ACLS65

This is reassuring (not): "An A340 operator reported the loss of ISIS followed, during the same flight, by the loss of all Electronic Instrument System (EIS) Display Units (DU). The failure mode of ISIS has been identified as a time counter fault after 145 hours of ISIS continuous power supply. AD/A330/28 (DGAC AD F-2004-117) mandates the operational procedure to recover lost EIS Display Units, pending a new EIS standard addressing the misbehavior. Situations where EIS DUs and ISIS are simultaneously unavailable may result in hazardous consequences in certain phases of flight. This Directive requires a periodic on ground-reset procedure to prevent the ISIS failure by clearing the time counter." AD/A330/42 Integrated Standby Instrument System 13/2004 The temp fix was to cut power to the ISIS for more than 5 secs so it would reset.

Dutch Bru

Yes, there was quite a discussion before on the mobile crew rest (MCR), which posts have been deleted, including mine, one of which gave the link to the similar MCR on the 340 now reposted by PJ2. Thanks for that.

In the previous posts it was confirmed by a fellow poster, obviously in the know, that MCR on Air France's 330's are positioned below deck at the hight of the 3L/R doors just aft of the wing box.

The debris photo of part of the MCR outside panel as reposted by lomapaseo seems to show heavy underside damage, possibly caused by this part coming loose from the structure that is locked in the floor of the hold.

Just as lomapaseo theorises that this supports inflight break up, I'm still there with a theory of a low vertical and horizontal speed (mostly) intact, tail first impact (making the V/S breaking away, while damaging the rudder base), followed by a further deceleration in which the a/c broke aft of the wings, releasing the cargo container there (inter alia).

Obviously the real and heavily filled cargo containers crushing the MCR, allowing the part on the photo to break free.

Just trying to see what could have happened.

einhverfr

@acls65:

I am surprised that timer wraparound faults are not addressed better.... It is possible that this is a similar issue here which lead to a tragic ending because it happened in bad weather.

captainflame

Capeverde2008:

I was merely refering to weather analysts pont of views. TO THEM it wasn't any particular CB cluster. As you say they happen frequently...you mean like in "everyday" ?

It was in response to a previous post from a non aviation professional who was mentioning typhoons or hurricanes !!

I was certainly not implying that these things shouldn't be avoided.

So relax, avoid as you wish, retire soon, happy landings !

ClippedCub

PJ2, ttcse, others,

Not familiar with the referenced book, my experience comes from having an interest in the manner since I like aerobatics, working in the industry 30 years, and proprietary manufacturer data.

The tips of swept wing transports stall first due to induced flow upwash angles. You all know the leading edge produces negative pressures. Pick a wing station inboard. As the air passes next to that inboard station, but outboard from it, it senses that negative pressure and starts flowing upward. Therefore, that wing element sees a slightly higher local AOA. This follows for each chord wise strip, or tiny section of wing, and by the time you reach the tip, the induced upwash can be 2-4 degrees higher than the reference airplane AOA at cruise conditions. Wing twist is set by this for best efficiency. As you increase AOA in the clean configuration, the tip will stall, with an attendant nose up pitch break, and is a terrible stall characteristic. Good reason for not stalling a swept wing transport in the cruise configuration. True now as it was in the beginning.

Tip stall is mitigated by the use of flaps and slats for the T.O. and lndg configurations, and are used to tailor acceptable stall characteristics.

Flat spins are sensitive to cg position and power effects. The aerobatic single engine airplanes I've flown won't go flat without almost full power. This is due to prop wash increasing elevator power, forcing the tail down. Some of them are driven by prop centrifugal force and torque, like the Russian monster, the Yak-52. Some will recover very quickly just by chopping power and applying corrective controls. The Yak-52 wants a few more turns before it recovers once it's wound up.

You don't have a prop on an airliner, but there are underslung engines providing a nose up moment. Is it enough, maybe, nobody's ever done the math. One thing though, you don't want these big airplanes spinning. Boeing experience on the 707 taught everyone that. The 707 did a snap roll a couple of different times. They recovered the airplane due to the design requirements of the time. Think they certified under CAR 4a. Much more demanding than the current regs, so the 707 is a tough bird, still significantly damaged, but it returned the crew.

Not hard to envision current airliners tearing themselves apart in a spin. The engines would go, moving cg aft, maybe it goes flat. And/or the empennage would go, and/or the nose breaks lose. Who knows. The crew would be under tremendous g load in a fully developed spin. Due to inertia and mass, it would take a couple of turns to really wind up.

Hope this helps.

DorianB

Very interesting article on the US Coast Guard providing drift analysis technology and support:

Air France Flight 447 Recovery Assisted By U.S. Coast Guard And

PJ2

ClippedCub - yes, it does, many thanks for your time and effort. We'll see what else emerges in the next weeks.

Sorry, the book I'm referencing is "Handling the Big Jets" by D.P. Davies, first out in 1967, last ed. 1971 - can be found online and is the best (and, perhaps aside from "Flying the Wing"), only book on handling large transports which doesn't cater to manufacturer spin but is instead written with great knowledge and even humour, to airline pilots.

It is both the easiest thing and a very difficult thing to conjur a series of circumstances from the first hint of difficulty to initial degradation to the beginning of the actual accident sequence. One thing is for sure - there's a lotta people out there including cockpit crews, "into the books".

ClippedCub

If it's stood the test of time, must be good. The physics don't change so the book will still be valid.

Another geometric characteristic that would promote flat spinning in airliners would be the fuselage cylinder ahead of the wing. At extremely high AOA, it would produce a drag force in the positive lift direction ahead of the cg.

BOAC

What is apparent here is that no-one DOES seem to know what can go wrong with it, so the answer to your second question is a resounding YES!

Wth such a spider's web of computer cross-chat, it is surely vital to have some sort of independent backup for that 'Titanic' moment when the infallable safety net breaks. Even basic uncorrupted attitude would be a start. TAS and ALT are not essential, you can live without profile rate for quite a while.

- up until the last week or so I had, in my blissfull ignorance, assumed that ISIS would have a mini-platform, thus requiring those "accelerometer" and "??gyrometer??" in the 'box'. Now I'm left wondering.

icarusone

ISIS Attitude information is independant of all three of the ADIRUS.

Regarding the ADIRU 1 + 3 inputs into the ISIS:

The ISIS has 2 possible power sources (ESS and Hot Bus)

When passing through 50kts the power source for the ISIS changes from the ESS to the Hot bus. How does it know when to do this? The airspeed information from ADIRU 1 + 3. Hence the reference to ADIRU 1 + 3 inputs.

This thread has gone interglactic in its amount of information. It reads like 80 people yelling at once. Good luck to you all and I hope we find the cause soon.

A33Zab

ISIS INTERNALS:

ISIS BREAKDOWN:

ISIS INTERFACES:


AIB Presentation 2003,

Considering connection with IR Bus ADIRU 1 / 3,
besides the supply switching.

AMM:

Two ARINC 429 high-speed buses provide the ISIS indicator with the ADIRU ground speed information.
This information is used as a backup source for ISIS ground/flight condition.

and,

Optional magnetic heading.

The magnetic heading provided by the IR1 or IR3 enables the display of the magnetic heading information.
.......
In case of detected failure of the selected ADIRU, a red HDG Flag is displayed in place of the magnetic heading indications.


A33Tech.

Rananim

Someone please clarify.Is the ISIS a mini PFD or a genuine standby instrument?This is at the heart of the matter.Forget wreckage,concentrate on what the crew had available to them in the last minute.This will tell the story.

Above is ACARS summary from Lemurian.

Question(s) for Airbus pilots:

a)data for triple pitot heat failure is surely relevant here;not what happened but it amounts to the same thing.In that case,what it says is that one of the ADR's should be turned off to trigger the ECAM "ADR DISAGREE".If icing anticipated,turn off 2 ADR's?Why?Does it not follow that the 02:12 ACARS message "NAV ADR DISAGREE" was in response to pilot action?ie pilot de-selects ADR(s) OFF.If the multi-pitot failure/blockage theory is correct,theres nothing for the computer to find disagreement with.Theyre all three plain wrong.So the pilot must intervene to generate the ECAM warning.Is this correct?

b)It mentions it again for dual pitot heat failure;the inherent risk is that the computer will lock out the one remaining good ADR.So the pilot must turn off one of the faulty ADR's and trigger the NAV ADR DISAGREE which in turn will lead to ADR CHECK procedure.
So NAV ADR DISAGREE warning can either be computer-generated OR pilot-action generated(Pilot switches OFF faulty ADR).Correct?



d)Now that one ADR has been disabled and the ECAM warning generated what next?For ALT LAW to kick in,there must be disagreement between 2 remaining ADR's,either speed or AoA.If speed disagrees in the remaining two ADR's,crew fly unreliable airspeed procedure.If speed does not disagree,AoA is culprit,and crew ignores spurious STALL warnings(Air Caraibe).Is that about right?

e)The ECAM message in the Air Caraibe incident was FCTL ADR DISAGREE,not NAV ADR DISAGREE.What does this indicate?How telling is it that both 447 and Caraibe began with AP/FD disconnects?


f)In ALT LAW roll control is same as for Direct?Please enlighten.Sensitivity issues?

g)With a CAS of say 90 knots on all 3 ASI's at time of RUD TRAVEL LIMIT warning,could rudder deflection of 31 degrees been mandated by the computer as the limit?

h)What is the QRH procdure(2.20) for flight with all 3 ADR's out.ECAM apparently only covers single ADR op?Would 3 ADR's out=Direct Law?

I know Im out of my depth(never piloted/travelled on an Airbus) but no-ones asking questions that relate to what the crew had available to them in the last minute.What did they see??

p51guy

NO COMMERCIAL AIRCRAFT SHOULD GO WITHIN 100nm LATERALLY of CB - ICTZ (Tropical Cumulo-Nimbus clouds) activity.....for any reason...FULL STOP. NEVER, EVER, don't even think about it.... or trying to overfly it. That is from MANY years experience and many grey hairs of wisdom!!

I guess you never fly through Texas with thunderstorms as big then. You would have to deviate over Oklahoma with those parameters.

A33Zab

A33Tech.

HarryMann

Rananim

Good post there, these are the questions many others would like to see answered too, I suspect.

Hyperveloce

The Air Caraibe report shows that the NAV ADR Disagree is a consequence of clogged/iced Pitot tubes and static ports and slightly different patterns of errors on each ADM generating the disagreement, making air speed and maybe also barometric altitude erroneous/inconsistant (this may be a potential problem for the TCAS ?). The NAV ADR disagree did not require a manual intervention of the pilots on one of the ADIRUs to appear.
All the faulty instruments, the ADIRUs, the ISIS and the TCAS appear to be inputed by Pitots, static ports or barometric pressure sensors. What if these Pitots and the airframe (static ports) were rime iced crossing a cumulonimbus ? If in the mist of a tropical cell, subjected to turbulences, severe icing conditions, a gradual deprivation of all air references (air speed), followed by a cascade of faulted nav instruments, multiple alarms in the cockpit, unreliable/unconsistant indicators, loss of flight assistances/protections, maybe a stall due to erroneous air speed which surprised everyone on board ? (pilots trying to understand/cope with the multiple failures, air crew). A stall followed by a dive, resulting in an overspeed during which the vertical aileron (and the spoilers ?) may well have been ripped off (not protected anymore against excessive manoeuvers in alternate law) before a critical airframe break up.
Hope these black boxes will be recovered soon.
Jeff

augustusjeremy

The ISIS and IR2 faults were related to inertial reference faults/disagreements. I can't see how airspeed and barometric data would per se generate them.

If they were later caused by a "shaking to hell" a/c it's another story...

Iceman49

Aquadalte: "If this aircraft was equipped with ISIS, it was surely also equipped with Back-Up Speed Scale and Back-Up Altitude, that displays on PFD's and gets information from AoA probes and GPS, respectively.

This is very important, once Back-Up Speed Scale and Altitude are only displayed when ALL ADIRU's are switched OFF. The old procedure (for the stand-by horizon equipped aircraft), would call for the trouble-shooting of the IR fault, and in case of all confirmed to be faulted, one would be advised to switch off those connected to Captn PFD and FO PFD, i.e., IR1 and 2. This would allow the aircraft to revert to Alternate Law.

This system (although brilliantly designed) seams to work fine in rather smooth air and on flight simulators, but I find it very difficult to see it properly working in a heavy turbulent environment, as seemed to be the case of AF447."

Believe that the Back-Up Speed option was available in 07, the ISIS does indicate Speed/Altitude and Attitude...the instument is only about 1.5" square.

grumpyoldgeek

I have a very simple question and I've reviewed virtually all of the posts in both threads and not found an answer.

Of the pilots' flight instrument displays, what do they lose if all 3 pitot sources fail? Do they lose everything including bank, turn and vertical airspeed or do they just lose the airspeed display?

Will Fraser

I'd like to direct a comment toward the VS/Rudder separation. From AA587, we know the failure was complete, and purportedly caused by pilot input beyond structural limit. Here, we know the failure was complete, though the cause and sequence is unknown.

A short aside re: MD11 gear. In Narita, one noticed how astonishingly robust the NG and one of the MG were, the nose gear remaining attached even after that last horrific nose plant, the Main pushing up and through the main spar of the wing. BA038 lost the gear and saved the hull.

Here's the thing. The composite VS is exceedingly strong, as well as relatively light, and demonstrably resistant to failure except at the root, where it attaches to the fuselage. It remains pristine, not a crack in its skin.

Its failure at the root is obviously related (Lateral failure, if such was the case, as I believe) to its leverage on the root, a tall structure that is acting through its moment arm on a robust attachment.

I am pointing out that its only failure was complete; it resisted until the total failure was unavoidable. If however the structure was designed to have top down sequential failure, a component at a time, each loss of surface area would reduce its advantage on the root, allowing for the remainder of the fin to operate in reduced effectiveness, but allowing for directional stability. Something as simple as two or three parallel seams such that the first piece to depart might be the topmost one meter, the next failure point halfway, until the moment arm was short enough that the fuselage attachment would be virtually failure proof, though providing very limited D/S.

I can't take my eyes off that nearly perfect fin.

One can envision 587 retaining some fin if not rudder. Rudder is a trimming device, without it things are difficult, lose the VS and you lose the a/c.

WF