Canadian Survey Plane Accident Albany, NY

[Jean-Pierre]

Nothing against her. I just lost friends that got nothing other than TSB choose not to investigate.

[Squaretail]

https://www.youtube.com/watch?v=89eUKcEZxgM

Juan Browne Video

I’m not sure why so many are focused on a mis set trim. There is no cause to set the trim at full, not even in landing configuration like the video suggests. That’s never been the case in thousands of hours in light aircraft, including several thousand in Navajos. The only time I have seen the trim get set maximum deflection is when maintenance has been working on it. All other times, if you flew it last, the trim for landing is really close to the take off position. There is no reason to play with it in the way suggested. Not saying it wasn’t possible, but unlikely for normal causes.

[wcpilot]

NTSB preliminary. To any Navajo drivers, what's with the pitch and roll instability? Is elevator trim an electric switch on the pilot's yoke? Any ECL for a runaway?
https://data.ntsb.gov/carol-repgen/api/ ... 194477/pdf

These things stand out to me from the report.

The pilot had 22 hours on Navajos, among 1800 TT. Given that LIDAR days in good weather can run up 8 or more hours a day, it is possible that she had only three or four days of flying with the company.

When the report say's “ The operator reported that the pilot had accrued about 22 hours of flight time in the accident airplane”. I think they're referencing how many hours she had in the actual plane that crashed. Not how many hours in type. She flew the Navajo for almost a year and a good amount of her total time was in it.

[cncpc]

https://www.youtube.com/watch?v=89eUKcEZxgM

Juan Browne Video

Yeah, the one thing that stood out in that video is that it is very uncharacteristic of Juan to make the blurt out comment that the trim had not been reset from all the trim put in in the landing.

I think the explanation that checks all the boxes is seat slide back or seat back failure. I've made the initial comments over there and now others are starting to bring it up. What happened is pretty well exactly what happens when your seat lets lose, or the back fails. I spoke with my ex investigator friend. He had two different fatalities from seat problems. However, they were not slide back, but seat back failures from fatigue failure.

The trim was not full nose up when the take off roll started, or when the second stage climb started at runway end. The airplane was flying flat to runway end at 200 feet. (I used to do that, lots of pilots do. If you do have a power loss past the runway, you are well above VMC when it happens.) But, it is that pitch up that would start the seat slide or back failure scenario if a pin hadn't seated or if something snapped in the seat adjuster for the back.

There are plausible alternative explanations for the trim actuator position in the wreckage that are consistent with the seat problem scenario, what may have been done to regain control, and cable stretching in the crash sequence.

There is a photo of Ms Gillis sitting in the Navajo. It certainly supports the seat scenario. You can see she is a tiny woman. The seat is full forward, her knees are into the switches. Even if the spar makes it a shorter slide back than a Cessna, she doesn't need to move back very much at all before her feet come off the pedals and she can't make pitch down inputs. The rolls? Consistent with her trying to pull herself back into control holding the yoke.

So sad.

[pelmet]

https://www.youtube.com/watch?v=89eUKcEZxgM

Juan Browne Video

Yeah, the one thing that stood out in that video is that it is very uncharacteristic of Juan to make the blurt out comment that the trim had not been reset from all the trim put in in the landing.

I think the explanation that checks all the boxes is seat slide back or seat back failure. I've made the initial comments over there and now others are starting to bring it up. What happened is pretty well exactly what happens when your seat lets lose, or the back fails. I spoke with my ex investigator friend. He had two different fatalities from seat problems. However, they were not slide back, but seat back failures from fatigue failure.

The trim was not full nose up when the take off roll started, or when the second stage climb started at runway end. The airplane was flying flat to runway end at 200 feet. (I used to do that, lots of pilots do. If you do have a power loss past the runway, you are well above VMC when it happens.) But, it is that pitch up that would start the seat slide or back failure scenario if a pin hadn't seated or if something snapped in the seat adjuster for the back.

There are plausible alternative explanations for the trim actuator position in the wreckage that are consistent with the seat problem scenario, what may have been done to regain control, and cable stretching in the crash sequence.

There is a photo of Ms Gillis sitting in the Navajo. It certainly supports the seat scenario. You can see she is a tiny woman. The seat is full forward, her knees are into the switches. Even if the spar makes it a shorter slide back than a Cessna, she doesn't need to move back very much at all before her feet come off the pedals and she can't make pitch down inputs. The rolls? Consistent with her trying to pull herself back into control holding the yoke.

So sad.

I am not sure that the typical seat slide back also results in electric trim operation at the same time. I suppose anything is possible.

[Squaretail]

https://www.youtube.com/watch?v=89eUKcEZxgM

Juan Browne Video

I think the explanation that checks all the boxes is seat slide back or seat back failure.

I am not sure that the typical seat slide back also results in electric trim operation at the same time. I suppose anything is possible.

First a slide back in the normal pa31, there is not a ton of travel. The spar sits directly behind the short seat rails. This it should be noted is different in the chieftain (not the model in the accident) where the spar sits farther back. In flight if the airplane is trimmed this should also be a non event, unless the pilot hangs on to the control yoke. If that was the case there would have been a pitch up and it would have stalled/ spun the rest of the way down. However if auto was engaged it would have fought this and rolled all the way nose down, which isnt the case.

A seat break should have been obvious to the investigators, especially a seat back failure. When i survived one of these the back was in the aft compartment when the rodeo was done. Flying the airplane like you are sitting on a milk crate is difficult, but not impossible.

In the short video of the airplane crashing it looks like the airplane is stalled with the pilot fighting to prevent a wing drop. I still think this is either an autopilot fail or a mishandling. It would be interesting to know if the company aircraft all have the same auto or not, and how much training is dedicated to systems differences.

[Donald]

Many moons ago when I flew a PA31-350, I did a maint flight where an avionics tech attempted to fix the autopilot. With it engaged, it started oscillating in pitch, in increasing severity. It got to the point where it was going from 20-30 degrees nose up to same nose down. It doesn't explain the barrel roll, but it does make me suspect an autopilot issue. Low time pilot, low time on type, maybe a failure they weren't trained on.

[CpnCrunch]

Many moons ago when I flew a PA31-350, I did a maint flight where an avionics tech attempted to fix the autopilot. With it engaged, it started oscillating in pitch, in increasing severity. It got to the point where it was going from 20-30 degrees nose up to same nose down. It doesn't explain the barrel roll, but it does make me suspect an autopilot issue. Low time pilot, low time on type, maybe a failure they weren't trained on.

I struggle to see how an autopilot problem in day VFR could cause this. I think everyone has had an experience of the autopilot not doing what's it's supposed to do, and turning it off.

[Donald]

Many moons ago when I flew a PA31-350, I did a maint flight where an avionics tech attempted to fix the autopilot. With it engaged, it started oscillating in pitch, in increasing severity. It got to the point where it was going from 20-30 degrees nose up to same nose down. It doesn't explain the barrel roll, but it does make me suspect an autopilot issue. Low time pilot, low time on type, maybe a failure they weren't trained on.

I struggle to see how an autopilot problem in day VFR could cause this. I think everyone has had an experience of the autopilot not doing what's it's supposed to do, and turning it off.

I tend to agree, however startle factor, confusion, disbelief...and you can't discount the low experience.

Just look at the YT poster FlyTN that crashed after most likely getting distracted by an autopilot she knew was faulty.

[I WAS Pez]

Many moons ago when I flew a PA31-350, I did a maint flight where an avionics tech attempted to fix the autopilot. With it engaged, it started oscillating in pitch, in increasing severity. It got to the point where it was going from 20-30 degrees nose up to same nose down. It doesn't explain the barrel roll, but it does make me suspect an autopilot issue. Low time pilot, low time on type, maybe a failure they weren't trained on.

I struggle to see how an autopilot problem in day VFR could cause this. I think everyone has had an experience of the autopilot not doing what's it's supposed to do, and turning it off.

A pitch trim runaway, either on its own, or caused by an AP fault could have contributed.... Depending on failure mode, disconnecting AP with normal disconnect switch may not have stopped it. Too early to know for sure, and indeed we may never know - but it seems plausible to me.

[cncpc]

Just look at the YT poster FlyTN that crashed after most likely getting distracted by an autopilot she knew was faulty.

If I remember correctly, that was part of the mix in that Aerostar crash on approach to Nanaimo a couple of years back.

[Squaretail]

Many moons ago when I flew a PA31-350, I did a maint flight where an avionics tech attempted to fix the autopilot. With it engaged, it started oscillating in pitch, in increasing severity. It got to the point where it was going from 20-30 degrees nose up to same nose down. It doesn't explain the barrel roll, but it does make me suspect an autopilot issue. Low time pilot, low time on type, maybe a failure they weren't trained on.

I struggle to see how an autopilot problem in day VFR could cause this. I think everyone has had an experience of the autopilot not doing what's it's supposed to do, and turning it off.

Have they though? This is only something a lot of PIC time buys you. In my decades in aviation, I've had a genuine auto problem twice. That is to say, an actual failure. One of those also had the fun of a faulty disconnect switch. In those cases i had more altitude to play with, but lets say i failed to keep the +/- 100' while fighting with it. Sub 1000' in a lower energy state J can easily see this being a fatal problem. You'll jam that dc switch a dozen times before you figure it out. Meanwhile its going to take all your strength to fly the plane, because until you get auto disconnected, you cant reset the trim.

Now in a normally functioning autopilot, if the trim maxes out, and it cant hold the condition its in, it usually kicks off dumping the fully trimmed airplane in your lap like a hot potato. This is the part that i think happened. AP is engaged soon after take off with a mis set condition it cant do, but is going to try, an overly optimistic rate of climb for instance. Maybs they hit the VS button when they wanted IAS, who knows. Either way the simple computer doesnt know it cant do what it cant do. Most old basic APs, just know if they need more rate, it needs more pitch. It knows nothing about stall angles. So the trim could get maxed out really quickly. A smaller pilot perhaps isnt touching the trim wheel all the time and doesnt know its moving. Old autopilots dont have an audible "trim in motion" when its probably doing something you dont want.

So there are both pilot mishandling and actual auto malfunctions i can think of and permutations of that are possible contributing factors here. In training new pilots, I have seen many pilots not recognize the danger the automation can put them in until the airplane is in a dangerous state.

I also have to say the "barrel roll" probably didnt happen, thats just what some observers who arent pilots thought happened. You would need a lot of speed in the Navajo to execute one, and more altitude. I mean its relatively nimble, but not THAT nimble. More likely it did a vmc roll/spin on its initial pitch up, manually working the trim wheel, recovering would have taken some doing. Likely you would have had to pitch the plane down towards the ground to regain speed at a sporty attitude to have made the recovery. Something pilots are disinclined to do at low altitude.

[CaptDukeNukem]

My personal inclination on this one is a mistrim. At rotation, the forces required to fly the aircraft would have been extremely high, combined with human factors such as startle, and frankly lack of muscle… this airplane found its way to the ground. Bless her soul.

[rookiepilot]

Perhaps N/A, but from a GA POV, I have always refrained from engaging AP until 1000 AGL, usually more. I don't trust them.

[Old fella]

My personal inclination on this one is a mistrim. At rotation, the forces required to fly the aircraft would have been extremely high, combined with human factors such as startle, and frankly lack of muscle… this airplane found its way to the ground. Bless her soul.

+1

[cncpc]

My personal inclination on this one is a mistrim. At rotation, the forces required to fly the aircraft would have been extremely high, combined with human factors such as startle, and frankly lack of muscle… this airplane found its way to the ground. Bless her soul.

This is not a mistrim prior at takeoff. That is an insult to the pilot.

[CaptDukeNukem]

My personal inclination on this one is a mistrim. At rotation, the forces required to fly the aircraft would have been extremely high, combined with human factors such as startle, and frankly lack of muscle… this airplane found its way to the ground. Bless her soul.

This is not a mistrim prior at takeoff. That is an insult to the pilot.

Okay… you must be god then if you’ve never made a mistake. I say mistrim, you say seat slide. We’re each allowed to have our opinions.

[cncpc]

I also have to say the "barrel roll" probably didnt happen, thats just what some observers who arent pilots thought happened. You would need a lot of speed in the Navajo to execute one, and more altitude. I mean its relatively nimble, but not THAT nimble. More likely it did a vmc roll/spin on its initial pitch up, manually working the trim wheel, recovering would have taken some doing. Likely you would have had to pitch the plane down towards the ground to regain speed at a sporty attitude to have made the recovery. Something pilots are disinclined to do at low altitude.

Yes, I also wouldn't be turning the autopilot on at 200 feet in a control zone. Even at 1000 feet. I'd want my feet on those rudder pedals in case one calved at 200 feet. Not a time to stop being a pilot. I can check my phone at cruise.

I agree to some extent with Squaretail on the potential for barrel rolls to be observer misconceptions of what is really the vertical part of spin development. But the whole picture doesn't really support that, although it may well be that what was reported is not what a pilot would report.

The trim all the way back at takeoff is no longer a credible theory. Trim running back in the air, runaway autopilot, etc are possible because the trim IS full nose up in the wreckage. When that happened is not known, as is the initiating action. But, there is a finality about that, if once full up, it stayed there.

At the end the aircraft is making 94 knots groundspeed. It has gone 1.5 miles from the airport. It has engaged in aerobatic like maneuvers. None of that is caused by a takeoff with full nose up trim.

[cncpc]

My personal inclination on this one is a mistrim. At rotation, the forces required to fly the aircraft would have been extremely high, combined with human factors such as startle, and frankly lack of muscle… this airplane found its way to the ground. Bless her soul.

This is not a mistrim prior to takeoff. That is an insult to the pilot.

Okay… you must be god then if you’ve never made a mistake. I say mistrim, you say seat slide. We’re each allowed to have our opinions.

I'm not a god. I have made mistakes. There is not a rational equivalence between a mistrim (meaning full nose up for takeoff, and not a trim issue later in flight) and seat slide. If you've been reading the many good posts here, you'd know the following.

The trim all the way back at takeoff is no longer a credible theory. No one has bebunked the seat slide theory. It checks every box when explaining what happened, beginning with the pitch up. Squaretail has offered that the Ho seat doesn't slide back very far. That's it.

Trim running back in the air, runaway autopilot, etc are possible because the trim IS full nose up in the wreckage. When that happened is not known, as is the initiating action. But, there is a finality about that, if once full up, it stayed there.

At the end the aircraft is making 94 knots groundspeed. It has gone 1.5 miles from the airport. It has engaged in aerobatic like maneuvers. None of that is caused by a takeoff with full nose up trim.

Like many folks here, I contribute to accident discussions. When circumstances require it, I take an approach that is meant to fight the pilot's corner. I'm sure you didn't mean it to be an insult, but you would mean it if you repeated the same claim about "mistrim".

[CaptDukeNukem]

This is not a mistrim prior to takeoff. That is an insult to the pilot.

Okay… you must be god then if you’ve never made a mistake. I say mistrim, you say seat slide. We’re each allowed to have our opinions.

I'm not a god. I have made mistakes. There is not a rational equivalence between a mistrim (meaning full nose up for takeoff, and not a trim issue later in flight) and seat slide. If you've been reading the many good posts here, you'd know the following.

The trim all the way back at takeoff is no longer a credible theory. No one has bebunked the seat slide theory. It checks every box when explaining what happened, beginning with the pitch up. Squaretail has offered that the Ho seat doesn't slide back very far. That's it.

Trim running back in the air, runaway autopilot, etc are possible because the trim IS full nose up in the wreckage. When that happened is not known, as is the initiating action. But, there is a finality about that, if once full up, it stayed there.

At the end the aircraft is making 94 knots groundspeed. It has gone 1.5 miles from the airport. It has engaged in aerobatic like maneuvers. None of that is caused by a takeoff with full nose up trim.

Like many folks here, I contribute to accident discussions. When circumstances require it, I take an approach that is meant to fight the pilot's corner. I'm sure you didn't mean it to be an insult, but you would mean it if you repeated the same claim about "mistrim".

I have flown airplanes where pre-flight checks included manually running trim full up and full down manually. All it takes is a simple interruption.

But seat movement is of course very plausible.

[cncpc]

I have flown airplanes where pre-flight checks included manually running trim full up and full down manually. All it takes is a simple interruption.

But seat movement is of course very plausible.

Sure thing. As we've heard here, there are some here who have stated that was their practice. It was mine. Full forward, three back. Not sure it was that detailed on the checklist, but certainly a trim set was. There is nothing to say she didn't. All the other causes are still on the table. I doubt any of them will reflect poorly on this excellent young woman. I believe she was fighting to control her aircraft right to the end. One possible reading of that awful two seconds of video was that she almost had it, but ran out of altitude.

Peace.

[CpnCrunch]

Sure thing. As we've heard here, there are some here who have stated that was their practice. It was mine. Full forward, three back. Not sure it was that detailed on the checklist, but certainly a trim set was. There is nothing to say she didn't. All the other causes are still on the table. I doubt any of them will reflect poorly on this excellent young woman. I believe she was fighting to control her aircraft right to the end. One possible reading of that awful two seconds of video was that she almost had it, but ran out of altitude.

Peace.

Most GA autopilots have a pre-flight procedure which must be done prior to use. Whether that would have detected this issue, and whether pilots actually do the autopilot preflight is another thing. I would say that close to 100% of small planes I've flown with autopilots, the autopilot hasn't been working properly at some point in the time I've been flying the plane...albeit usually just crapped out servos which either don't do anything useful or porpoise, so easy enough to deal with.

[pelmet]

I am curious to know how anybody has been able to eliminate the possibility that the trim was not at full aft setting for takeoff.

It may not be a normal place for it but it could have been moved there for some reason.

[cncpc]

I am curious to know how anybody has been able to eliminate the possibility that the trim was not at full aft setting for takeoff.

It may not be a normal place for it but it could have been moved there for some reason.

As soon as the elevator became effective, the nose would have risen and continued to rise. The aircraft took off and flew over 6000 feet in level flight down the runway without any problems that would have been very evident almost immediately if that were the true scenario.

For that to happen, full nose up trim would have to have been applied in the previous landing. Generally, the Navajo does not require that, as a number of Navajo pilots have stated on here.

Full nose up trim does not cause "barrel rolls".

[digits_]

I am curious to know how anybody has been able to eliminate the possibility that the trim was not at full aft setting for takeoff.

It may not be a normal place for it but it could have been moved there for some reason.

As soon as the elevator became effective, the nose would have risen and continued to rise. The aircraft took off and flew over 6000 feet in level flight down the runway without any problems that would have been very evident almost immediately if that were the true scenario.

For that to happen, full nose up trim would have to have been applied in the previous landing. Generally, the Navajo does not require that, as a number of Navajo pilots have stated on here.

Full nose up trim does not cause "barrel rolls".

Would that not be exactly what you'd expect? Plane accelerates, wants to nose up. Pilot holds it down, tries to build up speed to avoid stalling, while wondering what is going on. As the speed increases, the force become too big to handle, plane climbs, noses up, stalls, VMC rolls and crashes?