NTSB Press Release Asiana 214

[Cherokee_Cruiser]

But can it be "on" but not "armed" so it won't capture the speed?

AT is usually on or off. The only 'armed' is during the takeoff roll with the power in Flex or TOGA or during a go-around in which case it is in TOGA. After thrust reduction altitude, the levers are brought back to the CL detent and stay there. AT is on. During approach you can leave the AT on and leave it in CL or take AT off and manually control the thrust yourself. If you're in speed mode with the autothrust on, then it will capture your approach speed and maintain it. If you bring the levers back to idle, AT turns off and you are in control of your thrust.

 

[seagull]

AT is usually on or off. The only 'armed' is during the takeoff roll with the power in Flex or TOGA or during a go-around in which case it is in TOGA. After thrust reduction altitude, the levers are brought back to the CL detent and stay there. AT is on. During approach you can leave the AT on and leave it in CL or take AT off and manually control the thrust yourself. If you're in speed mode with the autothrust on, then it will capture your approach speed and maintain it. If you bring the levers back to idle, AT turns off and you are in control of your thrust.

Precisely, and that is my point. Only Boeing has a corner-point where it can be on, and appear to be armed, but not capture ANY speed. AI, MD, L-M, N-G and all the others I know of do not have it.

 

[seagull]

I have read that before and found it absolutely mind blowing that there is a school of thought alive out there that says it's a good idea to decrease pitch attitude in the flare in the manner you described.

That is not the big part of it. The larger issue is the perceptual issue on height due to kinematic issues in large/long aircraft.

 

[mshunter]

They're not dangerous. With very, very few exceptions, the industry-standard stabilized approach criteria work very well.

Maybe I phrased it wrong. For some, they see it as a box. When it doesn't fit in that box, they don't know what to do, and that is a dangerous situation.

 

[Cherokee_Cruiser]

Precisely, and that is my point. Only Boeing has a corner-point where it can be on, and appear to be armed, but not capture ANY speed. AI, MD, L-M, N-G and all the others I know of do not have it.

Not really a corner-point. I have zero experience in Boeing aircraft but I surmise if I take the throttles back myself, and the FMA annunciates "HOLD" mode then I assume the thrust levers will now stay at the place I put them because I manually moved them myself and they are "HOLD."

Regardless, in Airbus, taking the thrust levers to idle (as this Asiana pilot did) turns the AT off. In the Boeing, the power is taken to idle and is left there in HOLD mode. I don't know about MD or LM planes, but both Boeing and Airbus if you manually take the power back to idle, it's going to stay there. I would think that's common sense. The CA should have known coming off the A320 that if the power is manually brought back to idle, he has to now manage his thrust himself.

 

[typhoonpilot]

Precisely, and that is my point. Only Boeing has a corner-point where it can be on, and appear to be armed, but not capture ANY speed. AI, MD, L-M, N-G and all the others I know of do not have it.

One can easily stall or overspeed an MD-80 by climbing/descending in vertical speed mode. It's happened more than once at U.S. airlines.


TP

 

[gotWXdagain]

Really? I haven't seen that at my airline. We are pretty clear about our unstable approach criteria though. 1000 feet AGL isn't a "gate." It is the point after which you must remain within fairly tight parameters or else you must go around. Maybe the problem is the way it is being taught at your carrier? That seems more like a training issue than a stabilized approach concept problem.

In my case it was at a flight school, and our criteria for the chickenhawk/seminole/arrow was stable at 200' or go-around. The problem became that the flight school expected even pre-solo students to meet basically an ATP version of stabilized: Speed +5/-0, Flight path correct and on centerline, landing configuration, Power settings normal, Sink rate normal, and checklists completed. Some standards instructors tried to force a continuous 2 red/2 white PAPI as well, which though it was never expressly written, was prevalent throughout training to the point where students became afraid to use runways that weren't equipped with visual approach indicators. When you're just learning how to fly an airplane, that gate becomes everything, especially when you have an instructor harping it in your ear all the way through your training.

In hindsight this seems like more of an institutional problem than a concept problem...

 

[PhilosopherPilot]

In my case it was at a flight school, and our criteria for the chickenhawk/seminole/arrow was stable at 200' or go-around. The problem became that the flight school expected even pre-solo students to meet basically an ATP version of stabilized: Speed +5/-0, Flight path correct and on centerline, landing configuration, Power settings normal, Sink rate normal, and checklists completed. Some standards instructors tried to force a continuous 2 red/2 white PAPI as well, which though it was never expressly written, was prevalent throughout training to the point where students became afraid to use runways that weren't equipped with visual approach indicators. When you're just learning how to fly an airplane, that gate becomes everything, especially when you have an instructor harping it in your ear all the way through your training.

In hindsight this seems like more of an institutional problem than a concept problem...

Yeah. I hear what you're saying, but that's not really what we are talking about for a 121 carrier.

 

[Bumblebee]

QUOTE]
staph infection.

 

[🦈💜]

In my case it was at a flight school, and our criteria for the chickenhawk/seminole/arrow was stable at 200' or go-around. The problem became that the flight school expected even pre-solo students to meet basically an ATP version of stabilized: Speed +5/-0, Flight path correct and on centerline, landing configuration, Power settings normal, Sink rate normal, and checklists completed. Some standards instructors tried to force a continuous 2 red/2 white PAPI as well, which though it was never expressly written, was prevalent throughout training to the point where students became afraid to use runways that weren't equipped with visual approach indicators. When you're just learning how to fly an airplane, that gate becomes everything, especially when you have an instructor harping it in your ear all the way through your training.

In hindsight this seems like more of an institutional problem than a concept problem...

Stupid, IMO. Fly the airplane you're in. If you're in a 737, fly it like a 737. If you're in a 172, fly it like a 172.

-Fox

 

[seagull]

One can easily stall or overspeed an MD-80 by climbing/descending in vertical speed mode. It's happened more than once at U.S. airlines.


TP

I am restricting this to 4th generation glass, MD80 is not.

 

[seagull]

Not really a corner-point. I have zero experience in Boeing aircraft but I surmise if I take the throttles back myself, and the FMA annunciates "HOLD" mode then I assume the thrust levers will now stay at the place I put them because I manually moved them myself and they are "HOLD."

Regardless, in Airbus, taking the thrust levers to idle (as this Asiana pilot did) turns the AT off. In the Boeing, the power is taken to idle and is left there in HOLD mode. I don't know about MD or LM planes, but both Boeing and Airbus if you manually take the power back to idle, it's going to stay there. I would think that's common sense. The CA should have known coming off the A320 that if the power is manually brought back to idle, he has to now manage his thrust himself.

No, that analogy does not hold. The system's autothrottles are normally engaged full time. Moving them does not disconnect them. It is not a "hold" mode as you describe. Under all circumstances other than the one described you can pull it to idle and it will still power up to capture the speed.

 

[🦈💜]

Half of me wonders if a lot of the apparent inability to revert to actually flying an airplane versus mashing buttons is a product of trying to teach the whole thing like a science/script, rather than developing a fluid understanding of the art of energy management.

Look, I've gotta say it—If modern airline pilots are just sitting in the seat to manage the automation, follow the script and keep the airplane within a rigid set of parameters for an entire flight, and if that's truly the path to safety, then humans have no business flying airliners and computers should do all of the flying all of the time, with a system monkey to manage the electronics like a modern automated train system.

Computers are capable of processing far more individual inputs far more quickly than a human, if you're only dealing with quantified and quantifiable data.

Personally, I'd prefer to fly on an airplane with a pilot engaged and active throughout the entire process of flying than on a fully automated airplane... but you can't really have it both ways—an incredibly tight box with strict checklists and sequencing, strict numbers for all phases of operations, etc, run by humans, without errors. In my opinion the failure modes we're seeing that we're classifying as "automation dependence" are the same failure modes you'd see in a fully automated system, where bad data, no data or lack of data leads the logic into an incorrect state... these are not the failure modes you see from a human-dependent system, which often relies on gut feeling, observation and intuition.

-Fox

 

[NewYorkophile]

In my case it was at a flight school, and our criteria for the chickenhawk/seminole/arrow was stable at 200' or go-around. The problem became that the flight school expected even pre-solo students to meet basically an ATP version of stabilized: Speed +5/-0, Flight path correct and on centerline, landing configuration, Power settings normal, Sink rate normal, and checklists completed. Some standards instructors tried to force a continuous 2 red/2 white PAPI as well, which though it was never expressly written, was prevalent throughout training to the point where students became afraid to use runways that weren't equipped with visual approach indicators. When you're just learning how to fly an airplane, that gate becomes everything, especially when you have an instructor harping it in your ear all the way through your training.

In hindsight this seems like more of an institutional problem than a concept problem...

I could fart and blow through that stabilized criteria.

 

[gotWXdagain]

Stupid, IMO. Fly the airplane you're in. If you're in a 737, fly it like a 737. If you're in a 172, fly it like a 172.

-Fox

Totes agree 100%

 

[Autothrust Blue]

One can easily stall or overspeed an MD-80 an airplane by climbing/descending in vertical speed mode. It's happened more than once at U.S. airlines.


TP

FIFY.

 

[dasleben]

Half of me wonders if a lot of the apparent inability to revert to actually flying an airplane versus mashing buttons is a product of trying to teach the whole thing like a science/script, rather than developing a fluid understanding of the art of energy management.

Look, I've gotta say it—If modern airline pilots are just sitting in the seat to manage the automation, follow the script and keep the airplane within a rigid set of parameters for an entire flight, and if that's truly the path to safety, then humans have no business flying airliners and computers should do all of the flying all of the time, with a system monkey to manage the electronics like a modern automated train system.

Computers are capable of processing far more individual inputs far more quickly than a human, if you're only dealing with quantified and quantifiable data.

Personally, I'd prefer to fly on an airplane with a pilot engaged and active throughout the entire process of flying than on a fully automated airplane... but you can't really have it both ways—an incredibly tight box with strict checklists and sequencing, strict numbers for all phases of operations, etc, run by humans, without errors. In my opinion the failure modes we're seeing that we're classifying as "automation dependence" are the same failure modes you'd see in a fully automated system, where bad data, no data or lack of data leads the logic into an incorrect state... these are not the failure modes you see from a human-dependent system, which often relies on gut feeling, observation and intuition.

-Fox

Thankfully, not true. Usually what happens in these JC threads is someone takes an idea like this and creates a story out of it, almost making it believable. But no, we don't sit there just to manage the automation, as I'm sure you're aware. We also don't fly rigidly, focused only on a stabilized approach "gate" as a student pilot might.

We are still professionals.

 

[PeanuckleCRJ]

They were above 100 feet. Also, above 100 feet, how slow will the Bus let you get below the target speed in normal law?

They were to fly by at around 100 feet at very low speed and do a steep climb out... just below A.FLOOR inhibition altitude to prevent that from kicking in. The FO was to manually maintain the thrust levers. They drifted down to about 40 feet, and saw the trees coming up quicker than anticipated. Well, he pulled up and was already at the low speed protection speed, so the plane said "no, you're already at alpha limit, you cant pull back anymore" while he waited for TOGA power to finally kick in.

Basically, on normal aircraft tech, he triggered the pusher while waiting for TOGA power to spool.

As far as the gotchas that typhoonpilot is pointing out with the protections, they very nearly mirror what is on the buses (320 generation and beyond). It's preferred to turn off the FD's to get you into speed mode, protection inhibition altitudes... they are all very similar.

I also agree with the 1000/500 gates. They are logical and work very well.

 

[PeanuckleCRJ]

Precisely, and that is my point. Only Boeing has a corner-point where it can be on, and appear to be armed, but not capture ANY speed. AI, MD, L-M, N-G and all the others I know of do not have it.

The bus has THR IDLE and THR CLB modes which activate during OP CLB and OP DES... which is identical to the Boeing FLCH and the MD equivalent (at least the 88/90...). During a FLCH or similar in the modern glass MDs, does it go into CLMP in any descent mode? If so.. that is the parallel to what we are discussing here.

The gotcha is the OP DES, where it goes into THR IDLE. It's the same gotcha with the THR HOLD or HOLD that the boeing goes into once back at idle in FLCH. If there isn't a vertical capture set up, it will appear to be active or armed, but will not capture a speed unless there is a vertical capture mode present as well.

NW slapped the tail pretty hard on an airbus about a decade ago going into DEN with a very similar situation.

 

[derg]

It will go "LO LIM" in a descent, which more or less means that the auto throttle has reached the lowest limit of it's authority and you can manually pull them back.