AF447

[Duck Rogers]

Not from where I'm sitting there isn't.

[vovachan]

From my days at the flying school and physics lectures, i seem to remember that these upsets are impossible to retrieve from. All T-Tail types are affected, 1-11, MD80 series etc....

I wonder what the "raw" stall characteristics of an A330 are, sans electronic protections, and whether the manufacturer could have taken some shortcuts in that area, since the plane is "unstallable" anyway...

[PEHowland]

TiberiusKirk said:

While it is indeed possible to extend radar coverage to every part of the globe...

Not with ground/land based radar it isn't. Radar is line of sight and the earth is curved. Even at altitude, an aircraft will often be below the horizon WRT to the nearest land.

Not true. Radar at microwave frequencies (1GHz+) is line of sight limited, but at HF (3-30MHz) propagation is either via refraction in the ionosphere to ranges between 500 and 3500 km (HF Over the Horizon Radar) or typically 0-500km when vertically polarised and over the sea (HF Surface Wave Radar). Thus it is indeed possible to extend coverage to every part of the globe, just not with the typical L-Band or S-Band ground based radars that you are probably familiar with. Several countries have operated or continue to operate HF OTHR systems, although none (as far as I am aware) are providing surveillance over the region of this crash.

Later edit: JDDE you replied to the above message disputing HF Surface Wave Radar and telling me to learn more about antennas and propagation. I replied to your reply, but this was deleted by the mods. Suffice it to say that I have been a radar engineer for almost 20 years and worked extensively on both of the above technologies. Your comments on NVIS are nothing to do with the technologies I am referring to.

[AstraMike]

Flat Spin

At first I discounted the possibility, but I now have to admit it is beginning to fit, given what has been explained by the investigation.

I cannot help but wonder how many of the Airline Transport pilots here have ever encountered a Flat Spin and, in fact, how may might know anything about recovering an aircraft that just happened to enter such a spin, by mistake. Three dimentional filght (as Duane Cole would put it) is not something that is practiced in airliners, is it?

I dont know, but I doubt anyone has any data regarding the Airbus 330 in any kind of spin; something tells me such out of control excursions are considered a thing of the past, but to return to the point, a flat spin has a descent rate approximately 50% of a conventional spin. Thus, from experience, would not be surprised if the descent rate for the aircraft in question was around 4,000 fpm. Isn't that about 45 mph? I admit it could be more, but depending on many aerodynamic factors it could equally be less. I dont know because I have not deliberately spun an A330, nor have i heard of anyone who has, certainly flying through an area of active CB buildup could be considered a good place to start, especially if you have no reference to speed and found yourself distracted by warnings and without the autopilot so many depend upon these days.

However, impact at a speed in the region of 45mph might just fit with the damage evidence recovered to date, what else does? I submit a full spin descendibg at around 8000 fpm would be too fast to produce the evidence found, a spiral dive even worse and a low level stall just does not seem to explain the apparent torque involved in the removal of the vertical stabilizer, whereas a flat spin, just might. It just might answer other puzzels too

Equally, I suspect no one really knows the attitude of an A330 in a "flat" spin and you should not assume "flat spin" equalls flat or horizontal attitude; it is quite possible that the nose could be above the "horizon" and if that is so it is equally possible that the aft fuselage may contact the sea first...

Sorry, but I couldn't help myself...

[Will Fraser]

The fuselage impacting 'flat' does not preclude any attitude up to flat impact, including nose up, as the 'tail' may have contacted first, destroying the tailcone and VS mounting hoops, releasing the fin from its forward mount. If there was rotational (radial) movement (not roll), that would explain the 'left' torsional component of the failure. Dragging the tail would slow horizontal velocity, derotate, and allow for a 'flat' aspect at contact (a la Schipol).

Having said that, irrespective of its possibility in fact, it is irrelevant.

The trail still needs unwinding to the moment of loss of control.

[PJ2]

IcarusRising, post 2889; http://www.pprune.org/rumours-news/3...ml#post5038394;

It is not uncommon for me to return from rest to find laptops out on the pull tray obscuring flight instruments or sorry animated discussions about work and pay conditions while the aircraft does her own thing an receiving looks of disbelief when voice my concern that monitoring aircraft flight performance and weather conditions mean just that and not getting sucked into complacency and side distractions. I am looked on by some crews as a relic and behind the times !

Yep. I occasionally found the same scene upon return and had the same "relic" feeling. With reduced standards in terms of experience and training, I suspect we'll see much more of the behaviour you are describing, (casual monitoring, laptops out, discussing everything under the sun, etc)

ralphie;

I am sorry to hear that you are retired. I would love to hear that Airbus or Boeing or whoever would grab you as a fantastic teaching authority. That is the same for all other retired captains on this forum. The airline industry should be using you big time!!!

There are a lot of pilots contributing here and who are in the industry with the same and far higher qualifications than I who are retired or retiring and who, because airlines don't pay well enough to attract the kinds of mentors the business needs, are for the most part lost to those coming along. I had a contract for specialized flight safety work for a year and was deemed "too expensive" so that ended that.

We shouldn't leave the impression that the new pilots are a different breed - they're not. They're as keen and as safe as they can possibly be. The key is, they have to have experience in aviation, not in schools which give out airline pilot licenses after 250hrs of simulator and a bit of line training. That's where the problem is. Most guys in my group had thousands of hours before being hired. We sat in the Second Officer's seat for a while if we were lower time, and we learned from the guys up front - the good and the bad. There are schools that can give an MCPL but there aren't mentoring programs, and mentoring only shares attitudes and stories; one must "do", to gain experience and quite frankly because of the way airlines have treated pilots as "expensive resources which need to be beaten down", fewer and fewer young people are choosing the career.

From Eurocockpit "RC analysis, posted by Squawk_ident, post 2897 http://www.pprune.org/rumours-news/3...ml#post5038676;

-> The descent of the aircraft involves a descent of the altitude of the cabin ordered by the main controller of pressurization (the other being functionally on standby). It is what occurs with each descent so that at the opening of the door, on the ground, the outside pressure is the same than the one inside the cabin. There is thus a regulation of the descent of the cabin altitude as the plane goes down. If the plane would start a very fast descent, the controller of pressurization would try to reduce very quickly the altitude of the cabin. This variation is by design limited to 750 ft/mn, which cannot thus generate an alarm corresponding to a value algebraically higher than 1800 ft/mn. If the plane was not in descent but unfortunately in a fall, it would then join the altitude of the cabin before the controller of pressurization could sufficiently reduce the cabin. In this case, the pressurization would be reversed: the pressure outside the plane would become higher than that of the cabin. To avoid the implosion, “safety valves” and “negative relief valve” open in the event of negative differential pressure (0,25 psi). In this case, the rate of variation of the cabin follows the fall of the plane to minimize the difference in pressure and the rate can thus be much higher than 1800 ft/mn.

This scenario assumes that the engines are running all the way to impact. I doubt very much whether they were and explain why in post #2921, http://www.pprune.org/rumours-news/3...ml#post5038998

Xeque, post 2902 http://www.pprune.org/rumours-news/3...ml#post5038792;

We have an object that appears to have fallen from the sky with normal, gravitational terminal velocity. We have no contact from anyone or anything inside it. There was no attempt (it appears) to prepare for a crash landing in the ocean.
The only explanation that makes sense to me is that they were all incapacitated at the moment of catastrophe whilst still in the cruise.
Now, what to infer from that? I have absolutely no idea and I would suspect that no-one else has either.

Contact would likely not be possible. The chaos in the cockpit due to warnings and probable turbulence would occupy the crew for the first minute or two; dealing with a rapidly developing emergency and prioritizing actions, handling a degrading aircraft uncertain of airspeed and probably dealing with the significant shock of what was happening would not permit communications. As dozens of professionals here have pointed out, we "aviate, navigate and communicate" even in such rapidly unfolding circumstances.

As described in the post above, with the possible/probable loss of electrical power after the stall had fully developed, only VHF1 would be available. VHF 1 is usually set to 121.5 when a flight is handed over to air-radio (who in turn, communicate with the ATC unit doing the actual controlling) and, with no CPDLC (google "FANS"), all communications are done via HF communications. There is no way in the circumstances this crew was in, that they would be broadcasting in the blind on HF - they could have broadcast on 121.5 which other aircraft would have heard, but didn't. We can only surmise why but I highly doubt that they or anyone in the cabin were incapacitated due to pressurization issues. Determining if the masks deployed might possibly be done through an examination of the pins holding the doors closed in the overhead panel where the masks are stored - that would indicate that the cabin reached at least, roughly, just over 14,000ft but the BEA said nothing about this. There are such panels in the wreckage.

Will Fraser;

The trail still needs unwinding to the moment of loss of control.

We have at least some clues, but no evidence.

vovachan;

I too was somewhat surprised that other ac were deviating quite considerably to avoid the weather while AF447 seems to have flown in a straight line until the last known position. They never requested a deviation. What is the protocol in that area where comms are flaky if you lose contact with ATC - do you deviate anyway and hope for the best? Stay the course damn the CBs?

Frankly, while none of us was there, I was astonished when I saw the satellite maps of the area, that some course deviation did not take place, and that impression has been subsequently supported by the actions of other flights in the area. Conditions don't change that much in half an hour. I don't think it is wise or fair to begin building a picture of the decision-making process but it is one factor among many that requires examination in my view; - 'nuff said for now.

As far as deviating around thunderstorms in areas of "flaky" communications, you do what you have to do - deviate as necessary to ensure the safety of the flight. You light the aircraft up with all landing lights etc, broadcast intentions on the common frequency, perhaps climb 300 to 500ft if you're really uncertain about traffic and continue to broadcast while deviating. CPDLC is such an improvement in flight safety that it should be mandated by ICAO for all countries which serve communications for international flights.
PJ2

[singpilot]

First off, THANK YOU DUCK!

This thread went seriously south at post #5.

They flew into a thunderstorm. The A330-203 was at max cruise altitude and at faster than normal mach (because of the high/heavy choice). Mach .82 and encountering moderate turbulence and most likely warming temps when they got a vertical gust of XX knots. Who knows, who cares. Iced pitots? Who knows, who cares, same result. The ADIRU's are programmed to 'fail' with a loss or gain of 60 knots, because that is thought to be impossible. They were entering an area of increasingly worse/impossible conditions for their planned performance prior to encountering/stumbling into a buildup.

The plane stalled. The engine/s probably flamed out/overtemped/burped because the inlet air also vectored off of normal due to the vertical gust. If only one was producing thrust, the plane may have even flatly spun, but it did so in a stalled condition. 208 metric tonnes falling from FL350 after a stall will look the same as a rock. They remained stalled all the way down. Vertical fin and pieces of a horizontal stab were found floating. How much tail remained on the way down? Who knows, not enough, obviously. None of this surprises me. The ACARS messages are nice to have, but fact is, they would be expected in this scenario.

I have flown this EXACT route several hundred times in 37 years, and no prudent aviator is going to ignore the weather in this region.

Sorry to be blunt, but anyone who has ever penetrated an active thunderstorm (for whatever reason) will do whatever it takes to never do it again.

Provided they survive their first encounter.

The Old/Bold rule.

Another small tidbit. Within an hour of the impact 7 other flights went within 100 miles of AF447 that night, and every night since, and not a single one of them has made CNN.

Sorry, I think Occams' Razor applies here.

[PJ2]

singpilot;

Sorry, I think Occams' Razor applies here.

Yep, and it doesn't paint a pretty picture, frankly.
So there's everything to learn, and nothing to learn...depends upon who's doing the learning. There is the pitot issue and the response, and the broader issues of cascading failures possibly overwhelming crews. I don't think there are structural or engineering issues here. Also, there is some very good thinking along the way to 3000 posts which have been of some benefit to others in our profession.

[Tailspin Turtle]

One Scenario

Based on my experience as an instrument and aerobatic instructor once upon a time, the easy way to replicate most of what I've recently read here is to lose attitude reference, enter a spiral dive, get going really fast (supersonic is not out of the question), finally figure out what level is, roll to level, and then pull as hard as possible because the altimeter is unwinding or you can see the water is nigh. The result is a high sink rate, wings level*, unsurvivable impact. *fuselage level too, perhaps, depending on the pitch rate of the attempted pullout versus the altitude remaining (added)

[L337]

Sorry to be blunt, but anyone who has ever penetrated an active thunderstorm (for whatever reason) will do whatever it takes to never do it again.

Amen to that.

The most frightened I have ever been in an aeroplane is in the middle of a thunderstorm, going down. Because I thought I could get over the top of it. I did not get over the top, and because of the turbulence and buffet boundaries had to descend into it.

Only once. Never again.

The satellite weather trace makes the hairs on the back of my neck crawl.

[BOAC]

Originally Posted by sing

The A330-203 was at max cruise altitude and at faster than normal mach (because of the high/heavy choice). Mach .82 and encountering moderate turbulence

- I appreciate you do not fly the 330, but you may wish to note that flying at .82 instead of .8 reduces the max (and opt) alt by 2000' and puts it well BELOW 35000 at anything warmer than ISA+17 at the 0210Z estimated weight. I suspect there may be a strong element of 'too high' here with the ITCZ to penetrate - I'm not sure, of course, what manoeuvre margin the graphs posted a while back allow for.

[Graybeard]

Thanks for the spin info, AstraMike.

If for some reason the cg were aft of the center of lift at time of stalling, the plane would tend to slide backwards, tail low, then level out as the tail gained lift, then do some other dance. I reckon it would be akin to a falling leaf in its maneuvers, rather than like a falling rock, which would indeed retard the descent.

As for stall and spin testing of the A330, I would be shocked if they didn't at least do heavy stalls, with a spin chute at the ready. I'm sure those test results are getting scrutiny right now.

GB

[sirgawain123]

loss of speed indication

i´m just a chopper flyer with no experience in airliners, but having read whole thread, there is a question i would like any AB330/340 driver to answer:

With such a heavy aircraft operating near coffins corner, in case of loose of Airspeed indication, Does the SOP procedure (as i recall reading here) of flying a slight nose up attitude and fixed power setting help? or..would it help inducing a stall?

Does SOP for loose of speed indication require a fixed AOA and Pwr setting, or...are those parameters dependant on weight,speed, and height?

[24victor]

ACARS Timing

From the English version of the prelim report;

The messages were at least five or six seconds apart, which can be explained by the limited rate of communication by satellite. There are two possible reasons for the longer gaps: either the aircraft did not have any messages to transmit, or it no longer had the means for doing so (loss of satellite communication performance, for example).

and

• the last message was transmitted to the aircraft at 2 h 14 min 28 s and was effectively received,
• the twenty-five messages transmitted by the aircraft were correctly received by the ground station,
• the gap observed between the message sent at 2 h 13 min 14 s and the one sent at 2 h 13 min 45 s is due, at least in part, to a temporary interruption in the communication link between the aircraft and the satellite,
• there were no satellite telephone communications during the flight.

It also clarifies that

17 The reception time given is that of the service provider’s server processor

The messages average six seconds between each one and this, as stated, reflects the timestamp applied at the SITA server, in other words the time received by the service provider, not the time generated.

If you look at the first fifteen messages you'll note they were all generated in the 0210 window, so the cascade was progressing very rapidly.

There are then four messages generated at 0211 and one, the NAV ADR DISAGREE, generated at 0212. Then come the biggies, the PRIM1 and SEC1 faults generated at 0213. There follows a maintenance status message on ADR2 and the CABIN VERTICAL SPEED advisory followed by an uplink message to the aircraft, probably an ACK of the last advisory, which was "effectively received".

The report noted that four transmit cycles were missed from
2h 13min 14s and attributes that, in part, to the aforementioned "temporary interruption" in the satellite link. No mention here of lack of messages to transmit. The most likely causes for this stated interruption (IMHO) would be temporary loss of power to the transmitter or, more likely, the SATCOM antenna being outside of its operating envelope. However, it was back and capable of receiving and transmitting an ACK up until at least 2h 14min 30s (as stated above).

Questions to the ACARS pundits;

1. Why the drop off in message traffic after the 0210 window
2. What is the interpretation of the 0211/0212 sequence
3. At what point would you surmise the upset may have occured.

Rgds.

24V

[EGMA]

Regrettably 'singpilot' is probably right ...

What we should learn from this, apart from the bleedin' obvious, is that the ACARS weather reporting system needs to become a little more proactive. The absence of messages told us something was amiss, but we didn't wake up to the fact for 6 hours!

Not good enough in this day and age ... it wouldn't have mattered to AF, but what if 100+ had survived a ditching?

It doesn't need more ATC, the infrastructure is there already.

[EternalNY1]

I have flown this EXACT route several hundred times in 37 years, and no prudent aviator is going to ignore the weather in this region.

Sorry to be blunt, but anyone who has ever penetrated an active thunderstorm (for whatever reason) will do whatever it takes to never do it again.

Provided they survive their first encounter.

The Old/Bold rule.

At this point, this would seem to sum the whole thing up nicely.

The radar returns showing CB tops above 50,000 feet along their route could easily put them in an impossible situation.

[fantom]

I have diverted around Atlantic CBs by more than 100nm - with or without ATC permission. They can see where I should have been/was anyway.

If the aircraft had been spinning at impact I do not think the fin would have separated as it did; there would have been tortional evidence.

Any tps/structural engineers to back up my opinion?

f

[Gringobr]

Watches

If any of the bodies had watches which stopped at the moment of impact, it might be interesting to know

[overthewing]

Gringobr, interesting thought, but I'm not sure if digital watches get stopped in quite the same way as the old analogue ones that were so useful in detective stories?

[JD-EE]

PEHowland commented:

Not true. Radar at microwave frequencies (1GHz+) is line of sight limited, but at HF (3-30MHz) propagation is either via refraction in the ionosphere to ranges between 500 and 3500 km (HF Over the Horizon Radar) or typically 0-500km when vertically polarised and over the sea (HF Surface Wave Radar).

To get an idea of the technology (and impact) of over the horizon radar look up the Russian Woodpecker on Wikipedia. There are some nice pictures and pointers to other nice pictures of these abandoned steel yards. The green you see far far below in the Wikipedia pictures are tall evergreens to give you an idea of the scale involved. ( File:Woodpecker array.jpg - Wikipedia, the free encyclopedia )

You also need to study radio propagation and antennas a little more. The 0-500km use is called NVIS, Near Vertical Incidence Signaling. It's used for emergency communications quite a bit. It's achieved with low horizontal antennas. Vertical antennas or VERY high horizontal antennas, as with the Steel Yards, are the ones with the low angle of radiation needed for long distances.

JD-EE

edit - added URL.