Clandestino

Because what was discussed is the part following the actual aerodynamic stall, where the ice has melted away, indicated airspeed was realistically measured to be below 60 kt and AoA exceeded 45°. Someone with not much knowledge about basic aerodynamics or flying dug out the certification requirement that stall warning should sound as long as the aeroplane is stalled, took umbrage at Airbus' decision to use 60kt instead of WoW to invalidate AoA measurements, thereby cutting off the stall warning and here we are. One step short of insisting that X-31 style probe should be installed on all airliners.

rudderrudderrat

Hi Clandestino,

Please explain how a ROD of 10,000 per min (about 98 kts) is not sufficient to validate the stall warning.
The invalid <60 kt logic must was meant for on ground nuisance warnings.
Using the same logic when airborne must have been designed by

xcitation

Coupling the stall warning to the IAS is clearly a problem. This cross dependency needs to be eliminated. If you lose/degrade IAS <60kts then you lose the stall warning! But when you lose IAS you need a stall warning more than ever. The only work around is to select the AoA display.

ChristiaanJ

We've been there before... and unfortunately we have no comment from whoever designed that logic, instead of just using WoW.
Somehow IAS <60kts at altitude was never considered....

Oddly enough, I still remember an [IAS <60kts] switch in the logic of a FCS I was involved in.... but it was an analogue system.... and it was a helicopter.... so not relevant here.

Machinbird

Clandestino
You must have missed the post where I stated I had personally tested an actual airline (In this case DC-9) AOA transmitter and that it came alive at 15-20 mph based on my vehicle's speedometer reading. Since I drove in a closed course and had consistent results, wind was a negligible factor.

I suppose it is possible that the A330 AOA transmitters require much higher dynamic pressure to provide accurate readings, but somehow I doubt that the manufacturers were that inept.

The term 'come alive' means that the transmitter has enough aerodynamic force operating on the vane to cause the vane to point in the direction of airflow and to maintain its direction while you turn the body of the transmitter. The difference between 20 mph and 60 knots is a factor of 12 increase in dynamic pressure.

The design and manufacture of electrical AOA transmitters has been going on for well over 50 years. A Korean war vintage jet I used to fly had one.

The design and manufacture are not trivial, but it is an art that is well understood.
The location of the AOA transmitters on an aircraft, requires understanding of the flow fields around the aircraft at various angles of attack and selection of one which will create the most representative indications.

The modern jetliner envelope is nothing unusual, you guys usually like to stay sub-sonic.

I do not know why Airbus allowed the AOA indications to be invalidated by speeds <60 knots, but "there is more than one way to skin a cat," and more than one way to achieve whatever objective they were attempting to achieve with that logic.

As I have stated numerous times, AOA is primary flight information and airspeed is primary flight information and they are each independent variables. To permit one to disable the other is setting up a loss of both pieces of information. Conceptually, this is poor logic from a reliability standpoint. It is just that the 60 knot point is so far out of the normal flight envelope, that the engineers never dreamed that it could potentially be a problem for someone.

We do not need X-31 style probes for airliners. The existing ones work well enough. Once AOA is well away from normal flight envelope, that is good enough information to work with unless you are running a full stall characteristics flight test.

But you do need to know that AOA is approaching flight envelope limits, preferably with a stall warning device that cannot be ignored and an actual indication to reinforce that information.

CONF iture

Of course ... Nevertheless they do not share all the facts starting with data and won’t necessarily feel the need to share all of their analysis, with Clandestino or CONF iture.

gums

I gotta go with 'bird and other pilots.

I realize that an airspeed below 60 knots ( TAS, EAS, IAS, CAS, whatever) is unlikely once airborne. OTOH, the AoA vanes or conical probes or whatever are independent of the pitot-static system.

If the pitot-static sensors go tango uniform, then do we ignore the AoA sensors?

Somehow, I get the feeling that the cosmic engineers never expected the jet to have unreliable dynamic pressure and then get close to a stall. Just ignore the AoA sensors, huh? This is despite the fact that actual EAS/CAS in the AF jet was above 60 knots and the AoA sensors were prolly still providing valuable data to the FCS.

Sounds like a fault tree analysis is in order.

I flew two jets before the Viper with the old-fashioned vane style AoA doofers. Then I flew the Viper that had the conical AoA things with slots in them. The Viper also had a pneumatic, hemispheric probe that provided another AoA indication and sideslip inputs for the FCS. If we lost dynamic pressure we used "standby gains" for "q", but still used the AoA data. Worked for me and thousands of Viper pilots.

If the AoA sensors are getting weird due to actual aircraft flight conditions, then no big deal. It means that the jet is doing something weird!! So don't ignore the sensor values.

And then there's all the "alpha prot" and other "laws"/"sub-laws" and on and on and on. Sheesh. Too many autopilot functions are embedded in the design. Maybe have reversion "laws" that are clearly defined and get rid of all those functions that should be allocated to Otto when it is engaged. And make no mistake, I am not advocating a reversion to "direct" law. There needs to be a basic "law" that uses basic flight dynamic data to provide the pilot a "basic" airplane to fly. If you lose dynamic pressure, then use AoA. If both go tango uniform, then pitch and power are the drill.

CONF iture

That what is Direct Law gums.

grity

is´n it the same for the AP? is it realy realy that an airbus in his auto-logic hold the altitude with the elevator+trims and hold the speed with the power ??? should the AP not hold the altitude with the power and the AoA with the elevators ? at least in case without dynamic pressure.....

Linktrained

NARVAL #885

Thank you for guiding me (and others) to Jacques Roysay's article in "Airbus Safety Magazine" of January 2011.

IIUC, he suggests that, in Direct or Alternate Law, one should fly 4000 ft. below the height recommended for FBW, especially in turbulence. (To keep the SW quiet ?)

He also discusses SW and the use of TOGA on flights at lower levels (in a 340)

Ian W

Clandestino
Clandestino it is not a stretch at all - and I am sure I could make you react in a similar way in a simulator where you 'totally fail to notice' something that is obvious to an outside observer. Overload a cognitive channel and NOTHING else will be processed by it. There is a famous experiment where a team of observers of a basket ball match did not notice a man in a gorilla suit on the basket ball court. Or for an aviation example a set of fighter pilots flying a 'highway in the sky' type display who on overshoot each flew through a blimp that was VERY obvious in front of their simulated aircraft. Then there was the YouTUbe video earlier in this thread showing a crew of three landing an aircraft wheels up while in the background for the entire approach the undercarriage warning horn was sounding. The examples are endless.

Unfortunately, people who disbelieve the existence of cognitive overload have been allowed to design, test and implement aircraft systems. Setting up the potential for just the type of human factors caused accident that we have seen in AF447.

This is why there is now a human factors team analyzing what happened. It would have been far better for the passengers of AF447 if the human factors analysis of what could happen had been before release to service

Lyman

Likely the release of the CVR record will explain the PNF's "What was that?"

Frankly, BEA have no choice, since they have ascribed his comment as "Probably related to the STALL SV."

Without actually hearing the noises in the cockpit, they will be stuck with "Probably". So that much more data will be forthcoming, what a relief.

Right, CONFiture?

IFLY_INDIGO

This accident shows that how much we rely upon the indications inside that we forget to look outside and fly using stick and rudder

GerardC

1) Does anyone know why AP disconnected at 02:10:05 ?

According to NTSB (and AI), see : DCA09IA064 : "According to Airbus documentation and analysis, if an airspeed discrepancy of more than 20kts or an altitude discrepancy of more than 400ft is detected between one ADR and the two others, the subject ADR is rejected. Then if a discrepancy occurs between the two remaining ADR, all auto flight functions are lost and autopilot, flight director, and autothrust disconnect.".

I do not see any discrepancy between the two recorded AS before, at least, 02:10:07.5.

In the TAM case : "the No.1 ADR airspeed dropped rapidly from about 260 knots to approximately 60 knots for a few seconds, then rose to 100 knots. At the same time, the recorded pressure altitude dropped by about 300 feet. About one minute later, the autothrust and autopilot disengaged..."

Why such a hasty AP disconnection in the AF 447 case ?

2) Why did the "NAV ADR DISAGREE... IF SPD DISAGREE... ADR CHECK PROC... APPLY " ECAM message came in as late as 02:12:XX ?

Lyman

GerardC

There is an additional circumstance. Without discrepancy among the ADRs, but with a drop or gain of >30 knots with the concurrence of all three, the automatics drop out. This is sudden, and would seem to relate to the need for immediate manual control, as in WIND SHEAR.

Turbulence is the prime candidate for even a conformal change in airmass speed (or direction, sensed as the 'same thing' by the a/c).

The sudden loss of a/p is consistent with the lack of a 'run of the mill' UAS.

It also argues against ICE having any impact on the Probes. Beyond this, airspeed variance of this order would also be associated with OVERSPEED PROTECTION.

Overspeed (a/c reaction to) explains the fixed position of the THS in the climb.

Similarly, a possibility of this sort explains why the speeds did not become discrepant until after the loss of a/p, and the degradation to AL2.

My understanding is that with WIND SHEAR, the a/c remains in NORMAL LAW. This also would explain why, if so, the PF was convinced he was in NL well into his climb, and felt safe in Pulling aft stick.

DozyWannabe

That is a complete falsehood. Half the reason for the somewhat arcane annunciation logic employed in modern airliners is for the precise reason that cognitive overload is so feared and to be avoided. This is why Stall Warning quite literally trumps a whole host of others.

@Lyman - give it up. The aircraft did *not* climb due to autoflight input and autoflight did not move the THS. It was pilot input, pure and simple. The reason for the delay in disconnect and the message is purely because it takes some seconds for an initial ADR discrepancy to become a confirmed ADR DISAGREE status, as was mentioned quite clarly in the earlier threads.

Lyman

Dozy. Nothing of any conclusive value has been demonstrated, other than to the threshold of this forum (!).

Autoflight did indeed initiate a NOSE UP at hand off (with the AoA at three degrees below cruise), and was climbing the a/c at 1000fpm (conservative) by the traces published by the BEA. This is in the record, and autoflight would not have climbed the a/c after that, where did I post that?

Neither did I posit a delay in "disconnect", where do you get these things? Of course the PF did not change the THS in the climb, it did not move at all. This is not consistent with anything other than a potential for Overspeed protection. You have eliminated that? You are making wild statements where BEA is still using "Probable", etc. Who is the rabid one here?

DozyWannabe

The point is that it appears that you're arguing that the autoflight should be considered in part responsible for the zoom climb initiated by the PF (as I read it), which doesn't look to be the case. You're also arguing overspeed protection coming into play, despite the fact that even CONF iture himself has said that in Normal Law, the Overspeed warning will sound before any protections kick in. There were no occurrences of overspeed warning on either the ACARS or the CVR, based on the material released so far. You've been continually trying to argue that the aircraft was responsible for the loss of control and you say you're doing this to "defend" the pilots. Defend them from what?

In the legal sense, it's always going to be a case of split responsibility between Airbus and the airline because of the known pitot tube issues and the subsequent mishandling of the aircraft.

Lyman

Autoflight is not responsible directly, no, though the a/c was given to PF with a need for correction eg climb, roll.

The CVR showed no warning for the simple fact that it has not been released, only a very few transcribed statements. The ACARS is not the Gold Standard, as you would believe, it simply is not.

You insist on eliminating possibilities out of bias, or worse, even pique. I am not aware of evidence eliminating the possibilities about which I inquire. If you have it, and will show it, let's move along, indeed.

BEA had their Pirate's opportunity to exonerate the airframer, which of course is not the case, in fact, or in evidence. They proclaim no new issues.

Why? Because they have not been announced? Have you taken a page from their book?

I submit that the beliefs of those who wish the a/c to be found innocent are spitting into the wind.

Until sufficient evidence exists from its source, let's just discuss?

I will repeat, you waste time in rejecting my opinions out of hand.

No evidence of computer aided PITCH UP? Not exactly a sound defense, at this stage.

I asked an honest question earlier, with a possibility that has not been addressed in detail....You overlooked it?

DozyWannabe

You use phrases like that and accuse *me* of bias?

You also mentioned "joining the dots" earlier, implying an effort to "exonerate" Airbus (even though, as I have said before they will have to carry at least some of the responsibility for the pitot problems), yet you provide no proof.

If there was an overspeed warning then the BEA would have mentioned it - there is absolutely no reason for them not to.

Sometimes a windmill is just a windmill.