BOEING 767F crash near Houston in the bay Atlas cargo

[Tantalum]

Does the TOGA button do anything to the FD or AP..?

 

[Arnold]

In the aircraft I have flown it disconnects the AP and sets the FD bars to wings level and a climb attitude.

 

[Greg Bockelman]

Well, those certainly look like TOGA buttons. I don’t have time in the 757/767 but it makes sense. Bottom line is it looks to me like it would be real difficult to hit the TOGA buttons accidentally no matter what airframe is involved;.

 

[Tarheelpilot]

Perhaps the FO mistook the negative Gs for a stall?.. assumed it was the plane falling and was attempting to break the stall?

*Would the outcome have been different in an Airbus? Airbus has envelope and "inappropriate" attitude protection, as well as I believe the ability to just select which stick is priority.. and in the case of opposing commands (without priority pressed) I believe it averages the two? Or would it have been the same as this / Air France

*The deck certainly seems to point to a failure of the FO.. but how was the captain, and the the jumpseater, so far behind the airplane?

..and yes, like someone else said, at least this was just 3 people in a bay, and not 200 in a city

The negative g was a direct result of him shoving the yoke forward. The guys behavior and lack of understand is what we call incompetence.

 

[Fearless Tower]

Wow. If that’s true, that’s a horrible design of the TOGA button.

Is this they? Sorry, no, this is the 777 I think.

That is the 777 not a 767.

Go to the Factual Report, which is the first item in the docket, for photos of the 767 throttle quadrant which shows the GA switches on the backside of the throttles (page 66). The NTSB 767 photos also go into detail about the location and travel of the speed brake lever and its proximity to the left throttle GA switch.

The factual report also goes into detail on the 767 GA and AP systems. GA is auto-enabled when the airspeed is below 250 and flaps are out of zero. So, when they selected flaps 1, the GA system was armed.

While in the descent to 3000', ATC told them to expedite at which point the speedbrakes were deployed. 4 seconds after the speed brakes were stowed, GA activated. So, speculation is that whoever moved the lever to stow speedbrakes bumped the left GA switch.

Note: nothing is said or noticed on the CVR that the engines were spooling up and increasing to GA power.

There is currently some discussion taking place on a pro-pilot board that the FO may have been disoriented by the nose coming up and sensation of acceleration (possibly combined with an issue with his flight display) and he interpreted that as the airplane was stalling and over-reacted Colgan style and forced the nose down.

Meanwhile, the CA is sitting next to him, hearing him say 'stalling', looking at his flight display and probably asking himself: 'WTF is going on?' about the same time they go to negative 1 G and the CA is still trying to figure out what is happening. The FO freezes and the CA pulls back on the yoke as they break through the clouds and are staring down at the ground.

What happened, happened fast.

 

[Fearless Tower]

Copilot was PF. He called for flaps, captain reached around the throttle quadrant to extend the flaps and accidentally hot the TOGA button.

Not the flaps, but the speed brake.

 

[Larry in TN]

Does the TOGA button do anything to the FD or AP..?

Yes, of course.

Autothrottle, vertical and lateral modes all go to "GA". Vertical and lateral modes are for both the FDs and APs.

Autothrottles in GA will set power for a ~2000 fpm climb

Vertical mode pitches for an airspeed though I don't recall what airspeed that is. Probably something in the Vref+20 range.

Lateral mode maintains current track.

You fix an inadvertent TOGA activation by disconnecting the autopilot and autothrottle, reducing thrust as appropriate, calling for the PM to cycle the F/D switches (necessary to get out of GA mode), then calling for the appropriate vertical and lateral modes. Once established, autothrottles and autopilot(s) can be re-engaged as needed.

 

[Tantalum]

Yes, of course.

Autothrottle, vertical and lateral modes all go to "GA". Vertical and lateral modes are for both the FDs and APs.

Autothrottles in GA will set power for a ~2000 fpm climb

Vertical mode pitches for an airspeed though I don't recall what airspeed that is. Probably something in the Vref+20 range.

Lateral mode maintains current track.

You fix an inadvertent TOGA activation by disconnecting the autopilot and autothrottle, reducing thrust as appropriate, calling for the PM to cycle the F/D switches (necessary to get out of GA mode), then calling for the appropriate vertical and lateral modes. Once established, autothrottles and autopilot(s) can be re-engaged as needed.

Thanks.. that's what I was wondering, even in the G1000 Perspective, while you won't have auto throttle, the FD (and AP if on) will command a straight ahead climb, generally somewhere around Vy-ish.. I don't know the exact figure. It's very perplexing how an inadvertent TOGA application would cause a stall, full blown panic, and subsequent crash

to include taking off with the speed brakes on

Certain professional exams have a max number of attempts and a required time frame between attempts. I always found it interesting that aviation allowed so much leeway. I think it's safe to say that at some point if someone messes up X amount of times on egregious things they shouldn't be flying, at least not commercially

 

[MFE]

Certain professional exams have a max number of attempts and a required time frame between attempts. I always found it interesting that aviation allowed so much leeway. I think it's safe to say that at some point if someone messes up X amount of times on egregious things they shouldn't be flying, at least not commercially

I mean, you read the evaluators' comments, in some cases following the same themes multiple times, and it would be hard NOT to predict this guy would be at the controls during a crash.

- Tunnel Vision
- Regressing Situational Awareness
- Trying too hard/going too fast/gets ahead of himself
- Improper settings and procedures
- Poor airmanship when not in auto modes (Exceeding limits in steep turns, difficulty maintaining speed, difficulty maintaining altitude)
- Incomplete/Improper communication in cockpit during other-than-normal situations

I mean, what could go wrong?

 

[Sinistar]

What does it mean in the report that the AP was on (or engaged) at the time of impact. Is this like a full on all 3 axis auto pilot? Or just this just mean the TOGA? Its hard to imagine a full AP when they are muscling the control columns like that.

Also just to understand better. All of this until about 2000msl was in IMC?

 

[Fearless Tower]

What does it mean in the report that the AP was on (or engaged) at the time of impact. Is this like a full on all 3 axis auto pilot? Or just this just mean the TOGA? Its hard to imagine a full AP when they are muscling the control columns like that.

Also just to understand better. All of this until about 2000msl was in IMC?

In the factual, there is a statement from Boeing that the A/P will remain engaged even with full force applied to the control column. So basically the A/P remained engaged up to the point of impact, but as long as the force was being applied to the control column, it was overriding the system.

 

[dmspilot]

If we're going to ask questions or post comments on documents in the public docket, can we please specify which document we are talking about? There are 3,000 pages of info here, including 9 "factual reports" from different working groups.

 

[Larry in TN]

It's very perplexing how an inadvertent TOGA application would cause a stall, full blown panic, and subsequent crash

It didn't cause a stall. The airplane did not stall. The acceleration and pitch-up from inadvertent GA mode activation made the F/O believe they stalled. He responded by pushing the nose down as far as 49° AND.

 

[35 AoA]

There is currently some discussion taking place on a pro-pilot board that the FO may have been disoriented by the nose coming up and sensation of acceleration (possibly combined with an issue with his flight display) and he interpreted that as the airplane was stalling and over-reacted Colgan style and forced the nose down.

To be fair to FO, "Colgan style" would mean he perceived a stall and hauled *back* on the controls to full throw all the way until impact, i.e. "Air France style". If he thought it was a stall, he was at least correcting in the right direction, though in many magnitudes more than were reasonable.

 

[kyleb]

To be fair to FO, "Colgan style" would mean he perceived a stall and hauled *back* on the controls to full throw all the way until impact, i.e. "Air France style". If he thought it was a stall, he was at least correcting in the right direction, though in many magnitudes more than were reasonable.

Imagine: You think the airplane stalled. So you push forward, looking for the airplane to regain lift and push you down in the seat. But, the airplane never stalled, and since you're pushing hard enough that the airplane is at fractional G's, you never get the seat of the pants "lift restored" sensation. So you continue to push... Until it is way too late. What a nightmare.

 

[35 AoA]

I agree that what the guy got was not what he was expecting, but this whole scenario is just beyond bizarre. Even more bizarre than the UCFIT (new acronym) events of Colgan or AF.

 

[denverpilot]

Doesn’t fit. It takes two conscious pilots going in opposite directions to shear the interconnect. In my opinion. As much as I hate to say it i think this is pretty clear.

Technically one could be slumped over but not likely. The documents say that only 25 lbs will shear it.

The docs are also claiming the TOGA activation triggers a visual annunciation but no mention of what the Board thinks the numerous audible warnings are yet. I assume an in flight TOGA activation includes an audible? Anybody know?

Mostly just curious how the Captain would miss it, no matter who hit the thing.

NTSB is leaning pretty heavily on the acceleration illusion, but even as low time as the right seater was there’s no way everyone in that cockpit hadn’t experienced the leans before. That their eyeballs would (or damn well should!) go straight to the AI and the ASI speed tape right next to it seems a no brainer for me.

So the “where did my speed go?” strikes me as serious.

Bad pilots or not, a trend toward 49 degrees nose down is usually noticed by someone. The usual CVR if not a formal call out about it is something along the lines of “You’re losing it!” if they’re not taking the aircraft.

Think the Captain got fooled by the straining right seater into thinking he was also pulling and the aircraft wouldn’t recover so he thought he was joining him instead of fighting him?

You have to say it. “My airplane.” “Nose up!” Something other than just pull.

Al Haynes always joked those were the dumbest three words of his career: “I’ve got it.”

He didn’t ... but his FO knew exactly what he was doing.

Maybe all the thuds in the CVR is crap flying around and somebody got clocked in the head or just whacked by something enough that it became a distraction.

 

[Tantalum]

It didn't cause a stall. The airplane did not stall. The acceleration and pitch-up from inadvertent GA mode activation made the F/O believe they stalled. He responded by pushing the nose down as far as 49° AND.

Right, sorry, the way I wrote that was misleading, I meant that more from the perspective of the first officer..

it's amazing how you can work your way up to the flight deck of a proper large airplane like that and still get that kind of severe tunnel vision and loss of CRM; it's insanity

 

[Sinistar]

But there is another piece to the puzzle...the FO indicated that he thought he had a PFD failure. Maybe it hadn't failed but just didn't match what he was feeling. Seems like some time (not sure how much) was spent during the final minute or two regarding the display issue and perceived (or actual) failure.

 

[35 AoA]

I think they are grasping at what few straws they have to work with. Because this mishap makes no logical sense from an outsider's perspective. I agree that the PFD/ND/whatever failure probably had something significant to do with the FO's perception of reality....whether it meant he didn't trust the truth data he was looking at (due to previous failure), or conversely that he was looking at a failed display that was giving bogus info and didn't realize it. Either scenario would at least start to explain the situation they found themselves in.

 

[Sinistar]

Wow, just think if the display failed in a way that mislead them into thinking it was stalling. Or if a input sensor into the display lead to "false" stall indication. In that case (except cross checking with other instruments) the FO may have been doing the right thing and excessive pitch down didn't fix it nor did the lack of normal G. Its too bad it was IMC or night or whatever didn't let them just look outside, see the horizon and do quick WTF and level it off. This could actually be a case (albiet rather remote) where a FO with a less than stellar training record was actually doing the right thing given his instruments. Wow, if this was another Boeing instrumentation issue...wow. But this case is so quiet. If I mention this accident to anyone outside of aviation and even several pilots they have never heard of it.

 

[Tantalum]

Pushing full forward on the control column for an extended amount of time should tell you that something is severely wrong.. especially if you combine that with engines at 110% and the obvious loud sound of air rushing past the cockpit

The guy, for whatever reason, seems like he just froze up.. we weren't there but that's certainly what it seems like happened

 

[Tarheelpilot]

Wow, just think if the display failed in a way that mislead them into thinking it was stalling. Or if a input sensor into the display lead to "false" stall indication. In that case (except cross checking with other instruments) the FO may have been doing the right thing and excessive pitch down didn't fix it nor did the lack of normal G. Its too bad it was IMC or night or whatever didn't let them just look outside, see the horizon and do quick WTF and level it off. This could actually be a case (albiet rather remote) where a FO with a less than stellar training record was actually doing the right thing given his instruments. Wow, if this was another Boeing instrumentation issue...wow. But this case is so quiet. If I mention this accident to anyone outside of aviation and even several pilots they have never heard of it.

Wow. What if it really was just pilot error.
Oh wait. We will still argue on this board.

 

[Rushie]

That is the 777 not a 767.

Go to the Factual Report, which is the first item in the docket, for photos of the 767 throttle quadrant which shows the GA switches on the backside of the throttles (page 66). The NTSB 767 photos also go into detail about the location and travel of the speed brake lever and its proximity to the left throttle GA switch.

The factual report also goes into detail on the 767 GA and AP systems. GA is auto-enabled when the airspeed is below 250 and flaps are out of zero. So, when they selected flaps 1, the GA system was armed.

While in the descent to 3000', ATC told them to expedite at which point the speedbrakes were deployed. 4 seconds after the speed brakes were stowed, GA activated. So, speculation is that whoever moved the lever to stow speedbrakes bumped the left GA switch.

Note: nothing is said or noticed on the CVR that the engines were spooling up and increasing to GA power.

There is currently some discussion taking place on a pro-pilot board that the FO may have been disoriented by the nose coming up and sensation of acceleration (possibly combined with an issue with his flight display) and he interpreted that as the airplane was stalling and over-reacted Colgan style and forced the nose down.

Meanwhile, the CA is sitting next to him, hearing him say 'stalling', looking at his flight display and probably asking himself: 'WTF is going on?' about the same time they go to negative 1 G and the CA is still trying to figure out what is happening. The FO freezes and the CA pulls back on the yoke as they break through the clouds and are staring down at the ground.

What happened, happened fast.

Thanks, found it! It’s hard to see how he could have accidentally hit the TOGA switch while going for the speed brake.

 

[Tantalum]

for what it's worth, the 767 has always felt like a beefy airplane to me, I'm impressed that this thing could handle over 400 knots indicated and positive 4 g's force without something breaking off.. nuts!

I had read anecdotal evidence somewhere that during testing Boeing did not bring the wing to destruction, they got to something like 168% but never actually broke the wing

 

[Fearless Tower]

Thanks, found it! It’s hard to see how he could have accidentally hit the TOGA switch while going for the speed brake.

The speed brake handle when deployed is all the way aft behind the throttles. Most likely scenario is captain goes to stow the speed brakes by grabbing the speed brake handle, it’s turbulent, and while pushing it forward the pinky knuckle bumps that TOGA switch.

 

[Rushie]

At 12:33:34.4 the captain says, “I don’t trust this thing” and then “I don’t trust all this automatic flight...” and later “I just clicked it off and flew it.” I’m not sure what he was talking about, but here is what bothers me about the FO: He was flying the plane, in IMC, and no matter what happened, should have been able to trust the instruments and fly it - make whatever correction was necessary, via the instruments, when something unexpected happened. So unless there was instrument failure (what did he mean where’s my speed?), where were his hand flying skills on instruments in IMC? Is it ironic that the captain had just bemoaned planes flying themselves, only to be done in by a guy that maybe could have used more hand flying in IMC?

 

[Rushie]

The speed brake handle when deployed is all the way aft behind the throttles. Most likely scenario is captain goes to stow the speed brakes by grabbing the speed brake handle, it’s turbulent, and while pushing it forward the pinky knuckle bumps that TOGA switch.

Ah yes, I can see that. Wow.

 

[Tantalum]

“I don’t trust this thing” and then “I don’t trust all this automatic flight...” and later “I just clicked it off and flew it.”

The TOGA and AP pieces are starting to fit, somehow it was activated, they were unaware of it, they were in IMC and turbulence, airplane wasn't behaving as they thought it would, assumed it was an airplane issue, got behind the airplane and made an "oops" situation catastrophic

 

[Larry in TN]

At 12:33:34.4 the captain says, “I don’t trust this thing” and then “I don’t trust all this automatic flight...” and later “I just clicked it off and flew it.” I’m not sure what he was talking about

He was talking about VNAV. They indicate that he didn't understand VNAV very well.

 

[Kenny Phillips]

Did you read the FOs training history?

This was Colgan 3407 all over again, but without a plane full of paying pax.

The Colgan thing was more a correct response to the wrong aircraft, then not realizing it in time. Peter Garrison wrote a nice piece about it.

 

[sourdough44]

All these pages and it seems to come down to screwed up piloting. I didn’t refresh on all the particulars. I guess in a more perfect world, the Cap would of taken over sooner.

 

[Doc Holliday]

Does the TOGA button do anything to the FD or AP..?

 

[flyingcheesehead]

Copilot was PF. He called for flaps, captain reached around the throttle quadrant to extend the flaps and accidentally hot the TOGA button. They were in IMC. Copilot took the sudden acceleration as a pitch up and rammed the yoke forward to the stops. Captain hauled back on his side (no positive exchange of controls) and broke the shear pin. Now they have a split elevator. Never retarded the throttle. By the time they exit IMC at 2,000 feet its too late.

Yep, sounds very plausible... But if the instruments were all working, how could they screw up so bad? Never trust the seat of your pants...

*Would the outcome have been different in an Airbus? Airbus has envelope and "inappropriate" attitude protection, as well as I believe the ability to just select which stick is priority.. and in the case of opposing commands (without priority pressed) I believe it averages the two? Or would it have been the same as this / Air France

Yup... The Airbus would have started downhill when the FO pushed, and when the captain pulled it would have averaged out to neutral. End result is the same.

*The deck certainly seems to point to a failure of the FO.. but how was the captain, and the the jumpseater, so far behind the airplane?

This is what blows my mind. If he did anything, it was too late... And he never said a word.

for what it's worth, the 767 has always felt like a beefy airplane to me, I'm impressed that this thing could handle over 400 knots indicated and positive 4 g's force without something breaking off.. nuts!

I had read anecdotal evidence somewhere that during testing Boeing did not bring the wing to destruction, they got to something like 168% but never actually broke the wing

I found that very hard to believe - If it got to 168%, it's too stout and useful load could be higher with a lightened structure. That's why when you watch the videos of the ultimate load failure tests, when they get to 150% all the beancounters and executives cheer, while when it breaks in the low 150s all the engineers cheer because they've done their job right in both directions.

So, I went looking. There isn't a lot of great information, but it sounds like for the 767 test Boeing attached the wings to a non-representative fuselage, and the fuselage failed when the wing was at 99.4% of ultimate load (149.1% of limit load) and the test was terminated. Surprisingly, it sounds like they never did complete that test to the full 150% of limit load. I guess the FAA must have said "close enough for government work."

 

[denverpilot]

I bet NTSB recommends video cameras out of this one. The tech is there now to record and store it in solid state.

The culture doesn’t want it. It won’t happen yet, but a video would tell this story faster and nowadays, cheaply.

If it’s not a recommendation that large transport category aircraft start to add it, in their final report, I’ll be surprised. Since it’s their job to push FAA.

 

[Tantalum]

Yep, sounds very plausible... But if the instruments were all working, how could they screw up so bad? Never trust the seat of your pants...



Yup... The Airbus would have started downhill when the FO pushed, and when the captain pulled it would have averaged out to neutral. End result is the same.



This is what blows my mind. If he did anything, it was too late... And he never said a word.



I found that very hard to believe - If it got to 168%, it's too stout and useful load could be higher with a lightened structure. That's why when you watch the videos of the ultimate load failure tests, when they get to 150% all the beancounters and executives cheer, while when it breaks in the low 150s all the engineers cheer because they've done their job right in both directions.

So, I went looking. There isn't a lot of great information, but it sounds like for the 767 test Boeing attached the wings to a non-representative fuselage, and the fuselage failed when the wing was at 99.4% of ultimate load (149.1% of limit load) and the test was terminated. Surprisingly, it sounds like they never did complete that test to the full 150% of limit load. I guess the FAA must have said "close enough for government work."

Like I said, what I wrote was anecdotal and I too found very little details when I dug around

Thanks

 

[Doc Holliday]

Yup... The Airbus would have started downhill when the FO pushed, and when the captain pulled it would have averaged out to neutral. End result is the same.

But there is a sidestick priority button (red) on the Airbus sidestick controller which over rides the input of the opposite stick.

 

[G-Man]

So, the CEO is out.*
Replaced by the Chairman of the Board. Not quite sure this is a harbinger of sweeping change or a guarantee of more of the same culture?

*Keeping pension? Last few years of paychecks? Stock options? Etc.

 

[hindsight2020]

Heck, for those kind of compensation packages, I'd be making a living at getting fired.

 

[denverpilot]

Heck, for those kind of compensation packages, I'd be making a living at getting fired.

Many of them do. Or worse.