Forgive me for not checking all sixty odd pages of this thread for the answers to the following.

Is there a temperature sensor anywhere near the fuel pumps - particularly the H.P. ?

Is that data recorded anywhere ?

What was the temperature of the fuel either side of the H.P. pumps at the time of engine failure ?

Goose
 

There is one Fuel Probe (Temp. Sensor) Data stored in FDR. It is located between Ribs nine and ten in the Port Wing, about one meter aft of the Boost Pump Pickup as I recall.

Airfoil

(Lowest Recorded Temp. Fuel was -38C.) And the Port wing was on the Sun Side of the A/C for that flight.

It is interesting that the FAA are now looking at the suction feed.Here is my post from way back, which may be an explanation:

There is actually one instance in which Boeing suggest that a fuel supply problem can cause engine thrust deterioration or flame out. This is when an engine has to revert to suction feed at altitude, following dual wing tank pump failure. (No mention is made of the low altitude situation.) The background to this advisory is that dissolved air in the fuel is released at altitude and can accumulate in the suction feed pipe. The implication being that low atmospheric pressure in the tank, plus air in the suction line will disrupt or cut off the fuel flow under suction conditions.

Note from the fuel synoptic posted by Jet11 that the suction pipe is connected directly to the engine fuel supply manifold, isolated only by a pressure bypass valve. Also that the tank pump nominal output pressure is 14psi, but can be much lower. For argument's sake, I will assume that the suction bypass valve opens when there is a pressure differential around 5 psi and it closed under normal conditions.

In Bejjing, BA038 takes on fuel with an unusually high dissolved air content. This is due to a combination of production/storage/transportation/pumping and weather variables. The a/c flies high, long and cold. The air is released from the fuel and the suction pipe is completely filled with (relatively) dense air. This air remains in the pipe during the descent and once below 6000ft is inceasingly pressurised by denser ambient air and rising temperatures. At 1000ft atmospheric pressure reaches 14psi, equalling the fuel manifold pressure. Fuel hydrostatic static pressure and adiabetic heating add another 6psi, causing the suction bypass valve to open against tank pump pressure.This releases pressurised air into the fuel supply manifold and disrupts the fuel supply. This happens in both tanks, but not simultaneously. The opening of the valve could be assisted by a momentary drop in fuel manifold pressure during the initial power demand.

Just another theory. My assumptions about the bypass valve are probably wrong. Also, why is this not a regular occurence? Maybe the fuel and meteorological conditions were unique to BA38. Maybe we have always been operating close to this scenario.........

Don't put to much on this AD, as this has been known about for over two years or more.
You will find that this task has been in operation for a few years and BA may have been doing it on its B777 fleet.
I was at a Boeing engineering committee a few years back on the B757 series and we were told that a new task was being introduced to carry out a fuel suction test, which we voted to accept.
I remember way back this being done on the old B737-200s and we were told at our meeting it had come about from findings on the B777 fleet. Due to the fuel system on the B757 being similar to that of the B777, it needed to be brought in. We discussed the interval that was proposed and made amendments to the interval and was as l said accepted.
We were told if we accepted the task, it would mean that it should not have to be made an AD. This task was introduced into our "C" checks two winters back.
So if the same thing has happened to the B777, l would guess it has been carried out for a couple of years.
The main reason for the task, is when you get low fuel quantity you may suck air in, if the fuel pipe seals are leaking. You need a very low fuel figure in each tank to carry out the check using engine runs or you have go into the tanks and use test equipment instead to check the piping.

Just a thought, FLY
 

But are you assuming the T7 has a separate supply, isolated from the pumps/line that "takes over" when Pumps inop? The mass of Fuel in the pipeline has great inertia at low/no velocity and with HP functional at 14psi (or slightly greater) could vaporise regardless of Temp. One atmosphere is ~ 14psi, right? In a "tail low" approach, the suction problem could be exacerbated by the inertia of the Fuel and its lethargic velocity in the Line due to Green descent. Haven't we been there? (Besides, the pumps were working, AAIB)


Airfoil

I am suprised no one has picked up on this.

http://www.ainonline.com/news/single-news-page/article/fuel-system-remains-focus-of-investigation-into-ba-777-crash/

Fuel system remains focus of investigation into BA 777 crash

By Ian Goold
June 1, 2008
Accidents


Safety officials probing the circumstances leading to the January 17 accident of a British Airways (BA) Boeing 777 at London Heathrow are continuing to focus on the fuel system. In particular, they want to know why the aircraft lost power when it was on final approach.

The 777, operating as flight BA038, was flying nonstop from Beijing to London Heathrow. The flight was uneventful and the engine operation was normal until the final approach. As the 777 approached to land, the R-R Trent 895 engines initially responded to an autothrottle command for increased thrust. At a height of about 720 feet, thrust on the right engine reduced to about 1.03 times the engine pressure ratio (EPR), followed about seven seconds later by a similar reduction (1.02 EPR) on the left engine.

The airplane touched down 1,000 feet short of Runway 27L’s paved surface before coming to rest in the undershoot area astride a taxiway junction near the threshold. In the heavy landing, the nosegear collapsed and the left main undercarriage was punched up through the upper surface of the wing, while the right undercarriage six-wheel bogie detached from the aircraft. One passenger was seriously injured and 12 other occupants received minor injuries. The airplane sustained heavy damage and was subsequently written off.

The UK Air Accidents Investigation Branch (AAIB) has determined that in the case of both engines, thrust reduction resulted from reduced fuel flow, which was reflected in all subsequent engine parameters. Although the fuel-metering valve responded to an engine-control system command to open fully, there was “no appreciable change” in fuel flow. Evidence indicated that both engines had “low fuel pressure at the inlet to the high-pressure pump.”

Following “sustained interest” in the accident, the first such event involving a 777, the AAIB last month published a special bulletin that said Boeing Commercial Airplanes division was conducting fuel-system testing in Seattle. Engine manufacturer Rolls-Royce had earlier completed “extensive full-scale engine testing” in a cell modified to replicate the actual response of the accident aircraft’s engine fuel and control system.

AAIB investigators say the primary challenge was for Boeing to create the environmental conditions the flight experienced while flying over Siberia at altitudes of up to 40,000 feet. The flight over the region took the 777 through particularly cold air, as low as -76 degrees Celsius (-104.8 degrees F). The minimum recorded fuel temperature was -34 degrees Celsius (-29.2 degrees F) and subsequent tests of onboard fuel showed a freezing temperature of -57 degrees Celsius (-70.6 degrees F).

Investigators are principally concerned with understanding the potential for
fuel-system restrictions to have formed. Additional work has aimed to improve understanding of fuel-flow dynamics between the tank and the engine.

Systems consultancy Qinetiq is reviewing recorded data from a large number of similar flights. Qinetiq analysts are concentrating on the identification of abnormal combinations of parameters, since no single reading from the accident flight has been identified as outside previous type operating experience.

I guess you couldn't be bothered to read through the 68 pages of this thread?:bored:

F111D - A summary of 68 pages... The evidence for lack of fuel flow appears to be limited to minor cavitation damage to the pumps (Minor in the sense that it did not stop them working). The issue is what caused the low fuel pressure/flow in the first place. It appears there was sufficient fuel on the plane.

Airfoilmode:

No, I am not suggesting that the suction system has its own fuel supply line. There is a short section of pipe that joins the main fuel supply manifold, via a check valve, just downstream of the LP tank pumps. I would not see fuel inertia as a problem there.

What I suggested was that air or gas that was trapped in the suction pick-up pipe might, under certain circumstances, become sufficiently pressurised to enable it to overcome the check valve differential and discharge like an aerosol into the main manifold. When this air/gas reached the engine HP pumps it would cause cavitation as the pumps ingested air and fuel supply to the engines would be reduced, similar to the situation where a diesel engine loses pressure at the injectors if air gets in the system. No doubt the LP pumps would have restored manifold pressure eventually, but in BA038's case there was insufficient time before the accident occurred.

Maybe the check valves have prevented this situation up to now and BA038 was the first unlucky one. Running a suction check from time to time would certainly remove any trapped gases in the system.

777fly-

Air bubbles in the fuel do not cause cavitation damage. The air cushions the collapse of the air bubble, so the bubble does not disappear when pressure is restored, and there is no damaging shock wave.

Bulletin S3 out in AAIB June report.

To save you rushing to download, there is nothing of note in it, and still no mention of a config change, and no rec. of any changes in operating procedures.

Indeed - in fact, it looks to me like it's exactly the same document as was published on 12 May 2008, just formatted differently to be incorporated into the Bulletin.

Rightbase is correct, air in the fuel stream will not cause cavitation damage, for the reasons he stated. This was caused by a fuel obstruction or restriction.

In fact, if you think about it, cavitation damage from suction would seem to confirm integrity of the fuel manifold.

When the freezing point of fuel is measured on the ground is the resulting figure corrected for the increased pressure found in the fuel system?

I think the freezing point of fuel (but not water) is raised by increasing pressure. In which case you might speculate on the possibility of having liquid fuel in the tank... which then freezes in the pipes/pumps when pressurised.

This effect probably isn't great enough though.... -34C to -57C is a bit of a stretch and the architecture of the 777 fuel system may also rule this out.

Edit: Initially when you compress something heat is given out which raises the temperature. At high flow rates the fuel might remain warmed. At low flow rates perhaps this heat has time to disipate leaving the higher freezing point to cause the problem. If that's correct it might explain why this didn't happen in cruise just on approach.

Or it might just be the difference between finding the fault on a complete and serviceable airplane and not finding or unable to duplicate the fault on an incomplete and twisted one?

For instance, if something was in error in the last few minutes of the flight, a sequence of system selections unique to the approach phase only, which generated a rogue electrical spike that did not make it to the recorders somehow? Maybe the temperature in the fuel tanks had nothing to do with the cause at all, perhaps humidity, pressure and temperature inside the hull and miles of wiring and electronic units did?


Green-dot

I think you will find that the airline industry has known about problems of transitioning to suction feed on Boeing airliners since at least 2001, a lot longer than two years ago:
http://www.pprune.org/forums/showpost.php?p=4000971&postcount=1101

I missed this.
Can you provide a link to this statement by the FAA that the AD has nothing to do with BA038?
Thanks. http://home.infionline.net/~pickyperkins/pi.gif

As has been addressed before, the Trent/777 does just fine in Suction only regime. Also, prior post (recent) explains that a non uniform Fuel supply (eg entrained air) would prevent cavitation, not cause it. Air/Water/Additives in the Fuel have been addressed pretty much ad nauseum. I think the chances the AD is related to BA038 are quite slim, except as it may apply to the Fleet in general, though not nil. (Consider 7,500 TBI per AD)

My sense of the cavitation problem would require cold Fuel in a tight system, devoid of easily vaporised contaminants, subjected to a sudden throttle slam shock with HP's straining at a volume of Line Fuel that has substantial inertia, creating cavitation, which in turn would inhibit the free flow of Fuel. A perceived restriction/block is in there as well.

Airfoil

cwatters, hi again,
A minor point, fuel cannot be compressed just pressurised although any air present would be compressed.

"PickyPerkins", your link has nothing l believe, to do with this AD.
As l pointed out this MRB task and now being made into an AD, is all about air being sucked into the fuel tank piping from within the tank, from worn seals where the pipes join each other.
This happens with very low fuel state and would not happen just after take-off.
What happened to the UA aircraft, is l believe not related to this extra task introduced from the MRB and now mandated.

Oldlae,

All liquids are compressible to some extent, even water.

There are some values for Jet A1 quoted here:

http://www.eng-tips.com/viewthread.c...=155581&page=1

I have no idea whether this is material to the BA038 accident, although with the low pressures involved, I suspect not.

Both the plumbing and the fluid therein (fuel, possibly with air or vapor) have elastic properties: they store energy when under stress, and give that energy back when the stress is removed. That's all it takes to have the potential for resonance, which was discussed a few pages back. I trust this is being studied until some better candidate emerges.

Agree

But what is the forcing function and the response :confused:

THAT, of course, is the $64 question (am I showing my age?)

One excitation possibility is the gear tooth passing frequency (assuming a gear-type HP pump).

Since both engines initially responded to the autothrottle with increased thrust, that meant the restriction or obstruction was not in effect yet, at least not initally. Then within 3 secs or so, the right engine rolled back to EPR 1.03, with the restriction or obstruct in effect, with low fuel pressure present and HP pump cavitation damage taking place. The same sequence occurred for the left engine, although the rollback in power to EPR 1.02 occurred 7 seconds later. In other words the restriction or obstruction developed slightly later for this engine.

This suggests to me that the sudden increase in fuel flow required for the power increase, is what actually triggered the formation of the restriction or obstruction (whatever it was), which apparently was not present in the first few seconds of increased fuel flow.

I'm not sure what that means, but this appears to be the sequence of events. High fuel flow was available initially, then the restriction or obstruction developed (perhaps precipitated by the sudden increase in fuel flow), then low fuel pressure and HP pump cavitation followed and remained in effect until impact with the ground.

Of course the million dollar question is, what was the nature of the restriction or obstruction, and how could a sudden increase in fuel flow rates possibly cause it?

Flight Safety,
Without any more AAIB input we're back to guessing games....
But your post makes for an interesting question.

What was the fuel flow in m/sec or ft/sec ?
In other words, if it was indeed some kind of obstruction, how far did the 'slush puppie / chinese noodles / wad of chewing gum' travel after being dislodged, to arrive after 3 secs on one side and 7 secs on the other side?
Maybe there's a clue there?

CJ

Bernoulli
 

Daniel Bernoulli, whose work with Hydraulics may have a hint to help the discussion. His most widely known discovery was that a fluid (inviscid Fluid, not "compressible") that was accelerated in a tube, created a concomitant reduction in pressure within. The Fuel line in 038 was subject to sudden demand. Cavitation at the HP (both) has been described and bears further thought. As an "incompressible" mass, the Fuel in the Line would transmit the focal cavitative results on the High side of the Pumps back through the system. In essence, the massive pressure at the Pump Lobes would communicate back through the Fuel and have an instantaneous effect on other mobile mecnanical devices in Line. Air "cavities" would have developed and perhaps failed backflow valves, Spar valves, and LP (boost) pumps back through the system. The Fuel in the Line is a "Hammer" that takes the shape of the System, but retains its abilities to hamper normal supply throughout the entire Fuel System.

Airfoil

(Suggesting there may have been no "obstruction", other than the Fuel itself, having created an insufficient flow due to inertial reluctance which may have failed other intersections)

airfoilmod,
Phenomenon also known as "water hammer" to your local plumber.
Known to every fuel and hydraulic system engineer.
As an explanation for what happened on BA038 it seems somewhat far-fetched.
Not to mention, that any "fuel hammer" effect is nearly instantaneous. It does not take 3/7 seconds to manifest itself.

Bernoulli was a plumber
 

And the theory has more to do with resonance and harmonics having an effect on mechanicals than a "Hammer". I used the term to convey the communicative effect of pressure differentials in an "instantaneous" way.

"Hammer" is an effective way to understand cavitation, notwithstanding your derisive retort to its use. I didn't think it would bother anyone, I'll be more careful with my vocabulary.

RGDS Airfoil

(BTW, if you've noticed Water Ram, you will also have noticed its brother, Recoil. That vibration, and "return,rebound" of the Ram, stops flow.)

Anyone got a phase diagram for jet fuel?

ubreakemifixem
 

Rather coy way of introducing doubt and suspicion of the Flight Crew, No?
Coupled with your statement on another thread, why don't you just man up and state your point.

cwatters: Awesome, now you're talking.

Airfoil

That may be hard to come by. Problem is that typical fuels are not an "it", but a complex and widely varying collection of complex components blended to meet a set of specs. A batch could consist primarily of ground-up gerbils and would likely not be rejected if it met all the specs.

Similar might be said for various of the aircraft systems involved -- they are designed to meet reasonable specs derived from science, engineering, and experience. Sometimes circumstances arise that have not been fully contemplated in the specs, and unanticipated results ensue. BA038 may well be one of those situations where tried-and-true standards and specifications did not prevent a unique special case from having adverse effect.

So did the UA 767 referred to in my link:
http://www.pprune.org/forums/showpost.php?p=4000971&postcount=1101

The engines on the 767 performed just fine in suction all the way from sea level to 41,000 ft.

What they could NOT handle was a cessation of boost pressure during the climb.
The engines spooled down 14 seconds after the boost pumps were turned off.

Exposed leaky pipes and seals were not involved because the tanks were full.

This seems a clear indication to me that the boost pumps were doing something which the HP pump could deal with so long as boost pressure continued.
But when boost pressure was discontinued, the engines spooled down after a delay of 14 seconds.

That something which boost pumps were doing could have been bringing air out of solution.
Which was OK with boost pressure on, but not OK when the boost ceased.

In the case of BA038, does it really matter whether air bubbles did or did not cause cavitation damage?

The AAIB says although there was cavitation damage, it was not sufficient to prevent the fuel being pumped. The damage was not significant wrt pumping capacity.

What matters is - why did the fuel stop advancing through the system?

More than 30 years ago a Royal Aircraft Establishment repot by WGS Lester, “Temperature and Fluid Effects on Cavitation in Aircraft Fuel Pumps” showed that air coming out of solution could cause the volume output of centrifugal type boost pumps to fall markedly.

He was not concerned about cavitation damage, only about whether the pump continued to deliver fuel.

With air dissolved in the fuel, his test pump didn’t stop pumping, but the output volume fell.
Sound familiar?

And the fall off in performance was greater when the fluid was cold.
Sound familiar?

In fact the pump output was greater when the test liquid was boiling (no dissolved air) than when it was cold (with dissolved air).

Some of his test results are shown in:
http://www.pprune.org/forums/showpos...postcount=1102

Fuel saturated at room temperature becomes highly supersaturated when cooled.
It typically remains supersaturated for many hours in aircraft fuel tanks at altitude.
This was mentioned by Lester, and confirmed decades later in Boeing test flights.

Bottom line
The difficulty in finding the restriction to fuel flow maybe because there was no restriction to flow, just a fall-off in boost pumping capacity due to air coming out of solution from cold fuel, and a mixture of air bubbles and fuel delivered to the HP pump.
Just my 2cents. Maybe 1 cent. http://home.infionline.net/~pickyperkins/pi.gif

PickyPerkins
 

You may be intermingling two separate causes of Fuel restriction or Flow reduction. Your reference to dissolved air or air (Gas) in the Lines reducing pump efficiency sounds a whole lot like "vapor Lock" (or Block). It is a serious problem in pressurised Fuel systems, but one wonders of its relevance here. Also, a "foaming" supply of Fuel would mimic starvation, even if Flow continued. Cavitation, having been demonstrated, has stolen all the attention, and is an unrelated (presumably) failure, unconnected to dissolved gases in Fuel (to my understanding).

Airfoil

Just a point if interest.

Fuel flow could have been restricted for a long period of time.

Think of Spar or LP valve being in a position between open and closed, Engine will run Ok upto a point, but when FMV/FMU needs more flow than avail, problems will start, problems may be delayed a little with the head of fuel down stream of restriction, but problems on the way.

Just out of curiousity -
 

What might the 777 cruise fuel flow have been just before TOD? (call this point A)

And what fuel flow might have been demanded on a normal close final? (call this point B)

Between these two points, fuel flow was at a very low level. Is it possible the restriction existed (without symptoms) at A, and didn't become evident until thrust advance at B? :confused:

barit1
 

Or Point B plus 3 and then plus seven seconds, the time needed to make the restriction critical ? Think you're on to it.

It never ceases to amaze me how "creative" individuals can be.
After all, this is the purpose behind brainstorming!

Only problem is that each "theory" has to be "understood" and either moved forward to the "possible" file or discarded due to some misunderstanding/lack of knowledge by the proposer.

I was particularly "impressed" by your "definitive" conclusion that there was adequate fuel on board after resolving the frost pattern on the underwing from a photo a few seconds before the crash landing.

Now the various "theories" put forward in the last few pages of posts, many seem to focussed around the cause of the "restriction" without any data in support of the nature of any such restriction.

In order to try to get you guys to be creative in a slightly different direction I wonder what are your thoughts are on the "size" of the restriction?

By this I invite your energy into determining the following:
1. What fuel consumption (flight idle?) would exist on approach prior to the incident. Ltrs/sec or whatever.
I don't have any idea what these big fan jets consume even at idle but its certainly "significant"!
2. What capacity (volume) exists in the fuel lines between pickup and engine spray nozzles?
Is this the same on both sides of the aircraft or slightly different and if so which side has the greater volume and why?
3. Is it possible to resolve this volume more accurately. i.e. what %age of the previous volume exists between pickup and lets say the spar valve, heat exchanger or HP pump.
4. How long would it take to use this volume with 100% "block" at whatever thrust setting was likely to have been called for during final to offset flap etc (which wasn't delivered when requested).
I appreciate that full thrust was at some point selected but it was a much lower power request that highlighted the problem..
DO THESE FIGURES MAKE ANY SENSE?
So we could deduce 75%, 50% etc restriction vs duration.
Unless these figures make sense, discussing the possibility of Bernoulli type restriction etc seems a little premature (although possible) especially since such an occurrence has not been experienced during multi million previous landings. So what was different this time around?

As a professional engineer, but unfamiliar with the "detail" of aircraft installations, the 7/8 secs is in my view highly relevant.
The difference between 7 and 8 secs on the other side may NOT be relevant but explained by deficiencies in the measurement (DFDR sampling rate) process or subtle differences in the level of thrust commanded. However it MAY be significant if indeed the capacity of the system is indeed different LH to RH.
However a "significant" amount of fuel would be used during a 7 secs period. So how about a few FACTS even if these are derived from from first principles.
The AAIB have access to the real data so I guess they will have already been through an exercise something like this.

Maybe you will just overlook my request / maybe not!

Finally, am I correct in my understanding that BOTH LH and RH boost pumps WERE WORKING and that this was actually confirmed by the AAIB...I'm too lazy to go and reconfirm! If so why the discussion about what would happen if the pumps weren't running and the possibility of air being drawn in with low tanks etc.

At first sight the 7 or 8 seconds difference could have been explained by a single fuel source going through a fuel crossfeed valve ...
I believe that kind of engine must be around 1000kg/H at ground idle, so flight idle would not be far from 1 Liter every 2 seconds

Thanks!
Half a litre per second is as good a ballpark figure as any.
Now what's the diameter of the fuel feed pipe?