Flybe AP has a 'Plummet' Mode?

[fenland787]

As I was only an occasional inhabitant of the Pointy End during flight my knowledge of Autopilot setting is limited to what I had to do to 'fly' the lab in Everett but I don't recall being able to set 'Plummet' mode?
(Edit - looks like the Beeb have decided 'drops' might be better than the original 'plummet' - they could be right but descends could be even better?)

BBC News (https://www.bbc.co.uk/news/uk-46137445)

[PAXboy]

As it is reported that the pilot responded to the alarm and disconnected the autopilot = Job Done! OK, there was an error of setting but, despite an uncomfortable dip, the pilots flew the aeroplane. Another useless media item picking something from 11 months ago. However, as if warning were needed, every carrier has to be ready with the correct story and to present it quickly and fairly.

[DaveReidUK]

(Edit - looks like the Beeb have decided 'drops' might be better than the original 'plummet' - they could be right but descends could be even better?)

Hmmm. At 1100 ft AGL, ROD was 4300 fpm, one can perhaps understand the BBC's choice of words, however misguided.

picking something from 11 months ago.

Picking up something from an investigation report released a few hours ago.

[Kerosine]

Mistake was made, very little actual risk unless pilots were asleep, looks good as a headline. Nothing to see here.

[WIDN62]

The AAIB thought there was something to see!

[jimjim1]

https://www.gov.uk/government/publications/air-accident-monthly-bulletin-november-2018

Direct link to Hi-Resolution pdf
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/749117/Bulletin_11-2018_Hi_Res.pdf
30MB file. Low res one does not display diagrams well.

DHC-8-402 Dash 8 G-ECOE 11-Jan-18

IMC, autopilot configured and then selected so that aircraft would crash into the ground/sea in 54 seconds unless corrected by crew. Crew oblivious until ...

EGPWS warnings alerted crew who reacted with 36 seconds to spare. Airspeed had been 163 KIAS in climb and reached 235 KIAS in unplanned and unexpected descent. [these times are too generous. Less time was actually available]

After recovery they tried AP again with same settings.

Hardly a benign little excursion I wouldn't have thought.

https://cimg5.ibsrv.net/gimg/pprune.org-vbulletin/695x639/g_ecoe_c299763fe4ebd0ce2d243b70e9397b4318582961.png

ALtitude curve interesting

[25F]

SLF here so please excuse if these are dumb questions - but why is there not some sort of sensible minimum altitude for the autopilot? And why isn't the EGPWS connected to the autopilot to disconnect it or override it and start pulling up?

[reverserunlocked]

It’s a Q400. At that point you stop asking sensible questions and just shrug. The Q400 has a lot of oddities and the autopilot is one of them.

[ajamieson]

"Staff reported that the plane had become visual with the ground."

In much the same way that a drunk becomes visual with the ceiling?

[Aso]

"Staff reported that the plane had become visual with the ground."

OMG Johnson! Did you see that? There is a ground underneath us!

[pinkpanther1]

As I was only an occasional inhabitant of the Pointy End during flight my knowledge of Autopilot setting is limited to what I had to do to 'fly' the lab in Everett but I don't recall being able to set 'Plummet' mode?
(Edit - looks like the Beeb have decided 'drops' might be better than the original 'plummet' - they could be right but descends could be even better?)

BBC News (https://www.bbc.co.uk/news/uk-46137445)

For takeoff the Flight director in the Q400 is set with the Vertical Mode ; GA (Go around) and the lateral mode HDG SEL (heading select). Further the assigned altitude for departure is entered with the ALT SEL (alt select) mode armed. In the Flybe case the crew forgot to input the assigned altitude and left it at 0ft. When they armed ALT SEL the flight director captured the assigned altitude of 0ft (as the aircraft was already at approx 0ft) This meant the vertical mode was in ALT. (i.e. the autopilot had been programmed to maintain 0ft. When the autopilot was engaged after takeoff at around 1600ft the flight director pitched down to returned to its assigned level of 0ft. This comes down to basic mode awareness of what the flight director is commanding given the stage of flight.

[infrequentflyer789]

"Staff reported that the plane had become visual with the ground."

In much the same way that a drunk becomes visual with the ceiling?

IMC

So more like: came out of the cloud, realized pointing at the ground.

[Sailvi767]

For takeoff the Flight director in the Q400 is set with the Vertical Mode ; GA (Go around) and the lateral mode HDG SEL (heading select). Further the assigned altitude for departure is entered with the ALT SEL (alt select) mode armed. In the Flybe case the crew forgot to input the assigned altitude and left it at 0ft. When they armed ALT SEL the flight director captured the assigned altitude of 0ft (as the aircraft was already at approx 0ft) This meant the vertical mode was in ALT. (i.e. the autopilot had been programmed to maintain 0ft. When the autopilot was engaged after takeoff at around 1600ft the flight director pitched down to returned to its assigned level of 0ft. This comes down to basic mode awareness of what the flight director is commanding given the stage of flight.

How at the moment of engagement does the flight crew not note the immediate nose down movement and disconnect the autopilot within 1 to 2 seconds. Basic airmanship dictates that any time a flight mode change is engaged or activated you verify the desired result is achieved.

[Vessbot]

Every once in a while I encounter a discussion that's a false positive hit against the typical notion of the sensationalist clueless media, and this is one of them. Nobody minding the store while it's plummeting (yes, plummeting) at 4300fpm below 1500 feet, finally to be woken up by WHOOP WHOOP PULL UP. Yes that is indeed a serious incident, and all of you pooh poohing it should imagine yourself in that circumstance, on a dead head or commute! Still "nothing to see here?"

That aside, can you have active FD modes on the Q with the FD hidden? That's the only possibility I can see, that can explain this. If the FD was set in a mode captured at zero altitude, and showing the severe nose down command to achieve that, who in their right mind would engage the autopilot?

[reverserunlocked]

How at the moment of engagement does the flight crew not note the immediate nose down movement and disconnect the autopilot within 1 to 2 seconds. Basic airmanship dictates that any time a flight mode change is engaged or activated you verify the desired result is achieved.

In fairness the report says the skipper responded almost immediately, but you can imagine that with engines at climb power (this was after the acceleration altitude) a nose down pitch to 0 feet would be pretty abrupt. Although the mode confusion wasn’t caught before it bit them, the report does commend the crew for their prompt actions after it happened.

[Vessbot]

the skipper responded almost immediately,

It was 15 seconds from AP engagement and beginning of the pitch down to the EGPWS and corrective action. Right now, get out your phone or kitchen timer or whatever is within reach, and start it for 15 seconds. As it counts, close your eyes and imagine flying and/or monitoring the airplane, right after the off, and imagine it continuously pitching down as it develops a 4300 fpm descent rate, over the course of 15 seconds.

Does that feel like "almost immediately?"

[Fursty Ferret]

What purpose does routinely setting the MCP altitude to 0ft serve on the Q400?

[reverserunlocked]

It was 15 seconds from AP engagement and beginning of the pitch down to the EGPWS and corrective action. Right now, get out your phone or kitchen timer or whatever is within reach, and start it for 15 seconds. As it counts, close your eyes and imagine flying and/or monitoring the airplane, right after the off, and imagine it continuously pitching down as it develops a 4300 fpm descent rate, over the course of 15 seconds.

Does that feel like "almost immediately?"

Ok I read it as they responded ‘almost immediately’ to the pitch down, not the ‘don’t sink’ cation. Although the report does say ‘Timely warnings of the hazardous flight path were provided by the EGPWS, and prompt corrective action by the PF returned the aircraft to safe flight’ so they are damned with some faint praise.

[Vessbot]

What purpose does routinely setting the MCP altitude to 0ft serve on the Q400?
I would guess that the last guy set it to zero after the flight, in an OCD fashion.

[DaveReidUK]

I would guess that the last guy set it to zero after the flight, in an OCD fashion.

SOP, not OCD.

"The first sector was uneventful, and the aircraft landed on Runway 22 at Belfast City. While it taxied clear of the runway, the co-pilot carried out the ‘after-landing’ checks which included setting the autopilot selected altitude to zero."

[Vessbot]

SOP, not OCD.

"The first sector was uneventful, and the aircraft landed on Runway 22 at Belfast City. While it taxied clear of the runway, the co-pilot carried out the ‘after-landing’ checks which included setting the autopilot selected altitude to zero."

Interesting...learning has occurred. Then I echo Fursty Ferret's question.

[hans brinker]

What purpose does routinely setting the MCP altitude to 0ft serve on the Q400?

The reason we are given is statistically it is more likely that seeing 0000 in the altitude select window will make you put in the correct number than you correcting the last altitude with the altitude from the departure clearance, and taking off with the wrong altitude in the window would often lead to an altitude bust.
Taking off with 0000 in the window should really not happen, and taking off with the FD in ALT mode instead of GA should not happen, and engaging the autopilot while climbing out with the FD commanding full nose down and ALT hold displayed should not happen, and waiting till you get to 4000+ FPM descent and GPWS before you disconnect should not happen.
There were way too many pointers something was wrong and none were noticed.
I definitely agree with our company policy of resetting to zero, yes, this wouldn't have happened if they had left the last altitude in but they would most likely have ended up at the wrong altitude, and the amount of things that went wrong here makes not setting the altitude the smallest error in the whole debacle.

[Herod]

"The first sector was uneventful, and the aircraft landed on Runway 22 at Belfast City. While it taxied clear of the runway, the co-pilot carried out the ‘after-landing’ checks which included setting the autopilot selected altitude to zero."

Question from a long-retired driver. Why?

[hans brinker]

Question from a long-retired driver. Why?
See my answer to Fursty Ferret's question.

[Martin_123]

SLF here so please excuse if these are dumb questions - but why is there not some sort of sensible minimum altitude for the autopilot? And why isn't the EGPWS connected to the autopilot to disconnect it or override it and start pulling up?

it's not a dumb question at all, I also fly Q400 and in our company the minimum AP engagement altitude is 1000 ASL.. that is the pure minimum and you only follow it if the wx or other circumstances are causing pain. For most of the times we hand fly it until clean and above. Before engagement of AP we have to make sure that your pitch and bank matches the one commanded by the FD, otherwise you shall not engage it. I can't imagine how come a PF can make a decision to engage the AP when the FD is actively pointing down in the brown. That is clearly noticeable and should be a big red flag


It’s a Q400. At that point you stop asking sensible questions and just shrug. The Q400 has a lot of oddities and the autopilot is one of them.


oh come on now, dare me asking how many hours you have on a Q400? I smell a lot of BS. The AP as in any other aircraft is doing exactly what you tell it to do. I agree that VNAV and RNAV bits are a bit hit and miss sometimes, but for more basic things, it works beautifully

SOP, not OCD.

"The first sector was uneventful, and the aircraft landed on Runway 22 at Belfast City. While it taxied clear of the runway, the co-pilot carried out the ‘after-landing’ checks which included setting the autopilot selected altitude to zero."

this is a stupid SOP. Generally the altitude selected would be zero if you start it up from cold and dark. As par of a normal cockpit preparation we set the altitude 100 feet below the SID clearance. Then once the clearance is taken, the PF adjusts it to the correct one. It works out wonderfully because, if not adjusted, the altitude will stay at something like 3900 or 3400 (most SID altitudes are nice round numbers) and it stands out as an eyesore on the PFD and upon taxi/instrument check you immediately notice you haven't done something right. If it is set to zero - you have tinyzero on the PFD and it is easy to miss.

Avherald also says the crew were reading a checklist at 1300 and were not actively paying attention to the monitoring of the flight. In our company you can do it either above MSA or 3000ft whichever is higher. MSA in Belfast City is between 2900 to the north and 3800 to south. Both pilots should have full attention on flying the aircraft below that point, what are you doing with a checklist in your hand just 1000 feet after departure is beyond me

[hans brinker]

Just made a long post about the merit of setting 0000, and now think you might have me beat. All I can say is in my aircraft the zeros do stand out.

[reverserunlocked]

ALT SEL is a big Dash gotcha for one, along with the the daft things like pressing the system page selector twice to get stuff onto the right screen, not setting the two remaining V speeds on the other side during the set up etc. It’s all smacks of it being a bit rushed during the design stages.

It has a lot of go though, fair play to Bombardier and P&W, they got that bit right.

[reverserunlocked]

Just made a long post about the merit of setting 0000, and now think you might have me beat. All I can say is in my aircraft the zeros do stand out.

RYR set it to 100 I think during the after landing flow. Makes sense.

[Martin_123]

ALT SEL is a big Dash gotcha for one, along with the the daft things like pressing the system page selector twice to get stuff onto the right screen, not setting the two remaining V speeds on the other side during the set up etc. It’s all smacks of it being a bit rushed during the design stages.


how is ALT SEL a big gotcha?
There is a very good reason why the speed bugs work the way they do, it allows for flexibility in airline SOPs
Can you elaborate what is the right screen that you get to see when pressing the sys button twice? I rather prefer hitting buttons twice on the ground to get to the page I want to see on the ground and only press it once in the air when I'm in the emergency, rather to do it the other way around.. Again, can I ask how many hours you have on the Q400 and what was your previous type? maybe habitation is causing some issues with you?

[Herod]

Hans. I see your point, but I have to disagree. Surely, setting the CLEARED ALTITUDE is part of the SET and CROSSCHECK. Sure, one pilot can miss it, but two?

[Flocks]

The answer is really simple ... Routine + the fact the brain is really good to see what he want to see ...

How many accident / incident are due because people after doing the same check and SOP for few years, finally got a day : they think they did it, they looked at what to check, they did the call accordign to SOP but in fact they didn't see the selection / switch position ... was wrong.

You really need to force you to do the check and even we are all Human.
I remember in the SIM few years ago, after a go around, I did the check list, I had a look to the no light of the landing geea and happily called "gear up, light out", during the radar vextov I thought the plane was noisy but it didn't trigger anything in my mind, when PF called for gear down during the funal, I realized the gear was already down with 3 green, I then apologize, the captain and the trainer realised at that time the gear were down all the approach, 3 people in a sim ...

Now I agree with many of you, was a lot of possibility to catch the problems sooner ... When they did the taxi check list, was a check of the AP mode, when they lined up, was again a check of the AP mode engaged, I m sure they did the right call and I m sure when investigator asked the crew, they both remember all was fine and set as they called with GA mode engaged (plane was in alt mode on the ground)

[Flocks]

What kind of dash 8 do you fly ? The NG ? Did the alt sel drop out if you touch the cobtcon wheel of autopilot while in alt* ?

I can guarantee you, if you press altsel with 0ft selected and the plane altitude near zero, plane will go to alt locked ...

[pinkpanther1]

Martin to clarify, you can arm ALT SEL below 1000ft. We regularly set 500ft ALT SEL on visual approaches as an SOP. But in this case it would not arm ALT SEL but go straight to ALT HOLD as the aircraft was already at it's armed altitude.

[Flocks]

Ok, i see the confusion.

In your airline you select the SID alt minus 100ft then press GA, HDG and alt sel.

​​​​​in flybe, the alt is set to zero, clearance is taken then the PF should set the SID altitude, then press GA, hdg and alt set then calling it, PM should check it.
what they did was, they took clearance, then went straight to press GA, HDG, Alt sel then set the SID altitude, so when alt sel with the altitude selector at 0 was pressed, plane went in alt locked mode.
Then now have a FMA saying HDG and alt locked with the altitude selector set at the SID. So you won't have any alt warning and altitude amber ...

Ofc, the FD was showing a horizontal bat at zero pithc in the ground and not +8 degree with GA mode, thet could have spotted it on the taxi checklist and they could have spotted it on the line up check list (check list ask to call FMA mode), was no alt sel arm white at all so it could have been spotted ... then when passing acceleration altitude, they could have seen the FD showing a deep descent and could have spot it there ... We end up about brain is good to see what is used to see and disregard what doesn't fit in his idea of reality.

How you explain the Emirates 777 with 3 crew trying to do a go around with idle ? They were used to have go around power set by autothrottle ... Expect in that case autothrottle was off due to the plane touching rwy...
Or how explain the airbus A320 in "mont Saint Odile" in France, instead of FPA 3degree they selected VS 3000ft/MN, they then we're focused of why the plane was speeding up ? I do remenrem one of the crew convincing himself saying, oh we have little bit of tail wing and plane really light today .... Brain see what he want to see ...

[Martin_123]

I agree, mind can play tricks and if I remember correctly, Flybe use crosshair FD, while we use chevron type. I never liked the crosshair type, I think it can disappear from your perception too easily. Don't mean to start an AUDI VS BMW or PC VS MAC debate here, just my personal opinion

[Flocks]

They have indeed the cross FD, no sure on the Q400 (not Ng) you can have the chevron ...

I m not an expert of the pro / con of each of them.
​​​

[Herod]

Martin and Flocks. I guess your bias depends on what you have the most experience of. 95% of my FD time is on crosshairs, and I never really liked chevrons. I guess someone with the opposite experience would say the opposite.

[JW411]

I once flew for an operator (in the USA) which had a mix of cross hairs and V-bars on the fleet. After a couple of weeks, you don't even notice.

[Vessbot]

A few separate points:

1. The question is still open: on the avionics in question, can you have active guidance modes with the FD hidden from view? Because my mind boggles at trying to imagine anyone - at any point along the automation dependency spectrum - merrily flying along in an initial climb with a visible FD commanding a pitch into the ground, and then engaging the AP in this state.

2. Prior to engaging the AP, if they were climbing with a visible FD pointing at the ground, I can easily imagine this (unfortunately). One of the aspects of the automation dependent culture is a taboo against not having the FD up, even if it's 100% inappropriate. I remember a sim session where we did a PRM breakout, and in the 90 degree turn away from the airport, (I was PF) I was turning against the FD, which was still commanding a turn back onto the localizer; and the PM was talking on the radio so I couldn't ask him to set it to heading and set the bug. So the FD at that point was counterproductive (i.e., worse than useless) so I reached up and turned it off to avoid the distraction. I got scolded for doing this. Yes, in the instructor's absurd conception, it was better in the interim to have it up and pointing the wrong way than to turn it off. "All available resources," right? :D:ugh::{ (Imagine explaining this scenario to the inventor of the FD.) So if this crew had gotten flogged in the past by instructors in a culture like this, I can certainly see how if they bungled the settings before takeoff, they would have felt more comfortable having a wrong-pointing FD than not having one.

3. Once they engaged the AP and were in (ahem) "plummet" mode, I wonder if they spent those 15 seconds distracted away from noticing the plummet, or were in the often-fatal "what's it doing now" conversation.

[SoFarFromHome]

Interesting incident.

Experienced crew on board, certainly the Captain, so I think its a bit quick to assume automation dependency.

If you consider no FMA annunciations, no GA, no ALT etc and just look at where the FD is?

Are the pink bars/mustache pointing to the sky?

No....

If its not pointing to the sky, its not correctly set. STOP and correct it.

Isn’t it the case however that when ‘we’ are feeling rushed, that is exactly the time to slow down, you know there will be errors, let us create the time we need to catch and fix them. Unfamiliar airport, tired from these preceding duties I would not be adding distractions with single engine taxi or other frivolous distractions.

TEM.

My biggest concern is that it took 15 seconds or so to notice the aircraft was not doing as it should be. Where was the scan, what were they doing? The PF’s eyes should be focused on the flight path and nothing else. PM should be backing him up, checklists can come later, much later.

[Vessbot]

I didn't mean to be automatically assuming automation dependency (even though being "experienced" is not an automatic exclusion from the category -- the famous "Children of the Magenta" video is 19 years old now) -- I'm just coming up with various scenarios that could explain the situation. In short, you can boil my last post down to: If the FD was showing a descent while they were climbing due to misset modes, why didn't they turn it off? And tried to answer that question with a plausible scenario from my experience.


If its not pointing to the sky, its not correctly set. STOP and correct it.

Just about this one point, if it's caught on the ground then I agree. But if caught right after takeoff, then any way I see it the right move is to turn it off, and untangle the mess later at a few thousand feet. Not during the initial sequence of gear, flaps, thrust, speed, freq change, etc.

[SoFarFromHome]

Vessbot,

Agree, if airborne FD’s off, then resolve the issue.

It is the missing 15 seconds that remain unexplained.

Any ideas?

[Flocks]

No, it is not possible to have FD mode but not seeing the FD in the dash 8, so yes they had a command bar showing a descent.
the initial pitch for take off is 15degree, at acceleration flaps are selected to zero, the nose is lowered to the FD GA bar (around 8 degree), then the autopilot is asked for engagement, what will engage the pitch mode, if the nose is not lowered to the GA bar, when engaging the autopilot the FD will synchronise to the current pitch and as said before go to pitch mode.

As it should done, if my FO call AP engagment too soon and not sync with the GA FD bar, I will ask them to go to the GA bar before engagement of AP, then explain why it is important.

You can try as much to understand what happened in their brain, you won t ... read the AAIB report (link at the beginning of the forum). In the world of Dash 8, I m sure it is not the first time pilot take off with wrong setting of FD, it is just probably the first time it went that far before reaction ... But they did appropriate reaction with the ground proximity warning and continued the flight safely.
we can all do mistake, we are all Human ... But we can all learn from mistake of other instead of saying : it won't happen to me, impossible, I can't do this kind of mistake ...

[Uplinker]

For those who don’t know: Pilots setting up a commercial airliner commonly program in the departure that is expected before obtaining the departure clearance. This is because there is plenty to do during the turnaround, and the clearance is often not available until 15 mins before scheduled off blocks time. We don’t normally wait and program in the departure at this late point, because there is too much to do with coordinating with ATC, ground crews, cabin crew, checklists etc to get away on time. What we do is put in what we expect from the filed flight plan coupled with experience and knowledge of the particular airfield. However, if the departure we read states the cleared altitude or level will be X, many operators/pilots will set X-100’ to remind themselves that the departure clearance and/or cleared level has not been confirmed by ATC.

Departure block altitudes or levels tend to be a whole number, so something like 6900 or 2900 instead of 7000 or 3000 respectively will stand out as being wrong and need checking. During departure if the cleared level was completely missed and not reset, at least the aircraft would climb away towards an expected level for that normal published departure.

This operator obviously sets 0000 altitude before clearance had been obtained, but in the light of this incident, and the way the Q400 flight director works; perhaps Q400 SOP should now be changed so that zero altitude is never set, and their before take-off checks re-written to include a positive check of the cleared level that has been set before lining up.

The incident is very worrying in that the FD pitch down command was either not noticed or not queried before AP engagement, and then once engaged, neither pilot appears to have noticed the abrupt pitch down of the aircraft until the GPWS alerted them. I would recommend some FD and AP misbehaviour scenarios soon after take-off during the next year or so of SIM tests to improve pilots’ scan and monitoring at such critical times.

[RVF750]

Some of us need to look at a google image for a second. Unlike many other types, there IS NO Flight ALt "Window" displaying the "000" or Alt at all. It's just a knob on the panel and all the actual figures are on screen in a Q400.

Alt Sel was always a Pain..... when I moved onto it the heading bugs weren't even synced on both sides. Still, a very capable and powerful beast.

3,500hrs on type before anyone wonders. Can't fathom how this happened though if the crew were paying attention. The Q400 requires concentration at all times. You simply cannot leave it alone. and you do need your feet on the pedals as well.

Happily a long way from the things now.

[Vessbot]

No, it is not possible to have FD mode but not seeing the FD in the dash 8, so yes they had a command bar showing a descent.
the initial pitch for take off is 15degree, at acceleration flaps are selected to zero, the nose is lowered to the FD GA bar (around 8 degree), then the autopilot is asked for engagement, what will engage the pitch mode, if the nose is not lowered to the GA bar, when engaging the autopilot the FD will synchronise to the current pitch and as said before go to pitch mode.


Thanks for the exposition, and it goes a long way toward explaining to me what happened. To make sure I understand you right, I'll repeat it in my own words: On a normal TO, the FD is only capable of showing 8 degrees, where you normally pitch to ~15. (So far so good?) So on a routine take off, (before acceleration) you're used to ignoring the pitch bar as you're flying ~7 degrees above it.

If you're used to following the FD (normal for my plane) it's inconceivable to not notice that it's down in the dirt.

But if you're used to ignoring it, it's a much smaller gap to bridge from "it's somewhere down there in the blue" to "it's somewhere down there in the brown." Not noticing is now a lot more plausible, and according to your normal-case scenario, it's easy to see how they could engage the AP at the routine time.

Of course all this only explains the pre-AP engagement phase.

[Vessbot]

You can try as much to understand what happened in their brain, you won t ... read the AAIB report (link at the beginning of the forum). In the world of Dash 8, I m sure it is not the first time pilot take off with wrong setting of FD, it is just probably the first time it went that far before reaction ... But they did appropriate reaction with the ground proximity warning and continued the flight safely.
we can all do mistake, we are all Human ... But we can all learn from mistake of other instead of saying : it won't happen to me, impossible, I can't do this kind of mistake ...


I read the AAIB bulletin which I had skipped over before, so thanks for pointing me to it. It explicitly answered one of my questions: "During the next 15 seconds, whilst the crew were completing the ‘after takeoff’ checklist, the aircraft descended at an increasing rate with the pitch attitude reducing from about 10° nose-up, to 8° nose-down."

This aspect of the scenario is well-addressed at my airline, where the after takeoff checklist is done entirely by the PM, silently; and I'm glad for that.

Another interesting find from the bulletin is that it lists 3 similar incidents from the same operator. In 2 of them, they saw the pitch down after AP engagement right away, and stopped it after about 150 feet. In the third, they saw the misset modes and fixed them before AP engagement.

As far as being unable to understand what happened in their brain, I'm not sure whether you mean in general, or about this incident. But eiter way I disagree. Sometimes you're left with no answer, sometimes you end up with a fairly good guess, and sometimes it's obvious. But at any rate, from our perspective as pilots it's very important to try to learn that lesson so we don't repeat it ourselves, as it is a cause of the accident chain. It's just as important as the guys who piece together the aircraft to find out what broke before impact, or read the FDR to see the sequence of mechanical events, etc.

As far as the brain-sequence of this particular incident, it's becoming clearer and clearer to me, thanks to:

- the description in the AAIB bulletin
- info from you guys who have flown the type
- similar experiences of my own

Something not yet mentioned in the thread is that they were rushed to takeoff with a late second engine start, a quick flurry of accomplishing the remaining procedures, one of which was interrupted by the takeoff clearance. Introspection moment #1: This happens to us all the time. Even though when I shake hands with the Captain after sitting down for leg #1 and he goes through the usual introductory spiel which most of the time includes something like "we won't ever let ourselves get rushed, that's when you miss stuff and mistakes happen. If we ever need more time, we own the parking brake and we'll make more time;" Yet it happens anyway.

So they do the last-second flurry of checklists, during which, by the way, they do catch that the modes are wrong! But unfortunately fix only part of the problem, leaving the modes still misset. Introspection moment #2: When rushed and you fix a problem, it's extremely easy to feel that mental reward of fixing it: "Aha! all good now" and don't necessarily take the step back and make sure the big picture is OK, which may include a still-unfixed second or third part of the problem. By getting that "aha!" you feel satisfied to continue on.

So they do that, continue with the takeoff, fail to notice that the pitch bar is "down there in the brown" as I noted in my previous post, and engage the AP at the routine time. Now that the AP is on, we can all rest easy right? ;) And this should be an introspection moment for all you guys where the SOP has both pilots involved in the after takeoff checklist.

For me introspection moment #3 comes where my type has a much simpler vertical mode scheme and it is literally impossible to make this particular mistake. If I ever move to a type like the one in this incident, can I bring some of my habits with me? Even if I fastidiously follow the new SOP, can my old habits rear up in a moment of hurry and/or high workload?

[Flocks]

Thanks for the exposition, and it goes a long way toward explaining to me what happened. To make sure I understand you right, I'll repeat it in my own words: On a normal TO, the FD is only capable of showing 8 degrees, where you normally pitch to ~15. (So far so good?) So on a routine take off, (before acceleration) you're used to ignoring the pitch bar as you're flying ~7 degrees above it.

If you're used to following the FD (normal for my plane) it's inconceivable to not notice that it's down in the dirt.

But if you're used to ignoring it, it's a much smaller gap to bridge from "it's somewhere down there in the blue" to "it's somewhere down there in the brown." Not noticing is now a lot more plausible, and according to your normal-case scenario, it's easy to see how they could engage the AP at the routine time.

Of course all this only explains the pre-AP engagement phase.


​​​​​​What you say is quit relevant, we are used to not follow the GA FD in case of take off or GO around with 2 engine, the FD when pressing GA is initial pitch for one engine ...
I was surprised about this comgic from a more advanced plane when I took my command ... But it is how the plane is designed ... And as said before there is a lot strangely designed in the dash 8.

Now I said that, still, I force myself to synchronise the pitch to the FD bar after acceleration before asking for the AP engagment and I also force myself to really read the mode FMA I see and not just what I think I pressed, it has saved my day more than once ... The dash is tealre basic, it will do what you ask for even if totally stupid.

A good example is when you go in IAS mode, my previous plane, if you had selected an altitude above and ask for a high speed climb with IAS mode but still at low speed when engaging, the plane would have level off, speeding up then resume climb with the speed requested, the dash 8 will start to descent (you ask for a high speed, so he want to reach the speed you want) then will climb when reaching the asked speed and realized you have lot of energy with engine and so can climb. Only way to no do that is avoided big speed change with IAS mode or monitor closely your engine set up (not auto throttle on the dash 8)

[reverserunlocked]

I replayed the incident flight on the Q400 desktop simulator, with the expected results.

https://vimeo.com/300191332

[IcePack]

We don’t normally wait and program in the departure at this late point, because there is too much to do with coordinating with ATC, ground crews, cabin crew, checklists etc to get away on time. What we do is put in what we expect from the filed flight plan coupled with experience and knowledge of the particular airfield.

says it all really. ATC ground procedural/restrictions having un- foreseen safety repercussions.
Also before I retired the number of readback errors ( not picked up by atc) was increasing. Ie reading back what you expect & briefed rather than what was said.

[pilotmike]

... I don't recall being able to set 'Plummet' mode?
(Edit - looks like the Beeb have decided 'drops' might be better than the original 'plummet' - they could be right but descends could be even better?)
Plummet would seem to be a very apt description here, given the evidence - very high RoD so close to the ground.

As for the rhetorical, indeed somewhat mocking question of whether the subject aircraft has a 'plummet mode': it would appear the answer is a resounding "YES". The questionable SOPs combined with the autopilot being able to be engaged with the FD bar 'on the ground', and the pilot choosing to engage autopilot with such obviously wrong FD command and failing to check the set altitude made this a very definite 'Plummet' mode.

The ability to engage the AP in this situation is of equal concern as the pilot's choosing to do so and then to fail to check it was doing what was expected. However the OP refers to their experience at Everett., implying experience of Boeing AP behaviour. Certainly the 737 AP will only engage if: a), the attitude is very close to the FD command, and b) there is negligible trim force on the column. Unless BOTH criteria are met, the AP will refuse to engage. It would appear that the design of the subject AP is not ideal, conspiring as it does with a pilot who fails to notice the daft FD command ,to spear the aircraft into the ground unless it is caught very quickly. Quite worrying.

[FullWings]

Having read the report, I’m finding it difficult to understand how the situation was allowed to develop for as long as it did. I’m also sure that the crew did not set out to have an incident like this either, so there should be some learning points we can take away?

It is not normal on any departure to start descending soon after takeoff, unless something is going badly wrong. This seems to have been missed/ignored for what is a considerable length of time, given the ~1s reaction time we are meant to have approaching V1 and the initial climb out being what most would call a critical flight phase. All I can think of is that the PF/PM relationship broke down to the extent that no-one was actually minding the shop and/or projecting forward the aircraft trajectory. If the GPWS hadn’t gone off, one wonders how low they would have got or even worse.

Were they uncomfortable with what the aircraft was doing or was it only brought to their attention by a warning system? I fly a fairly computerised plane (777) but those first couple of minutes after departure you are concentrating hard on basic data and what the automatics are doing plus what modes they are in. I’d love to know what distracted the crew on this flight enough to end up in this situation, so I can avoid it myself...

[airtrumpone]

Interesting incident.

Experienced crew on board, certainly the Captain, so I think its a bit quick to assume automation dependency.

If you consider no FMA annunciations, no GA, no ALT etc and just look at where the FD is?

Are the pink bars/mustache pointing to the sky?

No....

If its not pointing to the sky, its not correctly set. STOP and correct it.

Isn’t it the case however that when ‘we’ are feeling rushed, that is exactly the time to slow down, you know there will be errors, let us create the time we need to catch and fix them. Unfamiliar airport, tired from these preceding duties I would not be adding distractions with single engine taxi or other frivolous distractions.

TEM.

My biggest concern is that it took 15 seconds or so to notice the aircraft was not doing as it should be. Where was the scan, what were they doing? The PF’s eyes should be focused on the flight path and nothing else. PM should be backing him up, checklists can come later, much later.


yes very expierenced Captain, incrediabley clever, it was once said that if you asked him the time he would build you a clock ( or tell you how to build one), needless to say no one ever asked a second time.

[VFE]

The AAIB report went into some detail to highlight the Commander's roster that week, which included being contacted prior to the start of his/her standby duty on the day of the incident to position in a taxi from ABZ to GLA (approximately a three hour journey). The report also mentions the Commander's previous days which included plenty of positioning, hotelling and a twelve hour duty.

Obviously they feel this is worth mentioning.

Just saying.

[Nil further]

The clever Captain

Every airline has these guys , build you a clock instead of just looking at their watch .

Quite why he answered the ‘phone and has his rest disturbed. Is beyond me .

Why do Flybe feel it’s OK to contact people during their rest period ?

The real questions left un-answered

[Uplinker]

If you want proper sleep, turn your phone off !

Or even easier and quicker; Set it to flight mode. Then the alarm will still work, but you won’t be distrurbed by any texts, emails or calls. :ok:

[reverserunlocked]

If you want proper sleep, turn your phone off !

Or even easier and quicker; Set it to flight mode. Then the alarm will still work, but you won’t be distrurbed by any texts, emails or calls. :ok:

Or use the ‘do not disturb’ mode which allows numbers you select (elderly parents etc) to get hold of you but mutes alerts, texts and calls from everyone else. I use it every day.