Lyman

Hello AZR

quote

As we know that no overspeed condition was recorded on flight AF447, trying to explain the events as the consequence of an overspeed is null and void.

Well. Yes, we have been told no overspeed condition was recorded. I am not declaring there was an overspeed, and am merely musing, as you see: "IF in zoom, etc.."

I do not have the authority to declare, or conclude, it is not my job. I am writing here by the permission of the website's owners, and am grateful for the opportunity. Neither is it my intention to insult, or libel.....

As I write, I need to amend my remarks. I am unsure whether we have been told anything relative to a potential overspeed. I think assumptions have been made due to the absence of reportage of an a/c commanded ascent; My sense is that BEA considers their work incomplete, and the final will not even exclude some possibilities, remote though they may be. By its nature, exclusion of some admittedly quite unlikely occurrences cannot be used to conclude anything, perhaps ever...

Instead of taking offense, or personalizing the posts, one could comment on the possibilities of certain potentials, without judgment, or "agreement/disagreement", since as I say, they are conjectural.

A while back, I discussed methods relative to investigations in which I had been involved. Essentially adversarial in nature, they are not intended to offend, only to energize, and get the neurons firing. This thread is in no way regarding reputation or standing, though you may think different. At least for me. I will step on an A330 tomorrow, or any day, it is safe far beyond my parameters, or any rational person's. This accident was of course an amazingly rare event, as it is expected to be.

AlphaZuluRomeo

Hello Lyman
Well, you may be sure. We have been "told" or, even better, "shown". Please take a look at the annex 4 of the 3rd interim report. One of the parameters diagrams shows the overspeed with 0=no, 1=overspeed. The value sticks to 0.

Lyman

Hi again. Thanks for the reply. Yes, stuck on "O".

Still, I think worthy of a discussion, since we read of overspeeds rarely, and it is not our job to approve/disapprove of data written into the public record...

See, here I read those who have taken BEA data/releases, etc. and extrapolated, into areas of conclusive and passionate affirmations that not even BEA ventures. Partisans. I have no such allegiance, and if any of what I express serves to help keep the discussion going, I say fine, BEA need no defense/help, etc. The conclusions I read here belong to those who remain anonymous, along with the bile, anger, and humor, since BEA have none, as yet. Am I biased? Completely.

But then, I said so, and have repeated it often enough to earn immunity from further claims of that which I admit! My bias also reposes in the direction of what is true, in the end. Who cares, really, if Lyman doubts anything? Or believes it?

wozzo

Translation: I'm not necessarily saying that the people at BEA beat their wives, molest little kids and falsify data to protect Airbus, I'm just asking questions!

AlphaZuluRomeo

Well, Sir I'm sorry but I don't. Not here.

Why? Because the present topic is about AF447. Not about what might have been AF447, if this or that (that we know did not occurred) had happened.

I'm not that interested in "what ifs" based on conditions/events (namely: overspeed) that we know did not occur. Moreover, I fear that pursuing such "what ifs" is misleading.

Now, that's nothing more than my personal opinion. But I stick to it as the 0 value stick to the overspeed curve of the graph. Rest assured this has nothing to do with you, in the personal sense.

CONF iture

That’s it : Thrust provides performance.
But in no way thrust prevents stall, only AoA does. Keep idle stop increasing AoA and you won’t stall. Now, if you want to keep your altitude, you need performance, thrust provides performance.

Negative. Alpha Floor engages to provide performance.
Alpha Floor has no relation with stalling or preventing stalling, only Alpha Prot has.
Keep idle, forget about Alpha Floor, you will not stall.

CONF iture

There are all valid questions.

We have been told ALT2 was latched, but still the FD bars reappeared so 2 ADRs ceased their disagreement.
It is said that high speed stability is lost in case of ADR DISAGREE but is it back when ADR agree again ?
It is said that low speed stability is lost in case of ADR DISAGREE but is it back when ADR agree again ?

AP is also lost if ADR DISAGREE, could it have been reengaged to also follow the reappearing FD bars ?

Sorry, it’s all too complex for me … and I am fully qualified on the bird.

AlphaZuluRomeo

CONF iture, you really don't need to explain (to me) stall, thank you.

Yes, Alpha Floor is no more than an automatic TOGA. TOGA is thrust, thrust provide performance.
Now, why do Alpha Floor engages itself before the aircraft's speed decrease to Vs1g? The purpose is then to give max performance before the aircraft actually stalls on its current path.

To prevent stall or to exit stall, you need less AoA. We're all aware of that.

There is two way to obtain less AoA :
1/ lower the nose, less pitch will give you less AoA.
2/ if you can't lower the nose (at all / enough) (typical situation is you must maintain a path to avoid ground or another aircraft below), and provided you're not already stalled, increase speed. Same path, more speed = less AoA.

Why do you refuse to see the relation thrust/AoA in my point 2?

I agree that one can not use this relationship to recover from a stall, but it can be used to prevent a stall if the trajectory is constrained.

I agree that the point 2 was far too much advertised in old procedures, now revised. I agree that you can't expect to survive all (approach to) stall conditions if you know only about that point 2.

Now, if you know both points, you may find the second one useful sometimes.

Lonewolf_50

@ Clandestino:
http://flightsafety.org/ap/ap_nov89.pdf
Been a while since I read that, thanks for the walk down memory lane. The principles apply to Air France management as well as to any commander.
You left out a critical piece of the answer to that question: they practice flying so they know how the aircraft works, in all modes. That means you need to experience what happens near the edges, in a controlled environment. Doing is a critical part of training!
Actual flying is a skill susceptible to rust, unless all you do is monitor what the robot does for you.
Amen, deacon.
Good point.

CONF iture

Because you just keep mentioning performance :

• maintaining a path
• avoiding ground
• avoiding aircraf
• trajectory constraint
• current path

Alpha Floor just does that - provide performance.

Avoiding the stall is NOT the role of Alpha Floor but Alpha Prot.

Stop mixing both and giving Alpha Floor a role that’s not his.

DozyWannabe

CONF iture:

While it's understood that Alpha Prot and Alpha Floor are different (I've gone into considerable detail on this twice, which you haven't acknowledged), the fact is that one of the functions of Alpha Floor is to provide extra thrust to increase or maintain airspeed when a pilot is commanding nose-up pitch and bleeding off speed to the extent that the boundary of approach to stall will be crossed if the attitude and thrust settings are maintained.

Alpha Prot cannot command a change in attitude if the pilot is holding the stick back - it can maintain the attitude at the maximum safe level as the speed bleeds off, but unless the pilot lowers the nose eventually Alpha Floor will activate, increasing thrust which in turn increases airspeed, changes the current coefficient of lift and lowers the AoA.

Alpha Floor may not be designed as a stall-avoidance feature, but it is nevertheless an intentional side-effect when it is active.

Coefficient of Lift graphic from : The Lift Coefficient



Generic Coefficient Of Lift vs. AoA graph from : How Airplanes Fly: A Physical Description of Lift

NB : This is a generic illustration, intended to show only that a correlation exists between Coefficient of Lift and AoA - beyond that it is not pertinent to this specific case.

AlphaZuluRomeo

CONF iture: Yeah, OK, I see your point. You're right, sorry about that.

Allow me to change my words into :
Alpha Floor (TOGA) can not be used to recover from a stall, but it may be useful to try to prevent:
- on an unprotected aircraft(*): a situation where the crew will only have two (losing) choices: maintening a no-stall path leading to a crash too if the trajectory is constrained, or pulling the aircraft into a low-altitude stall (leading to a crash too)
- or, on a protected aircraft, the activation of Alpha Max/High AoA protection, leading to a crash if the trajectory is constrained (e.g. Habsheim)

(*) NB: academic case, at last re: Airbii

[Rem: DW just wrote that better than me]

Now =>

Owain Glyndwr

Dozy,

I am usually with you, at least in spirit, in your tussle with Confiture, but I really don't think you need to instruct him on what constitutes a stall

It might not be so bad if it were not for the fact that your illustration isn't actually representative of a modern airliner stall except at low Mach numbers and with flaps and slats deflected. For most of the speed range the lift coefficient doesn't fall back as suggested by that sketch - it staggers along at a more or less constant level without any particular pitching moment change to induce control loss either. What happens is that the separation spreads over more and more of the wing, producing heavier and heavier buffet until the test pilot calls a halt and says that's enough - we'll declare that to be Vs1g.

Whilst that is going on the drag is increasing rapidly and if there is no compensating thrust increase the aircraft is going to get into a severe low energy problem very rapidly.

Alpha floor was intended to make such a situation less likely - OK, if you want to call it a performance issue I won't disagree, but in my book it is more for maintaining margins than increasing speed and reducing AoA. AoA control is the province of the stick movement,

Would it be useful to consider Alpha floor as precautionary and Alpha max as prevention, rather than lumping both in AI's "High AoA protection"?

Lyman

I'd have thought by now you'd have noodled out that CONF iture flies the Bus, and has piloting skills far in excess of any "credentials" necessary for this foxhole, Dozy.

You are infringing on my territory. Speaking outside one's skill/knowledge domain, that is...

best, lyman

Oh fwiw: Alpha Prot: resignation. Alpha Floor: desperation.

Thrust/Altitude, Attitude/Airspeed. Or so I was taught.

UAS? PAUSE, ASSESS,THRUST,INSTRUMENTS,ENERGY,NOSE,CLIMB,ENERGY

PATIENCE......

gums

Salute!

Thank you , OG, for the observation that the classical stall depiction and AoA is more applicable to general aviation planes and very old straight-wing designs.

The top of the curve for just about any bent-wing plane is less severe. There's no abrupt loss of lift, and additions such as leading edge flaps, slats, slots, etc. can make the stall onset very delayed. With enough thrust, you can even remain controllable and flying well beyond the maximum AoA versus lift point on the curve. Wing loading and aspect ratio also contribute to "flyability".

The high-aspect wings increase drag faster than loss of lift. You can also see this with a plane that has great directional authority and a high wing-loading. So I flew one model that had classical AoA versus lift curves, but you could get that sucker up to 40 or 50 degrees AoA and sink like a rock without getting into a spin. The rudder remained very effective. To recover all you had to do was lower the nose. Hmmmm......

On the delta I flew ( like the Concorde and such designs), we did not have a sharp break in lift versus AoA, but drag increased dramatically. The Viper had similar characteristics due to the leading edge flaps and using the flaperons to shape the camber of the wing. So our alpha limit was more due to the mission requirements - maintain energy and maximize lift versus drag. The Hornet was not designed like that.

I must add that our FBW laws had many sensor inputs that kept the pointy end forward. Rate gyro inputs, acceleration inputs in all three axis, AoA until the three probes froze up, then things were based on last value or ignored and body rates became prime. Attitude was not used in any fashion, unlike the 'bus laws that limit bank angle and pitch attitude according to the mode and sub-mode of the system.

out,

DozyWannabe

I wasn't (and would never presume to do so), but more on that shortly...

I did say it was a generic example from basic classroom material! If not being representative of a jet transport is a problem I'll remove it.

What I suspect CONF iture misses is that while the calculation of AoA is not directly affected by thrust and/or airspeed, airspeed (and Mach at high speed) *does* figure in the Coefficient of Lift calculation from which AoA is determined (said calculation is what I'm guessing he means by "performance"). The presence of the graph was simply to show that a correlation exists - not the correlation of a specific type of aircraft.

So to summarise - I certainly wouldn't presume to tell CONF iture (or any other line pilot) of the correlation between AoA and Stall - but the relationship between AoA and Coefficient of Lift (and by extension, the variables involved in the calculation of Cl) shows an indirect relationship between Thrust, Velocity (i.e Airspeed/Mach) and AoA does indeed exist.

Fair enough, but if the pitch attitude cannot be reduced for whatever reason and the aircraft approaches a low-energy state (precipitating approach to stall), the only way to maintain a sufficient Coefficient of Lift and therefore AoA to keep the aircraft flying is to increase speed, which is done with thrust. Alpha Prot will limit pitch angle and ensure the aircraft won't stall, Alpha Floor will activate to keep the aircraft from bleeding energy until it hits the ground. I've understood it that way for quite some time.

Personally I'm not bothered by the semantics too greatly, as it is largely subjective. Coming at it from an engineering and systems perspective I see Alpha Floor as a function of autothrust and Alpha Prot as a function of the flight control logic - it just so happens that there are scenarios in which the two will interoperate.

(Which incidentally dovetails with at least some of what CONF iture is saying - he's right in that if you disable autothrust (and by extension Alpha Floor) but remain in Normal Law, the Alpha Prot will prevent the aircraft from stalling, but the aircraft will gradually lose energy until you hit the ground. Funnily enough I've never said that wasn't the case...)

As AZR said, there's a degree of quibbling over semantics going on here that to my mind adds little to the discussion and merely generates unnecessary friction.

Owain Glyndwr

Well I wouldn't be so sure of that - such a relationship is fundamental airmanship I think.

But anyway, I think you have it a bit wrong - one doesn't calculate AoA from CL but CL from AoA. CL doesn't figure directly in assessment of whether the aircraft is near stall or not - that is a simple function of AoA and Mach Number. If AoA is measured (as it is) then at low Mach that is enough to determine stall margin. When at higher Mach the boundary needs to be adjusted (which is why a simple AoA gauge cannot function as a stall warning when airspeed information is lost), but even then it would be possible to work with a default value which, even if it did not prevent "stall", would serve to avoid the sort of shambles that we see in AF447.

Assuming that is, there is enough thrust to maintain level flight and have something left over - which is not always the case.

DozyWannabe

I'm not sure of that, but to state that thrust and AoA have no relationship is simplifying to the point of inaccuracy.

I mangled my words a bit there - apologies (I think my brain needs a cooling fan). Strictly speaking AoA and Cl are reciprocal functions of each other - i.e. a change in one will result in a change in the other. I feared going into those kind of terms would cause folk to start dozing off...

Agreed - but because of the relationship between AoA and Cl and the V(elocity) component of the Cl calculation there is an indirect relationship between airspeed, thrust and AoA - that's all I've been saying.

But there we get into the quasi-religious aspect of pilot authority. While it would be possible to do such a thing, those who consider the Airbus FBW and modern FMS as a step too far* would likely kick up a hell of a stink.

Well yes - if you're already at maximum thrust and you can't decrease your pitch attitude then you're in a bad way!

[* - For all people say I defend Airbus and automation, I actually think it's gone about as far as it should with the current level of technology. Technically it would be possible to go further (after all, it's been 23 years since the A320 was launched - even longer for the B757 and B767), but given the fact that business practice seems dead set on abusing the current level of automation to enact cost-cutting measures that were never intended it would be foolhardy to do so. ]

gums

Salute!

I am gonna throw my lot in with OG.

When I talk about reversion, my only experience was with that little jet.

When speed went tango uniform, the system reverted to "STBY Gains", which was dynamic pressure, and not total pressure. So with gear up we had one value and gear down another. The values were used for gains and rate limiting and other stuff. AoA was still used unless only one sensor was still operable.

We also had two more static pressure sensors independent of the basic pitot-static system. Not sure if they came into play for mach calculations, but they were there. Those suckers also were used for sideslip functions and to check the basic pitot tube doofer.

My basic complaint is that too many "protections" and such are supposedly there amongst all the reversion modes. I do not advocate going directly to "Direct law".

What I advocate is a very straighforward reversion that provides AoA and Gee limiting. In other words, worry about stalls and ripping the wings off. Screw the bank angle and pitch attitude and such. Screw the overspeed. If dynamic pressure is lost, then use some "stby gain" values for many FBW calculations. Make more use of the body rates. Let the pilot know he is definitely in a degraded mode and all that it can/ can not do. Forget the auto-thottle stuff!!! No "TOGA" button! The levers become basic thrust controls and stay where they were when the malfunction occurred. The trim function of the THS becomes "manual" and starts where the position was when the malfunction happened.

I will bet that I can come up with a test for the 'bus pilots that includes the 24+ footnotes that may or may not apply to every reversion mode. I would bet a bottle of good scotch that 90% of the pilots would fail.

I don't buy the "system worked exactly as designed" argument. Lost two friends when the thing worked exactly like it was designed. And you know what? We changed the system!!! Sure, we improved our training and such, but we could see the problem, and the government and the company agreed that we needed to change/improve the design.

sorry to get emotional here....

PJ2

Hello Dozy;
I might be able to provide some input here, if only obliquely to your point.

I support the views that Owain is offering; - your example isn't representative of a jet transport and isn't even generic, basic classroom material for airline pilots.

Machinbird began a very helpful diversion in the thread by opening the discussion about stall AoA's.

You may recall that in the second AF447 thread, entitled, "AF447 - Search to Resume" (you can use mm43's excellent search tool, here), Machinbird HN39, Chris Scott, mm43, ChristiaanJ, fdr, Gretchenfrage, CONF iture and others had a lengthy discussion on the stall, AoA's, FPA's/Pitch and so on. I think that discussion is well worth reviewing.

In response to Machinbird, I introduced what I thought was "standard, basic knowledge" of stall AoAs as presented in Davies' Handling the Big Jets, I discovered that my understanding of stall AoAs at high Mach Numbers and at our typical cruise altitudes was incorrect, I offered that they were around "14 - 18deg" which was what all the graphics in Davies' book indicated. They were correct, but were for the approach phase where high-lift devices were in use. AoA's as low as "5deg" had never occurred to me, and I can tell you, this was never taught or discussed in any groundschool or recurrent simulator training I had ever received - another issue.

So a very helpful discussion and exchange with HN39 began and it changed my understanding of high altitude, swept-wing, high-Mach Number flight in transport category aircraft.

It was, I think, a re-interpretation of what we might understand in old-fashioned terms as "coffin-corner", (which, for practical reasons, does not exist for the A330).

I hope this pointer to some of our previous discussions before the crash site was discovered, is both interesting and useful. I found it refreshing to re-read many of the thoughtful contributions of the time. I think it is a good thing to do so as it lends an ongoing check on one's thoughts, as well as providing some perspective to current discussions.

Happy reading!

Cheers,

PJ2