stinsonjr

The accident chain of any accident is fascinating. The DC-10 in Chicago seemed to have a weird accident chain. That was obviously maintainance error with the engine change/forklift modification, but I think the airplane would have been flyable even after the engine flew off. Apparently (correct me if I am misinformed) the stick-shaker was an option on the co-pilot side, and the co-pilot was the pilot flying that take-off. As the slats retracted after the engine flew off, the left wing stalled, but the co-pilot did not have the stick shaker and wasn't able to recognize the impending stall. The small things that add up to big things are amazing.

III Corps

The AA DC-10 was, as you noted, a series of event but one factor was that at that time, if I have read correctly, training dictated that if you were above V2 with an engine failure, you decelerated back to V2. The -10 was flying when the engine separated and it was still flying when the slats retracted. The study I saw said that as speed decreased to V2, the airplane just ran out of aileron to keep it from rolling. Nothing said about rudder but again, training would again play a part in another accident later.

I was always taught that the airspeed became fixed when things left the airplane until you had a chance to see how the airplane flew at higher or lower speeds.

The guy who came up with using a fork-lift to change the engines is still very much in the airline industry to this day. But the fork-lift engine change was a procedure used to expedite and was NEVER approved by any agency. But that procedure caused fractures in the pylons and with enough time, it failed.

The NTSB report (DCA79AA017) cites the following :

PROBABLE CAUSE(S)
PERSONNEL - MAINTENANCE,SERVICING,INSPECTION: IMPROPER MAINTENANCE (MAINTENANCE PERSONNEL)
AIRFRAME - WINGS: NACELLES,PODS,PYLONS
MISCELLANEOUS ACTS,CONDITIONS - PREVIOUS DAMAGE
MISCELLANEOUS ACTS,CONDITIONS - OVERLOAD FAILURE
FACTOR(S)
PERSONNEL - PRODUCTION-DESIGN-PERSONNEL: POOR/INADEQUATE DESIGN
MISCELLANEOUS ACTS,CONDITIONS - ASYMETRICAL FLAPS
MISCELLANEOUS ACTS,CONDITIONS - HYDRAULIC FAILURE
MISCELLANEOUS ACTS,CONDITIONS - SEPARATION IN FLIGHT
FIRE AFTER IMPACT
REMARKS- #1 PYLON DMGD DURG MAINT PROC DVLPD BY OPRR.SLAT DISAGRMT & STALL WARNING LGTS INOP AFT ENG SEP.

Wikipedia has a discussion about slowing from the initial speed of around 165kts to the slower V2 and when the slats retracted, substantial lift on the left wing was lost and thus the rolling motion.

http://en.wikipedia.org/wiki/America...nes_Flight_191

Another discussion:
http://aviation-safety.net/database/...?id=19790525-2

As in the USAir 427 accident, the most likely response by pilots is to pull back on the stick rather than to unload the airplane and this only increases the problem and reduces any chance for recovery. (All this to only say what happened and faced with the same predicament, I don't know if I could have sorted it out in time. Some of the best aviators I ever knew just ran out of time when faced with a serious problem. )

Roll Inverted and Pull

The crew didn`t realize that they had truly "lost" an engine. They assumed that the engine quit. The previous poster is correct, they were taught to pull up to and maintain V2. BTW, if it was an L1011, the slats would have been locked in the out position with a hydraulic failure.

Cubdriver

75% is a widely debatable number. There is a big gray area between airline crashes that are due to poor systems design and are likely to lead to crew error, and those disasters that are clearly due to crew/atc malfunction without causes attributable to a particular equipment system. It is probably more like 65% overall, depending on what study you look at and what time period you include. However, it is clear that a majority of air disasters are caused by crew or atc no matter which study you examine. This is good news in the sense that these are controllable causes. Once and aircraft goes into service "that's it", the design has been certified safe whether it is or not. I see a number of fallacies in your logic though I admit the gray area here is large.


I am hopeful there may be a reduction of human factors as a player in the cause of airline accidents because a number of solutions are receiving attention among academia, NASA, and FAA. We all know that atc is overworked and something has to be done about that, and aircraft systems still lead to accidents that could be avoided with better design.

vagabond

Just curious if pilot/aviation training includes a study of case histories. For example, interns and residents in medical schools participate in M&M sessions where they discuss the week's mortality and morbidity cases. The hope is to learn from other people's mistakes.

It's true that there are few 1011s flying now, and the design of airplanes are improved versions of the old, but I think there is always value to the study of history.

III Corps

It is also clear that the majority of the accidents occur when an anglo-saxon male who is the Capt is flying. The obvious solution is to place minority F/O pilots in control. Right? Therein is part of the problem. The statistics are usualy biased. As for fallacies in my logic, please continue. I am here to learn as well as to espouse.

Yes, the airplane is certified but remember that training is always the rubber band that must compensate for whatever deficiencies occur in design. How many 'work-arounds' do you have to use on your FMC? And how did those work-arounds evolve? Because crews found that the handbook was basic instructions and often did not consider compound events.

Certified, yes.. reminds me of the story of when the F-14 was introduced into the fleet. It was supposed to be almost spin-proof but it was not nugget proof and soon after it entered the fleet, a nugget got it into a spin and punched. So much for being spin-proof.

A video of the Tomcat in a flat spin...
http://www.liveleak.com/view?i=78b_1172623495

Someone is going to find the weak link in the chain one way or the other, certified or not.

Finally I am not knocking NASA or the researchers but only noting that often one has to translate what the study is saying for it to be useful to the line pilot. Considering the complicated formula and number of factors that combine to produce an accident, the likelihood that same formula will be created is extremely remote. Some have gone so far as to suggest that NTSB reports only say what happened and now why it happened leaving the summation to the reader.

III Corps

What you are describing is 'hangar flying' where discussions occur and an incident/accident is explored. Doesn't happen that much in airline ops.. you sign in, go fly, go home. Even in training, so much is focused on completing the syllabus that little time is left for 'what if' and exploring what happened and why, if any. And no company is going to pay pilots to come in and discuss what happened the prior week.

That was one of the great things about being in the training dept.. you got to know what was going on out on the line and you got the story from the debriefs of who did what. Just when you thought you had heard it all, some guy would come up with a new one.

The best thing to come along is the FOQA programs which show the line pilot what is really going on in daily ops. Without a good FOQA program, you are just pokin' in the dark with a stick about what is really happening in daily ops.

rickair7777

The previous posters are correct. The crew pitched up for V2 and lost roll authority as the airspeed dropped off. I believe this was caused by a stall on the wing which had the inadvertent slat retraction, not by loss of aileron authority.

Since both pilot and copilot controls are normally mechanically locked together, a stick shaker/pusher on either should work on both. It sure as heck should prompt the OTHER pilot to do something quick. Actually I suspect the bad wing stalled before the protection thought it would...

IMPORTANT POINT: Stall protection (shaker/pusher) on turbojets is not like on GA airplanes...it often does not actually measure a stall condition on the wing. Rather it measures air-data conditions and AOA and COMPUTES the expected condition at which a stall will occur...if it senses those conditions it ASSUMES that a stall is imminent and activates sticker and then pusher. This depends on the programmed formulas being correct. The formulas usually work fine, but they depend entirely on the KNOWN aerodynamic properties of the wings. If the wing is not performing as expected, the stall protection is useless. This can (and has) happened when the wing is contaminated with ice/snow or the shape of the wing is physically altered (severe damage). The challenger and CRJ are notorious for fatal ice accidents where the stall protection didn't know that stall was imminent.

mjarosz

Just to answer Vagabond's question, when I was at Daniel Webster, one of our senior level classes was AE401: Flight Safety. We did, among other things, research and analyze various accidents. Each group took a case study every two weeks (about 6 for the whole semester.) As there were 4 groups, we were exposed to 24 various accidents. They ranged from Eastern 401 and AA 191, to AE 3379 (which led to the Pilot Record Improvement Act) and even JFK Jr. I believe I am a much better pilot for having studied these landmark accidents and learned from the error chain that led up to them. It is very unfortunate that something like this is not included in pilot training. It certainly is extremely valuable.

stinsonjr

Maybe we should have a section, or thread, about "landmark" accidents on APC. I for one would be really interested in that. With all the talent on the APC Board it would be awesome to hear what everyones thoughts were.