Back to Background
 

Twinops. Global, 12,14 hour sector over saltwater and/or Granite ops.
OEI ? OK, worst case. Hmm...Not after BA038. It is difficult not to oversimplify the problem. The fault? Mostly with Gross Fuel architecture.
Pipe Ice. A look at the difference in systems RR/GE? Not much, and given that both firms have been building powerplants since Frank Whittle invented the format, (Apologies to Messrs. Pratt, Whitney), it is not completely accurate to find fault with the engines. Wings Ice, it is a fact of Life. Fuel has water in it, likewise a fact.

Redundancy in ETOPS. Don't forget, that was what was eliminated. Not added. Safety no longer in numbers of engines, but in engineered dependability. At a certain point, the Fuel path is symmetrical, common, undifferentiated. Isolation of systems isn't the solution; the solution may be additive symmetry. If it turns out that Fuel simply is what it is, wet, then as the Fuel closes in on the nozzles, perhaps adding alt. fuel as an option on the Flight Deck may return some "fails safe" to the flight.

In other words, relying on Two Engines is a proven concept, proven.
A step backward to a bypass, or "branched" system, rather than pinning all reliability on a "single pass" system might be worth looking at.
Weight? Cost? Of course. But as the Fuel necks down to ever more critical pathways, an alternate Path might be attractive when Ice becomes a factor.

Earth, The Water Planet.

Discussions about Twin engine operations don't belong in this thread :=

I'm curious to know the bore of the pipes in the FOHE which the fuel passes through. Looking at the photos, I've guessed at 6mm, am I far out?

Airfoilmod - Symmetry?
 

My light hearted observation was prompted by an assumption that since there is no (significant) net heat gain, one fluid body loses as much heat as the other gains, and a second assumption that it would work as well whichever was the hot side, and whichever was the cold. Add the (uninformed) gut feeling that the tubes were more vulnerable than the shroud to blockage, and I'm there.
But reflecting on your post, the contact times for a plug of each fluid are not necessarily the same, so maybe the symmetry is broken...
There is no difference between us in the logical need for a bypass mechanism. I just liked the idea of a (very) simple one that does not rely on a moving part, and which puts any ice right in the heat exchange zone. Can I have it as well, please, just in case the (relatively) simple one with a moving part gets wedged open by a lump of the ice solid in below zero fuel when TOGA is needed. That scenario would be the very devil to diagnose from the expensive later consequences.

Rightbase
 

The design basis is that it is an oil cooler, not a fuel heater.

That is why the oil takes a convoluted path around the tubes, diverted back and forth by baffles, to ensure a (relatively) long residence time in the cooler and thus maximum cooling.

The fuel is supposed to go straight through the tubes and out the other end, with minimum restriction to the fuel flow. Very little heat from the oil passes to any particular molecule of fuel, but that doesn't matter because there is plenty more cold fuel coming along behind.

Of course, the guy who designed it that way didn't know that lumps of ice were going to be thrown at it. :uhoh:

If you swap the flow paths between fuel and oil, the result will be warmer fuel and much less well-cooled oil.

Sorry, that is thermodynamical nonsense.

If the fuel would be getting warmer, it would mean that more heat (energy) was transferred to the fuel. All energy that is transferred to the fuel, must come from the oil (disregarding other heat sources for the sake of the argument), which means that the oil would also be cooled more.

What you state above might be true if the flow rates would also be exchanged along with the flow paths. But that is not possible, since both the oil and the fuel flow rates are dictated by operational conditions.


Bernd

Bernd
 

The FOHE cools Oil. Fuel flows through 1,080 tubes with spaces between all of them between Face and Exit. This reduces cross section and increases Pressure and transit time. But the Oil gets cooled. There is a bypass for the Oil when further cooling is unnecessary, yet the Fuel must still transit the FOHE, regardless of Engine Demand. At this point, if max Fuel is commanded, and no Oil is being cooled, why, in your opinion, must the Fuel continue to flow through this necked down passageway, just to get to the HO side of the pumps for the FMU?

What is your opinion of the GE architecture: HP upstream the FOHE?

rgds.

AF

Not knowing, as we assume the designers did not know, that ice may be a problem, the question is: why not?

The fuel delivery system, including pumps, filters, and the path through the FOHE is obviously designed to be able to deliver maximum fuel flow needed for TOGA thrust, and then some.

Without our hindsight, nothing would have justified the additional weight and conplexity of a fuel bypass.


I'm not a fluid dynamics engineer, so I would have no opinion on that. I assumed from earlier pictures I saw that GE had a different layout of pumps and coolers than the Trents, but now I'm no longer so sure. I've read here with great interest about benefits and disadvantages of different designs, but I'm not in a position to judge any of that.


Bernd

Weight, weight, weight...

Honestly, weight is really not a problem these days on commercil aircraft when we are talking of a few pounds, seriously.

Think of just a crew putting on a few pounds each, let alone big blokes Vs small blokes, or the whole pax... or just a gallon or two extra fuel - honestly, things are NOT that critical that it would come into a design engineer's decision when safety is an issue.

Complexity Yes, weight, No!

I've seen parts fail on light aircraft because ONE pivot pin was made 0.020 " too small dia. when to have made a 1" long pin twice that diameter would add about a gram - and people die :ugh:

Both
 

Engine designs, GE and RR, have "bypass" for Fuel. It is at the FMU in the Trent, and the HMU in the GE. In this case, Fuel in excess of combustion requirements, (Spill) returns to the HP via dedicated plumbing.

I would agree "weight" isn't an issue for an additional bypass for Fuel around a clogged or malfunctioning FOHE (F/O HX in GE). I also don't think complexity is an issue; GE has some complexity that Trent does not.

The GE design has three separate Heat exchangers, Trent only one.
Two of the GE's units are included specifically to mitigate Fuel icing.
The Main HX and the Servo HX are paired in the Low Side, and the schematic states they "Prevent Fuel Icing". An additional HX uses separate sourced Fuel to cool the Gearbox Oil.

Here I need to back track and perhaps apologize. On different occasions I have read here that GE's HX is downstream of the HP Fuel. I now believe that not to be the case. It remains a question; the schematic I use is a simple one, the orientation of the Main F/O HX cannot be perceived conclusively. In bsieker's post #2316 he demonstrates his idea of the critical architecture, HP in front of HX. This I think if not wrong, is not accurate to certainty.

To be quite fair, though the schematics I study are from a common source, again they are not blue-line exactly, and I regret any mistaken conclusions I may have encouraged by being so convinced of conclusions by others and myself.

This leaves a major discrepancy. GE in its schematic clearly has addressed Fuel Icing (potential) in the text of their document. Rolls has not. This does not mean Trent didn't consider Ice in Fuel, just that they didn't mention it in their schematic's text. Include a reasonable vulnerability in the architecture relying on one FOHE, without Heat, (Other than Oil), or Flow Bypass.

NTSB's release and letter continue to shadow the 200-ER. I still don't see the proposed mod to the FOHE as a "Fix". There may be other steps in Rolls plan, but without knowing what they are, the change doesn't address the known Icing other than to try and keep it melted. A very abbreviated Fuel testing program brings up new worries; it didn't solve anything, and really hasn't even identified the specific problem or its potentials.

AF

Oops
 

Mea culpa. I forgot to allow for the flow rates.

However, I stand by the original description of why the flow paths are chosen the way they are.

Sooty

Just to get the facts straight. Idle fuel flow is less than 1,000pph (see first AAIB interim report, p.4, second paragraph), at which flow rate the difference between counter- and concurrent flow designs may become significant.

In the case of BA flight 38 the engines rolled back to between 5,000 and 6,000pph, which was still not as much as demanded, but significantly more than idle.

It is also worth noting that that reduced fuel flow was not low enough to allow the ice to melt sufficiently quickly (or at all), whereas all tests indicated that idle fuel flow was low enough.

Two more things to consider:

- would 6,000pph in a parallel flow design have been low enough to allow the hot oil at the cold fuel inlet to melt the ice?

- would a parallel flow design FCOC of the given size have enough oil cooling capacity at idle fuel flow?

I cannot answer either of these questions, but I assume RR engineers are already looking at them (and a lot more.)


Bernd

Bernd
 

What I had in mind, 1.15 EPR, not Idle. My mistake. Think about your 1,000pph. The 895 burns just 200 gallons an hour at Idle? Who'd have thought.

AF

Swedish Steve has pointed out that the T800 also has an air-cooled oil cooler - presumably this is intended to cover for any lack of cooling capacity at low fuel flow rates.

Sooty

sooty
 

I would think that the AOHE is a backup for a system that at times, works too well, the Fuel. Thermodynamically, air is not efficient compared to liquid. Up high, where Ice is forming, the Oil runs cool, and the FOHE isn't as critical as Low, where Thrust demands are high, variable, and the Fuel is warmer. My picture of the Delta flight is high cruise, very cold temps, Icy Fuel Pipes, and not much need for Oil Cooling. So when the FOHE packed up, it wasn't in danger of being unable to cool Oil, but in causing Fuel Starvation. Had the phenomenon been addressed, there would have been an Alternate Fuel path available, sensed by diff. Pressure at the FOHE "ends". IMHO

AF

I hear that they are breaking up the aircraft at LHR, does anyone have pictures? I understand that the cockpit has been cut away, perhaps to be converted into a simulator?

That's correct for high FL ("first few minutes of the descent") but at low altitude the idle FF per engine must be closer to 2200 pph and even 2600 pph when in landing configuration.

From a contractual standpoint, where is the physical boundary between Boeing's sphere of responsibility and that of Rolls Royce, the engine supplier? Would it be in the area where the engine nacelle attaches to the wing pylon?

Presumably, the contract under which RR supplies its engines to Boeing would include detailed specifications regarding the nature of the fuel composition/condition to be delivered by the 777's fuel system at the Boeing/RR interface. It seems unlikely that the contract would have specified the sort of volume of water/ice which arrived at the FOHE. If so, RR could argue that the fuel delivered to its engine was outside specification and, therefore, they (and their insurers) are not liable for the loss of the aircraft.

I doubt the contractual boundary is in the same place as the physical boundary.

Lay Theory
 

The 777 supplies (3rd Party) Fuel to any of several engine types. In the case of the GE, where fuel icing is addressed in the design, there have been no reported rollbacks, so at first blush, the uphill path is RR's. But it is not that simple. Sub contractors for design, materials, manufacturing, testing, delivery, spares supply, transit, indemnity, Fuel, SOPS, ............
The list is long. Welding subs, branched and subrogated claim, maintenance, etc. etc. The Room will be full.

AF

Not sure about your wording there, airfoilmod. I'm sure GE, RR and others all designed for the level of fuel icing anticipated/specified by Boeing.

If the GE arrangement really is better at accepting the unexpected quantities of ice dumped from the pipework, then that is luck, not design.

Two rollbacks over the entire 777 operation is a very small statistical sample. I'm not yet convinced this is only a RR problem.

Sooty

sooty
 

The Icing of Fuel was explicitly shown in the schematic for GE. It was not shown on the schematic from the same source for RR. I mentioned that before, and meant only that, not that RR hadn't addressed the problem as understood at the time. The presence of three separate HX on GE may or may not be more effective, and your noting that only two rollbacks have been documented for the Trent is quite valid.

airfoil

tanimbar

The 3C you quote is the required margin above freeze point, not an actual temperature.

With -47 fuel the minimum fuel inlet temperature would have to be -44 for RR engines.

Similarly GE could accept -47 and CFM -42 at the engine fuel inlet.

But perhaps you knew that already.....:confused:

Sir Richard - min fuel inlet temp is relative, not absolute
 

Thanks for making clear the figures in that document that I mentioned.

Not for sure. It was one possibility I identified from the words used in the document.

Thanks again.

Some pounds here plus some pounds there make up for two-digit tons in the end. I am sure, weight IS an issue in aviation. And it may even be an issue WHERE it is.

Mismatch
 

The prior ref. to weight involved a hypothetical increase for additional system design in the Trent. Specifically a bypass for Fuel around the FOHE. The Trent is already leaner by 2,600 lbs than its GE counterpart, so some room may exist for a contemplated addition of Plumbing. Of course weight is important. Safety frequently revolves around weight. Where it is, and how much there is.

AF

Mismatch
 

Thanks Airfoilmod for clarifying that.

Of course weight matters, but in context... if you had 1,000 similar linkpins or rivets holding something on, you would spend a lot more time optimising the design for size, weight, material and stress levels as well as on testing to assure your final release.

But if you have 2, one per elevator for instance, over-designing them by 50% wouldn't make a bat's hat of it, would it, and in fact, because being only two in a critical path, would probably require you to apply a larger proof/ultimate factor, a wider fatigue factor and for peace of mind, a larger Reserve Factor.

I've seen one or two rather crass and ultimately deadly results from not understanding or formally addressing a sensible weight/citicality balance... e.g. a single 7/32" shear pin might suffice and meet the design case, but one day fail, a 1/4" might never go anywhere near a failure mode. Weight difference approx. zero!

If an engine needs a by-pass around a heat exchanger, being on the C.G or close to it, even if it cost one pax, or two on MTOW, if it needs it - it needs it - FULL STOP

I am really surprised that Boeing let the R-R Triple7 out of the barn without very aggressive fuel heating capability. Boeing's experience with fuel icing goes back six decades - and they have a library full of lessons learned from B-47, B-52, 707 etc.

Look at 747 etc. fuel schematics and you'll see what I mean.

The relevant point about ETOPS is that twins fly about 4000ft higher than the equivalent 4 engined aircraft. That must mean lower fuel temperatures over the same flight time. Has anyone checked?

Any details would be appreciated.

Just to reflect on the ice theory. Last week I personally experienced something totally unrelated to this incident but then perhaps not.

While extracting some money from an ATM (automatic teller machine) my mobile phone was activated as someone made a call. At that exact moment I was logged into the machine and had punched in the first of 2 digits for the amount I wanted to extract (40 euro's) when the machine went berzerk. I never got to punching in the second digit or confirm the amount by pushing the "ok" button as it spontaneously returned my card and released 100 euro's.

I finished the call and then, confused, I made another extraction with the phone off and the ATM performed as advertised in a normal way.

The incident is now under investigation.

This brings me back to BA38 and how perhaps a one in a zillion chance of EMI/HIRF could have ended its carreer prematurely.

Green-dot

With Mr. G Dot's post, we've now come nearly full circle in the thread.
 

The PM's antenna farm radiations coupled with a nattering nymph's Nintendo DS
in row 13 caused both metal-encased & shielded ECC units to go bananas.

Ice was a related factor as there was not enough in the mother's drink in row 13 so there was just enough added vodka to knock her out keeping the woman
from such maternal duties as slapping the wayward child on the noggin when
he continued playing the small electronic aircraft systems jammer otherwise
known as a portable video game.

When reporting the ATM anomaly to the bank I was met with the same cynical reaction. The reaction was that it was impossible radiation from a mobile phone could cause the ATM to fail. "It was just a matter of coincidence . . . . ."

At least I'll give credit to the AAIB as they did investigate the possibility of EMI/HIRF affecting BA38. Question remains though if all conditions at the time could ever have been duplicated and if all essential data was recorded. For now I will wait until the final report has been released.

Green-dot

Barit1 - I am really surprised that Boeing let the R-R Triple7 out of the barn without very aggressive fuel heating capability. Boeing's experience with fuel icing goes back six decades - and they have a library full of lessons learned from B-47, B-52, 707 etc.

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

Any chance Boeing and R-R had a difference of opinion, and R-R stated "our engineering says it's good enough, and you're just operating on opinion" ?

Ice
 

757 is my speciality and I know why, when on the walk-round, the left wing always pours water down the back of my collar whilst the right wing remains frosty. Is the 777 walk-round similar? Without going all the way through the thread, I would assume that wing tanks were in use at the time. If so, all a bit odd?
Monom

Thought I read, Boeing purchased MMM, that 777 appears to be a tough old bird, great landing by the crew as well.

JoeTom wrote;
"Thought I read, Boeing purchased MMM, that 777 appears to be a tough old bird, great landing by the crew as well."

If Boeing did, then they bought it off the insurers and it was in small cut up sections!

May I ask a pertinent question? Anybody still breathing at AAIB? Have they been redundantated?

Will,

AAIB are still well within their averaged time frame.

Nothing sinister here.

I also believe (naively many will say, no pun intended) this thread actively informed AAIB of sensible possible lines of enquiry...........

All covered in great detail in earlier posts.

CW

chris weston - Thanks, and I agree this thread was food for thought. No acknowledgments are necessary, nor will they be likely forthcoming. That is as it should be. Honestly, I think some of the boffins at AAIB might resent some of the comment here, I probably would. I did not mean to imply any suggestion of conspiracy or anything sinister, just trying to foment a comment; thanks again. Also, I would definitely include your comments on thread on my list of very constructive contributions.