BA038 (B777) Thread

[Spaz Modic]

:* The ultimate update - it was outa gas! :bored:

[jackharr]

Is the AAIB required by regulations to publish its findings within a certain time scale? Or is it able to procrastinate indefinitely like a certain African leader?

Jack

[lomapaseo]

Is the AAIB required by regulations to publish its findings within a certain time scale? Or is it able to procrastinate indefinitely like a certain African leader?


The AAIB responds to both the government as well as ICAO.

Send your letter of support directly to them and CC either of the above.

[ribt4t]

I thought one year was the typical timeframe for accident investigations.

[Green-dot]

Reading the AAIB Bulletin S/1 2008, I found the spar valve description a little bit puzzling regarding the examination of the spar valve circuitry and which valves are actually addressed. The spar valves installed on the rear spar of the wing or the engine shutoff valves:

On page 5 of the bulletin:



"Spar valves

On examination, both engine spar valves were found to be OPEN, allowing the fuel leak evident at the accident site.

The spar valves are designed to shut off the fuel supply to the engines . . .etc. . . . .etc . . . .

The wiring on G-YMMM was as originally designed and manufactured, and such that when the fire handle was operated, it isolated the power supply to the run/cut-off relay. When tested, the run/cut-off relays for left and right engines were still in the valve OPEN position, despite the fuel control switches being set to cut-off. The fire handles had also been pulled and the engine fire bottles had been fired. Therefore the fire handles had been operated prior to the fuel control switches.

The left spar valve circuit breaker (CB) had been tripped. This was due to damaged wiring to the valves as a result of the left main gear being forced upward through the conduit at the initial impact. The tripping of the CB meant there was no means of electrically closing the left spar valve. Similar damage was also evident to the right spar valve wiring, however, in this instance the CB had remained set."

On page 6 of the bulletin:


"Examination and tests of the wiring identified that, in the case of the right engine, the valve CLOSE wire from the run/cut-off relay was still continuous. This could have allowed the valve to operate had the fuel switch been operated before the fire handles."

1. First of all, I assume the AAIB refers to the engine fuel spar valves in the aircraft fuel system, not the engine fuel shutoff valves (SOVs) on the engines? As I read the text, it leaves room for misinterpretation as to which valves they actually refer to. No mention is made of the spar valve control relays, which in case of operating the fire switches are the relays directly isolated, down stream of the run/cut-off relays. The run/cut-off relays referred to by the AAIB are multi-function relays. Depending on which relays are actually addressed, other run/cut-off relays control the engine SOVs and are also wired to the fire switches.

2. Furthermore, has the AAIB positively established (proven by evidence/testimony) that the open spar valves were a result of incorrect sequencing or is that an assumption purely based on the pre SB 777-28-0025 configuration? Or is the AAIB also trying to find other possible causal scenarios that may have lead to a similar effect?

3. As I show in bold text above, and assuming the AAIB refers to the aircraft spar valves in the engine fuel feed system, there is perhaps a contradiction in their findings. On page 5 the AAIB found the run/cut-off relays "were still in the valve OPEN position." However, on page 6, in the case of the right engine, "the valve CLOSE wire from the run/cut-off relay" was still continuous?

I assume the AAIB means the run/cut-off relays (Reset/Fuel Spar L-(R-) Eng) located in the left (right) power management panel. From that relay onward the normal wire paths have yet to pass through the spar valve control relays before reaching the spar valves. If so, how could the right wire path still be continuous when the right run/cut-off relay was found in the OPEN position? Under these circumstances this would imply that the right spar valve control relay should also have been found in the OPEN position, hence no continuity. The bulletin, however, does not mention any examination regarding condition/position of those control relays. My assumption is that they must have electrically closed them in order to test continiuity and subsequently found the right valve wire to be continuous. Or were the control relays (latched) closed without having to move them to the closed position for the continuity test. . . .?

The final report may perhaps answer that question.


Green-dot

[Sunfish]

Now for a new fact, a rumour and an apology.

BA have NOT made a claim on it's insurance policy for the hull loss yet. The insurers have of course been advised, and the aircraft has been assessed as a hull loss. The reason for not yet submitting the claim is that they still don't know what caused the incident - so the prospect of cross claims between Boeing's, Roll's Royce's or other insured entities is as yet unquantifiable.

The theory on the part of the floor at Lloyd's (as of last week) is that it ain't the fuel that caused it, it's "RF interference". The theory goes that after George Brown was dropped off at Heathrow, his driver, or some other entity either deliberately or accidentally forgot to switch off their "jammer" and that the aforesaid vehicle on it's return journey may have gotten within "range" of the affected B777 and caused the mischief.

According to the Lloyd's rumour, the "jammer" in Brown's convoy apparently does very, very considerably more then just "jam cell phones", doing bad things to guidance and detonation systems over short ranges and I will leave it to your imagination to work out what that means.

IF (and it's a very large "if") this theory has any credibility, then the AAIB and HMG are in a pickle since this stuff is classified, and a suitable cover story will have to be constructed.

So my apology to Green Dot, Lloyd's currently agrees with you, although I'm buggered if I can think how such a signal would get into what is almost a perfect Faraday Cage.

[wymonwold]

The AAIB report does not state the actual fuel quantity that remained contained within the fuel tanks after the accident. A significant amount of fuel spilled from the aircraft at the crash site but what is a significant amount? If a hundred kgs of fuel spilled, it would still be significant due to the fire risk.:suspect:

[airfoilmod]

Surprising no one considered that here. Insurors can have Subrogation clauses, that distribute responsibility and financial payments. I may as well put forward a thought I had as an offshoot of GreenDot's persistence and technical corroboratives. My intuition tells me if EMI, it is most likely Microwave. Though I didn't fly FAC, many friends did in Vietnam, I have heard much of interest about "Interrogating and Painting" while flying fast movers that were unarmed in forward sectors. I think that Fuel is still the immediate cause here, but perhaps not the procuring cause. I can't escape thinking about how ILS information is acquired and processed; this was a low level incident, so Glideslope and R/ALT. are obvious sources that could have been corrupted to the extent that the aircraft may have thought it was above/below glideslope or had actually landed, R/Alt. Or Even Tyre buttons telling lies. Again, thanks to Sunfish and Mr. GreenDot. Hats Off.

Airfoil

[PBL]

Sunfish tells us of the latest rumors at LLoyds.

First, a word of caution. Insurance people do not necessarily know any more than anyone else does, and they are not necessarily technically trained (indeed, most of them whom I have met aren't technical people at all).

Second, some sort of EMI effect has by no means been ruled out by any info available to us. Neither has the manifestation of a digital logic error of some sort (the EEC has of the order of hundreds of thousands of lines of code (LOC; a kLOC is a thousand LOC). The crude average error rate for safety-critical systems is of the order of 1 error per kLOC. The very very best measured rate of which I know in a largish system with hundreds of thousands of lines of code is 1 error per 25 kLOC. That doesn't mean the errors show up in any safety-critical way, but that cannot be ruled out either).

Concerning EMI with a source outside the aircraft, let's do a sanity check on this.

If somebody on the ground is EM-radiating in such a way as to cause well-shielded systems some 700 ft in the air to go haywire, the question is how that could happen without any other effects being noted. If that signal is broadcast, and it happens to be on a road, then lots of less well-shielded auto electronics are going to go crazy also, let alone stuff in buildings. Which means that if it is somebody travelling down the road, lots of people within 700 ft (oh, let's say 1000 ft) are going to experience problems with their cars and their company electronics. And our intrepid journalists would have had that all over the front pages within a day or two. Didn't happen. So not a broadcast.

Suppose it was directed. Then somebody would have been pointing a strong EM source at or near a landing aircraft for some reason. I think there is no chance at all that anybody working for the PM would have done such a thing.

External source does not cover the EMI hypothesis fully, of course. There is also the possibility of EMI from within AC systems themselves, which is the case which Green-dot has been pursuing, since he knows of instances.

PBL

[airfoilmod]

From within the A/C itself, unrelated to systems or Gordon Brown. Some extraneous pulse that affected FCU or FADEC, or the Spar Valves logic master. Or cables (conductors) phasing from energy fluctuations induced by exterior (transmitted) energy. Would be so nice for a crumb from the chaps.

[Desk Jockey]

BA have been paid out and the hull is now owned by the insurance company. Source- BA staff briefing.(Weeks ago, as previously reported here.):ugh:

[Green-dot]

Quoting Chris Scott:

What I have deliberately left out so far, in accordance with the AAIB Bulletin, are the spar valve control relays that you are concerned about. Are they relevant? You say that they are downstream of the run/cut-off relays, and are directly controlled by the fire handles. This seems to be at odds with the AAIB’s explanation. Are you saying that they are normally sequenced to the valve-open position by the run/cut-off relays? Are you proposing that they are possible candidates for RF interference?

Chris,
Yes, the spar valve control relays are relevant. The AAIB wrote that, i quote: "Examination and tests of the wiring identified that, in the case of the right engine, the valve CLOSE wire from the run/cut-off relay was still continuous." (Here i assume they mean the spar valve circuit not the engine shut off valve circuit because the header on page 5 refers to spar valves not SOVs)

Without reading anything more into it than what the AAIB wrote here, this could imply that from the run/cut-off relay (which is inside the right power management panel, part of ELMS), the CLOSE wire downstream, right up to the spar valve, had continuity. The relevance here, with the run/cut-off relay in the OPEN position, is that in order to be able to measure continuity on the CLOSE wire right up to the valve implies that the path has to go through the right spar valve control relay which had to be in a closed position in order to measure continuity right up to the spar valve.

Normally, under the conditions described by the AAIB, one would expect that control relay to be open, similar to the run/cut-off relay. They are in the same path (the control relay somewhere between the run/cut-off relay and the spar valve). The fire switch, when selected from "Normal" to "Fire", isolates the above mentioned relay circuit. The wire making the "Normal" circuit is directly connected to the control relay, not the run/cut-off relay. When the switch is selected to the "Fire" position it isolates the "Normal" relay circuit and feeds a close signal directly via a second (separate) wire path to the spar valve if all functions as expected (to be clear, this is the pre SB 777-28-0025 situation valid for G-YMMM as the AAIB explained and makes sequencing important). The 28V dc power source is connected to the fire switch. With the fire switch in "Normal" position, power is supplied to the normal run/cut-off circuit and the second wire path is isolated.

The AAIB, since they mentioned the run/cutoff relay, therefore did not refer to measuring continuity in the second (switch in "Fire" position) wire path but in the "Normal" wire path, which, as explained has to inevitably go through the control relay.

Regards,
Green-dot

[Sunfish]

Desk Jockey:

BA have been paid out and the hull is now owned by the insurance company. Source- BA staff briefing.(Weeks ago, as previously reported here.)

My sources are names at Lloyds who have to put their hands in their pockets if their syndicate has a share in BA's policy.

As of last week my sources tell me they have not had to put their hand in their pockets yet, and they would know.

BA has certainly advised Lloyds. The hull has been assessed as a total loss, however my information is that it will be necessary for the insurers to receive the accident report before a claim is made and accepted because of the possibility of cross claims against the insurers of the Beijing airport (if its fuel), Boeing (If its a system stuffup), contributory negligence by BA or Rolls Royce.

However it's all only rumour and I wouldn't know. As for lloyds, yes they are merely insurers, but the have an infinitely subtle and world wide intelligence network made up of every name on the floor. I would not ignore Lloyd's.

[Southernboy]

It seem entirely logical that insurers will withhold payment until the AAIB report is published. There will be clauses in the policy that depend on cause.

If, for example, negligence were to be proven, the insurer will either limit payment or block it altogether.

[PBL]

Sunfish,

I don't wish to question either your contacts or your very interesting information about what is going the rounds in the insurance business. But we must distinguish between effects that obey the general laws of terrestial electromagnetic physics, and magic.

....... getting way beyond my competence, there is stuff apparently being used in Iraq to defeat electrically detonated IED's. GPS can also be jammed,


You are right about both of those. It is well known how easy it is to jam GPS signals. I suspect you can buy jammers for a couple hundred dollars at most. It is also well known (if one reads Aviation Week, for example) that the military in Iraq is deploying electronic warfare devices to disrupt IED control. However, both these phenomena consist of disturbing radio signals. Causing a disturbance in well-shielded on-board electronics from a point at least many hundreds of feet away is something else entirely.

To employ an analogy I have used before: it is one thing to ruin the picture on a television set by turning on your vacuum cleaner; it would be something else entirely to get your vacuum cleaner to change the channels.

If there was any EMI coming from outside the aircraft, it remains to be explained how it can have happened without leaving any trace at all in the highly-instrumented and recorded electrical and electronic control systems of the aircraft. If you explode a grenade sufficient to collapse your garden shed, it leaves a hole. No one would mistake such an event for, say, your shed collapsing spontaneously. Exactly the same is true of electromagnetic radiation. And the AAIB has said they have seen no control anomalies on the recordings.

we are not talking about jamming the whole of London either, merely about 500 feet, I would guess, around a moving vehicle.


We are not talking about jamming, period.

As for lloyds, yes they are merely insurers, but the have an infinitely subtle and world wide intelligence network made up of every name on the floor.

I take it that none of the members of this infinitely subtle and world wide intelligence network, or any of their friends, has yet seen fit to spread a rumor of a physically plausible causal mechanism by which this disturbance may have occurred.

PBL

[PBL]

I am glad to see forget chip in with some experience. A decade ago I was chatting about Elaine Scarry's proposal that TWA 800 might have been brought down by stray radiation from a nearby P3 with a colleague of mine in the Physics department. He supposed that it all depended on how well-cooked the U.S. Navy likes its airplane crews.


.... but what about chips on pumps, valves, etc., out on the engines which, as somebody pointed out, are not really Faraday cages...

People have said all kinds of silly things on this thread and this may be one. Check the picture earlier on this thread (or some other) of the box in which the EEC sits on the engine. A better Faraday cage you will not see, unless it's an airplane hull.

Please also note, as bsieker and others have repeatedly pointed out, that any purported EM explosion left no visible traces in a highly-instrumented system!

Engineers who deal with electromagnetic radiation have spent most of their education learning how to stop the bits they work on being influenced by EM radiation from bits other people have worked on (or the sun, or more recently the background radiation from the earth and from cosmic rays). It's been like that now for, dare I say it, over a century. It didn't start with the invention of cell phones!

Please see forget's next post for a link to an EM test cell operated by QinetiQ. Note (a) how big the kit is, and (b) how close to the aircraft it must be to get suitable field strengths. Take it seven hundred feet away (the altitude of the purportedly affected aircraft) and it is going to have to be at least 50 times the power of what you see in the brochure, which amongst other things means a lot larger.

I believe some sort of EM interference testing of this sort is a requirement for aircraft certification, but I can't actually put my hands on either the requirements or on any B777 EM testing history at the moment.

PBL

[forget]

For those with no idea of the electronic hammering aircraft are subjected to as part of RF Immunity Certification a commercial test facility may be of interest. Note that these very expensive, and very large kits are to prove aircraft immunity. So, to produce something even bigger, but highly mobile, capable of breaking through that immunity, is waaaay beyond the wit of most. In fact, I'd say it was impossible. :hmm:

See HERE (http://www.qinetiq.com/ix/defence/test_and_eval_and_training_support/emctesting.Par.70779.File.pdf)

[airfoilmod]

If your vacuum is controlled by Infrared, like your Television, the maid may easily change your viewing pleasure. The Theory, as I see it, isn't about motors and static, but relays, slaves, (solenoids), and CONTROLS. Also, Radio seems to be the sole player in this one-up EMF discussion. As fragile as some postulated theories seem, no less fragile is a proffered indignant defense. It's all interesting, but to me getting defensive is annoying.

Would you be more engaged if this were re: Airbus? It has a computer "in the Loop". Introducing microamperage into 28vDC and with attendant interfaces is quite a project. Beyond my ken anyway. I'd be disappointed if this were moved to Tech, lomapaseo and Greendot along with Chris Scott get way over my head as it is. Feel a whole lot more like Watson than Holmes.

I prefer the Microwave Tx approach, more powerful and versatile. Where are the Relays for the Spar Valves? How close to the composite flooring are they? For that matter, how close to each other are they? Fuel seems to be the direct cause, the rest of the story is yet to be written I'd say.
As an aside, can any theorist connect the forward cargo hold Fire Bottle UA fiasco to Control/Interference issues-Surprising "fleetwide attention" is always interesting.

[FullWings]

Maybe someone with engineering schematics could comment but I was under the impression that the 777 spar valve sensing system has three states, i.e. Open, Closed and In Transit. If this is true, then a record would have been left of any malfunction and it would have been picked up quite early on by Rolls/Boeing/the AAIB, etc. As this doesn't seem to be the case, maybe the cause lies elsewhere?

[airfoilmod]

The failure happened very late in the recording cycle. What happened (if anything) could have gone unrecorded or corrupted by the signal folks are discussing. (Relative to FDR). Recorded or not, Flight Record isn't available to those outside, precious little Data from the chaps. Gawrsh, there's no film in the camera, Syl.

[lomapaseo]

No reason to be cycnical. If the data was corrupted or missing the AAIB would have stated as such in their releases. Also I haven't heard a thing about mysterious on-board happenings to such things as the Pa's, passenger mobiles, etc. or other consumer goods found in the cabin. One would have thought that at the very minimum all the address books and hard drives in these passenger electronics would have been the first to go.

[Sunfish]

It's only a rumour from lloyd's and it is rumour I couldn't work out how it's possible either....but that's the current rumour.

As for physics and impossibility.... my brand new Motorola Razr3 has come back to life after an accidental dipping in salt water!

[snanceki]

What is surprising to me is the FACT that the AAIB have failed to issue an updated statement for some time now.

This suggests that they really don't know what happened or alternatively they do but consider the chances of a repeat event to be EXTREMELY low.

There appears to be consensus that the engines were partially starved of fuel almost simultaneously.

So what are the most likely causes.
1. Spar / engine valves PARTLY closed.
2. Inadequate fuel in the wing tanks.
3. Inappropriate fuel.

1. Spar valves partly closed.
Possibility of some form of EMI (airborne) or power spike (wiring), likely emanating from adjacent systems on board.
However I would expect that the spar valve and engine cut off valve position to appear on the DFDR (due to importance of these devices on engine performance).
In which case this possibility is easily verified or discounted.
If indeed this valve was only partly open I would have expected the finding to be published even if the cause was still undetermined due to the significance of the event and the possibility of a repeat due to an unknown mechanism.

2. Inadequate fuel in the wing tanks.
Mechanism suggested (failure of scavenge system from CWT to wing tanks (due to temp/design/sensing) but the problem here is why the fuel level system did not warn (or subsequently record on the DFDR) that the wing tanks were empty. The AAIB stated that sufficient fuel was indicated as being on board.
Both engines experiencing the same problem within 8 secs is troubling since differences in engine fuel consumption and initial wing tank level / tank capacity are unlikely to be exactly the same. However this could be accounted for by the significant attitude change of the aircraft during finals.

3. Inappropriate fuel. The -57 C freezing point spec of the fuel taken post the event is interesting. Why was the spec so "good" and could this be linked to pump cavitation? Incorrect fuel would have impacted other aircraft unless the combination of fuel and engine type didn't mix well but why did the problem occur at the end of the flight rather than take off etc when the engines are under greatest stress.

I don't discount any of the possibilities but IMHO the faulty fuel monitoring system (incorrect calibration?) appears marginally the most plausible.

I suspect that the AAIB must be struggling even with the wealth of data around them or more likely confident that this really was a 1 in X million aircraft specific event, otherwise I would have expected an updated report by now.

Guess we will have to be patient!

[PBL]

What is surprising to me is the FACT that the AAIB have failed to issue an updated statement for some time now.

You shouldn't be surprised. It is standard procedure. ICAO conventions require a preliminary report within 30 days and a final report, and that is what most accident investigation agencies do, AAIB included. It is very rare for any information to be made public in between, except when safety issues are identified in the course of an investigation, in which case safety recommendations dealing with these issues may be issued.

There are very good reasons for this. If you publicise what you are *thinking about*, then all those people and organisations who feel they might be disadvantaged by your potential conclusions will activate their lobbying machines to try to get your ideas changed. This happened most notoriously during the AA587 investigation, and prompted a public reprimand to the involved parties from the NSTB Chairman.

This suggests that they really don't know what happened or alternatively they do but consider the chances of a repeat event to be EXTREMELY low.

This suggests only that the AAIB are holding to their normal procedures during investigations.

BTW, trying to come up with a list of all possibilities and reasoning about them is a very good exercise. It is one of the skills that I try to encourage students to develop in my course on causal explanation. It is a skill that requires considerable experience and practice, as well as apparently (dare one say it) a certain amount of analytical talent. (In mathematical terms, one would speak of devising a *partition of the space* of possibilities, "space" here meaning simply "set".) The main trick is not only to make sure you have included everything, but also to learn to apply some self-checks to see if you indeed have everything. The trick which comes a close second is to devise a partition that is explanatorily fruitful, that is, in which each category leads you down some relatively narrow set of possibilities. The third trick is to know when you don't have enough information to be able effectively to apply tricks 1 and 2.

I don't know how you judge the likelihood of your three "likely causes", but I think it is obvious to both you and me that this is by no means a complete enumeration of the possibilities at any level of generality. So I am not sure how it helps anyone.

PBL

[snanceki]

Hi,
Thanks for your reply.
The key difference on this investigation is that we have an almost intact aircraft rather than wreckage spread over a wide area / difficult terrain.

So either the cause is already known and the paperwork is just being completed or something very strange occurred. The absence of actions (couple of exceptions) to other aircraft of the type/fleet is also interesting suggesting it is aircraft specific.

My list was of course not exhaustive and all the "debate" on the subject has been undertaken with very few facts.

However, with the wealth of interest and the various "theories" that have been put forward by individuals with diverse backgrounds I still remain surprised that the AAIB has remained as tight lipped as they have on this occasion.

Of course doing so is good practice but the very nature of this event when combined with the fact that the aircraft remains largely undamaged I remain surprised that there has been some further interim statement.

Just a personal observation that may or maynot be indicative of something.
OK. It doesn't add anything specific but does anything else put forward on this site?

As I said we will have to be patient. Only then will we be able to draw conclusions and see whether any of the theories put forward prove to be anywhere near (by chance) correct.

[HotDog]

Snanceki,

However, with the wealth of interest and the various "theories" that have been put forward by individuals with diverse backgrounds I still remain surprised that the AAIB has remained as tight lipped as they have on this occasion.

Doesn't surprise me one bit. Forty pages of repetitive nonsense by armchair "experts" has proven to be absolutely useless to the AAIB and like you say, wait for the official result of the investigation.

[PBL]

Forty pages of repetitive nonsense by armchair "experts" has proven to be absolutely useless to the AAIB

I don't agree. I grant you it has been repetitive, but I have read some extremely sophisticated analyses of fuel physics and chemistry which no one can reasonably regard as "nonsense". As well as some interesting facts about EMI in fuel subsystems on other aircraft. We have also produced a Causal Control Flow Diagram of the EEC environment for RR-powered B777's, which as far as I know does not exist elsewhere, as well as a fuel pathway diagram which is available to anyone who wants it. Neither of these are nonsense; they are, to the contrary, useful, even essential, analysis tools.

I understand certain people may not be interested by any of this, but then one wonders why they would even be reading this thread.

PBL

[slip and turn]

Well I am not surprised but I am slightly annoyed that the AAIB has not released much information publicly about the cause.

As for the insurance angle, because large wodges of their cash are involved, it is often worth heeding 'medium-matured' rumours in the London Insurance market, and if any significant part of that market are settling on radio interference as one cause (it's the third that get's you, right?) then at this stage of the game I would heed it, because later on, if any element of it is true, I would expect it to have been sanitised for the annuls.

To lend context to my thoughts, I am currently reading Peter Wright's "Spycatcher" after somehow having previously failed to open it.

In the book I haven't come across airfoilmod's elusive word "tramp signal" yet, but Peter Wright was a practical electronics expert (a government bugger!) and his book is the kind of place I'd expect to find a word like that used to discuss some kind of feedback phenomenum perhaps, which he might have learned how to exploit 50 years ago - so goodness knows what's possible in SIGINT / problematical to civvie circuit designers in 2008 ... I predict that the state of both these arts will be for a time reconciled in the circuitry of the new Airbus military tankers (select customers only) :p

[HotDog]

PBL, none of those high tech hypotheses that you are referring to are of any use unless they pinpoint and prove the cause and effect of the subject. I agree there have been some sophisticated discussion put forward on fuel physics and chemistry which would be better placed in the Flight Testing forum.

[PBL]

PBL, none of those high tech hypotheses that you are referring to are of any use unless they pinpoint and prove the cause and effect of the subject.

Granted. I just enjoy reading the contributions of people who know so much more than I do about a subject which is relevant, even though I have no idea at this point whether any of it is causal to the accident. I also really don't care that much where such discussion takes place, although I understand that others might have preferences.

PBL

[Chris Scott]

Quote from snanceki [Today, 07:53, currently #796]:
As I said we will have to be patient. Only then will we be able to draw conclusions and see whether any of the theories put forward prove to be anywhere near (by chance) correct.
[Unquote]

Err… yes. To paraphrase part of something I think I was saying yesterday; those of us amateurs, who are mildly obsessed with trying to find a plausible cause for this thankfully non-fatal accident, find ourselves in the position of would-be diners at a top restaurant; aperitifs have been served and menus distributed, but we are still waiting for the Maître d’Hotel to come and tell us what the plât du jour is, and take our order. As time passes, and there is no sound emanating from reception or cuisine, we start to wonder WIHIH…

[CAAAD]

Chris Scott

I have not contributed to this thread very much but, like most, have been impressed by its general tone and interest.

But, like you, I do begin to scent a whiff of something peculiar in the investigation.

A Large Transport Aircraft was written off in dramatic circumstances three months ago, and nothing seems to have been done to prevent it from happening again. No ADs, Recommendations, ASBs, nothing.

Even if it had been 100% crew error we would have expected to hear a reminder about this and that in the Flight Manual.

I agree that the professionals should be allowed to get on with the job, but it is not acceptable to keep the travelling public let alone the industry uninformed for such a long time. Perhaps a bit arrogant in fact.

AAIB owe us an update ASAP.

[Bis47]

IMHO, one of the immediate recommandations that should be given :

Publish the best glide speed in typical flaps/landing gear configurations, and train the crew to react to full engine failure.

Because, for sure, let us putting it straight : the handling of the event was less than perfect and that is something that should be improved. I presume some chief pilots are already working at it ...

AAIB doesn't need to know the causes, for that first step.

Yes, it does happen, from time to time ...
Boeing should be able to provide the data ...
Crews need some numbers, some practice, too.

Long overdue ...

[michaelmarsh]

It's irrelevant. Large transport aircraft are not intended to glide in the landing configuration. In the case of BA038, it looks as if the approach path was extended to the utmost- any other procedure would have surely resulted in an even shorter landing-in the road.

[lowflare]

There is a best glide speed for everything flying including a brick...And 113 KTS is far from it because it is closer to a stall and probably beyond it. B777 is not a glider but still is a machine heavier than air, on which all aerodynamics laws apply. Judging (not sentencing) the crew action bringing it almost to a stall on A/P it could have been better handled and surely more followed up by Boeing in some AD, Urgent circular or something similar...

Lowflare

[CAAAD]

I was not presuming to comment on the conduct of the flight after the problem had been revealed.

My point is that it is high time that we had an update on the circumstances that led to the problem, together with preliminary recommendations to prevent a recurrence.

[michaelmarsh]

At 600', with less than 60 seconds to touchdown, no power and in the landing configuration, the ONLY action that won't make the situation worse is to hang on to CLmax (the edge of stick-shake) and if landing is inevitable, make the best touch-down you can. This is the standard procedure for extreme down-draft or terrain proximity recovery i.e. max lift and don't change the config. This action appears to be exactly what occured in the case of BA038.
To suggest that one might train crews for an event that has happened once in the history of jet aviation and is, as yet, unexplained, is to fail to understand training priorities.
Why not wait for the findings of the Enquiry?

[LLuke]

Without knowing the details of what happened, they did (imho) the right/ most natural thing. It all happened in a couple of seconds at low altitude (500 ft?), waiting for the damned engines to respond, going for minimum decent rate, but obviously not allowing it too stall. At the altitude it happened, in landing configuration, it probably wouldn't make sense to dive to the best glide speed. (just my opinion ofcourse) I am sure it has been replayed on the sim already many times...

If both engines go in relax mode at 4000 feet, it is obviously a different story.

[Green-dot]

What if the aircraft would have made it to the runway with this failure and had landed safely without any structural damage? Obviously the aircraft would have been taken out of service for some serious trouble shooting and would have undergone every operational and functional test available.

What value should be placed on the fact that in the current situation (with the aircraft relatively intact, minus engines and landing gear) it is impossible to do engine runs or landing gear related tests, hence impossible to check their functioning in conjunction (or interfacing) with other aircraft systems? Could something, due to their absence, be missing in the scheme of things? Bench testing components may not always reveal the problem which might have surfaced if the component was installed on the aircraft and tested.

If no definite cause is found, would it be worth to (if only temporarily) repair and complete the aircraft to a fully functioning vehicle to do such tests to possibly try and find the exact cause that way?

Regards,
Green-dot

[Bis47]

It is definitively false that this is the first occurence of total engine failure in an airliner ... Quite a few occured in the past ... This one is not the last one. Especially if the cause of the accident remains unsolved.

Considering the very special circumstances, and the basic limitations of a two pilots crew in such an extreme emergency, no one should blame the pilots. But ... that is not a valid reason to affirm that the aircraft was flown the best possible way to achieve a maximum glide and a safe touchdown.

The problem, as far as I remember, occured at 700 ft, that is one full minute before touchdown, not "a couple of seconds". The autopilot flew the aircraft down to the stall speed while - we may suppose - both pilots were busy trying to recover some engine thrust. This is the very classical trap indeed : both pilots busy with the hardware, nobody flying the aircraft.

Hard to believe that the auto-pilot had any kind of airmanship to deal with such an unusual situation. Bringing an aircraft to the stall, and letting it fall from 200 ft to the ground without any chance for a minimal flare is "zero airmanship". The auto-pilot was not programmed to achieve the best possible glidespeed, or to let go altitude for speed and recover the initial loss later during the flare ...

From a pure aerodynamic standpoint, we may assume with confidence that it was possible to bring the aircraft over the airport fence with a much lesser rate of descent. Same energy, but not the same trajectory.

Just basic piloting skills ...

Someday - perhaps - somebody will replay the event in a simulator, and experiment some other course of actions. Perhaps ... it is done already?

It should be, and whatever the results, they shoud be made public.

[PBL]

Biz47,

let me try to summarise.

Your question: was the trajectory optimal?
The current answer: who knows?

The next question: who cares? Is this going to happen again?
The current answers: few people; and, who knows?

The next question: should resources be devoted to finding and teaching the optimal trajectory, should this kind of thing ever happen at 700' again?
Answer: I dunno. Does anybody?

PBL

[M.Mouse]

Someday - perhaps - somebody will replay the event in a simulator, and experiment some other course of actions. Perhaps ... it is done already?


Indeed it has been done already.

Your post sums up entirely the basic facts of the last minute of that flight accurately and without emotion, something which has been notably lacking in many comments regarding the flightpath and control, or lack of, in the final 700'.

I am sure I would have been as stunned and possibly overwhelmed as surely 99% of us would have been given the sudden and almost complete unexplained loss of power in that last minute.

I hope and suspect that analysis of the flightpath and alternative scenarios will be made public in due course.

[LLuke]

I can't comment too much since I really didn't read this thread. I didn't know it was on AP all the way. However I don't think that diving to best glide from 700 ft would have resulted in better range. I always thought all airliners train 'loss of all engines'. Best glide obviously depends on weight. If it happens on short final, your speed isn't going to make a difference, you will *never* make it to the runway (unless you're in a glider a/c ofcourse). Don't even try it. Better spent some time with looking outside and find a nice place to ditch/land. That's exactly what these guys did; finding a nice place. Unfortunately that place is not always ahead of you. The time it took me to write this message is more time than the BA crew had to analyze what the f*ck was happening and what to do. The job they did resulted in a survivable outcome. If a better result was possible, I will be most pleased to read about it in my comfy arm chair, enjoying my drink, when the report comes out :-)

[16024]

Don't want to make any point at all regarding the correctness of the crews' actions.
Just to chuck in a couple of points about glide speeds.
Michaelmarsh says:
hang on to CLmax (the edge of stick-shake)
This will give a very high R.O.D. in the glide. It works in windshear or terrain avoidance situations because of the thrust vector in a nose up situation.
Min. sink is at about 1.1 to 1.2 Vs.
But here we are after best glide ratio, which is at about 1.3 to 1.5 Vs.
Lluke says:
it probably wouldn't make sense to dive to the best glide speed
Vref is close to 1.3 Vs so you are right there.
Glad I wasn't...

[M.Mouse]

I didn't know it was on AP all the way.

It was until the speed reduced to 108kts and the aircraft was about to stall.

However I don't think that diving to best glide from 700 ft would have resulted in better range.

Try it in a simulator it makes a significant difference.

I always thought all airliners train 'loss of all engines'. Best glide obviously depends on weight.

We do but usually from altitude where variations of speed are less critical and with far more time and distance to judge the descent profile.

If it happens on short final, your speed isn't going to make a difference, you will *never* make it to the runway (unless you're in a glider a/c of course).

I know for a fact that it has been done in a simulator (made the runway that is).


Sitting here we do of course have a luxury that was not afforded to the unlucky crew i.e. time and detailed analysis. For that reason it would be unfair and foolish to criticise them directly or indirectly. My statements are made with that qualification.

[LLuke]

Glad to hear all this. Guess it also depends on A/C type... On my type I fly with a typical thrust setting of 60-70% N1 depending on flaps setting, wind, weight, etc..; wont make it idle.

Maybe I will surprise my colleagues next time during recurrent (that's on the sim ofcourse) at 700 ft on the ILS :E

[Chris Scott]

There is much merit in the above discussion, if only to increase our understanding of L/D ratios, and their implication. We must recognise, though, that the crew must have been hoping for, or even expecting, a sudden recovery in the thrust shortfall. Double (all) engine failure is one thing; slow and/or partial engine acceleration is another.

One of the most relevant data strings that we would-be diners have been starved of so far is the actual thrust from TOD (top of descent) to the point at which the engines noticeably failed to respond to the autothrust command. This missing information has been a major inhibitor to our discussion from day one.

To spell it out: could this have been the first call for thrust above idle? In which case, how long had the problem been lying dorrmant?

[M.Mouse]

That question has been answered numerous times. It was not the first demand for more power.

[Chris Scott]

Thanks. Idle thrust always seemed unlikely, going into LHR. But what/who is your source?

[PAXboy]

Hhmmm, more criticism ...

:- of flight crew who had an unprecedented event and managed (with the automatics) to get everyone home alive when they had less than a minute to assess and react.

:- of the AAIB who cannot publish an update until they are SURE of what they have found. If they are working on something very serious, then anything less than a full explanation may cause panic in the travelling public. And the airline business is in a bad enough way as it is. If they are thinking of saying, "We have a bunch of ideas - and it could have been any one of them" then that will also cause a bad reaction. They do not 'owe' us anything, until they have something they can say with certainty.

And this thread was doing so much better of late.

[snowfalcon2]

Regarding the glide issue and autopilot.
I believe a modern computerized airliner already has almost all the data necessary to enable its computers to calculate in real time how it can best make the runway at the present thrust setting, be it partial thrust or zero.

The missing pieces of data are a) the airplane's lift/drag curve in various configurations, which the manufacturer should already know, and b) the wind gradient in the remaining flightpath until touchdown (usually the headwind decreases, which in this case is good and adds margin if we simply use the instantaneous wind data).

So in theory it's not that difficult to devise a software that automatically commands the autopilot into optimum glide mode at the first sign of a thrust deficit that prevents it from holding the commanded glideslope at Vref. In case thrust comes back, it would revert back to normal glideslope hold mode.

Whether that "auto glide mode" would be a truly useful feature or only serve to confuse the flight crew, is another discussion. Training would of course be essential. But a technical solution is certainly possible.

Just my $0.02

[Bis47]

The missing pieces of data are a) the airplane's lift/drag curve in various configurations, which the manufacturer should already know, and b) the wind gradient in the remaining flightpath until touchdown (usually the headwind decreases, which in this case is good and adds margin if we simply use the instantaneous wind data).

So in theory it's not that difficult to devise a software that automatically commands the autopilot into optimum glide mode at the first sign of a thrust deficit that prevents it from holding the commanded glideslope at Vref.

I fully agree. There might be as well a data base of obstacles, allowing for the best way to avoid obstacles ...

However ... I have a more readily available solution :

1. Just tell us where the best glide speed is, in % of the VRef. That percentage is a constant for a given aircraft configuration. For an aircraft with a high wing aspect ratio, in a normal landing configuration, I suspect that the best glide speed is somewhere between 90% and 115% of VRef (educated guess).

2. The pilots already have a fair idea of the wind gradient, because they know the instantaneous wind ... and the ground wind as well (being "wind minded" is a part of airmanship, isn't it?)

3. The stronger the head wind, the stronger the wind gradient (rule of thumb ...)

4. The stronger the head wind ... the higher the need to inscrease the speed somewhat above the best glide speed. Add one third to one half the head wind component. Additionnal benefit : the aircraft will be low sooner, where the head wind is less. Extra speed will convert back into distance. Extra distance covered in ground effect is a bonus, regarding drag reduction.

5. Let us know also for sure the real effects of partial flaps reduction.

6. The pilots can see the main obstructions, and deviate if necessary.

Nothing is sacrilege in such an emergency.
The first available software is in the pilot brain, its common name is "basic airmanship".

Perhaps there is a need to train from time to time for the maintenance of that precious software.

[FullWings]

1. Just tell us where the best glide speed is, in % of the VRef. That percentage is a constant for a given aircraft configuration.
That can be done but remember the BA38 was not a glider: the engines were still producing significant thrust, albeit not enough to maintain a 3deg slope in full landing config. This would have had quite an effect on the optimum speed to fly (whatever that was...).

A possibility would have been to take the aircraft into ground effect earlier but that's hard to model correctly as it is very dependent on the precise nature of the terrain and the aeroplane position relative to it. Simulators are good for many "what if?" exercises but they can throw up "false positives" unless you fully understand the limitations of the simulation environment and what the software is/isn't taking account of.

Nothing is sacrilege in such an emergency. The first available software is in the pilot brain, its common name is "basic airmanship".
Definitely!

Perhaps there is a need to train from time to time for the maintenance of that precious software.
Agreed, but in the context of this particular accident, I don't see much training value. A slight change of circumstances might lead to a very different method to achieve an "optimal" outcome; there's not much point in training for a 1x10-9 scenario when your resources are limited and you want to achieve the best "bang for your buck". There is also the point that, generally, we don't play out scenarios in training that have a crash as the most likely outcome, despite crew actions, for obvious psychological reasons! This is reflected in the Boeing manuals, where they only deal with a dual engine failure at height and effectively assume that you get one or both restarted...

[CAAAD]

PAXboy
A little sanctimonious. I thought society had moved on from unnecessary secrecy 'in case the Public misunderstand and panic ensues'.There should be periodic updates to keep us in the picture.
But we have had no news from anyone for two months now, which is unusual to say the least.
I agree that the tenor of this thread has been amiable, but being kept in the dark for so long on such a significant issue naturally leads to a bit of impatience and an impression that something peculiar is going on.

[Taildragger67]

CAAAD,

There should be periodic updates to keep us in the picture.
But we have had no news from anyone for two months now, which is unusual to say the least.

I disagree.

1. 'Periodic updates' do not give a full and complete picture (by definition) and so allow those who write headlines to publish banners which have little link to reality. Further, as an investigation progresses, things which may (earlier in the investigation) have appeared to be causal, may in fact not be whilst new things pop up; 'periodic updates' might therefore be contradictory or confusing if read in isolation.

The best 'periodic update' would be "we're still working on it and will let you know when we have the full story. Until then, naff off and stop making us waste time we could be spending on the investigation."

2. It is not at all unusual for investigations to stretch into several months (indeed, years). Not having heard anything for "two months now" is not unusual in the slightest.

[CONF iture]

In my view AAIB has been very (too) shy on basic data releasing:

1- Contrary to M.Mouse last comment, no mention of additional thrust request between TOD and 2 miles final. It has been rightfully suggested by Chris Scott here (http://www.pprune.org/forums/showpost.php?p=3943401&postcount=499) but that’s it.

2- How and when the APU door did open ?

3- "Fuel crossfeed valves indicated that they were closed …" but were they actually closed … ?

4- No mention of actual ZFW versus planned one.

5- No mention if the actual flight time of 10:35 was noticeably shorter than the planned one.

6- It is not clear if the planned flight level was restricted around 320 up to London, or if further steps were planned later on and if so where ?
Except from the early statement "to accommodate a request from ATC …" it is not clear if the crew choose to disregard a conservative low flight level flight plan computed to avoid "unusual low temp conditions compared to the average, but not exceptional"


I’m sorry but all these points are relevant when both engines fail to deliver almost simultaneously, especially in the last minute of a 10 hours flight … and following comments:

- "I don't discount any of the possibilities but IMHO the faulty fuel monitoring system (incorrect calibration?) appears marginally the most plausible" from snanceki

- "You don't get it much more serious than being OUTAGAS" from Spaz Modic

should not be taken lightly.

[Bis47]

After this one, I'll stop!

Thank You FullWings for showing some interest in piloting techniques considerations.


the engines were still producing significant thrust [...] This would have had quite an effect on the optimum speed to fly (whatever that was...).


Not significantly : Except for the wind, flying a jet at best L/D always provide the best angle ... whatever the thrust available. Best angle of climb, or lesser angle of descent. In a given configuration, flying best glide is simply flying minimum drag (in that configuration).


Agreed, but in the context of this particular accident, I don't see much training value. A slight change of circumstances might lead to a very different method to achieve an "optimal" outcome; there's not much point in training for a 1x10-9 scenario when your resources are limited and you want to achieve the best "bang for your buck".

Sure, there are priorities in training ... Shame on me ... I didn't care about dual engine failures when I was in charge, I have to admit. But today I would hate to be left without any information regarding this kind of situation. Or to let pilots of my company without any hints.

How much training is at shake? Some lectures, one hour in the sim, demonstrating typical situations ... Replaying the sequence of BA38 uses less than two minutes of sim time. That should be enough to get the facts ... No need of course to demonstrate 100% success in order to get a "pass" ...

Perhaps some kids are now becoming champions at that game on MS Flight Sim ... (Do not underhestimate the value of that king of cheap simulation ... as far as I know, some aircraft models are quite realistic, and the sofware is excellent)

Let us hope that dual engine failure never become "business as usual".

But the scenario leading to the discovery of an engine failure in short final is already quite standardized - CDA it is. After all, facing a lack of engine responsiveness at the very end of flight is the result of a standard CDA scenario : using idlle power (or so) from top of descent, the trouble, if any, will show up very late. Almost allways at the same point, i-e somewhere between 500 and 1000 ft AGL if pilots and ATC really work at it ...

Nowaday, I guess there is some stress in a 777 cockpit at the end of a perfect job done during a CDA ... Idlle thrust or so from top of descent, feeling the need for aditionnal thrust only a couple of milles towards touchdown. Will the engines spool up? Are there spooling up fast enough? Let us keep an eye on them, make sure they don't roll back! Keep on, threshold coming soon! Finally :D!

I guess that 777 crews todays are a little more suspicious about low temperatures, icing messages, exotic jet fuels and ... perfect CDAs ...

Any volunteer to replay flight BA38? Exactly, I mean, in the real world.

Good flights, Gentlemen!

[M.Mouse]

- "I don't discount any of the possibilities but IMHO the faulty fuel monitoring system (incorrect calibration?) appears marginally the most plausible" from snanceki

- "You don't get it much more serious than being OUTAGAS" from Spaz Modic

should not be taken lightly.

It has been stated ad nauseum that the aircraft had plenty of fuel on board. Around 10 tonnes I recall, certainly in that region.

Nowaday, I guess there is some stress in a 777 cockpit at the end of a perfect job done during a CDA ... Idlle thrust or so from top of descent, feeling the need for aditionnal thrust only a couple of milles towards touchdown. Will the engines spool up? Are there spooling up fast enough? Let us keep an eye on them, make sure they don't roll back! Keep on, threshold coming soon!

While in an ideal world idle power from TOD until 1000' would be the most fuel efficient method of descending it just does not happen that often in busy airspace and almost never will it happen into LHR. The emphasis on CDAs is reserved for the descent commenced at some point after the holding fix and that emphasis is due more to the reduction in noise for those below the flight path than for simply economy.

I guess that 777 crews todays are a little more suspicious about low temperatures, icing messages, exotic jet fuels and ... perfect CDAs ...

Er......no.

[PAXboy]

CAAADA little sanctimonious. I thought society had moved on from unnecessary secrecy 'in case the Public misunderstand and panic ensues'.
Last year in the UK, the Governor of the Bank of England said that, if the Bank had been allowed to arrange bridging finance for Northern Rock in private, then he was confident that the matter could have been dealt with. He wasn't and the bank was broken.

We have no means of knowing what would have happened if they had made a private arrangement. What we do know is that the British media are sluts for trawling anything that can be sensationalised. It is not the public but the Daily Mail that you need to be worried about.

It is quite possible that we shall never know the full reasons for this failure and loss, until it happens again. Consider how long it took to discover what was wrong with the Comet. For now, I am happy to wait and feel glad that I am not working on the AAIB team, or in any of the companies involved for there is no doubt that this is a very difficult, and intractable, mystery.

[snanceki]

@ M.Mouse
It has been stated ad nauseum that the aircraft had plenty of fuel on board. Around 10 tonnes I recall, certainly in that region.

Without double checking I believe you have overlooked two points.

1. The word INDICATED was used.
2. No statement as to where the indicated fuel was stored. i.e. Wing Tanks or CWT. The fuel needed to be in the wing tanks.

So why has this been left with so many possible interpretations.

Was there adequate fuel in the wing tanks? The wing tanks were undamaged plus detritus was found within the tanks (quite normal as I understand it). So how much fuel was in the wing tanks post landing?

They must know the answer to this question BUT for whatever reason they are keeping the info close to their chest! The CWT was damaged and therefore the amount in the CWT immediately prior to the event is unknown.

[bsieker]

It has been stated ad nauseum that the aircraft had plenty of fuel on board. Around 10 tonnes I recall, certainly in that region.

Without double checking I believe you have overlooked two points.

1. The word INDICATED was used.
2. No statement as to where the indicated fuel was stored. i.e. Wing Tanks or CWT. The fuel needed to be in the wing tanks.

The AAIB also stated that a significant amount was spilled, showing that there actually was fuel on board.

As you said, the tanks were checked.

Don't you think that the AAIB would have mentioned it if there had been a stark discrepancy between indicated fuel and the amount actually found in the tanks?

"Oh, there's no fuel in the wing tanks. Do you think that's worth mentioning in the report?" --- "Nah, just an unimportant little detail. It'll only confuse the public."


Bernd

[JamesCam]

Judging (not sentencing) the crew action bringing it almost to a stall on A/P it could have been better handled and surely more followed up by Boeing in some AD, Urgent circular or something similar...

Lowflare


I have no experience of landing heavy aircraft like that, but if I felt that the rate of descent was too high to allow a normal landing without breaking the gear, I would definitely try and put it down on soggy grass rather than a paved runway even if the impact speed was slightly greater. Braking effect of the grass would be greater, shock on point of impact less and the chances of fire greatly reduced. In summary, I'd probaby shorten the glide to land on the grass.

James

[Chris Scott]

Hi JamesCam,

The conventional wisdom, which I support, is that the hard pavement is better, provided its LCN is up to the job. The last thing you want is the wheels burying themselves into mother earth. Even on a smooth touchdown, the tyre pressures (over 200psi) of a large aeroplane mean that the latter is inevitable on grass, and this was the cause of most - if not all - of the damage to Echo-Echo.

[JamesCam]

Hi Chris, thanks for that. I take your point. I suppose in many ways the collapse of the gear under extreme heavy landings is probably the most signifcant factor in absorbing the shock of the impact?

James

[Chris Scott]

Hi James,

Yes: the gear definitely absorbs vertical kinetic energy, preferably without collapsing in the process! I'm afraid you are now taking this topic well beyond my expertise. I have suggested that the aeroplane would have remained intact, had it contacted the runway at the same VS; but the truth is, I don't know. The AAIB has not stated the VS.

If the landing gear collapses on a runway, my GUESS is that there is likely to be less secondary damage to the rest of the aircraft than on grass. When the engine nacelles strike a runway, they will not dig in; likewise the nose or a wing tip.

Chris

[infrequentflyer789]

2. No statement as to where the indicated fuel was stored. i.e. Wing Tanks or CWT. The fuel needed to be in the wing tanks.

So why has this been left with so many possible interpretations.


Erm, the AAIB state quite clearly that the total fuel was:

distributed almost equally between the left and right main fuel tanks, with a minor imbalance of about 300 kg.

Yes, they state quite carefully that this is indicated fuel, because that is what they have the precise figures for. They don't have precise figures for actual fuel remaining in the wing tanks because it ended up all over the crash site through the open spar valves. That's how we know the fuel was there, we don't know exactly how much, but if the indicators were right it was ten and a half tons.

Unless you are contending that someone filled a fire appliance with fuel instead of foam ?

[demomonkey]

Sorry to bring these comments up again

So in theory it's not that difficult to devise a software that automatically

As someone who has a Software Eng Degree, several years commercial experience with industrial strength, fault tolerant, 'mission-critical' applications (I hate that phrase too) and an ATPL I feel that it would be almost impossible to create such an application.

For a start that the aircraft would need to be fitted with so many sensors that its drag curve would be a vertical line. A database of global obstacles would be so onerous to create that it would be Terabytes of data and commercially un-maintainable.

In short we forget how wonderful the MkI eyeball attached to the Mk10 human brain actually is to sense inputs (topography, wind, aeroplane charactersitics) and meld these with prior experience to produce a safe and effective plan.

I do agree that the best the aeroplane's onboard computer can do is target Vref (or stick shaker = Vref + a bit for the technically minded) for the instantaneous inputs it is receiving.

Hope this helps. Oh and another well done to the BA 777 boys and girls! Those MkI eyeballs and Mk10 brains sure do work well under pressure!

[snanceki]

@bsieker and Infrequentflyer.

I was merely stating the importance of the words Indicated and that the fuel needed to be in the wing tanks in order to supply the engines.

I'm neither supporting or denying that the aircraft ran out of fuel due to lack of fuel in the wing tanks. I'm just surprised that the wording used seems to have left this whole issue "open".

Yes the spar valves were found open but is it reasonable that there was the correct amount of fuel in the wing tanks and that it all leaked out post accident or did the spill come from the damaged CWT where the fuel was not supposed to be?

We just don't have the FACTS to answer these questions but I find it hard to believe that the AAIB doesn't.

The spar valves were open but this doesn't define the rate of leakage from the wing tanks post event which would be influenced by wher/what the leak was.

We will just have to be patient.

[infrequentflyer789]

Hi James,

Yes: the gear definitely absorbs vertical kinetic energy, preferably without collapsing in the process! I'm afraid you are now taking this topic well beyond my expertise. I have suggested that the aeroplane would have remained intact, had it contacted the runway at the same VS; but the truth is, I don't know. The AAIB has not stated the VS.

If the landing gear collapses on a runway, my GUESS is that there is likely to be less secondary damage to the rest of the aircraft than on grass. When the engine nacelles strike a runway, they will not dig in; likewise the nose or a wing tip.

Chris

I don't know the design details on the 777 specifically, but from what I do know of design of similar sized craft, the oleos will absorb vertical forces up to a point, but when it comes to horizontal (drag) forces - eg. when landing on grass - the gear is supposed to detach. My understanding is that it also is supposed to detach cleanly and without damaging the main spar and fuel tanks (for fairly obvious reasons).

In this case, it looks like the gear (at least on one side) punched up through the wing structure prior to detaching due to the drag. From that observation, my fellings are that either:

a) the 777 gear didn't detach cleanly (as, I think, it will have been designed to do)
or
b) the vertical forces were way over the capability of the gear, at the point of impact, before drag forces could detach the gear - i.e. VS was beyond the limits of the gear.

If it's (b) then my guess is that the vertical forces will be the same on a runway (maybe more since soggy grass might have absorbed some) and you would still punch the gear up through the wing. The gear may or may not detach after that, but the damage to the wing would have been done - my guess is that the wing (main spars) is the major damage that has written the hull off. Also, while the plane (with gear or not) might slide easier on the runway, wet grass would be a lot less likely to create sparks, which might be a consideration when you've got fuel everywhere...

Overall, I think it is quite possible that where they came down was the best place (from a survivability point of view) to do so for that impact.
However, I'm not sure any of this is a useful comparison since I would be very suprised if, in the same circumstances, you managed to extend the flight as far as the runway and still have the same sink rate (but now I'm well beyond my expertise...).

[infrequentflyer789]

@bsieker and Infrequentflyer.

I was merely stating the importance of the words Indicated and that the fuel needed to be in the wing tanks in order to supply the engines.

I'm neither supporting or denying that the aircraft ran out of fuel due to lack of fuel in the wing tanks. I'm just surprised that the wording used seems to have left this whole issue "open".


I don't think it has been left as open as that, whilst they have been careful to say that the fuel was indicated to be in the wing tanks, they also haven't said (or implied) anywhere that those indications are suspect. . We know that they cannot have figures for fuel-on-landing from post-landing measurements because of the leakage.

Also in the report they [I]have said that the leakage was through the spar valves, which means (if I understand the fuel system correctly) that the fuel was from the main wing tanks, not the CWT.


both of the engine spar valves were found to be OPEN, allowing the fuel leak evident at the accident site.

[snowfalcon2]

thanks for your comment. However, I still believe it's technically doable with reasonable effort.


In short we forget how wonderful the MkI eyeball attached to the Mk10 human brain actually is to sense inputs (topography, wind, aeroplane charactersitics) and meld these with prior experience to produce a safe and effective plan.


The words in (my) italics are the key here. Complete engine failure on final approach is evidently judged to be such an unlikely event that it's not included in training or simulator exercises. That was most probably a contributing factor to the fact that the airplane got near the stall and in that process irrevocably lost some dearly needed energy that might have taken it to the runway.

And as others have said, the flight crew is not to blame for it. They were faced with a situation no one had prepared them for and did what they could - saving the lives of all aboard, which in the end is all that matters.

[Jumpjim]

I would suggest that the loss of energy is due to the double engine failure rather than crew actions?

Jeez, you people....

[Green-dot]

Quoting infrequentflyer789:
Also in the report they have said that the leakage was through the spar valves, which means (if I understand the fuel system correctly) that the fuel was from the main wing tanks, not the CWT.


Fuel leakage through the open spar valves, with the boost pumps switched off due to the fire drill, means there must have been a considerable amount fuel in the wing tanks. With the aircraft coming to rest wings level (which can easily be verified from published pictures), wing lower surface at the location of the spar valves has a considerable amount of dihedral downward to the rib separating wing tanks from CWT.

The spar valves are located somewhere outboard of the main landing gear and just inboard of the engine pylons and certainly not at the lowest point in the wing tanks because of spar valve location further away from inboard rib, separating wing tank from CWT, and dihedral sloping upward toward the pylon.

From the last picture taken of the aircraft while still airborne (taken from the left looking up toward the aircraft), one can clearly see the lower wing surfaces. The location of the rib separating the center tank from the right wing tank is clearly visible. The wing lower surface of the right wing tank has a frost layer beween that rib and engine pylon, meaning there is fuel in that tank, at least all the way up to the engine pylon. However, inboard of the rib toward the fuselage, the lower surface of the CWT has no frost which means no fuel is in contact with the lower wing surface in the CWT, hence the CWT is virtually empty.
Frost is also present on the lower surface of the left wing tank but barely visible because the left engine nearly obscures that particular wing surface area. You will notice the absence of frost on the left side of the CWT as well though.


Green-dot

[Feather #3]

Ahh...when the "report" is finally published or someone REALLY knows what's going on, would you please publish a separate thread on the subject?

Trying to get some sense out of this one is a bit like herding cats! :ugh:

G'day ;)

[Bis47]

Sorry to bring these comments up again

Quote:
So in theory it's not that difficult to devise a software that automatically
As someone who has a Software Eng Degree, several years commercial experience with industrial strength, fault tolerant, 'mission-critical' applications (I hate that phrase too) and an ATPL I feel that it would be almost impossible to create such an application.

For a start that the aircraft would need to be fitted with so many sensors that its drag curve would be a vertical line. A database of global obstacles would be so onerous to create that it would be Terabytes of data and commercially un-maintainable.



Are you really aware of the state of the art in flight simulation? Very realistic 3D rendering of landscape and obstacle around airports an beyong. No need for terrabytes of data ...

Obstacles data base are actually used (and thus maintained) to compute landing and take-off restrictions, taking into account all the relevant factors. That kind of software is in use for decades by airlines or specialised service providers. Both data base and software are "accepted" for those computations by the authorities for flight dispatching. (Grey zone, in any way).

Any modern airliner is fitted already by an air data computer. No need for additionnal sensors. Thanks to GPS integration, and differential GPS technology, the position (of the antenna) is known within 10 ft horizontally, 30 ft vertically.

So what? When a runway is selected, a small portion of the obstacle data base should be querried and the data set made readily available for either or both :
- "visual" presentation to the crew (like in a sim)
- emergency trajectory computation by any decent computer. No need for terraflops here ... most of the computations should be done "well before" and the results stored, ready to be retrieved at the time the runway is selected.

As it was stated before, the only important missing information (today) is the "future" wind. The present wind is wel known, the ground wind (somewhat averaged) is known by the tower, but there is no direct link with the onboard computers. And no information "in between" - but good models are available.

I think that the available technology is not put to its maximum use ... Cost is the factor, but also lack of imagination.

[snowfalcon2]

I would suggest that the loss of energy is due to the double engine failure rather than crew actions?

Well, in my world an engine failure means a loss of the energy supply. Which is a good reason to use the airplane's remaining energy in the optimal way.

[PBL]

Biz47,

obviously you have no experience with designing or building SW, let alone SW for what would be a safety-critical application. But you presume to tell a self-identified professional IT person that there is a lack of imagination. There may be and it might well be yours. In a world which does not yet know how to deal with map shifts, building reliable, fault-tolerant, fail-safe systems which do some of those "simple" things you mention, to the required standards of reliability and fault-tolerance, is beyond state of the art.

PBL

[Bis47]

obviously you have no experience with designing or building SW, let alone SW for what would be a safety-critical application

I have some ... Not on a big scale, I do admit.
But I have some deep knowledge of flight operations.

I confirm that reliable sofware is already available for terrain visualisation and terrain avoidance :
- Flight simulation software (all levels, PC driven and higher) is amazingly realistic (terrain visualisation and aircraft handling) => allowing approved "Zero flight time" aircraft type rating (no perceived difference between simulator and real aircraft handling).
- Military software allowing flight at low altitude inside valleys ... relying more and more on stored terrain data and less on radar.

And of course, reliable software already exists regarding aircraft handling, allowing auto land in zero-zero, and flight enveloppe protection ...

Please note that airbus flybywire software does not require supercomputers : an Intel 8186 does the job pretty well. 20 years old technology ... much cheaper and much more powerfull hardware is available today - about 1000 times more powerfull ....

Note also that more than 10 years ago, cheap (but certified) general aviation GPS receivers already included some kind of advisory terrain elevation data (grid mora).

And ... today, the most popular general aviation GPS manage to display a nice picture of the runway, and some landscape features as well, when approaching an airport to land.

Where is the lack of imagination? Well nobody did imagine that a modern airliner would meet a dual engine failure on short final, so nobody saw the need for a software to handle that - neither for a pilot training actually. That is lack of imagination.

[cats_five]

I would suggest that the loss of energy is due to the double engine failure rather than crew actions?

Jeez, you people....

At the point at which the engine failure became obvious, the plane had X amount of energy - some was potential energy, some kinetic energy. The need was to manage this energy to conserve as long as possible, which means flying the plane in the most efficient way e.g. at best L/D. Unfortunately that's not what the autopilot did - apparently it tried to maintain the glide slope, which will have frittered the energy away.

If you fly too slowly (which is what the autopilot did) you increase the induced drag, too fast and you incrase the form drag. Both will result in the plane not flying as far as it could for the original amount of energy.

See the total drag vs. speed diagram on this URL - where the two cross is the best L/D ratio, and the best speed to fly for the situation G-YMMM was in. Unfortunatly the crew had very, very little time to think about it.

http://pilotsweb.com/principle/liftdrag.htm

[finncapt]

I am a bit pickled, but you are all fantastic, hindsight is a wonderful thing.

I think with the situation they had they did their best in the time available.

Lets praise them instead of nitpicking about what they could have done better.

I wonder whether I and you would have done as well on the day!!!

[RomeoTangoFoxtrotMike]

:- of the AAIB who cannot publish an update until they are SURE of what they have found. If they are working on something very serious, then anything less than a full explanation may cause panic in the travelling public. And the airline business is in a bad enough way as it is. If they are thinking of saying, "We have a bunch of ideas - and it could have been any one of them" then that will also cause a bad reaction. They do not 'owe' us anything, until they have something they can say with certainty.
I've just crunched the data on published formal reports by the AAIB back to 2006 (as far back as I could be bothered to go). The average length of time from incident to final report publication is 25.6 months, i.e. a little over two years. This does not and has not stopped them issuing recommendations, where appropriate, before the final report.

I'm quite happy for AAIB to resist the rabid frothing of the news media for sensational information to fill their 24-hour outpourings, and instead concentrate on trying to work out exactly what happened... :=

[CONF iture]

Green-dot, you made a good point for materializing the presence of fuel in the main tanks.

Do you think the following selected area depicts also a frozen surface, which would show in the same time where the fuel extends outboard …?
http://i45.servimg.com/u/f45/11/75/17/84/mmm_on10.gif (http://www.servimg.com/image_preview.php?i=13&u=11751784)


Now, is the total frozen area large enough to confirm an actual fuel volume of 6500 liters each side ?

[bubbers44]

The crew did a remarkable job of using optimum flaps to get close to the runway end. Leaving them in final landing position would have resulted in a bad situation. They should be commended on their skills to save those people.

[L337]

As this is the rumour network... And I do understand the inverse square law, and all that good stuff. I have to report the "dark" rumors on the street are, that it was electronic interference of some sort.

I have a real struggle to see how it could possibly be, yet the rumor is persistent, and independent.

[Super VC-10]

RTFM - Spot on! :ok:

[PBL]

Biz47,

the difference between our views comes very likely because you are, obviously, amateur and I am expert in the field of embedded safety-critical digital systems.

PBL

[snowfalcon2]

I think you are sidetracked by this proposed obstacle-avoidance software. Let's look at the feasibility of the original proposal, which was an autopilot sub-mode that would fly the plane to its best glide speed in case of a serious engine failure.

From a flight safety perspective, it can IMO be treated much like the stick pusher stall prevention device. The functions are in fact quite similar.
The basic difference is that in this case the primary trigger signal comes not from the AoA sensor but from a thrust deficit sensor. The autothrottle already incorporates the function of such a sensor, presumably engineered to flight-critical reliability standards, thus it already exists.

This document from Transport Canada (http://www.tc.gc.ca/CivilAviation/certification/guidance/525/525-020.htm)specifies the reliability requirements for stick pushers in transport category aircraft. It says the probability for failure of the device to operate shall not exceed 10^-4. Furthermore, regarding the probability of unwanted operation, a very pertinent issue in this kind of device, no single failure shall cause this and the probability of unwanted operation from all sources shall not exceed 10^-5. IMHO that is easily achievable with well known methods.

In fact, the only additional component not part of standard airplane hardware would be the lift/drag polar curve data. I presume any transport category aircraft-certified computer would have error-correcting memory as a standard feature, so no additional reliability issue there. And the software algorithm is very simple.

Furthermore, this suggested device would only need to be armed at final approach and when at or below the glideslope, greatly reducing its potential to do any harm. Last but not least, it would only ever be triggered when a flight critical "10^-9" event has already happened.

To summarize, if industry can build safe stick pushers, building this device is not a problem.

So please tell me if I've forgotten something crucial. And to repeat: this is about the technical feasibility. The actual usefulness of this suggested device is another discussion. :)

[Green-dot]

Posted by CONF iture:
Do you think the following selected area depicts also a frozen surface, which would show in the same time where the fuel extends outboard …?

Yes, i believe that is also fuel induced frost.

Now, is the total frozen area large enough to confirm an actual fuel volume of 6500 liters each side ?

I will do some calculations using scale drawings of the fuel tanks.

At first glance, using a B777 front plan view and drawing a horizontal line between the bottom of the wing tanks just outside the left and right pylons with the line disecting the wing tank/CWT rib, the fuel column measured from the bottom skin up reaches two-thirds to the top of the wing tank/CWT rib (i.e. top wing skin). Only one-third of total wing thickness at the rib (from fuel level to upper wing surface) is filled with air.

I could place the drawing here, if I only knew how. Apparently i do not have an attachment option, which according to PPRuNe FAQ i should have.

Should not be too difficult to calculate fuel contents (taking the dihedral of the wing bottom surface into account) if exact dimensions of wing thickness at the ref. rib, chord of the rib between front and rear spar, and distance from rib to just outboard of the pylons are known.

Green-dot

[Chris Scott]

Interesting line of enquiry, you two, but I would suggest that the resulting figure might be slightly less than the amount on board. It seems likely that the wing bottom surface would have been experiencing positive TATs for at least 10 minutes before the photo was taken.

The aircraft left the Lambourne Hold, according to the AAIB, at FL90. The surface/sea-level temperature was +10C, suggesting something of the order of –8C at FL90. At a typical hold-exit speed of 210kts IAS, I think (in the absence of a Dalton computer) that this would produce a positive TAT. Perhaps someone can tell us? [It is conceivable that the TAT may have become positive prior to the speed-limit point, which would have been 8 minutes earlier. This may have been short-lived, however, and reversed as they slowed down into the hold.]

Time from LAM to touchdown is likely to have been 10 – 15 minutes, during which the skin below the dry part of the fuel tank would have been conducting relative heat towards the ‘wet’ part. This would be eroding the margin of the frost below the ‘wet’ part of the tank, I think. How much would depend partly on the fuel temperature; partly on the TAT.

You should treat whatever figure you arrive with as a MINIMUM estimate. Presence of frost means presence of fuel; conversely, absence of frost does not prove absence of fuel.

[PBL]

snowfalcon2,

pertinent comments, with which I do not entirely agree, but for reasons unrelated to this thread I probably won't be contributing further.

PBL

[snowfalcon2]

I agree fully with Chris. This is based on a survey of about 50 photographs of landing B777s showing the wing undersides. I can find only one picture possibly suggesting wing surface frost in the tank area (ANA-Star Alliance plane).
I would imagine the airflow in most cases prevents frost (or ice) to form until the plane has landed and slows down. The wing underside is also a high pressure area, meaning that moisture is less likely to condensate on that part of the wing than on a low-pressure area.

That said, most of the pictures are taken in conditions of less humidity than the weather BA38 had.

But if it's frost that is visible on the BA38 picture, it might be a fairly unique occurrence?

Here are some pictures:
http://www.airliners.net/photo/Korean-Air/Boeing-777-2B5-ER/1341078/L/
http://www.airliners.net/photo/British-Airways/Boeing-777-236-ER/1339905/L/
http://www.airliners.net/photo/Air-Canada/Boeing-777-233-LR/1342818/L/
http://www.airliners.net/photo/Continental-Airlines/Boeing-777-224-ER/1339702/L/
http://www.airliners.net/photo/Jet-Airways/Boeing-777-35R-ER/1338232/L/
http://www.airliners.net/photo/Star-Alliance-(All/Boeing-777-381-ER/1336739/L/
http://www.airliners.net/photo/British-Airways/Boeing-777-236-ER/1333455/L/
http://www.airliners.net/photo/British-Airways/Boeing-777-236-ER/1326063/L/ G-YMMM
http://www.airliners.net/photo/Korean-Air/Boeing-777-2B5-ER/1325100/L/
http://www.airliners.net/photo/Korean-Air/Boeing-777-2B5-ER/1321846/L/

EDIT: Based on some more searching, frost on the underside is not totally uncommon but still fairly rare. A search turned up a total of 3 pictures with visible frost under the tanks.

http://www.airliners.net/photo/Finnair/McDonnell-Douglas-MD-11/0991454/L/
http://www.airliners.net/photo/Virgin-Atlantic-Airways/Boeing-747-443/0764483/L/
http://www.airliners.net/photo/Gulf-Aviation/Vickers-VC-10-1101/0701662/L/

[Green-dot]

To Chris Scott and snowfalcon2:

I've checked A.net pictures as well today and came across almost the same pictures snowfalcon2 posted. I agree with you that actual fuel quantity is probably somewhat more than the frosted area for the reasons you explained. Making it only more likely there was sufficient fuel in the tanks before the moment of impact.

For it being a rare condition to observe frost on the lower wing surface, that may be when searching photographs for examples. But since I've performed countless post flight inspections and also participate in de-icing activities as a supervisor in winter seasons I come across frosted lower wing surfaces quite regularly if the weather conditions are right for it. Indeed the fuel level in the tanks usually exceeds the frosted area somewhat which can be observed by a wet thawed area bordering the frost at + deg. C conditions. This condition cannot be determined from the G-YMMM subject photograph but was probably present.

Therefore, I do believe the fuel quantity in G-YMMM's case was slightly more than the frosted area in the picture and will take it into account when making any calculations. However, I'm still convinced that the center wing tank was empty when observing the clear outline of frost starting at the rib, separating wing tank(s) from center wing tank. Also because of the absence of frost near the wing root.

Green-dot

[DozyWannabe]

Please note that airbus flybywire software does not require supercomputers : an Intel 8186 does the job pretty well. 20 years old technology ... much cheaper and much more powerfull hardware is available today - about 1000 times more powerfull ....
Yes, more powerful hardware is available now - but do you know *why* the decision was made to use older hardware, and indeed why previous generation hardware continues to be used in safety-critical situations?

In the home and business sectors of computing, the onus is on bang for buck - no-one particularly cares if a fast processor crashes once in a while if the worst that can happen is losing a report that's due in tomorrow. In safety-critical systems the focus is on predictability - knowing that barring a catastrophic hardware failure a certain input will *always* produce a given output.

In the case of the Airbus system, the 80186 was a 1982 update of technology that was released in 1977, and had been developed for several years prior to that. Which means that it was in effect 6 years old by the time the A320 was delivered (26 years old now), based on technology that was 5 years older than that (31 years old now). Then as now, 6 years is a long time in microprocessor development.

Processors of that vintage were very simple technology by today's standards - but that's what makes them useful in real-time, safety-critical systems. As a Software Engineering graduate you'll remember your first Hardware Architecture lesson and the diagram you were shown, illustrating the program counter, registers and arithmetic/logic unit. '70s and early '80s era processors were slightly more complex but still hewed to that basic design. Today's modern processors are nothing like that - they are essentially a CISC translation unit which takes the x86 instructions and translates them into a format that a highly-complex RISC back-end then uses to do the grunt work. Such a design is wonderful for today's desktops, workstations and servers but their complexity makes it a challenge to answer the safety-critical question of predictability with any conviction.

Hence the A330/340 was introduced 4 years after the A320, but still used the 80186 in the SEC. The Space Shuttle continues to use heat and radiation-hardened variants of the Motorola 68000, a design of similar vintage to the original 8086 (but logically structured more in line with the PDP-11 computers used in development in the 1970s). The latest Honeywell FMS uses an AMD design that's 20 years old now.

Anyways, that's how I understand it - I'm sure the Prof will correct me if I've made any glaring mistakes, seeing as my knowledge was only current when I graduated Uni 7 years ago.

Apologies to the mods if this post is too far off-topic.

[PAXboy]

Dozy, not off thread, wonderfully ON thread to straighten some of the wonky lines that people have been shooting in here.

I am sorry that PBL cannot continue (for whatever reason) and, whilst remaining fascinated by this mystery, I think that I shall wait the results that RTFM has kindly calculate at 25.6 months. My thanks to RTFM for that, it confirms what I have said so often.

Many can now draw a sigh of relief as I leave this thread to it's natural end ...

[LurkerBelow]

Dozy: slight correction - Shuttle flight computers are actually made by IBM and the architecture is an update of the S/360 developed in 1960s (and is still in use today albeit much enhanced on the IBM mainframes).
The actual processors had a well tried pedigree as an avionics processor in such aircraft as the B-52 (obviously the later variants).

[CONF iture]

I believe the main idea is:
there was actually cold stuff (hopefully fuel) in these main tanks, and obviously enough for another 2 miles.

But still, I’ll be interested if Green-dot is able to produce an estimate.

I would not bet on any figure, but to me, the area looks pretty tiny, at least more tiny that the one from ANA when blue sky (and Metar confirms temp / dew point) was not optimal for that frost formation. To have access to that ANA remaining fuel could provide some direction of thinking …

http://i45.servimg.com/u/f45/11/75/17/84/ana_0210.gif (http://www.servimg.com/image_preview.php?i=14&u=11751784)


Regarding BA038, humidity was effectively more important with that broken at 1000 ft confirmed by a 2 deg dew point spread on surface.
I do agree that TAT has an influence, but not that much in this case when you consider this remark:
“clear outline of frost starting at the rib, separating wing tank(s) from center wing tank”
Nevertheless I do consider the limited cold fuel influence on the outboard as volume is lesser, but I still find the outline quite marked.

Just one more thought:
Considering the huge size of the CWT close to 100000 liters (and also the architecture, as I understand it the tank is not in direct contact with the fairing skin ?) I’m not sure even 10000 liters would be made visible by any frost on the left or right side of that tank, as that volume would be exclusively in the lower part, just in the center of that center tank … ?

[Swedish Steve]

CONFiture

The wing fuel tank on the B777 goes a long way inboard of the strut.

I can't comment on how often frost is seem while aircraft are landing, but I do meet B777 every day on the ramp at ARN. On arrival on the ramp there is nearly always frost under the wings like on the picture above. The area under the wings inboard of the engine normally shows a clear line where the wing tank ends. Frost under the wings on long haul aircraft is there on arrival at the ramp about 90pc of the year. The air must be very dry for it not to form.

[DozyWannabe]

LurkerBelow:
Aye, you're probably right. I think the 68k was used in the engine control system rather than avionics.

[kick the tires]

A-320 is a totally different kettle of fish but, if I DONT see under surface wing frost on a turnround, I think something is wrong!!

Its a pure case of cold metal descending into warm air and picking up the moisture, its not an issue at all.

[Chris Scott]

Just how closely the under-wing frost defines the ‘wet’ part of a wing tank is a matter for some conjecture. I don’t disagree with anything that’s been said; nor amend my suggestions above.

Swedish Steve is our most experienced B777 witness on the ramp, and has stated that the inboard edge of the tank is clearly discernable. This is, of course, where the fuel is deepest; and the massive rib that forms the end of the tank has itself been refrigerated by the fuel for many hours. At the opposite end of the tank (the end we are considering), the shallow fuel is lapping the shore, so to speak; and not contained by a rib in the presumed fuel state.

CONF iture has reminded us of the centre-section (belly) fairing, which on most types is deepest at the forward end. My dial-up connection discourages me from downloading snowfalcon2’s excellent photos of the B777. But other types have enough depth to accommodate the air-con packs at the front. Further aft, the forward base of the centre tank is also well within the fairing. For example, tank-pump canister-plates are only accessible via an access panel on the fairing; likewise the magnetic fuel-level-indicator sticks for the centre tank.

snowfalcon2, could you tell us the extent of the B777’s belly fairing?

Subject to the above, I don’t think absence of frost on the belly fairing would tell us anything about centre-tank contents. As for the wing tank, see my post above. To paraphrase what I said: presence of frost indicates presence of fuel; conversely, absence of frost does not necessarily define precisely where the fuel is absent.

[Swedish Steve]

http://www.airliners.net/uf/536904336/phpElb51J.jpg

Took this this afternoon. A B777-200ER of Malaysia airlines just parked on the ramp at ARN after an 11 hour flight. This frost cover on the lower wing is typical. Total fuel on board 9700kgs.

Oh dear picture did't work. Will try again later.
Thats better.

[snowfalcon2]

snowfalcon2, could you tell us the extent of the B777’s belly fairing?
I'm not sure if I understand the question 100%. I believe there are several posters here with a much better knowledge of the B777 tank construction than myself.

What I can add is the regulatory requirement. FAR 25.963 "Fuel tanks: General" states that:
Fuel tanks within the fuselage contour must be able to resist rupture and to retain fuel, under the inertia forces prescribed for the emergency landing conditions in §25.561. In addition, these tanks must be in a protected position so that exposure of the tanks to scraping action with the ground is unlikely.

To me, this means that the centre tank is inside the structure and thus frost formation on the outside of the belly due to cold fuel is highly unlikely. Maybe our Swedish contributor can confirm that.


Subject to the above, I don’t think absence of frost on the belly fairing would tell us anything about centre-tank contents.

Agree, see above. However, the centre tank does reach into the wings, as this incident report (http://www.aaib.gov.uk/cms_resources/dft_avsafety_pdf_030051.pdf) shows, so in theory frost could form on the wing underside portion of the centre tank if it contains fuel.

It is interesting that while underwing frost is commonplace on parked airplanes, judging from the photos it's quite rare to form while the plane is airborne, as it did on BA38. How rare, is an interesting question.

[CONF iture]

Very good Steve !
As frost shape is quite similar to what we saw under MMM, I think it does validate the "indicated" 10500 kg from AAIB report.
Thanks

[GemDeveloper]

A bit off thread, but...

... jolly nice to see an MH arrival with a bit of spare fuel on board... :)

[Swedish Steve]

To me, this means that the centre tank is inside the structure and thus frost formation on the outside of the belly due to cold fuel is highly unlikely. Maybe our Swedish contributor can confirm that.

If you look at my picture, the centre tank starts where the frost ends, and goes right across the aircraft. The wing/body fairing covers the centre part of it completely.

[Machaca]

http://img61.imageshack.us/img61/6208/777200erfueltanksvh4.th.jpg (http://img61.imageshack.us/my.php?image=777200erfueltanksvh4.jpg)
http://img57.imageshack.us/img57/5748/777200erfueltanks2dm2.th.jpg (http://img57.imageshack.us/my.php?image=777200erfueltanks2dm2.jpg)

[Green-dot]

Quoting CONF iture:


Very good Steve !
As frost shape is quite similar to what we saw under MMM, I think it does validate the "indicated" 10500 kg from AAIB report.
Thanks


And Swedish Steve:


Took this this afternoon. A B777-200ER of Malaysia airlines just parked on the ramp at ARN after an 11 hour flight. This frost cover on the lower wing is typical. Total fuel on board 9700kgs.


Very similar to the frost area under MMM, see picture and drawings below:
http://pickyperkins.home.infionline.net/photo.jpg



http://pickyperkins.home.infionline.net/front.jpg


http://pickyperkins.home.infionline.net/tanks.jpg


With special thanks to a fellow PPRuNe-er for helping me get wiser on how to post images here.

Checked another B777-200ER here at AMS during a post flight insp. which had approx. the same area of frost under the wing.

Regards,
Green-dot

[DozyWannabe]

So in layman's terms we can put the "poor fuel management" theory to bed with the EMI nonsense - is that the case?

Oh, and another thing I forgot from my earlier post - older processor designs didn't use as much power and didn't require large fans to keep the temperature under control - in engineering terms fewer parts = fewer things to go wrong.

[autoflight]

I am very interested to know all the differences between advancing the B777 thrust levers vs advancing them to maximum.

[cats_five]

So in layman's terms we can put the "poor fuel management" theory to bed with the EMI nonsense - is that the case?

I suspect we could ever since one of the two AAIB reports commented that indicated fuel was more than adequate.

[Chris Scott]

Congratulations to our two ramp men - Green-dot and Swedish Steve - and CONF iture, for finally kicking the OUTAGAS theory into touch. [If the crew or anyone connected with the AAIB ever reads this stuff, no doubt they will have cracked a wry smile.]

Dozy Wanabee,

It's easy to dismiss other people's areas of enquiry as "nonsense". But in the absence of alternatives, was this accident purely an act of God? How about reminding us all of your more sensible theory?

[Green-dot]

This is an accident with very rare circumstances, likely never encountered before. In the order of 10 to minus 9 power: extremely improbable but with catastrophic consequences.

Chances are that the combinations making this happen are conditions that have never made it to the list of required DFDR parameters.

My guess is that if they find the cause to be a unique set of circumstances, the list of recorded parameters will possibly once again be extended.

As an example, such was the case when the B737 rudder PCU servo-valve had been identified as the cause of several accidents.

In case of the 737 rudder problem the conception was also that all swans were white, until presence of a black swan (the PCU) was identified.


Green-dot

[FullWings]

Do you mean something like this?

To focus on these _spontaneous words_ (http://news.bbc.co.uk/player/nol/newsid_7190000/newsid_7194700/7194714.stm?bw=bb&mp=wm&asb=1&news=1&bbcws=1) would have been very premature before the 30 days period.

But 2 months after the “*no anomalies in the major aircraft systems + fuel conforms to Jet A-1 specifications with no signs of contamination or unusual levels of water content*” Bulletin release and ensuing constructive but unofficial talking, with occasional but repetitive (?) deletion … these *spontaneous* words may fully sound …

I wouldn't take silence on the part of the AAIB/Boeing/RR/BA to mean that investigative work has slowed down; just that there is nothing more to report that impacts current operations. I'm sure if they turn up something safety critical, there will be an AD or special bulletin out PDQ.

Maybe the mods are just tired of the same old speculation? This second hand account actually ties in very well with the current AAIB report, given that the person relating it wasn't a pilot so could easily confuse loss of all engine power with loss of control of the engines... Who knows?

[infrequentflyer789]

As an example, such was the case when the B737 rudder PCU servo-valve had been identified as the cause of several accidents.

In case of the 737 rudder problem the conception was also that all swans were white, until presence of a black swan (the PCU) was identified.


That would be the one where the retrofitted rudder controls (to fix the problem) included duff control rods that could also lead to uncommanded hardover ? :uhoh:

I think the investigation in that case included a lot of serious test rig kit with rudder assemblies in cold boxes etc. - for G-YMMM you might want most of a wing (+engine) in a cold box. Yikes.

Comet hulls in swimming pools also come to mind.

I wonder if there will be that level of resource put into this incident at this point - at some point, I guess, they decide not to spend any more on investigating what is clearly a very very rare event.

That B737 rudder issue, and the Comet, weren't investigated in the level of detail required until after multiple hull losses - I wonder if we'll ever get the answer on G-YMMM... :hmm:

[safetypee]

The presence of fuel in the tank does not completely rule out mismanaged fuel. What happens between the tank and the engine, and if selectable might not be apparent to the investigators, i.e. not on the FDR.
There is evidence (or inference) that with the fuel metering valve open there was insufficient fuel flow to accelerate the engines; thus whatever is in between, could be mismanaged either manually, automatically, by design, or unintentionally.

[Flintstone]

This is what ruins PPRuNe and threads such as this. Anoraks and their ridiculous theories on subjects they know nothing about. I'm glad those posts were removed. 'Sunspots caused the crash' my @rse.

If you want to spout rubbish do it in Jet Blast or, better still, on another site and leave the grown up subjects to those of us with a vested interest.

Back on topic now please.

[FullWings]

There is evidence (or inference) that with the fuel metering valve open there was insufficient fuel flow to accelerate the engines
The reports seem to indicate that this was the case.

...thus whatever is in between, could be mismanaged either manually, automatically, by design, or unintentionally.
I've had a ponder about how you could deliberately restrict (not shut off) the fuel flow to the engines and I can't really think of a reliable way of accomplishing this without some rather noticeable secondary effects (!). The main tank pumps can be turned off without affecting the fuel flow (at sea level, that is) so you'd have to put both fuel control switches to cutoff or pull the fire handles, then pull the CBs for the spar valves to try and freeze them in a slightly open position, then quickly re-instate the FC/FH... I don't know what the transit times are, so it might not even be possible. At that point the FBW, autopilot, electrics, etc. would have all gone for a ball of chalk, not to mention the RAT deploying. I think we can put this scenario down as "unlikely", at the very least!

[Skydrol Leak]

Can someone please close this thread?
It looks like it is going into a dead end street (if it ever ends?) where they will give you a free pukie bag to get rid off all the theories and thesis gathered by mostly "experienced" and qualified researchers along the way.I think CAA will have the final word on it after all,right?
The only other thing I can match with this is a thread started by a BA union rep when it comes down to strike or something similar...100.000 views and all that, almost like a soap opera or a set at Beverly hills 90210 on the end.
Boring and mundane or just superficial who knows.

[Green-dot]

Quoting safetypee:

There is evidence (or inference) that with the fuel metering valve open there was insufficient fuel flow to accelerate the engines; thus whatever is in between, could be mismanaged either manually, automatically, by design, or unintentionally.


Or uncommanded . . . .


Quoting Flintstone:

This is what ruins PPRuNe and threads such as this. Anoraks and their ridiculous theories on subjects they know nothing about.


Sadly, what is worse, are biassed threads . . .

Biassed:

1. A statistical sampling or testing error caused by systematically favoring some outcomes over others.

2. To influence in a particular, typically unfair direction; prejudice.

Time for me to study a bit more about some specific issues in the T7 AMM, WDM, SSM and FIM regarding ATA 28-22-15. Understanding you are a professional, i will skip the details . . . . .


Just for the record, this post originally contained 3 replies to 3 different posters. Somehow 1 reply was removed/deleted minutes after i submitted this post on the 17th of april.


Regards,
Green-dot

[Sunfish]

I wish this thread could be locked and removed. In an effort to correct the record regarding the Comet, if my memory serves me correctly, it was a design defect, not a construction defect, that caused the in flight break-up. To whit, "square" windows in a pressurised hull, leading to stress concentrations at the window corners. It had nothing to do with the redox bonding process.

[Chris Scott]

Quote from Flintstone:
If you want to spout rubbish do it in Jet Blast or, better still, on another site and leave the grown up subjects to those of us with a vested interest.
Back on topic now please
[Unquote]

Quite. And your theory is...?


Skydrol Leak,

The most ludicrous posts on this thread are those asserting which topics are kosher; and which taboo. Thank you for your gratuitous rant. The solution is simply this: if you object to informed professionals (and even me, given the chance ;)) discussing the merits and weaknesses of hypotheses that you find counterintuitive, or unpalatable; don’t click on this thread, it’s not obligatory. And beware of Skydrol; it can blur vision.

Chris

[Flintstone]

Better still, have the mods delete the rubbish and have something approximating a reasoned and informed thread.

Oh look, that's already happening.

[safetypee]

“… also for turning up something that hasn’t been properly understood before …”

Like the Lightning gear collapse during landing attributed to a structural vibration (resonant frequency) which momentarily moved the mechanical portion of the electo-mechanical valve? This, IIRC, involved a very few events on one variant type, but was a rarity without logical explanation.

Are there any interesting vibrations, system selections, or characteristics which could contribute to such a theory during an approach in a 777?
If so then presumably with structural considerations, there might be an enormous combination of weight, cg, configuration, speed, turbulence, control input, intersystem activity, etc that might account for the apparent rarity of this event.

[airfoilmod]

I don't spend much time worrying about it, but note how baffling it is that there are those who post how irritated they are by other's posts. If one is annoyed, and insists on shutting a thread or that other's simply shut up, wouldn't it be easier for them to move along? Why would one wish to be irritated by other people? Seems somewhat immature.

Which brings me to this moment. I do not believe Websites have ANY obligation to allow full and free access to those they wish to exclude. This isn't a "Free Speech" environment; that would be the Public Square, which this venue is NOT. This is Private Property, and as such requires deference to, and respect for, the OWNER(S). I have believed that since the beginning of the Ether.

BA38. In 1988 (roughly) our company, with its progenitor, Boeing Aircraft, was looking at ETOPS. There were essentially two camps, as there are re: mandatory 60 year old retirement. Two crew Twin engine A/C were an assault on all who held dear the concept of "Four's enough, but not too many". Loss of the Flight Engineer was difficult as well. Many were very sceptical of two engines more than gliding distance from an Aerodrome.
As the BA38 issue unfolds, I continue to fully believe in ETOPS, though initially my faith was shaken. I continue to be confident because I think the incident was unrelated to ETOPS. A combination of reading, sifting through admittedly sparse Data, and Intuition tells me that. I am fascinated by the posts here, by those far smarter and more experienced than I, and I think to squelch vigorous and respectful debate would be a great Loss. As I have said, this post is now the property of the owner of this website, to do with or not whatever he/they may wish.

AFM

[Sunfish]

Against stupidity, even the Gods struggle in vain.

[chase888]

This is only my second post, so forgive me if I appear to go over old ground.

I fly for fun, but my real business is power.

I know elsewhere, that wax has been targeted as the culprit, and in my view, this is partially correct.

When wax forms in fuel, it is also the fraction of the fuel with the most heat, that has been effectively removed, and the heat left in the remaining liquid will be at a lower heat value, and this could be further aggravated if the remaining fuel has in fact separated into its other fractions, with the worst fuel( lowest amaount of energy) being at the bottom of the tank.

The result, in my viiew is that the FADEC worked fine, but the amount of fuel pumped into the engine lacked the required energy (kJ) to meet the power demanded.

Just my 2 cents worth from a dreaded PP.
Cheers

[broadreach]

Firelight and Infrequentflyer789

Yes, yikes! I don't think a "cold box" of the necessary size would be too difficult to find. Expensive yes; there are some pretty large reefer warehouses around but renting one out for a few months could put a considerable dent in the distribution of beef or frozen veg. Getting the temperature down to minus 50C would be a challenge (normal working temperature being minus 18-20C) and intake/exhaust another, assuming the engine would have to be working.

If the AAIB were to go the route of testing temperatures it would probably be more expedient to rig up a prefab coldstore around an existing airframe.

oh, and Brian... JS Mills?

[Brian Abraham]

...... and Von Hayek.

[soem dood]

AFM:

The reason the EMC zealots get short shrift is that as a rule, they violate the precept that 99% of other proposers of pet-theories manage to follow:

No matter how improbable your theory, please walk through it from cause to effect. Those opining on fuel waxing, FADEC failure, inner engine icing, pilot error, partial spar valve failure, etc all at least provide the courtesy of a scenario and some sort of support chain.

IMHO, only the proponents of mystical EMC feel that a bit of arm waving, appeal to the vagaries of past unexplained accidents, and exasperated frustration that their explanation is not just ratified as-is, can somehow suffice as support for their otherwise unsupportable stance.

This comes to mind:

http://i29.tinypic.com/k33z9c.jpg

That is simply not how scientific inquiry and accident investigation in particular is to be approached. Einstein said: "Keep an open mind. But not so open your brain falls out."

:E

[Green-dot]

Quoting safetypee:


Are there any interesting vibrations, system selections, or characteristics which could contribute to such a theory during an approach in a 777?


Taking the AAIB reports published into consideration, (perhaps the investigation has progressed beyond what has been published sofar) if no evidence of an anomaly has been found:

Approaching the issue from another angle, i would go for a combination of system selections (including interaction or possible interference with other systems) and characteristics.

A question that comes to mind would then be: which (sub) system components could (due to a yet undefined fault) migrate from commanded position without generating a fault indication to the flight crew but result in reduced fuel flow and cause engine rollback?

A question related to the one above is, why was the APU door open during approach if there was no mention in the AAIB reports of an electrical power failure? Isn't the APU normally an after landing checklist item?


Green-dot

[lomapaseo]

Approaching the issue from another angle, i would go for a combination of system selections (including interaction or possible interference with other systems) and characteristics.

A question that comes to mind would then be: which (sub) system components could (due to a yet undefined fault) migrate from commanded position without generating a fault indication to the flight crew but result in reduced fuel flow and cause engine rollback?



Just to be a tad more inclusive in this postulation, also add crew actions as a contributing/causal factor in some way that is not annunciated (time captured) on the recorders. Remember that some recorders don't latch everytime a bit or byte changes.

[slip and turn]

This thread needs some guidance and direction again. Come back PBL and bsieker - what should be discussed?

[mark exclamation]

Well here's a potentially new twist, as I haven't read all the posts maybe it's been mentioned B4.

Mate of mine who flys the 777 made the same arrival into Heathrow from LAM a couple of hours B4 the stricken A/C. Now, aficionados correct me if I'm wrong, the 777 has both auto engine A/ice and auto Airframe A/ice systems. My mate had obviously seen operation of the engine a\ice system many times but never the auto airframe a\ice. That morning, apparently, the icing in the LAM hold was so severe that for the first time that he had seen the auto airframe a\ice system operated. He was so surprised that the fact was mentioned to ATC and later the ATIS was updated to mention severe icing.
He also mentioned that those huge Trents can have a problem with ''core icing'' but didn't elucidate on the matter!
Could severe icing during potentially a textbook CDA approach from the hold with the throttles almost closed right down to 600' produce a lack of response from the Trents as seen on the BA A/C?

I commend this question to the house!

Take care

[bsieker]

mark exclamation,

I don't know if there's really a new twist.

Core icing in high-bypass turbofan engines has lately been the subject of some discussion, such as this PPRuNe thread (http://www.pprune.org/forums/showthread.php?t=321622), discussing a Wall Street Journal article, as well as a small article in the German Spiegel weekly magazine. But what I've read, the symptoms are usually flameouts, almost always with a quick relight.

From what I gather one of the problems is that core-icing at cruise altitudes has for a long time been thought impossible, but has recently been established as a likely cause for flameouts at altitude. The icing is apparently caused by very small ice particles and/or supercooled droplets.

This video (http://www.youtube.com/watch?v=6CTy_xVplUY) mentions that drops above a certain size are normally deflected outward by the fan blades and do not enter the engine core. The very small size of the droplets/crystals in question may facilitate their entering the compressor and accumulating there under certain conditions, which, as of now, seem to be unknown. When larger chunks of that ice break free they may melt in the combustion chamber (or the later, hotter compressor stages) and cause a flameout, or even damage turbine blades.

Another thing that's peculiar about this form of ice accretion in the engine is that it may happen in conditions, under which the airframe does not suffer from ice accumulation.

To get back to the topic, in the case of BA038 there was no flameout, but "only" a failure to sustain a higher than "somewhat above flight idle" thrust. I'm still uncertain if there may be a connection. It seems unlikely, but then again, so does every possible explanation so far.


Bernd

[M.Mouse]

....a textbook CDA approach from the hold with the throttles almost closed right down to 600'....

In BA that would not have been a 'textbook CDA'. It has been covered numerous times before ad nauseum.

Core icing is a known issue in certain defined conditions on the ground with Trent B777 engines and we have procedures to follow when it is a possibility.

In the air we have no knowledge or requirement for any procedures relating to core icing. As has been mentioned, it has now been discovered that it is possible in certain conditions but we have received no official communication on the subject.

[wilyflier]

ExclamationMark,
.......As a scenario it fits.
As I read it , core Icing when taxying in foggy conditions did not always lead to flameouts while running at low speed.If the engines had got any sort of ice in them you can bet your boots they would not be able to run properly or efficiently.Acceleration from idle revs would be bound to be somewhat reduced
.......This is the first I heard about severe icing in the LAM holding pattern before Final approach. Who says this? Whats his qualification to make the observation? Is it confirmable.Surely the AAIB interrogation would would have shown this up early on.

[F111D]

Looks like the investigation is focusing on slush or ice in the fuel system as the root cause of the problem. Is there a redesign of the oil coolers in the near future?

The article follows....

Re-Evaluating Long, Cold Flights

Crash May Prompt
New Safety Rules
For Boeing's 777
By ANDY PASZTOR
April 25, 2008; Page B2

Prompted by the crash landing of a British Airways (http://online.wsj.com/quotes/main.html?type=djn&symbol=bay) jetliner near London earlier this year, air-safety investigators are moving to recommend heightened cold-temperature safeguards for Boeing 777 aircraft flying long polar routes, according to people familiar with the details.
In a few weeks, these people said, investigators are likely to warn operators and pilots of Boeing (http://online.wsj.com/quotes/main.html?type=djn&symbol=ba) Co.'s long-haul 777 models to take extra precautions when monitoring fuel temperatures on a growing number of extended flights over the North Pole. If temperatures creep too low, pilots can descend to warmer air or speed up to increase the heat generated by air friction against the plane's skin.
• The News: Investigators are moving to recommend cold-temperature safeguards for Boeing 777s on polar routes.
• Background: Slush or ice build-up in a British Airways 777 fuel system may have led to a Jan. 17 accident at London's Heathrow Airport.
• Outlook: Operational warnings or recommendations could come in the weeks ahead.


"It's a possibility" that such operational warnings or recommendations will be issued within the next few weeks, said David King, chief investigator for Britain's Air Accidents Investigation Branch, or AAIB, which is in charge of the probe. In an interview Thursday, Mr. King said safety experts are "looking at a wide range of things" affecting fuel management and environmental conditions on polar flights.
Though some are arguing for design changes, a team of U.S. and British investigators hasn't reached a consensus on whether to recommend modifications to make the fuel systems on some Boeing 777s more resistant to unusually frigid conditions. Such a move could be seen as a public relations blow to Boeing and Rolls-Royce (http://online.wsj.com/quotes/main.html?type=djn&symbol=RR.LN) PLC, which supplied the engines and related hardware on the accident aircraft.
A spokesman for Rolls-Royce, which has declined to comment on the accident or the status of the investigation, said Thursday that the company "remains committed to working with the AAIB and Boeing to establish the root cause of the event" and is cooperating with other participants. British Airways, Boeing, the Federal Aviation Administration and the U.S. National Transportation Safety Board all declined to comment.
Investigators suspect that the Jan. 17 accident at London's Heathrow Airport occurred because slush or ice built up in part of the fuel system of the British Airways 777 during a long polar flight from China in unusually cold outside temperatures. As the widebody jet descended toward London, remnants of icy particles likely clogged a portion of its fuel system, starving the Rolls-Royce engines and providing minimal power despite pilot commands for more thrust.
The plane slammed down 300 yards short of the runway. All 152 aboard survived. The probe has been closely watched partly because the root cause remains elusive and complex. The accident has garnered attention because it is believed to be the first such incident involving both of a plane's engines while maneuvering at such a low altitude in more than 14 million 777 flight-hours since 1995.
Some U.S. experts have advocated changing the design of the oil cooler used on some versions of the Boeing 777 powered by Rolls-Royce engines, arguing that that is where ice or slush may have blocked proper fuel flow. Taking the unusual step of disclosing details of a pending investigations, Mr. King said: "I don't believe there is anyone on my team that would argue for a redesign" of the oil cooler.
Investigators have ruled out engine or computer malfunctions, and pilot slip-ups, as well as systemic problems with fuel-tank design, fuel pumps or temperature sensors, according to people familiar with the details.
Write to Andy Pasztor at [email protected] ([email protected])

[aviate1138]

F111D says in part.....
Quote "Looks like the investigation is focusing on slush or ice in the fuel system as the root cause of the problem. Is there a redesign of the oil coolers in the near future?

The article follows"....End Quote.

Aviate quotes from the said article....
"Taking the unusual step of disclosing details of a pending investigations, Mr. King said: "I don't believe there is anyone on my team that would argue for a redesign" of the oil cooler."

Smoke and mirrors and speculative journalism from Andy Pasztor?

[Navy_Adversary]

The aircraft "slammed down", but I just love that word "jetliner":)

[VnV2178B]

Perhaps one of the older generation:

Wiki quote "Also developed in 1949 was the Avro Jetliner, and although it never reached production, the term jetliner caught on as a generic term for all passenger jet aircraft"

The only other time I've heard it used is in an Arlo Guthrie song (?).

VnV

[HotDog]

Ater my retirement from a legacy carrier on 747 aircraft equiped with RB211-524 engines, I operated a 747-100F with JT9D engines which had engine de-icing systems providing controlled heating of the fuel when manually activated if the fuel temperature indicator showed fuel temperature at the engine to be 0c or below, or when the fuel icing light illuminated. Heating was limited to a one minute heater operation or two minutes if No.1 tank temeperature was below -40c, for every 30 minute cycle. Kept the flight engineer pretty busy. The fuel heater was situated between the boost and high pressure stages of the engine driven fuel pump and heating was provided by 15th stage bleed air. As far as I know, fuel heating is automatic on all modern high performance powerplants but the problem of fuel icing is old hat and well known.

[tanimbar]

Warning: I'm non-professional; not crew, not engineer - just scientist guest and thanks.

Hello again,
Referring to the Andy Pasztor article in WSJ, post 908, and my post 751, 'What if no proof of cause - what does the AAIB do?' when I wrote:
"I've been wondering what would the AAIB do if they could not demonstrate to a high probability the cause of the accident. If, for example, the AAIB were convinced, but could not prove, that the fuel quality had been lowered by cold-soaking for an extended period of time and that this had lead, in some unknown way, to the accident, what would they do?

Presumably the AAIB would operate on the precautionary principle.
Might they:
1) Recommend/demand/regulate the avoidance of extended flight-time in very cold air masses;
2) Demand that if aircraft had experienced extended flight-time in very cold air masses then they must
loiter for some time in warmer air before descending for landing?"

I'm going to assume David King of AAIB is fairly certain that fuel degradation due to cold-soaking for extended periods of time is the root cause of the accident and that his quotes in the above article are accurate.

Then David King would seem to be trying to gather support in advance for a set of operational recommendations that he knows, or suspects, the industry will resist. Presumably he is not acting alone and, hopefully, will already have significant political support, albeit presently hidden below
the palisade.

Fact is that demonstrating by experimentation that fuel degradation was the result of operating conditions will take years. In the meantime, the industry should soberly reflect on the continuing lives of 152 souls and consider the accident a clear warning that, as has been the case so many times previously, and in most arenas of human endevour, we don't know it all!

So I think that in a few weeks there will be recommendations from the AAIB (and others) that operations in 'polar regions' must change.

And my predictions for the years ahead:
1) Fuel stratification will, by experimentation, be shown to be the direct cause of the accident (see post 241).
2) Aircraft will have systems (probably laser based) to detect ice/sludge etc. in fuel tanks. These may be retro-fitted to aircraft that fly ETOPS.
3) New fuel heating systems will be mandated for all new aircraft designs, that is, the fuel will be heated in the tanks and will not rely any longer on the downstream heat exchange systems returning the fuel to a non-degraded state.

Regards, Tanimbar

[CONF iture]

Smoke and mirrors and speculative journalism from Andy Pasztor?I think so.
Obvious contradiction with initial report where fuel quality was just fine.
Any Fuel Pressure Pump EICAS message type ?
Any Fuel Filter Clog EICAS message type ?
What about an AAIB official update ... is it better to first advise Mr Pasztor ?

[Spaz Modic]

:D It takes a while to conjure up the necessary technical gobbledegook type spin to pad out a 1000 page report document which finally comes to the conclusion it was outa gas. :ok:

[greekdalek]

Core icing might expaln my previously expounded theory about rotating stall (which could be caused by failure of the airflow control system OR restriction into the IP compressor).

[steamchicken]

The theory on the part of the floor at Lloyd's (as of last week) is that it ain't the fuel that caused it, it's "RF interference". The theory goes that after George Brown was dropped off at Heathrow, his driver, or some other entity either deliberately or accidentally forgot to switch off their "jammer" and that the aforesaid vehicle on it's return journey may have gotten within "range" of the affected B777 and caused the mischief.

Maybe he's right; after all, whoever was dropping off the corpse of a long-dead Labour Foreign Secretary of the 1960s at LHR must have been up to something!

[Robert Campbell]

Gee! And anyone who speculates on such a conclusion gets sent to the "other" list. And we get to ride on the short bus, too...

[DozyWannabe]

I have a feeling that if there was any genuine weight to that theory then that wouldn't be the case. As it appears, the theory has only hung around as long as it has due to an uncorroborated posting on a US pilot board with absolutely nothing backing it up.

For random background, it was David King who oversaw putting the Lockerbie 747 back together and proved exactly where the bomb was placed, so it's not like he's unknown in the US, nor is he politically tied, having been the head honcho at the AAIB since long before the current administration (to use US terminology) was in charge. From what I've read the NTSB hold him in *extremely* high esteem.

[precept]

It would seem to me, in this instance, we should take action to invigorate the AIRB and their associates in America, Boeing and RR to come clean with it. This has gone on too long and our fellow airmen and the people they are charged with carrying are either too much at risk or not to be bothered. Make your thougts known to those who might encourage the blokes along.

Those of us long retired can be a nusiance among them. On the other hand, if the gentlemen on the line begin to growl, maybe ...........

I have sent a number of messages to the US FAA, the NTSB and Boeing. Amazingly, in three weeks, not one of them has responded. The world wonders.............................


Tom

[Mat Tongkang]

Didn't all the geopolitical horse trading with that rogue regime of Muamar Ghadaffi make that Lockerbie tragedy seem a bit " convenient " and ' aha-ish "?

The icing and slush scenario is indeed plausible but to have it happened to both engines ( albeit with some seconds of time lag ) was a little bit too coincidental. Was there any indication that the auto engine anti ice had been activated on that BA038 flight?

[FireLight]

The post by F111D (re: Andy Pasztor article) does support those of us who favour the slush/icing theory for the crash.

It is a theory that goes a long way to explain the following:

* near simultaneous loss of power in both engines. (a critical factor that all theories need to explain and a severe limitation on most.)
* slight differences in conditions in each fuel tank can explain the slight difference (seconds) in response.
* why the engines responded properly to start with then dropped down to a thrust level somewhat above flight idle shortly afterwards.
* fuel pump cavitation (from slush retarding fuel suction)
* minimal remaining evidence (remaining slush would melt)
* if slush was mostly sucked out of the system prior to the crash, it would also explain why the fuel would test within specifications. (Jet A-1 Chinese equivalent Jet Fuel #3) It probably wouldn't have taken much water content in the fuel to cause problems if it was in the right location. All (or most) of the water would have frozen out at the temperatures in the fuel tanks and migrated to where it could cause grief.

I don't see any notes in the last AAIB update about EICAS warnings (fuel pressure pump or fuel filter clogged) as noted by CONF iture, although it isn't clear that they would invalidate the theory if they didn't (or did) occur. The AAIB report notes that the high pressure fuel filters were clean after the incident.

It also doesn't rule out there being other factors involved, but it explains events reasonably well without them.

I think that the reason the next update is taking so long is that the incident has a number of subtle causes. AAIB needs to make very sure they have it right. If this theory is correct or mostly correct it could cause huge changes in procedures in long haul flights in extreme cold weather as well as have the potential for other operational or aircraft design changes. If the cause was simple and straightforward, the AAIB would most likely have let us know already.

[Green-dot]

"- Both engine low pressure fuel filters were clean.
- The fuel oil heat exchangers (FOHE) in both engines were free of blockage.
- The right FOHE was clear of any debris, however the left engine FOHE had some small items of debris on its fuel inlet bulkhead.
- The high pressure filters were clean.

- A sump sample taken from the left and right main fuel tanks shortly after the accident revealed no significant quantities of water."

Although the ice/slush theory is plausible, wouldn't there have been an increased level of water content found in the fuel lines near the fuel filters, fuel oil heat exchangers or servo fuel lines? And if there was a presence of ice/slush, wouldn't some of it have been collected as water in the main tank sumps? Some of the ice/slush, being a solid and heavier than fuel, would likely have found its way to the lowest points in the tanks.

The Andy Pasztor article reminds me of the "industry insider" who claimed for example on jan. 30th [quote] "the fuel crossfeed valve switches were found in the open position and only one valve was open" [unquote]. The AAIB report, published on feb. 18th, explicitly mentioned that both valves indicated that they were closed and they had not been operated during the flight.



Green-dot

[tanimbar]

Many posts comment on ice/slush as being the primary cause of fuel degradation but this need not be so. The fuel is complex and a prolonged cold-soaking, in a moderately high frequency environment, might produce physical amalgams and/or emulsion of hydrocarbon + H20 and any other material included in the fuel. Give these oddities a few more hours to develop and they may change their physical characteristics again.

Put those oddities back into a temperate climate (Heathrow ground level) and they might disappear altogether, leaving no clue to there former presence.

Difficult to investigate. And, if the AAIB cannot find another cause it might now have to show that the fuel was, or was not, the cause of the accident. I don't think it can leave doubt.


regards, Tanimbar

[wilyflier]

#923 Greendot
....There was a strange re-edit of a lot of posts which became redated 18Feb, so I cant find full versions of the original relevant posts now.
....It had been stated that a "The Captain now recollected operating X feeds" (Plural),with the implication that it was after the engines failed to respond. It was also stated that (I personally cant remember which )
....Either they were set to, open and one was not fully open-
....Or they were set to closed and one was not fully closed.
.... In any case I read that one xfeed finished in a part way position...............So???

[overthewing]

Wilyflier

That information didn't come from the AAIB; it was on another aviation site, and it was an 'informed sources have said' kind of report.

It seemed to contradict the subsequent interim report from the AAIB in several ways, so I think it just served to confuse the issue.

No reason to believe that the pilots opened x-feeds, per the AAIB.

[DozyWannabe]

Mat Tonkang:
This is off-topic, but just to confirm - the AAIB's work at Lockerbie was purely technical, determining the break-up sequence from the point the bomb went off. It had no jurisdiction on the who, only the how.

Back on topic, the AAIB *is* a branch of the Civil Service, but it is a very different beast to the NTSB. What they do have in common is that they are both politically neutral, so if EMI had not been ruled out, or at least downgraded, they would have mentioned it as a possible contributing factor in the interim reports regardless of where it came from. Their remit is to figure out what happened - nothing more - and by all accounts they are extremely good at it and are world-respected as a result.

[Green-dot]

To wilyflier #926:

The AAIB special bulletin 1/2008 does not mention any recollection of the crew switching the crossfeed valves or finding one of them open or partially open.

Quoting 1/2008, page 2:

"The total fuel on board was indicating 10,500 kg, which was distributed almost equally between the left and right main fuel tanks, with a minor imbalance of about 300 kg. The crossfeed valves indicated that they were closed and they had not been operated during the flight."

This implies the AAIB has been able to confirm this from the FDR data (and likely from crew interviews) or they would not have made the statement: ". . .and they had not been operated during the flight."

Those are the officially published facts sofar regarding the crossfeed valves. No more, no less.

In light of only 300 kg imbalance between left and right main tank, it is also ruled out that one of the crossfeed valves could have been (partially) open undetected.



Green-dot

[CONF iture]

In light of only 300 kg imbalance between left and right main tank, it is also ruled out that one of the crossfeed valves could have been (partially) open undetected.

I'm not sure about that Green-dot, if crossfeed is open but fuel pumps are not switched off on one side or the other, each wing tank will still feed its own side engine unless fuel pressure is somehow stronger on one side.

If I'm correct one crossfeed will automatically open under a certain type of electrical failure ... !?



... You mentioned that, I mentioned that, it would be nice to know exactly when the APU flap started to move ?

[Green-dot]

To CONF iture:

You are correct, with all boost pumps running in both main tanks there would be no crossfeed if a crossfeed valve were to be in the open position. However, if only the APU is running (with boost pumps off on the overhead panel) there would be a crossfeed. But that would be a temporary condition.

An undetected open position can be ruled out. First of all, an open crossfeed valve in disagree with the pushbutton would generate a valve light in the pushbutton and EICAS advisory and status messages would be displayed (after 10 seconds). No way it would go unnoticed by the crew.

I am not aware of any automatic crossfeed valve selection. To my knowledge crossfeed valve operation is a manual selection only.

With regards to when the APU flap (assuming you mean the APU door) started to move. That may be a tougher nut to crack. I can think of a few possible scenarios:

- APU started by the crew: Not mentioned by the AAIB in their report S1/2008. Could very well have been the case even if not mentioned by the AAIB.

- APU autostart: For that to happen, power would have to have been removed from left and right transfer buses in flight. That did not happen because the ram air turbine (RAT) would have automatically extended during an APU autostart. There is photographic evidence the RAT did not extend in flight. It extended after impact on the ground.

- APU door open due to other causal factors: Answer to this will have to wait for the final report. This is where several holes in the swiss cheese would have had to line up before a failure of this nature could occur.
As discussed before, i can think of a possible (but very unlikely) scenario where the APU fuel shut off valve control relay could have been affected. The same scenario would have affected the spar valve control relays. Temporary loss of power to the 28v dc hot battery bus, making both the spar valve- and APU fuel s/o vlv control relays susceptible to EMI reverse logic. If (i stress if) EMI was present it could have caused the spar valve control relays to close and the APU fuel s/o vlv relay to open, possibly latched in that position when power was restored until the EMI source disappeared, returning the system to a normal logic condition. The effects would be:

* Closing (or partially closing) of the spar valves restricts fuel flow to the engines.

* Opening the APU fuel s/o vlv control relay, instructing the APU fuel shutoff valve to open, may have triggered a signal to ELMS to open the APU door?

Ready to be flamed . . . .but . . . "when you have eliminated the impossible, whatever remains, however improbable, must be the truth", Sherlock Holmes used to say.


Green-dot

[soem dood]

greendot:

If (i stress if) EMI was present it could have caused the spar valve control relays to close and the APU fuel s/o vlv relay to open, possibly latched in that position when power was restored until the EMI source disappeared, returning the system to a normal logic condition..How?

Please provide a plausible scenario to drive the relatively high current DC components involved, via spurious radio signals, keep them engaged, and then release them when the radio signal is removed. I won't ask you to explain the presence or strength of said signal, merely to describe the electrical activity of the aircraft systems responding, esp why other systems were apparently completely unaffected, only the fuel control.

I believe it is time to take the strop to Occam's Razor.

http://img117.imageshack.us/img117/8192/listeninggz4.gif

http://img329.imageshack.us/img329/3820/clipboard01lx1.th.jpg (http://img329.imageshack.us/my.php?image=clipboard01lx1.jpg)
http://www.airmech.co.uk/forums/showthread.php?t=3650


This gives a good overview and history. CAUTION: link to PDF file:
http://www3.verticalgateway.com/portals/12/news/hirfnprm.pdf

Link to FAA Final rule on HIRF certification as of Feb 2008:
http://rgl.faa.gov/Regulatory_and_Guidance_Library%5Crgpolicy.nsf/0/28569ED29E178BE3862573FE0071E67E?OpenDocument

[Super VC-10]

OK, just assuming slushy fuel was the cause, one would think that tests would be carried out with a jet engine and fuel that had been purposely cooled to a state of slushiness. Anyone head of any such tests being performed?

[Green-dot]

soem dood:

Thanks for the links.


How?


That is what i wondered years ago as well, at a time when we were confronted with uncommanded closing of main fuel shutoff valves (MFSOV), also a 28v dc system but on a different aircraft type. At the time only the valve was affected, no other 28 v dc system components. Regarding a scenario: EMI does not necessarily imply an external source, in this case it was generated in the connector itself.

Something to add to the history:
http://www.corrdefense.org/CorrDefense%20Magazine/Spring%202007/pdf/feature.pdf

and:
http://www.dtic.mil/ndia/2001systems/horne.pdf



With regards to G-YMMM: without the data it remains a theory, nothing more. I do not intend to twist facts to suit theories, it is beter to twist theories to suit facts.

But since i have had first hand experience with a similar issue (as the above links explain), first time it happened it was also thought an impossible scenario, but happen it did.



Green-dot

[CONF iture]

I am not aware of any automatic crossfeed valve selection. To my knowledge crossfeed valve operation is a manual selection only.
So it could well be an only Airbus thing, which does not apply to 777, when all AC bus fail, one single fuel pump remains powered and therefore crossfeed valve automatically opens ?

Regarding the ice build-up fuel scenario, I’m very septic that the same conditions develop at the almost exact same time in two separate fuel tanks, especially when one of them had almost 400 more liters than the other one.

I don’t know the oil cooler / fuel warmer architecture, but I find it surprising that the restriction occurs into the very spot where fuel should get warmer ?

Also, still no official mention if thrust has been applied after Top Of Descent but before the event and how both engines did react ?

[soem dood]

Greendot, in line with your point that increased susceptibility can come from sources such as degradation of connector internal contact integrity, here is an interesting prior airworthiness directive on the 777:

RIN 2120-AA64 - prelim
http://www.thefederalregister.com/d.p/2003-12-08-03-30338

RIN 2120-AA64 - final
http://rgl.faa.gov/Regulatory_and_Guidance_Library%5CrgAD.nsf/0/9B665F89FA92B5708625697C006F4E03?OpenDocument

...applicable to all Boeing Model 777 series airplanes. This action requires repetitive corrosion/resistance inspections to measure the resistance of each wire bundle of the flight control system; and repair of the receptacle bond, repair of the bundle connector backshells, or replacement of the wire bundles with new components, if necessary. This amendment is prompted by reports of corroded connectors and numerous other discrepancies of the wire bundles, such as loose backshells and loose shield retention bands, due to the presence of moisture inside the wire bundles. The actions specified in this AD are intended to detect and correct such corrosion, which could reduce system protection against lightning strikes or high intensity radiated field (HIRF) events...Thanks for the back story and discussion. I am nothing more than an armchair observer when it comes to air accident investigation, but I have much interest, as I am in a not-too distant line. For those that want an instant perfect answer, and want it now, I will just offer that the NTSB and/or AAIB do good work, and the truth will out, but good original work takes time. One of the most difficult cases ever was on the Boeing "soda can" rudder activation valve that brought down US Air flt 427, and it took a lot of investigation, and frankly, some related incidents (Eastwind 517, United 585) before it was figured out, but they got there. For those who have their interest whetted for following air investigations, but can't wait for the AAIB report, let me suggest any of the following as good reading, from my own bookshelf:

Normal Accidents, Perrow
The Limits of Safety, Sagan
Inviting Disaster, Chiles
Inside the Sky, Langewiesche
Mystery of Flight 427, Adair
Aircraft Investigation Analysis: Final Reports, Walters
Aircraft Safety, Krause
The Logic of Failure, Dörner
Mayday!, Wadley
Human Error, Reason
Columbia Accident Investigation Board, NASA/GPO
Safety in the Skies, Sarsfield

Fiction:
Airframe, Crichton

[soem dood]

One of the links extolling the virtues of CPCs noted this about the F16 connectors:

...the main fuel shut-off valve
connectors involved a unique material combination of goldplated
pins mated to tin-plated sockets. This combination is
known to be susceptible to a form of degradation known as
fretting corrosion. While a lubricant was eventually shown to
be very effective in reducing or eliminating this degradation,
this material combination is believed to be unique to this
application, in contrast to the more usual use of gold-gold
pins and sockets in avionics...

Huge "no-no" using dissimilar metals in an electrical connection that you want to remain reliable. Galvanic action, and other bad things can happen. Pretty surprised to find that all contacts were not initially gold-gold on such a critical application.

As to CPC, the problem with many corrosion preventing compounds is degradation over time. It is hard to beat a good gold to gold mechanical contact. I remember we evaluated some stuff in the eighties call "Tweek" or "Stabilant 22" that was supposedly a 'contact enhancer but it was not proven to improve anything, and I notice in your other presentation, it is specifically not recommended... the water displacement properties of the CPC may be it's best feature, assuming a gas tight connection subsequently...

[Green-dot]

soem dood,


From one of the links you provided:

Ref. Docket No. 2002-NM-305-AD. . . . This proposal would require replacing four socket contacts on the four boost pumps of the main fuel tanks with new, highquality goldplated contacts, and sealing the backshell of the connector with potting compound. . . . . .


It seems dissimilar metals in electrical connections is/was used elsewhere as well.

Interesting book list.
Some of those are on my booklist as well. I am currently reading "Inside The Sky."

Some other noteworthy books on the subject:

Human Factors In Flight, Hawkins
Blind Trust, Nance
The Safe Airline, Ramsden
Pilot Error, Hurst

P.S. The above mentioned book titles should not be percieved as to suggest any link to the cause of this accident. They don't and are mentioned only because they explain the many factors related to accidents in general.


Green-dot

[wilyflier]

Green dot.
....Speak up in plain language then ,stop hinting.What is your inference that I and the others dont like to say?

[Mismatch]

Possible and not uncommon reason for the open APU inlet door:
I have seen it happen that the APU ilet door actuator was inoperative. In order to be able to provide APU power if neccessary, the aircraft may opearate under an MEL, with the door secured in its open state throughout the whole operation and an extra percentage of the tripfuel accounted for the aerodynamic loss. So open APU inlet does not neccessarily mean a running APU.

[Me Myself]

We were at the sim today ( 777 ), finished a bit ahead of time and figured, why not put ourselves on 27L at Heathrow, 160 kt til 4 DME, full flaps and throttle back the engines one by one at 700 ft to try to simulate what happened to BA 038.
The surprise and stress element was of course totally absent.
I disconnected the A/P, went under the G/P and ended doing a hard landing, actually more like a low level stall, just before the runway.
Very interesting experience.
I can't believe people could imagine they could / should have done something different. I was prepared and it just happened so fast I couldn't believe it.
Couple of miles back would have been very tragic.

[sky9]

Latest update pulished today 30 April
http://www.aaib.dft.gov.uk/cms_resources/S1-2008%20G-YMMM.pdf

[Me Myself]

Could flaps have been adjusted further to reach the runway? and Was the 777 gliding?

Not in a life time !! The aircraft just bled its energy so fast. Getting to the end of the runway was really fantastic.
You have to try to imagine the magnitude of such an event. As someone put very accuratly on the site earlier, no one knows what he would do in such a case. Trying to get the grip of what's happening and to think of a way to remedy ot it in a such a short period of time is hard to imagine.
I'm glad I had a chance to try it on the sim, it does give you an idea of what the unthinkable is about.

[golfyankeesierra]

Latest update pulished today 30 April
http://www.aaib.dft.gov.uk/cms_resou...8%20G-YMMM.pdf

and it ends with:

Published February 2008

..........

[PAXboy]

in a such a short period of time is hard to imagine.
I understand, Me Myself that the elapsed time from loss of power to impact was 50 seconds +/-. Could you verify that from your SIM session?

[pattern_is_full]

The "update" is a final report on a June '06 Nott. East Midlands Boeing 737-300 accident/incident - but to be fair the AAIB website has a screwy revolving link, so if you click at the wrong instant, you something different than you intended. And it does say "5/2008 Update" - just updating on the wrong subject....

Likely what happened to Sky9

[Me Myself]

I understand, Me Myself that the elapsed time from loss of power to impact was 50 seconds +/-. Could you verify that from your SIM session?

Afraid not and if it was, even in the sim, it did feel very very short. Bear in mind this is just something we did at the end of recurent training because we had 10 minutes to spare. No rocket science. We just wanted to get the feel of what these guys must have experienced. We started at 4 dme, configured full flaps and retarded the engine one after the other to end up almost precisely at the same spot. Didn't change flaps setting because we didn't think of it too busy flying, but then, no fence in the sim.
Again, caught in the real thing I really don't know what my reaction would have been and I just hope they would have been along Cpt Burkill's lines.
Any sensible pilot will think the same way.

All of us flying the 777 would really like Boeing to come up with some explanations so it can be fixed and allow us to relax a bit.

[Bearcat]

I have asked it again and again , what is the envoirnment envelope re temps for a 777? In the 320 it is SAT -70.

I note when they checked the fuel freeze temps at the accident site, they were -57c.

They had loads of juice coming into land (10 tons), fuel must have been like jelly for some reason....hmmmm.

[Xeque]

Me Myself - I can well understand how you guys feel.

A question from someone whose regular steed is a C150. The report states that the autopilot disengaged at 175 agl when the speed had decayed to 108 knots. Is it normal to keep the autopilot engaged so late in the approach? Or were they exercising a Cat 3 autoland where (I guess) the autopilot stays engaged.

The report seems to pretty well squash any thoughts of fuel contamination, freezing etc. Also, it seems to indicate that there was nothing wrong with the fuel delivery system apart from the mention of cavitation damage.

What's left? Way back in this thread I (tongue in cheek) said 'Windows fatal error'. Is this what its going to turn out to be - the old 'blue screen of death'? Dilbert on a bad day?

[wilyflier]

Bearcat
OAT +8 or +10.Like jelly or like froth? or both?

[Mariner9]

The freeze point of the fuel may well have been -57C when tested, but if a kerosene is cooled to wax precipitation, wax drops out and is not redissolved into the fuel, then the freeze point of the Kero will be lower after the event. In other words the freeze point of the fuel would have been above -57C prior to wax precipitation, if it had occured.

However, if that were the case with BA038, the question arises what happened to the wax and why were no traces found? (I presume the investigation has included GC analysis of the bulk fuel and residual fuel on the filters for comparitive paraffin component profiling)

[L337]

Xeque: BA fly a monitored approach. It was the FOs sector, so the Captain flew the approach for him. The FO took control at 750'. Now normally what would happen, the FO (in this instance) would have disconnected the autopilot, and continued to do a manual landing. What follows is conjecture, but it is a fair bet to say that, as the AT failed to respond, and the situation unfolded. Both pilots would have looked inside and gone "WTF". Best decision was, I would hazard to leave the autopilot in and try and work out what the hell was going on. During this small window a decision was made to retract the flap one stage. That saved the day. As the speed decayed the A/P disconnected, and they hit the ground.

Would hand flying a "fly by wire" aeroplane helped in these circumstances? No idea. Those who have tried it in the sim would suggest that they made the best of what they were presented with.

It is important to remember that they first had to notice the problem. Then they had to diagnose it, then they had decide what to do. All close to the ground in circumstances they would never have seen before, and never have been trained for. That they made the crucial decision to retract the flap that saved the day cannot be under emphasised.

What I do know is that, God forbid, If I am ever presented with something so unusual in such a parlous position, I can find it within me to do as well as those guys did.

/respect

[Oluf Husted]

Dear fellow PPRuNers,

Since the BA38 crew, unknowingly, flew a: "Green Approach" (idle) with the engine anti-ice system in "Auto" they probably had some core engine icing, that caused the hesitation.

That must scare, all involved in a B777 flight (app. 1000 times a day) with an average of 200 seats sold, equals 200.000 human beings and some in the leadership of BA, RR, AAIB and Boeing, sh..less.

That why, the AAIB did't find any signs of ice damage, it just melted away.


Oluf Husted

[davecr]

Nice bit of shameless self promotion there Oluf......:mad:

Thank you, also, for pointing out what "Engine anti-ice" is, as many of the airline professionals and the AAIB would OBVIOUSLY not have a clue as to what causes it or how to look for traces of it after an accident! :rolleyes:

Just because you can't find any ice in the engine doesn't mean you can't establish an icing condition using collected data!

"that caused the hesitation" - there was no hestitation, the engine was completely unresponsive and continued running at idle until impact.

[obie2]

PAXboy wants to know if the time from 4DME to touchdown equates to 50 secs, or there abouts. No one seems to want to tell him!

Assuming 180 kts average over 4nm= 80 secs.

Adjust the average, adjust the time...simple! :ok:

[borghha]

dear Oluf,

I read your anti-ice plea 'for your grandmother to understand' and found it not only very amusing but very interesting indeed. (I am a frequent SLF, and being an interpreter, was able to read your text in danish)
However, your analysis refers only to stator/rotor blade ice formation while taxiing and the way to get rid of it prior to TO and after landing. So if I understand you correctly, what you call 'the hesitation' of BA 038' engines could have been caused by ice debris, but can what happened to this flight (engines throttling down to somewhere above flight idle and staying there) be described as mere 'hesitation'? or do you think there wasn t enough time for the engines to resume normal operation? and how would this relate to the cavitation found in the HP pumps?

and, following from your theory, no 'green approach for you in certain weather conditions?

Fly safely.

[Me Myself]

Since the BA38 crew, unknowingly, flew a: "Green Approach" (idle

Surely you're joking right ?? Unknowingly ??? This approach has been going on since basically .........the ice age !! I find myself repeating " Roger, 160 kt til 4 DME " in my sleep.
Since you are coming from 210 kt down to 180 kt and requested to keep 160 kt til 4, this is basically a green approach.
I suggest that, like the rest of us, you wait for the final report instead you make any phony assumptions.

[davecr]

PS Oluf, how often do you fly 160kts IAS fully conifgured at idle thrust?

:ugh:

[Oluf Husted]

Dear davecr,

According to the AAIB report, page 2, from feb. 2008 (latest): "The Engines initially responded but, at a hight of about 72o ft, the thrust of the right engine reduced. Some seven seconds later, the trust reduced on the left engine to the same level"

That is "Hesitation" in my book.

The different Jet Engine Producers, Airbus & Boeing and The Airlines has over the last 50 years been "Out of Synch" about recommendations and actual Anti-ice and De-ice Procedures.

And not until 1994, it was never admitted that, at idle power, the anti-ice systems does not work.

Oluf Husted

[Oluf Husted]

Dear davecr,

Anytime I had speed to "bleed of" but at 1ooo ft, (or at the very latest) 500 ft, pilots have to be "stabilized" and so it seems to be with the BA38 crew.

Oluf

[davecr]

Dear Oluf,

As I too aim to be stable at 500' on a daily basis, I am well aware of this :)

It seems to be a language problem here, as you seemed to suggest that the aircraft was idle all the way down to being at its "stable" criteria. Furthermore, our interpretation of the word "hesitation" seems to differ slightly.....

Nevertheless, I think you are out of order to be speculating about the cause of this accident, especially using the words that you did. Excuse me for having more trust in the AAIB then an engine icing enthusiast.

Can I also point out that this type of approach is not carried out unknowingly, but is highly encouraged due to fuel savings and noise restrictions in this part of the World? As mentioned before, min 160kts until 4DME or somewhere along those lines is the standard at London airports.

Dave

[Oluf Husted]

Dear borghha,

Right, no (fully computerized) "Green Approach" to me, but a late descent and a less than idle power setting when the anti-ice systems "on" is OK!

It is also valid, when airborne, what i wrote about run-ups during taxi "out" and "in. You need MORE than idle power, to get hot air enough, to keep the metal/platinium in the front end of a jet engine above freezing point.

So, my guess is, that the BA38 flew at idle power from FL400 to 720 ft, and even if the selected anti-ice "On" it does not work at idle, even if it should according to certifications.

Oluf

[Oluf Husted]

Dear Me Myself,

"Unknowingly" means, that crews very seldom stay at idle power during the whole descent, since ATC req. some "level flight" or "holding".

But the BA38 crew had a very light plane and "not headwind" during descent,
they even entered and departed "Lamborne Holding" descending at idle power.

Dear fellow pilots, have you read any other (and better) possible cause of this 104 day old mystery?

Oluf

[Starbear]

Oluf

On a modern jet engine, which parts are actually anti/de-iced by "hot air"? The answer, I suggest, may help to focus some of the wilder speculation going on here.

And do you really think the absence of any "remaining" ice due to melting will have escaped the AAIB?

[Green-dot]

Oluf,


But the BA38 crew had a very light plane and "not headwind" during descent, they even entered and departed "Lamborne Holding" descending at idle power.


Is that your assumption or is it based on facts? The AAIB only mentioned that the aircraft entered the hold at Lamborne at FL110; it remained in the hold for approximately 5 minutes, during which it descended to FL90.

Although that would give an average 400 ft/min rate of descent, the aircraft may have maintained FL110 for 3 or 4 minutes before descending to FL90? No mention in the report how the hold was performed. Or does a descent in the hold include a standard descent procedure?


Regards,
Green-dot

[Oluf Husted]

Dear Starbear,

The jet engine "Lips", the compressors big-fan blades and the very first part
of the "core-engine" is kept above the freezing point by hot bleed air, if you remember to turn the system "on" and increase the power setting.

"The higher the temperature, the more severe is the icing situation" up to +10 degrees C. (since the cloud water content, at 100% humidity, is approx. the double at +10 C, than it is at 0 degrees C.)

The amount of ice needed to make the engines to hesitate, can easily be gone, before the AAIB gets around, when the ground temperature is/was 11 degrees C.

Oluf

[Dont Hang Up]

One thing is clear - the BA38 incident was self evidently an exeptional event. I am doubtful about how the engine icing theory fits this bill. If such a relatively unexeptional approach profile can cause this problem, why are B777s not dropping like flies?

[Oluf Husted]

Dear Green-dot,

I is my assumption, since also i only have the AAIB feb. report, but if BA38
just had some speed above "Holding Speed" when entering Lamborne at FL110 and was cleared to FL90 early, they could have stayed at idle power.

The AAIB report don't even mention if the engine anti-ice system was used,
very strange, since the actual weather was "Broken Clouds" over London.

Oluf

[Oluf Husted]

Dear Dont Hang Up,

During my 30 years of descents, I almost never did one without some level-offs or "Holdings", but I think, that was what the BA38 crew did.

And further more, I was almost always "Stabilized" no later than 1000 ft. and had my engine anti-ice system "on" when flying in and out of clouds.

Oluf

[3Greens]

Oluf

Wild,wild speculation my friend. And for the record is the anti-ice system on the 777 not automatic..i.e.doesn't require "switching on" in the traditional sense.
Also i understand that at idle thrust adequate engine anti-ice is provided by the switch from low-stage to high stage bleed air. Otherwise no anti-ice would be available from high flight levels for long periods of time.
Just leave it to the AAIB who i'm sure will have considered this at some point QED.

[Oluf Husted]

Dear 3Greens,

Any descriptive of the Anti-ice systems (on most jet engines) claims to work, also at idle power, but that is mostly to comply with the certification, not in the real world.

P & W send at notice to: "All costumers" in the year 1994, that you need more than idle power to have any/some effect from the anti-ice systems.

Oluf (read all about it here: www.whistleblowers.dk )

[PAXboy]

Don't Hang UpIf such a relatively unexeptional approach profile can cause this problem, why are B777s not dropping like flies?I think the point being made is NOT so much about the descent profile but the exceptionally low temperatures that were encountered in the cruise. Many other AC encountering the same cold air mass had to drop height. Some people are considering that the extended cold soak of the fuel was then met by icing conditions during the descent.

Which is why this one-off event is so perplexing. There may well be guidance from the AAIB and manufacturers, there may well have to be another five years elapse before it happens again and more data is captured. It may never happen again.

obie2Assuming 180 kts average over 4nm= 80 secs. The 50 seconds I read was quoted earlier in this thread but, as I recall, not more than 60 seconds elapsed from loss of power to ground contact. Thinking time for a unique event that no one had ever encountered before? And yet some (I am merely an observer) criticise the crew for not taking out A/P and all sorts of other things. It is thanks to them that we can be discussing data and a very different AAIB problem.

[spannersatKL]

I see some columnist in The (soaraway) Sun has postulated his theory today....no gas in the tank? Is this to be discounted as pure Journo speculation....or is the investigation leaking again?

[lomapaseo]

I'm sorry to read so many posts using the words engine icing and anti-ice in such an unsupportable matter regarding BA038.

Not for many years have these large fan engines used piped hot air to warm up the blades or vanes to keep ice off them. In the air they depend on the compression cycle temperature rise, centrifugal force and air air-velocity loads to shed accreted ice at low enough mass levels to not damage the engine internal parts or affect the engine surge stability.

Whilst tis true that the culprit ice does melt, its evidence on the engine that it leaves behind is unmistakeable to an investigator.

So to me the credibility of the engine teardown examination as well as the memory readouts pretty much eliminates this area still in consideration.

[Oluf Husted]

Dear lomapaseo,

You and a few others think, that I should wait for the AAIB or drop my theory of possible engine ice produced "hesitation"

OK, can you explain, why the FAA issued an AD on the 23. jan. 2008, only six days after BA38s "grass-landing" in order to:

"Prevent internal engine damage due to ice accumulation and shedding, which could cause a shutdown of both engines, and result in a forced landing of the airplane"

It was a 10 page AD called 2008-02-05 (or Directorate Identifier 2005-NM-263-AD) effective Feb. 27. 2008.
Valid for: Boeing Model 777-200 and -300 Series Airplanes Equiped With Roll-Royce RB211-TRENT 800 Series Engines.

Read it yourself, the improved procedure, calls for a more strict "engine anti-ice ON GROUND procedure" why this can avoid shut downs when airborne beats me. But it is the only change of procedures, since 17. jan.
2008.

So, at least the FAA thinks that engine ice might be the villain, why have the AAIB not commented on this AD.

Oluf

[Oluf Husted]

Dear lomapaseo,

You and a few others think, that I should wait for the AAIB or drop my theory of possible engine ice produced "hesitation"

OK, can you explain, why the FAA issued an AD on the 23. jan. 2008, only six days after BA38s "grass-landing" in order to:

"Prevent internal engine damage due to ice accumulation and shedding, which could cause a shutdown of both engines, and result in a forced landing of the airplane"

It was a 10 page AD called 2008-02-05 (or Directorate Identifier 2005-NM-263-AD) effective Feb. 27. 2008.
Valid for: Boeing Model 777-200 and -300 Series Airplanes Equiped With Roll-Royce RB211-TRENT 800 Series Engines.

Read it yourself, the improved procedure, calls for a more strict "engine anti-ice ON GROUND procedure" why this can avoid shut downs when airborne beats me. But it is the only change of procedures, since 17. jan.
2008.

So, at least the FAA thinks that engine ice might be the villain, why have the AAIB not commented on this AD.

Oluf

[forget]

AD here. Page 2 and on.

http://www2.lba.de/dokumente/lta/2008/2008115.pdf

[Sunfish]

This is pure speculation from someone no longer associated with airlines, but I am afraid it may not be so easy to discard Mr. Husted's theory.

I note that the latest AAIB report says:

Data, downloaded from the Electronic Engine Controllers
(EECs) and the QAR, revealed no anomalies with the
control system operation. At the point when the right
engine began to lose thrust the data indicated that the
right engine EEC responded correctly to a reduction
in fuel flow to the right engine, followed by a similar
response from the left EEC when fuel flow to the left
engine diminished. Data also revealed that the fuel
metering valves on both engines correctly moved to then
fully open position to schedule an increase in fuel flow.
Both fuel metering units were tested and examined, and
revealed no pre-existing defects.

Posters have jumped to the conclusion that if the fuel flow increase was scheduled, then, if engine thrust did not increase, therefore the scheduled fuel flow increase did not happen. But, unless I'm missing something somewhere else, that's not what the report says.

Absent independent data from a fuel flow sensor showing that fuel flow did not increase, then I wonder if it's possible that fuel flow did increase, but that Mr. Husted's core icing prevented the engines from generating sufficient thrust?

If this were the case, I suspect there would be considerable egg on a lot of faces in the industry.

P.S.: In a fully flapped (40 deg) Cessna on a 3 degree approach at 500 ft, at engine idle I can't reach the end of the runway either. Please think about it before criticising the pilots.:E

[Oldlae]

I don't think that increased fuel flow would be prevented by core icing, if fuel flow was reported to be increased with no response then icing could be suspected, but fuel flow did not reportedly increase. As I understand the term, the engines did not hesitate, they merely carried on in accordance with the fuel they received. Regarding the AD, as far as I am aware, freezing fog was not present nor poor vis.

[Oluf Husted]

Dear Oldlae,

The definition of a cloud could be: "Freezing Fog" or just "Fog" and the visibility is always lousy in fog, and that is what the AD warns about.

Oluf

[Sunfish]

Oldlae:

but fuel flow did not reportedly increase

No Oldlae, you are making an assumption. Nowhere in the report does it say that the fuel flow did not increase. What it says is that "the fuel
metering valves on both engines correctly moved to then fully open position to schedule an increase in fuel flow." You are making the causal assumption that since this occurred, and the engine did not respond, then the increased fuel flow did not occur.

As far as I can tell, the AAIB used those words because it does not have the data to show whether an increased fuel flow occurred or not.

To put it another way, if the AAIB knew for a fact that the fuel flow did not increase, they would have said the fuel flow did not increase, but that is not what they said.

If Mr. Husted's theory is correct, maybe an increase in fuel flow did occur, but the engine did not respond because of core icing.

[wilyflier]

Sunfish,
....Presumably an increase of fuel flow without rotational acceleration would have led to higher engine temperatures.Wouldnt these show on FDR or QAR ?

[airfoilmod]

"The EEC responded correctly to a reduction in fuel flow to the right engine"

Read your block quote again, there was a reduction in fuel flow. To both engines, followed by correct mechanical responses. Cavitation could not have happened if Fuel had been flowing within parameters.

"Followed by a similar response in the left EEC to a diminished flow..."

Since both engines continued to run, though not at commanded thrust, one assumes ignition was present, but insufficient Fuel to produce commanded thrust. Interesting thought though, and #2 did cease spinning prior to cowl crushing at impact. Hmmm.

[chase888]

If core icing did occur, then the fuel most probably was no longer homogenous, particularly if wax was floating on the surface.

Most probably everything worked as advertised, but the heavier parts of the fuel(and water) did not possess enough heat(kJ) to provide the required thrust, despite the amount of fuel in all probability being correct for fuel in its normal condition.

Not sure, but I estimate that the fuel delivery pipe would be around 4" dia to give enough fuel for max. power, so ice or anything else would not restrict flow, particularly as the fuel is under pressure.

My final 2 cents worth is that evidence of cavitation on the pumps is only significant in that it takes quite some time for cavitation to become evident on the impellor surface, so most unlikely to have been a contributor to this problem.

Cheers

[Green-dot]

Dear Oluf,


OK, can you explain, why the FAA issued an AD on the 23. jan. 2008, only six days after BA38s "grass-landing" in order to:

"Prevent internal engine damage due to ice accumulation and shedding, which could cause a shutdown of both engines, and result in a forced landing of the airplane"


After reading the AD, i think you are taking this passage out of context.

The statement in your post is incomplete. It is preceded by, i quote:
"This AD results from reports of engine surges and internal engine damage due to ice accumulation during extended idle thrust operation in ground fog icing conditions. We are issuing this AD to prevent internal engine damage . etc. . ."

Shutdown of both engines in this case could occur during or immediately after take-off, resulting in a forced landing. If there would be a relationship to the BA38 "grass landing" i would have expected the AD to have included prolonged CDAs in icing conditions also. Since the AD does not, i read this AD as valid in ground fog icing conditions only.

Also, there is a difference between ground idle (minimum idle) and approach idle. Approach idle is automatically selected in flight, resulting in a higher engine thrust setting when engine anti ice is on or when landing flaps are selected. For the RR Trent the EECs use EPR as primary thrust setting parameter. If core icing would have been a factor i would assume EPR would have started to deviate from normal settings in relation to the other primary engine indications, N1 and EGT. If not noticed by the crew, i think they would have noticed if this were the case, this would have certainly shown up in the DFDR and QAR parameters. No mention about such deviations is made in the AAIB reports.


Regards,
Green-dot

[airfoilmod]

see page 34 (this thread) post #662.

Captain Carbaugh, Boeing chief pilot, "The fuel pumps had been cavitating not long before impact, perhaps as a result of a fuel restriction or low temperature...etc." So Boeing places cavitation in a frame well within one incorporating cause. It is consistent with starvation or diminished flow of Fuel. It also places focus on Roller, not that that was anyone's intent.

[Oluf Husted]

Dear Green-dot,

Point taken, the AD 2008-02-05 does involve ground fog icing conditions only, and that in e x t e n d e d periods of idle power.

The BA38 probably flew a very extended period at "flight idle" power, from FL400 to 780 ft. Since engines have slightly different "lowest idle RPM" and the AD deals with "core ice" problems, I still think BA38 could have had some ice build-up, enough to make the engines "hesitate".

Furthermore this AD claims to be issued to lower the risk of both engine shutdowns, but all it does, is lowering the visibility range, where pilots should/shall do "Run-ups" from 1000meters to 300meters in freezing precipitation.

This is very strange, since the renowned Norwegian Meteorologist Åsmund Rabbe already in 1999 revealed his findings: "That ice danger goes UP by factor 100 when visibility, in freezing precipitation, increases from freezing fog to freezing rain, this due to droplet size." (if bigger, far worse)

Maybe the "timing" of this AD (six days after) was just accidental, but to start doing far less "Run-ups" has earlier resulted in engines being damaged by ice (on the ground and in the air) at Oslo Gardermoen 14. dec. 1998 (20 engines) and at Denver Airport in 2002 (15 engines)

Oluf

[woodpecker]

So it's just a coincidence that, while this "core icing incident" unfolded the evidence is that everything in the FADEC worked flawlessly except that the fuel flow never reached the commanded level.

I think not. Less than commanded fuel flow is a fact, core icing is not. Show how the two could be inter-related then you might have a case, but until then ......

[BOAC]

and #2 did cease spinning prior to cowl crushing at impact. Hmmm. - can you direct me to the source of this fact as I was unaware of it? As far as I can see the AAIB state both engines were running but delivering less than adequate power up to impact - or have I missed something?

[Shaka Zulu]

@ Oluf Husted

Interesting site you have however you are starting to misinterpret AD's and connecting it up to BA38. You may be right (can't exclude any possibilities at this stage) but what you are saying can't be connected up with the evidence given.
Reading the report you could have seen that the engines weren't operated at Flt Idle and App Idle from F400 to 780'AAL. And for the many BA38 flights I have operated into LHR, a flight idle approach is VERY VERY unusual.

The reason for the AD from RR was because more engine data had been collected over it's running life henceforth they amended the ground procedures for core ice shedding.
One thing I would like to say though: Why does Cathay operate the EAI and WAI manually, where so many 777 operators don't?

[Mariner9]

..the fuel most probably was no longer homogenous, particularly if wax was floating on the surface.

Wax is (a lot) denser than kero - it will sink not float. Though what actually happens in a cold tank is that the wax layer "grows" outwards from the cold surfaces in the tank, till eventually there is sufficient wax to form an insulating barrier, greatly reducing further cooling. :8

[tanimbar]

lomapaseo you wrote:


Whilst tis true that the culprit ice does melt, its evidence on the engine that it leaves behind is unmistakeable to an investigator.



Can you explain please or recommend a document/paper that covers this subject?

And, does the evidence indicate past icing events?

I ask purely out of interest.

Thank you, Tanimbar

[Me Myself]

Fellows

What if the AAIB and Boeing ended up not having a clue about what happened. Hard to come with a report when you have nothing to say. Don't you think ??
It seems to me that since there isn't anything conclusive they must be raking their brain to come out with something that isn't damaging too to Boeing, BA and finally the crew.
Still, I would think Boeing has a lot to answer for.
I am just patiently waiting for the final report. All these speculations are, in my view, a waste of time.
The only undisputable facts are :
- Both engines retarded to flight idle ( didn't stop !! )
- Crew saved the day.

[Oldlae]

Sunfish.

The AAIB report states that shortly after the FO took over the AT commanded an increase in thrust from both engines. The engines then responded initially, therefore fuel flow must have inceased albeit momentarily.

[oxo]

We've had nothing from the AAIB for some months now. Anybody know how long it might be before more information is forthcoming?

[Bis47]

The only undisputable facts are :
- Both engines retarded to flight idle ( didn't stop !! )
- Crew saved the day.

1. Both engines retarded somewhat above flight idlle. Too bad the AAIB does not state exactly how much ...

2. The crew saved the day? Can you bring some insight? Because, according to the AAIB :
- the "flying" (co-)pilot forgot to fly the aircraft - just doing nothing, letting the autopilot fly the standard glide slope, bleeding airspeed down to the stall.
- the captain failed to monitor copilot lack of responsiveness while the speed was decaying down to stall speed at 170 ft agl. Stalling the aircraft from 170 feet led to a very hard crash landing. Other options were clearly available.
- neither the captain nor the copilot managed to restore engine thrust.

Actually we don't now exactly what were the crew actions if any. The release of the VCR will bring some insight ...

You shoud know perhaps the basic principle of crew coordination : The Flyin Pilot FLIES the aircraft, The non flying pilot "manage" the flight. Both crosscheck each other. It works pretyy nice in normal operation.

In hard emergency ... no room is left for crosschecking. That is the "two" pilots crew shortcoming.

So, perhaps the two pilots were just "frozen" ... Nobody would blame them due to the exceptionnal nature of this incident, the lack of information available to them, the lack of briefing discussion and the lack of minimum training.

However, writing that they "saved the day" is a little far fetched ...

[Spaz Modic]

:D No Oldlae, you are making an assumption. Nowhere in the report does it say that the fuel flow did not increase. What it says is that "the fuel
metering valves on both engines correctly moved to then fully open position to schedule an increase in fuel flow." You are making the causal assumption that since this occurred, and the engine did not respond, then the increased fuel flow did not occur.

:= Can't schedule fuel to the injun when there ain't any in the tanks.:hmm:

[stickyb]

Posters have jumped to the conclusion that if the fuel flow increase was scheduled, then, if engine thrust did not increase, therefore the scheduled fuel flow increase did not happen. But, unless I'm missing something somewhere else, that's not what the report says.

I think that is what the report says.


At the point when the right engine began to lose thrust the data indicated that the right engine EEC responded correctly to a reduction in fuel flow to the right engine, followed by a similar
response from the left EEC when fuel flow to the left engine diminished.


The quote says specifically that there was a "reduction in fuel flow" to the rigght engine, and that the "fuel flow to the left engine diminished"

For those statements to be made something must have been measuring fuel flow, and those measurements must be in the hands of the investigators.

[wilyflier]

Bis47,
.......I think for whatever reason the stall did not occur at 170 ft . Free fall from that height calculates a far higher vertical speed than the 2000 ft per minute indicated by the damage.(Edit - I assumed stall to occur from an existing descent rate of about 1100 fpm)
......My rough calculations indicate a stall about half a second before impact, below 50 ft (I got a nominal17ft agl) with the nose just beginning to fall, such that there was no visible tailstrike mark.
....I guess if it was in a mushed but still steady descent at 2000fpm there would have been more tail down

[Me Myself]

Other options were clearly available.
- neither the captain nor the copilot managed to restore engine thrust.


Yeah ??? Like whot ???
I did what happened to them on the sim, very same spot and apart from the fact I disconnected the A/P right away, I just flew the aircraft to the very same spot they did, just didn't stall, or very close to the ground. At any rate the motion didn't trip meaning it wasn't too hard of a landing.
I knew what was going to happen.......they did not !! I don't know who you fly for, but I do not remember my outfit training me for the loss of 2 engines at 700 ft after a 11 hours flight........or maybe I missed class that day........again.

They bloody did.......and nothing happened.

Now let me tell you something about the F/O " lack " of reaction. 15 years ago while on approach on a south american airport, a traffic pulled out of a cloud, right on my right hand side and the only profound thing I managed to utter was " Oh **** " He missed us by 12 feet !! ( no TCAS in those days ). I was the handling pilot and thought I had reacted pretty swiftly. As it turned out, I disconnected the A/P 7 sec after the initial conflict.
The skipper only understood what happened once he saw the traffic on his side and I think we also got a " Oh **** ".
Were we under stress ??? ........" oh **** YES !! " I also remember how freaked out we were that we may have stuffed up and although facts pointed to the contrary we still doubted. It was most unpleasant ( huge understatement !! ). It also took me some time to become my confidant self again. What about now ??? I love my TCAS !! Best thing invented after the dishwasher.

I think you are being most unfair totally disregarding the " surprise " factor here.
The aircraft also bled its speed when I flew it on the sim. Apparently, according to previous posts, BA seems pretty strict about their SOP's. Where I work, I would have jumped right on the controls. No right or wrong here, just different rules which kick right in under stress. That's the way it"s supposed to be.
If BA wants to change its SOP's, they'll do it in due time if they see it fit, but flogging the crew won't bring anything to the AAIB report.

[atakacs]

I don't know who you fly for, but I do not remember my outfit training me for the loss of 2 engines at 700 ft after a 11 hours flight........or maybe I missed class that day........again.

Can we all concur that in fact there was very little they could do given the odd circumstances surrounding this event and it's timing ? As far as I am concerned I would be the last to blame them but I also wonder if it was wise to proudly stage them two days after the crash as heroes...

[L337]

Spaz Modic:

You are like a broken record:

Can't schedule fuel to the injun when there ain't any in the tanks

:D It takes a while to conjure up the necessary technical gobbledegook type spin to pad out a 1000 page report document which finally comes to the conclusion it was outa gas. :ok:

:* The ultimate update - it was outa gas! :bored:

:confused: Notwithstanding the B777 fuel system having been rebuilt several times over, the odds are still, it ran outa gas :cool:

know this is a long shot, but, perhaps, maybe, possibly, or just bloody factual that significant aeration and cavitation of all the fuel pumps almost simultaneously could possibly maybe that the thing was OUTA GAS?

Yet the AAIB report clearly says:

The total fuel
on board was indicating 10,500 kg, which was distributed
almost equally between the left and right main fuel tanks,
with a minor imbalance of about 300 kg

A significant amount of fuel leaked from
the aircraft after it came to rest

[Me Myself]

but I also wonder if it was wise to proudly stage them two days after the crash as heroes...

I'm not sure they had much of a say. They were basically taken out of the rubble right to head office. It looks it was more than likely " Uncle " Willie's decision.
The media and the travelling public demand immediate answers and crucifying a crew and.......don't forget the aircraft, isn't going to drive sales up is it ??
Tell you what, between the 2, I'd rather be made a heroe, much nicer when pushing my trolley at the supermarket I tell you.

Can we all concur that in fact there was very little they could do given the odd circumstances surrounding this event and it's timing ?

I really think we should.

[bsieker]

Summing up the answers to some recurrent ideas, and also some (seemingly) new ones.

1/ Mr. Spaz "outagaz" Modic

L337, regarding Spaz Modic, I might just say, "Please don't feed the Troll".

But in addition to what the AAIB report says about fuel on board, see the very good and insightful discussions (in particular Green-Dot's post #878 from April 14, thanks a lot for that!) on frost on the wing underside, and its relation to fuel levels in the main (wing) tanks.

Unless you're a diehard conspiracy theorist, this shows convincingly that the figure of just above 10 tons mentioned in the report is about right.


2/ AD 2008-02-05 and Core Icing

a- Timing

The release of said AD in temporal proximity to the accident we're discussing is coincidental. It deals with extended (in this case: more than 45 or 60 minutes) ground operations in freezing fog of low visibility. Said release is to reduce the threshold distance from 1000 to 300 metres visibility for the described procedures.

b- Conditions relevant to the AD

BA038 didn't perform extended ground operations in freezing fog.

What is the typical time, including a 5-minute hold, from TOD to touchdown for a flight on this route? I have no idea, but I guess it will be less than 45 minutes, a considerable time of which will be spent outside freezing fog, and above idle thrust.

Even if clouds in cold temperatures (AD talks about the three distinct cases of 0 to -6 C, -7 to -13C, and below -13C) could be called "freezing fog", BA038 will not have spent anywhere near 45 minutes inside them at near idle thrust. More to the point, takeoff in the conditions described in the AD is even permitted after up to 60 minutes without intermediate run-ups, i. e. engine damage leading to immediate failure is not expected by the ice accreted within that time, although slight engine damage may occur, so an inspection within 10 flights is mandated.

c- Core Icing

We cannot yet exclude core icing as a causal factor in this accident, but the dangers of core icing as described in AD 2008-02-05 are engine shutdown and/or engine damage when accreted ice chunks above a certain size brake loose. For BA038 we are not talking about a shutdown, but merely a reduction in thrust. It may be related, but we don't know yet.

3/ Fuel flow

The AAIB report states that the fuel flow reduced, first in one, and then in the other engine. It also states that the EECs responded correctly by opening the fuel metering valve. Actual fuel flow is measured after the metering valve, and we can be quite certain that the AAIB would have mentioned an increase in fuel flow caused by opening the metering valves, which was not accompanied by a corresponding increase in thrust.

Regarding ideas about a full flow rate of a liquid with just not enough energy content: I find it hard to believe that a liquid with an energy content just enough to deliver slightly above idle thrust at maximum possible flow would burn at all. We're talking about a fully open metering valve, which would, given normal fuel, equate to above TOGA thrust (TOGA will not be the limit of what the fuel system can deliver, but a limit imposed by the software in order not to damage the engine).

Fuel (or rather: energy) flow needed for thrust slightly above idle is only a small fraction of the total fuel delivery system capability. All fossil fuels have roughly similar energy contents, so the largest percentage of the liquid delivered in this scenario would have to be something uncombustible, dare I say water? Would a liquid of mostly water be able to sustain combustion?

Wouldn't it also leave highly supicious traces in the fuel lines, bound to be found by the investigators? How does that reconcile with the AAIB's stating that ...

[...]
there were no signs of contamination or unusual levels of
water content. A sump sample taken from the left and
right main fuel tanks shortly after the accident revealed
no significant quantities of water.

There are (or were) a couple of fuel chemists and other professionals of the field in the thread, I'd like to hear their view on this.

d- Hesitation

Maybe I have a strange book, but in all the dictionaries "to hesitate" means to "pause before doing something". Some add "uncertainty" as a reason for this kind of pausing. I cannot see anything in this accident that would fit that definition. A thrust increase was demanded, fuel flow and thrust increased (without any mention of a delay, or "hesitation"), and then fuel flow and thrust reduced, despite further increased demand, and stayed at that low level.

So you might say it "hesitated" to reduce thrust uncommanded. I don't think that would be a useful description.


Cheers,
Bernd

[FireLight]

Great summary bsieker. :ok:

The problem with any credible theory is that it has to fit those pesky facts.

[RomeoTangoFoxtrotMike]

We've had nothing from the AAIB for some months now. Anybody know how long it might be before more information is forthcoming?
I commented on this earlier, with some stats I'd come up with (on average, it takes a little over two years from incident to final report for a major investigation).

[Edited to note that my previous post was temporarily removed...]

[Green-dot]

Quoting FireLight:

The problem with any credible theory is that it has to fit those pesky facts.


And for facts you need the data and evidence to prove a theory.

If evidence remains unavailable, the next best thing is try and simulate the conditions and observe systems behavior against the data available. If the cause is still unknown at this time, apparently something happened which cannot be retrieved from the recorded data.

That brings me to bsieker's clear summary. Focussing on the fuel flow and referring to my posts #590 "Code One . . ." on page 30 and #810 "What if . . ." on page 41 of this thread:

If the investigation remains inconclusive, would the AAIB wage a temporary repair of the aircraft, sufficient to install engines and landing gear/flaps and simulate at least the last minutes of the flight with fully operational systems? Perhaps from the moment the aircraft entered the Lamborne hold.

The engines would only have to operate near approach idle, with the aircraft ground/flight sensing system simulating the in-flight mode. The gear installed to support the aircraft only. Gear retract actuators disconnected from the gear but fully functional with weights attached to simulate the gear extension (the same for uplocks and door actuators). Then perform (including all system selections i.a.w. the SOP) the complete final approach sequence of lowering flaps, gear and consequent engine response to the changing aircraft geometry, changing to the landing configuration. If it was a problem in the aircraft's systems which ultimately exposed itself in the fuel system, perhaps it started during this sequence. Somehow affecting fuel flow, but going unnoticed by the crew until the engines required increased fuel flow (apparently to above approach idle) at approx. 720 ft.


Green-dot

[Seloco]

A sad but inevitable aside is that MM is now surrounded by temporary screens as she gets scrapped; no fin or wingtips as of yesterday.....:{

[lomapaseo]

Green Dot

And for facts you need the data and evidence to prove a theory.

If evidence remains unavailable, the next best thing is try and simulate the conditions and observe systems behavior against the data available. If the cause is still unknown at this time, apparently something happened which cannot be retrieved from the recorded data.


Ah, a true scientist among us

However this kind of "let's see what happens" stuff has too many drawbacks.

You can never be sure that you have actually got all the varriables tinkered to perfection and thus any answer that you would like to draw is challengeable by all except yourself.

It's akin to running an experiement to see how it breaks. When all that we can really use is to confirm that it doesn't break enough times to be deemed safe.

In the long and short of it we can't build an airplane that won't break, all we can do is build one that won't break most of the time (there is a difference)

[airfoilmod]

But the search for zero defect, zero fault cannot be given over to cynicism, either. All of Human endeavor is a balance of some kind between expected outcome, intended outcome, and potential outcome. The search for perfection may be ridiculed in the Boardroom, or on the Line, but must never be abandoned by analysts, engineers, or manufacturers.

BA038 is a unique opportunity to analyze an as yet unidentified failure. Should this endeavour be left to only a few? We may want to identify others who have a stake in air travel, passengers, employees, families,etc. There is an unfortunate adversarial relationship at times in failure analysis. This is regrettable. Public scrutiny has been part of our cultures for as long as anyone can remember; that is why our aviation industries have such excellent safety records, and "closed societies" records are lamentable.

[airfoilmod]

Besides the apparent ongoing salvage disassembly, the impact without doubt precludes any reasonable expectation that MM could be rebuilt in any way to even approximate pre landing form. A better, and more obvious method would be to approximate conditions surrounding the incident with a duplicate, a sister ship, as Boeing has done (is doing?). That resulted in public comment from Captain D. Carbaugh, Chief safety Pilot at Boeing. His conclusions paralleled those of preliminary findings by AAIB. Until a new report from the Investigative Authority, perhaps a new thread to discuss ETOPS, SOPS, and type?

[Green-dot]

lomapaseo,


You can never be sure that you have actually got all the varriables tinkered to perfection and thus any answer that you would like to draw is challengeable by all except yourself.

It's akin to running an experiement to see how it breaks. When all that we can really use is to confirm that it doesn't break enough times to be deemed safe.


Basically, i agree with you. Although it might seem an unorthodox approach, my suggestion is not meant to test it until it breaks but utilise the original complete article while it is available. To run through all the systems and sequences in their interacting, normal, functional environment. This to collect/reproduce data from existing parameters, complemented with parameters that may be suspect as a result of the investigation but not recorded before and lacking evidence of a (possible) dormant failure. What if, while performing such unique excersize, the problem is reproduced?

Obviously, this would be of added value only, provided that at this time no conclusive evidence has been collected to correct whatever caused the problem.

As long as MM is relatively intact, it would be an opportunity. With the airplane systems basically intact, all its removed components are probably quarantined for the investigation, it is at least an option. Once the scrapper starts putting its teeth in it, the opportunity is lost. From what i read from Seloco's post today, that opportunity is quickly evaporating . . . . it could also indicate that the AAIB may very well have progressed further than published.


Green-dot

[gas path]

The fin was removed as an assembly at the request of the new owners, the insurance company! The fin is for sale...any takers?
No part of that aeroplane will ever be used on any BA a/c!
The aircraft will not be scrapped for some time as the AAIB are still very active on it. The fuel manifolding is going to be removed in one piece for assembly into a test rig.
The fuel removed from the aircraft, approx. 9 tonnes is currently quarrantined in a bowser and will be transferred into a specially made storage tank and kept until the investigation is over.

Core icing! That procedure has been in place for at least a year, and I can't see any difference in the latest release of the AD:confused:

[Green-dot]

But the search for zero defect, zero fault cannot be given over to cynicism, either. All of Human endeavor is a balance of some kind between expected outcome, intended outcome, and potential outcome. The search for perfection may be ridiculed in the Boardroom, or on the Line, but must never be abandoned by analysts, engineers, or manufacturers.

Exactly :)



Until a new report from the Investigative Authority, perhaps a new thread to discuss ETOPS, SOPS, and type?


Good suggestion :ok:

[Oluf Husted]

Dear beseiker, (Bernd)

Flight time from TOD to 780 ft. (incl. one holding) is app. 30 min, but the metal in the very front-end of the core engine was 10 degrees lower (due to the pressure drop) than the outside temp (minus 66 to +0 C.) during descent.
So if the engine anti-ice system is not "on" or not effective at flight idle, the cloud water content clings to stator and rotor-blades, even better/worse, than when on ground in freezing fog.

My theory then is, that BA38 had just enough ice to destroy the aerodynamics and hesitate, but not to shred and damage the blades, that ice would long be gone before the AAIB ever came around.

Oluf Husted (only PPRuNer by name)?

[TyroPicard]

Oluf
Can you please explain how "just enough ice" leads to reduced fuel flow? With the Thrust Levers fully forward?
TP

[airfoilmod]

Yes, exactly. Without the required Fuel, what additional impact would shed-ice have? Also, Oluf, what do you mean by Aerodynamic? As in "destroy the aerodynamics and hesitate." I am still having problems with your word "hesitate". The Engines responded initially, then spooled down, to become "unresponsive" to inputs from both A/T and flight crew. Nothing hesitant about any of the Powerplant issues. Not to me.

[Oluf Husted]

Dear TyroPicard ad airfoilmod,

I newer flew with RR engines, but my P & Ws on a MD-80 once had a very hard time getting up in RPMs with the front end of the engine all iced-up, it first "hesitated" then started to stall, first after three tries and very slow advancement of the throttles it worked normal again. (Eng. Anti-ice was ON)

This was in 1989, the year before my airline got compulsory "Run-ups" (every 10 minutes and always before T/O) also during taxi-in.

The BA38 never came to the point, when the stalls sets in, it "just" hesitated,
so until the AAIB comes up with something better, I hope all airlines will forget all about "total idle descents" (Green Approaches) without taking the "local" weather into consideration.

Oluf

[bsieker]

Oluf Husted,

the problem with that theory is that the engines did spool up, initially, until fuel flow reduced.

And I am still curious exactly which behaviour of BA038 you describe as "hesitation".


Bernd

[airfoilmod]

Both engines responded to A/T command for Thrust and ran at that setting for at least (unknown) seconds (AAIB). #2 then decelerated, then #1 five seconds after. No hesitation, no Ice (Had there been Ice, it was shed at the high thrust level, without damage to engine (AAIB). Now explain what you mean by hesitation at this point. Remember, the engines have both cycled to advanced thrust at this point, and had run there, so no core ice.

Note:
...for at least (unknown) seconds: Counting the response to power level set by A/T, then #2 abandoning setting, then 5 seconds for delay of #1 leaving the advanced and commanded thrust set by A/T.

Oluf: As long as you're at it, explain "destroyed aerodynamics".(your Post #1018). Thanks

Airfoil

[Oluf Husted]

Dear Bernd and Airfoil,

Bernd, my engines also did spool-up a bit before before hesitating and starting to stall, then I throttled back and tried again and again it worked fourth time and took a minute or two.

Airfoil, when ice builds on a curved surface (rotor or stator blade) it sticks
to the most curved part first, that is on the back-side, since it is here the pressure (and hence the temperature) drops the most and it is damaging to the fine tuned aerodynamics of a compressor.

Frankly, I am not a professor in aerodynamics, but this is common sense and a bit of experience.

Have you any better explanation as to why BA38 landed short?

Oluf Husted

[airfoilmod]

You think I am unaware of Blade profile and icing, I may be completely unable to understand your writing and discern some data from it. If so, the fault is mine. I admit I cannot easily understand your prose and what it is intended to convey. I may be too demanding in my definitions and usage of vocabulary incorporated in your posts. I can only say I took your reference to "Destroyed Aerodynamics" to mean other than airflow related to turbine compression. My mistake. Your issue was addressed long ago, and at length. I do not think it applies to this incident. Do I have a better explanation? It isn't my job to assess speculation and theory. You may be correct, I think no one is in a position to say. That's a pretty extravagant challenge, sir.