This incident reminds me of a 737 crash that occurred many years ago in either Columbia or Panama. It was documented in a NOVA episode on public television here in the US. The plane broke up in the air during a night flight over mountans. The first reaction was that a bomb had detonated, which seemed to make some sense because of the scattering of the debris and drug cartel incidents that had happened in the past. Once the FDR was recovered, it showed roll rates in both directions that made no sense. In fact, Boeing engineers said that they were inpossible rates. What the investigators finally determined was that there was failure in the gyro for one of the artificial horizon indicators. Once a bank angle was reached in a turn, the AI would lock in that position. Reverse input from the crew would indicate no change, so they would bank opposite even harder. With no visual reference, they assumed that the indicator was correct. This was compounded by the fact that in a 737, normaly, both the pilot and copilot AI's run on seperate gyros, but in this case, a selector switch was set to have them both operate off of one. Hence, they were both seeing the same thing failure. When the gyro would unlock, it would snap over to the actual bank angle of the plane, which at this point was a steep angle in the other direction. The crew would then correct back to the opposite direction again and the same thing would happen. The result was they rolled the plane over and entered a dive which exceeded the limits of the airframe.

The reason I bring this up is that it sounds like the FDR in this recent case recorded information that, on it's own, makes no sense. What kind of instrument failure could fool pilots into making incorrect inputs into either yoke and throttles, and at the same time give indications on the FDR that are also incorrect? I would not jump to the conclusion that the strange FDR outputs are the result of an explosion or breakup. They could also be an indication of instrumentation failures.

"I would not jump to the conclusion that the strange FDR outputs are the result of an explosion or breakup. They could also be an indication of instrumentation failures."

Patrickal

Pneumatic inputs of erroneous pitot/static pressures to the CADC (Central Air Data Computer) could have this effect... if the failure compromised the pitot-static system prior to the FDR losing power.

However, thinking outside the box a little here, both the non-transponder secondary (raw paint) radar data and FDR info would be consistent with the gyrations of an aircraft that had lost its tail and had then pitched up violently under climb power. I have said in a prior post on this thread that I consider that the aircraft gyrations following loss of the empennage would be a drastic pitch-up into violent end-over-end tumbling - because the engines would be still under high power.

The radar plots roughly tie in with the tail departing and then, some finite time later, the breakup of the remainder of the wings engine and fuselage. If you look at my prior posts you might see the picture more clearly. The Airworthiness Directive that I quoted is quite clear about this being a possible outcome. Unreported/misassessed/improperly repaired heavy landings or tail-strikes can initiate damage (cracking) in the aft pressure bulkhead area that can "work" over time and eventually give way under the greater pressure differential at height.

UNCTUOUS,

I disagree with the idea that a plane would pitch UP after loosing its stab or its tail. It is much more likely that it pitches down.
The horizontal stab normally produces downforce, although on modern planes part of this is produced by fuel weight in the stab, which is more drag efficient.
Taking into account the lost mass at the tail, the DOWN pitching moment is even larger.
So loosing the tail would explane the reverse heading (half loop forward) and the breakup, but not the altitude gain. If something broke apart and caused an pitching UP, it must have been the nose section.

There had been the idea, the crew might have allready been passed out during breakup due to pressurization malfunction and hypoxia.
Did the FDR tell us something about that ? Had there been normal conversation in the cockpit ? Have not read anything about words spoken by the crew, are there indications they were noticing problems and try to analyse and react ? Or are there just ´abnormal sounds´ and no signs of human life on the tape ?

"I disagree with the idea that a plane would pitch UP after loosing its stab or its tail. It is much more likely that it pitches down.

Volume
Fully expected someone to take me up on that aspect. It's wholly dependent upon whether:

a. empennage detached cleanly and instantly, or

b. it tore off from the bottom, hinging at the top, or

c. vice versa, but

d. momentary random (or hardover) inputs to elevator and/or rudder before detachment are also possibilities - as is prior loss of a single primary control surface and FCS or crew intervention with power and trim changes causing a pitch-up (and detachment occurring sometime thereafter). It is perhaps more likely that control would be lost some few finite seconds before empennage detachment - and then (logically) that the gyrations resulting from control loss would assist (or ensure?) complete detachment.

Who's to know? And with an ancient 18 parameter FDR, it may never be known for sure. My first guess was that after the loss of the minimal tail downforce (there likely being little tail-tank trim fuel in there on a short hop), the initial commanding couple would be Thrust/Drag and that four climb-thrusting underwing-podded jets would rotate it into a climb.


It's not feasible to upgrade old jets to modern DFDR's but it is quite practical to throw in some recording cameras in crash-proof boxes. They would just need to be mounted rationally and plumbed into power. IMHO cameras internally and externally are now necessary post 911. Investigations can drag on for five years or more and in the interim passenger confidence is eroded as they suspect:

1. maybe it was a terrorist act (they don't really seem to be sure and need all this time to tell for sure - and
would they tell us if they then knew (or do they know now - and it's all a time-gaining pantomime)?)

2. maybe old airplanes just aren't being maintained properly because they wouldn't then be so profitable to continue operating them; maybe they just keep their fingers crossed and hope that things like this DON'T happen.


So if you simply fin-mount a non-CCTV wide vista recording camera as a post-crash disclosure device, you need do no more than provide a one hour endless loop tape (or a hard disk), mount it in a transponder-fitted crash-proof housing and then functionally check it on a regular basis. It need not necessarily be cockpit monitorable. Improvements upon that external system would be that it was to be a CCTV which could be displayed in real-time on cockpit monitors as an aid to crew trouble-shooting and surveillance. Dollar cost of the basic system? US$150K per airframe maybe. Swissair was planning to be first to have such an in-fuselage system internally in their finalized post SR-111 MD-11 upgrade, but I would guess that Swiss Lines would've canned that project as they too are now struggling to remain afloat. You have to review for yourself what cameras might have disclosed immediately post-crash in Silkair, Valujet, TWA800, SR-111, AI182, the Helderberg, Panam103, KAL007, KAL 801, GF072, AA587, Egyptair Flt 990, China Northern MD-82 fire off Dalian and the CrossAir RJ-100 Zurich etc. Then you have to wonder what use to the crew fin-mounted cameras may have been (Concorde, the Air Transat A330 Glider, Alaskan 261, British Midland 737 at Kegworth etc).

But primarily weigh up that the significant percentage minority of the now post 911 non-flying public might be more likely to engage if answers were more readily available. It would also do much to dampen the cult of conspiracy theorists that now thrive upon airline accidents and foment public anxiety. That fearful ex-passenger minority are perhaps the difference between operating in the red or black.

Food for thought.

Unctuous, there is no fuel tank in the tail section of the B747 Classic airframe (i.e. -100,-200,-300). However, I think the camera is a very good idea.

I am skeptical of theories that either don't explain all observations
or which, if true, would have led to consequences that were not observed.

Unctuous, there was an increase in pitch, not much except
percentagewise, for the final seven seconds of FDR, which is the time
interval during which the climb rate appears from the FDR to have
increased. The final FDR climb rate and the final FDR altitude (at
7:27:58) would lead to an altitude 29 seconds later that would
be in the ballpark of the Taiwanese military radar reading
of 36911 at 7:28:27 (which is about the time of the breakup, if I
understand PB correctly). I am very skeptical of that radar reading
- it is not consistent with the preceding sweep - but unless and until
it is repudiated there it is.

If I correctly understand the superposition that the ASC made on one
of the FDR plots then there was a cockpit communication 93 seconds
before the end of the FDR recording, 80 seconds before the start
of the vertical and lateral acceleration anomalies, and perhaps
less than a minute before the decrease in EPR4. I am not knowledgable
about the onset of hypoxia: would that communication be consistent
with a hypoxia theory?

It would help to have a more precise understanding of the
chronology. Can anyone shed light on the three question marks below?

7:26:25 Cockpit crew communication - Was this the last recorded?

The next four instants of time are identified with vertical lines on
the ASC's time synchronization plot. But the last two are labeled only
in Chinese.

7:27:58 FDR Stop
7:28:03 CVR Stop and last beacon return
7:28:17 ?
7:28:29 Breakup?

bblank

IMHO, it would be difficult to build a case involving Hypoxia with the info we know. However, from memory, the key, USEFUL consciousness depends largely on the individual, but the general rule of thumb is that above FL300, you don’t have a lot of time. It varies, at FL300 perhaps better than a minute, but as you climb, the envelope becomes much smaller. FL340 less than 45 secs, and above 350, less than 30 secs. At FL400, perhaps 15 secs. These numbers may be further reduced if the loss of pressure is sudden because of the corresponding sudden escape of gas from the lungs. That why you have quick-donning O2 masks, remember.

bblank
Not sure what you're exactly being sceptical about. I assume, since you named me, that I am the offender. There are some suppositions in what I've said - but nothing that seems to diverge from your time-line.

7:26:25 - last cockpit communication
1min26 secs later at
7:27:51 - beginnings of a pitch increase
7 secs later at
7:27:58 - no further FDR data
5 secs later at
7:28:03 - CVR stop and final transponder return
14 secs later at
7:28:17 - unknown recorded event
10 secs later at
7:28:27 - peak altitude recorded on skin paint of 36,911ft
7:28:29 - breakup?

I'd agree with Wes_Wall on the hypoxia question. If hypoxia was involved I suspect it was more as an incident outcome than a contributor. The breakup occurring very noticeably just 2 secs after the observed max altitude is not normally a logical sequitur, the greater probability being always that a loss of control breakup will occur (per say, Egyptair MS990 and the Lauda 767) after it's wound up a little in a g-increasing spiral dive. However in this case (as my theory goes) the structural failure had already happened and so, minus its structural integrity (and tail), CI-611 was free to begin its breakup well short of gaining any additional downward knots or g's.

So, as far as "explaining all observations or which, if true, would have led to consequences that were not observed.", I think I've filled in the first box, but not sure where I ended up on the deficient unobserved consequences. The "unobserved" bit is actually the pretext upon which I introduced my fin-mounted (and cockpit) camera suggestion as a necessary solution to these type enigmas.

UNCTUOUS

After running short of further facts (the quiet time) I tend to concentrate on aligning the facts against possible scenarios under columns of Pro and Con.

The Pros and Cons can be analysis related (a better form of opinion) and then with the help of others with competeing theories/scenarios one can chip away at the pro & Cons trying to convert them to fact. Eventually one can pare the list down to only a couple of theories and pursue from there. The difficulty of doing this on a forum is the abscence of fact and the plethura of theory.

I see you already have some cons against hypoxia, even though its pro is that you can imagine all sorts of outcomes with it. Also some of the passengers were sure to excperience acute distress long before all of the flight crew. Surely this would have been evident on intercom calls on the CVR from the cabin crew.

Might I also suggest that some other scenarios might consider the lack of radio communication and intercom calls as a con.

Sorry Unctuous for the lack of clarity. I was only addressing you
in the one paragraph in which I responded to your previous
post in which you said "The rate of climb increase (and climb?)
recorded seems to be unsupported by any zoom-climb increase in
pitch attitude." I just wanted to point out that pitch did increase
during the last seven seconds of the FDR.

There is a lot we don't know (that is, those of us who only have access
to the data that has been made public). We don't know about the CVR sounds,
we don't have a map of the debris field, recovery of a large percentage
of the a/c, a model for the flight path after things started to go wrong,
models for the trajectories of the major pieces after the breakup, an
understanding of what was going on with engine 4 and whether it had anything
to do with anything, an explanation why the flight crew did not utter
anything in the last 12 seconds of the CVR (if that is indeed the case),
a history of this particular a/c, information about what experts in
the Taiwan military say about their radar data, ...

I'd like to have a better understanding of what happened before
considering any theory. You can only judge a theory after you
know everything that it must explain.

bblank

Firehorse in several messages back said the following:

Quote
The tail wreckage group was closest to the last radar return with the cockpit group being .way south and to the west, almost the furtherest group from the tail. The tail incidentally
is rigged and ready to lift any time now. Unquote

Can we read into this that the order of debris is the tail area first, then other parts, then the cockpit?

I wonder if Firehorse can shed any additional information on the location of key peices? Has anyone see a map?

I don't know if it has been done before as this is a long thread, but I would like to thank firehorse for his contributions so far. A bit of good, real time info sure cuts out the cr@p that sometimes surfaces on these threads.

Agreed, firehorse has been VERY generous to PPRuNer's who have been following this thread, in allowing us a FACTUAL tracing - step-by-step - into this (as yet) unresolved accident.

My initial "gut feeling" on this one, was an encounter with severe weather. However the issue of crew "interference" (ala the Silk Air B737 "accident") might now appear to be something that may have to be factored into this scenario, because of that which appears to be UNSAID (and unheard) immediately prior to the final sequence of events.

Firehorse

Thanks for the update on the debris field. Interesting, and I guess we can be thankful about ruling out the CWT scenario.

Kaptin M

Lets wait until we have more information before we begin to point fingers at the crew. That will be done in due time, you can bet, so why not let it rest until then.

Kaptin M, would you care to elaborate on the meaning of your second paragraph on your last post? I have a feeling you are leaving yourself wide open for a sh!t load of flak, legal and otherwise.:(

Firehorse - need some info please
 

1. Request status of two midspan latches of forward cargo door: May be missing.

2. Request status of two pressure relief doors in forward cargo door: May be jammed open or missing.

3. Request number of pieces of forward cargo door retrieved so far: Door may be shattered and broken longitudinally.

4. Any paint transfers above hinge of forward cargo door?

5. Any inflight damage to right wing leading edge wing fillet?

6. Any FOD to engine number three?

7. Request status of aft cargo door: May have opened in flight.

Firehorse

Are there any updates you can share regarding the recovery process and early physical findings? Also, has any additional info on the FDR tapes been forthcoming?

Thanks.

AWST 1st July:

'Relative Abnormality'
Near End of Flight 611
MICHAEL A. DORNHEIM/LOS ANGELES

China Airlines Flight 611 appears to have experienced no major problems before it came apart, but the flight data recorder shows small anomalies in the last 20 sec.

The cockpit voice recorder (CVR) reveals a few noises that experienced observers found unusual for a Boeing 747-200 and were unable to identify. One type started 7 min. before the breakup, and in the last second of the recording there is a "not very loud 'chahhh' sound," said Kay Yong, managing director of Taiwan's Aviation Safety Council (ASC), which is leading the investigation of the May 25 accident ( AW&ST June 24, p. 41).

There is no obvious correlation between the noises and anything on the flight data recorder (FDR), Yong said. The ASC is now studying the CVR with a spectrum analyzer.

Yong said the ASC has ruled out flight crew operational error as a cause. Also, there is no sign of impact with another object because the wreckage so far shows no marks of fire or explosion, he said.

The Flight 611 FDR recording started at 2:56:28 local time and stopped at 3:27:58.9. The CVR stopped 4 sec. later, at 3:28:03 when referred to radar site timing. The final radar beacon transmission observed by Taiwanese radar was also at 3:28:03, though mainland Chinese radar picked up three more beacon signals ending at 3:28:12. The FDR was powered by the essential instrument bus while the CVR and No. 1 transponder beacon were on the essential radio bus. The No. 2 transponder was on the secondary radio bus.

THE TWO BEACON TRACKS plus a Taiwanese military primary skin track indicate the airplane came apart at about 34,700-34,900 ft. It was climbing to 35,000 ft. and had taken off 20 min. earlier. The final altitude on the FDR is 34,573 ft. The FDR is a Lockheed 209F digital tape device with about 44 parameters. Some of the tape was crumpled around the reel, but the Flight 611 section was unaffected.

The FDR shows that once the airplane accelerated to 300 kt. at 12,000 ft., the vertical acceleration stayed well within a ±0.04g range until the last seconds. The lateral acceleration was within ±0.008g, except there is a curious +0.04g spike to the left at about 3:23:58 p.m. local time, or 4 min. before the FDR stopped, and at about 30,400 ft. It is not clear if this is just a glitch, but other parameters are smooth at that point.

Starting at 3:27:37, 21 sec. before the end of the FDR, the indicated thrust of outboard right engine No. 4 drops from 1.51 engine pressure ratio (EPR) to 1.49 EPR over about 6 sec. The EPRs of engines Nos. 1 and 3 rose about 0.05 in apparent response. However, this may not be unusual--4.7 min. earlier there was a greater drop by the No. 3 engine. In fact, if the airplane was flying a 300 KCAS/Mach 0.84 climb speed schedule, this is the altitude at which it would switch from holding knots to holding Mach. The FDR shows the airspeed decreasing and rate of climb increasing, which is consistent with this switch.

VERTICAL ACCELERATION was indicating a constant 0.95g approaching the top of climb, but increased to 1.00g at 3:27:48 without any change in body attitude or airspeed. (These are preliminary readings, uncorrected for instrument error.) Then the body pitched up about 0.8 deg., increasing vertical acceleration to 1.09g. As the nose came back down to its original 2.1-deg. attitude the gs dropped to 0.87, then returned to 0.94 a split second before the FDR ended. It is not clear where the accelerometer is located and how it is affected by body pitch rate. The positive load factor increased the climb rate to 2,800 fpm. from 1,500 fpm. at 34,500 ft.

Yong said these measurements are within normal tolerances and are only being looked at because of the accident. "They are only relatively abnormal. The flight was very smooth and well-behaved, then something changed. If it had landed in Hong Kong, these changes would be totally acceptable."

THE CVR WAS REVIEWED by two China Airlines pilots, a Boeing official, an FAA official, and two ASC investigators. They found seven heartbeat-like noises that occurred from 3:20:53-3:21:14 to be unusual. Also, the final 'chahhh' sound was unusual. The tape has a number of unidentified clicking and squelch break sounds, but they may be switches or other normal noises. The CVR stopped recording for 0.3 sec at about 3:12, but this was likely a problem with the CVR itself, Yong said.

Wreckage recovered from the Taiwan Strait includes the cockpit, a 72-ft.-long section of the forward right wing, the upper deck skin between fuselage station 620-780, and the No. 4 engine. The copilot's body was in the cockpit. The No. 4 engine, which showed the last-minute slight EPR drop, was found 0.8 mi. southwest of the main wreckage area that included the cockpit and right wing. Other parts of the right wing are still attached to the fuselage there and had not been recovered as of late last week.

There is a shape on this right wing that might be the No. 3 engine, but divers were not able to tell for sure. The left wing appears to have broken up more in flight--most of the floating wreckage consisted of control surfaces from that wing. The structural part of the left wing has not been identified. "We're pretty sure the left wing is not by the right wing, but the main wreckage area has lots of debris," Yong said. The No. 1 and 2 engines have also not been found yet.

From AW Week:
There is a shape on this right wing that might be the No. 3 engine, but divers were not able to tell for sure.
Also, the final 'chahhh' sound was unusual.

Well, yes I know that Firehorse reported fcd 'intact and locks in place'...

But..when the nose comes off when forward cargo door ruptures in flight, the nose turns to right and hits number three engine usually tearing it off early. (3 is also fodded from cargo compartment of passengers and on fire.)

Also, aft cargo door is identical to forward and possibility of mixup always there.

And, 'chahhh' is not 'thud' and closer to the sound of door unlatch motor trying to unlatch the ten latches...which only have eight locking sectors on bottom eight latches leaving the midspan unprotected from uncommanded opening. So...when one says 'locks' he may be referring to the bottom eight locking sectors which would be in place after AD 88 12 04 that strengthened them after PA 125 door was found open after landing and implementation accelerated after United Airlines Flight 811 door opened in flight

But...the nose should be first in sequence of debris from last radar contact, and sound should have been loud. If tail first and sound soft, then it may have been aft cargo door rupture or aft pressure bulkhead failure.

If shape on right wing is engine number three that may mean nose did not come off and hit it.

The evidence will clearly show where the hull rupture occurred. If forward area there will be inflight damage to wing and right stabilizer and engines 3 and maybe 4. If tail came off there should be no inflight damage forward of tail.

Question remains, hull rupture in flight, where and when and how?