AF447

[singpilot]

The Aeroperu 603 crash had something AF447 doesn't (yet).

They had recovered a piece of the fuselage that had speed tape covering the static ports, done to protect them during an aircraft wash. That was THE smoking gun. Case closed.

The crews' actions (and the planes' responses to crew input) were entirely consistant with the blocked static ports. The possibility of a triple redundant failure (all 3 static systems), was unheard of, and was never trained for, or even discussed. But it happened. If it had been during daylight hours, perhaps the tape would have been spotted in the walkaround, or a horizon would have been seen and used by the crew out the window.

[Razoray]

Thanks....singpilot.......

And the question is what happened this time...we dont know, but we can assume (or, I can) that it was something that was never supposed to happen. Thats why it is such a mystery!

But what is the smoking gun?
What happened that was not planed for, is it the unexpected that needs to be looked at?

[singpilot]

Well, Razor....

We don't have a 'smoking gun' yet.

[PJ2]

Jeff;

there will be multiple crews of similar composition & experience formed and tested in simulators under varied hypothesis.

I am sure you will know this intuitively but it requires stating. A simulator is precisely that - an electronic, software-driven "look-alike" with wonderful fidelity to the aircraft under normal and known abnormal conditions. It is programmed not through "sampled" data but using software that mimics, very well, the aircraft's cockpit arrangements. The aerodynamics of the aircraft are extremely well done and within expected flight regimes, I should think reliably mimic the aircraft.

However, the simulator's very best qualities are also its greatest fault. Its very veracity can mislead an experimenter into believing that "this is how the airplane would respond". These are very complex machines and we simply cannot know, in my view, that what the simulator does in all circumstances, is what the aircraft would do.

So, programming failed pitots and permitting the speed to bleed sufficiently to enter an initial stall, as one experiment, with very heavy gusts from either side or from below, possibly with hail, (programmed noise), may or may not yield interesting answers in terms of aircraft stall behaviour under these specific circumstances.

I am emphasizing limitations here on purpose - these are very good machines, and I suspect Airbus's has special algorithms for taking the airframe beyond expected airline operational limits. Still....it is not an airplane and the storms and the radar are not real. Just a caution, that's all.

[Hyperveloce]

Ground Simulators

PJ2
Aren't there several kinds of ground simulators: those used by pilots for their regular training, and some Airbus industrial simulators ? for design & research, with a finer modelling (possibly also accurately describing the aerodynamics outside the flight enveloppe) and maybe a greater flexibility to set up some specific scenarii ?
A clue about how the plane departed from its high altitude cruise would be a first good step ? (even if we don't how it transited from H.A. to surface.)
Razoray
This Aeroperu had lost both altitude & airspeeds, would rather look to birgenair (six month earlier I think), but it did not occur suddenly and at high altitude.
Did you read the Air Caraibe safety report or the Air France Paris-Antananarivo flight report ?
Jeff

[Art-Deco]

ACARS interruption & rain attenuation
Art Deco, what are the rain attenuation values you have used for your estimation ? Do you make use of a cloud model (a spatial distribution of water, ice,...) ?
Jeff

Sorry for a late answer:
Re: my post 3397 on page 170 regarding ACCARS attenuation.
My calculation was approx 1,8 dB per 1000m (1km) at given elevations for L-band freq. The book you referre to is Ka-Ku link budgets that are in a much higher band 10-25Ghz, so the attenuation then is much higher.

link: http://www.pprune.org/rumours-news/3...ml#post5052419

[Robert Campbell]

Simulators

Aren't there several kinds of ground simulators: those used by pilots for their regular training, and some Airbus industrial simulators ? for design & research, with a finer modelling (possibly also accurately describing the aerodynamics outside the flight enveloppe) and maybe a greater flexibility to set up some specific scenarii ?

How could an accurate simulation be created of unknown situations unless the AB test pilots intentionally put the aircraft into those situations, which, in my opinion, might lead to vacancies in the test pilot ranks?

I think it might be prudent to just put notices in the flight manuals saying, "Unknown territory, don't do this!"

Perhaps some results of wind tunnel testing would be informative.

[Tree]

Has the 777 been grounded ? They fall out of the sky with both engines rolled back, remember (and we still don't have final report on why - in fact the silence is...). Well, one 777 fell out of the sky anyway - do you have your "feeling" another one is "on the way", or does that feeling just apply to one particular aircraft manufacturer ?

This should answer your query as of 13 July 09:

http://ad.easa.europa.eu/blob/easa_a...AD_2009-0142_1

EASA Airworthiness Directives Publishing Tool

[PJ2]

How could an accurate simulation be created of unknown situations unless the AB test pilots intentionally put the aircraft into those situations, which, in my opinion, might lead to vacancies in the test pilot ranks?

Well I tried to convey this notion - you can't just go into a "fancier" simulator and do better - it's electronics and software, all the way down, not a molecule of air to be had... It's good, but it's not an airplane.

Wind-tunnels work well I am informed but have their limitations as well, as I am similarly informed. In short, there are no magic bullets in this kind of research/experimentation, just hard work slugging through a lot of data, with few true insights.

In pondering this over the past few days while staying away and discussing this with others we were wondering if an initial stall entry, perhaps through vertical/lateral gusts sufficient to take one wing further into the stall than the other, leading to a developing partially-stalled spiral - some forward motion, very high vertical motion, slight nose down, possibly hobbled by reduced/no hydraulics due to high incidence of airlfow over the inlets and the RAT, but really, who knows? We know it was at 350 and that later it hit the ocean. The rest of the conjecture has filled the thread.

[Smilin_Ed]

There Are No Old, Bold Test Pilots

How could an accurate simulation be created of unknown situations unless the AB test pilots intentionally put the aircraft into those situations, which, in my opinion, might lead to vacancies in the test pilot ranks?

Aircraft design engineers provide their best estimates of the limits (maximum airspeed, stall speeds in various configurations, bank angles, g-loading, etc.). The manufacturer's test pilots, who are usually also have engineering degrees, then approach these limits slowly, being alert for bothersome (squirrely) behavior. Then they make a judgment as to how close the typical line or fleet pilot should be allowed to come to the limit. How close varies with the type of aircraft. Then the user's (airline/military) own test pilots verify the manufacturer's proposed limits and decide what additional limitations, if any, should be imposed on line or fleet pilots. Aerodynamically precise simulators are designed by extrapolating the experiences of test pilots so, of course, they can't be perfect and can only be expected to predict aircraft behavior within certain limits. In the past 50 years, very few test pilots have bought the farm. That only happens when they rush the program.

[HarryMann]

How could an accurate simulation be created of unknown situations unless the AB test pilots intentionally put the aircraft into those situations, which, in my opinion, might lead to vacancies in the test pilot ranks?

If you're talking aerodynamically, then of course, behaviour outside the normal flight enveleope can be simulated. All the data e.g. mass/inertial distributions, lift, drag pitching moment of indiviudal components and the whole integrated a/c as well as all the pertinent stability derivatives, thrust and even aerolastic effects - would be available from initial CFD work, w/t refinement/cross checks and corrected within the flight tested envelope, a fair bit wider than certified limits.
The predicted beahviour could be simulated well past any flight test/certification flight limitations

There wouldn't be too many what ifs behaviour-wise, the whats are what the actual character and level of disturbance was, what were the auto and/or crew responses/inputs.

[PJ2]

HarryMann, Smilin'Ed, very helpful thank you. So the roll I did in the 320 sim with all the FACs and SECs off was what it woulda been like...good to know -

[HarryMann]

In the past 50 years, very few test pilots have bought the farm. That only happens when they rush the program.

Very much due to the accuracy of CFD/Wind tunnel and other methods of quantifying whole a/c behaviour up to and beyond any likely flight conditions, test or operationally. The high Mach upper certification limits might possibly produce the most uncertain (& disturbing) behaviour, but still pretty well mapped out beforehand.(think the A380 was about M.92 ?)
Not taking anything away from test pilots, but of course the job must be by nature very different to say the early 50's when much less was known, and CFD was just a dream!

[JD-EE]

singpilot

Indeed we don't have a smoking gun. I just wish some of those bent and twisted bows and arrows strewn around here could be definitively debunked and tossed aside.

JD-EE

[Mad (Flt) Scientist]

Originally Posted by PJ2

HarryMann, Smilin'Ed, very helpful thank you. So the roll I did in the 320 sim with all the FACs and SECs off was what it woulda been like...good to know -

Not necessarily "like"....

BUT had AB's aerosim guys been stood in the back of the cab, and told you that they had taken data gathered during the exceedences part of the cert programme and used it to SUBSTANTIALLY extend the sim matching beyond the VERY rudimentary requirements of Level D (or whatever we're supposed to call the top end training devices now) then maybe it would be "like". They'd certainly be able to review the manoeuvre post-session and tell you where you were in the interpolated aero data envelope, where it was extrapolated, and where it was likely nonsense.

The Level D QTG basic requirements are very thin indeed. Most OEMs will go a bit further, because there are qualitiative tests to pass too. But there's a LOT more goes into validating the math models for specific engineering purposes than the basic sim models, and not all of that necessarily gets into the production sim.

[Mad (Flt) Scientist]

Originally Posted by HarryMann

Very much due to the accuracy of CFD/Wind tunnel and other methods of quantifying whole a/c behaviour up to and beyond any likely flight conditions, test or operationally. The high Mach upper certification limits might possibly produce the most uncertain (& disturbing) behaviour, but still pretty well mapped out beforehand.(think the A380 was about M.92 ?)
Not taking anything away from test pilots, but of course the job must be by nature very different to say the early 50's when much less was known, and CFD was just a dream!

I'd attribute the somewhat safer flight test environment to changing attitudes towards test safety as much as anything else, rather than to any great whizz-bang technology. There's been more than enough flight test casualties in the last generation, in any case. (And if you factor in the very much reduced number of development programmes these days, and that much of the development that there is is systems driven, rather than flight envelope stretching ...)

[PJ2]

MS,

The Level D QTG basic requirements are very thin indeed. Most OEMs will go a bit further, because there are qualitiative tests to pass too. But there's a LOT more goes into validating the math models for specific engineering purposes than the basic sim models, and not all of that necessarily gets into the production sim.

For my money that is really valuable knowledge, thank you. So modeling is far more advanced in the specialized sims than the production D models, (the ones the airlines have to certify checkouts for new crews on new aircraft (to them) and to renew IFR and PPC qualifications. I've learned something new and I thank you all. Jeff, I stand corrected.

They'd certainly be able to review the manoeuvre post-session and tell you where you were in the interpolated aero data envelope, where it was extrapolated, and where it was likely nonsense.

That would be absolutely fascinating to see. It wasn't that difficult a maneuver - 250kts, a lot of push at the top and a gentle pull-through to avoid alpha-prot. I always wondered if it was "as if".

Given the chances of finding the recorders, (I still think they'll find them, and the main wreckage), and this capability, almost certainly, a number of scenarios have been run at Toulouse then.

[singpilot]

From JD-EE

Indeed we don't have a smoking gun. I just wish some of those bent and twisted bows and arrows strewn around here could be definitively debunked and tossed aside.

I watch everything going by in here, a lot of it repetitious, but some of it very insightful. I have learned a lot of the 'why's' of the interactions between hardware that, while completely explained in school, training and sims, falls short of the detailed failure analysis in here. After verifiying some of what I've seen in here, I am much better at these hardware and software interactions.

The bent arrows, while plentiful, can usually be examined by source and motive, and not too much expended effort to dismiss.

I have spent a great deal of time trying to put myself there (as I was last summer at this time) and trying figure out how this magnificent airplane and crew/pax got so far afield. I think back to times approaching the ITCZ when distractions would arise, be dealt with and prepare for the hour or two of real, 100% work to get thru the band of weather. My instructions were that meal service was to end prior (giving the cabin crew 4-ish hours from T/O) to crossing the wx, and that we were not to be disturbed until we signalled that we were clear. I was always up front when I was junior, and learned that lesson when I got senior.. to be there as well. There would be plenty of time on the other side to rest, eat and deal with Dakar. The same for the reverse (southbound) leg.

Some of the F/O's I flew with took this area as seriously as I did, and some did not. I know I'll read/hear about the ones that did not someday.

We are all truly hoping for a miracle of some kind, an answer with a smoking gun. If I was asked what could be my 'perfect solution/resolution'...

The CVR.

I'd want to know what was going on in that cockpit prior to anything going wrong, and what it was that they encountered. We all know we will never hear the actual sounds and voices, but only a transcript. I firmly believe the answer is there.

P.S. I wanted to add that this (or anything I've posted prior) is not necessarily an endictment of the flight crew per se. I think they encountered something beyond what they were expecting, and discovered, too late, how best to deal with it. That IS a wide open statement that covers what we know so far.

[TIMA9X]

I agree with singpilot, what has been said on this thread, some of it great and some not so great and some of it so confusing veteran pilots may be thinking "enough is enough!"

At the end of the day we simply don't have enough evidence to come to any conclusion no matter how much we use our bank of combined knowledge.

Ar-men!

[SPA83]

NTSB safety report

Safety Report on the Treatment of Safety-Critical Systems in Transport Airplanes

http://www.ntsb.gov/publictn/2006/SR0602.pdf