Could one control a plane only using CG changes?

[mardlamock]

From the perspective of someone fairly far into studying an Aerospace Engineering degree, Fel's understanding is incorrect - although appeal to authority would be a very bad idea here, as their misunderstanding is the same as what I normally see coming from pilots. I've seen many arguing on the internet, that think speed is somehow directly related to current lift, and that too little is an automatic stall - and to preserve that misunderstanding many seem to think stall in turn is somehow a fundamentally different phenomenon to level flight.

But, this isn't particularly related to the original question. The original question can be summed up as: if the controls are left in free-floating trim or fixed position (which would be unusual for an Airbus due to its automatic controls), then it's very easy to control an aircraft using CG changes. There are many small aircraft designed to be flown this way, and many pilots of light aircraft have attested that in free-floating trim they can control their plane just by shifting in their own seat. However, if there's anything that would actively fight you, then aerodynamic control will always beat you - the most you can do is change the plane's stability margin, but if that remains positive, that only changes how twitchy the plane is to control, not whether it can be. And if the stability margin becomes negative, then there's only two possibilities - pilot remains in control with unstable aircraft, or pilot loses control and the plane tumbles out of the sky. Neither of which would save anyone, so your time would be better spent trying to talk the pilot out of it through the door.

Not necessarily, couldnt it be thatthe plane was long enough and the passengers fat enough so as to completly shift the CG such that it is always behind the CL, no matter what the pilot did? And also, couldnt they have all waited until the last moment and then pitched up with the slope of the mountain, reducing the airspeed and hitting the ground at lower speed? Sort of like what this guy does

[WestAir]

Mardlamock,

There is a drastic difference between 20 knots (what the guy in your video is doing) and 300+ knots (what that airbus was doing). The Asiana 777 in San Fransisco hit at 100 knots with gear deployed on a level seawall and it was shattered.

To the OP, nothing the PAX could have done, short of breaking down the cockpit door, would have saved that airbus. Period.

[mardlamock]

Mardlamock,

There is a drastic difference between 20 knots (what the guy in your video is doing) and 300+ knots (what that airbus was doing). The Asiana 777 in San Fransisco hit at 100 knots with gear deployed on a level seawall and it was shattered.

To the OP, nothing the PAX could have done, short of breaking down the cockpit door, would have saved that airbus. Period.

I know, what I am talking about is starting a very steep climb going along the mountain slope just as the plane is about to hit it. If they were going at that speed and dropped to say 140 knots they would have climbed like 300 meters. The pilot could roll the plane for sure, but lets assume he doesnt because he is just waiting for it to crash. What do you think?

[Nibb31]

No because the FMS would compensate the CG to maintain the orders it was given. Nobody was holding the stick. These planes are flown by computers, not by the stick.

[mardlamock]

No because the FMS would compensate the CG to maintain the orders it was given. Nobody was holding the stick. These planes are flown by computers, not by the stick.

The flight computer cant change where the CG is, and if it is too far out then it doesnt have much control authority.. The real question is wether or not you can change the location of the CG such that the control surfaces cant control pitch.

[WestAir]

Unlikely.

Often we do move PAX around once we do weight and balance (especially with a lot of no shows) to keep us within CoG limits, but if I were a betting man (and I live in Las Vegas, so I am), I'd bet that the hor. stab can be trimmed to handle any CoG changes brought on by PAX movement. You might not even need to move the elevator at all.

[mardlamock]

Unlikely.

Often we do move PAX around once we do weight and balance (especially with a lot of no shows) to keep us within CoG limits, but if I were a betting man (and I live in Las Vegas, so I am), I'd bet that the hor. stab can be trimmed to handle any CoG changes brought on by PAX movement. You might not even need to move the elevator at all.

Do you know of any good sites to get the technical info? Such as usual position of the CoG, with PAX, where you start losing authority, etc.

[Iskierka]

The problem is that moving CG out of rear limits doesn't create a problem whereby the plane will always pitch up no matter what the pilot does - the elevators are very, very powerful and can always remain in control. What happens when CG is out of limits in normal flight is that the aircraft becomes very difficult to control, as you can't just hold it in a position and have it go forwards. If you're pointing a tiny bit too high, it'll go up as you imagine, but if you're pointing a tiny bit too low, it'll pitch down just as violently.

This is because an aircraft is a dynamic vehicle trying to keep control by deflecting air, not simply balancing like a see-saw as you might imagine. There are some aspects of stability that are see-saw-esque, but they're more around when you have a stable aircraft and are watching its behaviour - moving the pivot has very, very different effects, which are far more reminiscent of trying to throw a dart forwards for a stable aircraft, or backwards for an unstable aircraft. Wikipedia has an article if you're interested, though there isn't a simple-English equivalent. The better KSP/FAR plane design tutorials should also explain how it works.

The end result of it though is that if the plane is being actively controlled, by a pilot or autopilot, then in theory you can put the CG extremely far back, considerably beyond aft limits - and so long as the controller is quick and precise enough in its response to movement, the plane can be kept in control, and so if you were trying to crash it, you'd be able to succeed. You can control an aircraft using CG if that's the only input, but aerodynamic input overrides CG. CG movement just causes some annoyances for aerodynamic control, which in their extreme can become lethal, when it becomes impossible for either to control the aircraft, as CG is too far back and controls don't respond fast enough.

[WestAir]

Do you know of any good sites to get the technical info? Such as usual position of the CoG, with PAX, where you start losing authority, etc.

I don't believe so.

[mardlamock]

The problem is that moving CG out of rear limits doesn't create a problem whereby the plane will always pitch up no matter what the pilot does - the elevators are very, very powerful and can always remain in control. What happens when CG is out of limits in normal flight is that the aircraft becomes very difficult to control, as you can't just hold it in a position and have it go forwards. If you're pointing a tiny bit too high, it'll go up as you imagine, but if you're pointing a tiny bit too low, it'll pitch down just as violently.

This is because an aircraft is a dynamic vehicle trying to keep control by deflecting air, not simply balancing like a see-saw as you might imagine. There are some aspects of stability that are see-saw-esque, but they're more around when you have a stable aircraft and are watching its behaviour - moving the pivot has very, very different effects, which are far more reminiscent of trying to throw a dart forwards for a stable aircraft, or backwards for an unstable aircraft. Wikipedia has an article if you're interested, though there isn't a simple-English equivalent. The better KSP/FAR plane design tutorials should also explain how it works.

The end result of it though is that if the plane is being actively controlled, by a pilot or autopilot, then in theory you can put the CG extremely far back, considerably beyond aft limits - and so long as the controller is quick and precise enough in its response to movement, the plane can be kept in control, and so if you were trying to crash it, you'd be able to succeed. You can control an aircraft using CG if that's the only input, but aerodynamic input overrides CG. CG movement just causes some annoyances for aerodynamic control, which in their extreme can become lethal, when it becomes impossible for either to control the aircraft, as CG is too far back and controls don't respond fast enough.

I know my static stability, what you are saying is right though, if the AoA was negative then it would just pitch down very violently, but that would only be the case if the plane was accelerating downwards. The plane was IIRC descending at a constant speed, meaning that it had a positive AoA at the time, therefore making it possible to pitch up only using CG changes. The control surfaces wouldnt be generating enough of a moment to counteract that generated by the wings, so long as the CG was behind a certain location, which is what i am trying to figure out. Ill draw a picture to better illustrate what i am saying

[Iskierka]

The trouble is that the control surfaces were under automatic control, not just set in place to hold the descent - if you moved the cg back, the flight computer will sense the changing loads and slight movements, and automatically compensate by holding the nose down. Like I said - this is possible if the controls are in fixed position, but in the suggested case, they are not, as they are either under automatic computer control or pilot's control, both of which counteract anything passengers could do.

[mardlamock]

The trouble is that the control surfaces were under automatic control, not just set in place to hold the descent - if you moved the cg back, the flight computer will sense the changing loads and slight movements, and automatically compensate by holding the nose down. Like I said - this is possible if the controls are in fixed position, but in the suggested case, they are not, as they are either under automatic computer control or pilot's control, both of which counteract anything passengers could do.

You dont seem to be getting it, the control surfaces can only go so far, if the CG can change significantly what the control surfaces do wont really matter, what will matter will be the attitude the plane was in before the change. Give me a minute to draw it

[Iskierka]

No, I fully understand it - but you wouldn't have the plane "controlled" by CG because of this, and you don't realise that the control surfaces are very, VERY powerful if they are attempting to control the aircraft. CG change will not overpower them to keep the nose up, only potentially cause the aircraft to completely stall out and crash if instability movements become faster than the controls can move.

[mardlamock]

No, I fully understand it - but you wouldn't have the plane "controlled" by CG because of this, and you don't realise that the control surfaces are very, VERY powerful if they are attempting to control the aircraft. CG change will not overpower them to keep the nose up, only potentially cause the aircraft to completely stall out and crash if instability movements become faster than the controls can move.

"very powerful" Is not useful data when seeing wether it could be possible or not. The control surfaces can only provide so much lift at a given airspeed, therefore can only keep an aircraft level within a certain range of locations for the CG. If the movements are faster than the autopilot's reactions, then the aircraft will pitch up as I said before. And you have also got to take into account that if its an airbus, the control surface's AoA will never exceed what is allowed by the structure, making it even harder for the autopilot to remain under control of the aircraft.

[Iskierka]

I could demonstrate how powerful they are, but let's fall back to this - at the farthest back allowable CG position, they'll still be pushing down, but very gently. A few metres forward of this you'll find the forward CG limit, at which they'll be pushing down quite hard (though still with plenty of room to manoeuvre). At the point where the elevators are at zero load, the aircraft will be at zero stability, already behind its aft CG limit. Given that control is roughly symmetrical, they can hold the nose down even if the CG moves several metres back past this point, without using its full control range. The trouble here is, regardless of if at this point the elevator has reached full loading, or it can't keep up with the unstable movement of the aircraft - either case will result in a loss of control of the aircraft. As in, total loss of control - the far back CG might (might, as it depends on conditions of control loss) cause pitch-up, but it will cause extremely violent pitch-up well into deep stall, and likely an unrecoverable descent into a very hard crash. Once the aircraft is stalled, the only hope to recover would be move fully forward again and have the control surfaces recover the aircraft - which then puts them back in control, and so whatever malicious action was being attempted can be continued, and CG movement has done nothing to prevent it.

The question posed is "can an aircraft be controlled by changes in CG?" The answer is, if the controls are doing nothing, yes. If you're trying to fight the controls, no - you can cause the controls to lose control, but you can't take control for yourself with CG, only remove it from both of you. Which is very much not the desired outcome.