I just watched the Air Crash Investigation-"Cutting Corners" which highlights the Alaska Airlines flight 261 crash approximately 10 years ago. Something that struck me is that towards the end of the episode, it was mentioned that it was discovered that the MD80 horizontal stabilizer jackscrew is not failsafe. Now, to my knowledge that is the only crash attributed to the failing of a jackscrew on an MD80(and I believe DC9, B717, etc), but I was still curious as to some of your thoughts about the design? I understand that it really is impossible to make everything failsafe on an airplane, but it's apparent that on modern airliners, redundancy is huge.

 

A jackscrew is getting about as simple as a mechanical motion device can be. It should be failsafe in day-to-day operations as long as it is lubricated and inspected properly.

If it was NDT'ed for manufacturing defects, is lubed and in wear spec before flight, it is almost inconceivable that it could fail under design loads.

By failure, I mean allow the stab to flop around without restraint. A motor failure or mechanical jam should normally just cause the trim to be stuck in that position...not catastrophic. I think there may have also been a restraining strap on that design which snapped under the sudden force of failure, but I may be mixing my airplanes up.

Any alternative mechanism involving say hydraulics or gear boxes would be complicated and more likely to fail, and would need a separate brake/locking mechanism to hold the thing in position if the mechanism let go. The jackscrew, in addition to being simple, has the benefit of being self-locking once the motive power is removed.

I think that if the AK crew had stopped monkeying with the trim and landed immediately, the jackscrew would have held together long enough to get them on the ground. Of course they had no way of knowing that, but I think that accident has spring-loaded most of us to declare and put the thing on the ground at the first hint of primary flight control abnormality.

 

I met one of the investigators on the crash a few years ago, and she said that the crew was under "signifacant pressure" to get the airplane to somewhere with Alaskan Airlines mx, I guess the captain was actaully mad about this and parts of the exchange were caught on the CVR, they wanted to return to their departure airport, if I remember correctly.

She also said that when examining the jackscrew upon its recovery the threading of the shaft was so worn down that it had become detached from the actual shaft and was, in her words, "like a giant slinky."

The "defect" with the MD-80 jackscrew, as she explained it, (many different designs use jackscrews) is that it is a simple "acme nut" set up. The shaft, or think of it as a giant bolt, is directly in contact with the nut, or the portion that is connected to the horizontal stab, with only a layer of grease between the two parts to prevent metal-on-metal contact. The Boeing jackscrews use a ball bearing set where a row of bearings are placed between the jack and the screw allowing the two parts to transfer energy via the bearings rolling along the threading of the jack.

And yes, i do belive that the DC-9, MD80, MD90 and 717 all use the same design.

 

That same show highlighted the fact that the NTSB has consistently recommended that the FAA write into Company Flight Manuals that once trim problems were discovered in flight, "to land the aircraft at the nearest suitable airport". That recommendation, at least for the DC-series, has not been acted upon. As an accessory drive it was not considered structural. And most trim problems, like said earlier, lead to a locked up screw which can be an armful but not a structures failure. They mentioned the same jackscrew was used for all the DC-series aircraft, and that the DC-8 had two jackscrews.

 

One human factors lecturer used this accident to show how we accumulate data to show we can do things safely. As I remember, the original inspection of the jackscrew was done at 300hrs and by the time Alaska went into the water, that inspection cycle had increased to nearly 3000 hours. Also, the jackscrew had been inspected and was near limits but was still WITHIN limits posing the problem of when do you change out parts.

The inspection was also problematic in that lubrication cannisters, once in the gun, were difficult to identify so it was possible to use the wrong lubricant. Also how it was to be applied, how much and where was fuzzy.

This was all used to show that EVERYONE was involved... the regulating authorities, mechanics, airline, OEM.. everyone said it was safe to continue and had the data to show it could be done. Like Challenger and Columbia.

Yes, the DC-8 had two jackscrews but it was decided the -9 could operate with one much like the 727 had two yaw dampers but you could dispatch the 737 with NO yaw damper so it got only one.

Finally, I have always been taught with a flight control problem you remain in that config is at all possible. The crew had lost control once and regained control before continuing to troubleshoot and losing control the second time.