PA-31 lost due to mis-installed trim tabs

[punkalouver]

Originally Posted by LOMCEVAK

One problem related to cockpit control checks is that some control surfaces are not visible from the cockpit, especially elevator and rudder trim tabs. The only way that you can then check these is during the pre-flight walkround having set full trim deflection first. However, this requires careful thought because the tab will deflect in the opposite direction to a control surface and will require you to go back to the cockpit to then set the opposite direction. Also, if an aircraft has a mechanically linked nosewheel steering system then full rudder checks can only be performed whilst taxiing and that is not the time to look over your shoulder at the rudder, which possibly you cannot see anyhow. So, there are a few practicalities which do require robust engineering procedures and pilot observations cannot always be used as a mitigation.

I believe that the above is much of the solution.

For light aircraft, I don't like checklists that say 'Controls.....Free'. The response should be free and correct. Admittedly, it is difficult to check the rudder on some aircraft like Pipers where you have to be moving but it is unlikely to be a fatal fault.

As for the 'correct' part of Free and Correct, that may have to be checked on the walkaround as mentioned in the quoted post. It is usually fairly simple to just reach into the cockpit and move the control column fore and aft while standing outside and looking at the tail.

As for the trim tab, it can be the same thing. For example on the C182 I flew the other day, I move the elevator tab to full nose down and see tab up. Then full nose up and see tab down. Now you have confirmed full trim movement and proper direction. Then I move the rudder trim tab a bit and check that the rudder pedals move in the proper direction. Doesn't take much time.

After a serious maintenance check, a more thorough flight control check must be done, especially in larger aircraft where visual checks are typically not done. It can be quite complicated in an aircraft like a 727 with number of slat panels versus flap selections combined with number of ailerons deflecting at different flap selections all needing to be taken into account.... and spoiler deflection differences as well. Of course a good post maintenance checklist using flight crew plugged into remote jacks and communicating with the cockpit while observing control movements should be used.

Checking maximum tab deflection angles with a protracter and comparing it to the type certificate is an interest thought. I suppose one could do that. If it were a typical flight school aircraft with others parked nearby, I might just make a comparison.

Best to be paranoid about this stuff.

I actually enjoy the walkaround(in good weather). The ones requiring the most detail in order to be considered reasonably thorough are........older biplanes. Got current on one recently that I hadn't flown in a few years. The instructor suggested an hour long walkaround for the first check of the day on that type. While I think I do it faster, he is not far off, in my opinion.

[India Four Two]

There was a similar accident in 2015 in Spokane Washington, where a Malibu crashed on a post-maintenance test flight, because the right-wing aileron cables were mis-connected.

https://aviation-safety.net/wikibase/175978

What surprises me, apart from the lack of proper pre-flight control checks, is that the cables had identical connectors - a major design flaw.

The sections of the two interchanged cables within the wing were about equal lengths, used the same style and size of termination swages, and were installed into two same-shape and -size receptacles in the aileron sector wheel. In combination, this design most likely permitted the inadvertent interchange of the cables, without any obvious visual cues to maintenance personnel to suggest a misrouting. The maintenance manual contained specific and bold warnings concerning the potential for cable reversal.

[DaveReidUK]

Murphy strikes again.

[Pilot DAR]

The prevailing design requirement says:

Sec. 23.685

Control system details.

........
(d) Each element of the flight control system must have design features, or must be distinctively and permanently marked, to minimize the possibility of incorrect assembly that could result in malfunctioning of the control system.

But, that doesn't prevent it happening.

In addition to confirming the correct direction of motion of each flight control following any maintenance, it's also important to confirm the correct range of motion (travel stops and freedom). I've had a few post maintenance (and radio installation) planes where full control travel was not available - obstruction behind the panel. I also test flew a C 206 once (which I'd flown lots prior to the maintenance) with a serious mis rig. The elevator trim tab moved the correct direction, which I visually confirmed myself before flying. But, its travel stop limits had been reversed. Being as they are 5 and 25 degrees, I had 4 degrees where I should have had 25 degrees, and the nose up control force was barely manageable for a terrifying circuit. The maintenance manual for that 206 was somewhat ambiguous, later Cessna service manuals for the 206 are very bold about this, as I know I'm not the only pilot to blunder into this very unsafe condition.

So, if you're doing the first flight post maintenance, you would like the maintenance staff to demonstrate to you that the control and trim tab travels are all as per the type certificate data sheet specifications. If you're not sure, keep asking until you are, this is not complicated stuff!

[punkalouver]

Originally Posted by Pilot DAR

The prevailing design requirement says:



But, that doesn't prevent it happening.

In addition to confirming the correct direction of motion of each flight control following any maintenance, it's also important to confirm the correct range of motion (travel stops and freedom). I've had a few post maintenance (and radio installation) planes where full control travel was not available - obstruction behind the panel. I also test flew a C 206 once (which I'd flown lots prior to the maintenance) with a serious mis rig. The elevator trim tab moved the correct direction, which I visually confirmed myself before flying. But, its travel stop limits had been reversed. Being as they are 5 and 25 degrees, I had 4 degrees where I should have had 25 degrees, and the nose up control force was barely manageable for a terrifying circuit. The maintenance manual for that 206 was somewhat ambiguous, later Cessna service manuals for the 206 are very bold about this, as I know I'm not the only pilot to blunder into this very unsafe condition.

So, if you're doing the first flight post maintenance, you would like the maintenance staff to demonstrate to you that the control and trim tab travels are all as per the type certificate data sheet specifications. If you're not sure, keep asking until you are, this is not complicated stuff!

One need simply to place the elevator tab to the takeoff position at the beginning of the pre-flight when you are doing the cockpit checks, and then observe its proper position when checking the elevator area.

It should be in the normal, near neutral position. Then it need no longer be touched before takeoff and the aircraft should initially handle fine upon getting airborne.

More detail of the flight would be welcome and if there were any unusual indications that you noticed prior to takeoff, such as trim indications when checking the trim wheel on the ground, as well, how the trim forces came to be as the flight progressed, the effects of flaps and power, etc.

[Pilot DAR]

One need simply to place the elevator tab to the takeoff position at the beginning of the pre-flight when you are doing the cockpit checks, and then observe its proper position when checking the elevator area.

When the tab travel limits are not the same (up vs down), the "takeoff position" is not necessarily in the middle, and the tab not necessarily trailing to the elevator at an indicated position. This was what I did not investigate enough during my control check. I did see the tab move in the correct direction, through what appeared to be the operating range. What I did not realize was that the long travel range of tab movement was on the side I did not need it.

The text of my original story about this:

A maintenance test flight was required, prior to the Transport Canada approved test flight for the evaluation of an external installation to the aircraft. I had flown this aircraft months earlier, for a very similar mod evaluation purpose, including spins and dives to 1.1 of Vne. Those flights were fine. It had not flown since, while it underwent inspection for a commercial C of A. This was the maintenance test flight following that inspection.



The aircraft is a Cessna U206F, with a Robertson STOL kit, and additional external equipment mounted. Following a review of the documents for the maintenance, a thorough preflight inspection, and normal start, I taxied out. Just before taking off, a final check of control freedom and direction, including the elevator trim tab – I have just read too many stories about flight control problems on test flights… Everything looked as it should from the pilot’s seat.



The aircraft was light weight at takeoff, and promptly leapt off the runway… Then immediately leapt more. A swift and large nose down control input seemed to help, but still it was heading for space! I confirmed that the pitch trim was set for takeoff (and it had been) then I rolled it all the way nose down. That helped a bit, but not really much.



By this point, landing back on the remaining runway was no longer an option. A circuit now lay ahead of me, which was going to be a muscle builder! I could lie on my back and hold my 30 pound daughter at arms length over me for a few minutes at a time. This flying was a lot more demanding than that! Flying with one hand so as retract flaps (very carefully) and adjust power, was manageable for only a few moments at a time. In downwind, I found that full flaps created the least uncomfortable configuration, probably simply because slower speed, less control force. I did not dare fly too slowly, being quite uncertain about how the aircraft would handle if stalled this way. During downwind, I was thinking about how I might jam my knee into the control wheel so as to relieve my tired arms – there really was no practical way..



Flaring for landing was an exercise in how to appropriately reduce the massive downforce I was applying to the controls. It worked. I taxied in, alternating thoughts of shock, and the old joke: I just flew in from – here! – and are my arms tired!” I also reminded myself how lucky I was there were no seats, occupants, or flight test ballast in the back for this first flight….



Knowing that I had had a serious pitch control problem, I asked for a check of the elevator and trim tab travels. The elevator was as it should be. The trim tab offered 3 degrees, where I would have expected to find five degrees according to the type certificate data sheet. I asked that the maximum travel in this direction be provided. I got 7 degrees. There was still a lack of certainty as to what the problem was, because I could not see how such a small travel limit error could produce such a dramatic effect. I invited the mechanic to join me in my next test flight. He reluctantly agreed, knowing that if he would not ride in it, why would I fly it. This time I was well prepared to abandon the takeoff, if things were not right. The takeoff was better, but the pitch control problem was still there. what had been a 40-50 pound push, was now 15-20. The mechanic now had no doubt that something had been very wrong on the first flight, as was still wrong now. I landed back.



After a rather puzzling review of the maintenance accomplished since my flight a few months earlier , the answer was found. The maintenance personnel had put a little too much thought into what they were doing…. The required maintenance had included the required overhaul of the trim actuator, thus it, and all the chain and cable, had been removed, and reinstalled. During the re-rigging, the technician had read the travel requirements for the tab in the maintenance manual. Instead of setting the tab for an up travel limit of 25 degrees, he set the travel to a “nose up” (tab down) limit of 25 degrees. This left the travel limit in the other direction of only 5 degrees, which I hereby attest is not anywhere close to the requirement!



By trying to “figure out” what the system required, the actual instruction was not followed as written. An unsafe condition was the result. The safety system further broke down, when the second signatory for the work accomplished did not detect the error. This was also a maintenance failing designed in by Cessna, as it was possible to mis-rig the system in the first place, and the manual did not give any warning to check for the mis-rigging.

[rotorfan]

I feel strongly about the “free and correct”, due to good instructors, and much reading of accident reports. When flying light GA planes, my technique for a control check is to move full aileron, then control full-forward to full-aft and back, reverse the aileron, repeat the pitch.
When doing a run-up and final checks prior to takeoff in a 172, I had full aileron and pulled back, only to have something catch behind the panel. With the typical winds here in the Midwest US, stiff crosswinds are common, requiring a lot of aileron while pulling to flare. I can imagine the above problem preventing me from fully flaring, and pranging the nosegear. I heard that a bundle of wires was restricting the pull, but only with a lot of aileron. I was amazed- and stunned- to think of all the pilots that had been flying the aircraft before me that had not caught the problem, and flown with it.

A well-known lawyer in my city died in a takeoff crash, post-maintenance. They found that, the more he trimmed nose-up, the more the plane nosed down. I believe it was a Baron, and I understand the elevators can be installed on opposite sides on the type, so the trim worked in reverse. Due to a post above, I will now conduct a full-trim exercise on the ground on my preflight checks, too.

[First_Principal]

In further reading these posts I'm reminded of the C-47 I once had something to do with.

When doing the run-up one of the pilots in particular would do a 'free and correct' with power on, the resultant movement of the fuselage was very evident - and presumably obvious from the cockpit had something been wrong.

I never tried it myself, so can't report on the control forces etc, however for a craft where you can't see what's happening behind you I guess it had some applicability. For me I was a little reluctant, thinking of possible damage to control surfaces from stones etc, particularly as we did fly it a lot from grass strips and those big props on full noise sure kicked up a lot of cr*p.

FP.

[punkalouver]

Originally Posted by rotorfan

A well-known lawyer in my city died in a takeoff crash, post-maintenance. They found that, the more he trimmed nose-up, the more the plane nosed down. I believe it was a Baron, and I understand the elevators can be installed on opposite sides on the type, so the trim worked in reverse. Due to a post above, I will now conduct a full-trim exercise on the ground on my preflight checks, too.

To be honest, I do a full trim check just to see that a reasonable amount of trim movement happens. But I cannot be sure that the amount of trim deflection that happens is not excessive(or not enough). It is really just a general 'seems reasonable' check in terms of amount of trim wheel movement. if one is checking for proper deflection direction, only a small amount of trim wheel input may be necessary to compare the tab movement direction in the proper direction as compared to what was seen a few seconds earlier. In reality, I don't plan to carry around a protractor and type certificate values for all the aircraft I am flying. I can see a value for a post-maintenance check flight.

But I still don't see why the C206 was leaping into the air. I would expect that the tab was quite far from the aligned with elevator position to do that. That said, I have never flown a 206. While elevator trim tabs on the Cessna's I have flown are normally a little bit off of being completely aligned with the elevator when set to the takeoff position, one would think that if it was close to or at the aligned with the elevator position for takeoff, there would only be a small but not unreasonable amount of control column force required. I could see a situation where mis-rigging cause an erroneous cockpit indication where the pilot thought the tab was set for takeoff but was not.

[blue up]

‘The Squadron had several new pilots now, including another American, Nicky Knilans, a droll youngster from Madison, Wisconsin, with precisely the quality of nervelessness that Cheshire wanted in 617. Knilans had already done about twenty trips with 619 Squadron and been in strife on nearly every one of them. Several times on the way to the target he had had engines shot out, and more shells had ripped chunks out of his aircraft, but he had always pressed on and bombed and had a D.S.O. to commemorate that laudable habit. Once his rear gunner had been cut in two by a night fighter, and it was such a terrible mess that, when they landed back at base, the ambulance driver who met them had had hysterics and largely left it to the nerveless Knilans to get the remains out of the turret.

Knilans had joined the Royal Canadian Air Force before America came into the war and had just recently been transferred. Now a '’Lootenant’ in the U.S. Air Force, he wanted to stay and finish his tour in the R.A.F., and had a row with his crew when he had them posted with him (without telling them) to 617. They claimed it was a suicide squadron, but, as Knilans pointed out, few people on 619 had ever finished a tour either, so it didn’t make much difference. The crew was even more unhappy when Knilans suddenly seemed to develop into an exceedingly hamfisted pilot. He was given a new aircraft, ‘R Roger’, when he joined 617 and could not make his usual three-point landings any more; even the take-offs were frightening, as ‘R Roger’ seemed most reluctant to leave the ground, and when she did leave climbed like a tired duck. ‘Give the game away, Nicky,’ one of his gunners said. ‘You’re getting flak-happy. You can’t even fly any more.’

‘Doggone, it’s not me,’ said the badgered American. ‘It’s this bloody-minded aircraft. You don’t have to fly it, you have to understand the son of a bitch.’

At length, the causes of R Roger’s terrifying tendencies were ascertained. Paul Brickhill’s The Dambusters, continues:

‘On 1 June Avro experts fitted new automatic pilots in the Lancasters for the D-Day operation, and Nicky Knilans at last found out why his much-cursed ‘R Roger’ flew like a lump of lead. They found it needed longer elevator cables than the others, inspected to find out why and discovered that the elevators had been put on upside down at the factory. Knilans had been flying it for months like that and, as Cheshire said, ‘Only you and God, Nicky, know how you stayed up.’

‘Not me, sirrrr,’ Knilans said in his American drawl ... ‘Only God. I didn’t know.' At any rate he was very relieved, but not so much as his crew. ‘R Roger’ had so often frightened them.’

[Expressflight]

When I worked for a small GA engineering company in the 1970s I was the only licenced airframe engineer apart from the Chief Inspector who was an elderly gentleman who worked part time. It was difficult for him to carry out some duplicate inspections in confined spaces so I always duplicated my own initial inspection while talking him through it alongside me. I was also the approved flight test engineer and both I and the usual flight test pilot took the job very seriously, which I know was not always the case sadly. I think it's always a good idea for the engineer signing off the work to do the flight test, which is easier to arrange in a small organisation. Over the years the only major problem we had was an insipient engine bearing failure due to oil contamination as a result of sand blasting of the cylinders not being cleaned out prior to re-assembly. We just made it back to the airfield with a very noisy engine as we turned off the runway.

One time, during an annual inspection, I found the wiring harness behind the panel of a Cessna 150 or 172 would jam the control column at an extreme position while checking 'full and free'. I think it came about during an avionics refit some time before and had just never been noticed in normal operations where the daily check was probably rather cursory. Luckily the full range of movement had not been required during that period.

[stevef]

I think most of the intermittent control restrictions I've found have been caused by sagging cables/looms and fouling hose clips, cable ties and locking wire pigtails behind the instrument panel.

[red9]

On collecting our PA31 G-PIZZ from its annual maintenance check , as I taxied away from the maintenance hangar I noticed that the control yokes ( without any pilot input) sat right "out" ie away from the coming and instruments where as previously as any PA31 driver will tell you they sit flush, ie "in" by the coming until aerodynamic speed is sufficient for the tail plane to start " flying" and they start to come out ( does that description make sense ??)
You cannot see the tail plane and elevators from the pilot seat.
In taxied back convinced something was wrong.
Very embarrassed engineers now set about now re-rigging the elevators correctly - they were crossed cables giving the opposite control to the input given....

[stevef]

That was a close one. R9. I'm surprised that two licenced engineers signed duplicate inspections confirming correct sense and travel without checking. Did you file a Mandatory Occurrence Report?
By the way, is that the right registration? G-INFO shows G-PIZZ as a de-registered Lindstrand LBL105A hot air balloon.

[DaveReidUK]

G-PZIZ, presumably?

[fitliker]

I always do the pre-flight as if my life depends on it .

[zac21]

Quite a few years ago I was watching a few sprog pilots milling around the rear of a C210 while talking to the rostered pilot, who was conducting a daily inspection.

The left door was wide open as he walked around the back and lifted the elevator to the full up position, looking underneath it then lowering it down.

From where I was sitting I could see that the control lock was STILL IN, that realy got my attention, so I told the pilot what I saw. He laughed until I showed him, then was confused. Told the engineer who imediately inspected it, then admitted he removed the autopilot without adjusting the elevator cables afterwards, so they were just laying slack in the in the bottom of the empenage.

The realy strange part is, the same pilot had flown the same C210 daily for over a week after the engineer removed the autopilot and never noticed any difference?

[Uplinker]

Might already have been mentioned, but do engineers use checklists?

I ask this because many years ago I innocently asked an engineer the same question, as his hands were flying all over the overhead panel, while we waited to close up on stand. (BAe 146). "No he said, we use common sense".

It strikes me that a hard pressed engineer could conceivably make a mistake*, and giving the trim a twirl, sees the trim tab moving and 'confirms' that it is connected and working correctly. But crucially, it might be rigged in the opposite sense. Ditto primary flight controls. Working on many different types, it could be possible to make a mistake. Or a junior engineer, asked to confirm movement might confirm correct movement but not actually be 100% sure about the direction.

Likewise on a control check, the correct direction of movement of the flight controls needs to be positively confirmed, not just that they move freely. Several times, during control checks, the PM has said "full left" when I was holding full right. (Airbus - where you can't easily see the other pilots side-stick to sub-consciously 'cheat' in the control check).

It would be comforting to know that engineers would use a check list after work on flight controls, that specified for example; ' move the XX trim wheel in YY direction, and confirm the trim tab moves up'.

*We all make mistakes - I've made some howlers - and I mean no disrespect to any engineers.

[stevef]

Originally Posted by Uplinker

Might already have been mentioned, but do engineers use checklists?

I ask this because many years ago I innocently asked an engineer the same question, as his hands were flying all over the overhead panel, while we waited to close up on stand. (BAe 146). "No he said, we use common sense".

It strikes me that a hard pressed engineer could conceivably make a mistake*, and giving the trim a twirl, sees the trim tab moving and 'confirms' that it is connected and working correctly. But crucially, it might be rigged in the opposite sense. Ditto primary flight controls. Working on many different types, it could be possible to make a mistake. Or a junior engineer, asked to confirm movement might confirm correct movement but not actually be 100% sure about the direction.

Likewise on a control check, the correct direction of movement of the flight controls needs to be positively confirmed, not just that they move freely. Several times, during control checks, the PM has said "full left" when I was holding full right. (Airbus - where you can't easily see the other pilots side-stick to sub-consciously 'cheat' in the control check).

It would be comforting to know that engineers would use a check list after work on flight controls, that specified for example; ' move the XX trim wheel in YY direction, and confirm the trim tab moves up'.

*We all make mistakes - I've made some howlers - and I mean no disrespect to any engineers.

There's a manufacturer's and company procedure to follow after disturbing anything that affects flying and engine controls. They include wire-locking, split pinning, cable tensions, tight locknuts, correct ranges of movement, operation in the correct sense and full & free. Each item is itemised on worksheets and the actions carried out are checked by two licenced engineers, with the first inspection carried out by an engineer licenced on type and the second by a licenced engineer who may - or may not be - type-rated It's a fool-proof procedure if followed correctly. If the adjustments are carried out as a line defect, the duplicate inspections are signed in the Tech Log and cross-referenced to a works order. If the rigging is carried out down-line, the captain (ATPL if I remember rightly) may be the second signatory if another engineer isn't available.
Hope that relieves you guys at the sharp end.

Note: that's the UK CAA system, other countries may have different procedures.

[HOVIS]

The short answer is that LAEs are 'supposed' to follow the manufacturer/NAA/Operator's approved data. The problem is that much of this data, including checklists is buried within mountains of superfluous catchall warnings, cautions and instructions. We don't have for example a QAR sheet to follow when powering up the aircraft or running engines.
One airline I used to work for had Engineer's Engine Run Check books. Superb, simple to follow and fool proof. The airline decided that it was too expensive to keep these updated and scrapped them. 🙄

Not following the approved data is breaking the law but sadly, especially on the line during an ever shortening turn round time, needs must. Telling the airline that there will be a delay because the approved data is a bit tricky to follow doesn't go down well. They just want an on time departure. 😕