belgique and others, I believe, have remarked, the sensor data resources integrated into the existing A300 airframe are too filtered and / or slow to provide adequate control data response for this new task. If so, then maybe one would have to tap a couple more holes to augment the sensor set.

Perhaps the way to make entire generations of aircraft safer is to require some standardized air-data sensors and processing tools as a generic airframe component, part of the backbone, developed to a common performance standard. That might make for economies of scale, concentration of greater economic effort in quality development in this complex and expensive part of airframe design, and might pave the way for mid-life fleet upgrades by replacement of the generic portion from time to time - feasible if it maintains a standardized interface to the airframe and proprietary stuff up front.

Lu Zuckerman
20th Nov 2002, 14:21
To: arcniz

SCAS systems in general are extremely fast reacting. The point I was trying to make is that SCAS systems or any other electronic control and monitoring systems are only as effective as the imbedded control laws. If the writers of those control laws did not foresee the problems that must be countered in event of the penetration of a highly unstable air mass then the SCAS or other systems are either blind to the condition or they are ineffective in countering the conditions. I was not implying that this type of system was installed on the A-300

A SCAS system (generally) monitors external conditions (air data computer) and various acceleration devices (autopilot) and electronic control inputs (LVDT or RVDT) and limits the control inputs by the pilot at the control servo to minimize the stresses on the airframe during maneuvering.

Speaking of monitoring systems installed on aircraft, the DC-10 was originally equipped with a system that monitored all of the systems on the aircraft in order to locate potential problems in the systems being monitored. This information was downloaded via a radio connection to the maintenance facility at the turnaround site. The system had many problems and was removed from the aircraft and not installed on new build aircraft.

Most US military helicopters now or will soon be equipped with HUMS (Health and Usage Monitoring System) that do the same as the DC-10 system but without the radio down link. With HUMS the readouts are down loaded to a computer for monitoring and detection of problems.