B777 - FMA Modes: THR vs SPD

[TheBusFlyer]

Hello,

I have some questions that I would appreciate the answers to.

1) In flight school, we are taught that pitch controls speed and thrust controls flight path angle. For example, if one is slightly below the GS, in order to get back on the GS, the theory is that one would first increase thrust in order to regain the vertical flight path. Then, this would require a compensatory increase in pitch in order to maintain the speed. Is this how it is done in airliners as well? Or are airliners flown with pitch for flight path and thrust for speed since their engines take longer to spool up? I recently talked to a real-life A330 pilot who told me that airliners are flown like Cessnas - IE pitch for speed and thrust for flight path.

2) I am confused as to the difference between THR and SPD on the 777. If airliners also inherently follow the "pitch for speed, thrust for flight path" philosophy, does that mean that in SPD mode on the 777, the autopilot still uses pitch for speed and thrust for flight path? For example, assume that one is flying in V/S mode at -1000 FPM, speed 250 knots. Assume that one wants to increase speed to 270 knots. In order to accomplish this, does the plane first pitch down to accelerate to 270 knots, and then increase its thrust to maintain the v/s at -1000? Or does the autopilot use thrust to accelerate to 270 knots and reduce pitch to maintain the -1000 FPM?

3) In the 777 FCOM, it says this:

"THR – autothrottle applies thrust to maintain the climb/descent rate required by the pitch mode ".

So if I got it right, in a climb, when FLCH is selected on the MCP, the MCP speed window opens at the current
airspeed and the aircraft’s speed is still controlled through the elevators but the FMA pitch
mode changes to FLCH SPD. The FMC considers the MCP altitude setting and commands
the autothrottle to provide thrust (THR) as required to achieve the resulting altitude change in
two minutes (120 seconds).
If a climb is commanded that exceeds the ability of the autothrottle to achieve the change in
two minutes the thrust goes to the current thrust reference maximum setting.

If the airliners like the 777 are indeed flown like a GA plane (i.e. pitch for speed and thrust for flight path), then wouldn't this mean that there is essentially no difference between the THR mode and the SPD mode? In that case, what exactly is the difference between THR and SPD then?

4) Finally, what is the difference between ATHR on Airbus and AT on Boeing? I heard they worked quite differently.

Any answer would be highly appreciated.

Thanks a lot.

[Fly3]

If thrust is at a limit, either max or idle, then pitch must control speed.

[Capn Bloggs]

In flight school, we are taught that pitch controls speed and thrust controls flight path angle. For example, if one is slightly below the GS, in order to get back on the GS, the theory is that one would first increase thrust in order to regain the vertical flight path. Then, this would require a compensatory increase in pitch in order to maintain the speed. Is this how it is done in airliners as well? Or are airliners flown with pitch for flight path and thrust for speed since their engines take longer to spool up? I recently talked to a real-life A330 pilot who told me that airliners are flown like Cessnas - IE pitch for speed and thrust for flight path.
Depends! Methinks your "real-life" A330 pilot must be a flight simmer because on final approach, where you will be most likely to handfly, he's wrong. On final approach, control speed with thrust, and pitch with the control column. Jets are not Cessnas. Actually the correct way to fly a Cessna is the same; it's just that it seems "easier" to do the opposite ie pull the throttle off to go down more because of the lift effect of the propwash over the wings.

Just fly like the autopilot does on the ILS. If it gets below slope, it pulls the stick back to get back on the glideslope then corrects with thrust if it gets slow. If the speeds increases, the autothrottles pull back to slow back down. It doesn't pull the nose up,off the glideslope, to slow down! Think of the primary effects of controls. Then it will be obvious how to fly. We don't use the secondary effects of controls in jets because the reaction takes too long (unless you fly a 737...:ouch:).

If the airliners like the 777 are indeed flown like a GA plane (i.e. pitch for speed and thrust for flight path),
They aren't, depending on what you are actually doing...

I don't fly the 777 but my machine does a similar thing if only climbing a couple of thousand feet ie lower thrust than full climb thrust. However, I wouldn't use this particular unique scenario to generalise. For a prolonged climb, FLCH will set and hold climb thrust, and the pitch controls the speed. Similar for an Idle/FLCH descent; throttles put to Idle by the autothrottle and stay there (unless moved, but that is another story), pitch controls the speed.

For example, assume that one is flying in V/S mode at -1000 FPM, speed 250 knots. Assume that one wants to increase speed to 270 knots. In order to accomplish this, does the plane first pitch down to accelerate to 270 knots, and then increase its thrust to maintain the v/s at -1000? Or does the autopilot use thrust to accelerate to 270 knots and reduce pitch to maintain the -1000 FPM?
This applies to all aircraft: the thrust will be increased to increase the airspeed. The VS will be held by the AP at 1000fpm. A similar thing happens when the AP is following a VNAV path: it will use the throttles to control the speed because the pitch control is being used to stay on the path. I believe Boeings also have VNAV SPEED. Don't ask me how that works or why it exists... I never have been able to understand it.

"THR – autothrottle applies thrust to maintain the climb/descent rate required by the pitch mode ".
Sort of. For a short climb (as you described, where full power would result in too much VS and an overshoot of the altitude), yes. For a climb (or descent) in VS, yes. For a long climb, No (throttles set to max, speed controlled with pitch). For a descent in VNAV PATH (I understand), Yes. For a descent in FLCH, No (throttles at Idle, speed controlled with pitch). Note the terms Idle and Max: they are the "limits" Fly3 is referring to.

You're probably an FO in the right seat. Just remember the old adage when flying an approach: "Always remember and forever take heed, right hand for flightpath and and left hand for speed!" :ok:

what is the difference between ATHR on Airbus and AT on Boeing? I heard they worked quite differently.
I'm interested in the answer to that too.

Knuckleheads who flew onto carriers need not reply! ;)

Now I'll get my hat and coat!

[RVF750]

Basically the autopilot is a machine working on code. It's a simple device really.

As said above, pitch for flightpath and thrust for speed control. if you're low pitch up. it's far easier for a machine to understand so it's done that way.

As to modes. For Boeings, you will either be climbing level or descending. A LVL CHG command is the simplest, it will either give you climb thrust or idle and then pitch to maintain speed. VNAV climbs or descents are similar except the computer is being clever and switches from speed to mach and protects the envelope by changing speed at 10,000 or whatever is programmed into the FMC.

In V/S, again, it's simple. Pitch to give the correct VS and thrust to maintain speed as selected.

An Airbus is physically not much different. The programs and controls look different but achieve the same results.

Where you can get confused is the throttles- a Boeing will use the Autothrottle to move the thrust lever and the thrust lever tells the EEC to give you the required thrust.

An Airbus lever is effectively a thrust limiter. It will just use it know the limits you've selected. As far as I can figure out anyway.....

[FlightDetent]

because on final approach, where you will be most likely to handfly, he's wrong. On final approach, control speed with thrust, and pitch with the control column. Jets are not Cessnas. Actually the correct way to fly a Cessna is the same; There are whole world's regions where the opposite is preached, in line with the original post. And it has served them well for generations.

Indeed, for a B737 (the global largest jet fleet) it is the easier way, due to strong pitch/power couple and non-augmented flight controls. I have no opinion on other types which would be relevant to this particular thread.

[Banana Joe]

What airline does this A330 pilot fly for? I'll make sure not to board on any of their flight.

[FlightDetent]

What airline does this A330 pilot fly for? I'll make sure not to board on any of their flight. Based on ... that he might know how to fly your plane better?

[Banana Joe]

I am a very mediocre pilot after all:O

[Check Airman]

The age old discussion. You’ll use a combination of pitch and power to get the vertical speed and airspeed that you want.

Now I’m no A330 pilot, but from my understanding of the control laws, using power to maintain a vertical rate on final would make for an interesting ride down the glideslope.

[Check Airman]

There are whole world's regions where the opposite is preached, in line with the original post. And it has served them well for generations.

Indeed, for a B737 (the global largest jet fleet) it is the easier way, due to strong pitch/power couple and non-augmented flight controls. I have no opinion on other types which would be relevant to this particular thread.

Are you saying that you use pitch for speed on final in a 737? No experience on the type, so genuinely curious.

[Bus Driver Man]

Based on ... that he might know how to fly your plane better?
Based on the fact that that A330 pilot doesn’t know how the FBW on his aircraft works.
How can you fly with thrust for flight path if the FBW system trims for any speed or thrust changes?

[Vessbot]

Based on the fact that that A330 pilot doesn’t know how the FBW on his aircraft works.
How can you fly with thrust for flight path if the FBW system trims for any speed or thrust changes?

AP on in OP CLB/DES with manual thrust? :ok:

[itsnotthatbloodyhard]

For anyone that uses pitch for speed down final - if your speed’s bleeding at 100’, do you still pitch down?

[Capt Pit Bull]

OP hasn’t been taught basic principleS of automatic control. Any discussion is built on sand.

[Capn Bloggs]

AP on in OP CLB/DES with manual thrust? https://www.pprune.org/images/smilies/thumbs.gif
Obviously, but that is just clouding the basic issue.

[AerocatS2A]

What I have found as I’ve moved into bigger, faster, and more slippery machines is that in a jet on approach, a small change in pitch causes a moderate change in flight path but a only a small change in speed, therefore pitching for speed doesn’t work very well. Point it where you want it to go and adjust thrust for speed makes the most sense to me! On the other hand, in a C172 a small change in pitch only causes a small change in flight path and causes a moderate change in speed, so pitching for speed in something slow and draggy works if you prefer to think of it like that. The truth is, as said up-thread, it is the combination of pitch and thrust that gives you your performance and if you change one then you will probably have to change the other as well.

[pattern_is_full]

"Power for altitude, pitch for speed" is a good basic rule for newbie pilots.

It tends to avoid stalls if inexperienced people are taught to not simply "pull the UP lever (the yoke)" for climbs, but add power. And generally it is the way aerodynamics work anyway - reduce or increase power without a trim change, and most fixed-wing aircraft will tend to nose down and descend, or nose up and climb, all by themselves. (High thrust-line aircraft sometimes only after a brief excursion in the other direction).

And without a power change, pitch will certainly change the speed - sometimes dangerously.

However, the dynamics of tracking a glideslope, with a heavy high-inertia aircraft, with slow-spoolup turbines, on a constant descent angle, generally work more responsively with small corrections being done with "pitch for altitude, power for speed." If you are 400 feet AGL and below glideslope, you want to GO UP RIGHT NOW - faster that turbines/thrust/power can respond. Pitch responds faster.

It is just an adjustment in habit that is part of that advanced training.

[FullWings]

Especially on a FBW aircraft like the 777 where adding power does nothing to the pitch: you would have to wait for a change in speed before seeing a pitch change as trim takes effect. Rather messy if you’re trying to follow a glide slope and possibly worse than that for a low-level go-around...

[FlightDetent]

Since the stoning has ceased, perhaps it's good courtesy to return some pieces for another round of throws:

Airbus FBW has a reactive auto-trim, hence there always is displacement due to pitch/power couple. Can't be flown hands-off although the deviation is small. I'm also told that for landing at 100' RA that feature stops completely for the A330.

I think the passengers of AF447 wish they had a pilot who would pitch for speed first, instead of firewalling the thrustlevers. A330.

I think the owners of EK521 wish they trained the pilot to use thrust forward to climb and vice-versa. B777.

There are many non-jet A/C where there is no pitch-power couple, ATR to name one. Using power to manoeuvre in pitch is pointless, and insisting not only silly but just would not work. On those, yes.

On a machine with a decent power-pitch couple, using pitch (elevator) for V/S and only then thurst for speed is sloppy. But also very easy as the PFD tells you exactly that. Flight director command first, and later the speed trend arrow. CoM?

[itsnotthatbloodyhard]

Coincidentally, posted on AvHerald on 24 March:

”An Alliance Airlines Fokker 100 on behalf of Virgin Australia, registration VH-UQN performing flight VA-1251 from Brisbane,QL to Rockhampton,QL (Australia), was on final approach to Rockhampton's runway 33 when the aircraft encountered moderate turbulence at about 300 feet AGL, which caused the loss of airspeed. While attempting to arrest the loss of airspeed, the airspeed fell below the minimum approach speed because the flight crew failed to push the throttles forward.”

[rudestuff]

The age old discussion. You’ll use a combination of pitch and power to get the vertical speed and airspeed that you want.

Exactly this. In a Boeing if I want to descend at 1000fpm I can use V/S (thrust for speed, pitch for r.o.d.) or I can use FLCH or VNAV SPD (Pitch for speed, adjusting thrust for r.o.d.)

In either case the aircraft does the same thing because it's always a combination of pitch and power, the only difference is how you choose to think about it.

[Capn Bloggs]

In either case the aircraft does the same thing because it's always a combination of pitch and power, the only difference is how you choose to think about it.
Seriously? If you wanted to descend at 1000fpm why you would even think of using FLCH? I know you can, but why?

[rudestuff]

Seriously? If you wanted to descend at 1000fpm why you would even think of using FLCH? I know you can, but why?
No idea. My point was to show that flight path is defined by pitch AND power.

[maui]

Blogs.

Because it will lock the SPD, which in many areas (not WA) is more important than precise V/S control. Think busy airspace and tight separation.

Cheers Maui

[FlightDetent]

Does this one look like pitch for profile and thurst for speed mis-applied? I am genuinely curious if Q400 does have a relevant pitch-power couple at normal speeds.
airplane was at an airspeed of 131 knots. FDR data showed that the control columns moved aft at 2216:27.8 and that the engine power levers were advanced to about 70° (rating detent was 80°) 1 second later.18 The CVR then recorded a sound similar to increased engine power, and FDR data showed that engine power had increased to about 75 percent torque. FDR data also showed that, while engine power was increasing, the airplane pitched up; Power forward (to add energy enabling certain altitude control) and pitch down (against the flight director) to keep safe flying speed - preferably in the opposite sequence. That is the agreed prevention strategy against re-occurence of that body-count.

I doubt strongly the stall recovery technique for B777 is any different. Sure as hell the A3x0 it is exactly as above. I like to think, within the scope of this thread's discussion, it's better not to unlearn the basics which your passengers might need to survive some rainy night when the pilot will only be reacting instinctively.

The statement "On jet A/C you pitch for profile and thrust for speed (change your habits)" with or without an exclamation mark irks me then.
- It suggests a complete change of perception, removing a vital learned skill. Even for situations where very valid, it sholud only be an overlay finesse how to best manipulate the flight controls.
- It suggests applicability for all jet aircraft but even when valid it does not apply to them all, jets with underslung engines specifically not (unless perfectly path stable which the A320 just isn't).
- It suggests non-applicability for prop aircraft but when valid it does actually apply to them.
- It encourages flight director laziness once a routine builds in.

[AerocatS2A]

You are way overthinking this FlightDetent.

[Vessbot]

This is a complex issue that deserves some thought, but I think the one thing we can all agree on is how far we are from putting the final nail in its coffin ;)

Pitch for flight path and thrust for airspeed is the system that:

comes intuitively first to mind
simplest to understand
most direct in mechanic
has the quickest reaction (but not always, for thrust for airspeed. This could even act in reverse)

I’ll call this “System 1” from now on for brevity.

Pitch for airspeed and thrust for flight path (System 2, if you will) is the one that is:

less direct
takes longer to act (except pitch for airspeed, that is immediate)
is tougher to understand

All of the above are negatives, but it has one tremendously important thing going for it: it is the more universally applicable one, since it is based on the fundamental flight dynamic relationships that in steady flight A) excess thrust determines climb angle, and B) AOA determines airspeed. These are true regardless of whether you’re talking about prop or jet, straight or swept, clean or draggy, heavy or light wing loading, which way the thrust pitch couple goes, manual or FBW, which control law, or any other distinction.

After all, you can set any thrust setting from idle to full, and with it fly at any airspeed from stall to Vmo/Mmo. Bearing that in mind, I don’t see how one can very confidently say that thrust controls speed ;). However, if you set one AOA, it will yield one and only one airspeed.

While the above paragraph seems to pose a conundrum for the easy and common sense notions of System 1 (the control inputs don’t do what they’re supposed to!), the astute reader may notice that the conundrum is solved by noting that these System 2 relationships are true in the longer-term, steady-state arena, thus allowing System 1 to act immediately, for quick changes.

(But even in the immediate arena things aren’t so clear-cut for System 1. With a “normal” thrust pitch couple, a thrust increase can yield a quite significant speed decrease, as we’ve explored in so many 737-related discussions.)

Ultimately, both systems must be fully understood, embraced, and used together.

Transitioning from level flight to a climb according to only System 1 (pitching up to go up, then noticing an airspeed loss, and reacting to it with thrust) is sloppy, reactive, and dull-witted. You KNOW the airspeed loss is going to happen, so why not act on it earlier and prevent it by simultaneously increasing thrust? Using only System 2 might be even sloppier. Increasing thrust by itself to inject yourself into the middle of a phugoid (thrust causes a speed increase, which causes a lift increase, which makes you go up, which causes a speed loss, which overshoots and then does everything again but with loss instead of gain) it will take many cycles over many minutes (if at all) to settle down in the climb, and the result would obviously be atrocious.

But a simultaneous combination of thrust increase, along with an attitude increase with the practiced use of trim to lock it into the new attitude (which itself is known from experience) is the way to go. But even this isn’t so clear-cut. Even something as simple as raising the nose, might not involve pulling back on the stick at all, depending on the thrust pitch couple. My current plane has a “reverse” one (thrust gives you nose down, a source of displeasure to me) but back when I was flying a normal one, I took great pleasure in the elegance (which I define as accomplishing what you want with the fewest inputs possible) of starting a climb by increasing thrust slowly (so as to pitch up via the TPC and inject myself into only a mild phugoid), and giving a well-timed blip of nose-down trim. That’s all it took! I started the climb without ever pulling back or trimming nose-up. And if done well, it would go straight to the new attitude and climb angle, at the original airspeed, with no overshoots of anything.

An even more open-and-shut example against only System 2, is being on the glideslope. Even though I’m one of the most outspoken proponents of System 2 awareness and usage, it’s obvious you can’t simply nose down to get speed. (You’ll get speed allright, but...) In this case a System 1-only response is also inappropriate, but more mildly so. If you only increase thrust, it’ll soon lead to a glideslope deviation, which if then corrected, is sloppy and reactive (it’ll then cause a speed deviation, which if then fixed with only thrust puts you back to step 1, in an ongoing cycle of unnecessary deviations and corrections). I think this is a clear demonstration of the simultaneous inputs being the most proactive and neat: If the initial deviation is only a speed one, then the fix is an increase in thrust along with a decrease in AOA, which requires a trim composed of a nose-down component for the AOA, and a component for the thrust pitch couple which could go either way. In my plane, for example, the TPC component dominates. So if I need to speed up a bit, I give it some thrust and some nose-up trim. If the initial deviation is glideslope only, then you need a simultaenous thrust change, attitude change, and trim for the TPC.

Now I’ve gone over (in possibly too much detail) a few different examples in regular flight, but the greast importance of System 2 comes at the edges of the envelope and emergencies. In AF447 or Pinnacle 3701, where there was no excess thrust but they were trying to get altitude, everybody’s fate was sealed by the pilots’ brains being locked into System 1, and treating the stick as the go up/go down control. Whether System 2 was ever even mentioned in their earlier training, (at best it was probably treated as a passing curiosity and never taken seriously) years and years of everyday flight in the middle of the envelope, using the stick as the go up/go down control, wired up their reflexes to only think of it that way. So when the emergencies unfolded and their field of vision shrank to a soda straw, eliminated their capacity of any deliberative thinking, and allowed them to act as only automatons by their wired-in reflexes, there was then no realistic hope of anything happening other than what happened. The going got tough, and they did the only thing they knew in their gut: they wanted to go up so they pulled on the stick.

These accidents will continue to happen, I see no realistic way around it. One way, that’s not realistic, is what I would do if I was emperor of the world: nobody would be allowed to touch a powered plane without some number of hundreds of hours in a glider first. Hopefully then primacy might lock in the correct response to any low airspeed problem, that the immediate way to safety is stick forward and nose down. (The increased thrust to counter the descent is only the cherry on top, that might or might not be available.)

[Capn Bloggs]

Very well written, Vessbot. :ok:

[vilas]

but the greast importance of System 2 comes at the edges of the envelope and emergencies. In AF447 or Pinnacle 3701, where there was no excess thrust These are not examples of the edge of envelope but well beyond it. You don't follow a procedure of stall recovery for normal approach. Whenever one is tracking a flight path it is maintained by pitch and speed managed by thrust. It's straight forward. When thrust is constant whether at idle or climb you are not tracking a flight path but maintaining a certain speed so obviously it has to be maintained by pitch.

[Capn Bloggs]

Vilas, well said.

For those of you advocating/thinking of using System 2/secondary effects of controls for normal operations, have a read through these incidents, both involving use of/demonstrating the trap of using FLCH on an approach:

https://www.atsb.gov.au/publications/investigation_reports/2011/aair/ao-2011-086/

https://www.atsb.gov.au/publications/investigation_reports/2007/aair/ao-2007-055/

Maybe they thought they'd get more precise speed control...:rolleyes:

[maui]

Bloggs



I do not believe anyone has advocated use of FLCH on approach. Descent yes, approach no. In fact, in my recollection, the only person that has even mentioned it in an approach context, is you.



The Boeing 777 FCTM refers to it thus: (my bolding and underlines)


Non-ILS approaches are normally flown using VNAV, V/S, or FPA pitch modes.

The use of FLCH is not recommended after the FAF. Recommended roll modes

are provided in the applicable FCOM procedure.

The cases you cited are not relevant. Both were related to standards / training, rather than “normal ops”.

BTW you still haven’t answered my earlier question about the existence or otherwise of IAS mode in your 717 AFCS.

Cheers. Maui.

[Vessbot]

These are not examples of the edge of envelope but well beyond it. You don't follow a procedure of stall recovery for normal approach. Whenever one is tracking a flight path it is maintained by pitch and speed managed by thrust. It's straight forward. When thrust is constant whether at idle or climb you are not tracking a flight path but maintaining a certain speed so obviously it has to be maintained by pitch.I wrote some subtle and detailed things in that post, that you seemed to skip right over to argue against something I didn’t say. What I did say about a normal approach is in the paragraph that starts with “An even more...”

But if subtleties aren’t the order of the day, I’ll say something blunt here instead:

In a “stall recovery” your AOA is too high and you must reduce it. Period.

In a “normal approach” with low airspeed, your AOA is too high and you must reduce it. Period.

(Also it is very much not “straight forward” that speed is managed by thrust, whether you’re tracking a flight path or not. If it was, then if your speed was low then you could straight forwardly fix it by simply increasing thrust. And that is simply not the case. I remember vividly one event flying with a student I’d flown with for years, at a new airport in a slightly stressed situation. His airspeed got a little low, I prodded him “watch your airspeed,” and in a startle he simply added some throttle, and we lost more airspeed. That’s when I did a Captain Picard double facepalm and lost some hope that everybody can be really taught to fly.)

[tttoon]

Pitch for speed and thrust for altitude can in certain situations delay the re-stabilization of the aircraft due to an incorrect initial action. Take, for instance, an aircraft that is high and fast on approach. If pitch for speed is used, the aircraft will be pitched up and thrust will be reduced. As the aircraft slows down, it becomes more efficient, and the descent rate will be (much) less than in the other case, delaying the stabilization of the approach. If, on the other hand, the aircraft is pitched down towards the glidepath and the thrust is reduced, the speed will remain higher (but below the applicable placard speed) and as a result the drag will be higher. The descent rate will be higher and a stable approach will be achieved in the shortest possible time.

Pitch for speed etc. is a simplification that may work for small aircraft (and delta wings, but they are a special case), but if you think about controlling the aircraft as a form of energy management (the amount and balance of kinetic and potential energy) one thing that becomes clear is that you will initially always need to pitch towards your desired flightpath, and then adjust thrust for speed. This assumes that you are in control of the aircraft, so not in a stall.

[FullWings]

Vessbot, I like your explanations a lot.

I think this subject causes confusion, as you rightly point out, if someone doesn’t have a complete understanding of static vs. dynamic, primary and secondary effects of controls, flight envelopes, AoA, performance limitations, energy management and path constraints. Quite a list and it helps to be at least slightly mathematically minded.

As someone who started flying in gliders then went into power, when introduced to the “power” method of getting out of a stall I was :confused:. Later, we changed to the more sensible option of reducing AoA before doing anything else.

Going back to the original thread title, understanding what each FMA annunciation implies and how the systems interact in various phases of flight is essential knowledge but generally buried deep in multiple sections of the FCOM. Well worth reading and digesting at a time of low workload...

[Vessbot]

you will initially always need to pitch towards your desired flightpath, and then adjust thrust for speed.You are advocating that if low and correcting up then you should lose speed first, and then increase it back to the target value. Why?!

The necessity of thrust to fix the vertical flight path is primary (in importance, though not necessarily sequence) as, again, excess thrust dictates climb angle.

Had the pilots of Asiana 214 really grokked that, they might have come in with thrust as they sank further and further below the glideslope, instead of leaving it relegated to the back of their minds as some side effect to clean up later once the main problem had been solved (and then, in actuality, forgotten). No, they pulled more and more on the stick (ending up with dozens of pounds of pull force, IIRC) not realizing that by doing so they were commanding a decrease in airspeed. (As, again, AOA dictates airspeed).

Back when I was teaching I remember flying with countless of other people’s students checking them out in new types (along with a few other factors such as that it’s tailwheel, minimal instrumentation, etc., that took them out of their comfort zone and loaded up their task capacity) and with regularity would watch their flight path fall off toward a point short of the runway, and then watch their response of pulling up their nose without power, which after the first few seconds’ initial (and temporary) success of pulling up the flight path, (and thereby tricking them into thinking the job was done) the induced drag at the lower speed would increase and the flight path would consequently fall off even further, then they’d respond with pulling the nose higher, feeding back into the same loop of higher AOA -> lower speed -> more induced drag -> more sinking tendency -> higher AOA. Each of these was a mini Asiana 214 in the making. And each of these could have been avoided with a thorough and early understanding of System 2 and that thrust is the primary control of vertical flight path.

Let’s consider the consequences of a stressed pilot initially forgetting one or the other (of thrust vs. pitch). If they forget thrust, the consequence is described in the above paragraph.

If they forget pitch, the consequence is much less severe. If the plane has a normal TPC, then the low plane would pitch up, beginning the glideslope correction with the same input. May be sloppy, but can be cleaned up later, and not dangerous. If it’s a reversed TPC, then the plane would start pitching down, which would surely catch the pilot’s attention and he’ll then pitch it up. Sloppy and slightly more dangerous than the normal TPC, but far less dangerous than if they forgot thrust. And it’s far less dangerous because total energy is gained which increases the maneuvering ability, thus increasing their available time to react (whether this ability is used toward a re-stabilization, or toward a goaround).

[tttoon]

You are advocating that if low and correcting up then you should lose speed first, and then increase it back to the target value. Why?!

I'm not saying that at all, a pitch up to correct the flight path should be coupled with an increase in thrust to anticipate the increased drag due to the pitch increase if needed. Would you say that if low and fast the correction should be to add thrust and pitch up? That will result in the aircraft getting back on G/P, still fast, and closer to the runway, opening the door for a whole range of other threats. An incorrectly trimmed aircraft will still plow into a field with any amount of excess thrust.

Any pilot should have at least an underlying reflex to increase thrust when they pull back on the tick, and the opposite for the reverse. They should be aware of the energy state of their aircraft and apply the required thrust after making the required pitch change.

To be clear, what I'm describing is the normal way (in my opinion) of controlling the aircraft, and is not applicable for stall recovery, where control is lost and the FCTM recovery should be applied.

Unfortunately, as seen in the Asiana incident and similarly in the botched EK go-around, over-reliance on automatics has removed those reflexes from many pilots. Couple that with a bad understanding of AFDS modes and equally bad monitoring of FMA changes and you have an accident.

[Vessbot]

Tttoon,

We’re on the same page then. I agree with you a hundred percent that a low and fast is fixed by a pitch-up alone, and that a low and on-speed is fixed with simultaneous pitch and thrust; and I agree a thousand percent that everyone should have a reflex to adjust thrust with pitch.

But I’ll add that this reflex is lost not only due to automation overuse, but was never developed in the first place in many people, including all the students I mentioned in my last post, most of whom had only bugsmasher experience. I think two main factors go into that together.

First, there is instinct from non-aviation life, where you point the vehicle (or your human body) in the direction it is to go. It’s common sense. Then when someone starts learning to fly, there is the much more direct-acting and easy-to-understand System 1, which this previous instinct slots right into. If you’re headed a bit short of the runway, it’s extremely easier to make the mental connection from pointing your nose up a bit to the flight path correspondingly going up that same bit, than to abstract heady stuff like total energy rate and excess thrust and AOA and all these things that it takes to really understand System 2.

So the above is instinct, and the first urge of what control to react with. The other aspect is which problem is more apparent. The flight path (associated with pitch in System 1) has a big visual cue, and is much more apparent than airspeed loss, which, for small and moderate amounts, is just a number on a gauge (out of the main vision field) and some small amount of pull force on the stick, which is very easy to miss for an unfamiliar pilot (or a familiar one who’s stressed and death-gripping the stick of his low-stick-force bugsmasher... or even his high-stick-force 777). For a task-saturated pilot, the perception of the flight path falling off takes up all the mental space and doesn’t leave any to share with the perception of the airspeed falling off.

When you combine these two factors, the flight path is seen and takes up all the attention of what needs fixing; and then the common-sense System 1 fix of using pitch is the one that comes in as a reflex, forgetting the thrust. And this is how we enter the loop of all the mini Asiana 214’s from my last post. System 1 edges out System 2 from both ends of the situation: the perceived problem, and the mentally available solution.

System 1 is in no danger of being forgotten. This is why I’m such a huge proponent of System 2 thinking taught from the get-go as primacy: not to overshadow System 1, but rather to bring it (System 2) from itself being overshadowed, up to some level of parity. It is not reflexive, it is not common sense, it takes some brainwork to really understand, and anything with these qualities needs all the help it can get to be recalled and used when necessary. And it’s when it’s the most necessary, that it’s furthest from easy recall due to mental load.

[Vessbot]

And to address your high and fast example from a previous post: Yes you have to pitch down initially. Pitching up and flying even higher above the intended path would be senseless. But the AOA is still increased. It’s just a a slow increase, corresponding with the airspeed bleedoff rate. An immediate increase would result in a balloon, same as a flare with excess speed and a pre-programmed stick pull (that would have been correct for a lower entry speed).

When you’re talking about these tings, you attach labels like “pitch for airspeed” that represent more subtlety about rates, etc., than can really be contained in 3 words. Also that by “pitch” we really mean AOA rather than directly pitch. But the elevator controls AOA and we sometimes use “pitch” as a synonym for “elevator usage” and there you have some confusion.

[maui]

Vessbot.
I agree with all you have said but jeez I didn't realise it was so complicated. I always thought it was point and push or pull as appropriate.

The one thing though that is missing from your analysis, is any mention of the importance of a good scan. Without it, all talked about above, will add to the complexity of the operation. With it, all becomes manageable.
Do like the "death grip" reference, which along with deficient scan, are the two principal causes of manipulative difficulties.

Cheers. Maui

[Vessbot]

Vessbot.
I agree with all you have said but jeez I didn't realise it was so complicated. I always thought it was point and push or pull as appropriate.

The one thing though that is missing from your analysis, is any mention of the importance of a good scan. Without it, all talked about above, will add to the complexity of the operation. With it, all becomes manageable.
Do like the "death grip" reference, which along with deficient scan, are the two principal causes of manipulative difficulties.

Cheers. Maui

​​​​​​Completely agreed with the good scan. Definitely can't process or act on the relevant information, unless you've perceived it to begin with. Cheers!

​​​​​​

[vilas]

In a “normal approach” with low airspeed, your AOA is too high and you must reduce it. Period Not necessarily. It depends. If you are on correct glide path your low speed is due to insufficient thrust and just adding thrust will do the job. Pitcthing down for speed will take you below the path then you pitch up for GS adding thrust then when on correct path you will have to reduce thrust again. If the speed was dropped due going above then pitch down and pick up the speed and path. Flying correct path with sustained low speed is insufficient thrust, just add it.

[Vessbot]

@Vessbot: a very detailed and thoughtful analysis but in day-to-day ops the task is either to follow a defined flight path (pitch for path, thrust for speed), or otherwise pitch for speed and thrust for vertical performance.

V/S and FPA modes can be considered as defined flight paths.

Most of us senior aviators find it incredible that none of the three pilots on the flight deck of Asiana 214 noticed the decreasing airspeed until too late.The NTSB, and common discussion about that crash, describe what happened as “insufficient monitoring of airspeed.” While true in a dry and technical sense, it doesn’t come close to capturing what was really going on. It’s only part of it. It’s like saying that if someone crashed their car straight into a wall at 90 degrees, with their foot on the gas pedal down to the floor, that he had “insufficient monitoring of closure rate to the wall.” I mean OK but... not really.

The failure was at an earlier point in the brain/control loop. By pulling on the stick with XX pounds, he was commanding the airspeed to decrease by that much! Just “monitoring” implies that they failed to notice some external event that might or might not have come, and they had to be on their guard to catch it in case it did. If they lost airspeed due to a sudden downdraft or tailwind shear, that’s monitoring. If they lost fuel due to a leak, that’s monitoring. If the FA called asking to turn the temperature down because the thermostat went crazy, that’s monitoring. If someone called them on 121.5, that’s monitoring.

If they lost YY knots because they were pulling on the stick with XX pounds, that’s not just monitoring. AOA determines airspeed, and they didn’t merely fail to monitor, but rather they caused that airspeed decrease by pulling on the stick and increasing the AOA.

Someone with an understanding of System 2 sees that clearly; while according to System 1, pitch control is merely for flight path and the airspeed loss was a separate incidental event that they would have caught had they just been monitoring better, like the fuel loss or the temperature in the cabin. No, the connection is causal.

And even in regular everyday flight in the middle of the envelope while following a defined fight path, AOA (even though primarily being used for tracking that path) still determines airspeed just the same. And that fact must be embraced like lives depend on it, because they do.

System 1, while simple and easy for an introductory student to see the direct effect of the controls, is not sufficient to truly understand flight.

[Vessbot]

Not necessarily. It depends.

This is not true. It does not depend on anything. If airspeed is to be increased for a constant flight path, AOA must be reduced. If the latter does not happen, the former will not.

[itsnotthatbloodyhard]

If airspeed is to be increased for a constant flight path, AOA must be reduced. If the latter does not happen, the former will not.

I’m not sure if it’s what you mean to say, Vessbot, but it sort of seems you’re implying that AoA must first be reduced before speed can increase, whereas I’d argue that one goes hand-in hand with the other. (Of course, if you’re far enough on the back side of the drag curve that thrust is no longer adequate, then the only way that airspeed will increase is to get the AoA off first - but that’s not the case with a normal approach.)

The alternative statement to yours is that if a constant flight path is to be maintained with increasing airspeed, AoA must be reduced accordingly - which I’d suggest is what we see day-to-day when following any defined path (e.g. accelerating in level flight).

in my younger days I used to fly something that went down final on the back side of the drag curve, using AoA as a primary reference. If for some reason you got increasing AoA/bleeding speed close to the ground, reducing AoA as an initial reaction wouldn’t help much - you’d probably just hit the ground with a higher rate of descent. The only effective response was an immediate increase in thrust, which came naturally as it was the normal technique used anyway. Same goes for flying an airliner. If you feel the arse falling out of it at 100’, are you going to lower the nose, or give it a handful of thrust?

[Vessbot]

I’m not sure if it’s what you mean to say, Vessbot, but it sort of seems you’re implying that AoA must first be reduced before speed can increase, whereas I’d argue that one goes hand-in hand with the other. (Of course, if you’re far enough on the back side of the drag curve that thrust is no longer adequate, then the only way that airspeed will increase is to get the AoA off first - but that’s not the case with a normal approach.)

The alternative statement to yours is that if a constant flight path is to be maintained with increasing airspeed, AoA must be reduced accordingly - which I’d suggest is what we see day-to-day when following any defined path (e.g. accelerating in level flight).

in my younger days I used to fly something that went down final on the back side of the drag curve, using AoA as a primary reference. If for some reason you got increasing AoA/bleeding speed close to the ground, reducing AoA as an initial reaction wouldn’t help much - you’d probably just hit the ground with a higher rate of descent. The only effective response was an immediate increase in thrust, which came naturally as it was the normal technique used anyway. Same goes for flying an airliner. If you feel the arse falling out of it at 100’, are you going to lower the nose, or give it a handful of thrust?Did not mean to imply any sequence order in that comment. Just the final state in both variables, compared to the initial. Certainly if slow and draggy, leading with nose down will destabilize things downward. A simultaneous change, or even leading a bit with thrust, would be right. But ultimately, AOA will have to be lower at the new higher speed. (If it isn’t, then the flight path will go up.)

To expand on this (not for you, you seem to be fully familiar especially given your background; but for the thread generally, as I didn’t have much time last post): it comes from the lift equation, which says that:

Lift = CL times V^2 times a few constants that we can ignore.

Since CL is a function of AOA, you can essentially take “CL” to mean “AOA” as a factor that varies under your control of the elevator. So now we’re left with essentially Lift = AOA times V^2.

For a constant flight path, weight is constant and therefore lift is constant. Since lift is constant, and equals two things multiplied by each other, those two things are inversely related. In other words, airspeed and AOA are inversely related; and if one goes up then the other goes down, and vice versa. Every value of AOA, after flight is stabilized, will yield one and only one airspeed (for a given configuration, weight, altitude, etc.)

Like this representative airplane here:
https://cimg0.ibsrv.net/gimg/pprune.org-vbulletin/1639x1075/e7ebd2c3_492f_486a_bebd_408c023e80c6_84f09da5f81715ebf9b2d35 a0b078c9bbdef1299.jpeg

And yes my “arse falling out” response is thrust, which is part of what I’ve been trying to say the whole time. The real question is are you gonna give it a handful of thrust or raise the nose, which is the reaction of the so-called “straight forward” pitch-for-flight-path thought. (If we’re making a forced choice of thrust or pitch... which is not a forced choice in reality)

[Capn Bloggs]

The real question is are you gonna give it a handful of thrust or raise the nose, which is the reaction of the so-called “straight forward” pitch-for-flight-path thought.
No it isn't. With system 1, you raise the nose so you don't hit the ground. You simultaneously put the thrust up (if you need to) to to stop the speed decreasing. If you like (and you'd have to do the same with system 2) you can put on full thrust. As soon as you approach the normal slope, stuff the stick forward to maintain it. Or would you (system 2) just pull the thrust off a bit after it had kicked in and just hope you end up going where you want to go?

I see it often. At 200ft, the @rse falls out of it because of a bit of a hole. On goes the thrust (overriding the ATS), we get lower and lower below the slope going faster and faster, the thrust now kicks in and we're now going too fast and land long. Simple fix: raise the nose then regain the slope! If the ATS doesn't hold the speed, put on some thrust yourself.

[Vessbot]

No it isn't. With system 1, you raise the nose so you don't hit the ground. You simultaneously put the thrust up (if you need to) to to stop the speed decreasing.

Yes, if you’re doing everything right. I admit I was a bit unkind in my inference that a System 1 person would necessarily go pitch only. But I was cheekily trying to goad anyone who is still dismissive of the whole thing and maintains that pitch for flight path is so “straight forward.” After all, the deviation of an arse-dropper is a lowered path, no? So the foreknowledge that you will need to add thrust (as you sensibly suggest, instead of waiting for a speed loss and then reacting) is a complication on the supposedly straight forward.

If you like (and you'd have to do the same with system 2) you can put on full thrust. As soon as you approach the normal slope, stuff the stick forward to maintain it. Or would you (system 2) just pull the thrust off a bit after it had kicked in and just hope you end up going where you want to go?

The System 2 description of the GS recapture after the upward correction, is that you’re simultaneously (or close to it) A) reducing the thrust to reestablish the excess thrust value for -3 degrees path, and B) lowering those nose to maintain the correct AOA for the speed (flight path down and nose down the same amount). It’s a less intuitive and direct thought pattern than “lowering the nose to point the path back at the thousand footers” but it is what’s happening just the same.

I see it often. At 200ft, the @rse falls out of it because of a bit of a hole. On goes the thrust (overriding the ATS), we get lower and lower below the slope going faster and faster, the thrust now kicks in and we're now going too fast and land long. Simple fix: raise the nose then regain the slope! If the ATS doesn't hold the speed, put on some thrust yourself.

If they were zooming below the GS with no pitch to go with their thrust, then yeah that’s senseless.

[vilas]

The cause of the Asiana prang was inadequate control of airspeed rather than inadequate monitoring. To control speed you have to look at your speed and thrust.They never looked at speed which reached 31kt below Vapp with thrust stuck at idle all along. SFO is a bad example to discuss normal approach. Normal approach within +_10kts or half a dot below or above is straight forward. You pitch for slope and as thrust comes up you will change AoA to maintain it. In airbus FBW the TW couple is not going to take you to GS.

[Vessbot]

To control speed you have to look at your speed and thrust.They never looked at speed which reached 31kt below Vapp with thrust stuck at idle all along. SFO is a bad example to discuss normal approach. Normal approach within +_10kts or half a dot below or above is straight forward. You pitch for slope and as thrust comes up you will change AoA to maintain it. In airbus FBW the TW couple is not going to take you to GS.

AOA determines airspeed, therefore pulling on the elevator controls the airspeed to decrease. This is just as true for 10 knots on a normal approach as it is for 31 knots at SFO; there is no difference.

(Yes if you want to also maintain a straight vertical path then you have to consider thrust, and the smoothness and timing of AOA vs. thrust changes, but that is a separate matter.)

[vilas]

Day in day out people are flying pitch for GS and thrust for speed. If you raise the pitch off course the speed will decrease so thrust will be increased simultaneously unless your speed is high and you want to loose it. SFO they were on verge of stall so let's not talk about it.

[Vessbot]

Day in day out people are flying pitch for GS and thrust for speed. If you raise the pitch off course the speed will decrease so thrust will be increased simultaneously unless your speed is high and you want to loose it. SFO they were on verge of stall so let's not talk about it.Thee are two separate but related things going on in is discussion, that it would help to untangle and look at separately. (Though they are occurring simultaneously)

First is the pilots’ intent of what the pitch is to achieve, which is something occurring in their heads. Yes day in and day out they are pitching for path, I don’t doubt that.

But, second, is the physical fact that AOA determines airspeed and pulling on the elevator controls airspeed downward. This is true 100% of the time, whether in a minor GS deviation in a normal approach, or what happened at SFO. I stress again that there is no difference in concept, only a difference in degree which is irrelevant. So the SFO event must be talked about, because it is a function of the same laws of flight dynamics, long before they approached a stall.

On a normal approach half a dot low, if the pilot pulls the elevator back a tiny bit to raise the nose a fraction of a degree, his intent is to adjust the path, which will happen successfully; but, simultaneously, the increased AOA is commanding an airspeed reduction. Whether he intends it or not, whether he’s aware of it or not, it is happening. If everything goes normally, he will increase thrust commensurately with the shallower path and return the elevator to the original position, commanding the wing to the original AOA, and the airspeed to the original value. Hunky dory.

But if he doesn’t come in with thrust and continues to pull the elevator back commanding a higher AOA and a lower airspeed, even by 5 knots, that is the SFO sequence. There is no dividing line. Where would such a line be, between 5 knots and 31? 10? 15? 16? No, The SFO sequence was a manifestation of the exact same flight mechanic as normal flight. Every time you pull on the elevator, no matter how slightly and how day-to-day, you are commanding an airspeed reduction and potentially entering that same loop. And to disregard the relevance of the fact that you’re doing that (regardless of that you’re primarily using AOA for something else at the same time), is to make the same mistake that the Asiana pilot did.

[Capn Bloggs]

So what are you suggesting, Vessbot? Half a dot low, push the power up? Wait a few seconds to see if it was the correct amount? Meanwhile, you've just landed short. In an Airbus (as Vilas pointed out) nothing happens because there is no pitch-power couple.

I wonder if the FO was doing that in this incident?

https://www.flightglobal.com/safety/lack-of-alertness-improper-operation-behind-china-airlines-freighter-undershoot/137667.articleAs the freighter approached Runway 05L, it appeared that the altitude was higher than the approach glide path. While the altitude got corrected immediately, at around 137 feet the aircraft again began to fall below the glide path.

Investigations revealed that the co-pilot had failed to maintain the normal speed and rate of descent by utilising pitch control and thrust. He failed to determine his mistake in time, which led to the hard landing before the touchdown point.

I agree with Vilas; AOA has nothing to do with this discussion which is now going off on a AOA tangent. Speed is what we see in the cockpit. You pull the nose up a bit, you might have to increase thrust. Might. You pull it up a lot, you will have to. But you don't teach people to fly by using the power to stay on the GS. Maybe in a 737, but not in anything else.

[Vessbot]

So what are you suggesting, Vessbot?
This:

But a simultaneous combination of thrust increase, along with an attitude increase with the practiced use of trim to lock it into the new attitude (which itself is known from experience) is the way to go.

I think this is a clear demonstration of the simultaneous inputs being the most proactive and neat: [...] If the initial deviation is glideslope only, then you need a simultaenous thrust change, attitude change, and trim for the TPC.

We’re on the same page then. I agree with you a hundred percent that [...] a low and on-speed is fixed with simultaneous pitch and thrust;


Yes, if you’’re doing everything right.
This “yes” is to you:
you raise the nose so you don't hit the ground. You simultaneously put the thrust up (if you need to) to to stop the speed decreasing.

I agree with Vilas; AOA has nothing to do with this discussion which is now going off on a AOA tangent.

Well the discussion is about airspeed, and since AOA determines airspeed, it has everything to do with it.

Again, you’re conflating A. whether an elevator input is made with the intention of an airspeed control, with B. whether it physically constitutes an airspeed control. They're two separate questions, and the answer to A is “probably not” while the answer to B is “yes.” And while most elevator inputs will continue to be made with the intention of controlling path and not airspeed (A), my aim is to make pilots aware that B is simultaneously occurring in the background.

B is part of the bare essentials of the reality of flight. And when stress, surprise, and task saturation cut away everything extraneous in the Asiana pilot’s mind, the bare essentials he was left with in his mind, were a mismatch to the bare essentials of the reality of flight, resulting in a fatal crash. I want to add B to the bare essentials in pilots’ minds so this isn’t repeated.

[Capn Bloggs]

My head hurts. How about we just monitor the @#$%^ speed?? :rolleyes:

[vilas]

B is part of the bare essentials of the reality of flight. And when stress, surprise, and task saturation cut away everything extraneous in the Asiana pilot’s mind, the bare essentials he was left with in his mind, were a mismatch to the bare essentials of the reality of flight, resulting in a fatal crash. I want to add B to the bare essentials in pilots’ minds so this isn’t repeated. I will say again leave Asiana alone. All they needed to do was scan their speed regularly which is a must in every approach, every type of aircraft, with ATHR or manual thrust. If there was a call SPEED when the the speed was Vapp-5 they would have added thrust and landed properly and eventually retired happily without knowing anything about AoA. It was poor knowledge of AFS and Skill. If you think your AoA theory would have saved them sorry I can't agree. We do things simultaneously. Let's not count nano seconds.

[Vessbot]

My head hurts. How about we just monitor the @#$%^ speed??

​​​​​
I'm not satisfied with that. Monitoring something as if the factors that go into it are a surprise, is not enough when we have an understanding of their cause (especially when that cause stems in our hands). The AOA-airspeed relationship may not be 2+2, but it's not particle physics either. (Actually, it's y=1/x)

I will say again leave Asiana alone. All they needed to do was scan their speed regularly which is a must in every approach, every type of aircraft, with ATHR or manual thrust. If there was a call SPEED when the the speed was Vapp-5 they would have added thrust and landed properly and eventually retired happily without knowing anything about AoA. It was poor knowledge of AFS and Skill. If you think your AoA theory would have saved them sorry I can't agree. We do things simultaneously. Let's not count nano seconds.
​​​​
Why do you say to leave Asiana alone? Is there some difference between it and a more moderate airspeed loss that makes it inapplicable?

No doubt had someone seen the airspeed indicator they would have noticed the problem. I agree. But I disagree that that's "all they needed to do." They also needed, at a more basic level, to not cause that airspeed loss to begin with... Which entails an understanding, that they lacked, that aft stick force does that.

And I'm trying to understand why you're resisting that. Back to my crashing a car straight into a wall analogy, you're saying that "all they needed to do" was to watch more carefully how close they're getting to the wall, and make a callout; and they'd eventually retire happy without knowing about how mashing the gas pedal accelerates them toward the wall. And that such knowledge would not have saved them. Why? On the contrary, that's exactly what would have saved them, at a stage prior to monitoring. Monitoring is supposed to add an extra layer of safety, not serve as a cover to remove the basic mechanism underneath.

You called it "poor ... skill," but that doesn't explain anything, it just puts a label on what's poor. And I'm explaining what actually constitutes that poor skill, (so that people can actually do something about it instead of merely tsk-tsk'ing over its poorness) but you're saying no.

You scoff at counting nanoseconds like I'm counting the angels on the head of a pin... no, I'm describing the extreme basics, or at least what should be.