Overspeed Question

[3WE]

Several weeks ago I had a former A320 pilot on the jumpseat. He had been furloughed by a major US carrier and flown Airbuses in a third world country. He talked about the skill of the pilots and, let's just say he did NOT put them in the same category as Chuck Yeager and Bob Hoover. That's when I realized why the Airbus system is designed the way it is: to keep the weak pilots out of trouble... Not necessarily a bad thing when you get right down to it.

Yeah, but as you said earlier- let the computer monitor EVERYTHING and warn and tell you what to do, and have the bestest autopilot with moving maps and synthetic vision, and maybe even keep score and add arcade noises woka woka, but at least make the airplane act like an airplane.

Pull back on stick till you see the vertical speed you want...relax the stick... (to hell with that speed-attitude-vertical speed equilibrium-seeking/overshooting behavior whatever that weird term is that Gabriel talks about ad-nauseum that paper airplanes do when you throw them too hard)....

Airbus logic is so wrong. HAL might take a break...and then you have an airplane...something that takes upwards of over 40 hours of training to learn how to fly and do stuff like hold speed, climb, descend, turn or even go straight and level...

[snydersnapshots]

I am not fond of the Airbus philosophy either, but it has its strong points. In an evasive maneuver as you propose, the best you can do in an Airbus is (as long as normal law is in effect) to fully deflect the flightstick to one side, fully pull up and advance the throttles to the stops. Whatever the initial conditions, the computers will ensure the max performance and, at the same time, that you don't exceed the G-limits, that you don't stall, that you don't roll over and that you don't blow your engines. Maybe, with a very good stick-and-rudder pilot in command, there will be a time where that extra 10° of bank or flying past the stickshaker but below the stall will save the day. But if you look at the accident records, it looks that it will save much more accidents by limiting the pilot authority than it will cause because of the same thing.

I agree that my scenario is pretty far-fetched and the successful outcome might be questionable. Accident reports do reflect that flying "outside the envelope" more often than not results in unwanted paperwork for the FAA and NTSB...

Airbus system is designed to diminish the workload and take some concerns during maneuvering out of the pilot's mind (like overstressing or stalling). However, and this is where I don't like it (and why I like Boeing approach more), the airplane response to the pilot's inputs are different than those of traditional planes, where the pilots learn to fly and make their first lots of hundreds or maybe few thousands of hours (before stepping into an Airbus). And, the pilos might get used to that: Plane's just can't exceed their envelope. When there is a problem like a high altitude unreliable airspeed event, maybe just firewalling and pulling up is not such a bad idea. The computer will take care of the rest (except that, with unreliable airspeed, the computers say "I don't like it, now it's your time to fly and ensure that the plane remains in its envelope" (i.e. they quit from normal law). Maybe that had something to do with the Air France accident, where what the pilot did since the beginning of the event was so extremely counterintuitive, counter-basic flight knowledge, counter-basic pilot training, counter-basic stick-and-rudder skills (and I mean VERY basic, pre solo as you say), and yes, against the airlplane's procedures too.

Several weeks ago I had a former A320 pilot on the jumpseat. He had been furloughed by a major US carrier and flown Airbuses in a third world country. He talked about the skill of the pilots and, let's just say he did NOT put them in the same category as Chuck Yeager and Bob Hoover. That's when I realized why the Airbus system is designed the way it is: to keep the weak pilots out of trouble... Not necessarily a bad thing when you get right down to it.

Another thing that I don't like about Airbus is the lack of control feedback. The sticks don't move when the autopilot does some input. The other pilot's stick doesn't move when the flying pilot makes an input, the throttles don't move when the autothrust makes an input.

In the AirFrance accident, when they were already below 10,000 ft (and descending more than 10,000 fpm), the pilot that was not flying says "try to pull up". The captain, that was standing behind the pilots, said "we don't have any choice now, we are too low to tray anything else", and the pilot that had been flying said "But I have been pulling up the whole time!" I can not imagine the "Duh!" faces of the other two guys when they learned that seconds from impact (a thing that would have hardly remained unnoticed had they pilots had two big yokes that move together).

I emphatically agree with your opinion on control feedback. If Air France had been a Boeing, it would have been obvious to the pilot trying to make the correct input that the other guy was holding the yoke back. Perhaps that would have made the difference. Unfortunately we'll never know.

[Gabriel]

I'll readily admit to not much familiarity with Airbus systems beyond things I have read and pilots I've talked to. In all honesty, I haven't read the systems manual so I'm not an expert on the various control laws. That being said, what happens if I need 80 degrees of bank?

You either switch all the flight computers off (which takes flipping a few switches) or accept whatever limit they impose to you.

I am not fond of the Airbus philosophy either, but it has its strong points. In an evasive maneuver as you propose, the best you can do in an Airbus is (as long as normal law is in effect) to fully deflect the flightstick to one side, fully pull up and advance the throttles to the stops. Whatever the initial conditions, the computers will ensure the max performance and, at the same time, that you don't exceed the G-limits, that you don't stall, that you don't roll over and that you don't blow your engines. Maybe, with a very good stick-and-rudder pilot in command, there will be a time where that extra 10° of bank or flying past the stickshaker but below the stall will save the day. But if you look at the accident records, it looks that it will save much more accidents by limiting the pilot authority than it will cause because of the same thing.

Airbus system is designed to diminish the workload and take some concerns during maneuvering out of the pilot's mind (like overstressing or stalling). However, and this is where I don't like it (and why I like Boeing approach more), the airplane response to the pilot's inputs are different than those of traditional planes, where the pilots learn to fly and make their first lots of hundreds or maybe few thousands of hours (before stepping into an Airbus). And, the pilos might get used to that: Plane's just can't exceed their envelope. When there is a problem like a high altitude unreliable airspeed event, maybe just firewalling and pulling up is not such a bad idea. The computer will take care of the rest (except that, with unreliable airspeed, the computers say "I don't like it, now it's your time to fly and ensure that the plane remains in its envelope" (i.e. they quit from normal law). Maybe that had something to do with the Air France accident, where what the pilot did since the beginning of the event was so extremely counterintuitive, counter-basic flight knowledge, counter-basic pilot training, counter-basic stick-and-rudder skills (and I mean VERY basic, pre solo as you say), and yes, against the airlplane's procedures too.

Another thing that I don't like about Airbus is the lack of control feedback. The sticks don't move when the autopilot does some input. The other pilot's stick doesn't move when the flying pilot makes an input, the throttles don't move when the autothrust makes an input.

In the AirFrance accident, when they were already below 10,000 ft (and descending more than 10,000 fpm), the pilot that was not flying says "try to pull up". The captain, that was standing behind the pilots, said "we don't have any choice now, we are too low to tray anything else", and the pilot that had been flying said "But I have been pulling up the whole time!" I can not imagine the "Duh!" faces of the other two guys when they learned that seconds from impact (a thing that would have hardly remained unnoticed had they pilots had two big yokes that move together).

[snydersnapshots]

I'll submit one last off-topic comment here in reply.

I'm going off topic but I wanted to set the record straight.

It's not true that I don't think that the rule "throttle = altitude, pitch = speed" is not important. In fact, I do think that it is important, and in fact too important to leave it at just that. If anything, it is an over-simplification. Speed and altitude control requires coordinated use of both elevator and throttle in a way that is more complicated than that rule, because they vary in a dynamic way with throttle and elevator inputs, in a motion known as phugoid.

If you add throttle and do nothing else, the first thing you'll see is a speed increase. That speed increase will create an increase in lift and the plane will start to accelerate up. That upwards movement will reduce the lift of the tail (or increase its downwards lift) causing a nose-up pitch. The upwards movement and pitch-up will overshoot the equilibrium, so the airplane will start to loose speed down to a speed that is even lower than the initial one, and the whole process will reverse, and everything starts again, and so over and over in an oscillatory motion that is lightly damped. Eventually, the plane will stabilize in a climb at the same initial speed (disregarding any effect of the thrust in the pitching moment).

Pilots don't even realize, but when they start a climb what they do is add power, pull back until the plane reaches the needed climb attitude, and then release the pull-back returning to the original elevator position. If properly done, the final effect will be the same as before, only that without the oscillation in the middle. In this way, the elevator has two main functions in the speed and altitude management: Establish the trim speed and damp out (or prevent) the phugoid.

Once all that is understood, I have no problem summarizing it with "power controls altitude, pitch controls speed".

I agree...

Now, I did invent an alternative to this rule (or at least I re-invented it independently because I've never heard or seen it anywhere except by myself): Throttle controls power available, elevator controls power required. Like it?

(so, 3WE, I did start a discussion on the matter after all)

Hmmm....going to have to ponder this one after a little sleep (it's been a long day), but this one sounds like it has merit.

Well, in fact that's very much what Airbus does. The plane will let you bank past 60° (I don't remember the exact value, other than it was sixty-something). Beyond I think it's 35° it will alert you by applying artificial positive lateral stability, which means that the plane will return to 35° if you let go on the stick, but it won't prevent you from banking up to the sixty-something limit.


... "What is it doing now?"

I'll readily admit to not much familiarity with Airbus systems beyond things I have read and pilots I've talked to. In all honesty, I haven't read the systems manual so I'm not an expert on the various control laws. That being said, what happens if I need 80 degrees of bank?

[snydersnapshots]

I'll submit one last off-topic comment here in reply.

I'm going off topic but I wanted to set the record straight.

It's not true that I don't think that the rule "throttle = altitude, pitch = speed" is not important. In fact, I do think that it is important, and in fact too important to leave it at just that. If anything, it is an over-simplification. Speed and altitude control requires coordinated use of both elevator and throttle in a way that is more complicated than that rule, because they vary in a dynamic way with throttle and elevator inputs, in a motion known as phugoid.

If you add throttle and do nothing else, the first thing you'll see is a speed increase. That speed increase will create an increase in lift and the plane will start to accelerate up. That upwards movement will reduce the lift of the tail (or increase its downwards lift) causing a nose-up pitch. The upwards movement and pitch-up will overshoot the equilibrium, so the airplane will start to loose speed down to a speed that is even lower than the initial one, and the whole process will reverse, and everything starts again, and so over and over in an oscillatory motion that is lightly damped. Eventually, the plane will stabilize in a climb at the same initial speed (disregarding any effect of the thrust in the pitching moment).

Pilots don't even realize, but when they start a climb what they do is add power, pull back until the plane reaches the needed climb attitude, and then release the pull-back returning to the original elevator position. If properly done, the final effect will be the same as before, only that without the oscillation in the middle. In this way, the elevator has two main functions in the speed and altitude management: Establish the trim speed and damp out (or prevent) the phugoid.

Once all that is understood, I have no problem summarizing it with "power controls altitude, pitch controls speed".

I agree...

Now, I did invent an alternative to this rule (or at least I re-invented it independently because I've never heard or seen it anywhere except by myself): Throttle controls power available, elevator controls power required. Like it?

(so, 3WE, I did start a discussion on the matter after all)

Hmmm....going to have to ponder this one after a little sleep (it's been a long day), but this one sounds like it has merit.

Well, in fact that's very much what Airbus does. The plane will let you bank past 60° (I don't remember the exact value, other than it was sixty-something). Beyond I think it's 35° it will alert you by applying artificial positive lateral stability, which means that the plane will return to 35° if you let go on the stick, but it won't prevent you from banking up to the sixty-something limit.


... "What is it doing now?"

I'll readily admit to not much familiarity with Airbus systems beyond things I have read and pilots I've talked to. In all honesty, I haven't read the systems manual so I'm not an expert on the various control laws. That being said, what happens if I need 80 degrees of bank?

[Gabriel]

I'm going off topic but I wanted to set the record straight.

It's not true that I don't think that the rule "throttle = altitude, pitch = speed" is not important. In fact, I do think that it is important, and in fact too important to leave it at just that. If anything, it is an over-simplification. Speed and altitude control requires coordinated use of both elevator and throttle in a way that is more complicated than that rule, because they vary in a dynamic way with throttle and elevator inputs, in a motion known as phugoid.

If you add throttle and do nothing else, the first thing you'll see is a speed increase. That speed increase will create an increase in lift and the plane will start to accelerate up. That upwards movement will reduce the lift of the tail (or increase its downwards lift) causing a nose-up pitch. The upwards movement and pitch-up will overshoot the equilibrium, so the airplane will start to loose speed down to a speed that is even lower than the initial one, and the whole process will reverse, and everything starts again, and so over and over in an oscillatory motion that is lightly damped. Eventually, the plane will stabilize in a climb at the same initial speed (disregarding any effect of the thrust in the pitching moment).

Pilots don't even realize, but when they start a climb what they do is add power, pull back until the plane reaches the needed climb attitude, and then release the pull-back returning to the original elevator position. If properly done, the final effect will be the same as before, only that without the oscillation in the middle. In this way, the elevator has two main functions in the speed and altitude management: Establish the trim speed and damp out (or prevent) the phugoid.

Once all that is understood, I have no problem summarizing it with "power controls altitude, pitch controls speed". Now, I did invent an alternative to this rule (or at least I re-invented it independently because I've never heard or seen it anywhere except by myself): Throttle controls power available, elevator controls power required. Like it?

(so, 3WE, I did start a discussion on the matter after all)

Don't limit my bank angle to 25 degrees because YOU, Mr. Computer, don't think it's appropriate to go over 25 degrees. If I NEED to bank more than 25 degrees, then by God LET ME. Warn me, tell me it might not be smart, but let me do it!

Well, in fact that's very much what Airbus does. The plane will let you bank past 60° (I don't remember the exact value, other than it was sixty-something). Beyond I think it's 35° it will alert you by applying artificial positive lateral stability, which means that the plane will return to 35° if you let go on the stick, but it won't prevent you from banking up to the sixty-something limit.

If I need 60 degrees of bank to avoid traffic and you only give me 25, then the last words you hear on my CVR are going to be...

... "What is it doing now?"

[snydersnapshots]

...well, I have read Stick and Rudder, and would be glad to sign your copy of it...that is if you would like a parlour-talking, know-nothing ass-hat signature.

From one parlour-talking ass-hat to another, I'd be honored to have you sign my book!

While I did not write that book or rule, I would bet that someone other than Langeweische came up with it.

I suspect you're probably right.

So, do you consider pitch-airspeed, power-altitude to be a very important rule of thumb and that is fundamental to good airmanship?...

I do. Do I think about it on a daily basis when flying? No. But I think it's extremely important to learn the concept early in training. It can be demonstrated as simply as trimming the airplane for a specific airspeed in straight and level flight and adding or decreasing power and/or increasing or decreasing pitch attitude. Once the student gets this concept down, other maneuvers will be much more easily grasped, in my opinion.

This concept is especially helpful when it comes to working on precision approaches. Trim the airplane for the approach speed and decrease the power. The nose will lower but the speed will stay the same (i.e. you're controlling altitude with power). If you get below the glide slope, add a bit of power. Above? Decrease power. If the airplane is trimmed properly and the power changes are done smoothly, the airspeed should remain the same.

It's also a handy concept to know in the event of an unreliable airspeed indication and the concept is the basis of the "flight with unreliable airspeed" charts. The chart gives you a pitch and power setting and cross references that with what airspeed it will maintain in level flight. Kind of handy to have when the feces hits the air-movement device on a dark and stormy night...

...or is it of little value because it has fallacies and the Airbus control system takes care of all of that.

I put the Airbus fly-by-wire control system in the same category as the city of San Francisco telling me I can't have a toy with my Happy Meal or Mayor Bloomberg in New York limiting my soft-drink size to 16 ounces or less: Let ME make the decisions that affect ME! Don't limit my bank angle to 25 degrees because YOU, Mr. Computer, don't think it's appropriate to go over 25 degrees. If I NEED to bank more than 25 degrees, then by God LET ME. Warn me, tell me it might not be smart, but let me do it! If I need 60 degrees of bank to avoid traffic and you only give me 25, then the last words you hear on my CVR are going to be very impolite comments about the French for designing it, Airbus for building it, the FAA for certifying it, and my airline for buying it...

Sorry...ya struck a cord with me there... I prefer the Boeing FBW philosophy, as you can see.

I call it "my rule" because Gabriel doesn't care about it, he only cares about AOA management, while I think that speed and attitude management are more imporant.

I haven't read a lot of Gabriel's posts, but I will agree that AOA management is important as well. The airplane doesn't know or care about where the earth is in reference to it. An airplane will stall in any attitude or airspeed depending on the load factor. (Of course, you reach a point where you'll break the airplane before you get the wings to stall, but that's not wise and will usually result in the owner not letting you rent the airplane again). It's just as important to realize the concept that an airplane can stall with the nose below the horizon (as in pulling too hard when performing a stall recovery) as it is to understand Mr. Langewiesche's theory on pitch/power/airspeed/altitude.

So, in short, both theories have their merits and are important to grasp, but rigid adherence to either is not necessarily the right thing to do. How is THAT for a wishy-washy, political answer?

[3WE]

Seems to me I saw "your" rule in Wolfgang Langeweische's book STICK AND RUDDER a few years back. OK, it was when I had hair, so it was a LOT of years back. Of course, if you're actually Wolfgang Langewiesche, would you autograph my book???

...well, I have read Stick and Rudder, and would be glad to sign your copy of it...that is if you would like a parlour-talking, know-nothing ass-hat signature.

While I did not write that book or rule, I would bet that someone other than Langeweische came up with it.

So, do you consider pitch-airspeed, power-altitude to be a very important rule of thumb and that is fundamental to good airmanship?...

...or is it of little value because it has fallacies and the Airbus control system takes care of all of that.

I call it "my rule" because Gabriel doesn't care about it, he only cares about AOA management, while I think that speed and attitude management are more imporant.

[snydersnapshots]

Oh, you'll always start a discussion on stalls (your favorite thing) but you use my rule without a courtesy PM?

Seems to me I saw "your" rule in Wolfgang Langeweische's book STICK AND RUDDER a few years back. OK, it was when I had hair, so it was a LOT of years back. Of course, if you're actually Wolfgang Langewiesche, would you autograph my book???

[3WE]

...there is a saying in aviation that goes "stick controls speed, throttle controls altitude" (no, I don't want to start a discussion on the truth of this saying)...

Oh, you'll always start a discussion on stalls (your favorite thing) but you use my rule without a courtesy PM?



There are those who say it's not really true because there are several flight modes where it does not work well.

[TeeVee]

ok, i'm beginning to see. thanks again for the time!

[Gabriel]

thanks for taking the time Gabriel.

what i still can't comprehend is this: in a bus, if the pilot commands a bank that HAL thinks is too steep, HAL simply ignores the continued sidestick pressure at a point which HAL determines the bank angle to be within operational limits. in other words, HAL countermands the pilots inputs directly. however, when it comes to level flight overspeed due to excessive throttle, HAL says, climb, to reduce the effect of the throttle "input."

of course the sum total my knowledge of aircraft operation and aerodynamics can be written on a 2" x 2" piece of paper. but is still think that HAL would be better off retarding the throttles as opposed to climbing.

I think you are making a too close link between "speed" and "throttle".

With no data to support this, I'll bet 1 cent that, when related with "pilot technique", there are much more cases of overspeed due to wrong pitch than wrong throttle. Examples: When descending from cruise, even with the speed at idle the plane cannot descend with a great rate without increasing speed. Exceed that rate and you will depart from your target speed and risk an overspeed. During climb, don't point the nose high enough and you'll overspeed (especially before the airplane is clean).

Not for nothing there is a saying in aviation that goes "stick controls speed, throttle controls altitude" (no, I don't want to start a discussion on the truth of this saying)

An other thing to consider is Airbus philosophy toward throttle management (which is nearly "don't touch this lever"). Very shortly after take-off you reduce the throttle levers from the "take-off" setting to the "climb" setting and let it there for the remainder of the flight until you are 20ft above the runway, when HAL must remind you of your mental condition for you to pull the levers back to idle ("Retard, retard!"). The "climb" setting is nearly the "one-for-all" setting: Climb, cruise, descent, approach. Obviously the engines themselves don't stay at climb all the time. Just the thrust levers do and HAL takes care of the engines. If you move them to TOGA (or any setting past climb) you are effectively telling the plane "I have the throttles", so the plane has no choice but to think that you have compelling reasons to do whatever you are doing with them. You are already overriding HAL, at least the "thrust" aspect of it. If HAL still wants to prevent the overspeed, then it must work with whatever it has left.

Also, maybe you are overjudging the "power" of the Airbus version of HAL.
It will still let pilot screw it up big time. Especially with the thrust levers.
Example: Fail to completely idle one of the thrust levers during the landing (it should happen at 20ft). The spoilers won't deploy upon landing, the autobrake won't engage, the autothrust won't disengage, and when you start to brake manually while asking "what is it doing now", the still engaged autothrust will increase thrust to try to hold the speed selected for the approach.

With the stick HAL is a bit harder to beat. It takes a few clicks or some failures (like Air France).

And finally, again, I see no compelling reason to declare one of the two choices the "best" one.

[snydersnapshots]

thanks for taking the time Gabriel.

what i still can't comprehend is this: in a bus, if the pilot commands a bank that HAL thinks is too steep, HAL simply ignores the continued sidestick pressure at a point which HAL determines the bank angel to be within operational limits. in other words, HAL countermands the pilots inputs directly.

Which is why, in my opinion, "If it's not Boeing, I'm not going." If, for some reason, I need to roll into a 60 degree bank (traffic avoidance perhaps?), I don't want the damn computer telling me I can't. Sure--squawk "Bank Angle, Bank Angle" all you want (my Boeing already does), but let me fly the airplane and do what I need to do.

OK...off my soapbox now...

[TeeVee]

thanks for taking the time Gabriel.

what i still can't comprehend is this: in a bus, if the pilot commands a bank that HAL thinks is too steep, HAL simply ignores the continued sidestick pressure at a point which HAL determines the bank angle to be within operational limits. in other words, HAL countermands the pilots inputs directly. however, when it comes to level flight overspeed due to excessive throttle, HAL says, climb, to reduce the effect of the throttle "input."

of course the sum total my knowledge of aircraft operation and aerodynamics can be written on a 2" x 2" piece of paper. but is still think that HAL would be better off retarding the throttles as opposed to climbing.

[Gabriel]

yes, i understand that autothrottle would normally prevent overspeed. but if for some stupid reason the pilot disconnected AT operation, selected max thrust in level flight, i still don't understand why airbus logic dictates "climb" to correct overspeed. theoretically, if the plane were lightly loaded, climbing alone may not solve the problem until it passed its max operating altitude.

yes i know my scenario is a weird one. i'm just trying to understand why airbus chose climb over retard. it's not as if the system can't and doesn't completely ignore other "stupid" pilot inputs...

First, rest assured that the plane won't climb past it's climbing ability. It seems tautological, but not so because there is a way to climb past the ceiling, like Air France did.

The thing is that to do it, you'll have to trade speed for altitude, and that's not what this protection does. What it will do is, upon reaching Vne/Mme establish whatever climb it takes to convert all the excess power to altitude (excess power is the power that the engines are producing minus the power needed to keep the plane flying straight and level, excess power WILL make the plane increase it's mechanical energy, be it in the form or kinetic energy -speed- or potential gravitational energy -altitude-, or a combination of both. You use the stick to govern where that energy goes). So it will keep keep climbing at Vne/Mme as the engines loose power (due to altitude) until it reaches the point where all the power the engines are producing is exactly what is needed to keep the plane flying straight and level at that speed.

As why doesn't the plane just say "screw the pilot, even if he disconnected the autopilot and wants full thrust I won't allow it", well, it's a sort of philosophical question. Who has the last word? The pilot or the computer? Do we allow the pilot to disconnect HALL? He already overrode the autothrottle but not the flight control computers which can prevent the overspeed in this way. If the pilots go further to override the FCCs (what he can do), then nothing at all will prevent the pilot from overspeeding (we can almost say that you are not in an Airbus any longer).

Other than letting the pilot judge that the computer is crazy (for example, that the overspeed is not such), I can think of only one scenario where it would be good do it as it is now, that I concede is too crazy: the plane approaching high terrain and the pilot applying TOGA but not immediately climbing for whatever reason. Let the plane accumulate energy in whatever form that the pilot can then use to trade for latitude. When the "speed" energy tank is full, start filling the "altitude" energy tank but please don't stop accumulating energy.

On the other hand, I don't see anything wrong with letting the plane climb except the "little" problem of traffic separation.