Sunfish
 

"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.

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

Dear Oluf,

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

Chase888
 

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.

BA38s engines "Hesitated"
 

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

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 ......

- 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?

@ 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?

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

How do you detect past ice accumulation on engine?
 

lomapaseo you wrote:

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

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.

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.

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

Indisputable, really ?
 

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 ...

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

I think that is what the report says.

Quote:

---

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.

stalling and guessing
 

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

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.

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...

Spaz Modic:

You are like a broken record:

Yet the AAIB report clearly says:

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.

I really think we should.

Summary
 

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 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

Great summary bsieker. :ok:

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

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...]

Quoting FireLight:
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

R I P G-ymmm
 

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.....:{

Green Dot

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)

Lomapaseo
 

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.

Green-dot
 

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?

lomapaseo,

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

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:

airfoilmod
 

Exactly :)


Good suggestion :ok:

BA38s engines "Hesitated"
 

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)?

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

Tyro
 

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.

BA38s engines "Hesitated"
 

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

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

OK
 

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

BA38s engines "Hesitated"
 

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

Mr Husted, If
 

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.