A/T: is it really helping you ?

[CVividasku]

Hello,

What follows is more oriented towards engineering, airplane flying qualities, rather than learning how to fly.
Any qualified pilot is able to deal an A/T especially if the A/T is available on their type. But does the A/T make the job easier, or harder ?
This topic is more related to knowing if airbus or boeing airplanes are easier to fly.

We're going to talk about airplanes longitudinal aerodynamic modes.
As you may know, you have two main modes in longitudinal aerodynamics. The phugoid and the AOA oscillation (or whatever its name in english)
The latter is less know since it's more difficult to observe. It's highly damped.
I'm no expert at certification requirements but quite obviously, an airplane with a diverging AOA oscillation would not be flyable by a human hence would not be certified.


The following aims at showing how the A/T can modify the aerodynamic modes of an airplane. The phugoid.
Let's suppose you have a completely conventional airplane. A piper PA28.
You have a trim, you do not have an A/T. Suppose you're in stabilized flight. You trim down the airplane. You do not touch the controls. What happens ?
The airplane keeps its power, accelerates, descends, goes over the new equilibrium trim speed (let's call that trim reference speed or TRS), then climbs back.. It's a phugoid. Eventually it will stabilize in a slow descent, with a descent angle that gives just enough "gravity thrust" to help the current thrust reach the new TRS speed.

Now let's suppose you have an A/T engaged.
You have a speed bug. At first you stabilize the airplane in level flight. Perfectly trimmed. So TRS = speed bug. Let's create a gap between the TRS and the speed bug. You can trim down, but since not many planes show the TRS, let's reduce the speed bug, this will allow us to make a small reduction that we can measure. Let's say a few knots. 5 knots. How much time have you got until you're in deep trouble ?
At first, almost nothing happens, and the instructor in the sim wonders why you asked to do this weird exercise.
There is only a 5kt difference so the A/T slowly reduces thrust, the speed reduces slowly too, it reduces even more slowly because the airplane is trimmed to keep its current speed at TRS. But the A/T continues to see a speed difference and continues to reduce thrust. Soon, the airplane behavior becomes exponential.
In a time that I could not evaluate in all configurations, and that is quicker if the speed difference is higher, but in any case much quicker than you would imagine, your airplane is stabilized in a descent with idle thrust. If you were on approach, in less than 30 seconds you end up with idle thrust and terrain warning to pull up.

The same goes for the other way around. If you increased the speed bug, the airplane would soon enter a climb with max power.
You have a hidden speed bug, the trim reference speed, and it must be in sync with the A/T, displayed speed bug.
Imagine they're in sync but you encounter a wiind gust, your speed reduces. The A/T will increase thrust, the trim will pitch down, both work to stabilize the airplane.
But if you're just 1kt off, the airplane will continuously want to enter an exponential dive to idle descent or climb to max power.

On an airbus, the hidden speed bug moves automatically with your speed. So you can have a difference between speed bug and current speed, but it won't cause any problem of a similar nature. The airbus is purely stable.
The boeing FBW have a similar fly by wire law, except that the introduction of the trim speed changes everything.

However, this issue is entirely manageable, of course. It's called flying the plane. It's just a bit more difficult, you cannot spend extended periods of time looking somewhere else if your plane is not in perfect trim.

Hence, you can find it easier to fly a trimmable airplane without the A/T.
Does anybody on a trimmable airplane like to fly approaches without A/T ? I hear most pilots at my airline almost always keep it on.

[321XLR]

Fly whatever airplane model you fly, in accordance to how the manufacturer and factory test pilots designed it to be flown.

[WhatShortage]

On a pa28 you won't have a phugoid, you'll have several sinuidoal ups and downs that will eventually stabilise towards whatever trim you put.

Not to mention airbus that will stabilise by itself so yeah, you're basically inducing everything to prove your "theory" right which is not.

[Capn Bloggs]

If you were on approach, in less than 30 seconds you end up with idle thrust and terrain warning to pull up.

Imagine they're in sync but you encounter a wiind gust, your speed reduces. The A/T will increase thrust, the trim will pitch down, both work to stabilize the airplane.
But if you're just 1kt off, the airplane will continuously want to enter an exponential dive to idle descent or climb to max power.

Eh? You're a pilot. Either YOU stay on the approach path/GS by pulling back on the stick, or let the AP coupled to the GS do it. This concept of letting the nose/flight path go up and down naturally because the actual speed is different to the TRS is nonsense. Something is (should be!) actively controlling the pitch attitude and therefore flight path so you won't end up diving into the ground at Idle power.

My last jet, the 717, was a classic example of this. Slack glidepath control, allowing the nose to wander, resulted in the ATS staying up when it should have ben coming back, and staying back when it should have been coming up. However, with tight slope control eg if you get hit on the face with a gust and you balloon, you stay on the slope with a push (ie don't let the nose come up) and the instant you start to apply forward pressure the ATS says OK! and pulls the power off. The speed then comes back to the TRS/bug speed. But all the while, I have actively stayed on the desired approach slope.

I always used the ATS. It was a fantastic system in the 717.

Jets should/can't be flown like Pipers on final. You don't pull the power off to make the jet down down by relying on the secondary effects of controls. You push the nose down, then adjust the power if you need to (or let the ATS do it for you).

You're obviously an FO: "Always remember and forever take heed, right hand for glidepath and left hand for speed!". Or, if you have an ATS, it can do the left-hand bit for you.

[rudestuff]

The OP is thinking to much. There was a question in there somewhere, and the answer is yes. A/T makes it easier. That's why we have them.

[FullWings]

However, this issue is entirely manageable, of course. It's called flying the plane. It's just a bit more difficult, you cannot spend extended periods of time looking somewhere else if your plane is not in perfect trim.
That’s what the AP is for? ;)

TBH this is a feature on the 777/787 not a bug (C*U). The aircraft (through the yoke, thrust levers, PFD and/or view out of the window) is telling you that you are not at the speed you trimmed for and to pay a bit more attention. If you are on approach near the ground and it becomes nose heavy you know before you scan the speed that it has dropped, plus the AT will help you by increasing thrust to compensate. It’s behaving dynamically much as a non-FBW aircraft would under these circumstances, so allowing it to feel more natural, save the lack of pitch/power couple. Control the path with the elevator, control the speed with power (AT does this) and keep it trimmed.

When you fly the actual aeroplane, it makes more sense.

[CVividasku]

If you are on approach near the ground and it becomes nose heavy you know before you scan the speed that it has dropped, plus the AT will help you by increasing thrust to compensate. It’s behaving dynamically much as a non-FBW aircraft would under these circumstances, so allowing it to feel more natural, save the lack of pitch/power couple. Control the path with the elevator, control the speed with power (AT does this) and keep it trimmed.

No unfortunately it doesn't.
It behaves as a non-FBW aircraft with A/T would, but not the same as a PA28.
If you have your TRS above your speed bug (that is, being nose heavy) the A/T won't help, the A/T will progressively retard the levers and crash the airplane if you do nothing.
Without an A/T, if the trim speed is above the desired speed, and the power is adequate for the desired speed, the aircraft will descend slightly faster than desired. It won't retard the levers since it can't.
And your flight path will be much closer than desired.

The A/T will only help if the airplane is perfectly in trim.

We usually don't do this type of maneuver during sim sessions, they're not line pilot maneuvers, they're engineers or test pilots maneuvers. Doing it just once was extremely enlightening.

[rudestuff]

You're making the assumption that AT only controls speed. In a Boeing at least, the AFDS is a very complex system with a multitude of modes, many of which rarely get used.
For most of the descent the thrust should be at idle, with the speed controlled by elevator. In the final stages of the approach other modes may come into play such as ALT HOLD, V/S, FLCH, SPD PATH (both geometric and approach) etc... but that all happens as part of an integrated system with AP or FD. In a Boeing, when you turn off the AP you turn off the AT as well and, you know... fly the airplane.

[BraceBrace]

If you have your TRS above your speed bug (that is, being nose heavy) the A/T won't help, the A/T will progressively retard the levers and crash the airplane if you do nothing.
...
We usually don't do this type of maneuver during sim sessions, they're not line pilot maneuvers, they're engineers or test pilots maneuvers. Doing it just once was extremely enlightening.

If I understand correctly, the A/T will reduce thrust, however the pitch of the aircraft will move down as well as the pitch channel is chasing a higher reference speed.

Pilots are trained on the basis of target attitudes and power. The change of aircraft attitude will tell the pilot something is off, and you will naturally want to bring the nose up. It's a very natural reaction. Any pilot will ask questions if on a glide the nose passes through the horizon. He will raise the nose which is a natural "request" for pitch up trim, which reduces the TRS. At least it should be for every pilot. That's why I hope you're not a pilot...

The Airbus FBW on the other hand is another peace of unnatural response, as it is tries to maintain a path. So with speed reduction on the A/T, the nose will rise, and many starting pilots will feel this is unnatural pitch behaviour of the aircraft.

[CVividasku]

You're making the assumption that AT only controls speed. In a Boeing at least, the AFDS is a very complex system with a multitude of modes, many of which rarely get used.
For most of the descent the thrust should be at idle, with the speed controlled by elevator. In the final stages of the approach other modes may come into play such as ALT HOLD, V/S, FLCH, SPD PATH (both geometric and approach) etc... but that all happens as part of an integrated system with AP or FD. In a Boeing, when you turn off the AP you turn off the AT as well and, you know... fly the airplane.
That's exactly the kind of information that I'm looking for.
At my airline, most of the time the plane in manually flown with the A/T still engaged.
If I understand correctly, the A/T will reduce thrust, however the pitch of the aircraft will move down as well as the pitch channel is chasing a higher reference speed.

Pilots are trained on the basis of target attitudes and power. The change of aircraft attitude will tell the pilot something is off, and you will naturally want to bring the nose up. It's a very natural reaction. Any pilot will ask questions if on a glide the nose passes through the horizon. He will raise the nose which is a natural "request" for pitch up trim, which reduces the TRS. At least it should be for every pilot. That's why I hope you're not a pilot...

Of course pilots are trained and are able to fly a plane with an A/T and a conventional trim.
But being able to do so doesn't say if it easy or easier than another airplane.

Your last sentence shows that you didn't understand the goal of my post.

Depending on the airplane characteristics, it can be easier or trickier to fly. The manufacturer usually tries to make the airplane as easy as possible. At least in my opinion they should.
If you can look at something else for an extended period of time, and the airplane hasn't changed it's trajectory, it's easy to fly. It means that if you pay reasonable attention you will achieve a very precise flight path.
If the airplane has an instability that can develop exponentially with time, it means both that if you look at something else for an "extended" period of time (it can be as short as 10 seconds) the airplane can already be in a very upset situation, and that you need to pay close attention to the airplane to correct this unstability everytime it appears.

Airplane manufacturers during flight test have to demonstrate the aircraft behavior with the controls released, to observe the different modes of the airplane. Dutch roll, spiral, phugoid, aoa oscillation... There are requirements about the stability of the airplane's behavior.

What guarantee do you have that the trim reference speed is exactly at the speed bug ? As little as 5 kt can create an instability. What can guarantee that if you want to trim for a speed bug at say 150kt, you're not going to oscillate the trim speed between 151 and 149kt each time you attempt to match closely the speed bug ?
Apparently some of the instructors at my airline have a supplementary document, not included in the manuals, that describes more precisely the trim system on the B777. Apparently it will say that the trim should match the speed bug when given a order consistent with doing so.

[BraceBrace]

That's exactly the kind of information that I'm looking for.
At my airline, most of the time the plane in manually flown with the A/T still engaged.
...
Your last sentence shows that you didn't understand the goal of my post.

Depending on the airplane characteristics, it can be easier or trickier to fly. The manufacturer usually tries to make the airplane as easy as possible. At least in my opinion they should.

1) we are talking about fly-by-wire systems. Boeing advises the use of A/T with A/P off on FBW aircraft. The statement "A/T OFF" is incorrect and applies to none FBW Boeing. On a FBW Boeing you are allowed to fly A/T off, but this is mainly done for proficiency and preferably in good weather conditions and otherwise standard operations (ie standard ILS approach into known airport).


2) yes I do understand your question, but it is not a question of "easy" or not. It is a question of "natural" or not, because with natural behaviour there will be natural pilot response. I also don't understand your questioning when talking about "easyness" and then only consider 1 single item in stability. You open the door of stability, there is a lot more to discuss. The reality for the pilot is that it is a natural behaviour. The nose will drop in a "seems natural direction" way, in a "seems natural slow/fast response", and hence we will react and support it (if you want to include stability, then all it a "perceived stability"?). There is also a reason why there is a PF and a PM. The PF is not supposed to look anywhere else for 10 seconds, especially not when he's reducing the speed. Why would he look anywhere else? This is basic piloting skills. Hence the question if you're a pilot or an engineer.

I'm a pilot (trainer) and an engineer in aviation. It's interesting for the instructors to have all the background information, it's even more interesting to be able to show it. But it doesn't change anything. When you go into the simulator and demo the behaviour, a behaviour that is considered "natural" whatever is causing it, it requires a natural response of the pilot. So yes, A/T ON is easier on a Boeing FBW as it reduces workload and reduces the chance of overstimulation in case of problems.

As far as the answer to your "hard/easy" question: it is more than easy enough as it doesn't require thinking. It is a trained skill that doesn't need adaptation. It's simply more relax.

[BBK]

Fly whatever airplane model you fly, in accordance to how the manufacturer and factory test pilots designed it to be flown.

This! I’d add in accordance with company SOPs.

[CVividasku]

1) we are talking about fly-by-wire systems. Boeing advises the use of A/T with A/P off on FBW aircraft. The statement "A/T OFF" is incorrect and applies to none FBW Boeing. On a FBW Boeing you are allowed to fly A/T off, but this is mainly done for proficiency and preferably in good weather conditions and otherwise standard operations (ie standard ILS approach into known airport).

Yes, there is a lot more to talk about regarding airplane design, we could spend several days debating.

I don't understand what you typed here. Which statement is incorrect ?
You mean having the A/T OFF should not be done, or a statement that I made regarding what happens with A/T OFF is incorrect ? If so, which one ?

The PF is not supposed to look anywhere else for 10 seconds, especially not when he's reducing the speed. Why would he look anywhere else? This is basic piloting skills. Hence the question if you're a pilot or an engineer.
Of course the pilot is supposed to be flying the plane. "Supposed to"... An airplane that is more stable, is easier to fly.
In some fatal accidents, pilots have been seen not to look at their displays for "extended" period of times.
Eastern 401 is the most prominent example. Much more recently, prime air/atlas air 3591, one was disoriented, the other one wasn't properly monitoring.
Another very good example is Turkish 1951. The A/T retarded and it wasn't noticed until too late. All of these are proof that pilots are fallible at maintaining a perfect constant monitoring of flight parameters.
Hence, a stable aircraft is a better idea than one that you have to keep a very close eye on.

In this debate you could simplify or caricature my position as one of the following :
- advocating for airbus law without the speed in the FBW law
- advocating for an easier trim. Display the trim reference speed (I'm told that the A220 displays it, which I couldn't verify), or have a way to synchronise the trim with the speed, or reduce the stick force to make the unstable behavior slower...

Then, easy or easier is subjective.
Imagine you're flying the airplane on final, you know the power setting for this phase of flight. Usually around 55% on most jets. Imagine you perfectly matched your speed bug and your TRS.
Your plane should be stable, right ? Still not !
If you pitch down ever so slightly, the A/T will reduce the thrust to keep the speed. You will lose energy. When you notice the glide error (be it small or large), the thrust will increase to help you climb back to the glide, then decrease again. All of it is destabilizing.
If you flew A/T OFF, with the correct power setting, if you pitch down slightly, now your airplane will accelerate. Hence you now have a second cue on your PFD that you're low : the glide and the IAS. You can just pitch up, you will trade speed for altitude, and you will be back on the glide at the correct speed, without thrust variation.

Some people will find the first method easier, because they focus on one cue and don't have to worry about knowing power settings. Other ones will prefer the second method.

[Capn Bloggs]

This is morphing into jets should be flown like PA28s.

[FullWings]

No unfortunately it doesn't.
It behaves as a non-FBW aircraft with A/T would, but not the same as a PA28.
If you have your TRS above your speed bug (that is, being nose heavy) the A/T won't help, the A/T will progressively retard the levers and crash the airplane if you do nothing.
Without an A/T, if the trim speed is above the desired speed, and the power is adequate for the desired speed, the aircraft will descend slightly faster than desired. It won't retard the levers since it can't.
And your flight path will be much closer than desired.

The A/T will only help if the airplane is perfectly in trim.
I think you are describing a situation: aircraft trimmed nose down beyond bug speed, pilot not on the controls, AP off and close to the ground, which is not something that is normal operation or even desired, in any type. It is still respecting the commanded speed. The only times where you are deliberately out-of-trim and make no attempt to change it are rotation and flare, and that’s only for short periods. The primary function of the AT is speed control, not path control - that’s up to the pilot when flying manually, which should not be news to anyone?

It’s akin to saying “I landed in a crosswind with my feet not on the rudder and we went off the side of the runway!” Not optimal but not surprising.

Boeing would say: ”Fly the Airplane”.

[CVividasku]

I really wonder why you would think I don't intend flying the plane.

With the same reasoning, we would not train for stall, approach to stall, or GPWS maneuvers in the sim.
This maneuver (being out of trim with or without A/T) is not a way of flying but a way of observing the airplane behavior.
Much like the manufacturer will make maneuvers without control inputs to observe dutch roll, phugoid, spiral...
It's a way of understanding the airplane behavior, and I think it could be used to demonstrate and prove how important being in perfect trim is.

On the long term, thrust will control flight path and trim will control speed.
With the A/T ON the flight path loses all of its long term stability, and that's worth noting, in my opinion.

When we tried that in the sim, the instructor had a similar reaction to the ones here. "Just fly the plane". My sim partner however, who was previously an engineer, was amazed with the speed at which events unfolded.
By the way, it's not easy to predict that the A/T will retard all the way to idle. A different speed tracking law, for example proportional, could not retard to idle, or increase to TOGA, for a small speed difference. This experiment hence proves that there is an integral component in the A/T speed tracking law. That is, the longer the plane senses a difference in speed, the bigger the thrust variation.

[BraceBrace]

This maneuver (being out of trim with or without A/T) is not a way of flying but a way of observing the airplane behavior.
Much like the manufacturer will make maneuvers without control inputs to observe dutch roll, phugoid, spiral...
It's a way of understanding the airplane behavior, and I think it could be used to demonstrate and prove how important being in perfect trim is.

You observe airplane behaviour in a sterile computerized environment. That it helps in "understanding" certain aspects I agree. As a demo it certainly has its place in a training environment to understand how fly-by-wire works, however it does not connect with the question of "how easy flying is" in reality.

Reality is that you will very rarely have this situation as "identifyable" as you will be flying through changing disturbances all the time and it becomes hard to tell what causes what when the changes in pitch _initially_ are small. From a pilots point of view there is no point in trying to make a distinction. A pilot will make a correction.

You ask the question if A/T ON is really helping us by pointing out a "problem". I'm saying the issue isn't making our life worse, and A/T ON has a lot more advantages than the subtle disadvantage you are trying to focus on here, a disadvantage that usually ends up "unidentifyable" in the "noise" of other natural inputs from a changing atmosphere outside of the world. If we reduce speed we know how to trim at the same time. My 777 time has been a while, but I think the value is 2 sec trim per 10kts speed change? How easy do you need?

Your conclusion here of "how important being in perfect trim is" has always been true. It's a bit strange to then compare to a system that trims itself completely. The reality is that many pilots transferring to Airbus fly-by-wire need more "thinking" as some aircraft reactions are contradictory to what is experienced as "natural". On a Boeing fly-by-wire, the reaction is natural.

To explain the "incorrect statement": on any non-fly-by-wire Boeing, Boeing will always recommend to fly with A/T OFF, when the A/P is off because of the pitch-power coupling (leaving the A/T ON will create unexpected pitch changes with power changes). However, on fly-by-wire Boeing, Boeing will recommend to fly A/T ON with A/P OFF as well, as the fly-by-wire will correct for pitch-power coupling (and pitch changes due to config changes).

[Vessbot]

. The phugoid and the AOA oscillation (or whatever its name in english)
The latter is less know since it's more difficult to observe. It's highly damped.


It's known as the "short period oscillation"

On a pa28 you won't have a phugoid, you'll have several sinuidoal ups and downs that will eventually stabilise towards whatever trim you put.
.

This is what the phugoid is.

[CVividasku]

If we reduce speed we know how to trim at the same time. My 777 time has been a while, but I think the value is 2 sec trim per 10kts speed change? How easy do you need?

To explain the "incorrect statement": on any non-fly-by-wire Boeing, Boeing will always recommend to fly with A/T OFF, when the A/P is off because of the pitch-power coupling (leaving the A/T ON will create unexpected pitch changes with power changes). However, on fly-by-wire Boeing, Boeing will recommend to fly A/T ON with A/P OFF as well, as the fly-by-wire will correct for pitch-power coupling (and pitch changes due to config changes).
So, you mean the trim will trim a continuous 5kt per second ?
So, if you had a speed trend of 50kt up, you would need to trim down continuously, and if you had a speed trend of 5kt up, you have to trim one second every ten seconds ?
Interesting but it's not something that you can count on.

As for the other part I quoted, I never said boeing recommended something or something else. I doubt they would recommend looking at the airplane without touching the controls, anyway.

[Fursty Ferret]

777/787 has a blip trim system that will set the speed term in the C*U law to your current speed if within 5 kts, so you just need to flick the trim switches every few seconds and you'll stay in trim.

[BraceBrace]

Interesting but it's not something that you can count on.

You don’t count on it, you use the feedback from the yoke giving you trim requirements like any normal aircraft because you would see the nose pitching down unwanted. Very natural.

But you could calculate to have an idea...

[Vessbot]

I really wonder why you would think I don't intend flying the plane.


It's as if we're discussing how the electrical system works, and how exactly the automatic cross tie behaves if one side power is lost... and the reply is "how could you possibly consider shutting off a generator in cruise, this is dangerous and irresponsible!"

[CVividasku]

A friend of mine tried the maneuver in the B777 PMDG sim (general public, how reliable is it ?)
For a sufficient difference in speed, the behavior is exactly as described on the first post.
However, if the speed reduction less than 5 knots, the airplane will trim itself !! My friend displayed the trim reference speed on the PFD, and the TRS was moving by itself !
However, for a further reduction of 5 knots, the TRS stopped moving by itself and the plane descended as described.

So, we found out that Boeing 777 can trim itself, a little bit.

[Obama57]

Eastern 401 was not an A/T issue. A/T on the Tristar was for approach only, and in this case, a low altitude radar box pattern to troubleshoot the gear not down and locked indication. The “flaw” was no aural warning when the A/P went from Alt Hold to V/S after the captain bumped the yoke while fiddling with the nose gear light. A/T maintained speed as aircraft entered slight descent into the Everglades. Many CRM issue, but it was, I believe, 1970 or ‘71.

[Speed_Trim_Fail]

Eastern 401 was not an A/T issue. A/T on the Tristar was for approach only, and in this case, a low altitude radar box pattern to troubleshoot the gear not down and locked indication. The “flaw” was no aural warning when the A/P went from Alt Hold to V/S after the captain bumped the yoke while fiddling with the nose gear light. A/T maintained speed as aircraft entered slight descent into the Everglades. Many CRM issue, but it was, I believe, 1970 or ‘71.

Correct other than the point that I believe it was CWS mode that the Tristar reverted to when force was applied to the control column.

[CVividasku]

Eastern 401 was not an A/T issue. A/T on the Tristar was for approach only, and in this case, a low altitude radar box pattern to troubleshoot the gear not down and locked indication. The “flaw” was no aural warning when the A/P went from Alt Hold to V/S after the captain bumped the yoke while fiddling with the nose gear light. A/T maintained speed as aircraft entered slight descent into the Everglades. Many CRM issue, but it was, I believe, 1970 or ‘71.
I used it as an example to emphasize that sometimes the crew doesn't scan enough the instruments. This case being the most extreme.
However, it would also be true to point out that without automatics engaged, and with just a slight inadvertent push of the stick, Eastern401 would have entered a "level phugoid", that is, would have started to oscillate around its original altitude. And we wouldn't be talking about them.
Automatic flight has its catches.

[Cough]

Manual flight also has its catches if you don't scan the instruments....

[CVividasku]

It's not really a matter of scanning or not scanning.
It's a matter of having a stable airplane or less stable.
Past some point, an airplane instability will become illegal.

[Uplinker]

Not really sure what the OP wants to know. Flying with manual thrust IS a matter of scanning. It makes the PF job more intense, since it requires very close scanning, over and above the manual flying.

In busy airspace, such as the London TMA, or JFK approach, or chaotic ATC such as Cairo, there can be a lot going on and it is not really a place to be manually flying, unless you have no choice owing to equipment failure. Both PF and PM need to be on top of their game; listening to RT calls, other aircraft, and making sure they are complying correctly, and building an overall picture of the airspace. If PF is manually flying, half or more of their brain is taken up with that task, leaving say, 1.3 brains dealing with ATC instructions and surrounding traffic instead of 2 brains.

With a conventional aircraft; changing thrust and or speed can/will change pitch, which has to be accommodated and re-trimmed for. An aircraft with a FBW pitch channel can automatically adjust for thrust changes, compensate for the pitch-power couple, and trim. This takes away a chunk of the PF task, allowing them to use more of their brain listening to ATC, which in turn makes the whole process less intense and easier.

I am type rated for and have flown both conventional and FBW aircraft on the line for many years. The Boeing Classic 300/400 has auto-throttle but no FBW, so when flying manually with auto-throttle, PF has to constantly adjust pitch and trim inputs when the auto-throttle changes, and they constantly have to react to these changes "without causing them". This is why auto-throttle is not recommended by Boeing when flying manually.

If the aircraft does have a FBW pitch channel, auto-thrust can be compensated for and trimmed automatically, so the pitch-power couple effectively disappears, and PF can fly manually with auto-thrust active, making this a much easier task.

With the B777, even though it has FBW you, (I think), have to repeatedly push the trim pickle switches to stay in trim ?, which seems like a waste of the FBW, (it didn't work at all for me during a recent SIM test, so I must have been doing something wrong), but the Airbus FBW and auto-thrust system is very well designed, integrated and implemented.

Basically, in my hands-on experience, Airbus have done a fantastic job of designing and creating a properly integrated FBW and auto-thrust system, which works extremely well.

[vilas]

The difference in FBW of Airbus and Boeing has commercial aspect as well. Boeing was introducing FBW for existing large customer base of 737, 767, 757. They wouldn't appreciate a new Boeing radically different. So whether stab was really trimmed by pilot or by FBW they kept the concept traditional. Airbus had no market presence. They were trying to create a new customer base so they went with different philosophy based on high level of automation with pilot mostly monitoring and over the years they have proved right.

[CVividasku]

Not really sure what the OP wants to know. Flying with manual thrust IS a matter of scanning. It makes the PF job more intense, since it requires very close scanning, over and above the manual flying.
.
With the B777, even though it has FBW you, (I think), have to repeatedly push the trim pickle switches to stay in trim ?, which seems like a waste of the FBW, (it didn't work at all for me during a recent SIM test, so I must have been doing something wrong), but the Airbus FBW and auto-thrust system is very well designed, integrated and implemented.

Basically, in my hands-on experience, Airbus have done a fantastic job of designing and creating a properly integrated FBW and auto-thrust system, which works extremely well.
Well that's the point. Actually no. At least for me.
If you want to follow your ILS, just know your power preset, and display it. Trim the airplane properly (that's only for boeing) Then your airplane will have a trajectory that cannot diverge quickly from the intended flight path. It just cannot.
If you have the A/T on, the airplane can diverge from the intended trajectory, for example if your nose is slightly low or high. It's easier to display a power preset to a tenth of a percent (it's displayed like so) than to display a pitch preset (the pitch indicator is more than 1° large and the scale is precise to 2.5°..) Also I noticed that, at least in the sim, the vertical speed has a half second/one second delay that can generate PIO if you try to fly the glide with vertical speed rather than pitch. Contrarily to airbus.

Flying an airbus manually requires very little scanning if you know your figures. Just display the power setting, set the proper pitch, monitor speed. You're all set. It will work the same with unstable winds. The airplane won't develop an instable behavior such as the idle descent that Boeing can do.
Also, the A/T (both of them) can fall in the "windshear trick". Reducing the thrust to idle when a headwind increases, only to find itself woefully unprepared when the windshear turns to tailwind, while thrust is at near idle.
At least, Boeing allows to see this phenomenon more easily.


Indeed, the airbus system is very well designed. Personnally, I don't see the point of having to trim an airliner. I don't see it as so major since I started real flights, because the airplane is actually much softer on the controls. The out of trim efforts are much smaller, actually very similar to that of a light aircraft, I felt.

In my opinion, the ideal system would be a small mix between airbus and boeing. Airbus laws, but with more emphasis on what trim is doing.
So, a small feedback of where the trim is by moving the neutral stick position (which boeing 777 doesn't have either), both sidesticks moving together when manually flying, and some improvements of the airbus laws regarding not allowing the airplane to use the entire trim range.

[pineteam]

Even tho I agree the Airbus A/THR works very well and reduces the workload; I find the landing with A/THR off to be easier and more consistent especially in area where the wind can varies a lot in short final. But saying this, landing with A/THR still work like a charm most of the times.

[Uplinker]

Ever since starting to fly, I have always thought that pitch trimming for speed, and a pitch-power couple, is something that should be designed-out if it can be. It couldn't be done originally, until technology and electronics provided a solution. And Airbus did exactly that - and more - with their revolutionary FBW system, which is an order of magnitude better than what came before. (I have flown many conventional types). Not having to pitch trim manually during hand flying is an advantage to me, not a problem. And the Airbus auto-thrust most definitely does help me.

I think that pilots who have got into trouble with it have been mishandling the aircraft in some way, e.g. AF 447, or through misunderstandings of how the FBW, or the Airbus Ground Speed Mini system works; and the subsequent auto-thrust behaviour.

But I am being a bit dim and still don't really understand your criticism, CVividasku. Without being too technical, what is it that an Airbus FBW aircraft does or does not do during a gusty ILS that you don't like or gives you a problem ?

[CVividasku]

Oh no, on the contrary, I like the Airbus FBW a lot.
The only thing I don't like so much is the ATHR, but not for the same reason. Like pineteam said, landing without it is easier.
Also, I like to manage the airplane energy by myself. I might want to reduce speed with a bit of power left, to slow down slower, or to be able to spool up the engines quicker.

The only thing I would change about it :
What I said above + trim indicator, and why not an emergency trim neutral button for upset recovery.

[tdracer]

All I know is that - when we were getting ready for the first flight of the 747-8, the chief test pilot was very adamant that the autothrottle needed to be functional, or he wouldn't accept the aircraft was being ready.
Hence we spent a lot of time testing the FADEC/FMC/Autothrottle interface prior to first flight to make sure there would be no surprises...

[Uplinker]

(CV), Is it the auto-thrust or actually the Ground Speed Mini rapidly changing the speed bug on turbulent finals that you don't like ?

I ask this because I have flown with pilots who think that Ground Speed Mini doesn't work or think it "gets it wrong", or they think it "is slow to react", so they prefer to take out the auto-thrust, and "do the job properly".

But G/S Mini is none of these things; they are thinking about conventional constant IAS systems, whereas G/S Mini works the opposite way round - it will increase thrust when the headwind increases.

But when going manual thrust in an Airbus FBW these pilots often do not obey the G/S Mini speed demands, and this appears to them to make the flight smoother and less hectic on turbulent days because the engines are not hosing up and down so much. But it actually reduces their safety margins, and reduces the aircraft energy, which is potentially dangerous.


I should also have mentioned earlier that the excellent Airbus auto-thrust works very well because it is so well integrated with the FBW pitch channel. And its engagement and operational logic is so much more intuitive and reliable than other older auto-throttle systems I have flown.
.

[CVividasku]

I tried the maneuver described here in the FFS. I tried an out of trim condition of only 3 knots, we were at 7000 feet. In less than time than necessary to explain what I expected to see to my two colleagues (pilot and instructor), or in other words, in less than the time you just took to read this sentence, the airplane was already in a full blown idle descent.
So between a small jolt on one of the two speed references, during stable flight and idle descent, there was less than 15 seconds.

More importantly, between the first noticeable effects of the descent and the full blown descent, there was maybe only 5 seconds.

With a speed difference of only 3 knots.
With a larger out of trim condition it would be even quicker.

This may be remotely related to this incident :
https://www.flightradar24.com/blog/southwest-excessive-descent-hawaii/
(CV), Is it the auto-thrust or actually the Ground Speed Mini rapidly changing the speed bug on turbulent finals that you don't like ?

I ask this because I have flown with pilots who think that Ground Speed Mini doesn't work or think it "gets it wrong", or they think it "is slow to react", so they prefer to take out the auto-thrust, and "do the job properly".

But G/S Mini is none of these things; they are thinking about conventional constant IAS systems, whereas G/S Mini works the opposite way round - it will increase thrust when the headwind increases.

But when going manual thrust in an Airbus FBW these pilots often do not obey the G/S Mini speed demands, and this appears to them to make the flight smoother and less hectic on turbulent days because the engines are not hosing up and down so much. But it actually reduces their safety margins, and reduces the aircraft energy, which is potentially dangerous.


I should also have mentioned earlier that the excellent Airbus auto-thrust works very well because it is so well integrated with the FBW pitch channel. And its engagement and operational logic is so much more intuitive and reliable than other older auto-throttle systems I have flown.
.
No, I think the A/T does a pretty good job with ground speed mini, it accepts speed variations, if I remember correctly, which is what I expect of it.

What I didn't like was, in order of importance :
- Especially during flare, there was a risk that the plane would add thrust and cause a long landing if you were a tiny bit late to retard. Also, it means that you cannot really keep your power at 50% for longer if you need to (for example, on the 318, you need to)
- Not being able to control my overall energy. With the correct power setting, the overall energy of the airplane will evolve in a predictable way. With A/T on, a half degree of pitch is nearly invisible on the PFD (I'll let you guess how many milimeters that is), but cumulated over some distance will give very noticeable altitude errors.
- Not being able to keep some power if I wanted to. 40% minimum to keep a very quick reaction time if needed. Or 40% to slow down in a softer manner.

I don't recognize what you write about GS mini. Airbus is the only brand to implement it. So, other airplane manufacturers make unsafe airplanes ?
In my memory, GS mini function, on the contrary, leads to very little variation of thrust during the approach (except if the wind has extreme gusts, which I don't recall having). Because when the airspeed increases due to a wind gust, the airplane's higher drag doesn't have time to slow down the plane back to its original speed before the wind gust dies down. So in my memory, following GS mini consisted mainly in keeping the thrust constant, or almost constant.

[BraceBrace]

If you have the A/T on, the airplane can diverge from the intended trajectory, for example if your nose is slightly low or high. It's easier to display a power preset to a tenth of a percent (it's displayed like so) than to display a pitch preset (the pitch indicator is more than 1° large and the scale is precise to 2.5°..) Also I noticed that, at least in the sim, the vertical speed has a half second/one second delay that can generate PIO if you try to fly the glide with vertical speed rather than pitch. Contrarily to airbus.

To repeat the point so you really understand, not to show off: it is possible to trim a 777 sim from manual level flight onto the glide, and fly the ILS hands off to the runway. Autothrottle on, raw data manual ILS. I have done it. The feedback is precise enough to know when pitch is in trim.

It is almost impossible in reality because the atmosphere is not like a simulator.

To get back to the original point: You fly data and think the world is like this sterile unchangeable data. It is not. Pilots are trained to set data as a starting point but continuously adapt to the reality which is not simulatorlike but full of little outside disturbances.

Your original question turns out to be a quest for confirmation that A/T is making our life more difficult. It is the opposite. Because reality is not a simulator using datapoints. Reality is everything from subtle to big sudden changes, where you need to have a hand on the throttle and a hand on the stick and use the feedback provided. Your ’stability problem’ is small compared to the real world disturbances but you don’t seem to understand that as you’ve never been at the controls of a calm wind ILS, let alone a gusty one. Your problem literally disappears into the wind if I may say so. With all due respect, so would your speed if you decided to disconnect the A/T because you think it is easier. The A/T does a fantastic job at it. It doesn’t even need additional wind corrections which makes life over the threshold a lot easier...

If you can’t accept that reality the solution for you is easy: wait for the first A350 that can takeoff with the autopilot and land without a single disconnect. Problem solved.

[Vessbot]

Your ’stability problem’ is small compared to the real world disturbances

Big or small, I'm glad you're at least acknowledging that the stability problem exists... which, for most of the thread until now, it seemed that it doesn't. I'm surprised that most of the posters either refused, or didn't recognize it, to address it. I would have thought that it was much more widely understood as a general basic precept in aviation, that airplanes are supposed to be stable.

I mean yeah pilots are supposed to make corrective inputs, duh. But even with that, aircraft are still supposed to be stable, other than some super specialty application like NASA's Flying Bedstead or whatever.

[Capn Bloggs]

This may be remotely related to this incident :
https://www.flightradar24.com/blog/southwest-excessive-descent-hawaii/

​​​​​​​Nothing remotely similar.

[stilton]

To repeat the point so you really understand, not to show off: it is possible to trim a 777 sim from manual level flight onto the glide, and fly the ILS hands off to the runway. Autothrottle on, raw data manual ILS. I have done it. The feedback is precise enough to know when pitch is in trim.

It is almost impossible in reality because the atmosphere is not like a simulator.

To get back to the original point: You fly data and think the world is like this sterile unchangeable data. It is not. Pilots are trained to set data as a starting point but continuously adapt to the reality which is not simulatorlike but full of little outside disturbances.

Your original question turns out to be a quest for confirmation that A/T is making our life more difficult. It is the opposite. Because reality is not a simulator using datapoints. Reality is everything from subtle to big sudden changes, where you need to have a hand on the throttle and a hand on the stick and use the feedback provided. Your ’stability problem’ is small compared to the real world disturbances but you don’t seem to understand that as you’ve never been at the controls of a calm wind ILS, let alone a gusty one. Your problem literally disappears into the wind if I may say so. With all due respect, so would your speed if you decided to disconnect the A/T because you think it is easier. The A/T does a fantastic job at it. It doesn’t even need additional wind corrections which makes life over the threshold a lot easier...

If you can’t accept that reality the solution for you is easy: wait for the first A350 that can takeoff with the autopilot and land without a single disconnect. Problem solved.


That was demonstrated some time ago

[Fursty Ferret]

Especially during flare, there was a risk that the plane would add thrust and cause a long landing if you were a tiny bit late to retard. Also, it means that you cannot really keep your power at 50% for longer if you need to (for example, on the 318, you need to)

Just pull the thrust levers back about half-way and Bob's your uncle. Thrust capped at 50%...

[BraceBrace]

Big or small, I'm glad you're at least acknowledging that the stability problem exists...

Because calling it ”stability” is a ”book people” explanation that doesn’t help pilots at all. To a pilot it is a side effect of bad trimming and it is easily ”catched” in the every day life without realising it.

AF447 is an excellent example: you can debate a lot about the technical background of it and FBW laws and logics and causes and effects... if the pilots had realised that at high altitude you can’t ”ask” more than 5 nose up, everybody would have been alive today. And hence that is what we focus on, and every tool that allows us to focus on those basics makes our life easier. So yes, A/T on helps a lot because it allows us to ”monitor” speed rather than ”control” it. If we sense the A/T is not doing what it is suppposed to, we will disconnect it.

Which is what I’ve been trying to say. Remarks like ”what if he wasn’t watching..." is a desk people remark who don’t know how to prioritize in a dynamic environment. A pilot flying manual is not allowed to do that. If he needs something, he has a PM. Because the outside world changes every second. Solving your ’minor’ issue you call a stability problem doesn’t make real flying ”sterile”. He still needs to monitor his instruments. Because we also HAVE to monitor CORRECT behaviour. You can’t see something is wrong if you don’t know what is right.

As a company you might want to create awareness, and include a demo in your initial training program, however the problem is resolved by good piloting practice. It falls under the same category of ”awareness” as the limited pitch authority of the elevator compared to THS on a 737. And on a 737 that danger is much bigger than the ”stability” issue with the 777 FBW (something many people claim to be also a stability problem, but it is not). The limited elevator authority in low speed TOGA situations has already caused crashes (RAM Amsterdam), and still no training program explains it. It only explains how to resolve the issue worst case (stall recovery procedure changed).

[Uplinker]

CV; if you are flying Airbus FBW and the auto-thrust adds thrust during the flare, it is because you were late retarding the thrust, because you flared incorrectly. We have all made this mistake from time to time and we learn to do it right !

My point about G/S Mini is that many pilots do not understand how it works and how it will correctly react with gusts. Even worse, some pilots want to take it out of circuit and control thrust and speed themselves, while still not understanding what G/S Mini does.

Conventional aircraft have to use a constant IAS strategy because when IAS changes, so does the wing lift and pitch trim. So without FBW, anything other than constant IAS will destabilise the aircraft.

Airbus designed a fantastic aircraft which incorporated fly-by-wire. Now they could look again at speed and stability during finals, and because the FBW will automatically compensate for the pitch power couple and for differing IAS, they could use a system based on energy instead of IAS.

Conventional constant IAS systems add a speed increment to keep the aircraft safe when a headwind gust disappears. The constant IAS system takes energy away during a headwind gust. Airbus adds thrust during a headwind gust which increases its energy.

When the headwind gust disappears, the constant IAS system energy drops to below where it was and the engines need to quickly spool up to regain the lost energy. The Airbus energy, on the other hand, only drops to what it was before the headwind gust occurred so the engines only need to spool down to where they were before. So the Airbus G/S Mini system never loses energy whereas the constant IAS system does.

The constant IAS system is not dangerous because a speed increment is added, but the Airbus system is better and is made possible owing to its FBW.

The real danger occurs when pilots override the Airbus GS Mini strategy without adding a safety margin of increased minimum IAS. So by controlling thrust themselves and ignoring the GS Mini, they actually remove the Airbus safety margins, and this could put them in a precarious situation.

[Capn Bloggs]

Conventional constant IAS systems add a speed increment to keep the aircraft safe when a headwind gust disappears. The constant IAS system takes energy away during a headwind gust. Airbus adds thrust during a headwind gust which increases its energy.
Off on a tangent, but that doesn't make sense. On final ie on a fixed approach path, there is only one parameter that is being controlled, that is the IAS, and that means only one energy state. An Airbus cannot "increase it's energy" without increasing speed. I'm not familiar with GS Mini but it sounds similar to the 717's VGust feature, where the FCCs would increase the target speed without changing the IAS bug. In other words, in gusty conditions, the jet would fly faster than bug. This speed would be similar to "bugging up" in the constant-IAS system. If the gust disappears, the Constant airspeed system would only have the speed reduce to the normal bug speed anyway. The advantage of VGust and I assume GS Mini is, of course, that when/if the gusty conditions settle down, the jet would slow down to the no-gust bug speed, thereby improving landing performance.

The constant IAS system is not dangerous because a speed increment is added, but the Airbus system is better and is made possible owing to its FBW.
It's got nothing to do with FBW. The system works perfectly well in the 717 which is as far from FBW as you can get, flight-controls-wise. The pitch-power coupling is a separate issue.

Interesting comments above re throttles in the flare; doesn't sound very user friendly to me.

And for the avoidance of doubt, the ATS in the 717 is excellent and a great reliever of workload and Boeing procedure is to use it at all times.

[BraceBrace]

Just 2 remarks:

1) The idle thrust during the flare works differently. On an Airbus, there is a retard callout and the PF puts the thrust lever to idle. I'm not familiar with this functionality in flight. on a Boeing A/T system, the system is designed to be very aggressive below Vref, it reacts a lot slower when speed increases above Vref. The retard function is automatic, the PF does not have to do anything. Hence the regular speed correction for headwinds and gusts is left out. Works pretty good, only once in 6 years I had to disconnect.

2) Concerning G/S mini
Again, not an Airbus driver, only studied Airbus systems 3 decades ago and kinda "read about it" once in a while. But when I read about a system like "g/s mini", a detail that I think is often overlooked is that it is looked at "independently" from the Airbus and sold as a "safety enhancement". From what I know, I call it a safety necessity.

Airbus pitch control is path related. So if there is a headwind gust, the FBW will pitch the nose down without pilot input. This will not happen on a Boeing, on a Boeing FBW, the aircraft will pitch up and will want to climb. If the headwind gust is gone, on an Airbus, the nose will pitch up as the aircraft "sinks away" from the path. If you want to talk about stability... well I guess you then have to call this unstable behaviour. This is dangerous as the aircraft loses speed and pitches up at the same time. So you need extra protection on the speed side of things.

It's a bit funny to see this discussion surface in a discussion that tackles the "stability" of Boeing FBW. A Boeing FBW that is designed to push the nose down when the speed decreases... (or "unstable" in your words, but I've already understood we don't need to make a distinction between static stability, dynamic stability, control stability,...)

Then again, 3 decades of +FL300 probably degraded my grey cells as well... ;-)

[Vessbot]

Airbus pitch control is path related. So if there is a headwind gust, the FBW will pitch the nose down without pilot input.

It doesn't hold path, that's a simplifying abstraction. It holds a blend of pitch rate (mostly at low speeds) and G (mostly at high speeds).

[CVividasku]

To repeat the point so you really understand, not to show off: it is possible to trim a 777 sim from manual level flight onto the glide, and fly the ILS hands off to the runway. Autothrottle on, raw data manual ILS. I have done it. The feedback is precise enough to know when pitch is in trim.

It is almost impossible in reality because the atmosphere is not like a simulator.

To get back to the original point: You fly data and think the world is like this sterile unchangeable data. It is not. Pilots are trained to set data as a starting point but continuously adapt to the reality which is not simulatorlike but full of little outside disturbances.

Your original question turns out to be a quest for confirmation that A/T is making our life more difficult. It is the opposite. Because reality is not a simulator using datapoints. Reality is everything from subtle to big sudden changes, where you need to have a hand on the throttle and a hand on the stick and use the feedback provided. Your ’stability problem’ is small compared to the real world disturbances but you don’t seem to understand that as you’ve never been at the controls of a calm wind ILS, let alone a gusty one. Your problem literally disappears into the wind if I may say so. With all due respect, so would your speed if you decided to disconnect the A/T because you think it is easier. The A/T does a fantastic job at it. It doesn’t even need additional wind corrections which makes life over the threshold a lot easier...

If you can’t accept that reality the solution for you is easy: wait for the first A350 that can takeoff with the autopilot and land without a single disconnect. Problem solved.

Sorry but you really don't understand my point. I'm not saying at all that the airplane is impossible to fly. I'm just saying that it requires a more constant attention than a plane without the same characteristics.
Trim without an A/T is easier to fly than trim and A/T (Boeing with A/T on)
A/T without trim (airbus) is easier to fly than trim and A/T
No trim and no A/T is also easy to fly for any properly trained pilot. I wouldn't call myself a pilot if I wasn't able to fly a manual flight.
Also not to show off, but I always get above standard marks for the "manual flying" competency. If it can reassure you that it's not a matter of my personnal skills.

My airline requires us to be able to fly approaches in all these conditions. FD or A/T OFF, the other ON, or both OFF.
And I don't have any particular problem doing it. I just think that this is still a design issue.
And the price to pay for this design issue is the type of incident that we just saw.

Also, the proof that this is a design issue is the existence of the "clic clic". Instructors taught us that on this airplane, there is a function that will make the trim speed stick to the A/T speed, which is the only condition that allows an A/T stability. If you trim the airplane within 5 kt of the target speed, and you're not too much out of trim, the trim speed will go to the target speed. I believe it behaves similarly a mouse pointer (like i have on my ipad now writing this text) that sticks to buttons it goes over : a way to improve the finger's precision by sticking to targets.

https://www.flightradar24.com/blog/southwest-excessive-descent-hawaii/
Why is it linked to this problem ? We don't exactly know what happened, but this sentence it's very relevant to the issue here.
" During the go around the first officer “inadvertently pushed forward on the control column while following thrust lever movement commanded by the autothrottle.”"
Followed the thrust lever movement commanded by the A/T. There was an out of trim condition. It was either a trim speed to high, or a force on the stick (which is also an out of trim condition), and the A/T reacted by reducing thrust during a go around, which is not a desirable behaviour.
With a fixed thrust at TOGA, had the F/O pushed, the speed would have increased very quickly, and the normal reaction to that is to pull. Also, without a reaction, it would cause a very conspicuous alarm, the overspeed alarm, which, even if the pilot did not react, would cause the other pilot to react.

To finish with, the stability issue is much larger than real world disturbances. Try to follow an approach in the sim with a voluntary out of trim condition, of just 3 knots, you will see...
Which other disturbance can make you go from normal approach speed to idle or full power in less than 5 seconds ? That order of magnitude is similar to a large windshear.
And I'm not refusing it... I'm however refusing that people would not see what exists.



With all due respect, so would your speed if you decided to disconnect the A/T because you think it is easier.
No, that's outright insulting. I can only wonder why you would write that.
Just pull the thrust levers back about half-way and Bob's your uncle. Thrust capped at 50%...
Very good idea but I was strongly advised against it by my first instructors.

[CVividasku]

It's a bit funny to see this discussion surface in a discussion that tackles the "stability" of Boeing FBW. A Boeing FBW that is designed to push the nose down when the speed decreases... (or "unstable" in your words, but I've already understood we don't need to make a distinction between static stability, dynamic stability, control stability,...)


Not talking too much about airbus because it's not really the point here. However let's just say that the airbus feels very natural to fly, and the FBW does not do what you talk about during headwind gusts without using GS mini (which is what we had to do sometimes because other planes didn't have GS mini, and there is a risk of catching up the previous plane)

I'm not talking about the airplane stability, I'm talking about the throttle levers with A/T ON. With A/T ON, the throttle levers are unstable if there is any out of trim condition, even the slightest.

Airplane stability is indeed divided in static and dynamic, but this "stability" is mainly a speed stability. With a positive static margin, a conventional airplane gains static stability. But it's not a path stability.
If you get a headwind gust in a conventional airplane, the plane climbs. The speed is stable, the path is not.*
If you get a headwind gust in an airbus you will keep a constant path, with an increase in speed. Speed is not stable, path is.
I believe that path is more important to a pilot than speed.
Then, once your path is stable, A/T is a positive addition to the airplane without A/T.
On boeing, if you have an out of trim condition, A/T is not a positive addition to the airplane without A/T.

Airbus revolutionized the concept of airplane stability, moving the underlying concept from speed stability to path stability.
That's the most important lesson that I concluded from this 10 seconds maneuver in the flight sim that's much more enlightening than it seems.

*approximating for the fact that the flight control law is indeed a blend of path and pitch. However while at the controls, if you're not told that it's a blend with pitch, you could easily believe it's just path.
This is a whole other topic but it would be interesting to know why they had to blend the pitch in. Maybe it didn't feel natural enough without it ?

[Uplinker]

Off on a tangent, but that doesn't make sense. On final ie on a fixed approach path, there is only one parameter that is being controlled, that is the IAS, and that means only one energy state. An Airbus cannot "increase it's energy" without increasing speed. I'm not familiar with GS Mini but it sounds similar to the 717's VGust feature, where the FCCs would increase the target speed without changing the IAS bug. In other words, in gusty conditions, the jet would fly faster than bug. This speed would be similar to "bugging up" in the constant-IAS system. If the gust disappears, the Constant airspeed system would only have the speed reduce to the normal bug speed anyway. The advantage of VGust and I assume GS Mini is, of course, that when/if the gusty conditions settle down, the jet would slow down to the no-gust bug speed, thereby improving landing performance.


It's got nothing to do with FBW. The system works perfectly well in the 717 which is as far from FBW as you can get, flight-controls-wise. The pitch-power coupling is a separate issue.

Interesting comments above re throttles in the flare; doesn't sound very user friendly to me.

And for the avoidance of doubt, the ATS in the 717 is excellent and a great reliever of workload and Boeing procedure is to use it at all times.

This is probably not the place for a detailed explanation of Airbus G/S Mini if you haven't studied it or flown it, but yes, it does add IAS speed during a headwind gust, (keeping the ground speed pretty constant). So the engines do add energy. When the gust goes away, the Airbus still has this extra energy which it slowly bleeds off. Whereas the constant IAS system - which reduced energy during the headwind gust by reducing engine thrust - has to spool its engines back up after the gust has gone.

I don't know anything about the B717, so I cannot comment on that, but the Airbus FBW adjusts for the extra IAS so the pilot doesn't have to and the flight path is not destabilised, hence the Airbus auto-thrust can be used when flying manually, whereas when flying the Boeing B737 manually; it is recommended not to fly with auto-throttle engaged.

[Capn Bloggs]

but the Airbus FBW adjusts for the extra IAS so the pilot doesn't have to and the flight path is not destabilised, hence the Airbus auto-thrust can be used when flying manually, whereas when flying the Boeing B737 manually; it is recommended not to fly with auto-throttle engaged.
I think you will find most modern types are designed to be flown with the ATS in at all times. Airbus is nothing special in this regard. It's only the 737 that has an issue with pitch-power coupling and hence it's recommended to not use the ATS when handflying.

[Speed_Trim_Fail]

I think you will find most modern types are designed to be flown with the ATS in at all times. Airbus is nothing special in this regard. It's only the 737 that has an issue with pitch-power coupling and hence it's recommended to not use the ATS when handflying.

747 and 757/767 are the same, for the same reason. AT is used during manual climb and takeoff but is off for manual flight in other phases.

[CVividasku]

I think you will find most modern types are designed to be flown with the ATS in at all times. Airbus is nothing special in this regard. It's only the 737 that has an issue with pitch-power coupling and hence it's recommended to not use the ATS when handflying.
On that topic, i'm very surprised because the same reasoning doesn't seem to apply to engine out situations.
When you're engine out without TAC (for example airbus which doesn't have a real TAC) or b777 with tac failure, the instructors at my airline don't recommend at all to disconnect the A/T on the remaining engine.

Even though, there is in this situation, a huge "yaw power" coupling.

[BraceBrace]

No, that's outright insulting. I can only wonder why you would write that.
Because it is true. That's not insulting, it's a reality.

I'll reply one last time, because I must have read the original question 20 times... I still don't get it. This is the world upside down.

This is the situation that happens so many times in training or during line checks, where something doesn't really quite works because of a skillset that lacks. However, because the automation was unable to solve it, the crewmember ends up blaming automation for the lack of skillset.

First of all:

The origin of the problem is inadequate trimming. There is nothing extraordinary in trimming the B777, on the contrary, it is dead easy. The FBW is an easy system, that helps out a lot. The problem you describe is not an A/T problem, it is a “trim skill” problem. If you reduce the speed on the A/T, every pilot flying the B777 knows or feels instantly he needs to trim, he can even calculate how long, but he can also feel it. He feels the throttles reducing, he feels the stickforce increasing (or sees moving pitch). You don't need a “number” for that. That's only excessive information that takes away attention from other things. The info is there, in both your hands (tactile or kinestetic feedback, whatever you want to call it).

I'm a training captain, in my company we make trainees do raw data approaches with A/T off. When people have a choice between A/T on or off, the difference is visible in the FDM statistics. Lacking the skill to properly trim the aircraft is one thing, and for trainees I can accept flying full raw data A/T off can be a bit overwhelming. It's just something you need to do once in a while. People who know me know I'll gladly take it to the bar and discuss over a beer, have a laugh and move on with life. No harm done.

However secondly:

If you come back to me during the debrief and state this is an A/T problem that made your life difficult, and on top YOU STATE IT IS GOING TO BE EASIER WITH A/T OFF, and that was the reason why you decided to disconnect the A/T, we are not making it to the bar. You can't trim with the A/T? You certainly won”t be able to do it without.

This is the typical pilot excuse blaming automation as the cause for his inability (VNAV is another one of those examples, but let's not go down that rabit hole). When people disconnect the A/T on gusty days, I strap my seatbelt. Because you take one vital parameter, and you take over control as opposed to “monitor”. You feel the engines move, your hands are on the throttle. It's instant, it's immediate feedback without looking at numbers. Now you want to do it yourself? You have the wise people who decide to use A/T on gusty windy days as it is easier. Or you have the “I don't trust” or “I doesn't work properly” people who disconnect and decide they can do a better job without. You should see it from the jumpseat. Very interesting. I don't know why so many people keep holding on to the fact they have better speed control than the A/T.


So 1) stop blaming the A/T for what you can't achieve, or use the excuse “I glanced away”. Because if you're not capable of trimming the aircraft on the approach, you probably couldn't in the departure, you couldn't do it anywhere.

And 2) people who state it would be easier without the automation are usually the ones who feed the FDM statistics

I will leave it at that and stay out of the thread, nothing against you, I prefer beers at the bar. I'll take A/T off approaches any time weather permitting up to a certain level, don't worry, but the A/T is a consistent reliable system. No man's/woman's brain is. There is always a lapse, a miss, ... and a chase to get back on whatever parameter we ended up missing/chasing.

[Vessbot]

Remarks like ”what if he wasn’t watching..." is a desk people remark who don’t know how to prioritize in a dynamic environment.
This is the fundamental basis behind aircraft stability. I am surprised that anybody would pooh-pooh the concept like this.

A pilot flying manual is not allowed to do that.
This is like saying that the hydraulic system is "not allowed to" fail, so there is no need for a second one.

[BraceBrace]

This is the fundamental basis behind aircraft stability.

A badly trimmed aircraft is not an unstable aircraft.

You can aileron trim the heck out of a 737 autopilot and it won't move at all the aircraft as the autopilot will counteract. But it ain't going to be staying there when you disconnect the autopilot because the aircraft is actually out-of-trim. It is perfectly stable though, it will violently look for it's new balance.

It's on this basis many people claim the 737 is an unstable aircraft. It is not. It is a stable aircraft with a control issue. Before anyone jumps on that 737 bandwagon again (we already did I noticed). Pitch-power couplings don't make an aircraft unstable, it does shift the balance. We already learn this ie on a Cessna with something called P-factor. Increasing thrust does not make your aircraft unstable, but the balance changes (albeit in a different axis on the Cessna).

And that's why the eyes need to stay on the attitude. Because "control" of balance is our responsability and we call it trimming. And you have a PM to act as a back-up primary back-up. You do realise the PM has to do deviation callouts? Because yes, eyes and brains of the PF wander around sometimes... (and autopilots disconnect... or get bad inputs...)

Keep it simple people, stop mixing around to complicate... I’m off to the bar now...

[Jonty]

747 and 757/767 are the same, for the same reason. AT is used during manual climb and takeoff but is off for manual flight in other phases.

not really a list of modern aircraft

[Vessbot]

A badly trimmed aircraft is not an unstable aircraft.
As has been explained thoroughly already, and demonstrated in the sim, yes it is, in this circumstance. I don't have anything new to add but sometimes different phrasing works so I'll try:

A statically stable aircraft returns to its setpoint after a disturbance. (A merely out-of-trim aircraft, as you're insinuating this is, has a setpoint different from where you are, and will seek that. But once reached, it will return there after disturbances.)

A plane with a trim speed different from autothrottle speed, does not return to its original flight path or attitude, but diverges all the way until throttle is full or idle. This is because there are two setpoints, one each in a different parameter: The autothrottle holds the autothrottle speed, and the flight path holds the trim speed. If they're the same, then no problem. But if they're different, they interact in a way that makes the flight path unstable:

Let's say autothrottle speed is 150 and trim speed is 160. The flight path (seeking trim speed) will go down to seek 160. But this causes the autothrottle speed to exceed 150, and in seeking it, it retards the power. This brings the speed lower than 160, which is good for the autothrottle speed but bad for the trim speed, which causes the pitch to go further down to seek 160, which causes the autothrottle to further retard to seek 150, etc. It is in a runaway feedback loop. Or, put another way, the pitch is stable on speed but unstable on flight path, because the autothrottles are continuously pulling the speed rug out from underneath us.

It's on this basis many people claim the 737 is an unstable aircraft
I have not heard this claim before.
And that's why the eyes need to stay on the attitude. And you have a PM to act as a back-up primary back-up. You do realise the PM has to do deviation callouts? Because yes, eyes and brains of the PF wander around sometimes... (and autopilots disconnect... or get bad inputs...)
If you yell at the hydraulic pump harder that it's not supposed to fail, maybe it won't?

[BraceBrace]

Let's say autothrottle speed is 150 and trim speed is 160. The flight path (seeking trim speed) will go down to seek 160...

I'm well aware of this explanation, as said, not sure how many times I've read it.

How about this:
* Push AP A CMD, confirm CMD is active on FMA.
* Wait hands under your pants, then put hands on controls again
* Push AP A CMD, confirm FD is active on FMA.

Is that better? Either you restore the in-trim situation, or if you can't, you can hope the AP can do it (within limits). It's called aircraft control.

Remarks like your last one are pretty ridiculous. Using the possibility of fail or lapses as an excuse not to be responsible. The responsibilities of PF and PM are clearly defined in the FCOM. There is a difference between responsibility, failure and blame. You start with a responsibility, you might fail somewhere along the line, and worst case be called in the office. Aviation has thankfully moved away from the blame in many companies and revised to learning from mistakes. But you will not leave that office with less responsibility in the role of PF or PM.

[Vessbot]

I'm well aware of this explanation, as said, not sure how many times I've read it.


Then why did you write that it's "not an unstable aircraft?" A stable aircraft returns to the setpoint, and this one does not.

[BraceBrace]

Because stability is used as excuse to claim it is "harder" to fly with A/T on than with A/T off.

If the AP would create such a situation in which it's trim speed would be "stuck", fine, I'll join you in the room to debate with the engineers behind the desks and hit the engineer button. In the 6 years experienced on the 777 I have never ever even touched the alternate pitch trim levers. I can't even recall if any checklist refers to the use of these levers.

We are pilots. From a pilot training perspective stability has never been the problem in the original question as you can perfectly trim out the aircraft. A pilot with average flying skills will not even notice. Stability is a false excuse. Out-of-trim is the problem, and any pilot can easily solve that. You just have to set the "setpoints" equal to use your terms.

[Vessbot]

Because stability is used as excuse to claim it is "harder" to fly with A/T on than with A/T off.


So your answer as to why you say it's stable, is not in reference to whether it's stable or not

[BraceBrace]

It’s only unstable in a failure case with a stuck trim reference speed for me as a pilot. Which never has happened as far as I know. The alternate trim levers are useless. My engineering thinking stayed on the ground pre 2000. The effect is unnoticeable.

And again... don’t blame the A/T is making it more difficult, you have to trim properly. That’s it and that’s easy.

[CVividasku]

Because it is true. That's not insulting, it's a reality.

I'll reply one last time, because I must have read the original question 20 times... I still don't get it. This is the world upside down.

This is the situation that happens so many times in training or during line checks, where something doesn't really quite works because of a skillset that lacks. However, because the automation was unable to solve it, the crewmember ends up blaming automation for the lack of skillset.
.
I will leave it at that and stay out of the thread, nothing against you, I prefer beers at the bar. I'll take A/T off approaches any time weather permitting up to a certain level, don't worry, but the A/T is a consistent reliable system. No man's/woman's brain is. There is always a lapse, a miss, ... and a chase to get back on whatever parameter we ended up missing/chasing.
Yes, you really don't understand the goal of this topic.
This topic is not a pilot's topic. It's an airplane design engineer's topic. Maybe that can clarify.
It's not a practical discussion. It's a philosophical one. My hand flying skills have nothing to do with it.

Is that better? Either you restore the in-trim situation, or if you can't, you can hope the AP can do it (within limits). It's called aircraft control
I'm going to put you in an airplane that introduces random disturbances requiring up to full back or full forward stick to keep a constant flight path, then tell you that you're not keeping close enough attention to the pitch...
(Yes, this is a "reductio ad absurdum" reasoning")

If you want, we can make a diagram of the B777 throttle stability.
You can create a diagram of potential for every different value of (trim speed - speed bug).
If the difference is 0 knots, then the curve is convex, centered on the equilibrium thrust.
If the difference is positive, the curve is increasing, meaning there is an equilibrium thrust at idle.
If the difference is negative, the curve is decreasing, meaning there is an equilibrium thrust at toga.

The B777 throttle is stable if and only if there is no difference between speed bug and trim speed.
And a 1 knot difference is enough to change the curve from stable to unstable.

To build up on that, yes, you can very easily trim the airplane at a stable speed, thanks to the "clic clic" function or whatever the name of this undocumented but crucial function.
But what about times when you're not going to be in trim ? For example if you're reducing speed, you're not going to be in trim for a few seconds. The time it takes to trim. During this time your workload is uselessly increased. Let's say you're established on the ILS, and reducing speed. You ask the PM to reduce the speed bug. The airplane wants to dive. At this time, there is no simple way to keep a constant flight path. You're going to have to increase progressively your pitch, and increase progressively the stick force (except when you will start trimming, you will have to adapt the force).. Then let's say you properly managed to reach Vref+15 like so and everything's going well for now. You trim your airplane at Vref+10, while the airplane is slowing at Vref+5. Only 5 knots difference won't be very noticeable while you're still in speed reduction. Let's say you let the out of trim situation develop for 5 seconds. No effects are visible for now because they are exponential, so very little at first and much larger a few seconds later.
Then the PM asks you to complete the landing checklist or whatever other disturbance. Now if you wait 5 seconds more your engines can already be at idle.

An airplane that can go by itself from normal flight path to idle is less safe and more difficult to fly than an airplane that has a natural path stability.
I really don't see how anyone could claim the contrary.

[Vessbot]

An airplane that can go by itself from normal flight path to idle is less safe and more difficult to fly than an airplane that has a natural path stability.
I really don't see how anyone could claim the contrary.
Well, such a claim could be made if the workload added by the instability, is less than the workload removed by having 1 thing to control instead of 2. I can't speak to that myself*, but it makes sense in theory; and, the 777 has flown for decades now (I'm sure the extreme majority of those flights with the autothrottle on) without a popular complaint of difficulty to control in pitch. So from a practical everyday standpoint, it seems to be plausible.

*I can, only a little bit. My only plane with autothrottles is the 737, and I have only ever turned the AP and AT off at the same time. But some captains will leave the AT on for a little while before turning them off, and I don't see anything weird going on during that time.

[Capn Bloggs]

An airplane that can go by itself from normal flight path to idle is less safe and more difficult to fly than an airplane that has a natural path stability.
I really don't see how anyone could claim the contrary.
Don't be ridiculous. Compare apples with apples. If you want to go on about stability, disengage the ATS. If you want to use the ATS, fly the aeroplane in the appropriate manner. Don't look away for 10 seconds.

You seriously need to either just fly the aeroplane like real pilots do or change jobs. You're holding on too tight.

[BraceBrace]

Yes, you really don't understand the goal of this topic.
This topic is not a pilot's topic. It's an airplane design engineer's topic. Maybe that can clarify.
It's not a practical discussion. It's a philosophical one. My hand flying skills have nothing to do with it.
...
For example if you're reducing speed, you're not going to be in trim for a few seconds. The time it takes to trim. During this time your workload is uselessly increased.

Let's get this clear: I'm an engineer in aviation. Only touched briefly on automation in books on old autopilots, I'm more of an engine guy. Stability is teached at university, applied and tested by engineers in factories. This behaviour is known. However it does not hold it's place on a flightdeck as the "flaw" is well covered and the aircraft is well protected. It's certified and tested.

You ask the questions at pilots, not engineers, and the question is litererally "is it really helping you?" Flightdeck answer: "we don't feel it" (and the engineers will proudly whisper "we got that nicely covered from those dumb pilots... let's drink beers as well"). Everybody happy.

Flying is a learned behaviour. If you stop flying, you die. In multiple possible ways. Valid for the Airbus as well. What you do is academic re-thinking what the world knows already, and Boeing covered in an economic acceptable way. It’s a nice case study at the university :-).

[CVividasku]

Long story short, this plane may be the only non-Airbus airliner I will ever fly :p

[Uplinker]

I have read the OP again and I really don't understand what CV is talking about:

Now let's suppose you have an A/T engaged.
You have a speed bug. At first you stabilize the airplane in level flight. Perfectly trimmed. So TRS = speed bug. Let's create a gap between the TRS and the speed bug. You can trim down, but since not many planes show the TRS, let's reduce the speed bug, this will allow us to make a small reduction that we can measure. Let's say a few knots. 5 knots. How much time have you got until you're in deep trouble ?
At first, almost nothing happens, and the instructor in the sim wonders why you asked to do this weird exercise.
There is only a 5kt difference so the A/T slowly reduces thrust, the speed reduces slowly too, it reduces even more slowly because the airplane is trimmed to keep its current speed at TRS. But the A/T continues to see a speed difference and continues to reduce thrust. Soon, the airplane behavior becomes exponential.
In a time that I could not evaluate in all configurations, and that is quicker if the speed difference is higher, but in any case much quicker than you would imagine, your airplane is stabilized in a descent with idle thrust. If you were on approach, in less than 30 seconds you end up with idle thrust and terrain warning to pull up.

my bold

You seem to be suggesting that auto-thrust looks at the TRS/THS speed and compares that with the IAS ? Which auto-thrust systems look at TRS/THS ? I have flown both B737 and Airbus FBW on the line and their auto-throttle/auto-thrust both respond to IAS.

In both aircraft types: if the IAS changes, the trim will change and either the pilot, the auto-pilot or the FBW will adjust the pitch trim, depending on type and whether they are flying manually or automatically. What's the problem ?

Why would the auto-thrust keep reducing thrust when aircraft speed is below the IAS speed bug, and "crash the plane", or "trigger a GPWS warning", or end "in deep trouble", in your words ????

Auto-thrust looks at IAS, not the TRS/THS, so why would the auto-thrust keep reducing thrust when the aircraft IAS is below the IAS speed bug ? Does the B777 do this differently ?

Or are you talking about Boeing Auto-throttle systems dropping out of active mode without warning and without the pilots noticing, as in the SFO B777 crash ?

[FullWings]

I think what he is pointing out is that if you trim for a greater speed than you have bugged on a Boeing with autothrottle (and this would have the same effect in a non-FBW aircraft of any type), then let go of the controls, it will fly the trimmed speed and the A/T will take power off as it “sees” that the speed is in excess of what you have commanded.

What others in this thread have been saying is that this is a feature, not a bug. If you push forward and accelerate past the bug speed, the same thing will happen - why would you want the thrust to stay up and drive you towards limiting speeds? The assumption is that if you are flying manually, you are in control of the pitch and have put it where you want. Also, that the AFDS has been correctly programmed and you are satisfying FD commands or have switched them off.

The advantage of a C*U implementation is that even if you’re mostly focussed outside, you will get an indication of speed and speed trends from the control loadings, e.g. if you find you need back pressure to maintain the attitude you want, you know the speed has reduced before you even scan it, much in the way of a conventional aircraft. Although the aircraft does indeed do what was observed in the sim, it is only doing what you asked it to while replicating the behaviour of non-FBW aircraft, so it shouldn’t really come as a surprise?

[CVividasku]

Yes, indeed Fullwings, thank you.

It shouldn't be coming as a surprise, however very few pilots can guess what happens when you describe the scenario. Most will correctly guess that the aircraft descends, but a large part incorrectly talk about the onset of a phugoid, and very few find the conclusion of the scenario, which is thrust being idle, even with guidance.

It's because it's actually not so intuitive that a 3 knot difference would lead the A/T to go to idle.
However, the explanation is that there must be an "integrating filter" in the A/T automatics that sense the speed difference building up over time and reducing to idle in reaction to it.

You get an indication of your speed from the control loadings, but you have no indication about your flight path, or the real energy of the aircraft.
If you want an indication of aircraft energy you have to look at thrust. Which is far to the left in your visual circuit*.
I find it personnally easier to set a fixed amount of thrust according to the flight phase, then only have to look at my speed, which is very much in the center of my visual circuit.
This works because I'm keen on understanding how much thrust is exactly required. I studied the airspeed unreliable tables to make sense out of them, I often look at which angle of glide I get from idle, and what's the power setting for my flight phase.
*On the other hand, the advantage of boeing is that you can also rely on your hand to get an indication of thrust.

[Vessbot]

Auto-thrust looks at IAS, not the TRS/THS, so why would the auto-thrust keep reducing thrust when the aircraft IAS is below the IAS speed bug ?

You might want to look at my example, with numbers, in post 58: https://www.pprune.org/11680143-post58.html

It reduces thrust because speed is *above* the speed bug, because trim speed is above the speed bug.