Sponsor Message:
Aviation Technical / Operations Forum
My Starred Topics | Profile | New Topic | Forum Index | Help | Search 
Question On The Accident Of BA B777, G-YMMM  
User currently offlineAjaaron From United Kingdom, joined Nov 2000, 113 posts, RR: 0
Posted (5 years 11 months 3 days 6 hours ago) and read 8185 times:

I have read the interim report regarding the incident on Boeing 777-236ER, G-YMMM in Jan 2008, http://www.aaib.gov.uk/publications/...ports/boeing_777_236er__g_ymmm.cfm

and I know it sounds obvious, so I VERY humbly offer this question:

If ice accretion caused restricted fuel flows to the engines, how could it be that both engines suffered from restricted fuel flows at exactly the same time, i.e. just 7 seconds apart, when each engine was being supplied independently from the other by its own main tank in its respective wing?

Statistically is it possible that that ice accretion would have occurred (and sufficiently so), and precisely at the same time on both independent fuel supplies to the engines?

Does it sound a likely occurence?

41 replies: All unread, showing first 25:
 
User currently offlineStarlionblue From Greenland, joined Feb 2004, 17054 posts, RR: 67
Reply 1, posted (5 years 11 months 3 days 4 hours ago) and read 8149 times:



Quoting Ajaaron (Thread starter):
Statistically is it possible that that ice accretion would have occurred (and sufficiently so), and precisely at the same time on both independent fuel supplies to the engines?

I'm going to nitpick a bit here. You can't really say "statistically possible". If there is any possibility, it can happen, however unlikely. Statistics and probability describe possibilities and results. Statistics and probabilities do not rule out improbable events. They just describe how unlikely such events are.



"There are no stupid questions, but there are a lot of inquisitive idiots."
User currently offlineStarglider From Netherlands, joined Sep 2006, 678 posts, RR: 44
Reply 2, posted (5 years 11 months 3 days 2 hours ago) and read 8129 times:



Quoting Ajaaron (Thread starter):
Does it sound a likely occurence?

Not really. After reading all the reports issued by the AAIB regarding this accident, chances of ice accretion being the cause are very remote but not impossible.

However, 5 litres of water spread out over 3 tanks in 100.000 litres (79.000 kg) of fuel just does not add up as a likely cause. The water content in the fuel was below average for a B777. Furthermore, business jets with much less wing volume fly at 40.000 ft in the winter and never encounter such a problem. Cold soaking the fuel on a business jet is reached much sooner than on a jet the size of a B777. In this particular case of BA038, the fuel temperature never went below -34 C, and stayed well above fuel waxing or freezing temperatures.

Any accumulated ice in the center tank (no more than 1.6 litres according to the AAIB) when departing from Beijing, most likely would have melted and have been consumed during the climb. Even if the ambient temperature was -6 C on the ground and fuel temperature in the left main tank was -2 C after refuelling, the fuel in the center tank must have been warmer (was +5 C when refuelling) during taxiing and climbing to initial cruising altitude with most of the center tank enclosed between the airconditioning system below it, the cargo compartments in front and behind it and the cabin above it. The report states that the left main tank fuel temperature remained at -2 C until cruising altitude, center tank fuel most likely was much warmer. Firstly because the fuel entered an empty center tank during refuelling (contrary to the fuel in the wing tanks, where the added, + 5 C fuel mixed with the cold soaked fuel remaining from the previous flight). Secondly because the fuel in the center tank remained relatively warm for a longer period of time until well into the cruise because it is enclosed as explained above.

The engines are fed from the center tank until it is almost empty (800 kg) before switching to the main (wing) tanks which contained no more than 3.4 litres of water in the fuel spread evenly over the two main tanks according to the AAIB.

Ice being the cause is a hypothesis, nothing more.

Personally I believe there are more factors in play here (not ice) which acted simultaneously at a very inconvenient moment.


Starglider


User currently offlineTdscanuck From Canada, joined Jan 2006, 12709 posts, RR: 80
Reply 3, posted (5 years 11 months 3 days ago) and read 8086 times:



Quoting Ajaaron (Thread starter):
If ice accretion caused restricted fuel flows to the engines, how could it be that both engines suffered from restricted fuel flows at exactly the same time, i.e. just 7 seconds apart, when each engine was being supplied independently from the other by its own main tank in its respective wing?

Because both engines throttled (or tried to) at the same time.

Based on the AAIB's initial report, ice was slowly building inside both systems while on descent. So, although the accumulation rate would probably vary subtly from side to side, the exchangers were both becoming restricted by ice formation. Since the fuel on both sides was essentially the same, the amount of ice formation should be roughly the same. The restriction wasn't enough to cause a problem at flight idle, but when they commanded a thrust increase on both sides it was enough to starve the engines.

Quoting Starglider (Reply 2):
After reading all the reports issued by the AAIB regarding this accident, chances of ice accretion being the cause are very remote but not impossible.

Of course it's remote...otherwise it would have happened before. The ice formation isn't time critical...the triggering event was throttling up, which happened simultaneously on both sides.

Tom.


User currently offlineStarglider From Netherlands, joined Sep 2006, 678 posts, RR: 44
Reply 4, posted (5 years 11 months 2 days 16 hours ago) and read 8019 times:

Quoting Tdscanuck (Reply 3):
Of course it's remote...otherwise it would have happened before. The ice formation isn't time critical...the triggering event was throttling up, which happened simultaneously on both sides.

That remains a hypothesis that has yet to be proven. Many tests described in the AAIB report end with "further tests are required." Most of the water in the main tanks (max. 1.7 litres per tank) was entrained in the fuel so soon after refuelling. When frozen as entrained ice crystals a large portion must have been consumed before it could be a factor in ice accretion. This means that a considerable less amount of ice could have accreted. The report for instance only found 0.25 ltr and 0.1 ltr of free water in the left and right main tank respectively (probably checked hours if not days after the accident). This quantity was considered to be relatively low for a B777. If the FOHEs were clogged by ice I would assume the water contamination in the FOHEs or downstream fuel lines would have been significantly higher. No mention of it in the AAIB report other than a few tiny water droplets in the fuel filters and housings which might have been free water which naturally settles in these areas.

All BA B777s were taking part in a sumping program which revealed that none of the 43 aircraft showed evidence of significant quantities of free water. G-YMMM was actually sumped under ideal conditions during the last 2 stop-overs at Heathrow only days apart.

The report refers to recorded data not revealing any anomalies in the aircraft systems. Question remains what is and what is not recorded and if data is recorded, could a parameter have been been corrupted by some "byzantine fault" scenario? Furthermore, I have trouble with the AAIB statement regarding the spar valves. Especially that if a spar valve were to fail, an enunciated warning would have been presented on the flight deck. When reading the maintenance documentation on this subsystem I come to the conclusion that there could be a scenario where this may not be the case. But for that to happen several holes in the Swiss cheese would have had to line up. From what I have read in the report the spar valve control system has been thoroughly examined and tested. Question remains, were these bench tests or tests of the actual spar valve system in the accident aircraft?

The report mentions and i quote: "Extensive testing [of the spar valve control system] to induce an uncommanded movement, that remained unrecorded, could not identify any such failure modes." My perception after reading this is that there are failure modes which can close the spar valves by uncommanded movement. But this would have been recorded on the FDR, provided the recorded parameters were not corrupted.

If due to some electrical or electronic spike the spar valve control system and FDR recordings were corrupted, that would open up other scenarios for the cause to this accident.

I agree, such a scenario is remote but no more remote than two donks rolling back due to ice with the minute quantities of ice found in the systems of this particular aircraft.



Just my   
Starglider

[Edited 2008-11-02 05:58:29]

User currently offlineOldAeroGuy From United States of America, joined Dec 2004, 3527 posts, RR: 67
Reply 5, posted (5 years 11 months 2 days 14 hours ago) and read 7969 times:



Quoting Starglider (Reply 4):
Furthermore, I have trouble with the AAIB statement regarding the spar valves. Especially that if a spar valve were to fail, an enunciated warning would have been presented on the flight deck. When reading the maintenance documentation on this subsystem I come to the conclusion that there could be a scenario where this may not be the case.

I think that the simultaneous build up of ice crystals in both FOHE's is more probable than a simultaneous failure (ie closure) of both spar valves. This is especially true since it has not been possible to detect or reproduce the postulated spar valve closures.

After all, the fuel for both engines was an identical mixture. Water entrained in the wing fuel would have been crystalized for nearly the entire flight as wing tank temperatures were less than 0 deg C. Since fuel flows for both engines were the same through out the flight, it's quite probable that a fuel flow restriction due to ice on one engine would be mirrored by an identical restriction on the other engine.

The application of Occam's razor in this situation would seem to be relevant.



Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
User currently offlineStarglider From Netherlands, joined Sep 2006, 678 posts, RR: 44
Reply 6, posted (5 years 11 months 2 days 8 hours ago) and read 7903 times:



Quoting OldAeroGuy (Reply 5):
The application of Occam's razor in this situation would seem to be relevant.

True.

Bust just to put everything in perspective, both spar valve control relais are located adjacent to one another on the same panel, millimeters apart.


Starglider.


User currently offlineOldAeroGuy From United States of America, joined Dec 2004, 3527 posts, RR: 67
Reply 7, posted (5 years 11 months 2 days 6 hours ago) and read 7868 times:



Quoting Starglider (Reply 6):
both spar valve control relais are located adjacent to one another on the same panel, millimeters apart.

And the failure mode was?

I doubt that the AAIB has jumped to an unsupportable conclusion.



Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
User currently offlineTdscanuck From Canada, joined Jan 2006, 12709 posts, RR: 80
Reply 8, posted (5 years 11 months 2 days 6 hours ago) and read 7857 times:



Quoting Starglider (Reply 6):

Bust just to put everything in perspective, both spar valve control relais are located adjacent to one another on the same panel, millimeters apart.

Yes, but the power and sense lines are on totally different sides of the aircraft and are on separate circuits. In order for this to be a cause, our unidentified fault would have had to effect both relays in the same way (i.e. hit both circuits in the same way), and both position sensors would have had to fault the same way (they're on opposite sides of the wing), and both data channels to the FDR (which is a digital system independant from the actuation system) would have had to be corrupted in the same way, and the EICAS would have had to miss both actuators doing it.

Tom.


User currently offlineStarglider From Netherlands, joined Sep 2006, 678 posts, RR: 44
Reply 9, posted (5 years 11 months 1 day 23 hours ago) and read 7802 times:



Quoting Tdscanuck (Reply 8):
our unidentified fault would have had to effect both relays in the same way (i.e. hit both circuits in the same way),

But (hypothetically) if an unidentified fault were to influence the polarity of both relais (open to close), the spar valve actuators would follow this command and there would be no warning on the flight deck because the logic is such that an alert would only be generated if the spar valve actuator would not follow commanded position, e.g. in case of a jammed actuator.

Regarding fault message on EICAS (or according to above scenario the absence of one) see AMM/SDS 28-22-00 "Fault Indications"


Starglider


User currently offlineTdscanuck From Canada, joined Jan 2006, 12709 posts, RR: 80
Reply 10, posted (5 years 11 months 1 day 3 hours ago) and read 7571 times:



Quoting Starglider (Reply 9):
Regarding fault message on EICAS (or according to above scenario the absence of one) see AMM/SDS 28-22-00 "Fault Indications"

Good point...seems like an odd design to me. How would you know if your relay had frozen?

Tom.


User currently offlineStarglider From Netherlands, joined Sep 2006, 678 posts, RR: 44
Reply 11, posted (5 years 11 months 10 hours ago) and read 7451 times:



Quoting Tdscanuck (Reply 10):
How would you know if your relay had frozen?

On the B777 in this, and I stress very remote condition, you won't know until you are confronted with the consequences. Only if the crew would have pre-selected the fuel synoptic or the fuel management maintenance page the spar valve position is visible to them (provided their attention is actually drawn to their position).

There are no spar valve in-transit lights.

Although the fuel synoptic is occasionally selected during a flight, chances are that it is not for most of the flight phases and usually not during approach unless there is a specific reason to do so. The maintenance pages are usually the domain of the maintenance department, again only selected by the flight crew if there is a reason to do so.


Starglider


User currently offlinePlaneWasted From Sweden, joined Jan 2008, 527 posts, RR: 0
Reply 12, posted (5 years 10 months 4 weeks 1 day 5 hours ago) and read 7240 times:

Why use relays in such a modern aircraft as the 777? A semiconductor based solution is cheaper, lighter, smaller, more reliable and generates less heat.

User currently offlineTdscanuck From Canada, joined Jan 2006, 12709 posts, RR: 80
Reply 13, posted (5 years 10 months 4 weeks 1 day 4 hours ago) and read 7212 times:



Quoting PlaneWasted (Reply 12):
Why use relays in such a modern aircraft as the 777? A semiconductor based solution is cheaper, lighter, smaller, more reliable and generates less heat.

The 777 went into service in the mid-90's, which means it's technology got defined in the late 80's/early 90's. Solid-state was around then, but much less established than it is now. I also have no idea of the certification status of solid-state relays at that time.

Tom.


User currently offlineStarglider From Netherlands, joined Sep 2006, 678 posts, RR: 44
Reply 14, posted (5 years 10 months 4 weeks 7 hours ago) and read 7084 times:



Quoting OldAeroGuy (Reply 7):
And the failure mode was?



Quoting Tdscanuck (Reply 8):
position sensors would have had to fault the same way (they're on opposite sides of the wing), and both data channels to the FDR (which is a digital system independant from the actuation system) would have had to be corrupted in the same way

In theory:
It crossed my mind that on the B777 spar valve control circuitry, when selecting the fuel control switches from cutoff to run, each fuel control switch positions several multi-functional relays in the respective (left / right) power management panel (PMP) from cutoff to run. One of these relays in each PMP provides an AIMS position signal (run or cutoff) by which the aircraft information management system (AIMS) is provided with system status and likely provides the FDR with parameter signals regarding fuel control switch (thus indirectly spar valve) position.

From one of the relays in each of the PMPs a run signal goes to the respective left / right spar valve control relays which are not located in the PMPs. The PMPs are part of the electrical load management system (ELMS). The PMPs are located on the left and right side of the aircraft. Spar valve control relays are placed at a separate (central) location. With the spar valve control relays in run (open) position, the signal commands the spar valve actuators to go from closed to open position. When the actuators reach commanded position, limit switches cut power to the actuators.

My perception is that with the fuel control switches on the flight deck and the relays in the PMPs in the run position, if (hypothetically) the spar valve control relays further downstream in the circuit were to change polarity due to some unidentified fault/influence, this could perhaps not affect the relays upstream in the PMPs. If that would be the case, the position signal to AIMS would still be a run signal with the spar valve control relays in cutoff (closed) position.

Suppose our unidentified fault was of temporary nature. The spar valves take approximately 15 seconds to transit from open to closed position and another 15 seconds back to open position. Hence, if the fault lasted anything short of 30 seconds, the spar valves would have only partially closed but would have restricted fuel flow considerably before opening again as the spar valve control relays returned to normal (run/open) condition . . . . . .

Just a theory of course.


Starglider


User currently offlineJetlagged From United Kingdom, joined Jan 2005, 2556 posts, RR: 24
Reply 15, posted (5 years 10 months 3 weeks 6 days 14 hours ago) and read 7050 times:



Quoting Starglider (Reply 14):
Suppose our unidentified fault was of temporary nature. The spar valves take approximately 15 seconds to transit from open to closed position and another 15 seconds back to open position. Hence, if the fault lasted anything short of 30 seconds, the spar valves would have only partially closed but would have restricted fuel flow considerably before opening again as the spar valve control relays returned to normal (run/open) condition . . . . . .

You are assuming that as soon as the spar valve moves towards closed a significant restriction is created. If not, the window for the fault to have such an effect on fuel flow is much shorter than 30 seconds. The failure scenario requires a significant and fairly constant restriction over a period of time. A varying restriction would effectively throttle the fuel flow and so vary engine power. Power didn't vary, it simply didn't respond to acceleration demand. The power levers remained at maximum throughout.

This speculation also ignores what phenomenon would move the relays to CLOSE then return them to OPEN with just the right timing. I suspect the AAIB have been through all the possible valve failure scenarios.

The AAIB ice accretion scenario is the most likely that has been identified, but as yet has not been reproduced in testing AFAIK.

Quoting PlaneWasted (Reply 12):
Why use relays in such a modern aircraft as the 777? A semiconductor based solution is cheaper, lighter, smaller, more reliable and generates less heat.

The reason for using relays is often to control a high powered motor from low voltage logic. Can solid state relays switch 115V AC?



The glass isn't half empty, or half full, it's twice as big as it needs to be.
User currently offlineTdscanuck From Canada, joined Jan 2006, 12709 posts, RR: 80
Reply 16, posted (5 years 10 months 3 weeks 6 days 8 hours ago) and read 6923 times:



Quoting Jetlagged (Reply 15):
A varying restriction would effectively throttle the fuel flow and so vary engine power.

Not necessarily...the spar valves feed low-pressure fuel from the boost pumps to the engine pump. You only get a throttling effect at the engine if the spar valve restricts the flow below what the fuel metering valve on the engine wants to deliver. Most of the time, the spar valve isn't the major restriction. The spar valve is sized for considerably more flow than it ever sees in cruise/descent.

Quoting Jetlagged (Reply 15):

The reason for using relays is often to control a high powered motor from low voltage logic. Can solid state relays switch 115V AC?

You can get them to do 440VAC now, but I'm not sure if the weight is adequate for aircraft use.

Tom.


User currently offlineLitz From United States of America, joined Dec 2003, 1765 posts, RR: 0
Reply 17, posted (5 years 10 months 3 weeks 6 days 7 hours ago) and read 6911 times:
Support Airliners.net - become a First Class Member!



Quoting Jetlagged (Reply 15):

The reason for using relays is often to control a high powered motor from low voltage logic. Can solid state relays switch 115V AC?

115VAC is no problem ... I have several relays at home (for an interface project between a pinball machine and a traffic light) that switch 115VAC at 15A apiece. The huge extra capacity means no need for cooling the units when switching the relatively small wattage bulb in the traffic light. They switch on anything from 3 to 20 vdc input.

They are rated, actually, to 230VAC, but at half the current from 115VAC. This makes them compatible in Europe too.

Considering these are off-the-shelf items, I'm sure in the aerospace world if they needed a 400+ volt solid state relay, it would be easily acquirable as a custom made item, if not already available.

The only considering you have to take into effect is a normal relay doesn't create much heat; you just energize the coil, and the contacts snap!, and the current flows.

Solid state relays on the other hand generate heat proportionate to the voltage and current you intend to pass through them. If you don't dissipate the heat, the unit will melt down and fail.

- litz


User currently offlineJetlagged From United Kingdom, joined Jan 2005, 2556 posts, RR: 24
Reply 18, posted (5 years 10 months 3 weeks 6 days ago) and read 6846 times:



Quoting Tdscanuck (Reply 16):
Not necessarily...the spar valves feed low-pressure fuel from the boost pumps to the engine pump. You only get a throttling effect at the engine if the spar valve restricts the flow below what the fuel metering valve on the engine wants to deliver. Most of the time, the spar valve isn't the major restriction. The spar valve is sized for considerably more flow than it ever sees in cruise/descent.

I realise that, but if the spar valves were closed enough to prevent acceleration then already they are throttling the fuel flow. If they continue to close further before re-opening then the restriction will increase, further reducing fuel flow and so RPM. But that didn't happen in this case as far as is known.

Even with these slow moving valves, the restriction will only be effective when the valves are nearly closed, therefore for a very much shorter time than the 30 second valve close/open cycle.



The glass isn't half empty, or half full, it's twice as big as it needs to be.
User currently offlineStarglider From Netherlands, joined Sep 2006, 678 posts, RR: 44
Reply 19, posted (5 years 10 months 3 weeks 5 days 20 hours ago) and read 6820 times:



Quoting Jetlagged (Reply 15):
The failure scenario requires a significant and fairly constant restriction over a period of time. A varying restriction would effectively throttle the fuel flow and so vary engine power. Power didn't vary, it simply didn't respond to acceleration demand. The power levers remained at maximum throughout.

This speculation also ignores what phenomenon would move the relays to CLOSE then return them to OPEN with just the right timing.

I emphesize it is a theory, in this scenario it is indeed a question of timing but if the valves were to close while the engines increased their demand for fuel, engine behaviour could very well appear to be not responding to acceleration demand. I put it in the same unlikely category as two engines rolling back within seven seconds due to the current hypothesis of ice accretion.

Furthermore, what happens to the fluid dynamics once the fuel has passed a more than 75 to 80% closed spar valve with an engine accelerating but then deprived of the required fuel flow? The fuel flow (what is left of it) most likely would become turbulent in the near empty fuel lines which might have upset the balance, making the fuel flow unstable in the engine feed system. The mass of an accelerating engine, suddenly deprived of demanded fuel, would not decelerate immediately, resulting in cavitation.

Once in this scenario, the instability reaches the FMVs and burners it might have taken too much time for the engines to recover and accelerate. Before they could, they hit the ground.

Another issue is FADEC. As long as a FADEC equipped engine gets fuel, it will tenaciously cling to life. This article in the link below might be interesting regarding that aspect. It deals with another type of turbofan engine but the principle is similar:

http://www.codeonemagazine.com/archi...rticles/jan_90/flameout/index.html

With regards to the phenomenon, without a trace of evidence and being of a temporary nature, that would be like trying to find a needle in a haystack. It could have been aircraft system related such as poor bonding or moisture in the electrical circuits. An electronic spike eminating from one of the other aircraft systems during the approach phase. Or it could have been a source alien to the aircraft and don't even rule out a combination of several distubances at the same time, lining up the holes in the Swiss cheese where, if only one of the factors did not contribute, nothing would have happened to cause this accident and there would have been nothing to report in the news

Quoting Jetlagged (Reply 15):
The reason for using relays is often to control a high powered motor from low voltage logic. Can solid state relays switch 115V AC?

The spar valve control system a 28V DC system.


Starglider


User currently offlineTdscanuck From Canada, joined Jan 2006, 12709 posts, RR: 80
Reply 20, posted (5 years 10 months 3 weeks 5 days 13 hours ago) and read 6762 times:



Quoting Starglider (Reply 19):
The fuel flow (what is left of it) most likely would become turbulent in the near empty fuel lines which might have upset the balance, making the fuel flow unstable in the engine feed system. The mass of an accelerating engine, suddenly deprived of demanded fuel, would not decelerate immediately, resulting in cavitation.

I don't think this could actually happen...you'd certainly get turbulence between the spar valve and the engine fuel pump, but the pressure rise across the engine fuel pump is a few thousand psi...that's going to squash any cavitation bubbles and instabilities long before they get to the fuel nozzles.

It would get you cavitation at the engine fuel pump inlet, which was observed on the BA airplane as I recall, but that just tells you there was a restriction...it's consistent with either a spar valve or fuel exchanger restriction.

Quoting Starglider (Reply 19):
I put it in the same unlikely category as two engines rolling back within seven seconds due to the current hypothesis of ice accretion.

I wouldn't put them at the same level of unlikeliness...we know there was water in the fuel and we know that flight was unusually cold...icing is a reasonable result and explains all subsequent events.

In the case of the hypothetical spar valve event we've got no idea what condition would cause it.

Tom.


User currently offlinePlaneWasted From Sweden, joined Jan 2008, 527 posts, RR: 0
Reply 21, posted (5 years 10 months 3 weeks 5 days 11 hours ago) and read 6738 times:



Quoting Jetlagged (Reply 15):
The reason for using relays is often to control a high powered motor from low voltage logic. Can solid state relays switch 115V AC?

Yepp. But maybe the 777 was to early for solid state.
I read that the A350 will have only solid state circuit breakers.


User currently offlineJetlagged From United Kingdom, joined Jan 2005, 2556 posts, RR: 24
Reply 22, posted (5 years 10 months 3 weeks 5 days 9 hours ago) and read 6715 times:



Quoting Starglider (Reply 19):
I emphesize it is a theory, in this scenario it is indeed a question of timing but if the valves were to close while the engines increased their demand for fuel, engine behaviour could very well appear to be not responding to acceleration demand. I put it in the same unlikely category as two engines rolling back within seven seconds due to the current hypothesis of ice accretion.

Your theoretical scenario relies on a transient restriction being able to reproduce the cavitation. The ice accretion theory allows for a constant restriction to take place. According to the data time history in the AAIB interim report EPR and EPR command began to diverge about one minute before the crash. The maximum time for a complete cycling of the spar valves is 30 seconds, full open thru closed back to full open. If a 75% closure of the valves is required to reproduce the effect, that only provides a maximum event duration of 7.5 seconds, assuming the valves fully closed before they open beyond 75% again.

http://www.aaib.dft.gov.uk/cms_resources/G-YMMM%20Interim%20Report.pdf

I regard the AAIB theory (ice accretion triggered by fuel flow demand) as being much more likely than the unexplained simultaneous failure of two relays coinciding with increased thrust demand. With respect, the AAIB have studied the evidence, and the entire fuel system, first hand.

Quoting Starglider (Reply 19):
Furthermore, what happens to the fluid dynamics once the fuel has passed a more than 75 to 80% closed spar valve with an engine accelerating but then deprived of the required fuel flow? The fuel flow (what is left of it) most likely would become turbulent in the near empty fuel lines which might have upset the balance, making the fuel flow unstable in the engine feed system. The mass of an accelerating engine, suddenly deprived of demanded fuel, would not decelerate immediately, resulting in cavitation.

However the fuel line is blocked, the resulting cavitation is the same so I don't see how this supports your theory.

Quoting Starglider (Reply 19):
Another issue is FADEC. As long as a FADEC equipped engine gets fuel, it will tenaciously cling to life. This article in the link below might be interesting regarding that aspect. It deals with another type of turbofan engine but the principle is similar:

Even FADEC can't do much about insufficient fuel pressure.



The glass isn't half empty, or half full, it's twice as big as it needs to be.
User currently offlineStarglider From Netherlands, joined Sep 2006, 678 posts, RR: 44
Reply 23, posted (5 years 10 months 3 weeks 5 days 7 hours ago) and read 6698 times:



Quoting Tdscanuck (Reply 20):
we know there was water in the fuel and we know that flight was unusually cold...icing is a reasonable result and explains all subsequent events.

Recorded minimum fuel temperature was -34C (-49F). Even if minimum TAT was -45C and minimum SAT was -73C, the fuel (at -34C) did not reach low enough temperatures to cause the fuel to wax.

Fuel temperatures on the accident flight were low but not unique with other flights experiencing lower temperatures which was revealed as a result of data mining 13,000 flights performed by B777 with RR engines.

Quantities of water in the fuel (40ppm dissolved, 30ppm entrained or free) estimated to have been present on this flight are considered a very low level for a B777 according to the AAIB.

Data mining revealed that fuel flows during approach on the accident flight reached more than 12,000 lbs per hour, although not unique, this was at the edge of the family for data analysed.

The theory according to the AAIB is that perhaps the temperatures (low but not the lowest) and high fuel flow during approach (at the high end of the family) could have released accreted ice at stagnation points somewhere in the engine feed fuel lines and blocked the LP pump inlets or FOHEs.

My remark on this theory is: 12,000 lbs/hr fuel flow may be at the high end during approach but many similar flights under similar conditions (perhaps with relatively higher water content in the fuel compared to the low water content of the accident flight) must have made go-arounds which require higher fuel flows but apparently do not release accreted ice and do not encounter this problem?

Those go-arounds, also occurring on a regular basis but without similar effects have not been covered in the AAIB reports sofar. Without any evidence or clear explanation to this accident in 13 years of B777 operation, this is why I place this "ice" scenario in the same category and as unlikely as the possibility of the uncommanded closing of the spar valves.

Quoting Tdscanuck (Reply 20):
It would get you cavitation at the engine fuel pump inlet, which was observed on the BA airplane as I recall, but that just tells you there was a restriction...it's consistent with either a spar valve or fuel exchanger restriction.

Cavitation was found on the HP pump pressure outlet ports on both engines caused by restricted fuel flow, leading to low inlet pressure at the HP pump.



Regards,
Starglider


User currently offlineTdscanuck From Canada, joined Jan 2006, 12709 posts, RR: 80
Reply 24, posted (5 years 10 months 3 weeks 5 days 6 hours ago) and read 6678 times:



Quoting Starglider (Reply 23):
Quoting Tdscanuck (Reply 20):
we know there was water in the fuel and we know that flight was unusually cold...icing is a reasonable result and explains all subsequent events.

Recorded minimum fuel temperature was -34C (-49F). Even if minimum TAT was -45C and minimum SAT was -73C, the fuel (at -34C) did not reach low enough temperatures to cause the fuel to wax.

I didn't say the fuel was unusually cold, I said the flight was unusually cold. A key thing to keep in mind is that the fuel temperature probe is only in one location and there is very limited fuel circulation on a 777 (as compared to some other designs), so the fact that the fuel probe only measured -34C doesn't tell you much about the temperature anywhere else, most especially at the heat exchanger.

Quoting Starglider (Reply 23):
Quoting Tdscanuck (Reply 20):
It would get you cavitation at the engine fuel pump inlet, which was observed on the BA airplane as I recall, but that just tells you there was a restriction...it's consistent with either a spar valve or fuel exchanger restriction.

Cavitation was found on the HP pump pressure outlet ports on both engines caused by restricted fuel flow, leading to low inlet pressure at the HP pump.

You can't get cavitation on the outlet of the HP pump...that would require something downstream sucking fuel and there's no such entity. Cavitation requires you drop the pressure below the fuel vapour pressure...the actual pressure on the outlet side of the HP pump is in the thousands of psi. The HP pump is pushing fuel against the combustor pressure and the outlet port is the highest pressure point in the entire fuel system. Low inlet pressure will cause cavitation on the impeller and inside the pump, not on the outlet.

Tom.


25 Post contains links Starglider : See page 18 of the AAIB Interim report, top of first column: http://www.aaib.gov.uk/cms_resources/G-YMMM%20Interim%20Report.pdf Apparently the AAIB g
26 Jetlagged : I doubt they got it wrong, they were able to replicate the cavitation marks during testing. According to the AAIB report the restriction would have t
27 Tdscanuck : I have to assume not...they're a lot better at this than I am. However, I'm totally baffled by the physics that could make this happen. Prior to my a
28 Starglider : On page 11 of the AAIB interim report it is explained how the OEM previously undertook tests with racks of thermocouples and determined that the cold
29 Jetlagged : In the test you refer to the engine flamed out after a series of auto-restarts. On the BA 777 the engine rolled back to a steady condition above flig
30 Nomadd22 : I've seen cavitation damage on the output sides of high pressure hydraulic pumps where nobody thought cavitation would be possible. Cavitation bubbles
31 Starglider : I don't intend to claim that the test in the article is identical to a posible failure in the BA777. It does show that unexplained incidents/accident
32 Jetlagged : It is conceivable, but it requires repeated unexplained intermittent faults on both spar valve control relays. It seems even more remote a possibility
33 Osiris30 : Maybe I'm not getting it here.. I've read the thread and haven't seen anyone explain in a credible fashion why ice isn't likely to be the source of th
34 Starglider : That was my point as well. Sequence was mentioned, not to blame the crew, just to explain a pre and post SB configuration and why (in a pre SB config
35 Osiris30 : So because you have never won the lottery means it is impossible (assuming you play of course)? Not exactly the most stellar line of logic I've heard
36 Starglider : You are taking my words out of context here. I said I differ in opinion on one issue regarding flight deck warnings in relation to spar valve positio
37 Tdscanuck : Part of the issue is that *none* of the causes of this accident are likely...any that were would have been seen long ago. Since all the options are u
38 Osiris30 : Apologies, you're right. I did mix two posts and that was wrong of me, so I officially ( ) withdraw criticism c above. My issue has never been with y
39 Jetlagged : I don't think that having posts removed about this accident on another forum is any indication of anything sinister about it so why does that cause y
40 Starglider : Apologies accepted. It was a plain and simple request for information about the cabin layout (3 or 4 class config.). Other topics removed which I pos
41 Post contains links Starglider : Another engine rollback is under investigation: http://www.flightglobal.com/articles...w-like-trent-800-engine-issue.html Starglider
Top Of Page
Forum Index

Reply To This Topic Question On The Accident Of BA B777, G-YMMM
Username:
No username? Sign up now!
Password: 


Forgot Password? Be reminded.
Remember me on this computer (uses cookies)
  • Tech/Ops related posts only!
  • Not Tech/Ops related? Use the other forums
  • No adverts of any kind. This includes web pages.
  • No hostile language or criticizing of others.
  • Do not post copyright protected material.
  • Use relevant and describing topics.
  • Check if your post already been discussed.
  • Check your spelling!
  • DETAILED RULES
Add Images Add SmiliesPosting Help

Please check your spelling (press "Check Spelling" above)


Similar topics:More similar topics...
Why No Extra "cockpit" Windows On The Top Of A-380 posted Tue May 27 2008 04:10:37 by L1011CPH
Strange Control Surface On The Wing Of The 767 posted Sun Mar 30 2008 11:49:59 by Pyrex
Quick Question On The AT-CTI Program (ATC) posted Tue Dec 4 2007 15:55:06 by Boston92
"Ridges" On The Intakes Of Early JT9D's posted Sun Apr 8 2007 00:26:00 by N231YE
What Exactly Is This On The Front Of The A/C? posted Sat Jan 20 2007 12:06:16 by RootsAir
Question On The ILS System posted Fri May 19 2006 03:16:48 by LTU932
Holes On The Cowlings Of The 747? posted Sun Apr 16 2006 02:58:58 by CalAir
Please Help: A Question On The Directional Gyro posted Tue Apr 4 2006 04:09:07 by FLY2HMO
What's This On The Wingtip Of An HP 752? posted Tue Jan 3 2006 13:09:04 by TupolevTu154
Question On The MD80 Series Autopilot posted Tue Nov 8 2005 15:51:24 by Oli
The Panel On The Side Of This 767-What Is It? posted Sat Jun 5 2010 08:54:45 by c5load
What Is The Vent On The Side Of This DC-10? posted Tue Jun 1 2010 19:18:58 by Gulfstream650
What Is The Cone On The Back Of A Jet Engine? posted Sun May 9 2010 08:55:23 by c5load
What Are These Openings On The Belly Of The 757 posted Wed Mar 10 2010 09:27:35 by c5load
Intake On The Side Of DC-8s Noses? posted Mon Mar 1 2010 12:31:20 by 747400sp
"Growths" On The Bottom Of Wings posted Sun Sep 7 2008 17:46:33 by Boeing777228
Why No Extra "cockpit" Windows On The Top Of A-380 posted Tue May 27 2008 04:10:37 by L1011CPH
Strange Control Surface On The Wing Of The 767 posted Sun Mar 30 2008 11:49:59 by Pyrex
Quick Question On The AT-CTI Program (ATC) posted Tue Dec 4 2007 15:55:06 by Boston92

Sponsor Message:
Printer friendly format