How Did MH370’s Flaperon Come Off?

In my last post, I reviewed Malaysia’s analysis of the MH370 debris its investigators have gathered. Not included in that study was the flaperon found on Réunion, as it is being held by the French. So today I’d like to look at what the damage patterns seen on the flaperon suggest about the crash, based on the work done by IG member Tom Kenyon and by a reader of this blog, @HB.

On February 3 of this year, Kenyon released an updated version of his report “MH370 Flaperon Failure Analysis” in which he gives an overview of the flaperon’s structure and how it was damaged. He notes that of the six main structural attachment points of the aircraft, the two biggest and most significant are the flaperon hinges (pictured above). They snapped in the middle:

The lesser attachment points failed in a similar way. That is to say, they did not rip away the flaperon structure to which they were attached.

Kenyon observes:

The location of the failure points of Flaperon hinges is consistent with a large singular lateral force or repetitive lateral (or torsional) movement of the hinges in the inboard/outboard direction. If Flaperon was separated from the Flaperon hinge with forces in forward/aft direction or by applying forces to the Flaperon in the extreme rotated up/down direction (beyond structural stops) then deformation of the Flaperon structure due to such forces would be evident. Significant and permanent deformation of the Flaperon structure does not appear to be present in photographs of the Flaperon.

Recall that two scenarios have been proposed for the flaperon coming off 9M-MRO: either the plane hit the water, or it came off as the result of flutter in a high-speed dive. Neither event could reasonably be expect to produce a primarily lateral (that is side-to-side) force on the flaperon of the kind Kenyon describes.

To raise the level of perplexity, Kenyon points out in other crashes involving 777s, failures didn’t occur at the hinges; rather, the hinges remained intact and the material to which they were attached broke. That is to say, hinges are stronger than the flaperon proper. Here’s an example from MH17:

Kenyon concludes that:

No significant evidence of secondary structural damage excludes a massive trailing edge strike and leads the author to conclude that the Flaperon separated from MH370 while in the air and did not separate from the wing due to striking water or land.

In other words, since the damage isn’t consistent with a crash into the sea, we can deduce that flaperon must have come off in the air. The only conceivable cause would be high-speed flutter. However, on closer inspection the evidence seems to rule out flutter, as well.

The reader who goes by the handle @HB is an expert in quantitative risk assessment in the transportation industry and has extensive experience with composite materials. In a comment to the last post, he observed:

For the flaperon… the lift/drag load is normally passed on the honeycomb panel over the exposed surface area then the primary stucture of the component (the aluminum frame of the flaperon) then the hinges then the primary aluminum stucture of the wing.

In a nut shell, those panels are not designed to sustain any in-plane loads either compressive or tensile. They are just designed to resist bending due to uniform lift load on the surface (top FRP layer in tension and bottom FRP layer in compression, the honeycomb is basically maintaining the distance between the layers without much strength).Those panels can arguably take a little bit of shear load due to drag forces on the top skin (top and bottom forces in opposite direction) but not much due to limits in the honeycomb strength.

The second thing to consider are the GRP properties. The GRP is very tough in tensile mode, much stronger than Steel. In compression mode, it buckles easily and only the honeycomb is preventing this. This is by far the weakest failure mode. If it fails, you will see fibres pulled out link strings on a rope failing under tension. For the skin under compression, you will see sign of compression on the honeycomb but the fibres will have to be pulled out as well. Also, a perfect manufacturing does not exist, there are always delamination (small bonding defects) between the honeycomb and the GRP skin to weaken further the compressive strength.

The hinges are usually much stronger as all the load is passing through them (analogy door hinges).

So you could imagine, if there is a large impact the hinges are expected to fail last. The part of the skin that will buckle is expected to fail first.

@HB here is agreeing with Kenyon: it is baffling that the flaperon came off the plane due to failure within the hinges. But @HB goes further, arguing that this type of damage is inconsistent with flutter:

I would tend to agree that the hinges have been subject to cyclic fatigue… the hinges appear to have been subject to cyclic lateral forces which are not expected in any accidental circumstances (take door hinges, for instance, and imagine the hinge fail after 50 times someone is trying to burst through – you can try at home but it is very unlikely to happen). This of course requires a closer look by experts to double confirm. I cannot think myself of any possible lateral force on this part in the first place but a lateral force that will fail the hinges and not the skin which is weaker is very hard to explain. Try with a wooden door and tell me if you manage.

In a followup comment, @HB observes that even under very strong oscillation the flaperon should not be expected to disintegrate. “If hydraulic power is on, fluttering is unlikely to cause any disintegration. If off, fluttering forces are up and down and the hinges are free to move. Lateral forces, I think, would be small in comparison with the vertical forces and not be strong enough to cause fatigue on the hinges.”

Kenyon concludes his report with a list of six questions and issues generated from his analysis. While all are worthwhile, one stands out to me as particularly urgent:

• Why are the official investigators silent on releasing preliminary reports on their Flaperon analysis? Why would France’s Direction générale de l’armement / Techniques aéronautiques (DGA) release photographic data to ATSB and yet chose not to make Flaperon Analysis findings public after such a long period of elapsed time?

To sum up, a close examination of the flaperon’s breakage points does not yield any comprehensible explanation for how it came off the plane, commensurate with a terminal plunge into the southern Indian Ocean.

This is baffling but unsurprising. Every time we look at the debris data carefully, we find that it contradicts expectations. The barnacle distribution doesn’t match the flotation tests. The barnacle paleothermetry doesn’t match the drift modelling. The failure analysis doesn’t match the BFO data. And on and on.

Something is seriously amiss.

296 thoughts on “How Did MH370’s Flaperon Come Off?”

  1. @HB: “The other question is whether the optimum AoA can be achieved or is likely in an uncontrolled situation may need to be addressed.”

    It has been addressed. The airplane will only achieve the AoA for CLmax with pilot control inputs. Without pilot or autopilot control inputs the airplane will stay at approximately the same AoA it had when control was lost. If the airplane does not bank, it will descend in a phugoid motion at constant AoA while rate of descent, airspeed, and loadfactor exhibit slow cyclic variations about constant mean values.

    If the airplane rolls to high bank angles, it will still descend at the same AoA, but rate of descent, airspeed and loadfactor will progressively increase. In the simulations that I have seen high values of bank angle and load factor were only achieve at the end of the descent, i.e. at low altitude.

  2. @Gysbreght,
    This is interesting as this information is a good starting point for us. The members of this forum are interested to determine the load profile on the wing components for a number of scenarios – high G due to high speed being one of them. From that data, a structural analysis could be done to evaluate the impact on, for instance, the control surface structure and overall the primary structure as well as the hinges.
    One more question, if you are aware of a good reference that can point our the load distribution on the wing profile (high speed, low speed) (which areas with positive pressures and which areas with negative pressures) could also be useful.

  3. @HB: Chordwise pressure distributions for a typical airfoil section are shown

    here (Click on the word “here”).

    Keep in mind that the airfoil sections used in the B777 wing design are different, are variable across the wing span, and the wing is swept back about 30°.

  4. @CliffG:
    “New debris from MH370 found in Seychelles, possibly from the engine cowling…”

    Good find. Don’t immediately recognise either piece as part of a modern airliner, although the brown piece could be part of the APU cover – wonder if the B777 used a variant of the compact X-jet to power this component? Not sure what the black carbon fibre piece could be. Perhaps an engineer from Boeing or a service company would be able to identify the parts.

  5. @all
    Per Victor, the new debris from MH370 found in Seychelles has apparently been ruled negative by Malaysia DCA – not from MH370/B777.

  6. @David,

    Certainly clues can be found from analyzing data gathered from previous aircraft incidents.

    However, when it comes to specifically understanding how the MH370 Flaperon primary connections failed, I feel the data used for analysis should primarily be focused on B777 Flaperons.

    The overall B777 Flaperon design, individual component dimensions, structure, materials, thicknesses, normal/abnormal mechanical forces, control surface functionality, and operability play such strong roles as to set of stresses and strains applied to the B777 Flaperon’s primary connections.

    Cautiousness enters when I attempt to compare the B777 Flaperon potential failure scenarios to examples of other aircraft control surface parts (diminished interest when non-Boeing, Non-B777, or non-Flaperon.)

  7. @Kenyon. Fair enough. One thought for you though is that these two aircraft would be certified to the same flutter clearance criteria, including the flaperons.

  8. Some posters on this forum seem to equate flutter failure to fatigue failure. I think that perception is wrong. A structure fails when subjected to loads that exceed its strength. Fatigue failure occurs when the material has been weakened by prior exposure to stresses. When a component fails due to flutter the process leading to overload is quite rapid and the number of load cycles much to small to cause fatigue damage to the material. The main difference between flutter failure and static overload is the greater role of inertial loads in the case of flutter.

  9. @Kenyon. Two delayed thoughts. The flaperons’ flutter clearance would be with no hydraulics (to cover the left) and part (one actuator).

    Boeing would have established that the flaperon’s flutter onset conditions are outside the 777’s “analytical clearance” (as in Silk Air NTSC report Fig 15). They would have an idea from the analysis what that speed would be. They would have a like idea from simulations as to the speeds MH370 might have reached. They could compare the two.

    Also I would expect them to have formed the view that if the aircaft exceeded its flutter boundary it most likely had a pilot.

    Likewise I do not think the ATSB, would be as resolute as it is in its assumption that there was no pilot if the French had advised there were signs of flutter or fatigue in the flaperon attachments, or if Boeing had indicated the aircraft may well have exceeded its flutter boundary.

    The indication is are therefore that there was no flutter evident or likely.

    One reservation I have about Fig 15 is that it makes no mention of load factor, possibly because in that instance it would not have been high.

    In V-n diagrams for these aircraft, were there a boundary due to flutter which was ‘n’ (load factor) dependent that could alter my 777 assessment. In other words if flutter could be ‘brought on’ by ‘g’, a pitch up could cause it in the same way a pilot could via speed.

    However I do not see this affecting the outcome of Boeing casting an eye over those simulation analyses which are outside the 777 envelope and simulation database, as above, except that besides aircraft speed, they would need to include’g’.

    @Gysbreght. You might care to comment on this, the last paragraph in particular.

  10. @David: I’m no specialist on flutter. I’ve never heard of flutter divergence caused by ‘g’.

  11. @Kenyon

    Imo the problem with your analysis is you start off with your Summary of Positive Statements and then try to prove them first with statements that are at least very debatable. You work mainly confirmation-biased all the way. Which has it’s use certainly but by the way you state things it can be (and is imo) deceptive.

    You start of with excluding scenarios like ditching and state deformation of the flaperon structure had to be obvious lateral in such a case (upward forces breaking the hinges and PCU connections excluded).
    A lateral force has to be the cause in your opinion.
    Argumenting the carbon reinforced skin has hardly any resistance to buckling and delaminating in compression for only supported by the honeycomb between them.
    This is just not thrue. The flaperon box-like carbon reinforced structure mounted firmly together on very strong alloy ribs have a tremendous resistance to deformation.
    Those flaperons are used as ailerons at cruise speeds and during landings. They have to have great resistance to deformation of the skin and its inner structure.

    The combined strenght of the flaperon’s skin and structure is far greater then the weakest spot in the hinges or PCU attachments.
    The weakest spot in the hinges were where they sheared off from back to front as did the PCU-attachments.
    Just like the big outboard flap section hinge snapped at its weakest point which was also not from the skin and did not deform the outboard flap structure.

    I can go on with inconsistencies in your analysis. But for now leave it here.
    Hope you show the ability to consider other arguments than your own that haven’t changed since december 2015 it seems.

  12. @KarenK

    Maybe it’s your name that triggers me more than usual. Like a narcissistic Karen I knew, you project the same devaluating crap without ever contributing anything constructive ever.
    I at least have the guts to call someone out and in a way I know I pass ‘gentle borders’ sometimes in emotion. I excusse me for it to @Kenyon in a way but this is serious business.

    I expect he knows he will get confronted with his analysis. And I hope he’s inviting it.
    Till now I’ve seen no sign he is taking up the challenge. Glad he responsed lately.

    I’m eager to debat with him on the details.
    But you are just devaluing all this way. And it seems it is your only way for I never ever read a constructive informative contribution of you.

    And regarding @Victor’s blog or @ Jeff’s I do my own thing with great respect for them over the years. We all make mistakes and have the most strange assumptions everyone in this very complicated case loaded with emotions sometimes. But I will always try to consider the arguments of others seriously within my capacities.

    Over the years I’ve tried to make a serious effort to contribute to the solving of this mystery. Jeff, Victor, anyone else may want to ignore me but I’m almost sure they know my objective is the same as theirs.
    Solving the mystery.

    Your sparse negative comments once in a while don’t attribute to this goal at all imo.

    Come up with arguments. This is not your regular face-book conversation.

  13. @David, thank you for your thoughts.

    To be clear I’m not fixated on ‘flutter’ as an anchor to MH370’s right Flaperon’s fate. There’s just not much relatable data available. However, Asiana Airways 214 does provide an example of what how the B77 Flaperon primary connections did not fail when exposed to significant damaging forces caused by impact.

    Hopefully the ongoing FEA study can shed some more light. The FEA could certainly bust my Rev 3 analysis, support it, open a new chapter, or be inconclusive. As always, I remain open…

  14. I have not contributed in a long time. Two years; I think. I spent a career at Boeing in preliminary design and manufacturing, so just a note on aircraft design for the 777 and similar designs.

    The airplane design process goes through a number of steps. To start: The basic design is decided on; configuration, range, cargo, etc. The airplane is sized for that mission. Basic loads and weights are calculated and sent to the various design groups. The actual design is a cyclical process where the strength of each component is sized for the current design loads. Loads are recalculated based the current actual design and the weight and strength of each component. These new loads are then again sent to the various design groups. The cycle repeats as the weight and strength of the design changes. The cycle stops when the weight, range and cargo targets are met or adjusted for business and other reasons. Each component is designed for its mission and the associated loads and conditions. In most cases; operation outside the design envelop is not included.

    So, parts on the airplane are designed to meet their flight envelop load and condition requirements plus a safety factor defined for that part. Primary structure must meet a specific safety factor. Secondary structure another safety factor based on its criticality to the mission. Reason: There is some redundancy in secondary structure. e.g. Left and right aileron/elevator. Other structures have their own factors to meet. E.g. Interior structure must meet a 9 G forward load requirement in a crash.

    Crash loads are not determined except for interior structure and components that support the interior plus parts that need to fail during a crash in a specific way. E.g. Engine strut pins fail before other structure to avoid a fuel tank failure; also landing gear structure for the same reason; all parts attached to the fuel tanks are checked to reduce the possibility of a fuel tank failure.

    The Flaperons must meet their design loads. They resist deformation and design loads to the point where they do not exceed distortion limits set by the design. The strength of the Flaperons is not far greater than their attachment hinges. The hinges are designed with the same load requirements. The hinges are attached to the rear wing spar which is part of the fuel tank; so that could be a weak link by design.

    It is true that designs fail at their weakest point. But that point of failure is determined by the loads and type of loads applied during failure. For a specific design, failures can occur at a variety of points based on the loads applied.

    In a crash; loads can come from all direction. I have seen crashes where the wings spin around and parts fail from a load from the rear or from above.

  15. @Ken
    Thanks for the insight.
    As a matter of interest is the design considerantion for natural frequency of parts and components. Is this considered as part of the Design Dive Speed case?
    The other point of interest is the design margin on the wing components. Literature suggest UTS 1.5 times design load of 2.5g as witnessed by wing load test. Is that understanding correct?
    Interesting comment for the engines, so if my understanding is correct the crashworthness principle here is to detach the engine in worst case to protect the fuel tank in case of crash. So for instance in case of ditch it is reasonable to assume a possibility of engine mounts shear off?
    For hinges and control surfaces, i agree that for design consideration it would be for the intended function and no allowance for impact then depending on the direction and force it will be done to the strength of the materials. Litterature suggests B777 composite strength problems after impact leading to undetected defects and weaknesses where the issue as been fixed for B787 later on.

  16. @Ken Godwin. Thanks for that.
    @Kenyon. I appreciate your preference now is to await what the FEA says. I am bound though to raise one correction and some further points which might be of interest to others.

    Most recently I implied that the ATSB would not retain its confidence in its assumption there was no pilot if from simulations,”…..Boeing had indicated the aircraft may well have exceeded its flutter boundary”.

    This is a non-sequitur so I wihtdraw it. Since the simulations assume no pilot, if Boeing had come across evidence that the aircraft had exceeded its flutter boundary during them that would have reinforced the no-pilot assumption, not the reverse.

    On fatigue, just going from the Silk Air report Fig 15 the highest frequency for, “antisymmetical” flutter (for an unstated component) was 22hz. Low cycle fatigue is that beneath about 10^4 cycles. If we say 10^2, at 22Hz that would take 4-5 seconds for a fatigue failure, to give an idea.

    My sense is that it is unlikely that flaperon flutter frequency, with its trailing edge attached or not, would be above this “buzz” sort of range, assuming that to be in primary mode, rotating about its hinges. If not, at a high descent speed in the dense low altitude air there might be insufficient time unless in a non-primary mode.

    Which doesn’t make it impossible though my earlier disagreement with your fatigue-of-all-attachments remains: the failure of a hinge, the most likely first failure, would result in immediate overload of the others.

    Also my analysis earlier suggested that a fatigue failure of the trailing edge would result in an immediate flaperon extension and attachment overstress, separation therefore being from a non-housed position. To be consistent with the ATSB’s finding of separation-from-the-housed position the trailing edge separation most likely would not precede the flaperon’s.

    Finally I include a point about flaperon fatigue sensitivity for general interest and another indicating a separation would not be beyond the foreseeable. First, the AMM notes at p67 of 27-11-00 that at on the ground at <85 knots, the ACEs put the flaperon PCUs in bypass; “This decreases the flaperon actuator fatigue cycles caused by the engine exhaust on the flaperons”. This makes it sound that the actuators could be ‘lifed’ to overhaul.

    Second, at p44 it notes that the flaperon hydraulic tubing has check valves to help prevent loss of hydraulic fluid, “caused by the loss of a flaperon”. Of course this does not mean that other components do not incorporate hydraulic fuses also or that either point is directly relevant to MH370.

    I intend to pursuing this as best I can in part because of its relevance to the right outer flap failure and the relevance of that to final flight. I am not entirely confident that the FEA report will resolve it or whether or when we will be apprised of that.

  17. @Gybreght. “I’ve never heard of flutter divergence caused by ‘g’.”

    What prompted my interest was the below at 7.1.4 which says, “Lifting surface flutter is more likely to occur at high dynamic pressure and at high subsonic and transonic Mach numbers”.

    However p148, para 5.7.2 of the below appears to contradict that: “At high altitude VD may be limited by high speed flutter”.

    Adding to the possibility that there can be an effect, para of the above FAA reference says, “Sufficient test conditions should be performed to demonstrate aeroelastic stability throughout the entire flight envelope for the selected configurations”. While not specifically addressing flutter it does include it.

    Incidentally I have encountered non-divergent flutter though that was due to slack in controls. Nominally catered for at 7.1.4 (testing should include, “expected levels of in-service freeplay”) of the FAA ref., in fact this was beyond “expected” levels. I think we can assume that would not apply here.

    All in all I can draw no conclusions from all this as to whether a pitch up could induce flutter, particularly if that exceeded the flight envelope load factor.

    The end of the road for comment on this aspect other than by Boeing I think.

  18. @David
    Was that statement actually made by ATSB ? Or was it just a comment on exceeding flight evelope?
    ” ”…..Boeing had indicated the aircraft may well have exceeded its flutter boundary”. Do we know what Boeing actually said? This is also very ambiguous statement as whether it was an opinion or the result of simulations or tests

  19. @HB. I do not know which ATSB statement you refer to.
    On the quote you might have missed the ‘if’ in my 4th line?

    The original in full was, “Also I would expect them to have formed the view that if the aircaft exceeded its flutter boundary it most likely had a pilot.”

    @Ken Gooodwin. I return the missing ‘o’.

    Your contribution earlier was valuable and I hope you will continue.

  20. @HB. Correction. In fact the original quote was, “Likewise I do not think the ATSB, would be as resolute as it is in its assumption that there was no pilot if the French had advised there were signs of flutter or fatigue in the flaperon attachments, or if Boeing had indicated the aircraft may well have exceeded its flutter boundary.”

  21. @Ken Goodwin

    Yes Ken, that was very useful. Insights into Boeing’s design philosophy are fascinating, and potentially very valuable in the quest for answers.

  22. @David, Gysbreght, Ken

    I am trying to summarise the issue and data, as I understand, here to check if mid air separation has occured or not.

    My understanding of the B777 flight envelope (please correct me if I am wrong) based on the data I have seen is:

    * Design Dive Speed = M = 1.04
    * Flutter margin (assumed – tbc) = M = 1.2 x 1.04 = 1.248
    * Load Factor (Positive) = 2.5
    * Ultimate Design Load for Wing = 2.5 x 1.5 = 3.75 g
    * Test to Failure Load for Wing = 2.5 x 1.54 = 3.85 g

    Now the issue if whether either the load factor was exceeded, the Design Dive Speed limit was exceeded/ Flutter Limit Speed, or Both or None.

    Flutter: All the simulation data I have seen show Maximum Mach in uncontrolled dive to be M less than 1, actually 0.8 (I ignore FSX which is computer game). Not sure what the Boeing simulator actually predicts (i hope the final report will tell). This on its own, if maximum dive speed is less than M=1 should rule out flutter as a failure mode on the wing component regardless of control inputs or not.

    High g overload: If we take the Ultimate Design Load for Wing of 3.75g as the starting point. Whether overload occurred or not depends on the lift coefficient calculation for assuming no control input. Appreciate if someone can calculate the Lift component on the basis that AoA is constant and assuming worst case of speed at sea level and high speed like in simulations M=0.8 ish.
    Then if overload is possible, the next step would be to determine what would be the impact on the primary structure, then the flapperon.

  23. @HB: “* Design Dive Speed = M = 1.04”

    I’m not sure that is correct. Please explain.

  24. David posted July 16, 2017 at 2:55 AM:
    “@Gybreght. “I’ve never heard of flutter divergence caused by ‘g’.”

    What prompted my interest was [AC 25.629-1B] 7.1.4 which says, “Lifting surface flutter is more likely to occur at high dynamic pressure and at high subsonic and transonic Mach numbers”.

    That doesn’t say anything about ‘g’.

    High dynamic pressure AND high Mach exist at the crossover altitude where CAS limit equals Mach limit, i.e. Vmo/Mmo or Vd/Md.

  25. I am asking confirmation. This is on the basis of Vd * 0.8 = design cruise speed specification. Since at cruise altitude B777 has the design cruise speed at M = 0:84 this implies design dive speed M = 1.04. I don t know whether this holds at different altitude levels.

  26. @Ken Goodwin @Kenyon @David

    Interesting you state the flaperon structure strenght is not far greater then its hinges. But it’s greater and that would be a crucial point in the sequence imo.
    For as you state all things directly attached to the fuel tanks would mean the flaperons attachements beyond its hinges on the rear spar of the wing (which is a side of fuel tanks) would not fail before breaking of the hinges and the hinges would break before the flaperon structure (for this structure is slightly stronger then the hinges).

    Concluding from this (your?) point of view the hinges would be the weakest point.


    Thanks for stating your opinion. And please excusse me if you feel I’ve been to harsh on you. That was not my intention.


    I think it’s interesting you mention the 22Hz flutter frequency and a failure after 4 or 5 sec. on this frequency rotating about its hinges.
    In what kind of situation could this frequency have occured during 4 or 5 seconds?

    Then consider again a ditch-like scenario. The flaperon (and other parts) could easily have hit up to 22 waves a second during 4 or 5 seconds. Causing the breaking of the hinges at their weakest point and imo after the trailing edge broke away which is the weakest point of the flaperon structure.

  27. @HB: Typical margins between Mmo and Md are 0.05 – 0.07 Mach (ref. FAR 25.335(b)).

    For normal operation without failures the airplane must be designed to be flutter-free in the Vd/Md versus altitude envelope enlarged by 15% (not 20%) (ref. FAR 25.629(b1)).

    A smaller “fail-safe” envelope applies for failure conditions (e.g. two engine inoperative). Refer to David’s July 16, 2:55 AM link to FAA AC 25.629-1B

  28. @Ge Rijn, @Rob, @Gysbreght and a few others. Perhaps it is time to take a vacation or perhaps find another ‘hobby’. This ‘analysis’ of parts failures is going nowhere. I continue to return to this forum with the hope that there is something new. There is not. At this point probably the best way to make progress in this case is to grab some official in the Malaysian government, put his head in a vise and keep turning until you begin to hear something useful.

  29. @Shadynuk

    Ha ha! I suggest the former Minister of Defence could be a usefull candidate..

    I feel something new will show up soon.
    It’s a kind of addictive too I admitt.

  30. @HB. As you imply the data on the relevant Boeing simulations (ie compliant with the final BFOs) have not been released: the ATSB has advised that they are proprietary.
    Even if they were released they might be of doubtful use. The ATSB has said that, “In some simulations, the aircraft’s motion was outside the simulation database. The manufacturer advised that data beyond this time should be treated with caution”. Further, “Some simulated scenarios also recorded descent rates that were outside the aircraft’s certified flight envelope”.

    Consequently the data might not be representative enough for analysis and not even Boeing might be able to use the outcomes for that. MAS also conducted simulations (no disclosure there either) but I doubt they would be any more useful. The purpose of the Silk Air 737’s simulations was to find matches of the aircraft’s radar track during its descent, the last paint being at 19,500 ft. As to the Boeing 737 simulator, “In the case of B737-300, the….software has been validated up to Mach 0.87, and extrapolated using computational data up to Mach 0.99”. That might give you an idea of the limits applying in the Boeing 777 simulation, though who knows?
    I notice they were confident in extending the 737 simulation into the extrapolation when looking for matches.

    Whether the 777, pilotless, reached overload due to a pitch up depends on its pitch stability, stick free, in high transonics. Qualitatively all I can do is repeat guess that the wing would fail before the flaperon. I doubt you will get anywhere quantitatively with that without a lot more information.

    While in the 737’s case the aircraft static load limits were not exceeded in their simulations they were with a pilot.

    @Gysbreght. High dynamic pressure, “….. doesn’t say anything about ‘g’”. No, but that started me looking into any effect high ‘g’, for which high dynamic pressure is necessary, would have on vulnerability to flutter, divergence and their severity. As yet I do not have the answer.

    @Ge Rijn. The 22Hz and thence 4-5 seconds for fatigue failure was indicative only. Since there was separation and yet no sign of fatigue, the separations apparently were in overstress. That speaks of divergence, which could be very quick. I am guessing that the Fig 15 “antisymmetric” Silk Air flutter mode might be the elevator twisting dynamically and resonantly feeding into stabiliser flexing and twisting, that in turn increasing elevator flexing etc. Ken Goodwin might be able to comment here. I do not pretend to know much about it and even if I did I would need to shed quite a few years and gain heaps of information from Boeing as to stiffness in flexing and torsion, mass distribution, lift coefficient slopes, areas, Mach effects etc., and probably access to a wind tunnel.

    If the retracted flaperon hit waves or just ‘bounced’ the damage to it would be unrelated to these in-air flutter frequencies, not least because we are talking about quite different components, the 737 stabiliser and the 777 flaperon.

    @Shadynuk. We are constrained by lack of information and various impasses and I agree that we have wrung about as much as can be gained from this topic, although I wish HB well in any continuing endeavours.

  31. @Shadynuk
    Where we are is waiting for the final reports from ATSB/MY which will possibly contain new info, at least new policy directions, which of course we may not agree with. I think we have some inklings that ATSB is piking 35S location and I have some hunches what MY is thinking.

    I do not think anyone (media, etc) can do a post audit without at first hearing the official position.

  32. @Shadynuk

    Yes I agree, it’s the frustrating exercise par excellence. As of now the Malaysian authorities are doing a very good impression of a brick wall (no change there) Not a chink on the horizon. I know it sounds duplicitous to say this here, but there’s some quite interesting stuff going on on the other channel, if you catch my drift.

  33. @Sajid UK

    Look up “Star Dust” for an airliner which crashed in the Andes and was found over 50 years later. There are many such stories where crashed aircraft have been found on land after 20 years or more.

  34. @David

    I also think we’ve come to the end of a line.
    An important line for it generally excluded the cause of flutter to the seperation of the flaperon.

    My intention was to explain a repeated cicle of impacts when hitting waves during several seconds (up to 22Hz) could induce stress-cracks in the hinges and rapid fatique failure.
    Fatique failure always starts with small stress cracks in the weakest points.

    But I agree lets leave it here. Without further details from the ‘officials’ (France, Malaysia, ATSB, Boeing) it has not much use speculating any further.

  35. Are the pins a different material? Could they be subject to earlier corrosion failure after being submerged in salt water?

  36. About the MH370 FLAPERON, I believe it was found almost in one piece because it is located exactly on the trailing edge and straight behind the 2.8mt diameter Rolls Royce TRENT 800 engine which I think at the moment of the water impact created a hollow or a vacuum so the FLAPERON didn’t receive the water impact directly. I think it was retracted when MH370 hit the water.

  37. There 8s no perplexity here. The lateral force is blatantly a consequence the 2014 damage to the plane when it’s wing tip was repaired. Obviously there was unattended lateral fatigue to the flaperon occurring from that prior crash. The plane should not have been flown following such a major runway collision.

  38. Something to consider, what speed curve would the aircraft experience after loss of power, given that it would start with a mach number around 0.875 and would have entered coffin corner fairly rapidly after power loss.

    Would brief (or indeed not so brief) supersonic wing loading be sufficient to cause this kind of structural failure?

  39. Theory
    Right wing flaperon came off mid flight and at same time right side fuselage damage caused catastrophic depressurisation.
    Plane subsequently climbed to 45000 circling to the left as since was now slightly heavier on it’s left hand side.
    Runway collision two year prior that caused complete loss of MH370 plane’s right wing tip which was repaired BUT right flaperon bolt fatigue and right wing/fuselage joint bolt fatigue took many months to become dangerous.

  40. Since this tragedy five airliners have piled into the desert. mountain or sea at around 45% to the near vertical. From what little I have gleaned the flaperons in all cases could be packed into a suitcase….totally mangled beyond recognition.

    From a laymans point of view it looks like a planned landing with flaperon extended, right engine takes a mouthful of water and rips from it’s pylon popping the flaperon like pea from a pod. The leading edge’s paint wasn’t even scratched which rules out any direct impact.

  41. Are there any comments which address the erosion of the flapperon trailing edge?
    This was the main evidence cited by the crash investigator author, which supports the extended flapperon being scoured by water during a ditch landing attempt.
    The previous damage could have stress fatigued the hinges, the engine protected the leading edge until it was ripped away taking the flapperon as surmised by others. That is the only scenario which explains no flutter damage, the pristine leading edge and the eroded trailing edge.

  42. Jeff, I agree that all the observable flaperon evidence is incompatible with separation from the aircraft in flight or by crashing or ditching. I also agree that the only obvious means of snapping the two main hinge brackets that seems plausible to me, is a severe lateral impulse to the flaperon, applied by the inboard end of the outboard flap face against the outboard face of the flaperon.

    This could have occurred when 9M-MRO was involved in a ground collision with a China Eastern Airline A340-600 at Shanghai Pudong International Airport (ZSPD) 09-AUG-2012 16:40 local time.

    A PPRuNe user based in Malaysia ( stated at 03:59 10th Aug 2012, (
    Damage seems bad and may be sitting in PVG for weeks awaiting repair.
    Another PPRuNu user ( stated at 06:30 10th Aug 2012, (
    A photo in the Chinese paper shows the detached B777 wing tip atop the A340 stabilizer.

    Images of the damage:

    Accident Database:

    The BEA listed the incident due to the fact that the other Aircraft involved was Airbus A340-600 Registered B-6050.
    (Refer to Page One – and the bottom of Page Four)
    The essential facts are that the wing tip of MH389 9M-MRO (B777-200) was broken, the elevator of MU583 (A340-600) was damaged.

    Media Report-1:
    “The tip of the wing of the Malaysia Airlines Boeing 777 was broken off and hung on the tail of the China Eastern Airbus 340-600, according to pictures posted by passengers on the Internet.”

    Media Report-2:
    Post #487 of 12th August 2012 – Quoting Shanghai Daily 10th August 2012.
    A TAXIING Malaysia Airlines passenger plane bumped the tail of a China Eastern Airlines plane waiting on the taxiway at Pudong International Airport yesterday afternoon. No one was injured.
    China Eastern flight MU583 from Shanghai to Los Angeles was waiting to take off about 4:30pm when a wing of the Malaysia Airlines plane bumped the left elevator on the tail of the Chinese plane, the China Eastern said later yesterday.
    The top of the wing of the Malaysia Airlines Boeing 777 was broken off and hung on the tail of the China Eastern Airbus 340-600, according to pictures posted by passengers on the Internet. Both planes are wide-body passenger aircraft.
    “It is like a sudden earthquake,” said a passenger surnamed Chai on the China Eastern plane.
    He said the plane was shook strongly when it was bumped and the shaking was more severe in the front of the plane.
    The Shanghai-based airline dispatched another plane for the flight and moved the damaged plane for inspections and repair, China Eastern said. The China Eastern flight was due to take off at 3pm but had yet to take off as of 9pm yesterday.
    The East China Regional Administration of the Civil Aviation Administration of China has sent a team to investigate, an official said. The operation of other flights at the airport was not affected.
    Malaysia Airlines’ flight MH389 was heading to Kuala Lumpur. The company could not be reached for comment.
    A civil aviation expert said the bump is likely to have happened at the beginning of the taxiway, when the Malaysia Airlines plane was turning around. He said the damage to either plane wound not be severe.
    “The pilot of either plane could have misunderstood the order from the air traffic controller and turned into the wrong taxiway,” a local pilot told Shanghai Daily yesterday. He said the incident is rare.

    Media Report-2:

    The collision was severe enough to rip off a few metres of the wingtip of 9M-MRO’s starboard wing. Therefore, there clearly was sufficient energy in the impact to break the outer end of the wing box, so obviously the wing must have been forced backwards and bent laterally in the process, compressing all the trailing edge devices (aileron, outboard flap, flaperon, inboard flap) together. All of those devices would have been substantially damaged and would have needed to have been replaced.

    Speculation-1: I don’t believe that the flaperon recovered on Reunion Island and held by the French Judicial Authorities was fitted to 9M-MRO when it pushed back from gate C1 at WMMK on the night of 7/8 March 2014.

    Speculation-2: I believe that the flaperon recovered from Reunion Island is the flaperon that was damaged in the Taxi Collision at ZSPD, and that it was removed and returned to MAS Maintenance for possible repair, but never was repaired. The fact that the flaperon recovered from Reunion Island had obviously had it’s data plate deliberately removed, strongly suggests that the item had been assessed as not worth repair and had been junked.

    Speculation-3: I believe that 9M-MRO was repaired at ZSPD by fitting another flaperon (which I am betting will be found at the bottom of the ocean – some day).

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.