Allegiant MD-83 Elevator Failure

[patrickal]

A report on Bloomberg today details an apparent elevator failure during takeoff on an Allegiant flight 483 from Las Vegas to Peroia, Illinois last week. It appears the plane started rotating early without pilot inputs, leading to a rejected take off. It sounds like possible shades of Air Alaska 261. Here is the link to the Bloomberg article.

Out-of-Control Takeoff Try by Allegiant Jet Spurs FAA Probe - Bloomberg Business (http://www.bloomberg.com/news/articles/2015-08-26/out-of-control-takeoff-attempt-by-allegiant-jet-spurs-faa-probe?cmpid=yhoo)

[Hotel Tango]

Allegiant said an inspection found that a device that moves one of the plane’s two elevators had become disconnected.

That sounds more like a serious maintenance oversight issue. Frightening!

[iskyfly]

"A preliminary investigation found that a nut on a component that moves the left elevator had fallen off, causing the control surface to become jammed in the up position."

The airline reported the left elevator boost actuator of N407NV had become disconnected. A fleet wide examination of MD-80 aircraft was conducted with no further anomaly found."

[FCeng84]

If one of the elevators was jammed in the trailing edge up position that must have occurred during the pre-flight freedom of motion check. One would hope that such a failure would be noted during the walk around. During and following the freedom of motion check does the flight crew have feedback as to the actual control surface positions to check for a failure like this?

The article linked speaks of "lift-off". Clearly this event involved pre-mature rotation with the main gear still on the ground. I wonder how close they got to a tail strike? We would have to know the weight, but does anyone have an idea how close they were to Vmu?

[peekay4]

If one of the elevators was jammed in the trailing edge up position that must have occurred during the pre-flight freedom of motion check. One would hope that such a failure would be noted during the walk around.
The elevators are also checked during taxi as part of the "free and normal" flight controls check.

On the MD-80 the elevators are checked first by pulling the control column full aft, then by pushing the control column fully nose down until the hydraulic augmentation (Elevator Power) light turns on.

It may be that one of the elevators jammed up during the first part of this procedure -- I don't know how the Elevator Power light is activated to see if this is a possible failure mode.

(The elevators move in response to relative headwind so they cannot be checked for movement prior to taxi).

[bafanguy]

"…a nut on a component that moves the left elevator had fallen off…" really doesn't provide much info at this point. It could've been any number of components.

Just as a basic anatomy lesson on the MD80 series, the elevators themselves are control-tab operated and not connected to the control columns…or each other. At the gate, it's normal (if the wind is from the tail) to see one elevator in the UP position and the other in the DOWN position.

The only hyd source to the elevators is the elevator augmenter system which isn't a normal operating mode but rather a deep stall recovery effort which uses hyd pressure to push both elevators full down if the control tabs are moved beyond a certain amount in the ANU direction. The ELEVATOR PWR ON light comes on to indicate that hyd pressure has been applied to the elevators…not where the elevators went. This light is observed during the normal control check during taxi by pushing the control column full ANU. This will normally move both elevators ANU which cannot be seen by the crew during taxi. The system has an accumulator so it can move the elevators at the gate if you're inclined to play with it to see it work.

There is no flight control position indication in the cockpit (other than stab trim position) like some airplanes have.

As for the Allegiant situation, the crew wouldn't have been able to see anything abnormal on a walk-around. Before castigating the crew, it might be prudent to wait for a lot more information about this.

I have no idea what would've happened if they'd taken the airplane into the air with one elevator jammed full ANU.

[DaveReidUK]

"…a nut on a component that moves the left elevator had fallen off…" really doesn't provide much info at this point.

It also sounds rather odd. Whatever happened to positive locking of all flight control linkages?

[highlandpark]

IF you do the hydraulic ram check at the gate, you can watch the reflection of the elevators moving in the glass of the terminal.

It is the norm to push full forward on the yoke as you take the runway to confirm hydraulic ram is working and to equalize the elevators.

[bafanguy]

"Whatever happened to positive locking of all flight control linkages? "

A MTC issue. Pilots have no answer to that.

[jack11111]

"It sounds like possible shades of Air Alaska 261."


Alaska 261 involved a stripped jackscrew of the stabilizer actuator.
.

[Gilmorrie]

I have flown on Allegiant round trips between Peoria and Phoenix. Never again. I'm not an aviator - just somebody that pays the freight. Poorly managed and not passenger friendly. Are they unsafe? I can't say, but seeing how they manage their operation and the attitudes of many of their employees, I'm skeptical.

[peekay4]

This light is observed during the normal control check during taxi by pushing the control column full ANU. This will normally move both elevators ANU which cannot be seen by the crew during taxi.
The light comes on when the control column is pushed full forward (nose-down). It should NOT come on when the control column is moved full ANU.

[Design Engineer]

AIUI, bafanguy has it 'spot on'.

The elevators respond to the trim tab, not the control column at rest and a ground walk-around won't help you.

Regardless of whether this was a nut that dropped off or a worn jack screw as in the Alaska Airlines accident, this is 100% a maintenance problem.

Damn good job by the flight crew. Without a RTO this jet would now likely be a smoking hole in the ground.

Beware of low cost operators and/or bad management ( how do you know you might ask about management standards ? ). 'Something' is likely to suffer. I hear anecdotally that Allegiant flight crew have a good reputation but what would have happened in the case of another operator ? I believe that the FAA has a duty in this case to inspect MX records for their entire fleet.

[bafanguy]

peekay4,


"The light comes on when the control column is pushed full forward (nose-down). It should NOT come on when the control column is moved full ANU."

You are quite correct...should've been "AND" Thanks for catching my typo. (haste makes waste !!)

[Organfreak]

I wonder, as a non-pilot, how they were able to get the nose back down in order to reverse the rotation, if one of the elevators was stuck NU. Or....wouldn't the airplane have been asymmetric? Sounds as if this was very near to a bad crash. If this is a stupid question, I can live with that.

[pattern_is_full]

how they were able to get the nose back down in order to reverse the rotation, if one of the elevators was stuck NU.

The elevator does nothing without airflow over it. As soon as the crew began an abort (reducing thrust or going into reverse thrust, speed brakes/spoilers on the wings deployed, wheel braking), the airflow/air speed over the stuck elevator drops along with the speed of the aircraft, and the down-force on the tail is removed. Down comes the nose.

[Organfreak]

Thank you, p_i_f! I figured that out after I posted, but you confirmed it. :O

[Terry Dactil]

how they were able to get the nose back down in order to reverse the rotation, if one of the elevators was stuck NU.
The other elevator would have been operating normally and able to oppose the NU force. I guess that's the advantage of having a split system.

[phylosocopter]

Applying the brakes opposes rotation RA-42434

[ACMS]

Closing the thrust levers would help as well

[FCeng84]

Thanks for the education about MD-83 elevator control.

Given that the MD-83 elevators are tab driven, is it known whether the problem for this event was with the elevator itself or with the elevator tab?

Regardless of where the local breeze positions the elevators themselves when at the gate, the tabs should be positioned per the flight deck controls. Does the pre-flight walk-around include observing elevator tab position to confirm that they are consistent?

I fully agree with the comments that this crew saved the day by recognizing improper response to their control inputs and electing to abort the takeoff. My interest here is from a lessons learned perspective with an eye toward recognizing such a problem in the future before leaving the gate if at all possible. The next crew might not be so sharp and might elect to proceed with the takeoff. Sitting at the gate, would there have been any visible clue (either observing the tail from outside or observing the cockpit instrumentation) that this problem existed?

[safetypee]

Assuming the design of the servo tab - elevator system enables the elevator to be ‘picked-up’ at the extremes of the tab movement, then with a jammed elevator there could be some force feedback to the crew at the end positons of the full and free check. This might only be a small change in the stick force and this could be masked by an adverse wind.
IIRC the MD80 pre takeoff checks require a final control check on entering the runway – into wind; this requirement arose after an elevator jammed due to a stone being blown up by a preceding jet entering the runway.

A jammed tab should be identifiable via a stick restriction or jam. It is unlikely that a visual check could be relied on as a safety check, even if the tabs could be seen (dark wet night).
If a tab is disconnected then it and the elevator should assume a free air trailing position, but this could change the overall elevator effectiveness such that the takeoff rotation trimmed condition is biased nose up.

The design requirements for the primary control systems require dual paths and an ability to separate them in the event of a jam. Thus a worst-case pitch-up elevator jam during rotation should be controllable via the other elevator; demonstration of requirements (CS 25-671) should enable the takeoff to be continued after V1 and that the aircraft can be safely manoeuvred for landing.

[peekay4]

@safetypee

I don't think there's a force feedback based on the elevator position per se. The feedback is based on trim position, to give (an artificially) heavier feedback when the aircraft is trimmed for high speed.

A pilot cannot "feel" a jammed elevator (vs a jammed control tab). (See incident below).

@FCeng84

In general (on any aircraft) control surfaces including tabs are checked during pre-flight. But the level of checks will realistically vary. Most visual checks might only detect obvious damage, or when things are seriously out of alignment.

Plus checks for the "first flight" of the day might be more thorough than checks during a quick turnaround, especially if there were no issues noted during the previous flight. Yet a maintenance or other mechanical issue can manifest itself at any time.

Also on the MD-80 series there are actually three different elevator tabs on each elevator. There is the main "control" tab, a second tab geared to the control tab, and a third "anti-float" tab geared to the entire horizontal stabilizer. No one will check all six tabs for correct operation during pre-flight.

(All this is assuming the problem is with one of the tabs -- but the jam is equally likely to be with the elevator itself. Because the elevators are free-floating, they are prone to damage if there are strong winds / gusts on the ground.)

In fact there was a previous incident where an MD-80 had a high-speed RTO because one of the elevators was jammed in the nose-down position. So in this case the plane could not rotate at Vr. Apparently the elevators were exposed strong tailwinds on the ground (100 mph gusts) causing one of them to jam past the nose-down limit.

In that case, both the maintenance crew and the pilots had checked the elevators prior to the next flight by exercising them from the cockpit. Unfortunately, as noted above a jammed elevator feels the same as a normal elevator so they could not detect the problem until the rejected take-off.