Pilots off hook?
It looks like the AAIB at least appears to be backing the pilots version of events; it seems the engines provided less power than commanded - the question seems to be "why?".
The Air Accidents Investigation Branch has issued an update (pdf) on its investigation into the 17 January crash-landing of a Boeing 777 at Heathrow, indicating that there "were no anomalies in the major aircraft systems" and that the "autopilot and the autothrottle systems behaved correctly and the engine control systems were …
Well at least this exhonerates BA and it's crew - which is great news. imho they did a great job getting the plane on the ground under really difficult circumstances without injury or death. It clearly vindicates BA's training and procedures and the crew's abilities.
What's a lot more worrying is that a well maintained, properly operated modern airliner crashed in normal flight - and there's no real indication of why, despite all the flight recording data being available.
But then again a cynic in the 21st Century where propaganda rules supreme what company is going to admit to the public at large that faulty system exist in large passenger aircraft given the number of legal sharks and super cleaners that exist in the wild for just such an event to clean up both companies big time !
I've been wondering:
Did they retract the flaps to try and make it to the runway?, I doubt it.
Do they have a drill for power loss on approach or practise it (on the simulator)?
The flaps allow the plane to fly more slowly but at the cost of greater drag, hence the need for throttle to maintain speed and/or a steeper glide angle (i.e. less distance made for height lost). Had they retracted the flaps they may have been able to make it to the runway. They probably did not have time to asses the risk/benefit of doing so, or maybe they didn't have enough speed for the plane to fly with reduced flaps.
Still, they did well to crash gently enough not to kill/injure everybody.
just my 2p worth.
Summarising:
Flight Crew acted properly
Autopilot behaved properly
Autothrottle behaved properly
Engine Control Systems behaved properly at all points
No evidence of FOD
Engine turbine vanes & stators all OK (before impact that is)
No evidence of Water contamination in fuel
The only problems appeared to be:
Debris in a single heat exchanger (shouldn't affect right engine).
Indications (not evidence) of aeration in fuel supply to both pumps (but both capable of supplying full flow)
Abnormally low temperature during parts of the flight.
Assortment of debris in all tanks.
I have a BEng in Aeronautical engineering, but have never worked in the the industry so mine's all theoretical. None of the problems appear to be that severe, which worries me. It'll be very interesting to see the final conclusions, though I agree heystoopid, I can't see anyone [wanting to] take full responsibility for this one.
For the sake of Boeing (and I'm not a major fan), they'd better find a definite cause, or else confidence in the 777 fleet will take a beating.
The flaps allow the wing to fly with a greater angle of incidence to the airflow, which means it can fly more slowly than normal without stalling - probably a good idea when landing.
If they'd have taken the flaps off I strongly suspect they'd simply have stalled and hit sooner.
Sure, the flap gives more drag, as a price to the increased lift (so does an increased angle of attack, byt the way).
Given the gliding-slope the plane had, the pilots did everything correctly. They bled off speed with the flaps to reduce relative speed to the ground (and thus reduce chances of friction-fires from speed, or massive breakup of the fuselage due to impact velocity.
disclaimer: I am not a commercial pilot.
disclaimer 2: I've been doing model airplanes since before the challenger accident. Dead-stick landing a plane is much more common with model planes, and the real trick to reducing damage is ALWAYS to get rid of speed.
//Svein
I'm an ex-RAF engineering officer, and my first thought when I saw the incident was to thing damn good job by the aircrew. The 2 worst things that can happen in an aircraft are engine failure after take-off and engine failure on landing. At around 700ft, you are pretty close to landing anyway, so the time to respond to a problem with engine response was minimal. To get the jet on the ground with no deaths or serious injury is testament to some outstanding airmanship. I suspect anyone on these forums who claims the aircrew could have done more doesn't really know what they are talking about. The pilots are well trained in simulation for this, but doing it right first time for real is hard, and they deserve a hell of a lot of credit. BTW, raising flaps at approach speed is one of the silliest suggestions I've ever heard. Without engine power the aircraft would have stalled and dropped like a stone - we would have been mourning the dead rather than investigating a non-fatal accident. If you are ever on a jet that has to go-around and you are in a window seat, watch the sequence of events. Engines will spool up, the aircraft will start to pick up speed, then flaps, slats and undercarriage will be retracted.
Now, it strikes me that this initial announcement basically results from little more than an examination of the flight data recorders. The comment about small items of debris probably comes from analysis of mag-plugs. These are small magnetic plugs that are fitted into things like oil and fuel pipes and systems to pick up debris and analyse where it has come from without stripping the whole engine etc. Also they may have picked some stuff up from fuel filters. Contaminated fuel is a very serious issue for aircraft operations - so its something they will investigate thoroughly. As yet though, with the evidence out there, I would say it is too early to make a judgement. I'll wait to see the next phase of the report before I make any call.
Wearing my pilot's cap for this one...
The plane stalled into the ground, which tells you all you need to know about their approach speed under limited engine power. If the pilots had reduced flap, the plane would have stalled at height and crashed in a much less controlled way, short of the peri track, and lives would have been lost.
The pilots ran out of options, air speed and altitude, in that exact order. Unless I completely misunderstand the many authoritative accounts of the incident, there is almost nothing they could have done.
The only slight option I can see would have been to retract the gear (assuming they still had hydraulics, which should have been the case). Done soon enough, that would have reduced drag, and _maybe_ got the plane to the start of the paved runway. The flight crew clearly considered that a gear-down mushy stall on grass was a less risky option than a gear-up belly landing on paved. I don't drive 777s, so I'm not qualified to make that assessment, but it seems to me that events proved that choice: they achieved the best possible outcome - a broken leg being the worst harm - from a very bad situation.
Retracting the flaps was discussed ad-nauseum over on PPrune. General consensus was that the extra drag induced by the change of configuration of the plane would have cost more than any possible saving. In any event as Neil Barnes points out as flaps are reduced the stall speed increases, and with insufficient thrust to maintain speed a catastrophic stall was much more likely.
There is generally no simulator training for complete engine loss on approach. Remember they had just seconds before reaching the ground. With more time e.g. if engines failed during the cruise, then it is possible to calculate best glide settings for the plane, but not when you are already below the height of Canary Wharf...
The pilots of the plane did an utterly exceptional job in handling the plane.
I base this on pprune
The flaps *were* reduced by the captain as the pilot in charge (the FO) continued other (SOP) rectification procedures.The flaps change probably changed a disaster into a bumpy landing
I understand that flaps control does not necessarily interrupt the Auto Pilot - which is programmed to maintain a specific flight parameter, such as descent angle, descent rate, or - in flight - a change in altitude or heading.
As an (ex) glider pilot, I can confirm that pulling in the flaps would increase stalling speed. With no engines, the only way to increase speed is to put the nose down, trading height for speed, but this must be done before retracting flaps, otherwise, as previously stated, the plane will drop out of the sky.
There was virtually no time to consider this. There was no height to trade. The mushy stall at zero feet previously referred to was an excellent option. Had the engines cut even 5 second earlier, we'd be looking at a very different scenario.
Careful reading of the published report will reveal the exact cause is still unknown. As long as that is the case nothing and/or noone should be excluded as the potential root cause. Also interesting: the autopilot was disconnected very late in the approach. So it seems manual pilot input was quite a bit less as what was being suggested immediately after the crash. It seems the autopilot (programmers) also deserve a round of applause!
My understanding is that cavitation damage is caused by small short-lived bubbles of (near) vacuum collapsing. Microscopic air bubbles may act as nucleation sites, but so do other impurities and microscopic features on solid surfaces in contact with the liquid, so I doubt that air in the fuel is the main problem. Such bubbles can form where there is very rapid and turbulent flow, such as inside a pump. Insufficient inlet pressure can increase the likelyhood of this happening. The space shuttle main engines use low pressure pumps to feed the high pressure pumps, in both the fuel and oxidiser lines, to allow the high pressure pump to operate without cavitation.
Such bubbles form in areas of liquid which are under negative pressure, such as behind an impeller blade. They contain a small amount of vapour of the liquid, and reach very high temperatures and pressures as they collapse to a very small volume.
Maybe the cavitation damage is a symptom of insufficient fuel flow from the tank to the pump. This could also restrict engine power. Did somebody say there was some debris inside the fuel tanks?
1) Well done cabin crew for landing a plane with no power at low altitude. Exceptional job.
2) They throttled up........and nothing happened. Confirmed.
So why?
Still had fuel, and pumps were capable of running.
Clue #1 : Engines didnt stop, they operated at reduced speed. ( like if your throttle cable snaps on your car )
ie, Pumps and fuel still running.
Answer
a ) The avionics decided to override the request for more fuel.
With so many failsafes, and duel circuitry on avionics designed for continued operation with single failure, the chances of a multiple failure on the same system are rare. ( not impossible, just rare )
What is more likely is the avionics overrode the request for more fuel due to something......... the question is what?
Now when I say Avionics, I mean the basic electronics, not the computer systems and the software.
The question I pose is - what in the Avionics setup has the ability to perfrom this override?
With Fly by wire, the computer will override certain conditions to compensate for bad flying ( ie like on the airbus that will not allow the tail to be scraped on takeoff ).
However the bulk of these inputs are collected on the black box. So with nothing in the data recorder showing misreading speed etc, it looks like a minor system caused it.
Any ideas?
BA is somewhat unusual in commercial aviation in that in addition to t he CVR and FDR (aka black boxes) they also monitor/record hundreds of other systems for analysis offline for engineering purposes. This is the Quick access recorders mentioned in some posts but strangely not talked about much by the main stream media (too complicated to understand?)
It was these records that helped get the Trident to land "hands off" in thick fog way way back.
Not all airlines use them (not required to do so)
re: post by Ian Peters
"..That the 777 is prone to issues where the flaps do not always work as they should and these early landings are not uncommon to correct this issue."
What utter poppycock ! Clearly you don't fly aircraft Ian !
ALL aircraft (with the exception of the Space Shuttle which doesn't have them) can perform flapless landings with ease. The pilots practice this regularly; it's not a big deal.
They do NOT 'land short' because of the non-availability of flaps !!!!
Let the AAIB do their work and we'll find out soon enough.
vacuum bubbles in fuel pump, and engines not delivering thrust, plus some debris in fuel tank - sounds to me like the pumps were unable to deliver enough fuel to the engines. If Boeing says the pumps were OK, perhaps the debris in the tanks restricted fuel intake and this was the root cause of the problem. It is strange though that it was a problem on both intakes and at the same time. If the tanks were close to empty (so more debris per liter of fuel), and if there is some kind of mesh or filter to stop the debris getting into the pump, the end of the flight would see the most debris blocking the filter....
disclaimer - I don't know much about fuel systems on airplanes, just a guess on my part...
Reading all the words in the latest report - the fuel metering unit valves, that is the mechanical bits, were fully open so there was no avionic system overriding the command for more fuel as far as I can see.
This is a conundrum for all at the moment and I hope the AAIB are close to finding the real answer for two reasons 1) I was involved with the FADEC software and 2) I am flying on holiday on a 777 fairly soon and the rest of the family are pretty nervous at the best of times...
I watch a lot of engineering disaster videos (especially relating to air disasters), and everything the AAIB is doing is consistent with all post-accident/disaster investigations. I fully expected that this accident would take a while to solve, and a good thing too. I don't want them to rush to a conclusion without fully exploring every avenue.
And regarding the comment about the manufacturer testing the pumps; Boeing did not produce the fuel pumps. And not only did they not produce it, they also likely didn't even _design_ it. Boeing--and for that matter, Airbus--simply provide performance requirements for most parts to manufacturers, and pick the one that best matches their requirements for the cost.
The fact that the investigation has so far shown that it doesn't initially look to be a problem with the flight crew, engines, or EEC means this could go on for quite a while.
I've seen all the comments here, and essentially there was nothing wrong with the aircraft systems, everything worked as it should have. When the crew opened the thrust levers to full power the fuel metering unit scheduled the fuel nozzles fully open. The only thing that was missing was sufficient fuel flow to reach the commanded thrust. There was 10.5 tonnes of fuel on board, it was not water-contaminated and had a freezing point of -57C, lowest recorded fuel temperature was -34C.
The FOD in the tanks was small and probably can't account for the inability of the fuel system to get sufficient fuel to the engines, finding such small items of debris is not unusual as it's difficult to account for every source of contamination during production and maintenance (although they should try harder, count everything issued and then count back in as the work is done maybe). The parts found were spotted by opening up the access panels, the mag chip detectors are there to catch and indicate the presence of metal particles to guard against bearing failures, they don't detect plastic, cloth or paper.
For there to be cavitation damage in the High Pressure fuel pumps (and it's not certain it happened on this flight) at some time there must have been a flow restriction between the wing tanks and the engine (fuel passes through two oil/fuel heat exchangers between pylon fuel pipe and HP pump inlet) that allowed either low pressure or aeration to occur, this might have been sufficient to prevent commanded thrust being developed. The key statement is that the pumps were deemed capable of maintaining full fuel flow, it does *not* say that they *did* maintain full flow. That's the smoking gun. Now what has to be found is how the flow restriction occurred.
The aircraft was flying through -76C air en route for some time, the fuel in the tanks never went anywhere near the measured -57C fuel freezing point; just maybe the pipes through the pylons to the engine did. If fuel partially froze in the pipes, maybe enough fuel was reaching the engine for cruise and descent, then as the warmer air began to warm the frozen stuff up it could have blocked the HP pump inlet enough to cause the damage seen and limit the fuel flow needed for more than a low thrust setting.
After the crash the frozen fuel would have melted and hence not be visible or detectable. Somehow AAIB and Boeing will have to identify how this happened (if it did), it may be that they never will, my comments above are only my theory.
The fire picture because with ruptured tanks, spar valves that couldn't close and ruptured oxygen cylinders the people on board are lucky there was no fire, if they had reached the runway and created a lot of sparks maybe there would have been. They hit the only piece of handy, glutinous grass for miles around, anywhere else and this would have been a disaster rather than an indident.
which is highly unlikely from what we know, it was the plane wot caused it.
In practical terms if my Boeing has technical issues do I to pay for an overpriced ticket with probably the finest crews on the planet, or a cheap ticket from a "budget" airline with presumably budget training and budget crews?
Suddenly losing all your luggage seems worth it. Respect to the crew on that flight and BA for putting them there in the first place. Amazing how worrying about that extra £100 (well technically 50p after all the hidden charges magically appear on the final booking page) suddenly seems so trivial.........
Flaps do not allow the aircraft to fly more slowly. This is a miscoception believed by many that do not understand flight.
Flaps create more lift. As a side effect, they increase the drag of the aircraft, causing it to slow down and a change in pitch when they are deployed.
Flaps are used to control the rate of descent. The aircraft needs to slow down during landing, which means the throttle levers need to be reduced ( the thrust needs to be reduced) to achieve a reduction in speed. The lift force generated by the wings is a a function of the speed of the airflow over them. If the aircraft slows down, then the lift generated by the wings decreases and the rate of descent increases, the aircraft loses height too rapidly.
To compensate, the flaps are deployed to increase the lift, and check (control) the rate of descent.
Flaps DO NOT change the stalling speed of the aircraft. The stalling speed remains constant.
Stall of the aircraft can occur under two circumstances: the speed of the airflow over the wing is too slow, ( ie. the aircraft isn't flying fast enough), or the angle of incidence of the wing relative to the direction of forward movement( angle of attack ) is too high, resulting in the airflow breaking up over the wing and a resultant loss of lift.
Flaps do nothing to affect this.
<<Flaps DO NOT change the stalling speed of the aircraft. The stalling speed remains constant>>
Not true. Have a look at an aircraft ASI (airspeed indicator). The white arc shows the speed range with flaps applied. It goes from the flap stall speed to the maximum allowed flap deployment speed. The main scale shows the stall speed without flaps. The stall speed with flaps is definitely lower. Not much lower, but it makes all the difference when landing.
The amount of lift generated by a wing is determined by airspeed, angle of attack and wing geometry. A stall occurs when the angle of attack (the angle between the wing and the airflow) exceeds the critical angle (usually around 15 deg for normal wing profiles). At this point, the airflow becomes turbulent, drag increased and lift is reduced (by about 60-70%).
When you fly slowly, you pull the nose up, so increasing the angle of attack in order to get the same lift at reduced speed. The (rather misleading) concept of "stall speed" is the point where you increase the angle of attack too far. The aircraft is perfectly capable of flying at that speed, it just can't generate enough lift to support its weight. So it's either going to descend, or (if you pull the stick back too far) it will stall - and then descend rather rapidly!
Flaps change the wing geometry, from one optimised for high speed to one optimised for low speed. Since you never get something for nothing, this new profile introduces more drag, so you need more engine thrust (or a greater rate of descent) to maintain the same speed. Very useful if you need to slow down for landing, but not so helpful if you've lost your engine.
Most aircraft will glide further without flaps than with flaps, because the "clean" wing is more efficient. So, if you lose power at altitude then you wouldn't normally deploy flaps until you were close to the ground, where they make the touchdown slower and more controllable. However, if you had an engine failure when the flaps were already out and you were low, then it would normally be a very bad idea to retract them. Although the long-term effect would be a more efficient glide, the immediate effect would be a sudden pitch change and a loss of altitude, amounting to several hundred feet. If you've only got 700 feet then that's the last thing you want.
By the way, I just completed my AFR (aircraft flight review - or periodic checkup) last week, albeit in a much smaller machine. It included a simulated engine failure. I passed, so I must have got something right!
Clive Harris
Given that the fuselage of the plane seems undamaged, will this plane be fitted with new wings and flown again, or scrapped, when the investigation is completed.
Or is the damage bad enough that it's not financially viable to repair?
Just interested in knowing what the usual procedure is for this sort of thing, Although I suppose most air crashes don't leave behind anything remotely air worthy.
It is almost certain that the plane won't fly again.
Aircraft bodies are designed to cope with the stress of taking off, flying, and normal landing. A mushy landing resulting in undercarriage colapse and wing damage is likely to stress elemets of the plane beyond their tolerances, leading to structural defects in many of the strength elements of the airframe. Damage will have been done to the wing fabric and possibly the roots, undercarriage, engine pods, and body (where it touched the ground).
If it were to fly again, there would need to be tests performed on all of the major structural elements to prove that they were not compromised. This would probably cost more than replacing the plane. In addition, it would need new wings and engines (which is possible, but expensive). For an example, see how much it has cost to return Vulcan XH558 to flying condition, and this was a much more simple aircraft.
In addition, the investigation will be probing all of the wiring systems and control systems, so these would need to be re-worked. If you have ever seen how much wire there is in a modern plane, and at what point in the construction it is put in, you would realise that it cannot be replaced. Hell, car companies don't like replacing the wiring loom in a car!
It is likely that most of the relevent parts of the plane will be kept until some time after the investigation is closed, and if they are ever released, re-usable parts will enter the spares pool of BA or Boeing (after being bought back from the insurers). The remaining airframe will probably become a engineering, fire, or evacuation training mule.
BA will not suffer, as the planes are insured.
An obvious probable cause, is as previously stated, fuel supply lines icing up, then the ice freeing up and blocking pipes or pumps, as the plane was landing in the warmer air at lower speeds. Although the likleyhood of this happening to both engines simultaneously is fairly low, it would explain the 7 second gap between the two engine problems, and also the fact that the engines continued to run, but failed to increase thrust.
Imagine the scenario, ice forms in the fuel supply lines (around the outside obviously) during flight cruise, enough to reduce flow but not below the levels needed for normal flight and descent, then we descend and the bonds between the pipe and the frozen fuel start to melt (weaken), then just before landing we call for large amounts of power (high fuel flow), which dislodges the 'ice', which is sucked into a solid mass, blocking or significantly restricting flow. No power, for a short time, just long enough to cause the accident, shortly afterwards the evidence just melts away.
When you have eliminated the impossible, whatever is left, however improbable is your answer.
The Penguin - just a jumbo jet designed by a committee
Do flaps change stalling speed?
More recent research I've undertaken indicates that it depends on the aircraft.
I would suggest that the effect on stalling speed is a side-effect, that this is not intended.
The discussion about stalling speed is a red-herring, because you should not be flying that close to the stalling speed for it to be an issue.
If you are flying that close to the stalling speed then your placing the aircraft and passengers at risk, should the aircraft be suddenly buffeted by wind, causing the aircraft speed to fall below stall speed.
In gliders, it is common to fly at 1 or 2 knots above stall speed to minimise the energy lost, to keep the glider up in the air for longer.
However, on approach to landing, the glider pilot will accelerate the glider to move away from the the stalling speed in case of any issues with wind on landing.
In a civilian airliner, they're flying too fast for landing and so tend to decelerate, but shouldn''t be flying at the stalling speed.
Major factors which affect stall speed are: Weight of the aircraft, - the fuel load on landing is taken into consideration and the landing speed will be higher with a heavier load, and whether the wings are dry or wet, if wet, the stall speed increases, in addition to flaps .
When an aircraft is gliding, the issue is to mimimise the energy loss, because it is when the energy is zero, that the aircraft will definitely be back on the ground and this requires flying at an optimal speed to minimise the losses, (and putting the aircraft into a minimal drag configuration) as this is how the energy is lost.
>Flaps change the wing geometry, from one optimised for high speed to one optimised for low speed
I don't fully agree with this statement, that implies that an aircraft that needs to fly at high speed does not have flaps or they're retracted, and would deploy them at low speed in order to fly slowly.
Variable geometry wings , such as that used by the Tornado and F1-11 serve that purpose, not flaps. I know you know what you're talking about, just the statement isn't literally correct.
Comments from RotaCyclic
"...Flaps DO NOT change the stalling speed of the aircraft. The stalling speed remains constant."
Wrong !! Flaps DO change the stall speed of an aircraft. What does NOT change is the angle of attack at which the wing of the aircraft will stall. Deploying flaps increases the angle of attack of the wing (as measured by the chord that runs from the leading edge to the trailing edge of the wing).
So with full flaps you can fly slower.....but you can't exceed the stall angle of attack of the wing. Reducing flaps means you can reduce parasitic drag, which in turn means you conserve energy more easily.
Brian Morrisons comments seem about the only plausible explanation I have seen. If the pumps show sign of cavitation then at some point they have been operated with insufficient fuel, I suspect that if the aircraft had operated with one or more tanks at low level in the past, the main obvious reason for insufficient fuel, it would have been an event that the flight crew would have logged and would have been picked by the AAIB. Although nothing is guarenteed on this planet if the Quick Access and FDR show systems demanding full throttle with correct fuel valve and pump commands then I would believe the systems tried do do what was requested - generally these data are monitored tale backs i.e. the actual commands provided to LRUs and not "internal" software variables i.e. for example the FDR wouldn't log what the FADEC said it was commanding but would log the "real" used command being sent. In anycase although FADECs are almost unique in not being dissimilar (I await the emails on this subject....) the idea that the same common mode failure could occur on different engines at different times with almost certainly some distribution in engine and throttle states is pretty far fetched (not impossible but certainly not as plausible as Brians explanation). But in anycase I am sure it will prove to be something else ...