Re-Heat

It almost makes one wonder if insertion of random, minor, marginal measurement and specification errors throughout a precision construction process would be desirable (thinking again back to the airway precision debate where random lateral offset - though still within the airway - is in current thinking).

I hastily add that I should probably patent this idea.

ChristiaanJ

Re-Heat,
You may have a point.....

Remember the static test on the A380 wing that broke at something like 148% of design load, rather than 150%?

Maybe, sometimes you would like to have that little bit of extra margin, because they did not mill out every single "superfluous" ounce...

Might be worth discussing, although it's more academic than anything else.
Accidents due to this kind of extremely rare combinations of circumstances still happen, agreed.

But as long as third-world airlines dump badly maintained aircraft in atrocious circumstances with a frightening regularity.... I think our priorities are elsewhere.

airfoilmod

Having to do with mirrored and duplicative systems (Fuel). has been addressed elsewhere also. In perfecting more and more dependable systems, the Human side of the equal sign becomes more and more..... Dependent. ETOPS must for purposes of economic feasibility, divert from isolation and engineered anomalous design to make more and more infrequent, the failures which challenge the philosophy. There is a definite difference between redundancy and anomalous isolation. Redundancy merely provides two systems to fail under similar circumstances; heterogeneous engineering separates the systems in ways that "prevent" mirrored fault.

Airfoil to Re-Heat: it's called engineered anomalous design.

As I offered before, 038 isn't about Etops, instead its about vulnerability of separate and equal systems.

Green-dot

trashie:
propaganda:
CONF iture:
barit1:
I assume trashie is referring to the boost pumps, not FADEC or HP pumps. If so, the pumps themselves are outside the hull (which functions as a Faraday cage) and very unlikely to have been affected.

Regarding the (pumps) electronic component locations relative to the mentioned seating position of the "official with his laptop", he would have been nowhere near any of those components if you know the T7 well enough.

Suppose the pumps were affected (highly unlikely), what effect could there have been? Stop the pumps? Suction feed would still feed the engines. Or make the pumps (3 phase AC) somehow run in a reverse direction . . . .and then all four of them at the same time with their LH/RH electrical circuits well separated? Again, highly unlikely.

If there is any substantiation in what the accident investigator discussed, the scenario could fit uncommanded transit of both spar valves. More holes in the swiss cheese had to have lined up though before such a scenario with a laptop could be considered as plausible.


Regards,
Green-dot

ChristiaanJ

The loud laughter and hilarious applause following this "revelation" have been edited from the tape in question.

johngreen

pls8xx

Keep up! Just to clarify the 777 fuel system, there are two booster pumps each with a non return valve located at the rear of each left and right wing tank. Each pair of pumps delivers fuel independently (unless the cross feed valves are opened) to its respective engine each of which is fitted with a heat exchanger and each with a mechanically driven HP pump.

Flight Safety

Your experience with the shower is a perfect analogy to the effect I have described.

The booster pumps are centrifugal and will operate constantly at their rated speed even when no fuel is actually being consumed as will happen for example during the initial stage of engine starting. As the resistance to flow is decreased, the volume of delivery will increase and the delivery pressure decrease accordingly. (A graph indicting values is on the catalogue for those interested).

The HP pump has to be a positive displacement pump – for example a vane or gear pump – such devices being able to produce the much higher pressures required for the burners. Regardless of the boosters trying to always deliver their maximum possible output, the HP pump will only allow a fixed amount of fuel to pass for each shaft revolution, the rotational speed being determined by its mechanical connection to the engine. It is this then that offers the restriction to flow that permits the build up of pressure fluctuations having the effect in this circumstance of being a valve that is opened or closed against the delivery from the booster pumps in accordance with the engine fuel demand.

Although it does not affect the resonance scenario, the HP pump is also fitted with a bypass arrangement. By this, the pump will always be delivering too much fuel for any given engine speed, the excess being recirculated in the vicinity of the pump. This also allows the pump to pressurise the engine fuel system against the closed fuel valve during the start procedure when the pump has no option to rotate with the engine although no fuel delivery to the burners is yet required.

Ailfoilmod

Enjoying your mind stretching additions. I keep reflecting on the possible pressure waves and other dynamics that might have been dancing together in the remaining bulk of fuel in those drum like fuel tanks…

Re-Heat

Exploring this incident - among others - I also have many thoughts about the relevance of precision on safety. It remains for me that if we were really meant to fly our bodies would be formed with wings, a tailplane and something hopefully more akin to a Merlin than a flat Lycoming between our legs. Having a personal lack of such equipment, given the choice, I think I would anyway still tend to choose the safety of today’s imperfect machines over too many adventures being carried aloft in the likes of the Doves and Herons of my youth. The fallibility of the human controller(s) nevertheless probably remains a greater threat than the fallible machine.

Regards
JG

Machaca

1. Low Pressure boost pump in wing tank, output is "constant"
2. Low Pressure - engine driven, output is direct relationship to engine rpm
3. High Pressure - engine driven, output is direct relationship to engine rpm







Low inlet pressure to the HP pumps is a focus of the investigation. There's a lot of cheese slices between the fuel tanks and HP pumps!

Harmonics? Perhaps.

Beyond FCOC and fuel tubing atheromata, I'm suspicious of excessive Prerotation and Backflow of unusually cold & viscous fuel reducing output of all the low pressure pumps.

An increase of intake vortices at the wing tank boost pumps may have been exacerbated by relative fuel surface level changes in relation to the pickups due to aircraft pitch changes during approach.

http://caltechbook.library.caltech.e...igs/fig408.jpg

http://caltechbook.library.caltech.e...igs/fig406.gif

-=MachacA=-

Re-Heat

I wouldn't suggest a return to past, inherently dangerous designs would be desirable!

airfoilmod

might involve the HP packing viscous fuel into the nozzle annulus and/or the bypass (relief circuit) while the Low side simply couldn't keep up. At this point the fact that the LP's were active would RESTRICT free flow, the alternative of suction being locked out. At some location (bearings at high side) is where the Fuel "Sheared" into dense "petals" and "voids" (cavitation). Alone, I think cavitation was insufficient to choke the supply, but briefly. At this point (when the Engines were ~1.57 EPR), I believe the coup de grace may have been introduced into the supply with an as yet undetermined combination of Resonance, Harmonics, flutter (at LP valves), and an inability of the system to "select" suction or boost. The reason I'm keen on flutter is that it would explain a halving of fuel flow during the time its cycle was "closed", against the LP port.

Re-Heat: A little too cryptic for me, sorry

rgds Airfoil

ChristiaanJ

Point taken.
However, being jumped on by the bean counters because it broke at 152% is not ideal, either....
Should probably be discussed somewhere else, because it's an interesting notion.

Flight Safety

Or it could be that when the fuel flow rate increased with the power increase, something got sucked into the pickup screens (that later disappeared) thus reducing the flow and was held there by the suction until impact. In theory this could fit the facts as we know them.

How many fuel pickups are there per tank, 3?

airfoilmod

posit feels nearest the fact given the Data. Restriction in the form of ice crystals was mentioned by D. Carbaugh, Boeing Chief of Safety Pilot. The restriction of the Fuel itself, as a too viscous mass or a too energetic mass, full of pressure waves and other considerations, all are important to explain the low fuel supply at HP.

Machaca, thanks for the visuals, great help.

Airfoil

Flight Safety

Yes, thanks Machaca.

It looks likes there are 3 fuel pickups for each tank (left and right). Each of the 2 boost pumps (per tank) has it's own pickup, and the suction feed has it's own pickup. So nice redundancy for fuel pickups, but this means 6 fuel pickup screens would have to be blocked, all at nearly the same time for this scenario to be the culprit.

airfoilmod

makes a more reasonable solution one that isolates one or two crooks or crinks, but those needing to be vulnerable to the cause of fault. And vulnerable in the sense of simultaneous failure (~8 secs.). I look suspiciously at FCOC, and the plumbing around it. What a swell place for ice crystals to cascade into a clump and choke the life out of our A/C. Or like a cocktail shaker to block flow due to agitation and foaming. Boeing has the answer in the test cell(s).

Airfoil

Flight Safety

The problem with the FCOC is heat is transferred from the oil to the fuel. This would seem to reduce our chances for ice formation between the LP pump and HP pump. Further, we haven't heard anything about any possible ice damage to the HP pump, only suction cavitation damage.

I don't know, maybe the fuel was as thick as mallases.

What's interesting is that both engines rolled back to virtually the same power level, EPR 1.03 and EPR 1.02. So not only did the restriction or obstruction occur at virtually the same time on both sides of the airplane, but the amount (quality, quantity) of the restriction was virtually the same on both sides. I'll bet the suction cavitation damage is virtually the same on both HP pumps as well. I wonder if maybe Re-Heat is on to something here.

airfoilmod

had to have been a transient, and fuel related. Scrapers, plastic, condoms, not possible, the block was a phantom, then a plug, then a phantom. What type of material would flow just fine at green descent, flow more slowly as flow and pressure increased, then mimic a solid, resisting all power and pumps to get it to move along. A change in phase? likely. A change in viscosity? easily. How about a change in energy content. John Green? (not caloric, but entropy)

Flight safety, I'm glad you brought up Oil Cooling. A thin oil, jet lube does not do well in very low temps. Fortunately it can be cooled by fuel, but wait, we need to heat it. Congealed oil in the cooler? Subject to convective cooling by the airstream at -70? Why should frozen oil bother 038? There's plenty left in the case. But if some is locked up in the FCOC and further cooling the fuel, well that's a horse of a different hue. Temps in the cooler? energy at 1.57EPR in the cooler. Where is the bypass? Does it matter? Who needs ice when the temp in the FCOC is -70C? Thaw? sure, give us twenty minutes. Pure speculation.

Flight Safety

Whatever the restriction or obstruction was, it was very consistent on both sides of the airplane. This should be an important clue because with this kind of consistency (on both sides of the airplane), this should a reproducable failure (since it already occurred twice the same way, on both sides of the same airplane).

Flight Safety

I have another idea.

I noticed looking at the drawings supplied by Machaca, that all 3 fuel pickups are located next to what appears to be a bulkhead between the main tank and the center wing tank. Since the center wing tank was basically empty at this point in the flight, what would the temperature of this tank have been? If this empty tank was colder than the fuel in the main tank, how would this have effected the portion of fuel in the main tank located right next to the bulkhead (where the fuel pickups are)?

airfoilmod

I think for there to have been a double failure so close in time, it's necessary to look for areas where parameters would be fairly close in range. The Spar valves shut at the same time could initiate cavitation but would make both environments (fuel lines) dependent on the "same" "set" of conditions. The volume of the lines, though similar, would have been susceptible to varying results from suction due to connections, tubing strength, indeed even the "position" of the valves themselves.

When I think of cavitation, I visualize a smallish tightly sealed space subject to great energy at one end and collapse close by (shear). I'm looking for a block close by the HP's, one that reacted to parameters present in spite of all engineered precautions and anticipated challenges.

On the other hand, the delay of 7 seconds could represent the "differences" in systems' reactions to the fault: partial "collapse" of the piping to fixed values, time for two spar valves to close completely having lagged, run up of the lobes of the HP's to max, all other junctions to reach maximum deflection due to demand, etc.

Question: Could cavitation itself have been responsible for loss of commanded power and failure to respool? Only in the sense that it represents the "onset" of fault, the plug would have been responsible for the transition from "cavitation as evidence" to starvation and ongoing low power. 1.02 and 1.03 EPR mean not much is happening in the thrust department.

FS: I think the empty center tank would have performed more as an insulator than a chiller. Think Thermos.

Green-dot

Posted by airfoilmod:

There may have been a time lag between the two spar valves closing, either at initiation or in travel time. One valve may have been slightly faster than the other. Depending on what hypothetically would have caused such an anomaly, one valve may have closed further than the other before restoring to normal condition.

Not necessarily so. Remember the engines were initially accelerating, demanding a considerable increase in fuel flow. Sudden deprivation of demanded fuel feed, depending on timing and rate of the reduction reaching the HP pumps could lead to the same results.

Fuel lines (LH and RH) between the spar valves and the engines are similar and almost a mirror image of each other. From the spar valve the fuel line initially runs parallel to the rear spar and then makes a 90 deg. turn forward following a rib then bends slightly inboard before making another turn outboard then forward and inboard again (going around the drybay) before finally going relatively straight forward, entering the engine pylon and finally going downward connecting to the engine fuel line. Viewed from the top the fuel line path roughly represents a Z shape followed by a U shape (around the dry bay) and then straightening out into the pylon area.

However, (depending on many "ifs" yet to be answered) if the spar valves were affected in the way i described at the very beginning of this thread, it should not be ruled out that the APU fuel shutoff valve (and perhaps the APU remote shutdown) electrical circuitry was also affected. If the valve was electrically opened in the same action the spar valves were closing, all of the fuel lines between (partially closed) spar valves (closed Xfeed valves) and boost pumps in the fuel feed manifold of the LH fuel system could have been acting as an accumulator providing just that much more buffer in feeding the LH engine before rolling back 7 seconds later than the RH engine did? The APU fuel supply line is a very long line going to the tail before ending at the APU fuel control unit. With a system out of wack, that APU fuel control would probably have continued to shut off fuel supply to the APU. Just as a reminder, there is still no official answer as to why the APU door was in the open or partially open position at "touch down", of which there is photographic evidence.

An empty center tank filled with air and fumes will adjust faster to ambient temperature than a main wingtank (partially) filled with cold soaked fuel. Furthermore, the center tank temperature is also affected by the heat radiated from the airconditioning systems directly below the center tank.


Green-dot