CONF iture

Apparently, part of the ab initio training at AF, the students spend one week in St-Auban.
min 41 in the TV program : AF en quête de sécurité

Clandestino

Yup, it seems that demise of Spanish crew, flying Spanish registered narrowbody, built in USA and out of production, plowing the skies of Africa, does not produce enough outrage to make the PPRuNe discussion worthwhile. Also it would somehow damage the positions of those who claim that kids today can't fly and are stalling aeroplanes because they are inexperienced to note that the commander was 14 000 hrs TRE.
It is not that simple, especially if we use AF447 as the starting point of the discussion since the report's greatest weakness was total absence of references to pilots' previous performance. That's something NTSB is very keen to explore, no matter whether they are investigating 4- or 400-seater occurrence but now with BEA we are left wondering whether the fatal crew breakdown came without previous warning, was at the end of unbroken chain of marginal performance or somewhere in between.

If I present you with three pilots who met their fate by a) flying Centurion into CB that tore it apart b) trying to pull his Decathlon through mountain pass on hot day at density altitude she just couldn't cope with c) being part of the crew that made 90 deg heading mistake, putting their aeroplane on the collision course with mountain, would you say "Now there are pilots who didn't know how to fly"?

Would you change your mind if I told you those folks were a) Scott Crossfield b) Steve Fosset c) Don Williams. Maybe it would help if I explain they were a) NAA engineer and X-15 test pilot b) pilot who flew solo around the world without landing c) USAF instructor of the year (1986. IIRC).

AF447 crew couldn't cope with the problem they had, at the time they had it. Whether they were fine pilots that were too fatigued from Rio layover to think straight for a fraction of a second or marginal ones, being lucky that far they never had to cope with serious malfunction, we just don't have a way of telling.

Even if one does not know anything about Airbus, mere fact that 40-something 330/340s passed through similar ordeal, with autopilots tripping and flight controls system degrading to ALTN law till landing, with no injuries or damage, is indication enough that aeroplane is no culprit.

Heck, if there would be just one thing worth learning from AF447 demise, IMHO it would be:

Maybe it should be written on QRH covers? Or on another cockpit placard?

gums

We are finally getting to the core of flying safety here, and we must take into account technology, human factors and the bottom line $$$$.

Late here, and will opine later, but first.

Few of us here arrived on this forum without having experienced one or more situations that were not planned or we could not avoid with all our skill and planning. Face it. Sierra happens. Mechanical things break. Weather is not what the troop told us. and the beat goes on.

The bottomline is how we handle the unexpected and how were we trained. Along the way, how do we fly day-to-day? Do we assume nothing bad will ever happen? Or do we think "what would I do if "x" happened this instant". I can tell you what this old survivor did. And I'll do it later, as just got back from TDY and hitting the rack.

Winnerhofer

Last paragraph:
Vol AF 447 Rio-Paris : reconstitution des minutes qui ont précédé le crash | Vanity Fair
AF says it vehmently denies that its crew were incompetent and that the article makes no mention of Airbus' systems that induced piloting errors.

DozyWannabe

Er, what systems do you believe "induced piloting errors"? At no point in the sequence did the crew act in a concerted manner to either apply the appropriate procedures or perform logical problem-solving operations.

TC-DCA

Is that true that in the case of AF 447, even if the stick was at neutral position AFTER the beginning of the stall by the PF, the aircraft would have maintained elevators AT FULL NOSE UP to maintain 1G as it's Airbus rule, and with stick full fwd, the elevators will be also in nose up but in half nose up position. True ?

So it'll be impossible to recover the aircraft even with full fwd stick, also aggravated by the THS that was at FULL NOSE UP due to pilot actions. True ?

I just have another question to avoid opening another subject concerning Airbus Load Factor Protections.
As we know, in clean config, the g load limit is +2.5g positive, so with a bank to 67 degrees for example, (or a high positive pitch), the load factor will be approximately 2.5g, so does it means that we'll loose authorities in pitch even by pulling it full back or full fwd as we are in g limit ? Is that true ? It's strange if that's true ? I remember the Gulf Air crash with that.

I don't understand exactly this function on Airbus FBW, as it's not detailled in the manuals of the aircraft, anyone to answer this Load Factor prot. with Airbus?

Thanks !

gums

@ TC

Best I can tell from the FCOM stuff I have gotten courtesy of the folks on this thread, until in "direct" law, the system tries to maintain one gee corrected for pitch attitide. So airspeed and AoA are not the biggest drivers even when in alternate laws.

And then the reversion sequence is not all that clear depending upon what sensor or computer you lost. Sheesh.

On the thread about the Indonisian event, the comments about Airbus flight contol law and reversion sequencces are confusing. If you look at the laws in the FCOM, they try to "protect" the plane ( not the pilot) as long as the computers deem the inputs reliable.

So we go from "complete protection" from anything to protection about some roll, pitch angles. Uhhhh, AoA? Gee? What else? Lemme get out that checklist or use the electonic display ( if available) to see where we are. Meanwhile the plane is climbing without a pilot command. The autothrottle might also be doing stuff we have not commanded. Sheesh.

Later, as this old pilot is weary of all the "gotchya" crapola of the new systems. And remember that this old fart flew the first fully electric jet and out reversion sequence was very simple.

TC-DCA

@gums

Thanks for the answer, but I dont have a real answer to my question it s a little complicated.

So as I said, is that true that in the case of AF 447, EVEN if the stick was at NEUTRAL position AFTER the beginning of the stall by the PF, the aircraft would have maintained elevators AT FULL NOSE UP to maintain 1G as it's Airbus rule, and with stick FULL FWD, the elevators will be also in nose up but in half nose up position. True ? If thats true, I think it´ll be stupid !

Can anyone explain this ?

Especially the Airbus Load Factor 1g demand, it´s horrific to hear that the plane computers demand full nose up in a stall by itself, with neutral stick, it wouldnt be the same with a conventional plane.

So it'll be impossible to recover the aircraft even with full fwd stick, also aggravated by the THS that was at FULL NOSE UP due to pilot actions. True ?

I don't understand exactly this function on Airbus FBW, as it's not detailled in the manuals of the aircraft, anyone to answer this Load Factor with this 1g, what is the link between 1g rule and elevators ?

Thanks !

vilas

TC-DCA
In pitching plane when the stick is out of neutral pilot is asking for some load factor from flight control computers and it is proportionate to how much the stick is out in fore and aft plane. The elevators and stab position themselves to give you that by pitching up. When the stick is released and is neutral the elevators will maintain 1G. In normal law when speed(AOA) goes below Valpha prot the stick demands AOA and not load factor as in that zone between Valphaprot and Valphamax AoA is more critical and relevant. How ever in alternate law there is only Vsw(stall warning) and no V alphaprot or V alphaMax. So below VLS is it load factor or AoA demand is not clearly mentioned in the manual. As far as stab going full up you have to understand that the pilot kept the stick back all the time even below Vsw giving the system no choice. However the system design would have gradually kept trimming back once the nose was up even with stick neutral to keep 1G. Similarly if the stick was pushed forward it would have translated in to negative load factor demand and the elevators and stab would position to give you that and ultimately recover from stall but the pilot never did that for any length of time. Every time stall warning came he pulled the stick back instead of pushing forward. AF447 was manoeuvred beyond its design considerations by a pilot who was not competent to deal with unreliable speed situation nor was he capable of recovering from the resultant self induced stall. Conventional airplane don't trim back but they have been stalled by pilots through mishandling and crashed because of faulty stall recovery procedure. AF447 pilot would have never recovered from the stall in any aircraft.

gums

It's hard for folks used to conventional flight controls to "appreciate" all the "protections" or "limits" as I prefer. Once you abandon a direct stick/yoke movement or pressure on a hydraulic valve and the corresponding action at the control surface, all kinda neat things are possible to smooth out the ride and then make it easier for a pinball wizard to fly a jet.

The gee command is hard to understand unless you see it during training. As in the Viper, AoA plays a role if the sensors are deemed reliable ( although the Viper AoA was deemed reliable if weight was off the gear - we had different criteria for "reliable" sensors than the 'bus). With a neutral stick and the jet trimmed for any gee, it would maintain the gee until the AoA reached the "limit", then ride the AoA limiter until speed increased/AoA decreased. We did not correct the gee for pitch attitude due to our mission requirements, huh? So we had a simple function that related gee to AoA. At 27 degrees or so AoA, max gee was 1 gee!!! At 15 degrees, max gee was 9. Sucker would even command zero gee to stay at 27 degrees AoA or below. So I would demo a loop to Joe Baggadonuts student by trimming to 3+gee, and let go of the stick. Sucker would smoothly pull up and then reach max AoA as it slowed, unload over the top, then increase gee as we came down the back side and AoA allowed. Due to our awesome vis, you could look back at the horizontal stabilators and see them almost maxed out as we came over the top ( nose down command just about as much as it could, but not quite. If nose wasn't moving and stabilators were maxed out, you were in a deep stall).

I mentioned this before, but our first group of youngsters in the Viper were of the Atari generation ( Xbox and Playstation 20 years away). We were worried that they would forget basic aerodynamics, as the jet was very easy to fly without getting into trouble. However, they had accumulated 250 or so hours in "conventional" jets that had no auto-anything!!! They did just fine, and went on to fly other jets after the Viper.

The key is to have a decent amount of flying in simple airplanes and no auto anything. Some "refresher" training is also highy recommended.

Finally. @engineers, quit trying to protect the plane so much when the primary modes go tango uniform. A straightforward reversion that can be easily understood seems logical, and I think I saw a comment to that effect in the accident report.


BTW, my pearl of wisdom I alluded to earlier was to do thousands of "thought" emergencies. What if? Sitting in the barbershop, waiting for a bus/train/plane. When the actual emergency happened, I had a personal procedure someplace in the back of my brain. I also had a philosophy of "hold what you got", at least if the plane wasn't tumbling end over end.

So what did I do when this happened shortly after lifting the gear handle?

flyingchanges

Eject.....

gums

C'mon, Flying.....

The jet is on the ramp, right where I started an hour earlier. Flight was about 15 minutes, and taxi for takeoff about 10. Taxi back was about 3 or 4 minutes, and you can see some on the video.

http://www.sluf.org/misc_pages/lef-landing.m4v

May have to find a good video player, but the landing was recorded in case I could not get down O.K. We had a limited FDR on the seat and the HUD video/audio was fairly survivable, and we got useable tape after several bad crashes.

TC-DCA

@gums

Wow ! I read a part of the LE problem, that's just amazing.
I'll continue to read the article, that's really interesting !

@vilas

Thanks for these informations !

So as we know, with stick at neutral, ELAC's try to maintain 1g with elevators, until we don't "request" another load factor demand by pushing or pulling the stick, but what I don't really understand is that, if we pull back the stick and then return it to neutral position at 10° for example, the aircraft will maintain it, so what is the purpose of the 1g rule here, the pilot normally demand a stable pitch of 10° "not a stable g" normally ?

That'll be strange for a pilot to see his aircraft pitching up with stick at neutral in a particular condition, like a turbulence, and then, they'll note it to the maintenance log as "uncommanded pitch and altitude changement in cruise/climb/descent", that's what I don't understand, so the aicraft is not really stable and doesn't maintain the pitch ?

Can you explain that complicated rule of Airbus Captain please, with a simple example ?

For another example, here the g limit, if we do a turn of 50 degrees with holding the stick, (or a pitch up to the maximum limit of 30 degrees), the aircraft will be at approximately 2-2.5g, and as positive g limit in clean config. is 2.5g, is that true that we'll nearly or completely loose the pitch during a steep turn for example, to don't exceed the limit imposed by the computers as we are already near 2.5g in a steep turn ?

EDIT: "However the system design would have gradually kept trimming back once the nose was up even with stick neutral to keep 1G."

How is it possible that computers orders nose up with elevators with stick in neutral, and when it does that, normally even in a stall condition, with stick at neutral, the aircraft must maintain the last commanded pitch, so it continue to stall but it normally should t command the g ? True ?

vilas

TC-DCA
I didn't explain 1G fully. In conventional aircraft you directly move the elevator that creates some load factor to raise the nose to 10 degrees of pitch and then trim yourself to keep it there, if you didn't trim and left the yoke aircraft will pitch down as it is not in trim. FBW treats side stick movement as a certain load factor demand and computers give you that through the required amount of elevator to raise the nose up and when you return the stick to neutral keeps it there by auto trimming. Maintaining 1G is same as maintaining same degree of pitch what ever you set it. So it trims to maintain the pitch but when the thrust is not sufficient only then FBW will keep pitching up to maintain 1G flight path, otherwise the nose will drop resulting in some minus G. The advantage of treating side stick movement as a load factor demand is that when you move the stick back one inch the load factor ordered is same at all speeds but the elevator movement is varied to give the load factor, thus the aircraft response to the movement of the side stick is same at all speeds unlike normal aircraft where you have to pull the stick more or less to achieve the same degree of pitch depending on the speed of the aircraft. There is nothing complicated while flying the aircraft you just fly like any other aircraft.
As far as 2.5g is concerned it is achieved in a level turn at 67 degrees of bank so if you reduce bank you can pitch up otherwise not. You are flying a commercial jet and not a combat aircraft. Conventional aircraft how much it can pitch up depends on speed at higher speed you can pitch it up beyond its structural limitations causing structural failure. In FBW you cannot.

PJ2

vilas, that's a very nice description of the Airbus system and should help a lot for someone trying to understand FBW.

TC, it's also an important point to emphasize that FWB is not the same as "protections".

Protections may be built into the system but they are separate from a FBW flight control system.

In terms of controls like the thrust levers, they could be dining-room dimmer switches for all the system cares! They are "shaped" like throttles out of convention and on the Airbus act more like "engine room telegraphs", setting power regimes in the various detents in which the autoflight/autothrust system then respond.

In the end it is most important to understand that the Airbus is just another airplane and can be flown just as any other aircraft. When one is trained on it and as a pilot, puts appropriate work into learning the airplane, it is a straight-forward machine and a treat to hand-fly both for the approach/landing and at cruise, (although not in RVSM airspace which legally requires the autopilot remain engaged!

The autoflight system for the A320 & A330/A340 is described at Smart Cockpit - a visit there would be enlightening. Happy reading!

Clandestino

If the stick were held at neutral, FCS would try to maintain constant vertical flightpath which would tend to exacerbate the stall but FCS does not use elevators alone to achieve demanded G, it uses THS full time - unless THS inhibition threshold is reached (e.g. less than 0.5G actual vertical acceleration)

False. Push forward and FCS moves THS and elevator towards nose-down position to comply with G demand.

It's true 67 degrees banked coordinated turn demands 2.5 G so no pitch up authority is available but high positive pitch does not automatically imply 2.5G. It's about pitchup rate, not angle.

Say, whaddaya need 67 deg banked turn on transport aeroplane for?

GF072 nevere came anywhere near G or bank limits. Have a look at the final report.

Because they were wrıtten by confused posters. Any similarity between this thread and that is not coincidental.

Lack of ejection seats on FBW Airbi makes pilot and aeroplane inseparable until the aeroplane has come to complete halt. Therefore protections are not at all ill-conceived.

What plane are you referring to? FBW Airbus will maintain constant vertical flightpath but to climb, demand must be made by the pilot and it was in AF447 case. Very emphatically.

It is vertical flightpath stable up to 33deg bank. It adjusts the pitch to maintain flightpath.

In real life, such a TLB entry would lead to tech refresher at the very least.

...intentionally, flying the line, in Airbus, you get fired and unemployable.

I suppose you mean "not following procedure", not procedure being faulty per se.

vilas

Clandestino
I mean both. AF447 did not follow proper procedure while previous procedure of slamming the thrust before lowering the nose itself was incorrect.

Clandestino

Hmmm.... could you please provide an example of an accident/incident where application of "Approach to stall recovery" instead of "Stall recovery" was a factor?

vilas

Earlier stall recovery procedure was based on approach to stall but after a few accidents in the US where the thrust increase prevented the pilot from lowering the nose FAA asked for a review of the procedure. There is a video by Airbus on the new thinking on stall recovery. Experts agree that even for a test pilot it is difficult to make out the approach to stall and full stall. So there is a standard recovery procedure now that recommends reducing the angle of attack first and once the stall warning has stopped increase thrust as required not necessarily TOGA.

gums

@ cland, et al ....

I go with vilas for the most part. Ditto for PJ.

My problem is folks describing to novices the "effect" of a control law versus the "intent", and actual code for HAL. PLZ convince me that the 'bus control law for pitch is primarily programmed for an attitude versus a gee. I realize that correcting the gee command for pitch attitude achieves the "percieved pitch" command versus a gee command. Our primitive system had no correction for attitude for fairly obvious reasons. So establish a 30 degree pitch attitude and stay at same power as the jet loses energy. Instead of 0.87 gee, the sucker tried to maintain trimmed gee, which was ususally one +/- a few hundreths. So pitch attitude gradually increased to maintain one gee ( Nz). Horizontal stabs "auto trimmed", just as AF447. Unless our AoA limiter failed, we were headed for a deep stall.

- I feel your pain, Cland, about no means to jettison the jet. Seems that should be a serious incentive to learn as much about the jet and its systems as possible and practice a few things when able.

- Our limits were intended to keep the pointy end forward and keep us from putting too many gees on the airframe and bending the wings if we rolled too fast under a decent gee load.
So our laws were a combination of "protect the jet" and "let the pilot command max performance without having one eye on the gee meter or AoA indicator or.........." So we could command maximum performance at any time and she gave you what she could. Maintenance troops loved it, as pervious jets could easily be over-geed. We would come back with ripples on the upper skin of the wings. Bad. "Bad pilot, bad", heh heh.

I like the 'bus clontrol laws in "normal" . They are about what I would expect for the mission and the mechanical design limts of the jet. So my problem is with reversion laws and lack of indications that the jet has reached trim l;imits or AoA limits or.......

- My comment about continuing a climb with stick in "neutral" stands. Until the Alpha stuff comes into play, PLZ show me where the jet will nose over or act like the "old" ones most of us flew years ago. In other words, if energy and AoA allow, the jet will continue to climb after you relax the stick. It will also trim the THS as long as it can.