Fluidic resonance in a tuned circuit tends to be low frequency and is far less likely to be attenuated by the aircraft structure.
Interestingly, this noise would be very similar to tire rumble at landing.

Cavitation noise tends to be high frequency and is likely to be attenuated by the aircraft structure.
Below is a spectrum for cavitation noise

http://caltechbook.library.caltech.e...igs/fig614.gif

Typical fluid resonance detectors operate between 2Hz and 300Hz.
Unfortunately, I can not find a spectrum for fluidic resonance that is not limited by copyright.
Those interested in further exploring this topic might Google: Bayesian Modelling of Fluid Flow in Pipelines.

Shawk - HPs part of engine
 

Shawk,
You say that "cavitation noise tends to be high frequency and is likely to be attenuated by the aircraft structure".

The cavitating high pressure fuel pumps (HPFP), that I mentioned might be the source of the reported strange noise, are integrated with, and driven by, the engines.

Lots of local pipework and other metal objects (nacelle etc) in the vicinity available to resonate with the HPFP, plus, a large hole at one end to let sound out.

Would much attenuatation occur and how would the other resonating objects alter the frequency range?

Of course, if your interesting fluidic resonance idea is correct then the sound is important evidence of causation. If it is the result of the cavitation then it is not important.


Regards, Tanimbar

One would have thought that any significant sounds reaching the ears of ground spotters etc. would have been recorded either as acoustic signals on waterfall plots on the CVR or as forced low-frequency responses on the DFDR from various inputs around the aircraft.

On the other hand less specific sounds (white noise) or those left in the memory of observers have a tendancy to only exist in the minds of theorists.

You may have experience with a water faucet that sometimes makes a loud, low rumble or rattle that is accompanied by a small volume of water that comes out of the tap in discrete spurts of water. That's fluidic resonance in a pipe. Generally, you have to stop the flow to stop the effect. With the resonance gone, water flows normally.

I suspect that high frequency cavitation noise similar to the frequency plot would be inaudible through a typical thickness of aircraft aluminum at 10 meters. A lot depends on the placement of the pump and the shape of the surrounding structure but not much high frequency sound is likely to escape an airframe. From what I've seen, aircraft piping is very well attached and damped to prevent vibration that will cause rapid metal fatigue, so it might not act as a good acoustical conduit for high frequency sound.

If the fluidic resonance theory has much basis in reality, the likely noise is somewhere between 16Hz and 32Hz. This sound frequency range has a very good chance of escaping an airframe and be audible on the ground.

But I'm not certain that a CVR will record audio at 16Hz to 32Hz. They seem to be optimized to record human voice and use brick wall digital filters to not record below 100Hz.
The DFDR may not be sensitive enough to record the airframe effects of a small mass of fuel oscillating at 16Hz to 32Hz.
I hope that experts will add their knowledge to what parameters could be recorded.

Thanks NoD

I was chatting to a FADEC "Expert" last week when the subject came up and he was of the opinion that as there are quite a few parameters in the FADEC system that are not logged on the DFDR and that if affected could cause a roll back to flight idle, then EMI still remains a possibility.

Does anybody know for certain the clock speed of the Trent FADEC?

I once knew of a FADEC clocking at 120Mhz, with predictable coupling to VHF comm frequencies. :eek:

The unusual sound heard by pax could have been the flaps retracting, which without engines spooled up would probably be the most prominent and "not normally heard sound' at that point on the approach.:E

And considering the obvious technical prowess of most people here I am surprised that you have difficulty with the OFF mode for your PC/MAC when you don't like what you are reading instead of this persistent fascist approach of telling people what they can and can't write, think or do. Get a life eh?:ugh:

Speculation or conjecture no matter how ridiculous is where discoveries and ideas come from, talk out the most ridiculous first and fine tune the way to the truth.:ok:

To whom?

It's fairly common knowledge that GOOD human hearing is in the range 50 Hz to 20,000 the upper frequency decreasing rapidly with age (and usually lower for the male gender)..

It rolls off rapidly below 50Hz to prevent damage from the energy content in low frequency sound.

Bottom string (E) on a bass guitar is around 42 Hz, and it's 2nd harmonic (84Hz) is usually more understood by the ear than the fundamental, which itself is more felt rather than heard.

32 Hz or less? Something by-standers commented upon, amongst the melee of other sounds from an aircraft so unusally close - nope, don't think so.

Neither do I find it credible that cavitation noise within a pump within an engine nacelle would be heard, or commented upon, whatever its frequency... over the mass of aerodynamic and other turbomachinery noise - IMHO of course!

From both tanks, a nebulous volume of cold fuel - within 8 seconds of each other?

The Fuel system worked fine
 

It sustained damage from Cavitation, but even with that was "able to function normally". The squishiest suspect is of course, Fuel.

1. Ten degrees "off spec.". Not "better than"; OFF SPEC.

2. Fuel contained some amount of water, deemed "less" than "excessive."

3. Water and Jet don't "mix" (non-soluble)

4. Tank Temp., at -34C would freeze water immediately

5. Boeing Captain Carbaugh posits "Ice Crystals", a "Temperature issue"

6. The Heat exchanger has yet to be exonerated from involvement.

7. Fuel and Water could have transited the LP pumps

Possible: Given a seriously cold-soaked engine Pylon and barely warm heat exchanger, ice crystals could accumulate in the exchange matrix of the exchanger, blocking flow to HP pumps.

For this to have happened, the timing sequence of engine thrust loss needs to be seriously considered to satisfy many possibilities.

1. The Fuel/Ice "mix" would have to be inordinately homogeneous between the two main tanks.

2. If the ice crystals had already been formed, further "freeze" may not have been possible anyway.

3. The Fuel/Ice emulsion, at flight idle, transits the Fuel works fine, no loss of supply or even thrust.

4. At Throttle up, the Fuel is moving more rapidly, ice is "collecting"
in restricted spaces, but the engines respond fully before the accumulation can get ahead of the supply.

5. At full flow, the ice crystals gather rapidly in the exchanger and elsewhere, constricting the Flow of Fuel. #2 drops power.

6. Drawn from Port Tank, the Fuel ice "emulsion" is perhaps 1.5C warmer than the Starboard Tank, and #1 drops 8 seconds later. (Port Tank is in sunlight throughout the flight, the inner portion of the Starboard Tank is in the shadow of the Fuselage.)

Why Crystals and not surface accretion? Agitation, and lack of nucleation. Any "Slurry" will shed its solids at turns and pressure foci differentials. It has to be Fuel Fault.

Airfoil

Lords, Ladies and Gentlemen...

In a nutshell...

How many times has this type of incident happened before?

How many times have aircraft flown through extreme weather conditions (inc freezing temps etc etc) with NO problem at any time during the flight?

Whats so different from this particular flight, to any other commercial flight?

As far as I know, Boeing and BA have issued NO change to operating procedures on the B777 (I stand to be corrected if they have)... Would Boeing have grounded the B777-RR powered, if anyone had lost their lives?

Maybe this is a case for the guys at CSI to solve, since the CAA,AAIB, RR, BA and Boeing seem to be at odds...

Unusual factors, exceptionnal set of circumstances
 

As far as I can remember early comments :

- Multiday rather cold winter operations
- Fuel tank icing messages
- Actual flight much higher than initial operational flight plan (to satisfy ATC request) => hence extra landing fuel.
- Very low OAT at a given time (- 76°c = ISA - 20). What was the TAT at that time ?Somewhere around minus 50 °c ? Quite a cold spot ... for how long?
- no descent to lower altitude over Germany while many other commercials were requesting FL 250 that day.
- no "low fuel temp" warnings, while other airliners got that same warning ...

So ... each fact not exceptionnal. But the whole set still quite infrequent ...

The culprit might be a deficient fuel temp warning system ... and crew putting too much confidence in that system, instead of "crosschecking" fuel temp information with indicated TAT and met temperatures forecast.

I guess (see previous post a few pages above) that crews didn't wait for Boeing or the CAA to take extra margins ... Any captain worth his pay would not want to replay that scenario !

Actually, the operational flight plan was calling for lower altitudes, avoiding very low temperatures ... Is the program already including some margins against limitating low temps? Perhaps ... BA knows the answer.

For Info.

I believe that the B777 fuel system has a suction feed system from tanks so that in the event of a.c fuel pump failure (of anykind),the engines will still receive a flow of fuel via suction ,albeit at a slightly reduced rate.

Suction Feed
-Suction feed line bypasses pumps.
-During climb air from fuel tank may collect in suction
feed line, inhibiting suction feed. Air will deplete
eventually after reaching cruise.
-Once air dissolved suction feed can occur with no
limitation.
-Time to deplete depends on altitude, fuel temp and
type of fuel.

One wonders if the fuel temp sensor was reading correctly?

But, seems too easy a solution.

Even if the tank fuel temp sensors used for flight-deck reporting WERE reading correctly, one may wonder if their reported readings were representative of the fuel temps actually in the dir0 ect path of flow to the engines?

As one very experienced in precision temperature measurement, I know uneven distribution of temps in a large volume (of anything) is a near certainty.

If pockets of colder fuel were present in the wing tanks, perhaps already gelled and stratified into layers or globs of denser fuel materially colder than the tank fuel temp sensors could see, then is it not possible that descent-related changes in aircraft attitude and acceleration could possibly shift the coldest and waxiest fuel into the feed path at the critical late stages of descent, with exactly the consequences observed?

A relatively simple enhancement to flight-deck data might be to add reporting of temps IN the actual flow path near the engines. Likely some of this data is already sensed and available within the FADEC -- and possibly available in a usable form.

To CONF iture, post #1138:

The 777 has no transit lights. Spar valve positions are displayed on the fuel synoptic and fuel management maintenance page. These have to be selected on the lower MFD (or left/right inboard MFD) to enable monitoring of the spar valves.

15 seconds.

28V DC

Spar valve control system (fuel control switches and control relays): yes

Spar valve actuators: no. Limit switches remove power from the spar valve actuators after commanded position is reached.

Continuous 28V DC power is provided to the control relay.

Fuel control switch moved to RUN energizes the open coil of the control relay which then sends a power signal to the spar valve actuator. When the spar valve reaches the open position, power through the limit switches is removed from the spar valve actuator.

Fuel control switch moved to CUTOFF energizes the close coil of the control relay which then sends a power signal to the spar valve actuator. When the spar valve reaches the closed position, power through the limit switches is removed from the spar valve actuator.

Yes , if a spar valve actuator jams while in transit. This wil generate a fault indication on EICAS due to a disagree between control relay position and spar valve actuator position.

Or (highly improbable but not impossible):

A temporary reverse logic to the control relay. Example:
With fuel control switch in RUN and temporary reverse logic to the control relay, this moves the spar valve actuator from open to closed position. No alert is generated because there is no disagree between control relay and spar valve actuator. Actuator moves to commanded position.

If the temporary reverse logic is corrected in less than 15 seconds, the spar valve moves [open->close->open] and will be in transit for almost but less than 30 seconds without ever reaching the closed position. Result, obvious limitations to the fuel flow. No alerts are presented and no visual indications if fuel synoptic and/or fuel management maintenance page are not displayed at moment of occurrance.



see my post #590 on page 30 of this thread:
Code One . . . .


regards,
Green-dot

Arcniz
 

From an ETOPS perspective, I think some folks have a hard time entertaining how both engines could respond similarly to a single or series of faults. The timing isn't so exclusive of what appears to be reality. There is a "range" of possibilities that could easily have trapped both Powerplants in a "simultaneous" cycle of failure. Subtle changes, or commands, or even trimming, result in subtle responses from the A/C. Large inputs can create their own problems.

Possible: BA038, established in flight, is exposed to fierce low temps.
There is some amount of Water in the Fuel, in some "state".
On descent, Equilibrium is established, the "system" is constant
Engines have established stasis, low operating temp. sipping fuel

1. At low altitude (~800feet) the Autothrottles command significantly increased thrust.

2. The engines respond, accelerate, and make commanded power.

3. Everything impacting the new thrust level is instantly at high level

4. Fuel flows rapidly increase, as does temperature in the engines

5. Stasis collapses, Fault (s) appear and overwhelm the system.

Possible: The isolation of the powerplants from each other works to prevent simultaneous failure, but given a demand in the nature of a sudden command for full power, the scale of the demand can "override" functional isolation. What I'm trying to say, is that though the Fuel was contained in separate tanks, it was of the same make-up, exposed to very similar conditions, and could have easily created unrecoverable circumstances in two engines within 8 seconds.

Back to stasis. The profile of the descent and the extended period of low power and uncommonly cold and perhaps contaminated Fuel created a homogeneous environment for the final hole in the cheese. I think the restriction was closer to the HP pumps than spar valves, LP stand pipes, or system piping upstream from the pylons and heat exchangers. The instantly present Vibration and fuel demand, including sonic energy and increased temp. may have served to melt/dislodge pack up the heat exchangers with the ice crystals that may have collected prior to throttle up. Within 8 seconds of either engine? Possible.

Airfoil

Fuel Cooled Oil Cooler
 

The Trent 800 FCOC (black) and fuel & oil tubes:

http://img57.imageshack.us/img57/8356/80079bxa3.th.jpg


http://img61.imageshack.us/img61/4224/80071bou9.th.jpg


http://img57.imageshack.us/img57/1477/80066bsc7.th.jpg


Cutaway of similar FCOC (fuel flows through matrix of small tubes):

http://img234.imageshack.us/img234/8...cutawayez9.jpg

Airfoilmod
 

I fully agree with your assertion that a core vulnerability in ETOPS redundancy is that the top-level systems (i.e. the aircraft, external nav-comm environment, crew) are unitary and therefore not able to be much more redundant.

As you observe, a variant of the same theme is that the recent history of the aircraft is also common-mode for even the redundant systems, esp the power train. When the two halves of the aircraft are somehow identical, certain unanticipated failure modes related to patterns of usage may affect them equally at roughly identical points in time.

Perhaps the logical extension of this observation - especially if the principle eventually is shown to be causative in the case of BA038 - would be a future requirement to ensure that the two halves of an ETOPS aircraft may use many common components but must be deliberately asymmetrical in regard to the form, fit, and function of sustaining flight-critical systems. A logistics nightmare, of course, but probably a reasonable extension of the no-fail-all concept.

Nature handles this problem quite neatly in many living things, including ourselves, by making the redundant critical appendages, sensors, and control systems from the same general parts and pieces, but different in right-left (etc.) symmetry. A substantially trickier design asymmetry might be needed to cover the many plausible common-mode failure symmetries possible with aircraft.

Arcniz,
Several hundred posts ago and I have spent some time trying to find them but it was too time consuming, it was explained that the jet pumps in the tanks are designed to keep the fuel in the tanks in motion to prevent cold spots or stratification from forming. As I understand it, the jet pumps were found to be working correctly.

In normal circumstances, MMM should have been around one minute from touchdown, so I figure out it was only 30 to 40 seconds from impact on this day when right engine reduced to 1.03 EPR
According to AAIB, both EPRs reduced and maintained steady values, 1.03 and 1.02, therefore I don't see as plausible scenario that spar valves were constantly in transit, which would have produced, I think, anything but steady values.

I'd rather go for your previous hypothesis:
And I totally support the following:
I would easily conceive a fuel flow restriction by mechanical means, like a partial spar valve closure, but I don't see how the fuel itself would produce its own restriction in a way that such a restriction stabilizes for the remaining of the flight and don't deteriorate any further ?

The slush (icy/waxy particle) theory can explain the fuel flow restriction with the following conditions met:

a) That there is a location for the "slush" particles to accumulate. (a screen, a partially open valve, a matrix of tubes (FOHE?))

b) That the particles are of sufficient size and properties to accumulate.

c) That the particles can move from their formative location, or that they form in a location with the right characteristics to restrict fuel flow.

d) That there is sufficient time for the particles to accumulate. (that there is enough time at the end of the flight and during the descent that the engines require no more than low/moderate fuel flow.)

The "slush" would allow fuel to flow through at a steady state as long as the fuel flow demand was low/moderate. At high fuel flow levels, the fuel would be restricted to near the low "steady state" value. (sucking fuel through ice using a straw so to speak)

---------------------------------

There are some hurdles to the theory:

a) That the AAIB report identifies that there were "no significant quantities of water." (Definition of significant?) This can be explained if the "slush" was mostly pulled through the system at the time of the crash. Even at ppm of water, 79,000 kg of fuel would result in a few kg of water/ice, which would be enough to cause mischief. Not to mention the legacy water in the system.

b) That the FOHE were clean of blockage and that the fuel feed lines did not reveal any defects or restrictions. This could be explained also with the "slush" being pulled through the system. There may also have been some loss of the fuel downstream of the HP fuel pump from the crash impact, therefore losing the evidence.

If the "slush" was more waxy components than ice, then they could have redisolved into solution, again leaving minimal evidence.

For clarity: 79,000 kg of fuel containing 1 ppm of water would equate to 79g of water (w/w)

In my own realm of fluid dynamics I have witnessed some few examples over many years of systems entering states of very unintended and undesired resonance created by the combination of a free moving valve component and some elasticity in pipework and/or structures.

Acknowledging many of the very informative and interesting previous posts on the subject of BA038, I would like to suggest that just such a resonating system as described could very well have propagated involving one of the non return valves in the booster pump and the associated pipework on the right hand side of the aircraft.
The liquid flow through such a system in resonance will be restricted in proportion to variations of the frequency and amplitude these both being influenced by changes to any physical property within the system and most likely very much exacerbated by any increase of differential pressure across the oscillating valve component.
Before any problem becomes apparent, all individual components function perfectly and none are damaged, nothing is dangerously cold, the fuel is within standards, there is no ice present or any EMI, but at some time in the last minutes of the flight a very rare and of course quite unintended combination of factors starts a small harmonic interference in the fuel flow. These might include all or any of a change in fuel demand, the fuel temperature and viscosity, the airframe temperature, the level of fuel in the tanks, the aircraft attitude and normal vibrations within the airframe etc.
Regardless of the exact moment that this resonance started, as the demand for fuel increases the differential pressure across the now oscillating poppet valve hugely exacerbates the resonance. The HP pump firstly manages to draw the bulk of the liquid fuel from the delivery pipe and the engine accelerates as required but the supply of fuel quickly becomes limited by the non return valve now jumping on and off its seat. During the last seconds of the flight this system is in stasis with the HP pump developing its apparent cavitation damage but delivering adequate fuel to the engine to maintain the recorded 1.03 EPR output.

Each side of the fuel system has matching components – and non return valves - and is constructed as a virtual mirror image of pipework and connections. Particularly given the ever increasing perfections and reduced tolerances of mass production is it unreasonable to therefore suggest that the two similar systems will have almost identical resonant qualities? Especially when we know that the left side has almost exactly the same amount of the same fuel at the same temperature etc, etc?
The range of vulnerability to harmonics created by all the variable conditions mentioned could well be adequate for the left side system to independently propagate the harmonic effect exactly as the right side.
Alternately, could it be that the resonance is transferred through the crossfeed pipework which creates a robust physical connection between the two boost pumps on the right hand side to the two boost pumps on the left hand side of the aircraft. Even though the crossfeed valves are properly closed, the resonance can be very effectively transmitted either through the metal of this pipe or the mass of liquid that fills it.
Whichever, within the period of 7 seconds the left hand system imitates the right but the fuel flow on this side reaches stasis with the engine at the slightly lower output of 1.02 EPR.

This scenario seems to me to very adequately explain all the phenomena that are apparent in this incident including the lack of any physical evidence beyond the damage found on the HP pumps.

I can think of two very pertinent questions that need to be asked.
Is it mechanically possible or actually quite impossible that one or other of the non return valves in the booster pumps could be induced to go into a state of oscillation?
Would any of the data recording systems be sensitive enough to identify and record the anomaly of pulsations in the fuel rather than just interpreting a reduced but steady flow?

Further considerations:
Given the very short period of the malfunction, it seems perfectly possible that no evidence of abnormal vibration within the fuel system would be apparent; in normal operational use all pipes and fittings will show signs or wear relative to normal and expected movement. The most likely appearance of unusual wear would be on the poppet valve assemblies but especially if the limits of movement are constrained by resilient materials this could be almost impossible to detect as being any different to normal wear and tear..

While it is very hard to accept the idea that such an event would happen on two such independent systems, very clearly something has happened. The similarity of the components and the mechanical connection of the crossfeed assemblies however do lead towards an illustration that perhaps the two sides are not quite as independent as they appear to be!

The frequency of the harmonics in the fuel system is most likely going to have a relationship to other natural frequencies within the airframe and hence it seems very reasonable that neither anyone on board nor any of the data recording equipment would detect any vibrations or noises created in the fuel systems as being either out of the ordinary or in excess of those to be expected of an aircraft in the landing configuration.


If the fuel delivery, rather than being a normal consistent flow - even though much reduced - is actually arriving at the burners in high frequency spurts, could this not perhaps explain the unusual engine noise suggested by onlookers, this likely to be far more apparent to those outside rather than those on board the aircraft?

Comments would be very welcome….

JG

Not sure here of units used but!!
1 ppm=1/1 000 000

79 000/1 000 000 gives alot less than 79kg!!

Firelight :-

......... There are some hurdles to the theory: .........


But you have to explain how the engines each separately initially responded and then each managed (at slightly different times) to have an almost exactly similar "problem" which limited the thrusts to 1.02 and 1.03 EPR.

.

Not Ready
 

To abandon ice in Fuel. I have not been 100% comfortable with what water levels in Fuel infer re: AAIB "less than significant" findings. Debating the Authority from a distance with perhaps 1% the information they possess has been frustrating, to say the least.

What the discussion has suggested to me is a different direction; without the data necessary to support it, I'm envisioning a theory that is at least compelling to me.

"contamination" can involve other than chemical or foreign substance intrusion in the Fuel supplied.

Fuel can also be contaminated by SOUND.

Fuel can be contaminated by TEMPERATURE.

Fuel can be contaminated by VIBRATION.

It seems to me that introducing ENERGY into a liquid system can create all manner of mischief. (Witness Cavitation, Foaming, Vaporisation)

Out of my field (by light years), I will suggest that I have seen gasoline "foam" at 1 atm. pressure simply by being subjected to vibration. The corollary, and of the most interest here, would be, given the VERY low temps involved (The lowest flying that day, arguably), what would it take to solidify very cold Fuel? What combination of factors can be entertained that would explain "Bubbling", "Congealing", "Waxing", etc. of tested Fuel?

1. Stasis at descent is demonstrated.

2. Sudden introduction of Massive energy into a balanced system

3. New (Commanded) Thrust level presents dozens of instant challenges to very cold Fuel, not the least of which is acceleration in the lines of the liquid fuel, reducing pressure.

4. Although snubbed against vibration and sound, Fuel passages create several local and isolated environments, each of which must react in its own way to intense vibration (mechanical) and harmonics (Acoustic) while containing extremely frigid Fuel, which has differing resistance of its own to the new "impacts".

5. Flutter. The ancient enemy of powered flight. Any articulating mass in a fluid medium can perform badly at certain levels of tuned energy. This would include, but not be limited to: Fans, Turbines, Pumps, Valves, Filters, Solenoids, and Fuel.

Holding, Airfoil

Yes, that is why I wrote 79g NOT 79kg as you did.

So once again, for clarity, 1 ppm of water in 79,000 kilograms (kg) of fuel = 79 grams (g) of water (w/w)

Cold pockets
 

Please note i'm not a pilot, i've just been following the discussion.

Does anyone know what the recommended minimum OAT is for a Boeing 777 and where the fuel temp sensors are located? There seems to be some suggestions that there might have been a localised region of cold fuel. Presumably for that to happen there would have had to be a preliminary failure of the systems in place to mix the fuel up. If the min rec. OAT for a 777 is colder than -76 (which i'm guessing it is), then there should not have been any concern about even localised regions of cold developing unless there was then a secondary failure of another safety system for warming the fuel.

Which seems a little tentative, none of the system failures have been reported yet, but if it is the case it might explain why it's such a rarity. Presumably, then there would need to be a reason why there was a delay of several hours in the effects of the cold....

Pedro
 

The Fuel Temp. sensing probe is located between ribs 9 and 10 in the left main tank. It is 12.6 inches from the lower wing skin, 40 inches outboard the aft boost pump inlet.

Credit Machaca, ref. to Boeing manual. (Thx).

Good analysis, johngreen.
A spring loaded diaphragm, such as found in a non-return valve, would make a very interesting fluid motor.
A drawing of the valve would be very helpful, as would valve specifications.

Even if resonance damping was applied to the spring or the diaphragm, a resonance might occur in such a valve due to gradual wear of the damping material or increased stiffness of the spring due to low temperature or some perfect combination of both.
Given the rarity of the event, it may be that some perfect combination of conditions and shock waves from other fuel supply components would need to occur at the valve to trigger this resonance.

Recording a low frequency resonance in a fuel flow probably depends on how often the fuel flow is sampled.
If the sample rate is less that four times the resonance, a fuel flow resonance will probably not be detected.

Sampling at two times the rate of the fundamental will suffice1, but your point is still well taken -- if the fundamental frequency of the system at resonance is, say 30Hz, and if the sampling rate of the sensor is, say, every second since a profile is not intended to be continuously constructed, but merely the point value sampled, then cyclic perturbations could be missed altogether, or aliased as noise/variance in the discrete samples, but the resonance signal would not be reconstructed.

I think the idea of resonance is interesting, but improbable. I think we will find a rare, but relatively mundane cause of dual fuel supply transitory obstruction.



1 "Nyquist (1928) pointed out that, if the function is substantially limited to the time interval T, 2BT values are sufficient to specify the function, basing his conclusions on a Fourier series representation of the function over the time interval T."

Location of temp sensor
 

How is it decided (apart from practical issues) where to install the sensor? Was the a/c once instrumented with tens of sensors in various places, then the worst case was chosen?

Thanks soem dood.
I oversimplified my description of the performance of a switched channel A/D by just combining the Nyquest minimum sample rate and the switching rate to sample two separate fuel system samples.
I plead laziness.
To atone, I looked up the performance of the 777 DFDR.

The DFDR in the 777 appears to record 128 samples / word per second. The parameters of interest to fuel flow and engine performance are sampled once a second.
This gives a maximum sampled frequency of 0.5Hz.
No fluidic resonance recording for you.

http://www.tc.faa.gov/its/worldpac/techrpt/ar06-11.pdf

The DFDR in the 777 appears to record 128 samples / word per second. The parameters of interest to fuel flow and engine performance are sampled once a second.
This gives a maximum sampled frequency of 0.5Hz.
No fluidic resonance recording for you.

The maximum sampled frequency is correct, but whether you can record something of interest is not so clear. If the sampling isn't low-pass filtered, you will still record aliases of the higher frequency. Basically you record a frequency on the difference between the signal frequency and the nearest integer multiple of the sampled frequency. So if there was a resonance at 21.25Hz, you'd record an alias frequency of 0.25Hz (21Hz being the nearest multiple of the 1Hz sampling). You couldn't tell from the record what the original frequency was, but you would get a recording of something unusual, so long as the sensors really are capable of doing point-sampling rather than one-second averages.

Awesome, shawk -- I love your concept -- that we atone for our sins here by providing heretofore unpresented information that is potentially relevant to the investigation!

And Fzz is right on, as well! :ok:

(Heck, if everyone did that, we could have this thing wrapped up in a week!) :)

Excuse the interruption in the jitters/flutters/harmonics/surges and PPMs, but from another thread comes 'confirmation' that a BA internal report says Flaps went 30 to 25. Hope the AAIB read this:p

When I flew
 

With a Johnson bar, I learned alot about flaps in ground effect and thereabouts. How to "pop over" an obstacle on very short final, how to make my brakes more effective for short ground roll, how to drop like a stone or float like a moth; this thread is about to get very interesting.

How's YOUR Johnson, Mav?

The profoundly toxic Honeywell Aerospace site lists the Digital Flight Data Acquisition Unit used in the Boeing 777.
http://www.honeywell.com/sites/aero/...74EACA8348.htm
To read the specification sheet, you need to register for a Honeywell ID. This will take several days to get.

The Honeywell unit is a 12 bit A/D with a single 12 bit sample per second for fuel and engine parameters.
A sub-harmonic that is 7 octaves below the fundamental is quite unlikely to be resolved with a 12 bit A/D and a single sample.

A better bet might be to analyze the audio track of any amateur video shot of the landing.
If anyone has a pointer to these, post it here and I can take a look at the audio track.

Shawk,

> This gives a maximum sampled frequency of 0.5Hz.
> No fluidic resonance recording for you.

I agree it wouldn't accurately record higher frequencies but that doesn't mean the samples would all appear "normal" - unless the resonance was somehow synchronised to the sampling.