Air France 447 - On topic only!

[Black Ram]

I think what made Sully's job a lot easier was the FBW system. The final approach towards the water was done at a slower than optimal speed, I think it could have been close to 20kias lower. Sully had already turned the APU on to enable the flight computers to work. They allowed for a perfect line-up and perfect control at a less than perfect speed (maybe close to the edge of the performacne envelope?), which is crucial for any water landing.

[Gabriel]

I know that, in Normal and Alternate Law, the g-load commanded is proportionate to stick deflection, so reducing the deflection would reduce the g-load request. I also know that the Airbus has neutral static stability in Normal and Alternate Law, so upon releasing the stick to neutral, the auto pitch trim will maintain FPA until a downward pitch command (-g) is introduced. That is how I arrive at my understanding of sidestick-to-pitch behavior. But correct me if I'm mistaken...

Well, I'd say that at least I don't like your wording:

...there may be a bug in the system, or this might just be the auto trim attempting to trim the downforce out of a very steep FPV as commanded by the sidestick inputs. It is designed to do this. Even while falling out at 10,000+ ft/min, it appears that the commanded PFV was something like a 10° climb gradient.

First, you don't command a FPV with the sidestick input, unless you consider "increase FPV", "reduce FPV" as FPV commands. The only way to hold an FPV (and not one given specific FPV but whatever FPV we have right now) is to release the sidestick (which I'd say is stop making sidestick inputs rather than doing inputs).

But more importantly, the underlined part.
In a first-order analysis, you don't need any more THS angle to trim a steep FPV than to trim a straight and level FPV at the same speed. (In a higher order analysis, you need less THS angle for the steep climb!!!)

You need a nose-up aerodynamic command (with the THS or elevator) to CHANGE the FPV from horizontal to the steep climb. Once in the steep climb, you need to bring it back to the original position to HOLD the new FPV (this applies to an Airbus or a Tomahawk).

If you don't have enough thrust to hold the speed, the speed will go down. And now you'll need to keep your lift (and hence to increase the AoA and hence to apply more nose-up elevator or THS) to keep the FPV, but that is the same whether the FPV is straight and level or a steep climb.

[ca787546]

I know that, in Normal and Alternate Law, the g-load commanded is proportionate to stick deflection, so reducing the deflection would reduce the g-load request. I also know that the Airbus has neutral static stability in Normal and Alternate Law, so upon releasing the stick to neutral, the auto pitch trim will maintain FPA until a downward pitch command (-g) is introduced. That is how I arrive at my understanding of sidestick-to-pitch behavior. But correct me if I'm mistaken...

Hey Evan, here's what I have observed in my experience flying airbii If you reduce thrust to idle and you're flying straight and level in a FBW airbus, the aircraft will not actually maintain the altitude without pilot input. As the speed decreases due to the thrust being on idle, the V/S will start to drop (ie, increase downwards) as the speed decreases. This means, every few seconds as speed decreases, I need to gently correct the pitch to maintain level flight. These are very small and subtle corrections, and they are more noticeable when, for example, one is maintaining altitude while decelerating during the downwind leg on a visual. The FDs are off, the AP is off, and you're decreasing speed from 250 knots to Green Dot (Best L/D airspeed) to start configuring.
So, in summary, in normal and alternate law, with the sidestick in a neutral position, the aircraft will not actually maintain a constant FPV when speed changes. Also, during strong roll disturbances (for example, I've noticed this when circling at high altitude airports in mountaineous terrain where moderate to heavy orographic and convective turbulence is found) if the bank angle is modified, the aircraft will not automatically return to the bank angle previously set by the pilot (this is again, with FDs and AP off). The aircraft will ride and correct for light turbulence (disturbances) and will help the pilot counteract things such as sudden thrust changes, spoiler extension, etc when in normal and alternate law.

Hope that sheds some light.

[Black Ram]

Btw, does AF447 remind you all of the 2006 TU-154M crash near Donetsk, or am I the only one? They tried to fly over a storm, exceeded their ceiling, stalled and entered a flat spin. I read somewhere they misused the trim controls.

But then again, there is a chance the BEA got it all wrong and what really happened was they got hit by ball lightning, causing the tailfin to separate due to cheap composites (sending an ACARS message and PFD warning "Vrt Fin Dtach...Again!" ). The pilots did everything right and commenced a climb to FL380 (the advised procedure by Airbus in the common event of a tailfin separation), but then they were hit by another lightning, this time a regular one, frying the flight computers and locking the trim 13dg up. Since in a Scarebus the french computers have authority, unlike in a good old stick n' rudder plane, and there is absolutely no way for the crew to override the computers, the pilots struggled to control the plane. All their chances were evaporated when a violent updraft caused a massive pitch-up, resulting in a stall and a flat spin, momentarily tearing off the rest of the control surfaces as well, making the aircraft completely uncontrollable.

[Gabriel]

But then again, there is a chance the BEA got it all wrong and what really happened was they got hit by ball lightning, causing the tailfin to separate due to cheap composites (sending an ACARS message and PFD warning "Vrt Fin Dtach...Again!" ). The pilots did everything right and commenced a climb to FL380 (the advised procedure by Airbus in the common event of a tailfin separation), but then they were hit by another lightning, this time a regular one, frying the flight computers and locking the trim 13dg up. Since in a Scarebus the french computers have authority, unlike in a good old stick n' rudder plane, and there is absolutely no way for the crew to override the computers, the pilots struggled to control the plane. All their chances were evaporated when a violent updraft caused a massive pitch-up, resulting in a stall and a flat spin, momentarily tearing off the rest of the control surfaces as well, making the aircraft completely uncontrollable.

Very likely. Your report is consistent with the statements of a sailor, who by chance is an expert aviation consultant, who by chance was in the zone and witnessed the accident. He said "there was a violent explosion, then then I looked up and saw the plane falling in a fireball with the engines doing a strange sound, the pilot was a hero as he maneuvered the plane to prevent crashing over my highly populated ship".

They tried to fly over a storm, exceeded their ceiling, stalled and entered a flat spin. I read somewhere they misused the trim controls.

I just thought I'd mention that exceeding your absolute ceiling is no excuse to stall/spin the plane. Trying to stay up there is.

[Theoddkiwi]

I try not to think about the 737NG slat track issue that burned China Airlines Flight 120 to a crisp.

That issue had nothing to do with the functionality of the Slats. One of the attachment parts came loose and through the action of the slats moving in and out pierced the fuel tank. that fuel leaked all over the engine. It only caught fire when the engine was shut down because up until then the exhaust from the engine was blowing the fuel away from ignition sources.

If the receipt person had seen the fuel leak he could have told the pilots to keep the engine running until fire services arrived to cover the area with foam.

But the Slats them selves never lost function or security. as the bolts remained in place, its just the nuts that came loose and fell off.

Inspection of the attachment security is now part of the normal 737 System of Maintenance. It wont happen again unless the inspections are not done by a dodgy operator

[Evan]

Hey Evan, here's what I have observed in my experience flying airbii If you reduce thrust to idle and you're flying straight and level in a FBW airbus, the aircraft will not actually maintain the altitude without pilot input. As the speed decreases due to the thrust being on idle, the V/S will start to drop (ie, increase downwards) as the speed decreases. This means, every few seconds as speed decreases, I need to gently correct the pitch to maintain level flight. These are very small and subtle corrections, and they are more noticeable when, for example, one is maintaining altitude while decelerating during the downwind leg on a visual. The FDs are off, the AP is off, and you're decreasing speed from 250 knots to Green Dot (Best L/D airspeed) to start configuring.
So, in summary, in normal and alternate law, with the sidestick in a neutral position, the aircraft will not actually maintain a constant FPV when speed changes. Also, during strong roll disturbances (for example, I've noticed this when circling at high altitude airports in mountaineous terrain where moderate to heavy orographic and convective turbulence is found) if the bank angle is modified, the aircraft will not automatically return to the bank angle previously set by the pilot (this is again, with FDs and AP off). The aircraft will ride and correct for light turbulence (disturbances) and will help the pilot counteract things such as sudden thrust changes, spoiler extension, etc when in normal and alternate law.

Hope that sheds some light.

Well, thanks, but it only makes things murkier for me. The FCOM tells me that in Normal and Alternate Law, in manual flight with the stick centered, the pitch mode will maintain 1G and adjust for changes in speed. The only way it can do that is to raise pitch, up to the point where protections will limit pitch and alpha floor will kick in (normal law) or all the way to stall (alternate law). The FCTM tells me that "The aircraft maintains the flight path, even in case of speed changes". But in your actual piloting experience, the system does not reliably hold 1G without manual pitch inputs.

The FCTM, mentions, under "Operational Recommendations" that "Since the aircraft is stable and auto-trimmed, the PF needs to perform minor
corrections on the sidestick, if the aircraft deviates from its intended flight path."

A bit of a contradiction. It seems to be telling me that the aircraft is designed to do these things, but has a tendency not to do them without minor corrective inputs. In other words, it doesn't actually hold 1G on its own.

This shouldn't surprise me, since we have seen other instances of systems not behaving as designed, but usually those are anomalous occurrences. Can you explain this discrepancy?

[Black Ram]

Very likely. Your report is consistent with the statements of a sailor, who by chance is an expert aviation consultant, who by chance was in the zone and witnessed the accident. He said "there was a violent explosion, then then I looked up and saw the plane falling in a fireball with the engines doing a strange sound, the pilot was a hero as he maneuvered the plane to prevent crashing over my highly populated ship".

In addition he also has hundreds of hours of flying RC gliders.

I just thought I'd mention that exceeding your absolute ceiling is no excuse to stall/spin the plane. Trying to stay up there is.

Fair point. They exceeded their ceiling, which is when an updraft caused an uncommanded pitch-up and prevented them from pointing the nose down, causing a stall (a stall is a condition when an aircraft loses lift because it's flying too slow). The stall made the aircraft unstable, and then a powerful air current caused a flat spin, which in turn tore off the vertical stabiliser with the horizontal stabilizer attached to it, due to low-grade legacy metal components. There was also a fire on board, engine failure, and the aircraft was struck by lightning.

[Evan]

You need a nose-up aerodynamic command (with the THS or elevator) to CHANGE the FPV from horizontal to the steep climb. Once in the steep climb, you need to bring it back to the original position to HOLD the new FPV (this applies to an Airbus or a Tomahawk).

Does a Tomahawk have an aft CG? In a Tomahawk you are trimming against positive stability aren't you? What defines the Airbus here (and the MD-11 and the 777) is the negative static stability issue. If you increase the FPA with pitch, that is going to produce lift which will create more pitch, right? So you need the THS to compensate for any further pitch moment beyond the commanded FPV and trim the aircraft to neutral static stability with the stick released. In a typical aircraft with a forward CG, the stabilizer trim mimics the pitch command. In a aft CG aircraft, it opposes it.

I assume the THS setting in the report refers to the amount of nose down stabilizer deflection required to trim nose up pitch in neutral stability, not the amount of stabilizer deflection required to hold the nose up. If the pilot made an abrupt and pronounced pitch input, could it not require this much to trim back to neutral stick? What about if the aircraft is falling at 10° ANU and something like -3G with an aft CG, with the THS trying to hold that pitch?

If I'm wrong in that assumption, then the THS setting baffles me.

[Evan]

In addition he also has hundreds of hours of flying RC gliders.

Man, this has become a mean-spirited forum.

Has it ever occurred to you, Black Ram, that the vertical stabilizer might have separated in flight, but stayed in close proximity to the fuselage due to the theory of constant planar inertia (see diagram) long enough to create an upset but also long enough for ball-lighting to weld it back on, thereby fooling investigators into thinking it never separated in the first place. We don't know much about ball lightning, but we know it is clever as hell.

[ca787546]

Well, thanks, but it only makes things murkier for me. The FCOM tells me that in Normal and Alternate Law, in manual flight with the stick centered, the pitch mode will maintain 1G and adjust for changes in speed. The only way it can do that is to raise pitch, up to the point where protections will limit pitch and alpha floor will kick in (normal law) or all the way to stall (alternate law). The FCTM tells me that "The aircraft maintains the flight path, even in case of speed changes". But in your actual piloting experience, the system does not reliably hold 1G without manual pitch inputs.

The FCTM, mentions, under "Operational Recommendations" that "Since the aircraft is stable and auto-trimmed, the PF needs to perform minor
corrections on the sidestick, if the aircraft deviates from its intended flight path."

A bit of a contradiction. It seems to be telling me that the aircraft is designed to do these things, but has a tendency not to do them without minor corrective inputs. In other words, it doesn't actually hold 1G on its own.

This shouldn't surprise me, since we have seen other instances of systems not behaving as designed, but usually those are anomalous occurrences. Can you explain this discrepancy?

Please read carefully FCOM1.27, FCOM3.4.27, and based on that information (which supersedes that of the FCTM) read again the FCTM OP020. I won't write what is says down because you're a researcher-type person and I know you have the manuals.
I will say, the aircraft gives G-load demands at high speed, and pitch-rate commands at low speed (I figure you already knew that though) and if you make a smooth deceleration at 1G it doesn't mean that for the V/S to increase downwards the aircraft is reducing its load factor below 1G. You may feel a further pitch down because your inner ear is being fooled by the reduction in forward speed (not actual G change), but the aircraft isn't.

[EconomyClass]

Hmmm. Can't imagine American corporations asking a US Court to cede jurisdiction to France. Globalization run amuck?

June 3 (Westlaw Journals) - The manufacturers sued in connection with the 2009 crash of Air France Flight 447 have fallen far short of showing that France is an appropriate forum for the litigation, the families of those who died in the accident have told a California federal court.

The families have asked the U.S. District Court for the Northern District of California to deny the dismissal request lodged by Honeywell International, Motorola Inc., Intel Corp. and General Electric Co.

The plaintiffs say the manufacturers’ “unilateral submission to jurisdiction in France is not effective to create jurisdiction under European Council Regulation 44/2001” because none of the parties live in France or the European Union.

The litigation stems from the June 2009 crash of Flight 447, en route from Rio de Janeiro to Paris. The Airbus A330-200 plunged into the ocean about 680 miles off the coast of Brazil, killing all 216 passengers and 12 crew members.

Family members of the victims filed lawsuits alleging the radar system did not provide accurate data and weather information. They also claim the plane experienced problems with the onboard software, microprocessors, deicing system, and various stick controls and sensors.

http://newsandinsight.thomsonreuters...r_France_case/

[Evan]

I will say, the aircraft gives G-load demands at high speed, and pitch-rate commands at low speed (I figure you already knew that though) and if you make a smooth deceleration at 1G it doesn't mean that for the V/S to increase downwards the aircraft is reducing its load factor below 1G. You may feel a further pitch down because your inner ear is being fooled by the reduction in forward speed (not actual G change), but the aircraft isn't.

Oh, you lost me. Load factor = lift / weight, right? I don't understand how an object, any object, can be falling at 1G. I'm certainly no physicist, but my understanding of 1G is that the force supporting the object (lift, in this case) is equal to 1 x the gravitational force exhibited by the object, and therefore the object neither rises not falls in relation to the earth's center.

Or are you saying that gravitational acceleration as perceived by the accelerometers will remain at 1G in a smooth deceleration and descent?

Can you explain that a bit further?

[Gabriel]

Does a Tomahawk have an aft CG? In a Tomahawk you are trimming against positive stability aren't you? What defines the Airbus here (and the MD-11 and the 777) is the negative static stability issue.

There is no such a thing. All these planes have positive static longitudinal stability.

If you increase the FPA with pitch, that is going to produce lift which will create more pitch, right?

Well, assuming an unstable plane, not exactly. The stability is a matter of AoA, not pitch. When an aircraft is stable and the AoA is disturbed from its equilibrium position, it turns to return to that equilibrium (trim). If it's unstable, then after disturbed it tends to amplify the deviation from the equilibrium (trim). While the AoA and pitch angle are related, you can have for example an increasing pitch with a diminishing AoA at the same time. Anyway, these planes are stable, not unstable.

So you need the THS to compensate for any further pitch moment beyond the commanded FPV and trim the aircraft to neutral static stability with the stick released. In a typical aircraft with a forward CG, the stabilizer trim mimics the pitch command. In a aft CG aircraft, it opposes it.

If you are talking about an artificial stability system, then I very much doubt that a THS like those we are discussing here would be of any help.

You would rather need the elevator linked to the stability function. That's because a gearbox with a jack-screw is just no fast enough. In an unstable airplane, a good bump of turbulence that creates a somehow strong disturbance in the AoA can bring you to an unusual "tail first" flight in a hurry.

You must differentiate between "trim" (equilibrium) and stability (what will the evolution if the system is disturbed from the equilibrium).

The THS and elevator position in a steady horizontal flight and in a steep climb or descent at the same speed are the same, regardless of whether the plane is stable or unstable, because they both require the same lift and hence the same AoA (i.e. same state of equilibrium). It is the transition between them that changes.

In any event, again, it's irrelevant since the A330 is stable even with the CG in cruise mode.

I assume the THS setting in the report refers to the amount of nose down stabilizer deflection required to trim nose up pitch in neutral stability, not the amount of stabilizer deflection required to hold the nose up.

I'm assuming it's exactly the opposite: The trim required to set the state of 1G equilibrium at a higher AoA in a diminishing airspeed situation. Not only am I very smart. Also:

The trimmable horizontal stabilizer (THS) passed from 3 to 13 degrees nose-up in about 1 minute and remained in the latter position until the end of the flight.

If the pilot made an abrupt and pronounced pitch input, could it not require this much to trim back to neutral stick? What about if the aircraft is falling at 10° ANU and something like -3G with an aft CG, with the THS trying to hold that pitch?

If I'm wrong in that assumption, then the THS setting baffles me.

I think you are overestimating the role of the THS. There is one thing used as primary pitch control, and that is the elevator. If "the airplane thought" it was pitching up too much, that is, the pitch rate was beyond the target value (given the inputs), then it would have simply lowered the elevator a bit.

The THS is there for a longer-term goal, which is to cancel the average force applied over the elevator, so the pilot (or the auto-pilot or the FCCs) doesn't have to keep a pull or push force. Or to put it in other words, to make that the equilibrium AoA (that about which the plane oscillates if the sidestick is released) is that needed to fly at 1G.

I'm no expert in Airbus systems, but that's my interpretation.

[Gabriel]

...if you make a smooth deceleration at 1G it doesn't mean that for the V/S to increase downwards the aircraft is reducing its load factor below 1G.

You are not saying that you can change from horizontal flight to a descent while keeping 1G even during the transition, are you?