Subject:  MD-11 Quality Assurance, Quality Control or Quality Quandary? 

Anyone? BabsF342@aol.com wrote: Does anyone have an opinion on this email that was sent to me? 

Hello again Barb, I hope that everything is (comparatively) well! One additional on the puzzle why sr111 finally went down: Despite the fact that there was smoke, possible open fire and massive break-down in the electrics during the last minutes, one can be quite assured that (as long as it was basically possible to steer and control the aircraft) the crew would have made every effort to bring sr111 into Halifax. I wonder if (another) design flaw prevented this: Due the unusual design of the tail (vertical stabilizer) of the aircraft - being smaller than normal to reduce drag while in cruise - maybe that the pilot was very dependent on good attitude references to be able to avoid entering oscillations in the vertical plane? Where, in other aircraft models, the pilot is aided by the aircraft being [so to speak] self-stabilizing, maybe in the MD11 it is very easy to enter a state of aggravating oscillations when hand flying without having good attitude references? If this is true in some way, maybe that the MD11 designers never reckoned that an aircraft could be as crippled as sr111 (probably) was, making it very difficult to avoid pilot-induced- oscillations. Maybe you can check this out further, with one of your expert contacts? Regards Lennart 

Answer

The less inherent (i.e. designed in) dynamic or static stability that an aircraft has, the more likely it is to be divergent, even in the trimmed state. Dynamic stability is the propensity (once disturbed by a gust) to dampen out any such upset's effects and return quickly, "hands-off", to the trimmed airspeed. If it tends to cycle around the trimmed airspeed it is said to exhibit neutral properties of dynamic stability. This can be very tiring if you are hand-flying. If it diverges further (as in a PIO or pilot-induced oscillation) it is said to have a divergent phugoid (i.e. negative damping)- and this is not good. In such an aircraft it's very difficult to avoid extreme PIO's.. You need artificial assistance (but more of that later). Heretofore the only aircraft to have dynamic instability intentionally designed in have been high performance fighters or sports aerobatic machines. Fighters enjoy a manoeuvring advantage because of it, but it is finely controlled by flight control computers. If the computers fail the pilots eject and there have been many examples of this. This flight characteristic is what the NTSB is concerned about in the MD-11's handling qualities, particularly at altitude. At high altitudes airliners ride on a knife-edge between aerodynamic stall and Mach buffet. They are said to be operating in the corner of their operating envelope. Unstable control characteristics can really complicate things up there. There have been a number of MD-11 handling upsets where passengers were killed or injured because of PIO's. Before reading further, you should re-read this (recently updated) article. It contains qualified comment by a current MD-11 pilot about the MD-11's poor handling qualities (and some accident history). http://www.airlinesafety.com/faq/faq9.htm 

The MD-11 complication was that the 40% lesser area of its empennage (tail surface areas), compared to its predecessor the DC-10, was artificially compensated for by a system known as the LSAS (Longitudinal Stability Augmentation System). This system's electric motors altered the angle of incidence of the horizontal stabilizer up and down depending upon thrust settings, centre of gravity disposition and pilot inputs. The reason why McDD 'designed in' this system (as a compensation for the reduced tail area) was in order to help it meet guarantees of payload, speed, range and SFC (specific fuel consumption). It fulfilled the performance guarantees (after initially failing them) by the reduction of tail surface area and trim drag and by having the LSAS manage the C of G at a quite aft position (i.e. keeping it more tail-heavy). The LSAS was also assisted by the FMS ability to pump fuel to and from a tailplane tank (from which fuel then had to be pumped up to the #2 tail engine). In a normal aircraft the tailplane's lift will be acting negatively against the resultant nose-down pitching moment of the thrust-drag/lift-weight couple (aka trim drag) i.e. in a conventionally configured aircraft in cruising flight, whilst you have positive upward lift on the main-planes the tail-plane will be producing a balancing downward force. In the MD-11 however, the electrically-driven LSAS and FMS acted in concert to maintain a balance by altering the tailplane incidence and adjusting weight distribution by pumping fuel around. So, as if in a designer's dream, in the long-term cruise you actually had both wings and tailplane providing positive lift with zero trim drag. This is very good for the economics of fuel burn but...., even when this system was working, MD-11 pilots have managed to "lose it" and kill passengers. This has ocurred as a result of pilot intervention upon the occasion of autopilot disconnects, high-level clear air turbulence or inadvertent wing-slat operation. It is just an aerodynamic fact of life that an aft C of G makes the aircraft "skittish" once you start making inputs. "Pilot-induced oscillation" is an unfair descriptor because it fails to acknowledge that you wouldn't have the condition if it hadn't been designed in. 

One unredeeming characteristic of line pilots is that they tend to live with (and work around) what they are given to fly. In the bad old days of early design you could tolerate some vices because much was determined by experiment (or by guess and by God). Nowadays it's inexcusable because designers have the computational skills to give an aircraft the handling qualities and performance they wish - plus or minus nothing. However when the selling points of a new aircraft are its payload, capacity, range and minimum fuel burn, they have to rob from Peter to pay Paul. This is what may have happened in the case of the MD-11. The manufacturer was desperately looking for performance advantages so structural strength (undercarriage and wing attachment) and handling qualities are raped in the name of performance. If the Test Pilots you employ to test the machine don't give it a tick in the box you go hire yourself a new batch. Working around FAR25 stipulations isn't difficult when you know how. Quality Control and Quality Assurance don't normally look at handling characteristics. These are a given - and in the province of the test pilot. QC & QA are more interested in the build standards and regulatory compliance. The risk of such unconscionable conduct is that you end up creating a handful for the pilots and it gets a reputation as a rogue. Eventually it will get a justly deserved name as a pax-killer. McDD didn't sell very many MD-11's because the word got out real quick about "the SCUD". After all, it was just a redesigned DC-10, and that aircraft had already made its own name - for its lethality. For a good idea of the MD11's gestation, look at this URL: http://www.geocities.com/Eureka/Concourse/7349/gestation.html

 So in the case of sr111, to put it simply, the electrical failure killed a number of systems but also took away many things that were artificially providing the pilots with a flyable machine. To summarise: a. The Pilot's Flying Displays (PFD's). The EFIS (Electronic Flight Instrumentation Sysytem) displays of attitude, heading, rate of descent, altitude etc on the six CRT's (cathode ray tubes) were most likely lost. The pilots were left with poorly positioned [and illuminated] standby analogue instruments (whose power supply was not integral and was also likely to have been affected by the power outage). b. The # 2 tail engine flamed out (without electric pumps to pump the fuel up to the engine, it must fail. The other wing engines can gravity feed). c. the balancing reposition of fuel (from tailplane tank to wingtanks) could have been stopped either by the elec fire or the checklist. d. The autopilot and LSAS and trimming (that all artificially compensated for the built-in aerodynamic instability) were all gone when the power failed. The LSAS may have been simply knocked out by the elec fire or checklist (or it may conceivably have shorted and motored the fin slab the wrong way).

 Handling qualities would have deteriorated but the major discrepancy on a dark night [without autopilot or a natural horizon] would be the lack of flight instrumentation (for aircraft attitude control). Without that, an unrecoverable unusual attitude would quickly develop. However, if the smoke and flames in the forward cabin ceiling became very evident after the cabin lighting went out and the fire flared up, it would be natural for forward-seated passengers to "head for the back of the bus". This would severely complicate the aircraft control problem as the already aft C of G was taken further aft. This would cause the aircraft to become tail-heavy to the point where control could have been lost. Further complicating this scenario, the loss of the #2 engine would remove the benign effect (i.e. the pitch-down moment) of the high centre-line thrust upon this control problem. What could a pilot do to try and restore some controllability (assuming that he could still fly an attitude reference of some sort - such as the lights of Halifax)? Well he could deploy some flap which would give a desired nose-down pitching moment - but not slats or spoilers. You'll have read in the first URL [above] that MD-11 spoiler deployment causes an undesired gross nose-up pitching moment. So what configuration did sr111 impact in? With flap but without leading-edge slats, a fact that has perplexed the investigators. Normally they go down together. To have one and not the other, they would have had to be intentionally selected as such. Why would the pilots have done that? Well, I've already explained that reason in this paragraph. 

The end result of a very tail-heavy aircraft? A high-speed, high g downward spiral. These were the factors at play, but in the absence of any indication that this was the scenario, it must remain as pure speculation. Only the investigators could possibly know and they are not yet saying. I'd expect that this aft-out-of-limits C of G scenario is never covered in any airline simulator drill. Why not? Simple, it's too far out in right field. It's the sort of emergency that's just not anticipated nor expected, so never was there ever a drill nor was it ever discussed or covered in Aircraft Handbooks (MD-11 or otherwise). But, think about why it could happen..... and then wonder about it. It's never the bullet you see or hear that gets you. The nature of accidents is that they come out of the blue, unannounced, bringing with them all manner of unforeseen "knock-on" complications. Always with the aid of hindsight we can see how they might have happened, might have been avoided.....Sometimes it's just not a chicken or the egg situation, you know that someone has thrown an even dozen at you and the chicken's nowhere to be seen.

 So if the MD-11 has these undesirable handling vices and structural defects, what will "they" do about it? Will "they" ground it? Well unfortunately you need something of the order of magnitude of the early Comet series of disasters for that to happen. You may recall that the Comet Mk1's started falling out of the sky because its design engineers didn't understand the characteristics of metal fatigue caused by pressurization cycles. The portholes were the wrong shape. There was no alternative to grounding that fleet. By comparison, in the MD-11 instance there are a number of disassociated anomalies unique to it. They've "fixed" the IFEN problem but there's still the silly electrical system design (smoke/elec/air switch and high current routed through cockpit switches), the Kapton wiring, the insulation blankets and the MD-11's poor handling qualities. There's also the MD-11's tendency to flip upside down and burn because of an underbuilt wing/undercarriage. This last misbehaviour is perhaps the most serious because of its effects upon passenger evacuation times - but they are all serious deficiencies in a certificated passenger-carrying aircraft. Well, why isn't anything being done you might ask? Why don't they make them all into freighters? Well accidents such as the ones at Halifax, Hong Kong, Shanghai and Subic Bay all take a long time to investigate and report upon. It's not sufficient to simply acknowledge that the deficiencies exist and pre-suppose that the accidents were caused by them - even though the Newark FEDEX accident was a good pointer to the structural cause of that particular crash. The authorities require proof positive and perhaps a higher body count upon which to base their sums. Otherwise you're being "unfair" to the operators and the manufacturer. It's not as if they are insured against grounding nowadays. And the MD-10 (large numbers of DC-10's being re-worked to a two-man crewed MD-11 standard) are coming down the developmental chute. So what exactly is going on, if indeed anything?

 Behind closed doors the FAA's "suits", ATA bean-counters, industry moguls and the manufacturer's engineers are debating fixes and flaws and the likely PR effect of doing nothing. Fixes are pretty well ruled out on cost grounds. It's very costly to take an aircraft out of service - you have to add the loss of revenue to the cost of formulating and enacting the fix. In the meantime the aircraft gets an even worse name and you may find that, by the time it's ready to fly again (or sell), no-one will touch them with a barge-pole (and the passengers are actively looking to avoid them). It'd be smarter to convert them all to freighters -except for the fact that there are so many freighters already firmly in the order-books. The bottom-line conclusion could inevitably be the usual non-decision. That's the well-known escape-route taken when it's all too hard, no-one can agree, it's politically unpopular and there is no-one in ultimate authority willing to make the hard decisions. Could be, but will it be? Not necessarily. Why do I say that? Well you think about it. The black box usually goes toute suite to North America. Keep thinking. Got it yet? 

No snowflake in an avalanche ever feels responsible. Stanislaus Lezczynski

I think there is only one quality worse than hardness of heart, and that is softness of head. Theodore Roosevelt

What is the best government?-That which teaches us to govern ourselves. Johann Wolfgang von Goethe

Return to HOT off the PRESS