Originally Posted by ventus
Loss of the VS is a lunatic theory
No, I do not see it as lunatic. I see it as unproductive to pursue because of its low probability and in your case relative unimportance.
I seem to have upset a few people.
No, you and many others replowing ground that is already quite well furrowed. We're not singling you out, even though I am not quite fisking your article. It just gets boring. Although I will admit that most people presume the VS departed the plane which started the whole mess.
The loss of the VS is a lunatic theory is it. So your collective assertion is that the VS on AF447 could not have separated in flight, and could only have separated at impact ?
I repeat, it is not lunatic. It is simply low probability given other facts as presented by BEA so far.
With all due respect to all of you, perhaps you have a different theory to the NTSB for AA587 ?
No. That is an entirely different event.
Note: - from the NTSB report:-
page 159
3. Conclusions
3.1 Findings
6. Flight 587’s vertical stabilizer performed in a manner that was consistent with its design and certification. The vertical stabilizer fractured from the fuselage in overstress, starting with the right rear lug while the vertical stabilizer was exposed to aerodynamic loads that were about twice the certified limit load design envelope and were more than the certified ultimate load design envelope.
Note the bolded text above. The shear from that event presented itself differently on the mounts than the AF447 damage we see in pictures. Just because the VS came off in both events does not mean they came off for the same reason. One must examine the damage to further diagnose the obvious fact that the VS and plane parted company.
Reference Document.
In-Flight Separation of Vertical Stabilizer American Airlines Flight 587
Airbus Industrie A300-605R, N14053
Belle Harbor, New York
November 12, 2001
Aircraft Accident Report: NTSB/AAR-04/04 PB2004-910404 Notation 7439B
http://www3.ntsb.gov/publictn/2004/AAR0404.pdf
Content References:
pages 31 to 42 inclusive.
pages 50 to 52 inclusive.
pages 55 to 60 inclusive.
page 159
3. Conclusions
3.1 Findings
6. Flight 587’s vertical stabilizer performed in a manner that was consistent with its design and certification. The vertical stabilizer fractured from the fuselage in overstress,
starting with the right rear lug while the vertical stabilizer was exposed to aerodynamic loads that were about twice the certified limit load design envelope and were more than the certified ultimate load design envelope.
Did you intend this apparent repetition. It still contains the significant sentence fragment.
Now, I re-emphasise:-
"The vertical stabilizer fractured from the fuselage in overstress, starting with the right rear lug while the vertical stabilizer was exposed to aerodynamic loads that were about twice the certified limit load design envelope and were more than the certified ultimate load design envelope."
(sigh) Three times does not make it true, Lewis Carroll notwithstanding.
I did not say, and I am surprised that many of you have apparently assumed, that I meant, or was implying, that the VS separated at initial upset. I did not, and I do not.
OK, on that much we do agree.
I actually think the VS came off somewhere between 5,000 feet and 0 feet, but probably closer to 5,000 than zero.
You are free to think this. I depart from agreement here. It is not the simplest scenario.
At the risk of repeating myself from a prior message the evidence in hand strongly suggests that the VS was still attached until the last 15 seconds or so of ACARS transmissions. It MAY have come off at the time the last message was being transmitted.
It is also fairly apparent that the plane was below about 6k to 8k' at the time. That leaves room for your scenario. It would be incumbent upon you to describe what might have happened between 0 and 5k', picking your numbers, that would be significantly different as far as the VS is concerned than what happened at higher altitudes. Furthermore you should be prepared to describe how the VS assembly took on an appearance that it had been pushed forwards rather than sideways given the visible clevis joint related fractures.
It also strongly suggests the plane was out of control and falling out of the sky at the time. It less strongly suggests that flight control recovery had been effected and pull out proved to be impossible and in fact the plane died in the middle of a frantic pullout attempt. I happen to like that scenario as it explains a few things. But it's too thoroughly based on coincidence to make me happy.
My thinking is that there are many plausible ways that AF447's VS could be (and I think were) "exposed to aerodynamic loads that were about twice the certified limit load" (if you will - and more) when the aircraft entered the "denser air" at lower levels.
That is part of an explanation. What loads? What was the plane doing at the time that created huge loads?
To begin at TOD.
For various reasons, I belieive that the initial upset was an underspeed stall upset, which resulted in the aircraft entering a fully stalled descent, with a flight path angle about 30 degrees off vertical (60 degree dive) but in a fully stalled condition, basically in a slow flat spin, slowly yawing (my theory is happy either way, but, lets say to the right if we assume that the BEA is correct with their "tail was moving to the left at impact" idea).
That is a fairly likely scenario. But if they are only 10 miles away from LKP was it God's hand that appeared in front of the plane and slowed it fairly rapidly and then let it fall in a stall? Facetiousness aside, what slowed the plane down so suddenly? It wants to fly and in four minutes it was in the water from 35000'. Where did all the plane's energy go? And more importantly, why?
The tail was moving to the left idea comes from the way the pairs of clevis joints tore off the plane. Study the pictures of the recovered VS REALLY closely. Then play with holes punched in paper with a fitted dowel through the holes. Pull it off different ways and see what form the tears in the paper take. It's only a rough equivalence. But you should quickly see a pattern. You have to get the tail or at least part of the tail moving forward up over the fuselage in order to break it that way. And I agree, how in heaven's name did it avoid getting bunged up by other parts of the plane?
Possibly there was considerable surface winds and they picked up the tail as it broke lose and moved it the critical distance.
This mode of descent would be obviously unstable, but it would also have been unrecoverable, due to total stall of both mainplane and HS, and aft-ish CG. So it was probably oscillating slightly in all axis on the way down, ie, probably oscillating in pitch from about level to perhaps 20+ degrees up, in roll a bit, and in yaw rate a bit, but with yaw always positive. In other words, sort of like a fully stalled spiral descent to the right with a bit of Dutch Roll-ish type motion happening.
Now, air loads are a function of dynamic pressure, and I am damn sure that the pilots, if still functioning on descent passing FL100, would, by that stage, have been open to trying "anything" to recover, including, full rudder inputs. But before you jump to the false conclusion that I am saying "AA587 First Officer", I am not.
If the plane was making very low headway at the time it was within 5k' of the water AND rudder travel was limited to in flight high altitude deflection limits, how would full rudder provide enough force? It was going forward too slowly to remain airborne. So full rudder deflection would not be quite as exciting as it would near 300MPH at 35000'.
I am actually thinking more along the lines of turbulent eddies shedding off the fully stalled HS imparting latterally asymetric cyclic loads on the VS, (sort of like the FA18 LEX vortex busting on the fins idea - if it helps your mind's eye) such that, at some point, both the forces and the frequency were sufficient to either break outright, and or generate a resonant flutter like condition, that snapped it off at the base. The latterally asymetric cyclic loads would be generated by the varying wake fields from the port and starboard HS's due to the oscillating yaw rate. If you think it's out of left field, fine. I don't. Read the DSTO reports on the LEX vortex busts. Interesting stuff.
Perhaps. And maybe the recorders MIGHT be able to sort that event out from the impact. It seems to be a huge stretch with coincidences.
And, as someone else observed, why is it important? The important information is that information which leads to learning why this whole sequence happened in the first place.
Additional forces, either in or out of phase, could be generated by pilot rudder input. The combination of both, as dynamic pressures increased with decreasing altitude, may well have been "more than enough".
For those who assert that the rudder arm needed 36g to break is proof of impact forces, is it worth considering the possible effects of resonant and flutter forces ?
I think the posts asserting that there were no ACARS messages indicating hydraulic failures, "thus the VS was definately present at impact" is also flawed logic.
It does prove that the VS was still attached down to a small fraction of AF447's reported flight level. It was provably still on until at least a few milliseconds before the start of the last transmission. Otherwise the word that it had come off would have preempted other messages, I believe.
That does leave room for double flame out and a short glide from under the altitude at which the pressure warning would be issued. And it cannot be issued if the plane is gliding. That's a "falling" rapidly sort of message.
First, the ACARS stream ended well before impact (some seconds at least) did it not ?
We do not know. And the simplest presumption is that it ended coincident with impact. There was a hiatus in the transmissions that is suggestive of the plane going through some interesting gyrations. (Or perhaps had turned 180 degrees from its flight heading so the antenna was aimed as far away from the satellite as possible. I believe the antenna lobe is large enough that a bank would be required to take it off the satellite. I wish I knew the dynamics of the antenna aiming algorithms. I suspect they are fast enough to track through a tight turn. So that loss could mean a very severe bank.)
Second, even if they did only end at impact, and since my scenario has the VS separating effectively only a handfull of seconds prior to impact, would any hydraulic messages have been generated, prioritised, and injected into the ACARS transmit stream in time to have been actually transmitted in those couple of seconds ?
Please see above. It would take milliseconds for the message to be reformatted to note the massive hydraulics failure. So until a very short time before that last message started transmission the VS was there. We don't quite know where "there" was.
It could be that the VS was attached at impact, but I think those who assert with a passion that it definitely had to be, and was, are being a bit closed mind-ish.
It would be nice to know one way or another. But, the really important thing is "what started the avalanche?"